REGNA AND GENTES

THE TRANSFORMATION OF
THE ROMAN WORLD
a scientific programme of the european science foundation

Coordinators
JAVIER ARCE . EVANGELOS CHRYSOS . IAN WOOD

Team Leaders Steering Committee

Miquel Barceló Gunilla Åkerström-Hougen
Mark Blackburn Volker Bierbrauer
Gianpietro Brogiolo Niels Hannestad
Alain Dierkens Przemyslaw Urbańczyk
Richard Hodges Mario Mazza
Marco Mostert H.H. van Regteren Altena
Patrick Périn Heid Gjöstein Resi
Walter Pohl L. Cracco Ruggini
Frans Theuws
Leslie Webster

Series Editor
IAN WOOD

VOLUME 13

REGNA AND GENTES

 REGNA AND GENTES
The Relationship between Late Antique and Early Medieval
Peoples and Kingdoms in the Transformation of the Roman World

EDITED BY

HANS-WERNER GOETZ, JÖRG JARNUT

AND

WALTER POHL

WITH THE COLLABORATION OF

SÖREN KASCHKE

BRILL
LEIDEN • BOSTON
2003
 This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

Regna and gentes : the relationship between late antique and early medieval peoples and
kingdoms in the transformation of the Roman world / edited by Hans Werner Goetz,
Jörg Jarnut and Walter Pohl ; with the collaboration of Sören Kaschke.
p. cm. (The transformation of the Roman world, ISSN 1386 4165 ; v. 13)
Includes bibliographical references and index.
ISBN 9004125248
1. Germanic peoples History. 2. Ethnicity Europe History. 3. Ethnicity Holy
Roman Empire History. 4. Europe History 392-814. 5. Europe Politics and
government. I. Goetz, Hans Werner. II. Jarnut, Jörg. III. Pohl, Walter, 1953 IV. Series.

GN549.G4 R44 2002

305.8'00943 dc21
2002034271

Die Deutsche Bibliothek - CIP-Einheitsaufnahme

Regna and gentes : the relationship between late antique and early medieval
peoples and kingdoms in the transformation of the Roman world / ed. by
Hans-Werner Goetz ... With collab. of Sören Kaschke. – Leiden ; Boston :
Brill, 2003
(The transformation of the Roman world ; Vol. 13)
ISBN 90–04–12524–8

ISSN 1386–4165
ISBN 90 04 12524 8

© Copyright 2003 by Koninklijke Brill NV, Leiden, The Netherlands

All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.

Authorization to photocopy items for internal or personal

use is granted by Brill provided that
the appropriate fees are paid directly to The Copyright
Clearance Center, 222 Rosewood Drive, Suite 910
Danvers, MA 01923, USA.
Fees are subject to change.

printed in the netherlands

 CONTENTS

List of Contributors .................................................................. vii

Abbreviations .............................................................................. xi

Introduction ................................................................................ 1
Hans-Werner Goetz

The Empire, the gentes and the regna ...................................... 13

Evangelos Chrysos

The Leges Barbarorum: law and ethnicity in the

post-Roman West .................................................................. 21
Patrick Wormald

Gens into regnum: the Vandals .................................................. 55

J.H.W.G. Liebeschuetz

Gens and regnum among the Ostrogoths .................................. 85

Peter Heather

The enigmatic fifth century in Hispania: some historical

problems ................................................................................ 135
Javier Arce

Pro patriae gentisqve Gothorvm statv ................................................ 161

Isabel Velázquez

The transformation of Hispania after 711 .............................. 219

Ann Christys

Gentes, kings and kingdoms—the emergence of states.

The kingdom of the Gibichungs .......................................... 243
Ian N. Wood
vi 

The relationship between Frankish gens and regnum:

a proposal based on the archaeological evidence .............. 271
Michael Schmauder

Gens, kings and kingdoms: the Franks .................................... 307

Hans-Werner Goetz

The Britons: from Romans to barbarians .............................. 345

Alex Woolf

Anglo-Saxon gentes and regna .................................................... 381

Barbara Yorke

Gens, rex and regnum of the Lombards ...................................... 409

Jörg Jarnut

The Bavarians ............................................................................ 429

Matthias Hardt

Avars and Avar archaeology. An introduction ...................... 463

Falko Daim

A Non-Roman Empire in Central Europe: the Avars .......... 571

Walter Pohl

Conclusion .................................................................................. 597

Bibliography .............................................................................. 629

Index of Peoples ........................................................................ 691
Index of Persons ........................................................................ 694
Index of Places .......................................................................... 700
Index of Subjects ...................................................................... 705
 LIST OF CONTRIBUTORS

E C is Professor of Byzantine History at the University

of Athens and Director of the Institute for Byzantine Research at
the Hellenic Research Foundation. His research interests include
Byzantium’s international relations in the early Middle Ages.

P W is a Research Lecturer at the Faculty of Modern

History, University of Oxford. He is a specialist in the law and leg-
islation of post-Roman Europe, his publications including The Making
of English Law, King Alfred to the Twelfth-Century, vol. 1: Legislation and
its Limits (Oxford 1999), and Legal Culture in the Early Medieval West
(London 1999).

W L is professor emeritus, and formerly head of the

Department of Classical and Archaeological Studies, at Nottingham
University. His principal interests are Late Antiquity and Roman
religion. His most recent books are Barbarians and Bishops (Oxford
1990), and The Decline and Fall of the Roman City (Oxford 2001).

P H is Fellow in Medieval History at Worcester College,

Oxford. He is a specialist in the history of the Later Roman Empire
and its successor states (c. 250–600 A.D.), with a strong interest in
the issues surrounding the so-called Migration Period. His publica-
tions include Goths and Romans 332– 489 (Oxford 1991), The Goths
(Oxford 1996), and Philosophy Propaganda and Empire in the Fourth Century:
Select Speeches of Themistius (Liverpool 2001).

J A is Research Professor in the Higher Council of Scientifical

Research (CSIC), Depart. de Historia Antigua y Arqueología at the
Instituto de Historia in Madrid, Spain. He specialises in Late Roman
History and Archaeology, and his recent books include: Centcelles. El
monumento tardorromano. Iconografia y Arquitectura, ed. J. Arce (Roma
2000), Memoria de los antepasados. Puesta en escena y desarrollo del elogio
fúnebre romano (Madrid 2000), El último siglo de la España romana (284–409
A.D.) (3rd edn., Madrid 1997) and Esperando a los bárbaros en Hispania
(409–507) (forthcoming).
viii   

I V is professor of the Latin Department in the

Complutense University of Madrid. She is a specialist in Late and
Medieval Latin and Epigraphy. She is Director of the Archivo
Epigráfico de Hispania and the review Hispania Epigraphica of the
Complutense University and Secretary of the Latin Studies Society
of Spain (SELat). Recently she has published Documentos de época visigoda
escritos en pizarra (siglo VI–VIII), 2 vols. (Turnhout 2000).

A C specialises in the historiography of Spain in the early

Islamic period and has recently published Christians in al-Andalus
711–1000 (Richmond, Surrey 2002). She works as an anaesthetist in
Leeds.

I W, Professor of Early Medieval History, University of Leeds,

has published numerous articles on Early Medieval History. His books
include The Merovingian Kingdoms 450–751 (London-New York 1994),
The Missionary Life (Harlow 2001), and, together with Danuta Shanzer,
Avitus of Vienne: Letters and Selected Prose (Liverpool 2002). He was a
coordinator of the ESF programme on the Transformation of the
Roman World.

M S is curator at the Rheinische Landesmuseum in

Bonn and teaches early Christian and early Medieval archaeology
at the Rheinische-Friedrich-Wilhelms-University in Bonn. His research
fields are late Antique, migration period and early medieval archaeol-
ogy. He wrote several articles on these topics. His book Oberschichtgräber
und Verwahrfunde in Südosteuropa im 4. und 5. Jahrhundert. Zum Verhältnis
zwischen spätantikem Reich und barbarischer Oberschicht aufgrund der archäol-
ogischen Quellen is just out.

H-W G is Professor of Medieval History at the University

of Hamburg and president of the German Mediävistenverband. His
main fields of research are the history of medieval mentality and
attitudes, historiography and social history of the Early and High
Middle Ages. His books include Life in the Middle Ages (London 1993),
Frauen im frühen Mittelalter. Frauenbild und Frauenleben im Frankenreich (Wei-
mar 1995), Moderne Mediävistik. Stand und Perspektiven der Mittelalterforschung
(Darmstadt 1999), Geschichtsschreibung und Geschichtsbewußtsein im hohen
Mittelalter (Berlin 1999) and Handbuch der Geschichte Europas, vol. 2: Das
Frühmittelalter (500–1050) (forthcoming).
    ix

A W is Lecturer in Early Medieval Scottish History at the

University of St Andrews. He has published a number of articles
relating to kingship and social transformation in early medieval Britain
and Ireland.

B Y is Professor of Early Medieval History at King

Alfred’s College, Winchester. She specialises in Anglo-Saxon History
and recent publications include Kings and Kingdoms of Early Anglo-Saxon
England and Wessex in the Early Middle Ages.

J J, born in 1942 in Weimar, Germany, from 1962–1967

study of history and German studies in Bonn (Germany), Caen
(France) and Perugia (Italy). 1970 Dr. phil. Bonn, 1977 habilitation
Bonn, 1980 professor Bonn, since 1983 professor for medieval his-
tory in Paderborn.

M H is Coordinator for Medieval History and Archaeol-

ogy at the Geisteswissenschaftliches Zentrum Geschichte und Kultur
Ostmittel-europas at Leipzig. He specialises in History of Migration
Period, Early Middle Ages and History of settlement structures in
Central Germany and East Central Europe.

F D is Professor at the Institute for Prehistory of the University

Vienna, Austria, and Director of the Vienna Institute of Archaeological
Science. He specialises in Avar studies and the archaeological evi-
dence for cultural exchange between Byzantium and its neighbours.
His recent publications include Das awarische Gräberfeld von Leobersdorf,
Niederösterreich (Wien 1987), Typen der Ethnogenese unter besonderer Berück-
sichtigung der Bayern 2, ed. F. Daim and H. Friesinger (Wien 1990),
Awarenforschungen, 2 vols., ed. F. Daim (Wien 1992) and Die Awaren
am Rand der byzantinischen Welt. Studien zu Diplomatie, Handel und Techno-
logietransfer im Frühmittelalter, ed. F. Daim (Innsbruck 2000).

W P is Director of the Medieval research unit of the Austrian

Academy of Sciences and teaches medieval history at the University
of Vienna. His books include Die Awaren (München 1988; an English
translation is in preparation), Die Germanen (München 2000), Werkstätte
der Erinnerung —Montecassino und die langobardische Vergangenheit (Wien
2001), and Die Völkerwanderung (Stuttgart-Berlin-Köln 2002).
This page intentionally left blank
 ABBREVIATIONS

CIL Corpus Inscriptionum Latinarum

CSEL Corpus Scriptorum Ecclesiasticorum Latinorum
MGH Monumenta Germaniae Historica
AA Auctores antiquissimi
Capit. Capitularia regum Francorum
Conc. Concilia
EE Epistolae
LL Leges
SS Scriptores
SSrG Scriptores rerum Germanicarum
SSrL Scriptores rerum Langobardicarum et Italicarum
SSrM Scriptores rerum Merovingicarum
PL Patrologiae cursus completus. Series latina, ed. J.P. Migne,
221 vols. (Paris 1844–1855)
This page intentionally left blank
 INTRODUCTION

Hans-Werner Goetz

Late Antiquity, no doubt, was a “time of transition or rather transi-

tions”.1 In spite of extensive research on the “Germanic” (or, from
the Roman point of view, “barbarian”) invasions and the successor
states of the Roman Empire, comparatively little attention has been
paid to the “transition of peoples”, or their “developing” into king-
doms. As far as we can see, research on the Later Roman Empire
and the late antique and early medieval kingdoms has focused on
four aspects: first, the “Great Migration”,2 second, the decline of the
Empire (including the role of migration and of the barbarian hordes
in this process),3 third, the ethnogenesis of the “Germanic” peoples,4
and fourth, the rise of (single) kingdoms.5 Meanwhile, we know a lot

1
Thus, for example, I.N. Wood, The Merovingian kingdoms, 450–751 (London-New
York 1994) p. 1.
2
Cf., for example, E. Demougeot, La formation de l’Europe et les invasions barbares,
vol. 1: Des origines germaniques à l’avènement de Dioclétien (Paris 1969); vol. 2: De l’avène-
ment de Dioclétien (284) à l’occupation germanique de l’Empire romain d’Occident (début du VIe
siècle), Collection historique (Paris 1979); W. Goffart, Barbarians and Romans A.D.
418–584. The Techniques of Accomodation (Princeton NJ 1980); J.D. Randers-Pehrson,
Barbarians and Romans. The Birth Struggle of Europe, A.D. 400–700 (London-Canberra
1983); Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989).
3
Cf. A. Demandt, Die Spätantike. Die Römische Geschichte von Diocletian bis Justinian,
284–565 n. Chr., Handbuch der Altertumswissenschaft III,6 (München 1989); id.,
Der Fall Roms. Die Auflösung des Römischen Reiches im Urteil der Nachwelt (München 1984);
A.H.M. Jones, The Later Roman Empire 284–602. A social, economic and administrative
survey, 3 vols. (Oxford 1964); Der Untergang des Römischen Reiches, ed. K. Christ, Wege
der Forschung 269 (Darmstadt 1970).
4
Cf. n. 7, and, summarizing, W. Pohl, Die Germanen, Enzyklopädie deutscher
Geschichte 57 (München 2000).
5
Cf. D. Claude, Geschichte der Westgoten (Stuttgart-Berlin-Köln-Mainz 1970); H. Wol-
fram, Geschichte der Goten. Von den Anfängen bis zur Mitte des sechsten Jahrhunderts. Ent-
wurf einer historischen Ethnographie (München 1979; 3rd edn. 1990); T.S. Burns, A
History of the Ostrogoths (Bloomington 1984); P.J. Heather, Goths and Romans 332–489
(Oxford 1991); id., The Goths (Oxford 1996); P. Amory, People and Identity in Ostro-
gothic Italy, 489–534, Cambridge Studies in Medieval Life and Thought (Cambridge
1997); J. Jarnut, Geschichte der Langobarden, Urban 339 (Stuttgart 1982); D. Geuenich,
Geschichte der Alemannen, Urban 575 (Stuttgart-Berlin-Köln 1997); R. Kaiser, Die Franken:
Roms Erben und Wegbereiter Europas?, Historisches Seminar N.F. 10 (Idstein 1997).
2 - 

about the political development of this period, and we may also have
sufficient knowledge concerning the political, social and cultural
(including religious) structures of this epoch. And, of course, there
are splendid surveys of the period, for example, by Herwig Wolfram,6
Herbert Schutz,7 John Moorhead,8 Patrick Geary9 and, most recently,
by Walter Pohl.10 What we lack, however, is a comparative view of
these kingdoms as well as an attempt to combine these four ele-
ments within a common perspective. One of the first attempts in
this direction was the Marxist volume “Germans are conquering
Rome” (Germanen erobern Rom) by Rigobert Günther and Alexander
Korsunskij,11 limited, however, to a short presentation of each king-
dom. Another, more recent attempt by P.S. Barnwell was restricted
to four kingdoms (Franks, Visigoths, Langobards, and Anglo-Saxons),
each dealt with under three aspects: kings and queens, royal house-
hold, and provincial administration.12 In his conclusion, Barnwell
demands a revision of our image of “Germanic” government (which
was less decadent than generally assumed).13 He rightly points out
our complete dependence on evidence which is, actually, totally
different for each kingdom. He lays further emphasis on the impor-
tance of “rank” for the Visigoths and Anglo-Saxons, and he dis-
covers a continuation of Roman traditions throughout in legislation,
administration (which was dependent on the extension of the king-
dom), minting, royal ceremonies, and Christianity. No doubt these
are important observations, which have been confirmed and refined
by numerous works on individual kingdoms. Nevertheless, we still

6
H. Wolfram, Das Reich und die Germanen. Zwischen Antike und Mittelalter, Das Reich
und die Deutschen (Berlin 1990) [English transl. The Roman Empire and its Germanic
Peoples (Berkeley 1997)].
7
H. Schutz, The Germanic Realms in Pre-Carolingian Central Europe, 400–750 (New
York 2000).
8
J. Moorhead, The Roman Empire Divided, 400–700 (Harlow-London 2001).
9
P.J. Geary, The Myth of Nations. The Medieval Origins of Europe (Princeton 2001).
10
W. Pohl, Die Völkerwanderung. Eroberung und Integration (Stuttgart-Berlin-Köln 2002).
11
R. Günther and A.R. Korsunskij, Germanen erobern Rom. Der Untergang des West-
römischen Reiches und die Entstehung germanischer Königreiche bis zur Mitte des 6. Jahrhun-
derts, Veröffentlichungen des Zentralinstituts für Alte Geschichte und Archäologie
der Akademie der Wissenschaften der DDR 15 (Berlin/Ost 1986; 2nd edn. 1988).
12
P.S. Barnwell, Kings, Courtiers and Imperium. The Barbarian West, 565–725 (London
1997). A first volume, Emperors, Prefects and Kings. The Roman West, 395–565 (London
1992), covered other peoples within a more strictly Roman context.
13
Ibid., pp. 172 ff.
  3

lack an overall comparison of these kingdoms, and the central ques-

tion, namely of the relation between gentes and regna, has so far only
been slightly touched upon and has never been explicitly and tho-
roughly discussed in a comparison of the single realms. As Karl
Ferdinand Werner observed, rex, gens and regnum formed a “triad”:
There were gentes which formed a vast kingdom, and there were oth-
ers which were absorbed by or integrated into these realms.14 However,
it is by no means clear whether existing gentes established kingdoms,
which would mean that the “foundation” of the “Germanic” king-
doms marks a development from gens to regnum, or whether gentes
resulted from the establishment of realms, or—the most probable
assumption—whether there was mutual influence, which in turn
affected both gens and regnum: how this all worked is equally unclear.
An important contribution to this problem has recently been made
by Hans Hubert Anton who, by considering the geographical ter-
minology, asked how the “gentile” communities/federations (or peo-
ples) developed into political and territorial ones. He showed that
extensive geographical terms (such as Hispania, Gallia, Germania, and
Italia) partly lost their political connotation in the new realms and
were overtaken by those of new segmentations (such as Aquitania,
Burgundia, and Francia), but survived (or were revived) as expressions
for the kingdoms in the case of Italy and Spain, and were also used
by “foreign” writers outside the respective kingdom.15 Thus, geo-
graphical terms lost and regained their political impact and (again)
superseded ethnic ones. This, however, is of course only one aspect
of a most complicated process.
Ethnicity and ethnogenesis meanwhile have come to be seen as
extremely difficult and complex phenomena. Since Reinhard Wenskus
published his great book on “The Growth of the early medieval
gentes” in 1961,16 it has become more and more clear and may now
be considered a nearly undisputed conviction that the gentes of the

14
K.F. Werner, “Völker und Regna”, Beiträge zur mittelalterlichen Reichs- und Nations-
bildung in Deutschland und Frankreich, ed. C. Brühl and B. Schneidmüller, Historische
Zeitschrift Beiheft N.F. 24 (München 1997) pp. 15–44, particularly pp. 15–6.
15
H.H. Anton, “Antike Großländer, politisch-kirchliche Traditionen und mittel-
alterliche Reichsbildung”, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Kanonistische
Abteilung 86 (2000) pp. 33–85.
16
R. Wenskus, Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen gentes
(Köln-Graz 1961).
4 - 

Migration period and the Early Middle Ages were not stable “eth-
nic” units (in the “biological” sense of an Abstammungsgemeinschaft), but
“historical”, that is, unstable communities that were prone to change.17
If former research identified “peoples” as communities of human
beings who spoke the same language, as members of a cultural group
represented in archaeological findings, as groups presented under a
single name in written sources, as ethnic groups of the same descent,
or as political groups under the leadership of a king or prince, we
have, in the meantime and to an equal degree, not only become
aware that these five elements do not correspond with each other,
but also that each of these elements is contestable.18 The key fac-
tors, however, according to Wenskus and his followers, were politics
and tradition. “The ethnogenesis of early medieval peoples, there-
fore, was not a matter of blood, but of shared traditions and insti-
tutions; belief in common origins could give cohesion to rather
heterogeneous communities. The early medieval kingdoms were, for

17
Cf. Wolfram, Geschichte der Goten; id., “Ethnogenesen im frühmittelalterlichen
Donau- und Ostalpenraum (6. bis 10. Jahrhundert)”, Frühmittelalterliche Ethnogenese im
Alpenraum, ed. H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985) pp.
97–151; Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern 1, ed. W. Pohl
and H. Wolfram, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 201. Veröffentlichungen der Kommission für Früh-
mittelalterforschung 12 (Wien 1990); Ethnogenese und Überlieferung. Angewandte Metho-
den der Frühmittelalterforschung, ed. K. Brunner and B. Merta, Veröffentlichungen des
Instituts für Österreichische Geschichtsforschung 31 (Wien-München 1994). An in-
structive overview and estimation of this research is given by W. Pohl, “Tradition,
Ethnogenese und literarische Gestaltung: eine Zwischenbilanz”, ibid., pp. 9–26; cf.
id., “Gentilismus”, Reallexikon der Germanischen Altertumskunde 11 (2nd edn., 1998) pp.
91–101; and, recently, id., “Zur Bedeutung ethnischer Unterscheidungen in der
frühen Karolingerzeit”, Studien zur Sachsenforschung 12, ed. H.-J. Häßler (Olden-
burg 1999) pp. 193–298. Cf. also After Empire. Towards an Ethnology of Europe’s Bar-
barians, ed. G. Ausenda, Studies in Historical Archaeoethnology (Woodbridge 1995);
S. Gasparri, Prima delle nazioni. Popoli, etnie e regni fra Antichità e Medioevo (Rom 1997).
For later periods: Concepts of National Identity in the Middle Ages, ed. S. Forde,
L. Johnson and A.V. Murray, Leeds Texts and Monographs. New Series 14 (Leeds
1995); Peuples du Moyen Âge. Problèmes d’identification. Séminaires Sociétés, Idéologies et Croyances
au Moyen Âge, ed. C. Carozzi and H. Taviani-Carozzi, Publications de l’Université
de Provence (Aix-en-Provence 1996); Medieval Europeans. Studies in ethnic identity and
national perspectives in medieval Europe, ed. A.P. Smyth (Basingstoke 1998). For a gen-
eral archaeological approach to the question, see S. Jones, The Archaeology of Ethnicity
(London 1997).
18
Cf. W. Pohl, “Franken und Sachsen: die Bedeutung ethnischer Prozesse im 7.
und 8. Jahrhundert”, 799—Kunst und Kultur der Karolingerzeit. Karl der Große und Papst
Leo III. in Paderborn. Beiträge zum Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann
and M. Wemhoff (Mainz 1999) pp. 233–6; id., Die Germanen, pp. 7–10.
  5

a time, a successful form of making such ethnic communities the

focus of states on the territory of the empire.”19 Seen from this angle,
modern ethnogenetical research has to investigate the subject in a
different way:

1. If gentes were not static units but prone to change, we are obliged
to investigate these changes in the course of the Early Middle
Ages rather than ask for the origins of peoples.
2. If gentes were political rather than “ethnic” units,20 and, conse-
quently, in many cases tended to establish kingdoms (within the
area of, but also outside the institution of the Roman Empire),
the relation between gens and regnum which is the theme of this
volume becomes not only a central, but also a crucial issue.21
3. If gentes were groups formed by tradition (Traditionsgemeinschaften)
rather than by descent, we have to inquire into their self-per-
ception as a gens.

These are central questions concerning the transformation of the

Roman world and the establishment of the late antique and early
medieval “Germanic” kingdoms. Without doubt, the Roman Empire
was not assassinated by the “Germans”, as André Piganiol, still
influenced by the burden of the Second World War, believed.22 But
we are now much less certain about the role of the “Germanic”

19
Thus W. Pohl, “The Barbarian Successor States”, The Transformation of the Roman
World A.D. 400–900, ed. L. Webster and M. Brown (London 1997) pp. 33–47, here
p. 46.
20
Recently, with reference to Bede, H. Kleinschmidt, “The Geuissae and Bede:
On the Innovations of Bede’s Concept of the Gens”, The Community, the Family and
the Saint. Patterns of Power in Early Medieval Europe. Selected Proceedings of the International
Medieval Congress. University of Leeds, 4–7 July 1994, 10–13 July 1995, ed. J. Hill and
M. Swan, International Medieval Research 4 (Turnhout 1998) pp. 77–102, again,
claimed a conceptual change of the gens in so far as the political concept of a gens
as a group of settlers under the control of one ruler was a secondary, post-migra-
tional one.
21
Cf. C. Brühl, Deutschland—Frankreich. Die Geburt zweier Völker (Köln-Wien 1990;
repr. 1995); M. Becher, Rex, Dux und Gens. Untersuchungen zur Entstehung des sächsischen
Herzogtums im 9. und 10. Jahrhundert, Historische Studien 444 (Husum 1996); K.F.
Werner, “Volk, Nation, Nationalismus, Masse, III–V”, Geschichtliche Grundbegriffe. Histo-
risches Lexikon zur politisch-sozialen Sprache in Deutschland 7 (1992) pp. 171–281.
22
A. Piganiol, “Les causes de la ruine de l’empire romain”, id., L’Empire chrétien
(Paris 1947) pp. 411–22 [repr. id., “Die Ursachen des Untergangs des Römischen
Reiches”, Der Untergang des Römischen Reiches, ed. K. Christ, Wege der Forschung 269
(Darmstadt 1970) pp. 270–85].
6 - 

peoples in this process. Orosius’s report that the Visigothic king

Athaulf planned to destroy the Roman Empire, in order to estab-
lish a Gothic one,23 seems completely anachronistic for his time. In
the end, however, historical development seemed to have reached a
state which came very close to Athaulf ’s plans, achieved by the very
“Germanic” kingdoms which (without knowing or planning it) had
become the “heirs” of a Roman Empire which in its turn had devel-
oped into an alienated figure far away in the East. Therefore, we
are forced to investigate very closely what had happened in the
meantime.
Examining the relation between regnum and gens is an approach
which, in this context, may reveal the (different) phases of political
changes and, even more important, the causes and consequences of
the “establishment” of new kingdoms. Moreover, it helps us to rec-
ognize differences and similarities between individual peoples and
realms. There are, however, (at least) six crucial problems inherent
in this question.

• The first problem is already inherent in the terms “peoples” and

“kingdom”, terms that can no longer be defined per se (despite
their necessary interrelation when “peoples”, too, are seen as hav-
ing a political connotation). Not only is “peoples” an ambivalent
term (and, what is more, the German Volk has become fraught
with ideologically incriminating connotations), but also a “king-
dom” does not simply emerge where there is a king, but can or
should be understood as a political order, or a “state” with a
sufficient measure of organization. Since the interest of this vol-
ume lies in the relationship between “peoples” and “kingdoms”,
our concern is focussed on those gentes which developed into, and
gave their name to (larger) regna, particularly with regard to the
successor states of the Roman Empire. In practice, however, it
is not at all easy to draw a clear line between “Germanic” com-
munities and “Germanic” kingdoms as successor states of the
Empire.
• A second, and particularly prevalent problem is the term “Ger-
manic” itself. After centuries of a seemingly clear distinction, fol-

23
Orosius, Historiarum adversum paganos libri VII 7,43,3 ff., ed. K. Zangemeister,
CSEL 5 (Vienna 1882; repr. Hildesheim 1967) pp. 559–60.
  7

lowed by doubts and restrictions, we have now reached the point

where we are not even sure any more what “Germanic” really
means. The Germani of the Roman sources seem to be a Roman
construction (and, moreover, the term was not used very fre-
quently), and there are no signs of a “Germanic” self-conception
among the “barbarian” peoples which the Romans considered
(or we think) to have been “Germanic”. In other words, the
“Germanic” peoples did not conceive themselves as being “Germa-
nic”, or at least did not attach any importance to this feature.
The term can be defined, of course, in terms of language, but
with regard to the early period we have little knowledge of the
language spoken by those peoples which, according to Reinhard
Wenskus, were conglomerations of different communities anyway.
Moreover, previous (German) research overemphasized many phe-
nomena (such as Herrschaft, Eigenkirche, Sippe, or Gefolgschaft) that
seem specific to the Early Middle Ages rather than being typi-
cally “Germanic”. It is important, therefore, to compare so-called
“Germanic” peoples and kingdoms with presumably non-Germanic
ones. As a consequence of these problems, it was suggested that
the term “Germanic” be dropped completely in this volume, but
to substitute it with “barbarian” would only mean adopting another
(Roman) ideology which, in the final analysis, is as inadequate as
“Germanic”. The only neutral alternative, therefore, would be to
simply speak of late antique and early medieval peoples and king-
doms, but, of course, this would merely be evading the problem.
Probably it is more important to remain constantly aware of the
problematic questions that are inherent in our topic. Moreover,
it would also be necessary to have some critical reflection on the
term “Roman”.
• The third problem concerns the difference in development and
structure of “Germanic” kingdom-building. Sometimes the for-
mation of a realm focuses more or less on a single act (such as
Theodoric’s Ostrogothic kingdom), sometimes it resembles a grad-
ual movement (as under both Visigothic kingdoms, the “Tolosan”
as well as the “Toledan” one), sometimes it consists of an accu-
mulation of territories and realms (as with Clovis’s “foundation”
of the Merovingian kingdom of the Franks). Moreover, we should
not forget that we are comparing developments that extended
over a considerable period, from the early fifth to the late eighth
centuries.
8 - 

• The fourth problem is obviously the state of research on the ethno-

genesis of the early medieval peoples. After four decades of inten-
sive work on this topic using a “modern” approach we now have
far more knowledge of what a “Germanic” gens was not, than
what it was: that is, we are aware of so many problems that it
seems nearly impossible to provide any straightforward answers.
We do know, however, as already mentioned in the beginning of
this introduction, that the gentes were not stable groups with clear
ethnic origins, but (constantly) changing (an extremely important
point), and that the political factor was at least as decisive for
these developments as the (usually fictional) consciousness of com-
mon origins. Ethnogenetical processes, therefore, can no longer
be considered without taking into account the political develop-
ment, that is, without considering the establishment of kingdoms.
However, the problem which remains is how to define ethnicity.
By which criteria, or by which historical evidence can ethnicity
be comprehended? We have to bear in mind that there are different
approaches to, and definitions of ethnicity, and in consequence
we have to make explicit what we, that is, each contributor respec-
tively, mean by using terms, or rather theoretical constructions,
such as gens or regnum.
• The fifth problem, accordingly, lies in the (newly emerging) king-
doms and their character. It is not so much a matter of the long
(but probably typically German) discussion whether (or when) these
kingdoms may be called “states” (which is a modern expression
anyway)—“state” in this sense may be used as a term for the sys-
tem of political order which has to be described under contem-
porary conditions, however it is labelled. Far more important, and
indeed extremely relevant, is the question of which bonds and
institutions (if there were “institutions”) the power of the “Germa-
nic” kings rested in. Were these “Germanic” or Roman elements?
or both? or neither? And was there a “state” (or a kingdom) that
was not exclusively dependent on the ruling of a (certain) king or
dynasty? What were, for example, the differences between the
“realms” of Marbod in the first, Attila in the fifth and Theodoric
in the sixth century?
• A sixth problem is the question of our sources and how to deal
with them. The discrepancy between the evidence we have for
each individual kingdom makes it inevitable that we should con-
sider the quality and range of sources for each contribution respec-
tively, when we aim at comparing the different kingdoms. The
  9

real problem, however, goes deeper. It includes not only the well-
known and lamentable fact that at least the early stages of the
“Germanic” peoples and kingdoms are almost exclusively recorded
by Roman sources and seen from a Roman perspective, but, even
more important, the more general question of whether there were
decisive differences between the actual historical process and the
way it was perceived by the contemporary authors of those times,
not to mention the authors’ bias, intentions, narrative structures,
or choice of events. This is not only a question of criticism of
our sources. As they did not (and could not) have our concept
of, and, moreover, our interest in ethnogenesis, that neither means
that they were wrong, nor does it mean that our theories are
inadequate. Although we have to go farther in our explanations
than contemporary writers did, at the same time we have to be
aware of the characteristic features of their perceptions because
it was their view (not ours) that was underlying the thoughts as
well as the deeds of the people of those times. Thus we are obliged
to take into account what they meant when they spoke of a gens
or a regnum and how (and if ) they saw any changes.

Looking at these problems, the relations between gentes and regna (or
between a certain gens and a corresponding regnum) are neither clear,
nor is it at all self-evident that there was an (explorable) develop-
ment from gens to regnum or how a people changed after the estab-
lishment of a kingdom. Neither is it self-evident that these changes
were perceived by our sources or what our sources made of them.
Certainly, however, there were alterations that we are able to observe
and compare, and it is the aim of the present volume to consider
these relations and developments as well as the political and “eth-
nic” structures in different peoples and regions.
This volume may be regarded as the result of a long process of
discussion that the majority of the contributors were allowed to enjoy
for five years supported by the European Science Foundation and
its project “The Transformation of the Roman World” (TRW). The
Working Group 1 (“Imperium and gentes”) of this project, chaired
by Walter Pohl, after discussions on the early kingdoms, gentile struc-
tures and other topics,24 aimed at clarifying the crucial question of

24
Cf. Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. W. Pohl,
The Transformation of the Roman World 1 (Leiden-New York-Köln 1997); Strategies
10 - 

the relation between gens and regnum to some extent in a compa-

rative approach. A first step was taken during the meeting at Barcelona
(October 30–November 1, 1997) where some members considered
one gens (or regnum) each under common leading questions. This dis-
cussion was continued in the working group’s last meeting in Manerba
del Garda (October 22–25, 1998) where the group decided to elab-
orate its results, to complement them by further articles, which were
to include non-Germanic developments, and to publish them in a
volume of the TRW series. Pre-final drafts of all papers were dis-
tributed among the participants and some invited experts and were
discussed at a meeting in Bellagio, sponsored by the Rockefeller
Center, which most of the colleagues involved were able to attend
(December 11–15, 2000). The contributors were given leading ques-
tions, previously agreed upon, which were meant to assist and war-
rant a comparative approach, although these questions naturally had
to be adapted to the special cases respectively. These questions were:

• (main and central question): Was there a development from a

“Germanic” gens of the Migration Period to a “Germanic” king-
dom? Or did a gens (or this gens) not exist until after the estab-
lishment of a kingdom?
• What sorts of changes and conditions led to, or represented the
development towards, respectively, the establishment of a “Ger-
manic” kingdom?
• What was the role of a gentile identity (Stammesbewußtsein) for the
establishment of a regnum?
• What sorts of changes in the “constitution” (Verfassung) of a peo-
ple and a kingdom (such as central organs of power, local power
structures, or links between the two) were linked to the estab-
lishment of a kingdom? How did socio-economic developments
contribute to this process?
• What was the role of kings in this development?
• What part did the Roman Empire play in this process?
• In all these points, special attention should be paid to change and
development.

of Distinction. The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Rei-
mitz, The Transformation of the Roman World 2 (Leiden-New York-Köln 1998).
  11

In pursuing this enterprise, this volume is deliberately not just confined

to the “Germanic” peoples (Anglo-Saxons, Bavarians, Burgundians,
Franks, Langobards, Ostrogoths, Vandals, Visigoths), but compares
these with the West and East Roman tradition (Byzantium and Late
Antique Spain) and also with non-Germanic peoples (such as Celts,
Huns and Avars), and even with the Islamic kingdoms in early
medieval Spain. It also seemed advisable to include a comparative
survey of the different Germanic laws. The editors are particularly
grateful to those colleagues who willingly agreed to join the “group”
at a later phase. They wish to thank the participants for helpful
comments on the introduction and conclusion, and particularly Ian
Wood for a last revision of those texts that were translated into
English. They would also like to thank Julian Deahl and Marcella
Mulder of Brill Academic Publishers for guiding their work and
preparing the volume for publication. Last but not least, they are
grateful to Sören Kaschke (Hamburg) who has transformed articles
that varied in form and footnotes into a standardized and legible
volume.
By following the leading questions mentioned above and concen-
trating on the topic of the relationship between gentes and regna, we
hope to contribute to an essential problem and help to fill a crucial
gap, both by presenting concise articles on the single kingdoms dealt
with here and by suggesting a basis for a comparative approach to
this subject which is not only central for the period of the transfor-
mation of the Roman world but, in a time of changing national
identities, also bears significant signs of actuality for the present day.
This page intentionally left blank
 THE EMPIRE, THE GENTES AND THE REGNA

Evangelos Chrysos

In this volume the route taken by the individual gentes in the process
of establishing themselves as regna in or at the periphery of the Roman
Empire is studied comparatively. In this short contribution I will try
to articulate some common characteristics of the influence the Roman
Empire may have exercised on this process.
In a seminal paper on the Gothic kingship Herwig Wolfram has
made the observation that the transformation of the gentes into regna,
as we can grasp it through the evidence in Greek and Roman sources,
could take place only in some sort of connection with the Empire.1
It was mainly the need to accommodate themselves politically and
economically in their new environment and in relationship with the
Empire that the migrating peoples shaped the structure of their poli-
ties as regna. Is this hypothesis correct and how are we to under-
stand it?
If we base our analysis on the well established and accepted assump-
tion about the gentes being not solidly formed and statically estab-
lished racial entities but groups of people open to constant ethnogenetic
change and adaptation to new realities, then it is reasonable to expect
that their relationship with the Empire had a tremendous impact on
their formation. We can grasp this impact in the following three
phases of their development. It is unneccessary to underline the point
that these phases as presented here are merely indicative of processes
that follow different paths with different speed.

First phase

The individual or corporate recruitment of barbarians during the

migration period in the Roman army2 offered them experience of

1
H. Wolfram, “Gotisches Königtum und römisches Kaisertum von Theodosius dem
Großen bis Justinian I.”, Frühmittelalterliche Studien 13 (1979) pp. 1–28, here pp. 1–2.
2
T.S. Burns, Barbarians within the Gates of Rome (Bloomington 1994); M. Cesa,
Tardoantico e barbari (Como 1994).
14  

how the imperial army was organized, how the government arranged
the military and functional logistics of their involvement as soldiers
or officers and how it administered their practical life, how the pro-
fessional expertise and the social values of the individual soldier were
cultivated in the camp and on the battlefield, how the ideas about
the state and its objectives were to be implemented by men in uni-
form, how the Empire was composed and how it functioned at an
administrative level. This knowledge of and experience with the
Romans opened to individual members of the gentes a path which,
once taken, would lead them to more or less substantial affiliation
or even solidarity with the Roman world. To take an example from
the economic sphere: The service in the Roman army introduced the
individual or corporate members into the monetary system of the
Empire since quite a substantial part of their salary was paid to
them in cash. With money in their hands the “guests” were by neces-
sity exposed to the possibility of taking part in the economic system,
of becoming accustomed to the rules of the wide market, of absorb-
ing the messages of or reacting to the imperial propaganda passed
to the citizens through the legends on the coins. In addition the
goods offered in the markets influenced and transformed the new-
comers’ food and aesthetic tastes and their cultural horizon. Further-
more Roman civilitas was an attractive goal for every individual
wishing to succeed in his social advancement. Persons like Flavius
Fravitta, “by birth a barbarian but otherwise a Greek not only in
habits but also in character and religion”3 set a remarkable para-
digm to be followed by others.

Second phase

Similar but more substantial and in depth is the path migrating gentes
took when they entered the wide orbit of the Roman world either
in accordance with a peace treaty as foederati or subjected to Roman
domination as dediticii.4 Their communication with the Empire on a
local level, with provincial governors, or generals and ultimately with

3
Zosimus, Histoire Nouvelle 5,20, ed. and transl. F. Paschoud, 4 vols. (Paris 1971–89).
4
See the papers of G. Wirth, P. Heather, W. Liebeschuetz and E. Chrysos in
Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. W. Pohl, The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997).
   15

the imperial government had by necessity to be in a Roman (Latin

or Greek) language and within the current legal and social concepts
and terminology which Rome had developed for her dealings with
her neighbours. The political terms concerning the regulations for
settlement on public, confiscated or derelict private land, the condi-
tions of autonomous (legal) conduct within the Roman system of
control and coercion, the framework for trade and the transfer of
foods for the people and—at a later stage—the work of the mission-
aries, who by preaching the Christian gospel opened to the converted
newcomers a gate to the Graeco-Roman culture of the Mediterranean;
all these and many other, more or less obvious channels of com-
munication and means of affiliation served as the instruments for
shaping the regna within or at the edge of the Roman Empire.
Furthermore an extensive nexus of kinships at all social levels, includ-
ing the leading figures in the gentes among themselves and with mem-
bers of the Roman aristocracy and even the imperial families,5 created
a new intermixed society despite the practically disregarded prohi-
bition of intermarriage between Romans and barbarians.6 However,
the way the Romans expected the gentile leadership to behave within
what was understood as the “international community”, created the
demand for access to standardised forms of political discourse. This
is obvious in the adaptation of imperial methods of diplomatic dis-
course by the regna.

5
A. Demandt, “The Osmosis of Late Roman and Germanic Aristocracies”, Das
Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989) pp. 75–86
with the attached table, and S. Krautschick, “Die Familie der Könige in Spätantike
und Frühmittelalter”, ibid., pp. 109–42. This “family of kings” was a real one and
had little in common with the fictitious system anticipated by F. Dölger in his
famous theory of “Die ‘Familie der Könige’”, in his Byzanz und die europäische Staaten-
welt (Darmstadt 1964) pp. 34–69. For a critical detailed analysis see J. Moysidou,
To Byzantio kai oi boreioi geitones tou ton 10o aiona (Athens 1995). Cf. E. Chrysos, “Legal
Concepts and patterns for the Barbarians’ Settlement on Roman Soil”, Das Reich
und die Barbaren, pp. 13–23, here pp. 13–4; id., “Perceptions of the International
Community of States during the Middle Ages”, Ethnogenese und Überlieferung. Angewandte
Methoden der Frühmittelalterforschung, ed. K. Brunner and B. Merta, Veröffentlichun-
gen des Instituts für Österreichische Geschichtsforschung 31 (Wien-München 1994)
pp. 293–307, here pp. 293–4.
6
H. Sivan, “The appropriation of Roman law in barbarian hands: ‘Roman-bar-
barian’ marriage in Visigothic Gaul and Spain”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transformation
of the Roman World 2 (Leiden-New York-Köln 1998) pp. 189–203. On the Hispano-
Romans see D. Claude, “Remarks about relations between Visigoths and Hispano-
Romans in the seventh century”, ibid., pp. 117–30.
16  

Following this demand several forms of imitatio imperii were placed

on the agenda. The court, the language, public ceremonies involv-
ing the king, court rituals, his titles and dress, forms of distinct
munificence to the people and many other expressions of power were
imitating Roman forms that were thought to safeguard and support
the position of the rex as dominus over his gens and the Roman pop-
ulation in his regnum.7

Third phase

The legal arrangement with the Empire in one way or the other
remained for quite some time the legal frame for the physical exis-
tence and the institutional consolidation of the new polities as regna.
As we know from the case of the Vandals, part of the agreement
with the Empire in A.D. 474 which created the basis for the insti-
tutionalization of the gens into a regnum in Africa was the agreed
“order of succession” (Nachfolgeordnung)8 and it was the breach of this
order by Gelimer that allowed Justinian to send Belisarius and the
fleet against him. Similarly the breakdown of Theoderic’s arrange-
ment with Anastasius after the death of his son-in-law Eutharic in
A.D. 518 opened the gate to a gradual destabilization of Amal rule
in Italy and hence the murder of Amalasuintha, who personified the
legitimacy of the regime, by her cousin Theodahad made the mili-
tary intervention of Justinian inevitable.9 For the organization of the
personal relations between individuals, members of the ruling gens

7
See M. McCormick, “Clovis at Tours, Byzantine Public Ritual and the Origins
of Medieval Ruler Symbolism”, Das Reich und die Barbaren, pp. 155–80.
8
D. Claude, “Probleme der vandalischen Herrschaftsnachfolge”, Deutsches Archiv
30 (1974) pp. 329–55, here pp. 329–30.
9
Ch. Schäfer, Der weströmische Senat als Träger antiker Kontinuität unter den Ostgotenkönigen
(490 –540 n.Chr.) (St. Katharinen 1991) pp. 240–1; J. Moorhead, “Libertas and
Nomen Romanum in Ostrogothic Italy”, Latomus 46 (1988) pp. 161–8. For a long
period of Amal rule over the pars occidentis people such as Cassiodorus and Ennodius
would propagate Theoderic’s image as the custos libertatis et propugnator Romani nominis
(CIL X 6850) in order to legitimise him in the eyes of the Roman population,
while in the years of decline (after 518 and more clearly after 526) Roman aristo-
crats dare to voice their hope for restoration of the libertas Romana anticipating the
ideological preparation of Justinian’s war of reconquista. Cf. J. Moorhead, “Italian
loyalties during Justinian’s Gothic War”, Byzantion 53 (1983) pp. 575–96.
   17

and the Roman local population, the two legal systems, i.e. the writ-
ten Roman Law and the orally transmitted barbarian code of norms,
had to be harmonized.
At the end of this development we find the creation of regnal leg-
islation, the so-called leges barbarorum.10 It was this lex, the “national”
code that gave the gens the necessary impetus for a maturing state.
In this sense what Orosius cites as Athaulf ’s dilemma concerning his
attitude towards the Empire—regardless of its historical reliability11—
becomes a classic (at least a classicized) paradigm. Athaulf is said to
have refrained from establishing a Gothic empire in the place of the
Roman one, because his people would have difficulties in abiding
by the laws (= leges, sine quibus respublica non est respublica).12 Apparently
it was known to everybody that in order to consolidate their terri-
tory and organize their people into kingdoms the kings needed the
existence of a functioning legal system.
When this stage of development was reached the successor states
experimented with what we can call the aemulatio imperii, an attitude
operating as an advanced form of imitatio. To emulate the Empire
meant for the successor states simply to present the merits of “national”
achievement and compare their own activities, building programmes,
political and military achievements and institutions, personal and col-
lective piety, social care etc. with analogue Roman realities and con-
sequently to claim equality and hence to be proud of their record.
On the other hand, political and/or denominational eulogists were
asked to articulate an ideological concept critical of Roman institu-
tions, behaviour and practices that were to be condemned and
opposed. Isidore of Seville, despite being consciously Hispano-Roman,
launched a negative image of the Roman Empire and its past achieve-
ments in contrast to that of the juvenile and pious peoples of his
time. In this sense he would recall intentionally that multae gentes a
Romano imperio recesserunt.13

10
P. Wormald, “Lex Scripta and Verbum Regis: Legislation and Germanic Kingship,
from Euric to Cnut”, Early Medieval Kingship, ed. P.H. Sawyer and I.N. Wood (Leeds
1977) pp. 105–38 and the paper by the same author in this volume.
11
W. Suerbaum, Vom antiken zum frühmittelalterlichen Staatsbegriff (3rd edn., Münster
1977) pp. 222–3.
12
Orosius, Historiarum adversum paganos libri VII 7,43, ed. K. Zangemeister, CSEL
5 (Vienna 1882; repr. Hildesheim 1967).
13
Isidore of Seville, Chronica maiora 238, ed. T. Mommsen, MGH AA 11 (Berlin
1894) p. 454. Cf. H. Löwe, “Von Theoderich dem Grossen zu Karl dem Grossen.
18  

More influential in the Middle Ages—and current even today!—

is Paul the Deacon’s perception of the Empire’s disturbing role in
the formative development of the regna. In his much-quoted Historia
Romana he replaces the regnal years of the emperors with incarna-
tion chronology after Theoderic’s takeover in Italy in A.D. 49214 (the
Imperium Romanum is replaced by the Imperium Christianum!). Hence
he regards Justinian’s reconquista as a subjugation (servitium) of the
“Romans” by the “Greeks” and consequently labels Justinian’s suc-
cessors as “Greek” emperors, and thus no longer “Romans”.15 Over
the next centuries questioning the Romanitas of the Empire of New
Rome will remain a frequently applied method for pushing Constan-
tinople to the periphery as an exotic and anachronistic survival from
the past.
The next two alternative stages would take us beyond the period
under investigation:

• The Franks, to be followed by others, set as their final goal, the

translatio imperii. On the basis of the ideas introduced with the
Constitutum Constantini, a text fabricated in the eighth century to
serve papal aspirations of political hegemony in Italy, but which
at the same time helped to re-arrange the world-system in favour
of the Frankish interests, the Franks claimed the imperium, at least
a Frankish imperium, and the nomen imperatoris, for their king. Many
centuries later, in A.D. 1202, this attitude was to be followed by
Pope Innocent III with his claim of the translatio imperii ad Germanos.16
(Two years later, the fourth crusade brought the sack of Cons-
tantinople and the establishment of a Latin Empire on the shores
of Bosporus!).

Das Werden des Abendlands im Geschichtsbild des frühen Mittelalters”, Deutsches

Archiv 9 (1952) pp. 353–401, here pp. 363–4; G.B. Ladner, “On Roman Attitudes
toward Barbarians in Late Antiquity”, Viator 7 (1976) pp. 1–26.
14
Paul the Deacon, Historia Romana, ed. A. Crivellucci, Fonti per la storia d’Italia
51 (Rome 1914): cessante iam Romanae urbis imperio utiliusque aptiusque mihi videtur ab
annis dominicae incarnationis supputationis lineam deducere.
15
On this see now H. Hofmann, “Roma caput mundi? Rom und Imperium
Romanum in der literarischen Diskussion zwischen Spätantike und dem 9. Jahr-
hundert”, Roma fra oriente e occidente, Settimane di studio del centro italiano di studi
sull’alto medioevo 49 (Spoleto 2002) pp. 493–556, here pp. 534–5.
16
H. Beumann, “Unitas Ecclesiae—Unitas Imperii—Unitas Regni. Von der impe-
rialen Reichseinheitsidee zur Einheit der Regna”, Nascita dell’Europa ed Europa Carolingia:
Un’equazione da verificare, Settimane di studio del centro italiano di studi sull’alto
medioevo 27 (Spoleto 1981) pp. 531–71.
   19

• Those kings of the early medieval states who could not claim or
dream of a translatio imperii for themselves would be consoled with
the principle of—to use a medieval expression—rex superiorem non
recognoscens, imperator est in regno suo.

The picture of the emergence and the shaping of the regna in con-
stant and close connection with the Empire as suggested in this paper
could create the wrong impression that the Empire remained through-
out the critical period a static structure with no changes in form and
life. On the contrary, it is true that the Empire was itself subjected
to these developments. The idea and the functioning of the emper-
orship, the social structure of the Roman population, the role of the
army, the principles and the trends of the policy towards the neigh-
bours, the cultural values and tastes, the impact of religion on every-
day life and many other practical and theoretical aspects of public
interaction underwent substantial changes. Some of these aspects in
the Empire were directly or indirectly affected by the existence of
the regna on formal Roman soil or at the edge of the imperial ter-
ritory. The political developments caused by the occupation and con-
quest of Roman soil and parallel to that the consolidation of the
successor states brought an equilibrium of growth and capabilities.
The road to the strengthening and the stabilization of the regna caused
what we can call the “regnalisation” of the Empire. The expression
regnum Romanum, already in use in the fourth century, acquired real
political meaning. The general use of the word Romania in the sev-
enth and later centuries by the Byzantines for their Empire expressed
the same attitude. In this sense the dispute with Constantinople over
the imperial title of Charlemagne after his coronation in A.D. 800
was merely a quarrel about the nomen imperatoris.
This page intentionally left blank
 THE LEGES BARBARORUM:
LAW AND ETHNICITY IN THE POST-ROMAN WEST

Patrick Wormald

One experience shared by almost every “barbarian” gens that sur-

vived as a post-Roman polity into the seventh century was the even-
tual adoption of a written lex.1 From this alone, it is clear that
possession of a lex came to be a highly significant element in the
making of any acknowledged political unit in the West. Further, the
leges have enough in common to suggest that the processes which
brought them forth were broadly similar. They entirely justify a
“cross-regnum” approach, such as has often been essayed, and not
only by legal historians.2
The salient difficulty here has been the historian’s usual bugbear
of hindsighted teleology. It has been normal to read back into these
texts the legislative assumptions espoused by legally literate western
cultures since the twelfth century: principally that a mutually identifiable
“law” is a fundamental ingredient in the personality of any people—

1
Exceptions: most obviously the Avars (though perhaps not the Bulgars: R.
Browning, Byzantium and Bulgaria: a comparative study across the early medieval frontier
[London 1975] pp. 124–5); also the Anglian and most Saxon peoples in Britain.
For present purposes, it makes no difference that some had to await the eighth or
early-ninth century (or in Scandinavia much later) for this.
2
Thus, H. Brunner, Deutsche Rechtsgeschichte, 2 vols. (Leipzig 1906, 2nd edn. 1928);
R. Buchner, Die Rechtsquellen, Wattenbach-Levison, Deutschlands Geschichtsquel-
len im Mittelalter, Vorzeit und Karolinger, Beiheft (Weimar 1953); most recently,
M. Lupoi, Alle radici del mondo giuridico europeo (Rome 1994), Engl. transl. The Origins of
the European Legal Order (Cambridge 2000). See also J.M. Wallace-Hadrill, Early
Germanic Kingship in England and on the Continent (Oxford 1971) pp. 32–9; and my own
approach—not, as has been observed, that of a legal historian—P. Wormald, “Lex
Scripta and Verbum Regis: Legislation and Germanic kingship, from Euric to Cnut”,
Early Medieval Kingship, ed. P.H. Sawyer and I.N. Wood (Leeds 1977) pp. 105–38,
reprinted and corrected in my Legal Culture in the Early Medieval West. Law as Text,
Image and Experience (London 1999) pp. 1–43; together with that volume’s Preface,
pp. xi–xv; and my The Making of English Law. King Alfred to the Twelfth Century, vol.
1: Legislation and its Limits (Oxford 1999), chapter 2. In compiling this retractatio, I
owe much to this volume’s editors and my fellow-contributors, notably Professor
Wood and Dr Christys.
22  

quis populus, eius lex, as it were.3 A law somehow emerges from the
society to which it applies, helping not just to resolve its tensions
but also to demarcate it as a community. Because the units that
developed out of the western Empire were recognizably the germ of
today’s European states, as this volume is concerned to analyse, the
lex that they came to profess must (it is thought) have been one of
their defining features: scarcely less so than is the law enforced by
the courts of modern governments.
On this supposition, the legislation of Germanic kings merely marks
their assuming responsibility to regulate the problems of the king-
doms created by their armies. The persistence of Roman law within
their regna would then amount to a concession on their part. Nor
were Romans the only ones covered by this allegedly prevalent prin-
ciple of “the personality of law”. Any people incorporated into a
realm must axiomatically have had an inherent lex of its own. That
people, therefore, could only feasibly be ruled by recognizing their
“law”. Hence, codes were issued for peoples other than the ascen-
dant “Volk”, which these peoples were then allowed to apply in cases
involving their members.4 This may indeed have become the posi-
tion by the time of the Carolingian hegemony. But it is very far
from clear that it was true anywhere before the seventh century,
while in certain parts of the West it was never the case. Nor is it
obvious how Roman law can have been something conceded, as if
to a minority group, when the aristocracies that adhered to it were
anything but marginal; or why the charters and formulae that gov-
erned property relations should have remained deeply influenced by
Roman concepts, when so few of the leges seemingly were. And how
did it come about that the Edict of King Rothari of the Lombards
(643) gives every sign of being an unambiguously “Germanic” doc-
ument with little to say of Romans, when the legislation of his suc-
cessors not only reflected the presence of Romans but was evidently
influenced by their legal ideas?

3
Note the discussion (following Regino of Prüm) of legal pluralism as a criterion
of ethnicity, and of its erosion from the twelfth century onwards, by R. Bartlett,
The Making of Europe. Conquest, Colonization and Cultural Change, 950–1350 (London
1994) pp. 197–220.
4
This article of faith, widespread in the “Rechtsschule” (Buchner, Rechtsquellen,
p. 5)—though controversial as regards Visigothic legislation—is epitomized in S.L.
Guterman, From Personal to Territorial Law. Aspects of the History and Structure of the
Western Legal-Constitutional Tradition (New York 1972).
  LEGES BARBARORUM 23

A central thesis of this volume is that the foundation of a regnum

was not simply a matter of displacing the structures of imperial gov-
ernment. To support that thesis, we need a view of early medieval
law-making which explains how extant texts developed out of estab-
lished Roman practice. In the argument that follows, two princi-
ples will be primary. First, chronology will be closely observed. The
course of legislative development will be plotted from the fifth century,
when it can be taken for granted that Roman procedures remained
dominant (except where, as in Britain, all aspects of Romanitas were
evidently at risk) through to the ninth, when something of a new
legislative equilibrium seems to have emerged. Second, the maxi-
mum possible attention will be given to geographical variation. Just
as some post-Roman regimes were clearly a lot more Roman in style
and machinery than others, so their legal pronouncements were
exposed to variable levels and types of Roman influence. At the
same time, it is regarded as not likely that the legal memorials of
post-Roman society were merely varieties of a Roman legacy. Much
in “Germanic” law can be adequately explained as sub-Roman provin-
cial routine. But a great deal of the extant material can only be
understood in terms of the customs imported by the West’s new
masters.

II

The legislation of the later Roman emperors was pre-eminently the

province of their “Quaestors of the Sacred Palace”.5 The Praetorian
Prefects of the Severan dynasty were among the great lawyers of all
time, with correspondingly enriching results on imperial rescripts.
Quaestors from Constantine to Justinian were not always legally
untrained but they were above all rhetoricians, with literary tastes
to match. The quintessential case is Ausonius, drafted from the lec-
ture-room of Bordeaux University to address and administer his ex-
pupil’s realm. Their legislative output was thus characterized by a

5
For what follows, see above all A.M. Honoré, “The Making of the Theodosian
Code”, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung 103 (1986)
pp. 133–222; followed up by his Law in the Crisis of Empire, 379 – 455 A.D. The
Theodosian Dynasty and its Quaestors (Oxford 1998); with, as background, his Emperors
and Lawyers (London 1981).
24  

new expansiveness and sonority, which has done much to persuade

historians of the depth of the empire’s problems—especially given
that the compilers of the Theodosian (435–8) and Justinianic (529–34)
Codes generally obeyed their instructions to preserve only what was
of general import, deleting all traces of any pronouncement’s specific
context. Two implications of this position are crucial for the under-
standing of sub-Roman law-making. First, appearances in the Codes
to the contrary, later imperial law continued to be made, as ever
since the first Caesars, in response to petitions and pressures from
citizens.6 That law not be generated by the solicitions of special in-
terests was one of the more spectacularly disregarded late Roman
prohibitions. Second, imperial decrees were “vulgarized”, like Roman
law throughout the post-Constantinian empire (except perhaps along
the Beirut-Constantinople axis). They were products of the educa-
tional system that qualified literary specialists to rule the state, a sys-
tem in which law of course had its place, yet subsidiary to the
imperatives of a properly Demosthenic or Ciceronian training.7
The Ausonian model of governance survived to steer post-imperial
regimes in the same direction. The Variae of Cassiodorus were com-
posed—and preserved—as literary showpieces by a highly trained
Italian rhetorician, but this does not mean that they were any the
less official pronouncements by Italy’s Ostrogothic kings.
The lamentable grievance of Constantius and Venerius has moved my
[King Athalaric’s] serenity: they complain that they have been deprived
by Tanca of their legal property [. . .] They add that, lest they should
try an action to reclaim their own, he is imposing on them the sta-
tus of meanest slavery. Therefore your mightiness, in obedience to this
decree [decretum], is to command the aforementioned person to attend
your court. There the whole truth of the case is to be examined, and
you are to dispense a justice that accords with law.
Tanca’s name implies a Gothic identity, as does that of the letter’s
addressee, Cunigast vir illustris, whose own depredations Boethius
claimed to have obstructed.8 A Roman official was making a Gothic

6
The classic account is now F. Millar, The Emperor in the Roman World, 31 B.C.–A.D.
337 (London 1977).
7
On all this, the fundamental account remains that of F. Wieacker, Allgemeine
Zustände und Rechtszustände gegen Ende des weströmischen Reichs, Ius Romanum Medii
Aevi 1,2a (Milano 1963); but note the reservations of Honoré, Law in the Crisis of
Empire, pp. 19–23.
8
Cassiodorus, Variae 8,28, ed. T. Mommsen, MGH AA 12 (Berlin 1894) [hence-
  LEGES BARBARORUM 25

king see that justice was done to Romans by Goths. The Variae also
included a “General Edict” in Athalaric’s name (533–4), decrying
sundry current breaches of civilitas: pervasio and other judicial mal-
practice, adultery, concubinage and bigamy, extorted gifts, sorcery
and homicide, and abuse of appeals procedure.9 The same topics
and many others no less typical of later imperial law-giving feature
in the well-known “Edict” of Theodoric himself (493–526), a short
code explicitly directed at “barbarians” and Romans.10 Now, Cassio-
dorus announced in Theodoric’s name that cases among the Goths
were to be settled by his “edicts”, those of Romans were to go
before “Roman examiners”, and those between Goth and Roman
should be “decided by fair reason in association with a Roman
jurisconsult”; “so each may keep his own laws, and with various
judges one Justice may embrace the whole realm”. The implication
here is that Theodoric laid down written law for Goths. That would
square with the remark of the “Valesian Anonymous” that “by the
Goths, because of his edict in which he established justice, he was
judged to be in all respects their best king”.11 Theodoric may have
done something of the sort. His Ravenna scribes produced the most
spectacular monument of early barbarian culture in the Codex Argenteus
of the Gothic Bible; not to mention charters signed manually in
Gothic script and language. Jordanes, probably following Cassiodorus’
lost Gothic History, knew of written belagines traceable to the dawn

forth: Variae], translations: Cassiodorus, Variae, ed. and transl. S.J.B. Barnish, Translated
Texts for Historians 12 (Liverpool 1992) p. 106 [henceforth: transl. Barnish]; Boethius,
The Consolation of Philosophy 1,4, ed. and transl. S.J. Tester, Loeb Classical Library
[The theological tractates] (2nd edn., London 1973) pp. 148–9.
9
Variae 9,18; 9,19–20 (transl. Barnish, pp. 116–21).
10
Most conveniently edited by J. Baviera in Fontes Iuris Romani Anteiustiniani 2, ed.
S. Riccobono et al. (2nd edn., Florence 1968–69) pp. 683–710: for its remit, see
Prol.; for injustice: 9; 55; 73–74; 91; 103; 129; on potentiores etc.: 43–47; 145; mar-
riage: 36–39; 54; 59–67; 93; slaves: 48–49; 56; 68–71; 78–87; 94–96; 117–118;
120–121; 128; 148; inheritance etc.: 23–33; 90; rape: 17–22; 92; homicide: 15–16;
99. Though it has been strongly argued that the Edict is that of one of the Visigothic
Theodorics in fifth-century Gaul, Dr Barnish has pointed out to me that clauses
10 and 111 seem to reflect Italian conditions, and that Variae 4,10 probably refers
to clauses 123–124; since there is no reason to suppose that Cassiodorus issued all
Theodoric’s decrees, the “Edict’s” lack of Cassiodoran rhetorical ornament is no
objection to its Ostrogothic provenance.
11
Variae 7,3 (transl. T. Hodgkin, The Letters of Cassiodorus [Oxford 1886] p. 321);
Anonymus Valesianus, Fragmenta 2,12, ed. and transl. J.C. Rolfe, Ammianus Marcellinus
12 (Cambridge Mass. 1939) pp. 544–5.
26  

of Gothic history.12 Yet all we can know for sure is that Ostrogothic
kings made law for Goth and Roman alike. Roman officialdom gave
barbarian kings the role of making law for Italy, as they were also
entrusted with its defence.
Italy was always a special case. But it is not unreasonable to extrap-
olate from Italian paradigms when sifting evidence of poorer qual-
ity from other Latin-speaking provinces. Sidonius Apollinaris was no
Cassiodorus, yet his poems and letters strongly suggest that his friends
took on Cassiodoran functions for the Visigothic and Burgundian
kings of southern Gaul.13 Unfortunately, Gallic legal texts are very
erratically preserved. The apparently earliest phase of Visigothic law-
making survives only as a fragmentary palimpsest. This may well be
equated with the first written record of Gothic law, dated by Isidore
of Seville to the reign of Euric (466–84). But if so, Euric seems to
refer to laws of Theodoric I (419–51).14 Likewise, the extant Lex
Burgundionum was most probably the work of Sigismund in 517. But
Gregory of Tours reported that Gundobad (474–516) issued “leges
mitiores [. . .] that there be no undue oppression of the Romans”,
and the Burgundian code was later called Lex Gundobada.15 The best
resolution of these conundra is that Theodoric I and Gundobad had

12
Chartae Latinae Antiquiores 20,704, ed. A. Bruckner and R. Marichal (Zurich
1982); Die nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700 2, ed. J.-O.
Tjäder (Lund-Stockholm 1955) pp. 95–6. On Jordanes and the belagines: Jordanes,
Getica 11, ed T. Mommsen, MGH AA 5,1 (Berlin 1882) pp. 73–5, see comment
by P.J. Heather, Goths and Romans 332–489 (Oxford 1991) p. 36. The fact remains
that belagines is not in Cassius Dio’s account of the Getic proto-hero, nor does it
seem to be a Greek word; whereas the word’s central syllable (cf. OE lagu!) makes
a strong case for regarding it as Germanic and legal: S. Feist, Vergleichendes Wörterbuch
der Gotischen Sprache (Leiden 1939) p. 91.
13
Sidonius Apollinaris, Carm. 7,311–13, ed. and transl. W.B. Anderson [Poems and
Letters], Loeb Classical Library, 2 vols. (Cambridge Mass. 1936/65) vol. 1, p. 144:
the future Emperor Avitus as assertor legum under Theodoric I; Carm. 5,562, vol. 1,
p. 108: Magnus of Narbonne dictat modo iura Getis when prefect in 458–9; Ep. 2,1,3,
vol. 1, p. 416: Seronatus leges Theodosianas calcans, Theudoricianasque proponens; Carm.
23,447–9, vol. 1, p. 312 and Ep. 8,3,3, vol. 2, p. 408: Leo of Narbonne eclipsing
Appius Claudius, and putting declamationes in Euric’s mouth which frenat arma sub
legibus; Ep. 5,5,3, vol. 2, p. 180: Syagrius a novus Burgundionum Solon in legibus dis-
serendis.
14
Codex Euricianus 277 (significantly on the sortes Gothicae et tertias Romanorum); 305;
and cf. 327?, ed. K. Zeumer, MGH LL nationum Germanicarum 1 (Hannover
1902) pp. 5; 16; 25 [henceforth: Cod. Eur.].
15
Gregory of Tours, Historiae 2,33, ed. B. Krusch and W. Levison, MGH SSrM
1,1 (2nd edn., Hannover 1951) p. 81. For this persuasive view of the genesis of Lex
Burgundionum, see Ian Wood this volume, pp. 253–4.
  LEGES BARBARORUM 27

issued individual laws on particular problems brought to their atten-

tion, as emperors and Ostrogothic kings did. Much of the Burgundian
Lex gives just that impression, and the topics covered by the Italian
texts recur there and in the Visigothic material.16 That these laws
and others should have been gathered into “codes” by their succes-
sors was simply a further case of barbarian-led regimes following the
Roman exemplum.
If the Italian analogy holds, there is no reason to think that Euric’s
or Sigismund’s codes were targeted only at their own peoples, any
more than were the Ostrogothic edicts of Theodoric or Athalaric
(i.e. Cassiodorus). The Burgundian Lex never said it was. Very much
more of it concerns Burgundians and Romans in inter-relationship
than Burgundians alone.17 The Visigothic palimpsest gives no war-
rant for the widespread view that Euric’s code consisted exclusively
of Gothic law for Gothic use. Isidore’s Historia Gothorum says merely
that “under this king [Euric], the Goths began to have instituta legum
in writing, for before they were governed by tradition and custom
alone”. That was not to say that Euric legislated only for Goths;
Isidore need have meant no more than that Goths were for the first
time integrated into the lex scripta regime of the old Empire.18 The
Roman law of the Visigothic kingdom was recodified in Alaric II’s
name in 506, and a shorter equivalent is extant for the Burgundian
realm. But neither of these points sustain the proposition that Euric’s

16
This is especially clear with the Novellae, cf. Liber Constitutionum 42–55; 62; 64;
74–81, ed. L.R. von Salis, MGH LL nationum Germanicarum 2,1 (Hannover 1892)
[henceforth: Lib. Const.]; Constitutiones Extravagantes 20, ed. L.R. von Salis, MGH LL
nationum Germanicarum 2,1 (Hannover 1892) [henceforth: Const. Extr.]. Lib. Const.
42; 52; 62; 76; 79 and Const. Extr. 20 bear their own dates, Lib. Const. 42; 76 and
79 being apparently laws of Gundobad. Among topics covered in n. 10, cf. on
injustice: Lib. Const. 1 and Cod. Eur., pp. 28; 31–2 for probable fragments of the
Codex Euricianus, with Lex Visigothorum, ed. K. Zeumer, MGH LL nationum Germani-
carum 1 (Hannover 1902) pp. 46–79 [henceforth: Lex Vis.] for a Spanish law of
546 on judicial expenses; on potentiores etc.: Cod. Eur. 276–277; 312; Lib. Const. 22;
54–55; 60; 84; Const. Extr. 21,12; on marriage: Cod. Eur. 319; Lib. Const. 12; 34;
36; 44; 52; 61; 66; 68–69; 86; 100–101; on slaves: Cod. Eur. 288; 300; Lib. Const.
6–7; 20; 39; 77; on inheritance etc.: Cod. Eur. 320 ff.; Lib. Const. 14; 24; 42; 51;
53; 59; 62; 65; 74–75; 78; 85; 87; on rape, Lib. Const. 30; 35; on homicide: Lib.
Const. 2; 10.
17
Figures from Lib. Const., p. 11.
18
Isidore of Seville, Historia Gothorum, Wandalorum, Sueborum 35, ed. T. Mommsen,
MGH AA 11 (Berlin 1894) p. 281; Isidore’s information perhaps came from the
lost prologue of the code of Liuvigild that finally superseded Euric’s (below, notes
50 and 51).
28  

or Sigismund’s own codes were solely for barbarians, nor of course

do they mean that Alaric’s Breviarium in any way “replaced” the Codex
Euricianus.19 In sum, the barbarian kings of Italy and southern Gaul
legislated in the first instance for all their peoples, not to supersede
but to supplement the law they had inherited from their imperial
predecessors. It had become an important part of their new job.

III

With the rise to dominance in Gaul of Clovis’ Franks, the legisla-

tive picture changed quite sharply. Their famous (one might almost
say infamous) Lex Salica is most plausibly credited to Clovis himself,
probably before he eliminated the Visigothic kingdom of Toulouse
(507).20 But the first thing to be said about it, and one of the most
important, is that it is not professedly a royal text. In what seems its
earliest form, that represented by the four “A” manuscripts (even
the earliest of which, however, is as late as the 760s), there is no
formal prologue at all. The so-called “Shorter Prologue”, clearly pri-
mary even though it is accompanied by the “Longer Prologue” on
all but one of its appearances, may date to the 590s, but it does not
need to be any earlier than the Liber Historiae Francorum of 727, which
quotes and amplifies it. As the clearest “ethnic” statement to be
found in any early Germanic legislation, it deserves quotation:
It has been accepted and agreed among the Franks and their leaders
[proceribus] that for the sake of keeping peace among themselves, all
intensified dispute [incrementa rixarum] should be curtailed, so that just
as they stand out among their neighbours for the strength of their
arm, so they may also excel them in authority of law, and thus put
an end to criminal behaviour [sumerent criminalis actio terminum] [. . .]
Hence, there came forward among them, chosen from many [electi de
pluribus] four men by name Wisogast, Arogast, Salegast and Widogast
who, assembling in three courts and carefully debating the sources of
litigation [causarum origines] gave judgement on each.21

19
Justified rebuttal of the case that the remit of the Breviarium was “territorial”
(covering all subjects) underlies the view that Euric’s code was “personal”: cf. H. Wolf-
ram, History of the Goths (Berkeley 1988) pp. 194–7 and nn.
20
I.N. Wood, The Merovingian kingdoms, 450–751 (London-New York 1994) pp.
108–13, supersedes all previous discussions of dating, in particular the view (adopted
by myself and many others) that the code must post-date 507.
21
Pactus legis Salicae “C”, Prol., ed. K.A. Eckhardt, MGH LL nationum Germa-
  LEGES BARBARORUM 29

The “Longer Prologue” almost certainly dates from the 760s and is
the product of King Pippin’s Chancery.22 It celebrates the Frankish
gens with rhythmic stridency, and it also credits Clovis (and his suc-
cessors) with the “amending” of “whatever seemed less fitting in the
Pactus”.23 But the four mysterious electi are still the primary authors,
though now rectores; either way they are still not kings. For its part,
the Liber Historiae Francorum placed them in villae beyond the Rhine.
It seems, then, that the Lex Prologues reflect a vision of the code as
deriving from a remote, “pre-invasion” Frankish past, of the sort
that the Liber itself liked to dwell on at length.24
Up to a point, the content of the original (i.e. “A”) code supports
this vision. It is singularly devoid of Christian traces. It is repeatedly
glossed by vernacular and presumably Frankish words, the “Malberg”
glosses.25 Roman influence has been detected, even in the most sur-
prising places.26 But it is a good deal less evident than in (for exam-
ple) the Burgundian laws, or indeed the “edicts” of the Merovingian
kings that were to follow (below, pp. 39–40). Yet if the code is (rel-
atively) un-Roman, for all that it was composed in Latin, how
specifically Frankish is it?
Given the discoveries of the last two generations of historians in
the field of Roman “vulgar” (or, for these purposes, “provincial”)
law, we can no longer suppose that a community of “free” men,

nicarum 4,1 (Hannover 1962) pp. 2–3 [henceforth: Pactus], and for Eckhardt’s view
of its date (criticized by Wood, as above), see his Pactus legis Salicae, ed. K.A. Eckhardt,
Germanenrechte 1 (Göttingen 1954) pp. 170–2; and cf. Liber historiae Francorum 4,
ed. B. Krusch, MGH SSrM 2 (Hannover 1888) p. 244. Eckhardt’s skilful con-
struction of a homogenized text tends to conceal variation between versions, and
it is often wisest to follow the texts arranged in his columns below. In this instance,
it seems that only “C5”, a late eighth-century Luxeuil MS, has the “Shorter Prologue”
alone—its text possibly taken from the Liber Historiae Francorum itself.
22
Lex Salica “D/E”, Prol., ed. K.A. Eckhardt, MGH LL nationum Germanicarum
4,2 (Hannover 1969) pp. 2–9; B. Merta, “Politische Theorie in den Königsurkunden
Pippins”, Mitteilungen des Instituts für Österreichische Geschichtsforschung 100 (1992) pp.
117–31.
23
The “Longer Prologue” here follows the “Epilogue”, cf. Pactus “A”, p. 253, a
doubtfully official text dating from revisions by Bishop Leodegar of Autun: Wood,
The Merovingian kingdoms, pp. 113–4.
24
See above, note 21.
25
Cf. Hans-Werner Goetz this volume, pp. 333–4.
26
As with the case that the historically fateful restriction of the heirs of terra Salica
to males, Lex Salica 59,6, pp. 222–3, derives from the Roman army’s needs: T. Ander-
son Jr., “Roman military colonies in Gaul, Salian ethnogenesis and the forgotten
meaning of Pactus Legis Salicae 59.5”, Early Medieval Europe 4 (1995) pp. 129–44.
30  

pre-occupied with its animals, its boundaries, and its mutual injuries
deliberate or accidental, is ex hypothesi “Germanic”.27 But there are
two good reasons to think that the law brought to light in Lex Salica
was that of a self-consciously “barbarian” culture. In the first place,
its keynote is compensation paid to an injured by the injuring party,
including the kin of each. The harmonics underlying that note are
those of bloodfeud: if payment was not made by the perpetrator
and/or his kin to the victim and/or his, then revenge would be
taken on any of the former by any of the latter.28 Social anthro-
pologists have made this process hugely better understood than it
was fifty years ago. We now see that it means much more than the
“interminable antiphony of violence” feared by the great nineteenth-
century English legal historian, F.W. Maitland.29 Yet, whatever its
peace-keeping validity, it had not been Roman law’s approach to
social discord since the time of the Twelve Tables. Under the empire,
initiative in placating or repressing disorder lay with government
authority, an agency which in Lex Salica is only a last resort.30 Perhaps
the presuppositions of feud had infiltrated Roman provincial justice
before imperial power collapsed in the West. If so, one must sup-
pose that ultra-frontier influence was already making itself felt.31 At
the same time, feud is so ubiquitous a feature of literature in the
Germanic vernacular, that to ascribe its prominence to anything other
than the custom of the invaders flies in the face of common-sense.32

27
Cf. the comments of R. Collins, Early Medieval Spain. Unity in Diversity, 400–1000,
New Studies in Medieval History (London 1983) p. 28, about the Byzantine “Farmer’s
Law”.
28
This is therefore the point not only of the many clauses defining payment for
damage, death or injury, Pactus 15–17; 23–34; 41–43 etc. (see Appendix below),
but also those regulating the extent and liabilities of kindreds: ibid. 58; 60; 62.
29
The key paper (in any language) remains that of J.M. Wallace-Hadrill, “The
Bloodfeud of the Franks”, Bulletin of the John Rylands Library 41 (1958/59) pp. 459–87,
repr. in his Long-Haired Kings, pp. 121–47; drawing on a famous paper by Max
Gluckman, “The Peace in the Feud”, id. Custom and Conflict in Africa (Oxford 1956)
pp. 1–26; for a sophisticated account of feud as reflected in Icelandic saga, see W.I.
Miller, Bloodtaking and Peacemaking: Feud, Law and Society in Saga Iceland (Chicago 1990).
30
E.g. Pactus 50; 56.
31
As, later, with a Roman law text from seventh-century Italy: C. Wickham,
Early Medieval Italy. Central Power and Local Society 400–1000, New Studies in Medieval
History (London 1981) p. 118. It ought conversely to be remembered, with refer-
ence to possibly Roman influence on barbarian codes, that imperial practices were
well capable of penetrating beyond the frontier before the late-fourth century, e.g.
through returning army veterans.
32
It is not, for example, a noticeable element in the “Farmer’s Law” (above, note
  LEGES BARBARORUM 31

The second consideration is less ambivalent. A famous clause of

the Lex distinguishes between compensations payable for killing var-
ious classes of subject:
If anyone kill a free Frank or barbarian who lives by Salic law [ingenuum
Francum aut barbarum, qui lege Salica vivit] [. . .] let him against whom it
is proved be liable for [iudicetur] [. . .] 200 solidi [. . .] But if anyone
kill him who is in the lord’s trust [truste dominica] [. . .] let him be liable
for [. . .] 600 solidi [. . .] But if a Roman man, a guest of the king
[Romanus homo, conviva regis] should have been killed, let him be liable
for [. . .] 300 solidi. But if a Roman man, a landholder [Romanus homo,
possessor] should have been killed, let him who is proved to have killed
him be liable for [. . .] 100 solidi. If anyone kill a Roman tributary
[tributarium] [. . .] let him be liable for [. . .] sixty-two and a half solidi.
Likewise, a Roman robbing a “Salic barbarian” (previously identified
as ingenuum) is liable for sixty-two and a half solidi, must find twenty-
five oath-takers to clear himself, or go to the hot-water ordeal; whereas
a “Frank” robbing a Roman clears himself with twenty swearers or
pays thirty solidi.33 The first thing these clauses make clear is that a
Frank, whatever else he or she was, was not a Roman. Second, a
Frank was also both a “barbarian” adhering to “Salic” law and an
ingenuus. The word “barbarian” does not otherwise occur; “Frank”
only in additions to the original sixty-five “titles”; and “Salic” five
more times with special reference to the Lex.34 On the other hand,
ingenuus is the standard qualifier for homo and appears hardly less
often. It is no outrageous deduction that its normal meaning is the
one it has in the clauses cited; hence, that “barbarism”, “Frankishness”
or (as it were) “Salicity” are specified in these clauses because a dis-
tinction is being drawn with Romans. In a further set of clauses,
killing a hominem ingenuum with a (war)band costs 600 solidi, or if he
be in truste dominica 1800 solidi, whereas slain “Romans or leti” are
worth half the amount (i.e. 300 solidi ). This is most significant because
laetus was the late Roman term for a barbarian settler-soldier. In
Æthelberht’s Kentish law-code, the læt is a figure of lower status than

27): W. Ashburner, “The Farmer’s Law”, Journal of Hellenic Studies 32 (1912) pp.
68–95, especially pp. 73–5.
33
Pactus 41,1; 41,5; 41,8–10; 14,1–3.
34
Ibid., 32,3–4 (it is again twice as expensive to tie up a Frank as a Roman);
38,2; 45,1; 47 tit.; 52,3; 57,1; 59,6 on terra Salica (above, note 26) is a “C” gloss,
while 63,1 is from Herold’s ghost “B” text. See also the next note.
32  

the ceorl/“free man”.35 In other words, the “free” (?Frank) is being

marked out from those barbarians previously settled on Roman soil.
The final and crucial conclusion to be drawn from these clauses
is that we here encounter the sort of “ethnic engineering” that is
also detectable in the legislation of the West Saxon King Ine (688/94).36
The upshot is that anyone not already laying claim to Frankish eth-
nicity would find that his (or her) legal position became up to twice
as secure if they proceeded to do so. It does not in the end seem
likely that the brilliant young southerners at the courts of Chlothar
II and Dagobert I, like Desiderius of Cahors or Eligius of Limoges,
were convivae regis, equivalent and so subordinated to Franks of the
king’s trust. But less socially self-confident “Romans” might well choose
to assimilate to a ruling-class defined in this way, and likewise their
social inferiors, whether possessor, tributarius, or letus. In this sense, Lex
Salica was not so much a reflection as a distillation of Frankish ethnic
identity.37
It looks, therefore, as though Lex Salica amounted to what its
Shorter Prologue implied. It was an attempt to encapsulate the law
whereby the new masters of northern Gaul believed that they and
their followers had lived hitherto, and whereby they wished that their
new subjects should proceed to live themselves. We should perhaps
not forget that if we are to understand the people we call “Franks”
as in some sense the residue of the garrisons that had guarded the
Rhine frontier, merged with those against whom they had guarded
it, then the essence of a Roman army was its “other ranks” and
“footsloggers”, not its officers. We should thus expect that the pre-
occupations of the Lex should lie with “rank-and-file” free men rather
than with a warrior “élite”. Given what else we know about the
composition of the last generations of the western army, it is no sur-

35
Ibid. 42,1–2; 42,4; Laws of Æthelberht 26, ed. F. Liebermann, Gesetze der Angelsachsen
1 (Halle 1903) p. 4. On the standard late Roman laetus, cf. A.H.M. Jones, The Later
Roman Empire 284–602. A social, economic and administrative survey, 3 vols. (Oxford 1964)
vol. 2, p. 620; vol. 3, pp. 186–7.
36
Laws of Ine 23,3; 24,2; 32–33, ed. F. Liebermann, Gesetze der Angelsachsen 1 (Halle
1903) pp. 100–3. Cf. B. Ward-Perkins, “Why did the English not become British?”,
English Historical Review 115 (2000) pp. 513–33, here pp. 523–4.
37
A conclusion that bears out that of M. Springer, “Gab es ein Volk der Salier?”,
Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher Personennamen, ed. D. Geuenich,
W. Haubrichs and J. Jarnut, Ergänzungsbände zum Reallexikon der Germanischen
Altertumskunde 16 (Berlin-New York 1997) pp. 58–83.
  LEGES BARBARORUM 33

prise that these men should envisage themselves as “barbarian”, or

should set out to distinguish their law from that of Rome. We may
reasonably conclude that what Clovis and the other authors of Lex
Salica sought to do was to assert the “respectability” of their own
traditions as against not only those of the empire but also of their
Visigothic and Burgundian rivals for the control of Gaul. This, then,
is not exactly a case of “Frankish” authorities carrying on where
emperors and Praetorian Prefects had left off. It is an effort to reg-
ister the presence of “Salic” Franks among the community of “civ-
ilized” peoples, whose law was codified. That being so, it is no
wonder that the Lex remained as good as unchanged by centuries
of social development in Frankish Gaul.38 As a vehicle of tradition,
adaptability was not its function.
The Frankish example may indirectly have influenced at least two
other essays in barbarian codification. The earlier was that of King
Æthelberht of Kent (c. 580–616/18), Clovis’ great-grandson-in-law.39
Like Lex Salica, his code is in fact anonymous: the first evidence of
his authorship is supplied by Bede over a century later.40 Bede
regarded what Æthelberht did “with the counsel of his wise men”
as of benefit to his gens, by which he presumably meant the people
of Kent. He reports that the laws were “written in English speech”,
that they were still “kept and observed”, and that Æthelberht acted
“according to the examples of the Romans”. The first of these three
assertions is conspicuously true. The code’s language is by far the
oldest surviving specimen of the Anglo-Saxon vernacular in extenso;
and even in its only extant copy, a manuscript of the 1120s, it is
studded with every sort of linguistic archaism. How and why it came
about that, alone of the West’s new rulers, those of Kent and (at
the end of the seventh century) Ine of Wessex legislated in a Germanic

38
Cf. A.C. Murray, Germanic Kinship Structure. Studies in Law and Society in Antiquity
and the early Middle Ages, Pontifical Institute of Mediaeval Studies, Studies and Texts
65 (Toronto 1983) pp. 117–225; and Wormald, Making of English Law, pp. 44–5.
39
Laws of Æthelberht, pp. 3–8. For a fuller statement, see P. Wormald, “ ‘Inter
Cetera Bona . . . Genti Suae’: Law-making and Peace-Keeping in the earliest English
Kingdoms”, La giustizia nell’alto medioevo (secoli V–VIII), Settimane di studio del cen-
tro italiano di studi sull’alto medioevo 42 (Spoleto 1995) pp. 963–96, esp. pp. 969–74
[repr. id., Legal Culture in the Early Medieval West. Law as Text, Image and Experience
(London 1999) pp. 182–6]; and my Making of English Law, pp. 93–101.
40
Bede, Historia Ecclesiastica gentis Anglorum 2,5, ed. B. Colgrave and R.A.B. Mynors
[Bede’s Ecclesiastical History of the English People] (Oxford 1969) pp. 150–1.
34  

language are almost unanswerable questions. What matters here is

that this point puts the fundamentally “Germanic” character of such
feud-focussed law beyond doubt. We can scarcely suppose that “Vulgar
Roman” (or Celtic) law was carefully transposed into an alien tongue
so as to suit the sensibility of a restricted “élite”. But what, then,
did Bede mean by saying that the code was “according to the exam-
ples ( plural ) of the Romans”? 41 A plausible answer is that Romani for
Bede encompassed the whole range of sub-Roman regimes; and hence
that what he (or Æthelberht’s Kentish wise men) had in mind were
codes known to have been issued by some of the emperors’ succes-
sors. If so, the genesis of written English law as a conscious emula-
tion of other law-making becomes especially clear—much as southern
English noblewomen seem to have adopted Mediterranean fashions
in the seventh century.42 Æthelberht’s great-grandson thought of his
code as the “just customs of the Cantwara”.43 Even in Britain’s largely
de-Romanized climate, his people had, like the Franks ruled by his
in-laws, joined the ranks of the civilized.
About a quarter of a century later, the Lombards of Italy followed
suit. King Rothari’s massive 388–clause edict (643) was perhaps itself
under Frankish inspiration and has resemblances to Lex Salica. Much
of its technical vocabulary is Germanic. It makes the feuding ratio-
nale of its compensation system explicit: “Feud is to be deferred”
( postponatur; not “limited”, still less “eschewed”).44 Rothari left little
doubt of his intentions in drafting his code:
We have sought out [. . .] the ancient laws of our ancestors which
were not written and, with the equal counsel [. . .] of our [. . .] lead-
ing judges, and all our most happy army assisting, we [. . .] have
ordered them to be written onto this parchment, so that we should
include in the edict what we could ourselves or through the old men
of the people [gentis] recover of the ancient law of the Lombards.45
This is from Rothari’s epilogue. The sentiments voiced by his self-
confidently royalist Prologue—sollicitude for his subjects’ “welfare”,
hope that his code will permit “each man to live quietly in secure

41
On this, see Wallace-Hadrill, Early Germanic Kingship.
42
See Barbara Yorke this volume, pp. 307–8, and references.
43
Laws of Wihtred Pr., ed. F. Liebermann, Gesetze der Angelsachsen 1 (Halle 1903)
p. 12.
44
Edictus Rothari 74–5; 45, ed. F. Bluhme, MGH LL 4 (Hannover 1868).
45
Ibid., 386.
  LEGES BARBARORUM 35

law and justice”—are more like what a sub-Roman society expected

of its lawmakers.46 They thus signal some of the differences between
his Edict and Lex Salica. It is a point that the Edict is dated. It is
authenticated by a “notary” Ansoald, and only copies so authorized
are to be used in disputes.47 Above all, as its very size implies, it is
hugely more comprehensive than the Frankish code. (Re-)marriage
is addressed by Lex Salica only in one of the mysterious anonymous
“capitularies” following the code in one or two manuscripts. In
Rothari’s Edict—as, regularly, in the Burgundian Code—it is cov-
ered from most conceivable angles.48 How are we to explain that so
“Germanic” a legal statement, one with even less to say of Romani
than Lex Salica, has such conspicuously “Roman” features? The nat-
ural answer is that Rothari and his “judges” could command the
services of men trained in the techniques of Roman law and legis-
lation: men like the notary Ansoald himself (for all his Germanic
name). In other words, the Edict, like Theodoric’s putative “edicts”,
was an attempt to write the Lombards into the still lively legal cul-
ture of post-Roman Italy.49 Rothari did not say how the result was
to relate to the Roman law by which most of his subjects may be
assumed to have lived. It would be a reasonable guess that some-
thing like Theodoric’s own rule applied (above p. 25).
The law of the Visigothic kingdom was heavily Romanized
from the outset (above pp. 27–8). In the sixth and seventh century it
followed this trajectory further. The next king after Euric credited
by Isidore with legislation was Liuvigild (569–86). This was one
aspect of a consciously Romanizing policy, which also ran to his
installation on a throne, his issue of his own gold coinage, his enrich-
ment of the fisc through confiscations, his foundation of a city in
his son’s name, and (thought his Catholic critics) his religious per-
secution of religious deviants: i.e. orthodox non-Arians.50 Again, a
precariously extant text seems to fit what Isidore describes: the 324

46
Ibid., Pr.
47
Ibid., 388.
48
Pactus 100–101; Edictus Rothari 178–204 passim; cf. Lib. Const. 24; 34; 44; 52
(a “Novel”, above, note 16); 61; 69; 74 (another).
49
For a recent example of many studies addressed to this problem, see N. Everett,
“Literacy and the law in Lombard government”, Early Medieval Europe 9 (2000) pp.
93–127.
50
Isidore of Seville, Historia Gothorum 51, pp. 287–8.
36  

chapters called Antiqua in the Visigothic code of c. 650 often appear

to build on the contents of the “Eurician” palimpsest. There is no
reason to doubt that Liuvigild legislated for all his subjects, like Euric,
so long as one allows that Roman law survived for its own purposes.
His best-known law lifts the ban on intermarriage between Goth and
Roman on the grounds that it was incongruous to keep asunder
those of equal dignity of birth. The most striking thing about this
law is not that it endorses what may long have been a “fait accom-
pli”, but that it most probably repeals a Roman proscription on the
marriage of barbarians and provincials, one going back via the Breviary
to the Theodosian Code.51 The emperor’s law could now be insou-
ciantly reformed by his barbarian successors.
What all of this comes down to is that law-makers of Germanic
origin—and with Germanic names, whatever language they now
spoke—were finding a variety of ways of accommodating what they
believed to be the customs of the followers who formed their mili-
tary power-base, within kingdoms whose legal culture was at least
residually Roman. It is hardly any easier to detect “Germanic” ele-
ments in the Visigothic Antiqua than in the fragmentary laws ascribed
to Euric. There is a handsome set of compensations for injury in
the Visigothic law-book, but these stand in the name of seventh-cen-
tury kings.52 Otherwise, the expedients adopted depended on the
level of Roman legal expertise surviving in the society in question.
In southern Britain, the one truly Roman aspect of legislation was
the fact that it was written. In northern Gaul, a rather crude and
far from systematic summary was made of aspects of apparently
Frankish custom; and not only was it made in Latin but it also con-
tained an invitation that Romans re-identify themselves as Franks.
In Italy, the traditions of the Lombard “army” (a significant term)
were gathered into an impressive corpus. None of these exercises
specifically excludes a continuing place for Roman law; and except

51
Lex Vis. 3,1,1 (Antiqua); cf. Codex Theodosianus 3,14,1 (Breviarium Alarici with the
usual Interpretatio), ed. T. Mommsen and P.M. Meyer (3rd edn., Berlin 1962).
Arguments that Liuvigild was repealing a Eurician prohibition (i.e. originally Gothic
legislation) tend to be circular: viz. that since Liuvigild’s concern was only with
Gothic law as decreed by Euric (as per the traditional interpretation of bi-polarized
Visigothic legislative history), there must have been a Eurician prohibition for him
to repeal.
52
Lex Vis. 6,4,1,3 (compare the Appendix below, pp. 47–53).
  LEGES BARBARORUM 37

in Kent, there is no reason to think that they actually did so. Post-
Roman legislation was moving towards a position where kings laid
down some sort of law for all their subjects. But only in Spain had
they got anywhere near making a systematic job of it by the mid-
seventh century.

IV

In fact, only the Spanish kingdom would ever truly realize such an
objective, in the work of the successive kings Chindaswinth and
Reccaswinth (642–72). Their law-book—later given the revealing
“sobriquet” Forum Iudicum (Court/Meeting-place [?] of Judges)—con-
sisted of a reissue of the Antiqua, complemented by 200 new or revised
laws of their own and their immediate predecessors, and arranged
in chapters, titles and books: twelve books, recalling the Twelve
Tables (which Isidore knew about) and Justinian’s code of a century
before (which he probably did not). This very Roman-looking enter-
prise had a very Roman purpose: it was to supplant all hitherto cur-
rent law, Roman as well as Gothic, every copy of which was to be
destroyed.53 It is no paradox that to abolish Roman law was a sin-
gularly Roman thing to do. Did not Justinian demand erasure of
every legal monument prior to his own? Hardly less (neo-)Roman
was another policy of later Gothic kings. Within a generation of
Reccared’s conversion to Catholic Christianity and the consequent
end of the Arian “persecution”, the heat was turned on another class
of nonconformists, the Jews.54 Byzantine emperors of the time were
doing the same. Spanish identity was to be homogeneous in obser-
vance as in law. It was a note that would sound all too often in
later phases of peninsular history.

53
Lex Vis. 2,1,10; cf. 2,1,3,5. Persuasively, P.D. King, “King Chindasvind and
the First Territorial Law-code of the Visigothic Kingdom”, Visigothic Spain: New
Approaches, ed. E. James (Oxford 1980) pp. 131–57. The code was again revised by
Erwig (681); an indication of multi-cultural Visigoth realities is that the very last
law, Lex Vis. 6,1,3 on the cauldron ordeal (Witiza, 698–710), sounds the least ambigu-
ously “Germanic” note in the whole code.
54
The first anti-Jewish law was Reccared’s own and much influenced by Roman
example, Lex Vis. 12,2,12; the next are those of Sisebut (612–21), Lex Vis. 12,2,13–14.
38  

Roman inspiration was no longer all that underlay the making

of written law in Spain. Hostility to the Jews must in part have
been prompted by the determination that Spaniards should replace
them as Chosen People of God.55 More than any regime before the
Carolingians, Visigothic legislators took their lead from the Church,
especially from the long series of councils of Spanish bishops held
at Toledo. The eighth and twelfth councils saw the publication respec-
tively of Reccaswinth’s and Erwig’s codes. The former undertook to
implement what his bishops thought conducive to justice; the latter
submitted his decrees for the assembled fathers’ approval.56 The Forum
Iudicum opens not with the usual prologue of legislative aspiration
but with a short book on the theory of law-making. The dominant
ideas are those of Isidore, whose pupil, Bishop Braulio of Saragossa,
has plausibly been thought its author.57 Law is “the soul of the whole
body politic”, guiding all its limbs from the royal head downwards.58
In Italy, law-making after Rothari followed a rather different course.
His Edict was in no way superseded, simply revised and extended
by regular legislative sessions, especially under Liutprand (712–44).
Since all of these new laws bore a date, like the Edict itself, it was
possible to see at a glance which law was current; this was just as
much the Roman practice as division of a code by book, title and
subject-matter, which Lombard law-makers eschewed.59 As one would
therefore expect, Romans and Roman law are much more promi-
nent than in Rothari’s laws. Lombards were formally (re-)introduced
to will-making to supplement gift by thinx or launegild, and to manu-
mission not only by gairethinx (bestowal of an arrow and whip at a
cross-roads) but also by leading a slave around an altar and/or draft-

55
On Old Testament inspiration in Visigothic royal liturgy, see M. McCormick,
Eternal Victory. Triumphal rulership in late antiquity, Byzantium, and the early medieval West
(Cambridge 1986) pp. 304–12.
56
Concilium Toletanum 8 Tom., p. 265; 12 Tom., p. 383, ed. J. Vives, Concilios
Visigóticos e Hispano-romanos, España cristiana 1 (Madrid-Barcelona 1963). See P.D.
King, Law and Society in the Visigothic Kingdom, Cambridge Studies in Medieval Life
and Thought (Cambridge 1972) pp. 23–39; but also Collins, Early Medieval Spain,
pp. 115–29, for more crudely political motivations.
57
C.H. Lynch, Saint Braulio, Bishop of Saragossa (631–51). His Life and Writings
(Washington DC 1938) pp. 137–40.
58
Lex Vis. 1,2,2; 2,1,4; cf. 2,1,2: Quod tam regia potestas quam populorum universitas
legum reverentie sit subiecta.
59
Cf. Codex Theodosianus 8,1,1,6 (Gesta Senatus).
  LEGES BARBARORUM 39

ing a charter that made freedom conditional on future service.60 Nor

was Romanization all that was involved. Some clauses in Rothari’s
Edict already look like ad hoc responses to particular pressures; with
much of his successors’ legislation, this is quite clear. Liutprand found
himself resolving what penalty should be paid by a man who had
removed a woman’s entire clothing while she was bathing in a river;
and that someone who had got himself killed by standing under a
water-hoist should have looked what he was doing.61 Lombard kings
were now supplying the legal needs of a whole society, whatever its
ethnic origins. Their law survived as their memorial long after their
kingdom had passed into abstruse history.62
Supplementary legislation took a different form again in the Frankish
kingdom. At least five kings in the century after Clovis’ death issued
decrees on current problems; and these diverged as markedly in char-
acter from the primary Frankish code as did their Italian counter-
parts from Rothari’s Edict. As in Spain and Italy, Frankish kings
were strengthening their sense of responsibility for law and order.
Some clauses in Lex Salica, as in Rothari’s Edict, already suggest a
more interventionist role.63 But Childebert I, Chlothar I, Chilperic,
Childebert II and Chlothar II were much more preoccupied with
the latro than was the Lex, and they organized pursuit of thieves by
“hundreds” from the trustis, of which the Lex said nothing.64 Following
perhaps in Visigothic and Burgundian footsteps and certainly in
Roman, Childebert II aggressively forbade the taking of revenge for

60
Leges Liutprandi 6; 19; 102 (cf. Edictus Rothari 172–5); 9–10 (cf. Edictus Rothari
224), ed. F. Bluhme, MGH LL 4 (Hannover 1968); Leges Ahistulfi 11, ed. F. Bluhme,
MGH LL 4 (Hannover 1968).
61
Leges Liutprandi 135–6.
62
C. Wickham, “Lawyers’ Time: History and Memory in Tenth and Eleventh-
Century Italy”, Studies presented to R.H.C. Davis, ed. H. Mayr-Harting and R.I. Moore
(London 1985) pp. 53–71.
63
F. Beyerle, “Über Normtypen und Erweiterungen der Lex Salica”, Zeitschrift
der Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 44 (1924) pp. 394–419;
Wood, The Merovingian Kingdoms, p. 110. Cf. E. Besta, “Le Fonti dell’ Editto di
Rotari”, Atti del primo Congresso internazionale di Studi Longobardi (Spoleto 1952) pp.
51–69.
64
Latrones (cf. Pactus 47): Pactus (Pactus pro tenore pacis) Pr.; 80; 83; 85–86; 89;
92; Pactus (Edictum Chilperici) 116; Pactus (Decretio Childeberti) 2,5; 3,1; 3,5; Centena
and trustis: Pactus (Pactus pro tenore pacis) 84; 99–102; Pactus (Decretio Childeberti)
3,1; 4–5. Cf. A.C. Murray, “From Roman to Frankish Gaul: ‘Centenarii’ and
‘Centenae’ in the Administration of the Merovingian Kingdom”, Traditio 44 (1988)
pp. 59–100.
40  

homicide, and laid down a death penalty both for killing and theft.65
More predictable additions were laws on sanctuary, consanguineous
marriage and Sunday observance.66 Yet the main text of Lex Salica
was hardly affected by these changes, nor was there any consistent
effort to incorporate them in manuscripts of the Lex so that they
could be seen to override it. More important was to maintain the
integrity of the traditions that the Lex symbolized.67
The other major departure of Merovingian kings, though in this
case seemingly in the seventh century, was also one notionally under-
taken by Rome: the issue of written laws for subject peoples hith-
erto denied that blessing. It is fairly clear that this was done for the
Alamans by Chlothar II and perhaps for the Bavarians by him or
Dagobert. If so, that might explain the otherwise unparalleled struc-
tural similarity between these codes, Rothari’s Edict and Æthelberht’s
laws.68 In most extant forms, the Alaman and Bavarian codes have
“gone native”: both were apparently revised and extended by “dukes”
of their own people (Lantfrid of the Alamans, Odilo and Tassilo of
the Bavarians) in the first half of the eighth century. This is as good
an illustration as one could wish of how a written law whose gene-
sis owed little to its own subjects’ aspirations could nevertheless come
to define them.
More important for the Franks themselves was the composition,
again probably under Dagobert, of the “Ripuarian” law. This was
allegedly the law of Rhineland Franks (or, viewed from another angle,

65
Pactus (Decretio Childeberti) 2,3,5; cf. Lex Vis. 6,1,8; Lib. Const. 2,7.
66
Sanctuary: Pactus (Pactus pro tenore pacis) 90; Pactus (Decretio Childeberti) 2,2;
irregular marriage: ibid. 1,2; Sunday: ibid. 3,7.
67
Cf. Wormald, Making of English Law, pp. 44–5.
68
Originally a theory of H. Brunner, “Über ein verschollenes merowingisches
Königsgesetz des 7. Jahrhunderts”, Sitzungsberichte der königlichen preußischen Akademie
der Wissenschaften 309 (Berlin 1901) pp. 932–55, now developed in my Making of
English Law, pp. 96–101. The Lex Baiwariorum Prologue (ed. E. von Schwind, MGH
LL nationum Germanicarum 5,2 [Hannover 1926]) is a mystery in its own right.
I argued at Bellagio that Theuderichus [rex Francorum] cum esset Catalaunis, at the out-
set of the text, can hardly be other than the hero of the great 451 battle with
Attila, who was a law-giver (above, note 14), and may thus have featured in the
prologue of the Codex Euricianus; it was certainly from the Codex Euricianus that the
draftsmen of the Lex Baiwariorum (Dagobert’s Midi glitterati ?) drew much of their
material; and the “Eurician” palimpsest is a S. Gallic MS (Codices Latini Antiquiores,
vol. 5, 624–7, ed. E.A. Lowe [Oxford 1934–71]). But see the more traditional inter-
pretation by Matthias Hardt this volume.
  LEGES BARBARORUM 41

of the Austrasians who were now coming to command the political

future of the Franks).69 Lex Ribuaria is in general so like Lex Salica as
to tempt one to wonder what mileage there was in drafting it at all.
The answer must be that merely to have a written lex had come to
be a point of critical political, if not indeed “ethnic” significance.
The compensation tables of the “Ripuarians”, as of the Alamans and
Bavarians, do in fact differ in detail from those of the “Salic” Franks
(see Appendix below). To contemplate the broadly similar yet invari-
ably divergent “tariffs” of the “Volksrechte” raises the thought that
these tables were in some sense the defining factor in any one lex,
expert knowledge of which thus identified the legal specialist. Certainly
suggestive is that the freshest research finds the most archaic lin-
guistic features of King Æthelberht’s code in its injury list.70 A tariff
recognizable as one’s own was perhaps itself an ethnic marker. At
the same time, Lex Ribuaria and the others have important further
features. One is treason law, found in Visigothic and Lombard pro-
nouncements but not hitherto in Frankish codes. Another is the atten-
tion given to written legal instruments.71 In other words, Frankish
legal culture was itself becoming more “Roman” at the time when
Lex Riburaria was composed. One most significant aspect of Lex Ribuaria
is the first explicit statement of what is often assumed to have applied
since barbarian leges appeared: that it was “according to Roman law
that the Church lives.”72 The Merovingian kingdom had thus by 650
established what we have no reason to think was a norm before-
hand: the principle that a people, sovereign or subject, had its lex;
while the law of Rome lived on in the Church.
English legislation in the seventh century continued to follow a
broadly Frankish pattern. With Ine’s code of c. 690, the West Saxons
joined the Kentishmen among peoples of written law. At least as
extant, however, this is unlike any other “primary” statement of post-
Roman law. Almost from the outset, it seems to address legal cruxes

69
The best recent account of this is Springer, “Gab es ein Volk der Salier?”,
pp. 71–3; and cf. Wood, The Merovingian Kingdoms, pp. 115–7.
70
See L. Oliver, “The Language of the Early English Laws”, Harvard Ph. D.,
1995, pp. 244–61.
71
Lex Ribuaria 72; 59; 61–62, ed. F. Beyerle and R. Buchner, MGH LL nationum
Germanicarum 3,2 (Hannover 1954); Leges Alamannorum 23–25; 1–2; 16; 18–19, ed.
K. Lehmann and K.A. Eckhardt, MGH LL nationum Germanicarum 5,1 (rev. 2nd
edn., Hannover 1966); Lex Baiwariorum 2,1–3; 1,1; 16,1–2; 16,15–16.
72
Lex Ribuaria 61,1.
42  

brought to the law-maker’s attention. Indeed, what is now known

as “Ine’s code” may in fact have been six or more separate and suc-
cessive pronouncements.73 There is no West Saxon injury tariff before
Alfred’s code of the early 890s, to which Ine’s is appended. Ine’s
concern was with current problems, and thus with the questions of
baptism, Lent and church-taxation as well as “ethnic” wergelds and
(above all) theft. The same considerations crop up in the exactly
contemporary Kentish code of Wihtred (695), which was probably
under Ine’s influence. In England as in Francia, written law had
generated its own momentum. Yet in each, the momentum then
gave out altogether for up to a century and was picked up only in
Carolingian conditions. The roots of lex scripta in northern Europe
were still shallow.

Before concluding this survey with a glance at what might be chris-

tened the “Carolingian equilibrium”, we should take note of a type
of legal document that was not legislative at all. Depressingly few
legal documents referring to particular transactions (for short, “char-
ters”) survive from the West before the mid-seventh century, and
then in rather marginal circumstances.74 Ravenna’s documents, the
obvious exceptions, have been sadly victimized by the enthusiasms
of eighteenth-century antiquarian collectors.75 At the same time, the

73
I have argued this at length (with table) in “Inter cetera bona . . .”, pp. 977–83
[repr. pp. 188–92].
74
E.g. Tablettes Albertini, actes privés de l’époque vandale, ed. C. Courtois, L. Leschi,
C. Perrat and C. Saumagne (Paris 1952): deeds on wood, dating 493–4, mostly
relating to sale, are all that survive from the Vandal kingdom (Wolf Liebeschuetz
this volume); and texts on slate mostly from Salamanca (Diplomatica Hispano-visigoda,
ed. A. Canellas Lopez [Zaragoza 1979]) represent more or less the sum total of
Visigothic diplomatic (though cf. R. Collins, “Conclusion”, The Settlement of Disputes
in early medieval Europe, ed. W. Davies and P. Fouracre [Cambridge 1986] p. 210).
75
The Ravenna papyri are edited by Tjäder (Die nichtliterarischen lateinischen Papyri
Italiens). Texts for sub-Roman Gaul are in the unsorted collection of J.-M. Pardessus
(Diplomata, Chartae, Epistolae aliaque instrumenta ad res Gallo-Francicas spectantia, 2 vols.
[Paris 1843/49]), to which key guides are K.H. Debus, “Studien zu merovingi-
schen Urkunden und Briefen”, Archiv für Diplomatik 13 (1967) pp. 1–109; 14 (1968)
pp. 1–192; and U. Nonn, “Merowingische Testamente: Studien zum Fortleben einer
römischen Urkundenform im Frankenreich”, Archiv für Diplomatik 18 (1972) pp. 1–129.
  LEGES BARBARORUM 43

corpus of formulae, mostly from Gaul but with a Visigothic outlier,

establishes beyond dispute that traditional procedures for drawing up
records of gift, sale, testament and judgement continued to be used.76
If we consider deeds of sale, the dominant pre-occupation of the
exiguous Vandal documentation, it is obvious that similar legal foun-
dations survived in each one-time western province. In Africa, Italy,
Spain and Gaul, there is regular phrasing, to the effect that the
buyer should “have, hold, possess, use and enjoy, for himself and
his heirs in perpetuity”.77 As might be expected, the formulae are more
regular, more obviously “legalese” in the more clearly Romanized
legal cultures of Italy and Spain. There can nonetheless be no doubt
that the origins of all are “vulgar” Roman. At the same time, by
no means all transactions have ecclesiastical connections. Even if the
formulae collections themselves must have come to rest in a bishop’s
or abbey library, we should look to the umbrose corporations of
post-Roman notaries for the drafting of this sort of document, just
as the variable levels of these officials’ professional skills may best
explain the erratic quality of post-Roman legislation.78 To say, then,
that Roman law in the post-Roman era was the “law of the Church”
is to miss the central point that it was the law of almost any significant
property transaction with a prospect of preservation, because Rome’s
law naturally dominated all forms of written business.
The seminal Formulary of Marculf and its derivatives make a fur-
ther no less important point. Included in them are formulae for the
pacification of feud.79 References to “Salic law” in these collections
are far from unknown, even if they apparently appeal to the whole
body of “Salic” custom, and are rarely if ever to specific points in

76
Definitively edited by K. Zeumer: Formulae Merowingici et Karolini Aevi, MGH
Formulae (Hannover 1886).
77
Tablettes Albertini no. 3,4b, p. 220 (493); cf. Die nichtliterarischen lateinischen Papyri
Italiens 36,2, pp. 116–7 (575/91), and Codice Diplomatico Longobardo no. 23, ed.
L. Schiaparelli, Fonti per la storia d’Italia (Rome 1929) vol. 1, pp. 89–91 (720);
Diplomata, Chartae, Epistolae no. 241, vol. 1, p. 227 (627); Formulae Andecavenses 8; 25;
27 (?earlier seventh century), MGH Formulae, pp. 7; 12; 13; Marculfi Formulae 2,19
(c. 700), ibid., p. 89; Formulae Visigothicae 11 (first half of seventh century), ibid., pp.
580–1.
78
I have argued this point for judicia in Making of English Law, pp. 89–92; cf.
W. Bergmann, “Untersuchungen zu den Gerichtsurkunden der Merowingerzeit”,
Archiv für Diplomatik 22 (1976) pp. 1–186.
79
Marculfi Formulae 2,18, pp. 88–9; Formulae Turonenses 38, p. 156.
44  

the code.80 Marculf even tackles head-on the “impious custom” of

exlusively male succession to alod, and the formulae make repeated
attempts to provide for succession in the female line.81 All this is of
more than merely “formulaic” importance. Like the novel elements
in Lex Ribuaria, it shows that at the landholding level there was some-
thing of an attempt to integrate Roman and “barbarian” traditions.
This is to say that even in the Francia of emergent “personal” law,
there were moves towards territorial homogeneity in documentary
practice.

VI

Close notice of the Merovingian, and indeed the Spanish and Italian,
roots of the Carolingians’ law-making reveals that, however vastly
they increased the quantity of documentary production, they did not
really change its character. Charlemagne’s own legislation was duly
affected by his experience in St Peter’s at Christmas 800. Within
months, he systematically revised Lombard law along self-consciously
Lombard lines, paying especial attention to the logic of compensa-
tion for injury. By the time two more years had passed, he is recorded
by his biographer and by one major series of annals to have set
about revising the leges of his subject peoples, and to have “given”
written codes to those who had had none before.82 The results are
indeed extant. So far as Lex Salica is concerned, there are three times
as many copies of Charlemagne’s Karolina as survive of all previous
versions. The new laws of Frisians, Saxons and Thuringians seem
to make an effort to distinguish these peoples’ customs as well as to
give them a common format (see Appendix, pp. 47–53). Over the
decades that followed, Charlemagne’s successors are regularly found
promising to “hand over” or “observe” the lex of each class of sub-
ject. But all this is to say that, empire or no empire, “ethnic” legal

80
My “Lex Scripta and Verbum Regis”, pp. 121–3 [repr. pp. 21–2].
81
Marculfi Formulae 2,10–12, pp. 81–3; Formulae Turonenses 21–2, pp. 146–7.
82
MGH Capit. 1, no. 98, c. 5, ed. A. Boretius (Hannover 1883) p. 205; Einhard,
Vita Karoli 29, ed. O. Holder-Egger, MGH SSrG 25 (6th edn., Hannover 1911)
p. 33; Annales Laureshamenses a. 802, ed. G.H. Pertz, MGH SS 1 (Hannover 1826)
p. 39. What follows in effect summarizes the argument of my Making of English Law,
chapter 2.
  LEGES BARBARORUM 45

practice still remained foremost. Society in Francia and its neigh-

bours may have changed almost beyond recognition in the three
centuries since Clovis. Yet the alterations made to the Salic and
other leges by his great successor were exiguous.
Here is the central paradox of Carolingian law-giving. At the same
time as Charlemagne and his successors were showing all possible
deference to the leges recognized by their regimes, kings and their
ideologists were seeking to remodel the behavioural profile of their
“Christian Empire” through a massive output of “capitularies”; edicts
that were by and large a lot more ecclesiastical in tone than their
Merovingian counterparts. The capitulary activity consummated in
Charlemagne’s solemn Admonitio Generalis of 789 actually had an
English offshoot, thanks to the visit of Bishop George of Ostia, papal
legate and in situ bishop of Amiens, to the courts of Kings Ælfwold
of Northumbria and Offa of Mercia in 786.83 The 789 Admonitio re-
issued the decretals of Pope Hadrian’s Dionysio-Hadriana in Charle-
magne’s own name. George (anyway in the extant 786 record) had
Ælfwold, Offa and their magnates subscribe decrees that he as
papal legate had drafted himself (perhaps with the assistance of
Alcuin). In each there was, needless to say, a marked scriptural ele-
ment.84 Yet it seems to have struck no legislator as unacceptably
anomalous that capitularies were declaring homicide to be an un-
acceptable blot on a God-fearing and Bible-reading society, while
the repeatedly endorsed leges continued to provide blithely for the
rhythms of personal vengeance.
Of especial interest is that the outcome of Carolingian efforts can
to some extent be checked, as in Merovingian times they cannot,
against extant manuscripts and recorded court-cases.85 The results
are much as might have been expected. The “personality of law”
regime meant that copies of multiple leges are a great deal com-
moner than those of a single lex; while copies of what was still the

83
MGH Capit. 1, no. 22, cc. 1–59, pp. 54–7; Alcuin, Epistolae no. 3, ed. E. Dümmler,
MGH EE 4 (Berlin 1895) pp. 19–29.
84
Cf. P. Wormald, “In Search of King Offa’s ‘Law-code’”, People and Places in
Northern Europe 500–1600. Essays in Honour of Peter Hayes Sawyer, ed. I.N. Wood and
N. Lund (Woodbridge 1991) pp. 25–45 [repr. id., Legal Culture in the Early Medieval
West. Law as Text, Image and Experience (London 1999) pp. 201–23]; C.E. Cubitt,
Anglo-Saxon Church Councils c. 650–c. 850, Studies in the Early History of Britain
(Leicester 1995) pp. 153–90.
85
Wormald, Making of English Law, pp. 30–70.
46  

standard Roman law text, Alaric’s Breviary, are as frequently found

associated with them as on their own. Capitularies, however, are by
no means so regularly associated with leges. They tend to circulate
as discrete collections, often with more explicitly ecclesiastical legis-
lation. The exception is, again, Italy, where capitularies issued by
Carolingian kings are almost always integrated with Lombard and/or
Frankish basic law. Further, it is only in Italy, and even more in
the Gothic, south-western, part of the Carolingian kingdom that one
is at all likely to find these texts cited in court. As in Roman times,
and as in the High Middle Ages, Italy, Spain and southern Gaul
were areas of legal specialism in ways that northern centres of juris-
prudential learning never quite managed.
The Carolingian kingdom was thus characterized by a vigorous
legal pluralism that was consistently at odds with the aspirations of
its Church towards uniformity. The attention given by Carolingians
themselves and by their predecessors in all their dominions to the
creation and preservation of a variety of legal traditions ensured this.
Most of these laws were at the outset, and considerable effort was
put into ensuring that they remained, repositories of ancient tradi-
tions, which it is not misleading to call “Germanic”. They are in
fact among the best evidence we have for the stolidity of a Germanic
culture. In this sense, a lex became a very distinct, even an indis-
pensable badge of post-Roman Europe’s emergent ethnicities. Yet
the main determinant of law as actually observed was the resilience
or otherwise of Roman practice. Where this made its influence most
felt, which is to say within the cultural climate of the Ancient Medi-
terranean, was also where we may expect to find law affected in
practice by law-making. Where Roman example was a vision of a
more distant and warmer South, rulers dreamt of a “civilization” of
written law, but had no real prospect of bringing one to fruition.
  LEGES BARBARORUM 47

APPENDIX: Tabulated injury tariffs in the “barbarian” leges

(in solidi/scillingas, and fragments thereof )
Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

blow 62½1 11 11 11 301 101 241

dumbness 182
arising
bloodshed 152 182 1½2 1½2 1203 102 14
hair 2½ 1
1206 453
headskin 61 33
brain 153 126 129 104 –5 187
– brain exposed 248
medical attention 407 64
skull 304 12 2
64 –5 32 1804 25 152
– breaks bone 363 301
– breaking bone 68 43 2405 303 126
so rings on
distant shield 68 452
– 2 bones 654
– 3 bones 755
– lesser bones 256
head or upper arm 65
swelling 67 602 ½3
blow on face 16 10
627

– accidental 1229 24
on beard 630
on nose 1611
on ear 1612
wrinkle 210
– lower 411
– third 212
punch 36 33 13
bloody punch 157
triple punch 98
side-wound/ 305 364 655 304
shoulder 306
neck pierced 628
neck tendons 10063
chest wound 2015
piercing breast 1228
– diaphragm 1829
– diaphragm 2430
pierced
lung exposed 4+60
– breath
escaping 861
48  

(table cont.)

Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

intestine 1256 1831

– intestine 2457 2432
pierced
– intestine 4058
mutilated
stomach wound 1225
bowel wound 1826
– if pierced 2427
abdomen 1228
– attacked 1829
with sword
– pierced 2457
guts and fat 4+30
spilt on top
rib 349 1222 1054
– dislodged bone 2055
belly 1241 3030
belly pierced 2042 2032
medical
treatment 3043
severe wound 3044
loin maimed 6048
– pierced 1549
– transfixed 3050
hand, foot, 1009 4010 66⅓56
eye, ear
nose
ear cut off 1525 1005 d½wer9 4010 2020 128 ?7209 1008 128 305
– no loss of 506 129 319 half 16
hearing
– loss of 4021 257 144010 10019 606
hearing
half of ear 611 622 610 quarter17
crippled 2011
ear pierced 319 39
eye, nose d
½wer3–4
eye struck 62½22 100 9
4015 5011 ?7207 1006 ½wer27 6 6 ⅓ 7
out
– laming of eye half 16
– still usable quarter17 20⅔26
– loss of vision 5010
2014
14408 228
(i.e. both eyes)
eye, mouth 1212
disfigured
  LEGES BARBARORUM 49

(table cont.)

Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

eyebrow 213
upper eyelid 612 323 214
lower eyelid 1213 625 214
nose cut off 4523 1007 4018 918 72011 1007 249 609
nose pierced 913
1515
unable to sneeze 1217 617 half 16
able to sneeze 508 616 quarter17
whiskers 216
cheek 315
617
– 2 cheeks 616
throat pierced 618 1215
tongue cut 10025 4025
1224 1009
as16
out eye
able to speak 2026
jaw 1513
chin 2019
1214
punch on nose 337
bruise 138
visible bruise 1½39
bruise beneath
clothes 140
wound visible
below hair,
sleeve or
trousers x247
lips 165
upper 619 324
lower 1220 623
tooth
knocked out 1526
teeth exposed 206
front teeth 167 620 218 810
upper front 621
canine? 322 1225 421 319 1512
other teeth 626 322
jaw teeth 88 123 123 420 411
lower front 1224
collar bone 625 421 2058
– lesser damage 1559
piercing arm 1613 626
– above elbow 631 616
– below elbow 332 317
– light bleeding 1½33
– heavy bleeding 334
50  

(table cont.)

Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

– unbroken skin 335

– below elbow 336
– above elbow 637
– damaged elbow 1238 1518
crippled arm 2039 68 627
2405
105

arm cut off 8043

at elbow 40 40

above elbow 1223

below elbow 624
both bones 1225 3019
at shoulder 8041 50⅓ 8046/51
57

striking arm 814

wrist blood 438
elbow blood 439
shoulder blood 440 3017
junction of 444
elbow/shoulder
arm lamed ½cut58
off
arm and hand as for59
thigh &
foot
loss of hand 10011 ½wer18 72012 10010 4545
– both hands 144012
lamed hand 62½10 5012 ½wer19 half 16 5011 4053
– still movable quarter17
hand wound 2052
thumb, big 5011 5013 1/6 wer201243 1212 2028 33⅓14 13⅓28 3020
toe cut off [thumb]
thumb nail 329 36018 521
thumb lamed 30 12
25 14 b
6 42
18019
upper thumb 241
joint
– lower 342
– junction 443
with arm
2nd finger 3513 3615 1621 1046 913 930 18024 33⅓15 729 1522
– 1st joint 2½44 ⅓
– 2nd joint 545 ⅔
– nail 323
3 other 45 14

fingers
2 fingers 3515
1 finger 3016
3rd finger 1517 522 649 514 431 12025 33⅓15 6½30 1224
– 1st joint 1½47 ⅓ 135
– 2nd joint 348 ⅔ 236
  LEGES BARBARORUM 51

(table cont.)

Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

– lowest 337
– nail 225
4th finger 918 823 852 514 632 12025 33⅓16 831 1726
– 1st joint 250 – 135
– 2nd joint 451 – 236
– lowest 337
– nail 427
little finger 1519 1624 as53 913 1133 24020 33⅓16 632 928
– one joint thumb 8021
– two joints 16022
– nail 139
all five 4133
finger nails 134
3rd finger 1254
disabled
militarily add
⅓15
palm 434
laming of ½ for56
hand, foot, cutting
finger off

disfigurement 335 2405 5018

worse injury 636
sinew 355 1260
– incurable 3061
small sinew 662
disabled 10027 2060 wer59
penis 645–6
one testicle 72014 10013 ½wer31
– both 144015 10012 wer30 8042
– testicle 6+32
restored
castration 20028 20019 4059 werx345
any limb 3618
hernia 369.
pierced hip 16 16
1261
striking hip 817
lamed knee 1262
leg pierced below knee 363
broken thigh 1247 2405 105 1233 3033
lamed thigh friends48
thigh pierced 650 3034
extent of stab 1s per
inch51–4
shin pierced 1235
52  

(table cont.)

Sala Rib Lombc Ale Baif Abt Saxg Thuri/j Frisk Af

– bone below knee 6–1234 3036

– cut off at knee 8057
lamed foot 4520 5017 ¼wer26
foot struck 62½21 10016 ½wer25 4066 5056 72013 10010 4535
off
– lamed half 16 5011
– still movable quarter17
– both feet 144013
leg struck off at knee 5067
leg struck off at hip 8068
big toe 1627 664 1057 ½thumb27 3017 86 2037
other toes ½finger ½finger28
58

2nd toe 628 365 747 1538

3rd toe 329 365 648 939
4th toe 330 365 549 640
5th toe 231 365 ½other 450 541
big toenail 1½59 toes29 h
other toenails ½59
rest of foot 1551
blood spilt as for52
fingers
Sources
Pactus legis 17,1–10; 29,1–18, ed. K.A. Eckhardt, MGH LL nationum Germanicarum
4,1 (Hannover 1962); Lex Ribuaria 1–6, ed. F. Beyerle and R. Buchner, MGH LL
nationum Germanicarum 3,2 (Hannover 1954); Edictus Rothari 45–73, ed. F. Bluhme,
MGH LL 4 (Hannover 1868); Leges Alamannorum 57,1–68 (see note d) below), ed.
K. Lehmann and K.A. Eckhardt, MGH LL nationum Germanicarum 5,1 (rev. 2nd
edn., Hannover 1966); Lex Baiwariorum 4,1–16, ed. E. von Schwind, MGH LL
nationum Germanicarum 5,2 (Hannover 1926); Laws of Æthelberht 33–72,1, ed. F. Lie-
bermann, Die Gesetze der Angelsachsen 1 (Halle 1903); Lex Saxonum 1–15, ed. K.F. von
Richthofen, MGH LL 5 (Hannover 1875–89); Lex Thuringorum 1–25, ibid.; Lex Frisionum
22,1–89 (and NB also Lex Frisionum (Additio Sapientum) 2–3 etc.), ed. K.F. von
Richthofen, MGH LL 3 (Hannover 1863); Laws of Ælfred 44–77, ed. F. Liebermann,
Die Gesetze der Angelsachsen 1 (Halle 1903). The Burgundian code has no table as
such (though cf., e.g. Liber Constitutionum 5; 11; 26; 48; 93, ed. L.R. von Salis, MGH
LL nationum Germanicarum 2,1 [Hannover 1892]); while Chindaswinth’s impres-
sive laws (Lex Visigothorum 6,4,1; 6,4,3, ed. K. Zeumer, MGH LL nationum Germani-
carum 1 [Hannover 1902]) are so much more elaborate as to frustrate comparison
with those tabulated here.
The tables are designed to show that the “barbarian” leges are all trying to do
the same sort of thing about personal injury, though in almost invariably distinc-
tive ways: that is to say, they vary in the injuries they identify (and their extent);
in the order in which they are listed; and in the seriousness—reflected in the com-
pensation specified—with which any one injury is viewed. Inevitably, given the
“multi-layered” structure of many of these lists (especially, e.g., Lex Sal.), thorough
consistency is not to be expected (e.g. see note a) below). I have as a rule omitted
bindings, blows struck on others’ weapons, or throwing into river (e.g. Lex Sax. 6;
  LEGES BARBARORUM 53

8–10), since these do not appear regularly enough to sustain comparison. I have
tried (so far as feasible) to follow the head-to-toe order essayed (if scarcely consis-
tently exercised) by all codes; the superscripted numbers following each compensa-
tion shows the place it occupies in any one code’s sequence (not the clause number).l/m

Other notes
a) Lex Sal. details are taken very largely from the oldest (“A”) group of MSS,
though with some details—signalled in Italics—from the “C” (later Merovingian)
tradition.
b) At this point, Lex Rib. enters the helpful generalization that if a limb ‘hangs
maimed’ composition is half what it would be for excising it altogether; cf. Lex
Fris. 58.
c) The Lombard compositions listed here are those of homines liberi as opposed
to aldii or servi.
d) The Lombard wergeld is not actually made clear until Leg. Liutpr. 62 (724),
where it seems to be established that a minima persona [. . .] exercitalis has 150s (¾
the Frankish ingenuus, cf. above, pp. 31–2), while the primus has 300s (50% more
than Frank, or half the Frankish antrustio). Perhaps the relevant point is that pay-
ment of half a wergeld for loss of eye, nose, hand or foot is exactly the same prin-
ciple in Frankish as in Lombard custom.
e) There are two sets of Alamannic compensations, those of the Pactus legis Salicae
being almost certainly the earlier, i.e. early seventh-century. But complexities in the
transmission mean that it makes more sense to list/analyse the tariff of the eighth-
century revised “ducal” code.
f ) Bavarians, like Lombards, distinguish compensations for freemen, freedmen
and servi; only those of liberi are analysed here.
g) Saxon compositions are given only for nobiles, apart from single clauses on liti
and servi; the former are what is listed here, which is of course why the figures are
so much higher than in other tables.
h) Lex Sax. here makes the interesting and otherwise unique stipulation that all
these compositions are to be doubled for women if virgins, though paid at the same
rate if already married.
i/j) Lex Thur. gives figures for both nobles and freemen; the latter, one-third of
the former, are those tabulated here.
k) Lex Fris., not content with providing by far the longest and most elaborate
tariff, proceeds to expand and modify this with a series of Additiones Sapientium; these
I have omitted, in order not to complicate and lengthen the table beyond tolera-
bility; indeed, I have omitted some of the original tariff ’s complexities, to which
there is no counterpart in others.
l/m) Oliver, “Language of Early English Laws”, pp. 247–9, offers instructive com-
ment on such tabling of injuries compared to other systems.
This page intentionally left blank
 GENS INTO REGNUM: THE VANDALS1

J.H.W.G. Liebeschuetz

I. Problems and sources

The theme of the ESF project has been the transformation of the
gentes into regna. All the peoples that settled and established kingdoms
within the empire were described in Latin as gentes. The meaning of
gens is however ambiguous. That emerges clearly from many of the
studies. Gens is generally translated as “tribe”, and a tribe is a large
association held together by a sense of obligation to a group which
is defined by common descent of its members. But this is where
complications begin. For the common descent of a tribe may be
entirely biological, in that the members of the group are demon-
strably derived from common ancestors, but far more often than not
it is almost entirely mythical, as for instance the descent of all the
biblical Israelites from Abraham, Isaac and Jacob.2 The implied kin-
ship is in fact simply a metaphor, even a very powerful metaphor,
prescribing how members of a group ought to behave towards each
other, namely as kinsmen to kinsmen. Between these extremes there
is a whole spectrum of conditions depending on the history and cus-
toms, and especially the degree of openness of a particular tribal
group. It must be remembered that until quite recently common
descent was not as strictly defined in terms of genetic transmission
as it has been since Darwin and Mendel. The Romans for instance,
for all practical purposes considered a son by adoption fully equal
to a biological son.
Even in tribal societies, which are ostensibly organised on a basis
of kinship groups, there is considerable flexibility with regard to the

1
I want to thank Mike Clover for helpfully answering questions of mine, and
Philipp von Rummel for allowing me to use his valuable Magister Thesis.
2
M. Weber, Grundriss der Sozialökonomik, Part 3: Wirtschaft und Gesellschaft (2nd edn.,
Tübingen 1925) p. 223: “Dieser Sachverhalt: daß das ‘Stammesbewußtsein’ der
Regel nach primär durch politisch gemeinsame Schicksale und nicht primär durch
‘Abstammung’ bedingt ist”.
56 .... 

admission of outsiders.3 The famous Berber historian and political

theorist Ibn Khaldûn (1332–1406) tells us that among Arabs mem-
bership of a tribe was something which one could acquire as well
be born into. “It is clear that a person of a certain descent may
become attached to a people of another descent, either because he
feels to be well disposed towards them, or because there exists an
old alliance or client relationship, or yet because he has to flee from
his own people because of a crime that he has committed. Such a
person comes to be known as having the same descent as those to
whom he has attached himself, and is counted as one of them with
respect to the things that result from common descent, such as
affection, the rights and obligations concerning talion and blood
money, and so on.” The same is true among Berbers today.4 In
practice the degree of openness in the case of a particular group
depends on political considerations, the priority given to monopolis-
ing material or other advantages over other priorities, for instance
that of maximising military manpower.
The tribal sense solidarity too is something problematical. It is
often assumed that it is adequately summarised as a sense of iden-
tity. But this is surely to define the whole in terms of only a part,
because the active component of this sentiment is not what the indi-
vidual tribesman feels about himself, but what he feels about a com-
plex of obligations, duties and sentiments shared with other members
of the group. What makes tribal feeling dynamic is that the indi-
vidual tribesman feels that his well-being is bound up with the suc-
cess of the group, and that he must therefore loyally subordinate his
own interests to those of the collective. But it is important to realise
that tribal feeling is not the same in every tribal grouping, but some-
thing whose strength, range and direction differs strongly from group
to group, and changes very much with the circumstances in which
a particular group finds itself.5

3
Ibid., p. 217: “Nicht nur die Tatsache, daß, sondern auch der Grad, in welchem
das reale Blutsband als solches beachtet wird, ist durch andere Gründe als das Maß
der objektiven Rassenverwandtschaft mitbestimmt.”
4
Ibn Khaldûn, The Muqaddimah, an Introduction to History, transl. F. Rosenthal, 3
vols. (London 1958) p. 267. See also M. Brett and E. Fentress, The Berbers (Oxford
1996) p. 230: “all that defines any such tribe as an actual social unit is the belief
in a common ancestor, whatever the actual genetic reality may be”.
5
Weber, Grundriss der Sozialökonomik, p. 222: “Der Inhalt des auf ‘ethnischer’ Basis
möglichen Gemeinschaftshandelns bleibt unbestimmt. [. . .] Ganz regelmäßig wird,
   57

Ethnicity is essentially an internal psychological state and as such

it is difficult of access to the researcher. This is true even when we
are dealing with a people who have left a literature to tell us some-
thing about the priorities and emotions of its creators. It is much
more difficult when, as with the Germanic gentes who founded the
regna, they have left little or no such direct evidence. In cases like
this, the method must be that of behaviourism: the strength and
character of the ethnicity of a gens must be deduced from both its
behaviour in recorded situations, and from institutions such as dress,
language, shared traditions, law, and religion, which make tribal dis-
tinctiveness visible, and indeed help to create and nourish the under-
lying emotions.6 But this approach too is not straightforward. For
the institutions were fluid. Without exception the gentes borrowed
heavily from the institutions of the Roman world in which they were
establishing their kingdoms. But the fact that the institutions which
expressed ethnicity were changeable, and that they had in any case
often been appropriated from the Romans, does not mean that tribal
solidarity which they expressed and upheld was necessarily weak and
insignificant.7 It is likely that, then as now, tribal solidarity was at
some times, and in some circumstances, extremely powerful, whether
for creation or destruction. After all the breaking up of the century
old imperial structure which these gentes brought about involved a
massive change.
The occupation of North Africa by the Vandals after 429 A.D. was
a decisive turning point in the disintegration of the Western Empire.

wenn diese Ausdrücke [Völkerschaft, Stamm, Volk] gebraucht werden, entweder

eine [. . .] gegenwärtige politische Gemeinschaft oder Erinnerungen an eine früher
einmal gewesene [. . .] oder Sprach- bzw. Dialektgemeinschaften oder endlich eine
Kultgemeinschaft, mit hinzugedacht.” In other words the content of “ethnicity” is
provided by the situation and history of the particular ethnos.
6
Ibid., p. 219: “Fast jede Art von Gemeinsamkeit und Gegensätzlichkeit des
Habitus und der Gepflogenheiten kann Anlaß zu dem [. . .] Glauben werden, daß
zwischen den sich anziehenden oder abstoßenden Gruppen Stammverwandtschaft
oder Stammfremdheit bestehe”. Ibid., p. 217: “Die Frage [. . .] ob die als auffällig
abweichend und also scheidend empfundenen Differenzen auf ‘Anlage’ oder ‘Tradition’
beruhen, ist für ihre Wirksamkeit [. . .] nomalerweise gänzlich bedeutungslos”.
7
Ibid., p. 223: “Praktisch [. . .] pflegt die Existenz des ‘Stammesbewusstseins’
wiederum etwas spezifisch Politisches zu bedeuten: daß nämlich bei einer kriegerischen
Bedrohung von außen oder bei genügendem Anreiz zu eigener kriegerischer Aktivität
nach außen, ein politisches Gemeinschaftshandeln besonders leicht auf dieser Grundlage
[. . .] entsteht.”
58 .... 

The organisation and character of the Vandal group which had such
a powerful impact on history nevertheless remain very obscure: unlike
Visigoths, Ostrogoths and Lombards, the Vandals have left practi-
cally no literary records of their own. The fact that we have no
Vandal laws makes it almost impossible to assess what changes in
the “constitution” of the gens were linked with the establishment of
the Vandal kingdom. The major accounts we have were written by
their enemies and are all—though in different ways—extremely one-
sided and incomplete.8 Victor of Vita’s History of the Vandal Persecutions
is as selective and one-sided as all histories of martyrdom. This means
that it is focused on the pitiful sufferings of martyrs and without any
attempt to give an account of the state of the Church as a whole,
much less to show the wider social and political background to the
persecution. The Life of Fulgentius of Ruspe is written to display its sub-
ject’s saintliness and asceticism, not to provide a picture of the soci-
ety in which he moved. Procopius’ history of the destruction of the
Vandal state by the Byzantine army led by Belisarius is the account
of an eye witness, but like most ancient historians of war, Procopius
was content to limit himself to military operations. Facts of the social
organisation of the territory in which the campaign took place may
emerge, but only incidentally. The account of the early history of
the Vandals, which Procopius wrote as an introduction to the cam-
paign-history proper, is scrappy and careless. This does not however
mean that everything in it must be discarded. Information like the
statement that there were 80 officials with the title millenarius will still
have been true when Procopius came to Africa with Belisarius’ force.

II. Did the gens exist before the entry of the Vandals into the empire?

Vandal (like Goth, Alaman and Frank) appears to be a collective

name shared by a grouping of sub-tribes. The components sub-tribes
that constituted the Vandal group may well have changed over time.

8
C. Courtois, Victor de Vita et son oeuvre (Algiers 1954); H.I. Marrou, “Le valeur
historique de Victor de Vita”, Les Cahiers de Tunisie 15 (1967) pp. 205–8; Y. Modéran,
“La chronologie de la vie de Saint Fulgence de Ruspe”, Mélanges de l’École Française
de Rome 105,1 (1993) pp. 135–88. Although mainly on a later period see also Averil
Cameron, “Byzantine Africa—the literary evidence”, Excavations at Carthage 1978 con-
ducted by the University of Michigan, vol. 7, ed. J.H. Humphrey (Ann Arbor 1982) pp.
29–62.
   59

Pliny (c. A.D. 23–79) tells us that the Vandals were one of the prin-
cipal subdivisions of the “Germans”, and that the Vandals were in
turn subdivided, Goths and Burgundians and others being their sub-
tribes.9 For Tacitus (c. A.D. 56–118) too, the Vandals were one of
the principal divisions of the “Germans”, though he does not list
their subdivisions.10 The origin of the relationship between the var-
ious collective gentes and their sub-tribes, and the precise nature of
the ties that gave the relationship meaning, are still fairly obscure.
They certainly need not have originated in the same way for each
tribal group.11
The difficult question of the relationship between gens and sub-
tribe is a problem of the same kind as that of the relationship of
the gentes to the “Germans” as a whole. According to Tacitus cer-
tain collective tribal names, e.g. those of the Vandili (surely our
Vandals?) and the Suebi are ancient. But the designation Germani
was originally only attached to a single tribe (the Tungri as he
thought) who had settled in northern Gaul. From being a descrip-
tion of a particular tribe it came into wider use to describe all gentes
who were known as Germani in Tacitus’ time, until finally the Germani
adopted the name for themselves.12 So the concept of a Germanic
people originated not with the “Germans”, but with their neigh-
bours, the Gauls and the Romans. It is a Fremdbezeichnung. But this
does not mean that it was without any basis in reality. Taci-
tus clearly thought that he was concerned with a distinct people.
He did his best to describe their common customs. He shows him-
self aware of their common language.13 He knows that there was

9
Pliny, Natural History 4,99, ed. and transl. H.R. Rackham (London-Cambridge
1949–68).
10
Tacitus, Germania 2, ed. R. Much and W. Lange with H. Jankuhn (3rd edn.,
Heidelberg 1967) [henceforth: Tacitus, Germ.].
11
One possibility: K. Strobel, Die Galater. Geschichte und Eigenart der keltischen
Staatsbildung auf dem Balkan des hellenistischen Kleinasien 1 (Berlin 1996) p. 131: “Die
Selbstbezeichnung des Kriegerverbandes hat die grundsätzliche Tendenz, die Grenzen
der unmittelbaren Stammeszugehörigkeit zu überschreiten, und als Namen der han-
delnden Kollektive vor die Benennung nach der jeweiligen Stammes- oder Abstam-
mungsidentitität zu treten. Die Benennung des Heerbannes wird durch die Übernahme
von Seiten der mit ihnen konfrontierten Gruppen zu einer Fremdbezeichnung, die
nun auf das Volk oder die Völkergruppen selbst übertragen werden kann”.
12
Tacitus, Germ. 2.
13
Ibid. 43: Gallic and Pannonian language disqualify respectively Cotini and Osi
60 .... 

a difference between the language of the “Germans” and that of

the Britons (i.e. Celtic).14 For him “Germanicity” was certainly more
than a matter of geographic location. For he can point out that
certain peoples living to the right of the Rhine, among the so-called
“Germans”, are really Gauls,15 and the Nervii and Treveri who lived
among the Gauls were, or at least themselves claimed to be, of
Germanic descent.16 Moreover Tacitus thought that some peoples
occupying the left bank, that is the Gallic bank of the Rhine, were
certainly Germani.17 So he did distinguish between “Germans” and
Gauls and others whom he thought non-“Germans”. “Germanicity”
was something real and recognisable. Moreover I think it can be
shown, though there is not the space to do so here, that other
Romans, who referred to “Germans” in the last decades of the
Republic and after, similarly visualized a distinct people. There
remains the question what the so-called Germanic peoples thought
about their “Germanicity”. In my opinion it is taking scepticism too
far to reject Tacitus’ express statement that they applied this name
to themselves. But once a collective name for these gentes existed, it
is likely to have made them much more conscious of what they had
in common.18
But if a certain sense of relationship, and some genuine, if rather
vague and not easily assessed emotional bonds linked the members
of the various Germanic gentes, these sentiments do not appear to
have been very strong. After all the concept of “Germanicity” did
not survive the Roman Empire. Even in Britain the word “German”
had to be rediscovered: it is not documented before the sixteenth

from being considered “Germans”. However in Germ. 28 the Germanicity of the

Osi is asserted. Perhaps there germanorum natione is purely geographical. Germ. 46
Bastarnae are Germanic in language as well as other respects.
14
Ibid. 45 (Aestii).
15
Ibid. 28 (Helvetii, Boii) and 43 (Cotini and Osi).
16
Ibid. 28.
17
Ibid. (Vangiones, Triboci, Nemetes).
18
Just as deutsch started as a Fremdbezeichnung. See J. Jarnut, “Teotischis homines
(a. 816). Studien und Reflexionen über den ältesten (urkundlichen) Beleg des Begriffs
‘theodiscus’”. Mitteilungen des Instituts für Österreichische Geschichtsforschung 104 (1996) pp.
26–40. At the oath of Strasbourg, February 843, the languages of the two Frankish
kingdoms were respectively lingua romana and lingua teudisca (Nithard, Historiarum libri
IIII 3,5, ed. E. Müller, MGH SSrG 44 [Hannover 1907] pp. 35–7).
   61

century.19 Indeed one wonders whether the various gentes whom the
Gauls and Romans described as “Germans” even had a word of
their own to describe what they had in common, before they appro-
priated the description used by their neighbours. The likelihood is
that they did not. Clearly in the Age of Migrations the active alle-
giance of individuals was to the gens, and especially to the sub-tribe,
that is in the case of a Vandal to the Silingi or Hastingi, in the case
of the Visigoths to the Tervingi or Greutungi.20 But if the strength
of the emotional bonding represented by shared “Germanicity” is
very uncertain, the cultural kinship of the bulk of the people involved
must have assisted the integration of such tribal federations as that
of the band which was to establish the Vandal regnum in Africa. In
short, it must have greatly facilitated ethnogenesis.
As for the Vandals, the fact that they were mentioned by Tacitus
(c. A.D. 56–118) is evidence that they are one of the older tribal
collectives. As such they are certainly considerably older than the
Franks and the Alamanni. The earliest occasion that the Vandals
are reported to have taken an active part in history was in the reign
of Marcus Aurelius (A.D. 161–180). At that time the name was
attached to two sub-tribes. One of the two, the Silingi, lived in what
were then known as the Vandal mountains,21 situated in what used
to be known as Sudetenland, in what is now the Czech Republic.22
The name of the former German province of Silesia possibly pre-
serves their memory. The second group the (H)astingi arrived at
around the same time on the western border of Dacia, that is around
the border of modern Hungary and Romania.23 Hastingi was the
name of the royal dynasty of the Vandals when they entered
the Empire.24 The earliest known ruler of this line was Visimar, a

19
The Compact Edition of the Oxford English Dictionary (Oxford 1971) vol. 1, p. 1134,
s.v. German. Previously the expression was Almain, or Dutch.
20
A. Lund, Die ersten Germanen: Ethnizität und Ethnogenese (Heidelberg 1998), and
the pioneer E. Norden, Die germanische Urgeschichte in Tacitus’ Germania (Leipzig-Berlin
1920), especially chapters five and six.
21
Cassius Dio, Roman History 55,1, ed. and transl. E. Cary (London-Cambridge
Mass. 1969) [henceforth: Dio]: the Elbe rises in “Vandal mountains”.
22
Ptolemy (A.D. 146–170), places them between the Sudeten Mountains and the
Oder; for their early history see L. Schmidt, Geschichte der Vandalen (2nd edn., Munich
1942) pp. 2–17; C. Courtois, Les Vandales et l’Afrique (Paris 1955; repr. Aalen 1964)
pp. 11–37.
23
Dio 72,12,1 (A.D. 172).
24
Jordanes, Getica 113, ed T. Mommsen, MGH AA 5,1 (Berlin 1882): Visimar
62 .... 

contemporary of Constantine.25 So the eastern component of the

Vandal group took their name from its dynasty. It is a likely hypoth-
esis that the subgroup originated as a war band that had split off
from the Silingi at the time of the Marcomannic wars, under the
leadership of this family.26
Subsequently Silingi and Hastingi appear to have operated inde-
pendently, though both described themselves, or at least were described
by others, as Vandals.27 We do not know what institutions, customs
and memories underlay this shared identity. But the problem of what
precisely it was that induced groups living in separate territories,
each having its own leadership, to think of themselves in some impor-
tant respect as one, is raised by the early history of the Franks,
Alamanni, Goths and Saxons no less than that of the Vandals. The
inhabitants of the numerous, constantly warring, cities of Greece
shared the belief that they were all Hellenes. Among Greeks, as
among the Germanic Sueves,28 the sense of shared identity was main-
tained by shared cult.29 Something of that kind might well have pre-
served the sense of shared “Vandalicity”. But we do not know. The
fact however that we do not have the information, does not mean
that we can assume that this Traditionskern did not exist, or that it
was insignificant. It may be that it played a significant, even an
essential, role in the “crystallisation” of the larger group in the course
of the Vandals’ progress through the Empire. But we simply lack
the evidence to judge.

[. . .] Asdingorum stirpe; Cassiodorus, Variae 1,9, ed. and transl. S.J.B. Barnish, Translated
Texts for Historians 12 (Liverpool 1992); cf. Courtois, Les Vandales et l’Afrique, pp.
391–404.
25
Jordanes, Getica 22,123.
26
According to Dio 71 (72),12,1 the Asdingi were led by Raus and Raptus, pre-
sumably the ancestors of the royal dynasty.
27
Deuxippus, Fragmenta 24, ed. C. Müller, Fragmenta Historicorum Graecorum,
vol. 3 (Paris 1849; repr. 1928) pp. 685–6) describes the tribe that invaded Greece
in A.D. 270 as Vandals—they clearly were the eastern group.
28
Tacitus, Germ. 39.
29
Modern scholars have identified the Silingi with the Naharvali, who accord-
ing to Tacitus, Germ. 43, controlled the sanctuary of the Lugian federation. But this
is mere speculation, based on the fact that the Naharvali appear to have lived in
roughly the same geographical position as the Silingi, and the fact that Tacitus
mentions the Naharvali, but not the Silingi.
   63

III. What sorts of changes and conditions led to the

establishment of the kingdom?

Each of the Vandal groups from time to time came into military
contact with the Roman Empire, the Hastingi more often than the
Silingi, as the latter lived at some distance from the imperial fron-
tier. But for around two hundred years neither group presented any
serious threat to the Empire—in this way they were quite unlike the
Goths. This changed quite suddenly. In the late 390s the Vandals
began to move.30 In 401 Stilicho expelled Vandals from Raetia,
before returning to Italy to deal with the first invasion of Alaric’s
Goths. Stilicho’s expedition into Raetia was not a major campaign,
though evidently a necessary one.31 Clearly the Vandals and their
federates were not yet the formidable band that they soon became.
Then on the last day of 40632 the Empire suffered one of its worst
ever disasters, when a huge band of barbarians made up both groups
of Vandals, as well as of Alans, Sueves,33 and even some Pannonians
from within the Empire,34 crossed the frozen Rhine and broke into
Gaul.35 The background of the invasion is obscure, but what made
it so formidable was an alliance between the two groups of Vandals
and the Alans.36 The attack was perhaps planned to exploit the

30
According to Procopius, Wars 3,1,22,1–5, ed. J. Haury, transl. H.B. Dewing,
Loeb Classical Library (London-Cambridge Mass. 1914–28) [henceforth. Procopius]
hunger was the motive. He relates that many years later the Vandals left behind
in their original home sent an embassy to their successful kinsmen in Africa ask-
ing them to renounce their property rights at home. Geiseric was persuaded to
reject the request on grounds of the changeability of human fortune. A good story,
but doubtful history. Procopius knew of no Vandals left in their native land in his
time.
31
Claudian, De bello Pollentino sive Gothico 363–6; 414–5, ed. T. Birt, MGH AA
10 (Berlin 1892; repr. 1962) pp. 259–83, here pp. 273–4.
32
The date: Continuatio Havniensis Prosperi a. 406, ed. T. Mommsen, MGH AA 9
(Berlin 1892; repr. Munich 1981) p. 299; Prosper of Aquitaine, Epitoma Chronicon
1229–1230, ed. T. Mommsen, MGH AA 9 (Berlin 1892; repr. Munich 1981) p. 465.
33
Isidore of Seville, Historia Gothorum, Wandalorum, Sueborum 71, ed. T. Mommsen,
MGH AA 11 (Berlin 1894); Orosius, Historiarum adversum paganos libri VII 7,38,3;
7,40,3, ed. K. Zangemeister, CSEL 5 (Vienna 1882; repr. Hildesheim 1967) [hence-
forth: Orosius]; Zosimus, Histoire Nouvelle 6,3,1, ed. and transl. F. Paschoud (Paris
1971–89).
34
Jerome, Lettres 123,16, ed. J. Labourt (Paris 1961).
35
By the time they came to Africa, and besieged Hippo, they are said also to
have included Goths (Possidius, Vita Augustini 30, PL 32).
36
The Alans too were made up of two groups, respectively ruled by the kings
64 .... 

difficulties created for the Empire by Radagaisus’ invasion of Italy.

In fact Radagaisus had already been defeated in August 406, but
the Vandals and their allies could not have anticipated that. The
invasion certainly required careful preparation and planning. The
organising and diplomatic brain behind it may well have been king
Godigisel of the Hastingi Vandals, who did not however live to see
the success of his scheme, because he was killed in a battle against
the Franks before the breaking of the Rhine frontier of the empire.37
The Pannonian provincials could have been drawn in to the allied
band by a group of Vandals who had been settled in Pannonia since
335.38 In time the Pannonians were no doubt absorbed into the
Vandals.
We have no figures at all, but the number of invaders must have
been large.39 This is a necessary deduction from their record. They
recovered from a defeat by the Franks, suffered before they had even
entered the Empire. They devastated the north-western frontier
provinces of Gaul,40 and even besieged and captured cities,41 in spite
of the fact that there must still have been a significant number of
troops, if only limitanei, or other garrison troops, in the area. After
that, they cut a swathe of devastation through Gaul as far as the
Pyrenees, and finally spread into the provinces of south-eastern Gaul
without the Roman troops being able to do anything about it.42 Each
tribal section must have been numerically strong, for each group was
in 411 given separate, and very extensive, settlement areas in Spain.43

Goar ( J. Martindale, The Prosopography of the Later Roman Empire, vol. 2 [Cambridge
1980] pp. 514–5) and Respendial (ibid., p. 940). The former went over to the
Romans (Gregory of Tours, Historiae 2,9, ed. B. Krusch and W. Levison, MGH
SSrM 1,1 [2nd edn., Hannover 1951] pp. 55–6).
37
Gregory of Tours, Historiae 2,9; cf. Martindale, The Prosopography of the Later
Roman Empire 2, pp. 515–6, s.v. Godigisel 1.
38
Jordanes, Getica 113.
39
For the following I owe much to J.F. Drinkwater, “The usurpers Constantine
III (407–11) and Jovinus (411–13)”, Britannia 29 (1998) pp. 269–98, though in my
opinion the invading force must have been very considerable. On the other hand
I am sure that “rather less than 200,000”, the estimates of E.A. Thompson, Romans
and Barbarians: the Decline of the Western Empire (Madison 1982) pp. 158–9, is several
times too high.
40
Jerome, Lettres 123,16; Salvian, De Gubernatione Dei Libri VIII 7,12(50), ed. F. Pauly,
CSEL 8 (Vienna 1883).
41
Jerome, Lettres 123,16; Salvian, De Gubernatione Dei 6,8 (39).
42
Orosius 7,40.
43
The Chronicle of Hydatius and the Consularia Constantinopolitana 297,17 a. 411 A.D.,
ed. R.W. Burgess (Oxford 1993) [henceforth: Hydatius, Ol.]. The settlement would
   65

Finally when the Sueves had been settled in Gallaecia,44 and the
Silingi were destroyed by the Visigoths, the Hastingi on their own,
together with the battle-reduced Alans, were still extremely formi-
dable, so that the Roman army in North Africa was not a match
for them. All this suggests a large invasion force.
The attack across the Rhine achieved complete surprise, and once
they had broken into the Empire the barbarians seem to have been
extremely mobile. Neither achievement would seem compatible with
an operation that involved the migration of three entire peoples,
including women, children and the old. It is likely therefore that the
invading force did not consist of three whole peoples on the move,
but rather of contingents, and presumably large contingents of cho-
sen, or volunteering, warriors, each war-band being led by its tribal
king. In the course of the campaign the band will have grown con-
siderably. We are told that the federate Honoriaci, who had been
supposed to guard the passes of the Pyrenees, joined them and re-
ceived land allotments, presumably at the same time as the invaders.45
It is possible that the Vandals and their allies were joined by sur-
vivors of Radagaisus’ band.46 Merging of warrior bands must have
been easier when both parties spoke Germanic dialects. But the cohe-
sion and discipline necessary for a long and strenuous campaign must
surely have been built around the ethnic solidarity of the original
ethnic units, and above all of the Vandals.
On the last day of 406, when the tribal band invaded Roman
Gaul, it was still a confederation of different gentes. No fusion had
taken place. Subsequent events showed this clearly. In 411 the gen-
eral Gerontius, acting in the name of the usurper Maximus, who
was briefly in control of Spain, reached an agreement with the
invaders assigning each group a different region of the country.47

seem to have been made with the usurper Maximus: See W. Lütkenhaus, Constantius
III., Studien zu seiner Tätigkeit und Stellung im Westreich 411–21 (Bonn 1998) p. 45.
44
Thompson, Romans and Barbarians, pp. 161–87.
45
Orosius 8,40.
46
This was deduced from Zosimus, Histoire Nouvelle 6,3 and Chronicorum a. CCC-
CLII 52, ed. T. Mommsen, MGH AA 9 (Berlin 1892) by F. Paschoud, Zosimus,
Histoire nouvelle, vol. 3,2, pp. 22–3, and accepted by Maria Cesa, Impero tardoantico e
barbari (Como 1994) pp. 124.
47
I am following J.F. Matthews, Western Aristocracies and Imperial Court, A.D. 364–425
(Oxford 1975) p. 311 and Drinkwater, “The usurpers Constantine III (407–11) and
Jovinus (411–13)”, pp. 287–8 against Thompson, Romans and Barbarians, pp. 155–6,
who interprets Hydatius strictly, to conclude that the barbarians drew lots for
66 .... 

The Hastingi received eastern Gallaecia for settlement, the Sueves

the lands of western Gallaecia bordering on the ocean, the Alans
Lusitania and Carthaginiensis, and the Silingi Baetica.48 Each gens
was ruled by a king, the Hastingi by Gunderic, the Alans by Addace,
the Sueves by Hermeric. The name of the king of the Silingi is not
known.49 It would appear that both Silingi and Sueves were in some
sense also subjects of the king of the Alans.50 To judge by the qual-
ity of the land assigned to them, the Hastingi, who were to be mas-
ters of Africa, were at this time actually the weakest of the barbarian
groups.51
Then in 417/18 the Visigoths under their king Vallia, and oper-
ating as allies of the Empire, destroyed the Silingi, and killed the
king of the Alans and greatly weakened his people. The Alani there-
upon placed themselves under the patronage of the king of the
Hastingi.52 The Alans long retained their tribal identity, so that the
title of the later kings of Vandal North Africa was “king of the Alans
and Vandals”. Nevertheless the events of 418 meant that the Alans
ceased to have any independent political or military existence. In a
sense they mark the beginning of an Alano-Vandal people. Sub-
sequently “Vandal” is used in two senses. Strictly speaking it refers
only to the Hastingi contingent, but it is also applied, and perhaps
by provincials and outsiders particularly, to the whole of the war-
band, Vandals, Alans and anybody else, who had been received into
their fellowship. The group did not straight away become a power-
ful military force. In 420 the Vandals, presumably now including
the surviving Alans, were expelled from Gallaecia by a Roman army,53
and after they had retired to Baetica, they were only saved from
further defeat in 422 when the Roman commander was deserted by

different regions of Spain without participation of the Roman authorities. On this

settlement see also J. Arce (this volume).
48
Hydatius, Ol. 297,17 a. 411 A.D., p. 86.
49
See Martindale, The Prosopography of the Later Roman Empire 2, s.v. Gundericus,
Addac, Hermericus.
50
The Alans saved the Vandals in the course of 406, when they had lost a bat-
tle against the Franks, and their king Godesil had been killed: Gregory of Tours,
Historiae 2,9. When they crossed the Rhine Gunderic was king: Continuatio Havniensis
Prosperi a. 406.
51
Unless, as argued by Thompson, Romans and Barbarians, p. 156, land assignment
was simply due to the luck of the draw.
52
Hydatius, Ol. 299,24 a. 418 A.D., p. 86.
53
Ibid. 299,26 a. 420 A.D., p. 86.
   67

his Gothic federate troops. This incident marked the turning point
in their fortune.54
Their recovery was remarkable. They acquired ships and raided
the Balearic Islands.55 In 428 king Gunderic, died and was succeeded
by his brother Geiseric, as the events showed a ruler of genius. Then
in 429 Geiseric took the fateful decision of shipping the whole of
his people Vandals and Alans and some other “barbarians”56 and
even some hopeful Roman provincials,57 with their families over to
Africa. According to Victor of Vita, Geiseric had his people counted
after the crossing, with the result that they were found to number
80,000 men, women and children. He insists on the comprehensive
character of the figures because the number of Vandal fighting men
at the time of writing—around 490 A.D.—was very much less, “small
and feeble”, in fact.58 The narrative of the Vandals’ ultimate defeat
confirms this estimate, since Belisarius succeeded in defeating them
with an army of only around 16,000.59
A very approximate estimate of the strength of the Vandals at the
time of their conquest of North Africa can be derived from the size
of the Roman army in Africa which they defeated. According to the
Notitia Dignitatum the garrison of the African provinces was composed
of ten units of limitanei and thirty six field army units. However thirty
of the latter appear to have been upgraded limitanei.60 If we take
the average unit-strength to have been five hundred, the imperial
army in Africa will have amounted to 23,000. The total that could
actually have been assembled to face the invaders will have been

54
Ibid. 300,28 a. 422 A.D., pp. 86–8.
55
Ibid. 301 a. 425 A.D., p. 88.
56
Some Sueves: Année Epigraphique 1991, no. 267 = Courtois, Les Vandales et l’Afrique,
Appendix II, no. 70; Mike Clover informs me that some more inscriptions of the
Sueves have turned up in North Africa. Goths: Possidius, Vita Augustini 28. Victor
of Vita, Historia Persecutionis Africanae Provinciae 1,19–21, ed. M. Petschenig, CSEL 7
(Bonn 1881) [henceforth: Victor of Vita] shows how the comes Sebastian became a
close adviser of Geiseric, refused to become an Arian, and was eventually executed.
57
For instance the four Spaniards who for a long time apud Gisiricum merito sapi-
entiae et fidelis obsequii cari, clarique habebantur (Prosper of Aquitaine, Epitoma Chronicon
1329 a. 437), until they were eventually executed for refusing to convert to Arianism.
58
Victor of Vita 1,2; in the face of this Procopius’ statement (3,18, part of his
unreliable historical introduction) that the 80,000 refer to fighting men, is to be
rejected.
59
Procopius 3,11,1.
60
A.H.M. Jones, The Later Roman Empire 284–602. A social, economic and administra-
tive survey (Oxford 1964) p. 197.
68 .... 

considerably less, say 10,000 to 15,000. The fighting strength of the

Vandals is likely to have been of the same order. It could well have
been smaller.
A plausible explanation of how the figure 80,000 arose can be
suggested. According to Procopius the Vandals were divided into
eighty units each commanded by a millenarius.61 Procopius knew that
the millenarii commanded military units.62 It is however likely, though
not certain, that the millenae were also divisions of the total popula-
tion, and that the mille in millena was a notional figure, like the hun-
dred of the centuria of the political assembly of the Roman Republic,
and that a millenarius was in fact the officer who was responsible for
the civil administration of his subsection of the Vandal people, as
well as being the commander of the tactical unit furnished by it on
active service.63 We are explicitly told that the Alans were included
in this registration, and so presumably were any other members of
Geiseric’s band who were not Vandals by descent.64 The bulk of the
local populations was excluded. Subsequently the population of Vandal
North Africa was made up of three peoples: Vandals, Romans and
Moors. That is how Procopius saw it. Nevertheless the description
“Vandals”, in the wider of its two senses still included some Alans
and others conscious of their non-Vandal origin. The king’s title
remained “king of the Vandals and Alans”.65

61
Procopius 3,5,18. Procopius was aware that the assumption of a fighting force
of 80 units of a thousand each resulted in an excessively high total, and quotes a
tradition that the total of Alans and Vandals at the time of the invasion, before
they had grown by natural increase and by associating other barbarians with them-
selves, had only amounted to 50,000.
62
Ibid. 4,3,8: each millenarius (Chiliarch) commands his unit. The fact that we
read about Vandal forces totalling round thousands (5,000 in 3,11,23; 2,000 in
3,18,1) does not necessarily mean that the basic tactical unit was 1,000 strong.
63
Victor of Vita 1,30,35: an anonymous millenarius, the only one known, owns
slaves, including a weapon-smith, and cattle. He looks like a great landowner.
64
E.g. the Sueve, Année Epigraphique 1951, no. 267 = Courtois, Les Vandales et
l’Afrique, p. 375 n. 70; and some Goths, Possidius, Vita Augustini 28.
65
Victor of Vita 2,39. The close association of these Alans and certain “other
barbarians” with the Vandals (in the narrow sense of the word) was something like
that of the Rugii with the Ostrogoths. At least Procopius uses apokrinesthai in each
case to describe the association of the client with the master people. See Procopius
3,5,22 and 7,2,2 following Thucydides 1,3). On the obstinate survival of submerged
tribal identities see P.J. Heather, “Disappearing and reappearing tribes”, Strategies of
Distinction. The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Rei-
mitz, The Transformation of the Roman World 2 (Leiden-New York-Köln 1998)
pp. 95–111.
   69

IV. What sorts of changes were linked to the establishment of the kingdom?

The Vandal conquest of proconsular Africa was followed by the

confiscation of land on a large scale, the exiling of many of the
landowners, and the granting of their estates to the Vandals. Lots
were assigned to the warriors, and large blocks of land to the king
and his sons, who needed a great deal of land in order to be able
to display generosity to their followers. “Byzacena, Abaritana and
Gaetulia and part of Numidia he kept for himself, Zeugitana and
the proconsular provinces he divided up as an allotted portion for
his people”.66 The confiscation was certainly not nearly as complete
as is suggested by our sources.67 There continued to be wealthy
Roman landowners.68 The location of the settlement areas is still
obscure. It would seem more likely that regions of Vandal settle-
ment were dispersed over the kingdom rather than consolidated into
one continuous block.69 But the Vandal areas were nevertheless dis-
tinct. For Catholic priests could be penalised for holding services in
the sortes Vandalorum.70 It also seems as if they lived on farms or in
villas rather than in cities.71 That would explain their policy of dis-
mantling the fortifications of cities, as well as the fact that when the
men were called up against the army of Belisarius, their families
accompanied them into the camp.72

66
Victor of Vita 1,13; Procopius 3,5,11–15.
67
The archaeology shows little trace of a widespread upheaval at the time of
the Vandal conquest, e.g. the production of fine pottery seems to have continued
with little change, M. Mackensen, Die spätantiken Sigillata- und Lampentöpfereien von El
Mahrine (Nordtunesien): Studien zur nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts
(Munich 1993) pp. 487–91.
68
Tablettes Albertini, actes privés de l’époque vandale, ed. C. Courtois, L. Leschi, C. Perrat
and C. Saumagne (Paris 1952).
69
G.G. König, “Wandalische Grabfunde des 5. und 6. Jahrhunderts”, Madrider
Mitteilungen 22 (1981) pp. 299–360; Courtois, Les Vandales et l’Afrique, pp. 218–9 with
map; P. von Rummel, Die beigabenführenden Gräber im vandalenzeitlichen Nordafrika, Magis-
terarbeit der Albert-Ludwig Universität zu Freiburg i. B. (Freiburg 2000), table 18.
In the opinion of von Rummel the grave evidence is insufficient to locate Vandal
settlement areas (pp. 132–3). See also C. Eger, “Vandalische Grabfunde aus Karthago”,
Germania 79,2 (2001) pp. 347–90.
70
Victor of Vita 3,4; cf. Ferrandus, Vita Fulgentii 6–7, PL 65. On taxes see chap-
ter VI below.
71
Procopius 4,6,9—which is highly moralising.
72
Ibid. 4,2,8.
70 .... 

After their settlement the “Vandals”, in the wider sense of the rul-
ing people, were a sharply distinct group. We lack the quantitative
information to assess how accessible to outsiders they were. There
is no evidence to suggest that there was a marriage bar between
Vandals and Romans, as there was between Romans and Visigoths.73
Certainly after the defeat of the Vandals their wives and daughters
were quick to marry Roman soldiers.74 It is moreover clear that in
the process of building up their formidable land and sea power the
Vandals absorbed numerous individuals from among stray barbar-
ians and Roman provincials.75 In addition some of the many Romans
who served at court or in the administration of the Vandal king-
dom are likely to have been admitted into the gens.76 But we cer-
tainly do not know that this happened. Again it is likely that the
Vandals became more exclusive in the course of their stay in Africa,
than they had been at the start. But once more we don’t know
whether this was so.
When we try to define the essence of what it meant to be a
Vandal we cannot get at the patriotic sentiments of the individual
member of the gens, but we can recognise some of the external signs
of their tribal solidarity.77 A Vandal could be defined by his enrol-
ment in a millena and the possession of a “lot”. The Vandals had a

73
See my “Citizen status and law”, Strategies of Distinction. The Construction of Ethnic
Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transformation of the
Roman World 2 (Leiden-New York-Köln 1998) pp. 131–52, especially pp. 139–41.
I am not persuaded by Hagith Sivan, “Roman-barbarian marriage in Visigothic
Gaul and Spain”, ibid., pp. 189–204. Endogamy may be, but is not by any means
always, an essential instrument of ethnicity. It is introduced for a particular pur-
pose, though the purpose is likely to be forgotten (Weber, Grundriss der Sozialökonomik,
p. 218: “Die ursprünglichen Motive der Entstehung von Verschiedenheiten der
Lebensgepflogenheiten werden vergessen, die Kontraste bestehen als ‘Konventionen’
weiter”).
74
Procopius 4,14,8–9. Courtois, Les Vandales et l’Afrique, p. 220 nn. 1 and 2 refers
to two possible cases of intermarriage (ibid., Appendix II, no. 153 and Victor of
Vita 3,24).
75
See above nn. 34, 45, 46, 52, 56, 57, 64, and below n. 125.
76
As I have argued for Alaric’s Goths in my Barbarians and Bishops (Oxford 1990)
pp. 75–8.
77
For a comprehensive discussion of “signs of distinction” see W. Pohl on “strate-
gies of distinction” and “telling the difference” in Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, pp. 1–69. But it is unhelpful to minimize the impor-
tance of these signs automatically and indiscriminately.
   71

Germanic dialect of their own, closely related to Gothic.78 It may

well have remained the principal language of Vandal religion,79 even
though Vandals increasingly found it more convenient to speak Latin.
If they had inscriptions on their tombstones, they were in Latin.80
In fact the Vandals could hardly have continued to make use of the
Roman administrative machinery in the way they did unless many
of them had fully mastered Latin. Nevertheless we read that Cyrila,
an Arian, that is a Vandal bishop, demonstratively insisted on speak-
ing the tribal language in discussions with the Roman, that is Catholic,
bishops, even though Cyrila was known to speak Latin.81 Most, though
certainly not all, Vandals appear to have had Vandal names.82 Of
the five Arian bishops, Iucundus and Antonius had Latin names, the
others Cyrila, Barbas, and Pinta are probably Germanic. Of the
lower clerics Anduit and Mariuadus had Germanic names. Felix had
a Latin name, but was of Vandal origin. The names of Fastidiosus
and Elpidoforus are respectively Latin and Greek, but both were
converts.83 There was a distinctive Vandal dress, though this might
also be adopted by Romans who served at the Vandal court.84 No
doubt it also happened that a Vandal wore Roman dress. The indi-
cators of “Vandalicity” were not obligatory and exclusive in the
way a particular uniform is obligatory for soldiers on duty in a mod-
ern army. But in general it must have been pretty clear who was,

78
Procopius 3,2,5; Anthologia latina no. 285, ed. F. Bücheler and A. Riese (Leipzig
1894).
79
Victor of Vita 2,4. We have no epigraphic evidence of the use of the Vandal
language, nor explicit reference to its employment in religious services in Vandal
Africa. It is however likely that the Vandals who received their Arian Christianity
from the Goths, continued to use Ulfila’s Gothic bible, on which see P.J. Heather
and J.F. Matthews, The Goths in the fourth Century (Liverpool 1991) pp. 155–97.
80
See references in Courtois, Les Vandales et l’Afrique, pp. 219–20, and the texts
ibid. in Appendix II. Most are undated, four out of 34 are dated by the year of
the king. Tombstones of Romans are dated much more frequently, and by the
provincial era.
81
Victor of Vita 2,55.
82
This could be a circular argument, after all we can only recognise a Vandal
if he has a Germanic name.
83
Courtois, Les Vandales et l’Afrique, p. 225 n. 3.
84
Victor of Vita 1,8. But P. von Rummel, Die beigabenführenden Gräber, pp. 78–91,
argues that buckles and fibulae found as grave gifts were too widely worn to serve
as adequate evidence for what was specific in Vandal habitus. The presence of some
grave gifts is probably evidence that the burial was Vandal, but not of specifically
Vandal attire. For evidence making likely a specific habitus for aristocratic Vandal
women see Eger, “Vandalische Grabfunde aus Karthago”.
72 .... 

and who was not, a Vandal. We do not have the evidence to see
at all clearly the nature of these “signs of distinction”, much less
how they were maintained or modified in the course of time. But
in the end, after the reconquest, the distinction was still clear enough
to enable the Byzantine authorities to carry out a policy of general
deportation of Vandal men.85

V. What was the role of the king in the establishment of the regnum?

Throughout the first half century or so of their stay in North Africa,

certainly as long as Geiseric was king, the Vandals were a formi-
dable people. Their fleet dominated the Mediterranean. No coast
was safe from their raids. In 455 Geiseric sacked Rome itself. In
461 Geiseric defeated a major attack by the western emperor Majorian,
in 468 a great Eastern expedition led by Basiliscus, the Eastern
Emperor Leo’s brother in law. Clearly Geiseric was one of the out-
standing leaders of the age of migrations, like Alaric and Theoderic
and Clovis, all of whom built up a gens into a major power within
the former Empire. But whatever the others may or may not have
done, Geiseric built his regnum around the core of an existing tribal
unit. Classical sources describe earlier leaders of the Vandals as kings.
But in the case of Geiseric we know that he succeeded his brother
Gunderic in 428, but he only counted his regnal years from 439,
the year of the capture of Carthage. So he evidently considered that
in an important sense it was the settlement in North Africa that
marked the beginning of the regnum of the Vandals.86
At the core of the regnum of the Vandalic gens was the tribal dynasty
of the Hastingi. The king’s sons had courts as well as the king. An
unusual rule governed the succession of the monarchy, according to
which the kingdom was transmitted along the line of the deceased
monarch’s brothers, before being passed on to the generation of his
sons.87 This rule is said to have been instituted by Geiseric. It may

85
See below chapter VIII.
86
F.M. Clover, “Time-keeping, monarchy and the heartbeat of Vandal and proto-
Byzantine Africa” (forthcoming).
87
Procopius 3,7,29; 3,9,10; 3,16,13; Jordanes, Getica 169; cf. Courtois, Les Vandales
et l’Afrique, pp. 238–42.
   73

however be traditional, since Geiseric himself had already succeeded

a brother. If the existence of a dynasty made for stability, the rule
of inheritance created a conflict of generations, and repeatedly resulted
in reigning uncles killing their nephews, and other prominent Vandals
thought to be suspiciously close to them. From the start the Vandal
regnum had a fixed capital at Carthage. That is where the king spent
his working life. As a rule the Vandal kings did not lead their armies,
but delegated this duty to their relatives or others. In this they resem-
bled contemporary Roman emperors. The kings resided in what had
been the palace of the proconsul of Africa. There he dined in Roman
style.88
There was a Vandal aristocracy, and we can be sure that some
its members were part of the inner circle of royal followers89 and
advisers.90 But this group included Romans as well as Vandals.91
Confidential envoys of the king had the title comes.92 Roman Land-
owners continued to be addressed with titles of Roman senatorial
rank. Like an emperor, a Vandal king displayed his munificence by
paying for public buildings. In return he would be praised in Latin
verse. A great deal of Latin was certainly spoken at court, where
there evidently were quite a few people, including some of the
kings, who appreciated Latin poetry.93 It is certain that the Van-
dals that could afford to do so adopted the way of life of wealthy
Romans, and enjoyed life in rural villas.94 No doubt both groups
enjoyed watching chariot races in the hippodrome and dancers in
the theatre. The court was where the leaders of the Roman and the
Vandal divisions of North African society met and conferred. To
make the system work, there must have been far more co-operation,
flexibility, and give and take, than our sources would have us believe.

88
Procopius 3,21,1–5.
89
Domestici: Victor of Vita 1,19; 2,42. Admission to court services was by oath,
Victor of Vita 2,39.
90
Mentioned by Victor of Vita 1,18; 1,19; 2,24; 2,27.
91
Prosper of Aquitaine, Epitoma Chronicon 1329 (four Spaniards).
92
Victor of Vita 2,14; 2,28. Is this a translation of a Germanic title?
93
For references to Latin poems dedicated to Vandals, and for other evidence
for the use of Latin at the Vandal court see Courtois, Les Vandales et l’Afrique, p. 222
n. 6. See also F.M. Clover, “Carthage and the Vandals”, Excavations at Carthage 1978
conducted by the University of Michigan, vol. 7, ed. J.H. Humphrey (Ann Arbor 1982)
pp. 1–22, especially pp. 20–2.
94
Courtois, Les Vandales et l’Afrique, p. 250.
74 .... 

Leading Romans were evidently quite ready to work with the Vandals,
and probably even to make concessions to their Arianism. The last
Vandal kings evidently assimilated their ceremony to that of an
emperor. A Latin poet, by name of Florentinus, describes the emperor-
like position of Thrasamund (A.D. 496–532), praised the dual her-
itage, Vandal and Roman, of Hilderic (A.D. 523–530), but puts the
emphasis on his descent from the emperor Theodosius.95

VI. The administration of the regnum

The genres to which our sources belong limit the information they
provide about the institutions of the Vandal kingdom. But this also
means that we must not be too hasty to conclude that a particular
institution did not exist because we do not hear about it. Since the
conquest the royal family owned a great deal of landed estate. The
king’s land was administered by a maior domus, that of the princes
by procurators.96 The king had a chief assistant who engaged in
negotiations on his behalf with the title praepositus regni.97 He would
normally seem to have been a Vandal. I have argued earlier that
the Vandals were subjected to the administration of millenarii. We
know practically nothing of the traditional custom that the millenarii
upheld, except that we know that the Vandals, like the Visigoths,
employed the non-Roman punishment of decalvatio.98
How much of the Roman administration survived has to be deduced
from disparate and fragmentary pieces of information, but the impres-
sion is that a good deal of it continued to function under the Vandals.
The king employed notarii. These included senior officials, perhaps
corresponding to the tribuni et notarii of the imperial court. One Vitarit
is the only holder of this office to be named, evidently a Vandal.99

95
Anthologia latina nos. 371 and 206. Hilderic’s father was married to Eudocia,
daughter of Valentinian III.
96
Victor of Vita 1,45 (Felix); 48 (Saturus); Ferrandus, Vita Fulgentii 1 (Fulgentius).
All have Latin names and were Catholics.
97
Ibid. 2,15 (Heldic); 2,43 (Obadus).
98
Ibid. 2,9, which probably involves not only shaving of the head but scalping.
See E.A. Thompson, The Goths in Spain (Oxford 1969) pp. 103–4 and Averil Cameron,
“How did the Merovingian kings wear their heir?”, Revue belge de philologie et d’his-
toire 48 (1965) pp. 1203–14.
99
Victor of Vita 2,3; 2,41; 2,54.
   75

Other notaries mentioned evidently were simple clerks.100 A primi-

scrinarius with the Roman name of Victorianus had influence over
appointments.101 He may be identical with the Victorianus who was
proconsul of Africa and executed by Huneric for refusing to be
rebaptised as an Arian. He is one of only two provincial governors
to be mentioned in the sources.102 But Huneric’s law of 384 men-
tions judges,103 and the provinces of Proconsularis, Byzacena, and
Numidia are mentioned too. Inscriptions commemorate flamines per-
petui and a sacerdos provinciae dated by the year of the Vandal king.
So the Roman system of provincial administration, and even Roman
provincial assemblies continued. However the fact that governors are
not reported to have played a role in the persecution—or prosecu-
tion—of Catholics suggests that this duty was in practice not assigned
to them, presumably because as a rule they must have been Catholic
themselves.
There was a public fiscal administration in addition to the private
property of the royal family.104 Taxation continued. The Romans
are said to have been heavily taxed,105 while the Vandals were
exempt.106 In practice this seems to have meant that taxes were col-
lected only over part of the territory.107 After the reconquest the
Byzantine administration had new tax registers made.108 We are not
told how taxes were collected. But Huneric’s law of 384 mentions
civic ordines among the institutions responsible for enforcing the law.109
So the ordines of decurions were still functioning, and it is likely that
they collected taxes, as they had always done. At Carthage the clerks

100
Ibid. 3,19; cf. also the epigram Anthologia latina no. 341 on one Eutychus, in
ministrum regis, qui alienas facultates vi extorquebat.
101
Anthologia latina no. 254.
102
Victorianus: Victor of Vita 3,27, the other proconsul is Pacideius (Dracontius,
Romulea 5, ed. F. Vollmer, MGH AA 14 [Berlin 1905] pp. 132–96, here p. 148).
103
Victor of Vita 3,13: iudices provinciarum; cf. 3,11: iudices; 3,9: officiales iudicum;
3,11: primates officiorum.
104
Fiscus and domus regia: Victor of Vita 3,11; fiscus: ibid. 2,23; Procopius 3,5,15.
105
Procopius 1,5; Victor of Vita 2,2.
106
Procopius 1,15.
107
Victor of Vita 3,22: sacerdotibus qui in his regionibus versabantur quae [. . .] palatio
tributa pendebant.
108
Procopius 4,8,25: the Roman registers are said to have been destroyed by
Geiseric. Presumably the Vandal replacements were unacceptable to the Byzantines.
The Byzantine registers were disliked by the tax-payers.
109
Victor of Vita 3,12. There is no mention of a royal representative in the cities
(such as the comes civitatis in the regna in Italy, Gaul and Spain) in Vandal Africa.
76 .... 

of local government organised the billeting of imperial troops in the

traditional way.110 The cursus publicus was functioning.111 When the
little town of Syllectus was taken over by Belisarius, the ordo appears
not to have played a part. The bishop and notables (dÒkimoi) rep-
resented the city,112 as they did by this time in most cities in the
East. A striking example of continuity between Roman and Vandal
Africa is the survival of provincial assemblies (or at least one such
assembly) together with some of the officials of what had been the
imperial cult, and was now presumably a ceremony of loyalty to the
Vandal king.113
That administration of the Vandal kingdom was far from “prim-
itive” is shown most strikingly by its coinage. The administration
needed money. As imperial currency only came into Vandal Africa
in very limited quantities, Gunthamund (484–96) and his successors
created silver and bronze coinage in their own names. This was
based on the imperial coinage, but it also in an important respect
improved on it. For in the second half of the fifth century imperial
silver coins had almost ceased to be issued, and the bronze coinage
had been reduced to small coins of very low value. So the Vandal
state was the first government in the Roman world to bring back a
stable currency of coins of small and medium denominations. The
size of these issues, and the range of their distribution outside Carthage
are still to be found out. The Vandal example was quickly followed
by Theoderic the Ostrogothic king of Italy and not long after by
the imperial government of Constantinople.114

110
Procopius 3,21,10.
111
Ibid. 3,16,12.
112
Ibid. 3,16,11.
113
On Astius Mustelus, Christian and flamen perpetuus (A.D. 526), and Astius Vin-
dicianus, vir clarissimus and flamen perpetuus, and Astius Dinamius, sacerdotalis provinciae
Africae, see A. Chastagnol and N. Duval, “Les survivances du culte impériale dans
l’Afrique du Nord à l’époque Vandale”, Mélanges d’histoire anciennes offerts à William
Seston (Paris 1974) pp. 87–118; F.M. Clover, “Le culte des empereurs dans l’Afrique
vandale”, id., The Late Roman West and the Vandals (Aldershot 1993) chapter 8, pp.
121–8.
114
M.F. Hendy, Studies in the Byzantine Money Economy (Cambridge 1985) pp. 478–90;
K.W. Harl, Coinage in the Roman Economy, 300 B.C. to A.D. 700 (Baltimore-London
1996) pp. 186–95; C. Morrisson, “Les origines du monnayage vandale”, Actes du
8ème Congrès international de numismatique 1973 (Paris-Basel 1976) pp. 461–72.
   77

VII. Arians and Catholics

Among the features that distinguished and divided Vandals from

their non-Vandal neighbours by far the most conspicuous, to us at
least, was their Arian religion. Whether the Arianism of Vandals—
or of the Visigoths or of the Ostrogoths—ought to be classified as
a “strategy of distinction” can be a matter for debate. On the level
of conscious decision conversion to Christianity can rarely have been
a strategy whose objective was “distinction”. It was rather a “strat-
egy of salvation”, and inevitably, in such troubled and dangerous
times, above all a “strategy of insurance”, aimed at ensuring divine
support in all contingencies. As far as the controversy between Cath-
olics and Arians was debated the argument was entirely theological.115
Nevertheless, as can be observed in the history of the Byzantine
Empire, and more recently of Greece, and Ireland,116 and Poland,
and many other countries,117 a particular brand of Christianity can
become an extremely powerful factor in creating regional, ethnic,
or national sentiments of solidarity.
In the Age of Migrations the Arianism of most of the Germanic
peoples separated its adherents from the worship of their Roman
neighbours. It took them out of the responsibility of the bishop. At
least potentially it attracted abuse, and subjected them to discrimi-
natory legislation. In the nature of things the facts of separation and
discrimination must have strengthened the internal bonds of the
immigrant group. The Arian form of Christianity became “their”
religion. It attracted group loyalty, and in turn further strengthened
the group. This development was probably not anticipated and con-
sciously steered from the start, but I find it very difficult to believe
that men with exceptional instincts of leadership, like Alaric, Gei-
seric or Theoderic, were not fully aware of the political uses of the
situation.
How far can it be said that the Arianism of the Vandals had
become an ethnic religion, that is an essential qualification for a

115
E.g. Victor of Vita 2,56–101: “The Book of the Catholic Faith”.
116
E.g. A. Clarke, “Varieties of uniformity: the first century of the Church of
Ireland”, The Churches, Ireland and the Irish, ed. W.J. Sheils and D. Wood, Studies in
Church History 25 (Oxford 1989) pp. 105–22.
117
See the essays in Religion and National Identity, ed. S. Mews, Studies in Church
History 18 (Oxford 1982).
78 .... 

person to be accepted as a full member of the gens? In terms of the-

ology the answer must be not at all. Vandals did not think of their
Arianism as restricted to the Vandal gens. They considered their
Church the true universal Church, just as the Catholics regarded
theirs as universal. On both sides that was the moral justification for
the coercive proselytising (which its victims called persecution), and
it was that which induced king Huneric to demand the right to unre-
stricted worship for those (mainly Gothic federates) who practise “our
religion” in the provinces of the Eastern emperor.118 Sometimes
Vandals did convert to Catholicism, but this did not mean that they
ceased to be Vandals.119 Numerous Romans certainly converted to
Arianism. They were still in Africa, and in public service (militia),
after the Vandals had been deported.120 From the point of view of
Justinian their conversion had turned these people into heretics,
whom he gave a chance to repent. It had not turned them into
Vandals.
But these qualifications do not affect the basic situation. In prac-
tice the Arian Church in Africa was the Church of the Vandals. It
had its own patriarch of Carthage, and bishops, though probably
only a few. For as far as we can see there was no attempt to set
up a bishop in every city, and the bishops we hear about usually
do not have the name of a city attached to their title.121 The Arian bish-
ops supervised parishes ( parochiae).122 These presumably covered areas
where Vandals were settled, perhaps a number of millenae. The ser-
vices were at least partly in Vandal “German”, otherwise Huneric
would surely not have demanded the right to use their own lan-
guage in services for the Arians in the East.123 It was the Arian bish-
ops and clergy who enforced the anti-Catholic measures. Unlike the
civic bishops of the Empire, and the orthodox bishops in even some
of the successor kingdoms,124 the Arian bishops in Vandal Africa

118
Victor of Vita 2,4.
119
Ibid. 3,33 (the wife of Dagila, cellarita regis); 3,38 (two anonymous wealthy
Vandals); 2,9.
120
Justinian, Novellae 37,10 (335 A.D.), ed. R. Scholl (Berlin 1895): rebaptizatos
autem militiam quidem habere nullo modo concedimus. On deportation of Vandals see below
chapter VIII.
121
But Cyrila was Arian bishop of Tipasa in Mauretania (Victor of Vita 3,29).
122
Ferrandus, Vita Fulgentii 6.
123
Victor of Vita 2,4.
124
Particularly under Ostrogoths and Visigoths.
   79

were not given any secular responsibilities, at least as far as we know.

The Arian Church in Vandal Africa was essentially concerned with
worship for Vandals. We may doubt whether conversion to Arianism,
whether forced or not, would by itself have been enough to trans-
form a Roman provincial into a full member of the Vandalic gens.
But it would surely have been an essential step.125 That may well
be why the Vandals were not content with simple attendance at
Arian worship, but insisted on rebaptism.
We do not know when the Vandals were converted to Christian-
ity. The earliest evidence that the Vandals were Christians comes
from the year 423, when they are reported to have chanted verses from
the Bible during their battle against the Roman general Castinus.126
They, or some of them, might well have already been converted in
the Balkans. The conversion of the king, whenever that happened
will have been decisive. But we are not told when that happened.
Hydatius knew of a report that Geiseric had been converted from
Catholicism to Arianism.127 However that may be, when the Vandals
arrived in Roman North Africa they were Arians, and they very
soon entered into severe conflict with the Catholic bishops.128
At the same time as they confiscated secular estates, the Vandals
confiscated some churches together with their property for use by
their own Arian clergy.129 The clergy who had lost their churches
were exiled. Subsequently relations got worse. The Catholics preached
anti-Arian sermons.130 The Vandals exiled numerous bishops and did
not allow those who died to be replaced. Victor claims that at the
time of writing, that is around 489—sixty years after the arrival of
the Vandals in Africa131—there were no more than three Catholic

125
I am assuming that the Vandals admitted outsiders (e.g. Procopius 3,5,20).
We lack the prosopographical data to even begin to assess the extent to which out-
siders were admitted.
126
Salvian, De Gubernatione Dei 7,11(45).
127
Hydatius, Ol. 301,79 a. 428 A.D., pp. 88–90.
128
I use “Catholic” and orthodox to describe the religion of the Empire as
opposed to the religion of the “Arians”. This usage is conventional, though biased.
Arians of course assumed that their Church was Catholic and orthodox.
129
Victor of Vita 1,9; 1,14–8. On churches confiscated at Carthage see L. Ennabli,
Carthage, une métropole chrétienne du IV e à la fin du VII e siècle (Paris 1997), especially the
summary pp. 150–1.
130
If the sermons attributed to Quodvultdeus of Carthage were really spoken at
this time, see Courtois, Les Vandales et l’Afrique, pp. 166–7.
131
Victor of Vita 1,1.
80 .... 

bishops officiating.132 Moreover while this and related actions were

justified in terms of theology, they had clearly recognisable ethnic
and political components. The Vandal authorities were strongly con-
cerned that there should be no Catholic proselytising among the
Vandals themselves.133 Members of the Vandal gens who were seen
entering a Catholic church—note that such people existed—were
attacked, and if they were women, publicly shamed.134 Moreover
Huneric, the most consistent persecutor, evidently suspected the Cath-
olics of disloyalty, for he tried to make them swear that they would
recognise the succession of his son Hilderic, and also that they would
refrain from sending letters across the sea.135 Presumably he feared
that they would intrigue against himself and his son with the im-
perial government. Throughout the period of Vandal persecution
bishops bore the brunt of the attack.136 In this the Vandal kings’
policy towards Catholics was no different from that of emperors,
whether Catholic or Arian, who took action against what they took
to be heresy. But the Vandals put pressure on individual laymen
as well, and significantly they focused on laymen who served the
Vandals as administrators or entertainers. A regulation was that only
Arians might hold office at the court of the king and those of his
sons, and severe pressure was put on Catholics already serving—
and there must have been many of those—to convert to Arianism
by rebaptism.137
As had been the case when the Christians were persecuted in the
pagan Empire, the discriminatory legislation was irregularly observed.
Numerous Catholics continued to attend the Vandal court,138 and
the pressure on the Catholics was relaxed from time to time; par-
ticularly when the Vandals were seeking to improve relations with
the eastern or the western Empire,139 Persecution reached a climax
under Geiseric’s son Huneric in 484. In the course of this action
severe pressure including exile, torture and confiscation of property

132
Ibid. 1,29.
133
Ibid. 3,4; Ferrandus, Vita Fulgentii 6–7.
134
Victor of Vita 2,9.
135
Ibid. 3,19.
136
N.B. not on monasteries. The period may even have seen an expansion of
monasticism.
137
Victor of Vita 1,19–22; 1,43–48; cf. Prosper of Aquitaine, Epitoma Chronicon
1329 a. 437 A.D., pp. 475–6.
138
Victor of Vita 2,8; 2,10; 2,23.
139
Ibid. 2,4.
   81

was employed against thousands of clerics, as well as against numer-

ous lay people to force them to convert.140 Under Huneric’s succes-
sor the pressure on Catholics to convert was gradually relaxed.141 It
is likely that the sources, by virtue of the kind of texts they are, give
a significantly exaggerated picture of the persecution of the Catholics
by the Arians, as well as of the steadfastness of Catholic resistance
to pressure to conform. It is nevertheless evident that for most of
the history of the Vandal kingdom in North Africa, that is over
something like eighty years,142 the regnum, and the gens of the Vandals
and the Arian form of Christianity were closely linked. This created
a real division between the Vandals and the bulk of the population
among whom they had settled. Paradoxically, after the state of the
Vandals had been overthrown, Arian Christianity was to create a
bond between Vandals and some of the Arian soldiers in the Byzantine
army.143

VIII. The end of the Vandals

In Procopius’ account of Belisarius’ campaign we have an eye wit-

ness report from Vandal North Africa, which even if it is largely
focused on military operations, nevertheless gives us glimpses of the
wider context of the campaigning. According to him Africa was
inhabited by three kinds of inhabitants. There were Libyans, who
had once been Romans,144 but who now were very far from identi-
fying automatically with the imperial army. Their support might go
to either side according to circumstance.145 Both sides made efforts
to win their support.146 Then there were Moors under their own
kings, who were still formally rulers of federates appointed by the

140
Ibid. 2,23 to end. Huneric was married to a daughter of Valentinian III and
at peace with the Empire. The motives of the persecution are not clear.
141
Procopius 3,8,6–91; though contrary to the impression given by Procopius,
Trasamundus (A.D. 496–23) forbade the ordination of Catholic bishops and exiled
those ordained against his command: Ferrandus, Vita Fulgentii 13.
142
Say from the capture of Carthage in 439 to the accession of Hilderic in 523
(Procopius 3,9,1).
143
Procopius 4,14,11–15.
144
Ibid. 3,16,3: Libyans are Romans of old, but now their allegiance depends
on how they are treated. Procopius consistently refers to them as Libyans.
145
Ibid. 4,1,8: “treason” of Carthaginians.
146
Ibid. 3,17,6 (Belisarius); 3,23,1,4 (Gelimer).
82 .... 

emperor,147 but who were in practice independent monarchs. In pre-

vious decades they had been extending their territories at the expense
of the Vandals. When faced with the choice between supporting the
imperial forces or the Vandals they consulted nothing but their own
interest.148 Lastly there were the Vandals themselves. These had their
own nobility.149 But Vandalicity was by no means restricted to nobles.
The Vandals were much more numerous than that. They consti-
tuted a large, ruling minority group, who were clearly distinguish-
able from the Libyans.150 After their defeat they had to seek sanctuary.
They could not just melt into the population.151 It was principally
the Vandals whom Gelimer called up to defend his kingdom.152 They
brought their wives153 and their treasures154 with them into the camp.
Evidently they did not think that these would be safe among their
Libyan neighbours. After his victories Belisarius had the Vandal pris-
oners of war deported.155 The wives of the Vandals were now remar-
ried to Roman soldiers, who expected to become owners of their
wives’ Vandal estates.156 Subsequently too, whenever Belisarius or his
successors came into conflict with Vandals, they had them deported.157
If the Vandals disappeared from history, it is because their gens was
deliberately broken up.

IX. Conclusions

The story of the Vandals in Africa is the story of barbarian con-

querors sharing the country with its native inhabitants. The king-
dom was a kingdom of the Vandals. The gens represented the military
power of the state, and the coercive power of the government. The

147
Ibid. 3,25,5–9.
148
Ibid. 3,25,1–10.
149
Ibid. 3,25, 2; 4,3,14; 4,4,32; 4,6,4.
150
Ibid. 3,17,11: the Vandals and others called on to defend Carthage.
151
Ibid. 3,20,1–2; 3,21,11; 4,4,10–13.
152
Ibid. 3,25,1.
153
Ibid. 4,2,8.
154
Ibid. 4,3,25–28.
155
Ibid. 4,5,1; formed into five cavalry units: 4,14,17–18; some avoided depor-
tation: 4,15,3–4.
156
Ibid. 4,14,8–11: an attempt to reclaim the lands as imperial property led to
mutiny.
157
Ibid. 4,19,3; 4,28,40.
   83

effectiveness of the state depended on the cohesion and military spirit

of the Vandal gens, as it had been built up during the migration.
However the administration of the kingdom depended on the old
Roman system of administration. This must have been run mainly
by Romans. The Vandals, the ruling people, were quite distinct from
other inhabitants of their territory. We don’t know the conditions
which had to be fulfilled for somebody to qualify as a Vandal. The
Romans were divided from Vandals by the fact that in the main
they adhered to the Catholic form of Christianity, while Arianism
was the religion of the Vandal gens and kingdom. Religious conflict
was sharper and more persistent in the Vandal kingdom than in any
other of the successor kingdoms. In practice there must have been
a great deal more compromise and co-operation than our sources
imply. The king was king of the Romans as well as the Vandals.
One imagines that many Catholics were prepared to become token
Arians, and thus made themselves acceptable to the king, whether
as members of the court, or of the administration. Culturally too there
was compromise. Romans wore Vandal dress, leading Vandals took
up the way of life of members of the Roman ruling class, including
in some cases at least the higher literacy. Without such compromises
the society could not have worked. Even so plenty of tension remained.
To counteract this required dynastic loyalty, and above all strong
leadership on the part of the king. It is probable that the kingdom
could not have been established without the genius of Geiseric. The
regnum survived under his successors. But the extraordinary dynamism
it had displayed under Geiseric faded fast. The weakening of the
regnum was demonstrated by the ease with which Belisarius overthrew
it. This weakness had become apparent earlier in the inability of the
Vandals to contain the expanding Moorish (i.e. Berber) kingdoms.
It looks as if the Moors were building up gentes and turning gentes
into regna just as the Germanic peoples had been doing before, and
during, their march through the Empire.158 Once the Vandals had
settled, and become accommodated to Roman society, they soon
became as helpless in the face of the gentes evolving along their bor-
ders, as the Romans had been in the face of the Vandals them-
selves. A comparable weakening was experienced by the Visigothic
regnum in Hispania, but not, however, by the regnum of the Franks.

158
E.g. CIL 8,9835 = Courtois, Les Vandales et l’Afrique, Appendix II, no. 95: pro
salute et incolumitate regis Masunae gentium Maurorum et Romanorum.
This page intentionally left blank
 GENS AND REGNUM AMONG THE OSTROGOTHS

Peter Heather

One of the most striking aspects of the career of Theoderic the Amal
is the number of times he was actually made king. He was declared
king first of all in c. 471, after his return from Constantinople, where
he had spent ten years as a hostage, king again upon the death of
his father Thiudimer in c. 475/6, and king still a third time after
his army conquered the forces of Odovacar to take control of Italy.1
Kingship among the Ostrogoths was thus multi-dimensional, involv-
ing the exercise of control not only over the governmental institu-
tions by which the Italian successor state was run, but also, and from
an earlier date, over the military force by whose arms that kingdom
was created. The use of regnum—in the singular—in relationship to
the Ostrogoths is in a sense misleading. Theoderic was king sepa-
rately of his army, and of the Italian kingdom: the fact that both
occupied the same geographical space after 489 should not lead us
to confuse them conceptually.
Whether this still remained the case in Theoderic’s later years,
and under his successors after 526 very much depends upon our
understanding of the military force by which the kingdom was estab-
lished. Did this force replicate itself in the generations after the con-
quest; did it have a sense of its own separate identity which continued
over time? Was it, in short, some kind of a gens, a grouping of peo-
ple with a continuous history which continued even after the settle-
ment in Italy, or was its existence no more than a brief phase in
the history of the rise of the Amal dynasty? It is these demanding
questions that this paper will attempt to address, taking full account
of the now considerable scholarly debate surrounding the whole ques-
tion of identity in the early middle ages.

1
D. Claude, “Zur Königserhebung Theoderichs des Großen”, Geschichtsschreibung und
geistiges Leben im Mittelalter. Festschrift für Heinz Löwe zum 65. Geburtstag, ed. K. Hauck
and H. Mordek (Cologne 1978) pp. 1–13.
86  

A. The Ostrogoths before Italy

The Ostrogoths did not have a long history as a united entity prior
to their entry into Italy in 489. Some of their number, a group of
refugee Rugi fleeing from the destruction of their kingdom by Odovacar
in 487, attached themselves to Theoderic’s train only immediately
before the Italian expedition set out in autumn 488.2 More substan-
tially, Theoderic’s conquering force was composed in large measure
of two militarized groups—both labeled Goths in their own right—
which he had brought together in the Balkans only within half a
decade of their departure for Italy. One of them comprised the army
which Theoderic and his father Thiudimer had brought with them
from Pannonia into East Roman territories in 473/4, so that, by
489, theirs was a history of reasonably well-established loyalty to the
Amal dynasty.3 The same was not true of the other Gothic group,
known to contemporaries as the Thracian Goths. In c. 470, they
appear in Byzantine sources as a force of allied soldiery with a range
of established connections to court circles in Constantinople. The
murder of their preeminent patron Aspar, in 471, led them into
open revolt, and it is very arguable that Theoderic and Thiudimer
brought their Pannonian Goths south in an attempt to convince the
Roman Emperor Leo I that they would make a more trustworthy
set of Gothic allies than their revolting Thracian counterparts.4
Whatever the case, the move initiated a decade of direct competi-
tion, in which the Pannonian and Thracian Goths occasionally con-
fronted each other directly, but more generally tried separately to
convince a sequence of imperial regimes that they were the better
recipients of the one available set of subsidy payments. This rivalry
was brought to an end only in 483/4, when Theoderic the Amal

2
Eugippius, Vita Sancti Severini 44,3–4, ed. R. Noll and E. Vetter, Schriften und
Quellen der Alten Welt 11 (Berlin 1963); Procopius, Wars 7,2,1–2, ed. J. Haury,
transl. H.B. Dewing, 7 vols., Loeb Classical Library (London-Cambridge Mass.
1914–28).
3
There are many modern discussions of the Pannonian history of the Amal-led
Goths; two recent ones are H. Wolfram, History of the Goths (Berkeley 1988) pp.
258–68; P.J. Heather, Goths and Romans 332–489 (Oxford 1991) pp. 240–51.
4
The history of the Thracian Goths needs to be reconstructed from more frag-
mentary references in a range of Byzantine sources, especially but not solely from
the Chronicles of Malalas and Theophanes, and the surviving fragments of the his-
tories of Malchus and John of Antioch. Fuller discussion and argument: Heather,
Goths and Romans, pp. 251–63.
   87

organized the assassination of the newly appointed leader of the

Thracian Goths, Recitach son of Theoderic Strabo (the Amal’s main
rival over the previous decade), and united the bulk of the Thracians
behind him.5 Indications in Byzantine sources suggest that both the
Pannonian and Thracian Goths numbered around 10,000 fighting
men each, and the general course of the narrative certainly confirms
that they were of a similar order of magnitude to one another.
Otherwise direct rivalry would never have continued for an entire
decade; one would have quickly been defeated, or swallowed up, by
the other.6 The Ostrogoths who set off for Italy in 488 were thus
the product of a very recent unification between two groups of more
or less equal size, who had not only not been united in recent past,
but had an immediate history of competition.
Moreover, while the sources make the details impossible to recover
in full, both the Thracian and the Pannonian Goths were themselves
the products of further processes of unification which had occurred
within what was, in c. 490, still living memory. The Thracian Goths
suddenly vault into documented history already fully formed as a
unit, as it were, with Theoderic Strabo as their king in one of the
earliest surviving fragments of the history of Malchus of Philadelphia,
which is clearly relating events of the year 473. Goths are a pres-
ence in Byzantine sources of the 450s and 460s, as army officers
allied to Aspar politically and matrimonially, and a major part of
the imperial palace guard, but there is no sign that at that point
they formed a united force under a single designated leader. No
individual is so labeled in sources prior to the 470s, and instead one
meets a whole series of prominent military officers with Gothic names,
some of whom are explicitly linked to the Thracian military. One
Anagastes, for instance, had been first a military comes in Thrace in

5
Assassination: John of Antioch, Fragmenta 214,3, ed. C. Müller, Fragmenta
Historicorum Graecorum, vols. 4 and 5 (Paris 1868/70). That this caused Gothic
unification has to be argued, but, although some Goths clearly preferred to retain
a Byzantine allegiance (e.g. Bessas and Godigisclus: Procopius, Wars 1,8,3; cf. Heather,
Goths and Romans, p. 302, for other possible examples), no large and distinct body
of Gothic soldiery remained in the Balkans after 489 and it is generally accepted
that unification did follow the assassination: e.g. Heather, Goths and Romans, pp.
300–3 (with further argumentation); Wolfram, Goths, p. 276; L. Schmidt, Geschichte
der deutschen Stämme bis zum Ausgang der Völkerwanderung: Die Ostgermanen (2nd edn.,
Munich 1933) pp. 267–8.
6
Pannonians: Heather, Goths and Romans, pp. 248–9; Thracians: ibid., pp. 253–4.
88  

the 460s and then MVM per Thracias; he was killed when he led a
rebellion in 469/70, but that had been substantial enough to involve
the seizure of some towns. It followed, moreover, his successful sup-
pression of a revolt of another Thracian military leader with a Gothic
name, Ullibos. Ostrys, likewise, commander of the Gothic palace
guard rebelled on the murder of Aspar, fled with his comrades into
Thrace. And, as we have seen, yet another Goth, Theoderic Strabo,
the son of Triarius, finally appears in Malchus of Philadelphia as
the leader of the revolting Thracian Goths in 473. To my mind, all
this suggests that the Thracian Goths were probably recognizable
enough within Thrace in the 460s—one source suggests, indeed, that
they may have been there since the 420s—but they did not func-
tion as a unitary entity with a single overall leader. They should
probably be envisaged instead as semi-separate groups under a num-
ber of separate leaders.7 Interestingly enough, this would appear to
be exactly the kind of arrangement that Theodosius I attempted to
enforce upon the Gothic groups he made his allies under the peace
treaty of 382. Then, too, no single leader was recognized and all
existing leaders of the Tervingi and Greuthungi were suppressed.
From the Roman perspective, this had the very desirable effect of
dampening the capacity for united action on the part of potentially
dangerous allies. And, like the Goths of 382 in the 390s, the formal
entry of the Thracian Goths into rebellion in the 470s involved the
overturning of this arrangement and promotion of a single preemi-
nent leader. The Goths of the 390s chose Alaric, the Thracians of
the 470s, Theoderic Strabo.8
The Thracian Goths may thus have been living in geographical
proximity to one another before the 470s, and may even have had
developed relations with one another (see below), but did not func-
tion as an autonomous entity under a single leader. The same was
probably also true of the Pannonian Goths before the 450s. When
we first encounter them in detail in the pages of the Getica of Jordanes,

7
Full discussion and argument: Heather, Goths and Romans, pp. 256–63. That
some at least were settled in Thrace as early as the 420s is suggested by Theophanes,
Chronographia AM 5931, ed. C. de Boor (Leipzig 1883).
8
Treaty of 382 and no overall leader: Themistius, Orationes 10, ed. G. Downey
(Leipzig 1965) p. 203, ll. 11 ff.; as interpreted by e.g. Wolfram, Goths, p. 133 (the
communis opinio) contra Schmidt, Ostgermanen, p. 419. On events 382–95, see Heather,
Goths and Romans, pp. 191–9.
   89

they were united, but operated a devolved political structure, where

each of three brothers—Valamer, Thiudimer, and Vidimer—had
their own followers living in separate zones of Pannonia.9 They could
and did sometimes act separately. Valamer was attacked at different
points by Huns and Sciri without his brothers participating in the
action,10 and Thiudimer alone attacked the Suevi.11 Valamer’s supe-
riority was acknowledged by the other two, and was inherited by
Thiudimer after his death.12 This suggests very strongly that the
departure from the group of the third brother, Vidimer, on the eve
of the move south into the Roman Balkans, was not the good natured
and politically neutral event portrayed in the Getica. Rather, it was
much more likely the result of tension within the dynasty as Thiudimer
manoeuvred to secure the succession of his son Theoderic—who had
already received his own followers and a first royal promotion—over
that of his younger brother, Vidimer, who was surely expecting to
succeed Thiudimer in turn.13 We also know that Theoderic had at
least one brother,14 but the power-sharing arrangements of the pre-
vious generation were not recreated.
This phase of centralization probably had a longer pre-history. In
what has long been recognized as one of its most problematic pas-
sages, the Getica preserves a series of stories about Gothic defeats at
the hands of a Hunnic king Balamber. The chronology of these sto-
ries is impossible—as their original composer clearly thought too,
since he inserted an unlikely forty year interregnum into the history
of Amal domination to make it work—and they also pose very sub-
stantial problems of historical coherence.15 The name Balamber is

9
Jordanes, Getica 52,268, ed T. Mommsen, MGH AA 5,1 (Berlin 1882). Their
exact geographical location is left uncertain by this passage, but was probably east-
west, rather than north-south, around Lake Balaton: Wolfram, Goths, p. 260.
10
Jordanes, Getica 52,268–9; 53,275–6.
11
Ibid. 56,283.
12
Ibid. 54,278.
13
Ibid. 56,283, with the discussion of Heather, Goths and Romans, pp. 250–1. The
early middle ages throws up many examples of brother to brother succession sys-
tems being interrupted at crucial moments by fathers’ ambitions for sons: not least
King Alfred favouring the claims of his own son Edward over those of the son of
his elder brother.
14
Theodimund: Malchus of Philadelphia, Fragmenta 18, ed. and transl. R.C.
Blockley, The Fragmentary Classicising Historians of the Later Roman Empire 2
(Liverpool 1882; repr. 1983).
15
Jordanes, Getica 48,246–52. There is a long history of discussion of these
90  

suspiciously similar to that of Theoderic’s uncle, Valamer (especially

in its Greek form: balamhr), and the problems all resolve themselves,
if one takes the view that the campaigns of Balamber are actually
confused reminiscences of Valamer’s early career, with a chronology
constructed on the basis of a confusion between Vinitharius (men-
tioned as Ermenaric’s successor in these stories) and Vithimiris men-
tioned in Ammianus Marcellinus (a text clearly known to the author
of this part of the Getica) as the successor of Ermenaric. Taken
together, the stories would imply that Valamer’s career was very sim-
ilar to that of Clovis among the Franks. Prior to his activities, what
were to become the Pannonian Goths actually comprised a series of
separate warbands under at least three dynasties of leaders: those
of the Amal brothers themselves, those of Vinitharius, and those of
Hunimund and his offspring. Valamer united these groups through
a mixture of successful military and political action—defeating and
killing Vinitharius,16 forcing Hunimund’s grandson Beremund to
flee17—and conciliation (marrying Vinitharius’ grand-daughter Vada-
merca, and being accepted as ruler by Hunimund’s son Gensemund).18
The Amal genealogy (figure 1) presents all of these figures as ances-
tors or close collateral relatives of Theoderic; it seems much more
likely that it is collating separate warband dynasties here into one.19
The creation of the Ostrogoths was thus a long drawn out affair.
The Pannonian and Thracian Goths were united only in 483/4 on
the assassination of Recitach, and, while the sources become less
good, and hence details more arguable, before that there are excel-
lent reasons for supposing that the Thracian and Pannonian group-
ings each had complicated pre-histories of their own. Such a picture

problems: Schmidt, Ostgermanen, pp. 253–7; H. Wolfram, “Theogonie, Ethnogenese

und ein kompromittierter Großvater im Stammbaum Theoderichs des Großen”,
Festschrift für Helmut Beumann zum 65. Geburtstag, ed. K.-U. Jäschke and R. Wenskus
(Sigmaringen 1979) pp. 80–97 with references. The problems include (amongst oth-
ers) an extremely unlikely 40–year interregnum in Amal rule and the claim that
the Huns had a king in the 370s, a fact denied by contemporary sources such as
Ammianus.
16
Jordanes, Getica 48,246–8.
17
Ibid. 33,174–5; 48,251.
18
Ibid. 48,249; 48,250–2.
19
Note though the Vadamerca marriage, which means that Vinitharius is in the
“right” place in the genealogy, although he was Valamer’s maternal not paternal
grandfather. The case is fully argued in P.J. Heather, “Cassiodorus and the Rise
of the Amals: Genealogy and the Goths under Hun Domination”, Journal of Roman
Studies 79 (1989) pp. 103–28.
   91

of the creation of the Ostrogoths of c. 490 from diverse origins, very

much within living memory, obviously raises fundamental questions
about the coherence and strength of any sense of identity which may
have operated amongst them at the time of their invasion of Italy
and beyond.
For one thing, it makes it quite clear that old nineteenth- and
earlier twentieth-century visions of the Ostrogoths as an ancient
Germanic “people” are entirely inappropriate. Jordanes’ sixth-cen-
tury account of Gothic history suggests that they had a continuous
existence as a coherent group from the third century, when they
were established north of the Black Sea, but this is part of the sixth-
century Getica’s generally anachronistic back-projection of the sixth-
century order of Gothic society onto the past.20 As we have seen, the
Ostrogoths that Theoderic led to Italy were created in the Roman
Balkans in the 480s out of two main population fragments, which
had had separate histories certainly from the 450s, very probably
from the 420s, and quite possibly for much longer. Indeed, although
it is impossible to prove such a negative point, there is no good evi-
dence to suggest that the Pannonians and Thracians had ever previ-
ously formed part of the same political unit, even if they might have
had other things in common (see below). The Gothic world north of
the Black Sea in the fourth century before the arrival of the Huns
was in all probability highly divided in political terms.21 I would sus-
pect, therefore, that the Pannonians and Thracians were descended
from groups (quite possibly more than one in each case) of Goths
who were already separate in the fourth century and who made
different choices in the face of the Hunnensturm which destroyed this
older world in the last quarter of the fourth century. The would-be
Pannonians may have remained east of the Carpathians until after
the death of Attila in 453; the Thracians were perhaps resettled in
Thrace by the East Romans as early as the 420s, when they were
detached from Hunnic hegemony. The Ostrogoths were straightfor-
wardly a new creation of the Migration Period. This conclusion fits
in perfectly well with both more empirically-led studies, which have

20
Jordanes, Getica 5,42; 14,82. Explored fully in Heather, Goths and Romans, cc.
1–2.
21
I have set out my views on the origins and “shape” of this world in The Goths
(Oxford 1996) cc. 2–3.
92  

stressed that the Germanic world on the fringes of the Roman Empire
was in more or less continuous political flux, and with more com-
parative perspectives deriving from the work of anthropologists and
sociologists, which have stressed that group identities can be remade
much more easily than older work—produced in the great era of
European nationalism—tended to assume.22
Nor, this account makes clear, was the creation of the Ostrogoths
dependent in any substantial way upon the exploitation of ancient
monarchical traditions. One response to the realization that Jordanes’
sixth-century account of unchanging political order among the
Goths cannot survive a detailed confrontation with more contem-
porary sources has been to suggest instead that continuity was pro-
vided by royal families—the Balthi among the Visigoths and especially
the Amali among the Ostrogoths—who ruled both before and after
the Hun-induced cataclysm among the Goths. But this idea also
comes from Jordanes—here clearly drawing upon materials from
Cassiodorus’ lost Gothic history—and is equally unconvincing. The
main piece of evidence in favour of such a view is Jordanes’ account
of a family link between Theoderic the Amal and Ermenaric, the
one supposed ancient Amal ancestor, who also appears independently
in another, more contemporary source: the Res Gestae of Ammianus
Marcellinus. The link is recounted, however, in the same highly
problematic passage of the Getica where the Hunnic King Balamber
appears (see above), and, looked at more closely, two conclusions
emerge. First, the Getica is clearly dependent upon Ammianus at pre-
cisely the point where Ermenaric appears in its text. Second, many
of the problems long-identified in this passage, particularly its chronol-
ogy, resolve themselves if one supposes that the original author of
this part of the Getica was attempting to integrate Ammianus’ account
of Ermenaric and his successors into what he already knew about
Theoderic’s ancestors. Such a view of the genesis of this text is obvi-
ously not susceptible to independent verification, but the hypothesis

22
The flagship of more empirical investigation is R. Wenskus, Stammesbildung und
Verfassung. Das Werden der frühmittelalterlichen gentes (Köln-Graz 1961); the equivalent
work in the comparative tradition is in many ways Ethnic Groups and Boundaries: The
Social Organization of Cultural Difference, ed. F. Barth (Boston 1969). For general introduc-
tions to the literature in the latter tradition, see Heather, Goths, pp. 3–7; P. Amory,
People and Identity in Ostrogothic Italy, 489–534, Cambridge Studies in Medieval Life
and Thought (Cambridge 1997) pp. 14–8.
   93

does solve all the passage’s problems. It also makes sense in two
more general ways. First, it fits in precisely with what the Variae tell
us about the working methods followed by Cassiodorus when con-
structing his Gothic history: namely that he supplemented oral mem-
ory with what he could find in Roman texts.23 Second, Ermenaric
is given a very significant role in the Amal genealogy, being made
the ancestor who provides the missing genealogical link between
Theoderic and his chosen heir, the Visigothic noble Eutharic, whom
Theoderic imported to Italy to marry his daughter Amalasuentha
when it finally became clear that the old king was never going to
sire a son of his own (figure 1). Cassiodorus, it would seem, found
in Ammianus’ Ermenaric a king of sufficient stature to play this cru-
cial genealogical role, and integrated him into the Amal story.24
All of this undermines the idea that there was a living tradition
of obedience to the Amali, stretching back from Theoderic to Erme-
naric, upon which the former was able to draw, and which provided
stability in an otherwise changing Gothic world. Looked at more
broadly, the general pattern of the evidence only confirms the point.
The separation and diversity in the history of those groups who even-
tually made up the Ostrogoths makes it extremely unlikely that they
would have been much persuaded to throw in their lot with Theoderic
the Amal just because he claimed a genealogical link with a famous
ruler or rulers of the past. Indeed, such was the level of competi-
tion between the two Theoderics in the Balkans in the 470s and
early 480s that the giving of allegiance was based upon perceptions
of practical leadership ability, not propagandistic claims. In 477/8,
some of the Amal’s followers transferred to Theoderic Strabo. At
that date, the Amal was young and untried (having just succeeded
his father and born in the mid 450s, he was in his early 20s), and
Byzantine subsidies promised in 475/6 had not arrived. Theoderic
Strabo, with much more established links to the court of Constan-
tinople, may well have seemed the better bet. When the Thracians
did transfer en masse to the Amal, likewise, it followed the acci-
dental death of their seasoned leader—his horse threw Strabo onto
a spear rack—and the accession of his son, Recitach, who proceeded

23
Cassiodorus, Variae 9,25,4–5, ed. T. Mommsen, MGH AA 12 (Berlin 1894).
24
Wenskus, Stammesbildung, pp. 471 ff. originally stressed royal tradition over any
broader continuity; the model was largely followed by Wolfram, Goths. I originally
explored the arguments summarized here in Heather, “Cassiodorus”.
94  

to murder at least two of Strabo’s brothers to secure his position.

Again, a younger leader probably failed to command confidence.25
Even in Italy, after the staggering practical success of Theoderic’s
reign, the Amal tradition, by then being fully developed both in let-
ters of the Variae collection, which stressed that the family was a
unique purple (i.e. “imperial”) dynasty, and in Cassiodorus’s Gothic
history (see above), played a very far from determinative role in
Ostrogothic politics. Theoderic’s choice of heir, after the early death
of his first choice Eutharic, fell upon Athalaric, his grandson by the
marriage of Eutharic and his daughter Amalasuentha. Nonetheless,
presumably because he was still a child (he was either 8 or 10),
Athalaric faced serious challenges not only from within the dynasty,
but also from without; the probability is that one group of opinion
saw a seasoned military veteran by the name of Tuluin as a much
better choice. Athalaric did eventually succeed, but even immedi-
ately after Theoderic’s death, there was no automatic loyalty to the
Amal dynasty. Subsequent events confirm the point. Athalaric died
young after which power passed to his mother and eventually, by a
complicated route, to his uncle Theodahad, Theoderic’s nephew.
Theodahad, however, was not very warlike and, as soon as his inca-
pacity was shown up by the Byzantine invasion in 536, the Gothic
nobility, we are told, organized his murder and elected in his stead
one Wittigis, who had previously shown himself a capable military
commander in Pannonia. In Wittigis’ case, an Amal connection was
maintained, since he married Theoderic’s grand-daughter Matasuentha,
but Procopius is explicit that he was chosen for his military ability.26
Wittigis’ own propaganda within Italy likewise stressed that he belonged
to Theoderic’s family because his deeds qualified him, and made
mention of the marriage only in a letter to the Emperor Justinian,
where the fact of it was deployed to attempt to dissuade Justinian
from continuing to pursue the war.27 After Wittigis, Ildibad, Eraric,
Totila and Teias were elected by different Gothic factions at different
points in the war; none of them had Amal connections.28 Broader

25
Transfer of loyalty: Malchus, Fragmenta 18,1. Death of Strabo and assassina-
tion of Recitach: John of Antioch, Fragmenta 211,5; 214,3. For commentary, see
respectively Heather, Goths and Romans, pp. 279–80; 297 ff.
26
Procopius, Wars 5,11,1–9.
27
Cassiodorus, Variae 10,32; 33.
28
Respectively: ibid. 6,30,4–17; 7,2,4–5; 7,2,10–13; 7,33,6.
   95

historical evidence thus confirms that the deconstruction of the Amal

genealogy and its associated pseudo-history suggested above is broadly
along the right lines, even if details remain arguable. A tradition of
obedience to one royal, Amal dynasty was not a major explanatory
force in shaping the course of Ostrogothic history. Theoderic’s suc-
cesses brought him huge personal charisma, which he clearly tried
to translate, via dynastic propaganda, into unchallenged preeminence
for his immediate heirs. To a very considerable extent, however, his
efforts fell on deaf ears. As in the Balkans in the 470s and 480s, so
in Italy from the 520s: kings of the Ostrogoths relied fundamentally
on their own merits for their election and survival, not an over-
whelmingly powerful and ancient dynastic tradition.29
If the Ostrogoths were neither an ancient “people”, nor the prod-
uct of unchallengeable dynastic tradition, how, then, should we explain
their creation? Part of the nineteenth- and earlier twentieth-century
vision of Germanic groupings of the Migration Period was to sup-
pose that they consisted largely of freemen. Differentials of power,
status, and wealth were always part of the picture, but, largely after
the picture of Germanic society drawn in the Germania of Tacitus, it
was generally supposed that political participation was very wide, in
contradistinction, say, to contemporary imperial Roman society, where
politics was the preserve of an oligarchy. This may or may not have
been true of Germanic society in the first century, but in the last
generation or so, there has been a well-conceived reaction to ideas
of broad social equality and attendant widespread political partici-
pation among Germanic societies of the late Roman period in gen-
eral, and the Goths in particular. In the 1960s, Edward Thompson
published a series of studies which argued that the institutions guar-
anteeing equality and participation in early Germanic society could
no longer be found among the fourth-century Goths. Different strands
of an in some ways parallel tradition in the German-speaking world
have argued along similar lines. Working from some of the ideas of
Reinhard Wenskus, Herwig Wolfram in particular has made the
idea of the Traditionskern a major feature of discussions of Germanic
society in the Late Roman period. Instead of political participation
being quite general, this view argues—on the basis of the massive

29
For fuller discussion of succession matters in Italy, see P.J. Heather, “Theoderic,
king of the Goths”, Early Medieval Europe 4 (1995) pp. 145–73, here pp. 168–72.
96  

discontinuities evident in the history of named Germanic groups such

as those behind the creation of the Ostrogoths reviewed above, where
even explicitly non-Goths (such as the Rugi) became involved in the
process—that group identity was both carried and determined by a
relatively small, elite group, comprising a few (how many is never
explicitly defined as far as I can see) royal or noble clans. Such ideas
have been highly influential, and have largely superseded romantic
nineteenth-century visions of egalitarian Germanic society.30
This reaction to older assumptions was entirely necessary, and, in
part, soundly based. A broader run of evidence suggests to me, how-
ever, that an intermediate position between the two views is both
possible and more accurately descriptive of the inner operation of
Gothic and some other Germanic societies in the fourth and fifth
centuries. To start with, there are some Graeco-Roman descriptions
of decision-making among Gothic groups which suggest that it was
not the preserve of a very small elite. As Edward Thompson himself
pointed out, Ammianus in particular among fourth-century sources
describes Gothic leaders arguing with and persuading followers towards
specific courses of action, and even being disobeyed, rather than sim-
ply issuing orders.31 For the fifth century, likewise, Priscus describes
a revolt of Gothic subordinates against Hunnic overlords in a con-
text where the Goths concerned had explicitly failed to preserve the
kind of independent centralized leadership provided by a king.32 Here
the decision-making clearly wasn’t the preserve of a small, royal elite.
Malchus of Philadelphia, likewise preserves an account of a con-
frontation between the two Theoderics at Mt. Sondis, somewhere in
the eastern Balkans, in 478, where a broader social group applies

30
E.A. Thompson, The Early Germans (Oxford 1965); id., The Visigoths in the Time
of Ulfila (Oxford 1966). The Traditionskern idea has been developed (with differences)
by Wenskus, Stammesbildung, passim, and specifically applied to the Goths by Wolfram,
Goths, passim, but esp. pp. 5–12 (cf. 92 ff.; 115 ff.; 231 ff.). It has proved highly
influential: see e.g. P.J. Geary, Before France and Germany. The Creation and Transformation
of the Merovingian World (New York-Oxford 1988) pp. 61 ff.; L.A. Garcia-Moreno,
“Gothic Survivals in the Visigothic Kingdom”, Francia 21/1 (1994) pp. 1–15.
31
Ammianus Marcellinus, Res Gestae 31,5,8–9; 6,3–5; 7,2–3; 7,8–9; 12,9–4; 15,2
ff., ed. C.U. Clark, 2 vols. (Berlin 1910/15). E.A. Thompson discussed this mate-
rial in “The Visigoths from Fritigern to Euric”, Historia 12 (1963) pp. 105–26, but
argued, overall, that social hierarchies were becoming more pronounced at this time
(cf. his slightly later works cited in note 30).
32
Priscus of Panium, Fragmenta 49, ed. and transl. R.C. Blockley, The Fragmentary
Classicising Historians of the Later Roman Empire 2 (Liverpool 1982).
   97

telling political pressure on the Amal to make him accept an alliance

with his Thracian namesake. This evidence is obviously rather un-
specific and sometimes has literary overtones: particularly, for instance,
in the speeches created by Malchus for the Mt. Sondis incident.33
It is, however, suggestive of a pattern of broader political partic-
ipation, which finds very much more specific and, to my mind at
least, convincing support in the picture of Ostrogothic society which
emerges from Procopius’ account of the Byzantine conquest of Italy.34
Procopius’ war narrative consistently distinguishes between two cat-
egories of troops in the Gothic army, the higher always picked out
by one of three synonyms: “the best” (aristoi ), “the notable” (dokimoi ),
and “the worthy” (logimoi ).35 All three terms are used of significant
Gothic individuals, and to small groups making important policy
decisions.36 These usages are consistent with what we might think of
in a Gothic “nobility”. More interestingly, all three terms are also
applied to much larger militarized Gothic groups. Logimoi alone at
one point in the war manned 47 ships in Totila’s raiding fleet.37
One thousand dokimoi attended the marriage of Theoderic’s sister to
Thrasamund, king of the Vandals,38 and dokimoi later comprised the
entire manpower of units detached for particular purposes during
the war: the garrison of Auximus, and Teias’ force sent against
Verona.39 Aristoi, likewise, could comprise entire garrisons,40 or detached
forces,41 while “many” ( polloi) of them manned the defences of the
Cottian Alps.42

33
Malchus, Fragmenta 18,2.
34
To introduce this evidence at this point obviously begs the question of whether
anything substantial changed in the workings of people named “Goth” between the
Balkans and Italy, and at least one major recent study by Patrick Amory (see note
22) would argue that it did. Amory’s arguments will be reviewed below, but I do
not find them convincing.
35
Procopius, Wars 6,20,2 and 6,23,8, respectively, apply the terms apistoi and
dokimoi to the same group of Goths; the usages are otherwise very generally syn-
onymous. What follows is based on my arguments in The Goths, Appendix 1.
36
Individuals: Procopius, Wars 5,4,13 (logimoi ); 6,1,36; 6,20,14; 7,18,26; 8,26,4
(dokimoi ); 7,1,46 (aristoi ). Policy makers: ibid. 5,13,26; 7,24,27; 8,35,33 (logimoi );
6,28,29 (dokimoi ); 6,8,9 (aristoi ).
37
Ibid. 8,23,10.
38
Ibid. 3,8,12.
39
Ibid. 6,23,8; 8,26,21.
40
Ibid. 6,20,2: Auximus again.
41
Ibid. 5,13,15: Marcias’ force sent to Dalmatia.
42
Ibid. 6,28,29 ff.
98  

These latter examples suggest that to use the modern word “nobil-
ity” for this group gives much too restricted an impression of their
number. In the entourage sent to the Vandal wedding, they formed
one fifth of its military manpower, the rest being armed servants. If
these proportions were sustained throughout Theoderic’s following,
which, as we have seen, probably numbered at least 20,000 fighting
men by the time he reached Italy, then upwards of 4,000 individu-
als fell into this category. I have argued elsewhere that this group
might be equated in legal terms with the freemen class mentioned
in some contemporary law codes (as we shall see, for ideological rea-
sons Theoderic never issued a law code), and the subordinate class
of fighting servants, again found in other contemporary Germanic
groups, perhaps with freedmen. Other evidence from the fifth and
sixth centuries suggests that this kind of social structure may have
been reasonably general in Germanic society of the period. Note,
for example, Codex Theodosianus, where foederati recruited into the
Roman army are expected to have fighting servants;43 this probably
refers to Gothic survivors of Radagaisus’ attack on Italy in 405/6,
who were recruited into the Roman army by Stilicho. Procopius like-
wise describes a sixth-century Lombard military contingent as con-
sisting of 2,500 “good” fighting men and “3,000” fighting servants.44
More generally, studies framed from a variety of perspectives have
begun to suggest that, while there were certainly richer and poorer,
greater and lesser landowners among the Germanic elites of the early
middle ages, nonetheless political power and participation was much
more widely spread than it was to be in and from the Carolingian
period, when the effects of manorialisation transformed social struc-
tures, producing a smaller, more tightly defined and much more
dominant socio-political elite.45 This suggests a compromise between
nineteenth-century romantic and nationalistic visions of Germanic
groups of the Migration Period, composed entirely of free and equal
freemen, and the much more restrictive social models constructed in

43
Codex Theodosianus 7,13,16, ed. T. Mommsen and P.M. Meyer (3rd edn., Berlin
1962).
44
Procopius, Wars 8,26,12.
45
See e.g. G. Halsall, Settlement and Social Organization. The Merovingian Region of
Metz (Cambridge 1995) (Franks); R. Faith, The English Peasantry and the Growth of
Lordship (Leicester 1997) (Anglo-Saxons); more generally, C. Wickham, “Problems
of comparing rural societies in early medieval western Europe”, Transactions of the
Royal Historical Society 6th series 2 (1992) pp. 221–46.
   99

some modern studies. In my view, the freemen of Gothic and other

Germanic groups formed an elite minority within these groups, but
a comparatively broad one. The Lombard evidence might indicate
that just under half of Lombard fighting men fell into this category.
Among the Goths, the elite fighters who attended the Vandal wed-
ding formed one fifth of the attendant guard of honour.
For present purposes, the main point is what the existence of this
broader elite suggests about political participation and decision-mak-
ing among late fifth-century and early sixth-century Gothic groups.
While there certainly was a small core of “the great and good” who
advised Theoderic and his successors, held major administrative posi-
tions and even commanded their armies,46 this was not the only
politically enfranchised group in Gothic society at this date. Alongside
them, there also existed a broader militarized elite, and there is some
good evidence that they were central to group cohesion and morale.
When all members of this class in a Gothic army sent to Dalmatia
were killed, for instance, the remainder (presumably comprising the
armed servant class) simply surrendered.47 Procopius also records the
highly deleterious effects which the loss of the elite fighters on board
Totila’s 47 raiding ships had upon Gothic confidence.48 Any account
of the creation of the Ostrogoths in the Balkans, therefore, needs to
be modeled in terms of the interests of this group. Theoderic’s super-
group could only come into existence because this numerous and
politically enfranchised group wanted it to.
The full story of how Theoderic the Amal persuaded the former
followers of Theoderic Strabo to join him after the death of Reci-
tach will never be known, but the sources and documented sequence
of events make clear the critical role played in the process by the
power of the East Roman state on a whole series of levels. First, it
was the tax machinery of the state which provided Theoderic with
the subsidies which underpinned his power as leader. As we have seen,
the three-way struggle which developed after 473 between the two
Gothic leaderships and the authorities in Constantinople was all about

46
Examples of such men are known from the Balkans (Anstat, Invila, and Soas:
Jordanes, Getica 56,285; Malchus, Fragmenta 20) as well as in Italy (the well-known
Pitzas and Tuluin), where Theoderic never again led his armies personally on cam-
paign after the defeat of Odovacar.
47
Procopius, Wars 7,7,26–37.
48
Ibid. 8,23,29 ff.; 24,3.
100  

securing Roman subsidies, which those leaderships then redistributed

to their followers. Some of the Amal’s supporters defected to Strabo
in 477/8 when the latter seemed the better bet, and it was the
Amal’s successes in 483/4—he was made both MVM Praesentalis and
Consul (the latter a distinction that even Strabo never achieved)—
which surely made the Thracian Goths ready to attach themselves
to his leadership. He was very obviously by this date the Gothic
leader most likely to succeed.49 More generally, it was probably only
the relative productiveness of the Roman economy, and the redis-
tributive power of its taxation system, which made it possible for
one Germanic leader to command sufficient powers of patronage to
keep so many followers happy at one time. Outside the Empire, lower
productivity and lesser taxation powers would have meant that any
leader had less of a surplus to manipulate and thereby manage his
following. Without coexistence of some kind with the more devel-
oped Roman world, therefore, the creation of Theoderic’s super-
group would have been impossible. This is also true, to my mind,
in the case of other kingdom-founding Germanic groups of the
Migration Period and the early middle ages. Alaric’s unification of
three previously separate Gothic groups—the Tervingi, the Greuthungi,
and the followers of Radagaisus—similarly proceeded on the back
of the wealth of the Roman world, and the same was probably also
true of Clovis’ unification of the Salian Franks. Leaders need patron-
age to distribute, economic productivity defines the surplus they have
available for generating loyalty, and hence the overall amount of
loyalty—as it were—that might be purchased.50
Aside from providing Theoderic with the wealth which made it
possible for him to attract an unprecedentedly large following, Roman
power also worked to create the Ostrogoths in another important
way. In the winter of 477/8, the Eastern Emperor Zeno agreed with
Theoderic the Amal and his Pannonian Goths on a plan for a joint
attack on the Thracian Goths. A total of 38,000 Byzantine troops
were to be committed to the plan, together with Theoderic’s own
forces (c. 10,000) with battle being envisaged somewhere in the vicin-
ity of the city of Hadrianople. In the event, no Roman troops ap-

49
See, in more detail, Heather, Goths and Romans, pt. 3.
50
Attila was able to do the same thing largely from beyond the frontier by
extracting Roman wealth in the form of tribute payments to sustain his Empire.
   101

peared and Zeno’s guides led the Amal straight into a confrontation
with Strabo somewhere in the Haemus Mountains, not south of them
in the environs of Hadrianople. One can only agree with the Amal’s
subsequent complaints to Zeno. He had clearly been deliberately
misled into a trap, to try to force him to fight Strabo on his own.
Zeno’s Gothic policy was duplicitous in the extreme. Rather than
favouring the Amal over his Thracian namesake, what the emperor
really wanted was for the two Gothic groups to fight each other and
inflict mutually significant casualties. The East Romans would then
have been in position to exert their military strength to solve the
Gothic problem once and for all.51 As well as the positive factor that
was Roman wealth, therefore, we must also take into account the
negative factor represented by Roman military power. The two Gothic
groups were playing a dangerous game in the Balkans in the 470s
and 480s. They were using their own military capacities to convince
the Constantinopolitan authorities to pay one of them rather than
the other. The Romans, of course, were always likely to get fed up
with these demands for money with menaces and to take appropri-
ate counteraction, as Zeno was attempting to do in 478.
In these circumstances, it made extremely good sense, if it could
be arranged, for the two Gothic groups to cease to compete with
one another, and to work as one unit. The great obstacle to such
an outcome, of course, was the fact that each group had its own
leadership. While the rank and file could (and did) swap sides,
the same was not true of the Pannonians’ and Thracians’ ruling
dynasties who would have found it difficult if not impossible to
occupy a subordinate position under a former rival, having once
been an independent king; nor, indeed, would their rival have tol-
erated them. They were committed to continuing rivalry, therefore,
but the danger posed by the Roman state made unification a much
more beneficial outcome for their followers. These factors, it seems
to me, dictated the means by which the situation was resolved. The
power of the Roman state meant it was better for the Goths to be
united, but also meant that a head-on confrontation between the
two leaderships was impossible; from 478 onwards, all were well
aware that only the Romans could gain from the casualties this would

51
Malchus, Fragmenta 18,1–3, with the commentary of Heather, Goths and Romans,
pp. 282–6.
102  

generate. When unification came about, therefore, it was through

the assassination of Strabo’s son and heir, Recitach. This removed
the major obstacle to unification without damaging the overall mil-
itary capacities of the Goths: a highly satisfactory outcome for every-
one—except Recitach.
Even Zeno was happy enough to accept Gothic unification in the
short-term, having fostered it by the unprecedented honours he
granted Theoderic: above all the Consulship for the year 484. He
was at this point much more concerned with the threat to his throne
posed by the revolt of the Isaurian general Illus in alliance with
members of the old imperial family. For Zeno, therefore, it was
imperative that there should be no disaffected Goths around with
whom Illus might intrigue. As soon as the threat from Illus subsided,
however, matters changed. Events from 485 to 489 cannot be recon-
structed in great detail, but it is clear nonetheless that the united
Goths were now too powerful for Zeno to tolerate. Such was their
power that their leader, buoyed up by his increased powerbase,
threatened to dominate court politics much as Aspar had done in
an earlier generation. Zeno wanted to get rid of him, therefore, but
didn’t want to fight a costly battle, while the Amal, we know, was
worried that the emperor would resort to assassination: a favoured
move for resolving political problems in contemporary Constantinople.
The end result of this impasse was that both sides, after revolts and
temporary solutions, agreed that the best way to solve their prob-
lem was at the expense of Odovacar. Zeno’s Frankensteinian mon-
ster, the newly united Goths, was shipped off to Italy: the problem
solved by exporting it to the west.52
It is possible, and indeed convincing, thus to explain much about
the creation of the Ostrogoths in terms of positive and negative inter-
actions with the East Roman state. That does not seem to me, how-
ever, a fully sufficient explanation. At least for the Thracian Goths,
or the politically-enfranchised freemen amongst them, a major alter-
native did exist to the option of throwing in their lot with Theoderic
the Amal. For certainly twenty, and perhaps as much as forty years
they had formed part of the East Roman military establishment in
the Balkans. Aside from joining the Amal bandwagon, therefore, they
might have reintegrated themselves into a Roman military world

52
In more detail, with full references, Heather, Goths and Romans, c. 9.
   103

with which they were extremely familiar. In the summer of 488, this
had to be compared to the perils and uncertainties of a huge trek
to Italy, part of it through substantially hostile territory. On the jour-
ney, Theoderic’s followers confronted both Sarmatians and Gepids.
The final achievement of an advantageous outcome in Italy would
also depend upon waging successful warfare upon Odovacar and his
army, which had just proved itself so successful in campaigning
against the Rugi, whose refugees had just attached themselves to
Theoderic’s forces. Some Goths, indeed, but clearly only a minor-
ity, took this option, preferring to remain in the Balkans and the
Roman army rather than following the Amal to Italy.53 Given the
uncertainties of the Italian expedition, why was this such a minor-
ity option?
No doubt much freeman decision calculation took the form of
weighing up possible and probable material advantages in Italy and
the Balkans, and the Roman option had the immediate disadvan-
tage that the imperial throne was currently occupied by Zeno. The
emperor was in origin an Isaurian military commander who had
been lifted to prominence as a deliberate counterweight to Aspar
and his Gothic military supporters.54 The current climate in Constan-
tinople may well not have been very attractive, therefore, for large
numbers of the Thracians. Nonetheless, the fact remains that most
of the Goths chose the option which promised best to preserve their
political autonomy as a distinct, Gothic political unit, rather than
disappear back inside, as it were, the East Roman military estab-
lishment. It is quite possible, therefore, that this also played a major
part in the decision-making processes of Thracian and Pannonian
freemen in the summer of 488, especially as they contemplated the
hazards and highly uncertain outcome of the proposed trek to Italy.
Given the nature of the source material, this thought is not sus-
ceptible to proper proof, and, to that extent, should not be pressed.
It is very important, however, not to collapse the range of options
that one allows to remain in play. The sources do not allow us to

53
See above note 5. The story of the trek is well told in Wolfram, Goths, pp.
279–81.
54
The story is very fully told in E.R. Brooks, “The Emperor Zenon and the
Isaurians”, English Historical Review 8 (1893) pp. 209–38, but the anachronistic nation-
alistic assumption that all Goths and all Isaurians would side naturally with one
another sometimes distorts his analysis.
104  

document that a sense of Gothic identity played a key role in the

decisions being taken in the summer of 488, but, equally important,
neither do they deny it. In such circumstances, sound scholarship does
not consist of reverting to the lowest common denominator—the
point at which sources cease to provide unassailable documentation—
but in defining the range of possibilities which the sources leave
open. And, looked at more generally, the behaviour of the mass of
the Thracian Goths in particular does seem less than fully explained
unless we invoke some factor beyond mere material calculation, like
the operation of an active sense of Gothic identity.
I don’t want to labour the point, but it is worth pointing out that
general circumstances would indicate that an active, shared sense of
Gothicness is not an impossible thing to suppose to have been in
operation. The Thracians and Pannonians are both labelled as Goths
in independent sources, before they united to create the Ostrogoths.55
I would not want to argue that the Pannonian and Thracian Goths,
or their ancestors, had ever previously formed part of a single Gothic
political unit. They had, however, a broader history in common.
Significant processes of migration from north-central Europe to the
Black Sea in the third century underlay the creation of the Gothic
realms of the fourth century.56 The groups caught up in these events
then subsequently also faced the fury of the Hunnensturm which again
provided a common stimulus to the Gothic world, even if different
Gothic groups washed up in different corners of Europe as a result.57
Some common memories of third-century migrations and the arrival
of the Huns, no doubt in the form of mythicising stories, would have
played a major role in the evolving identity of both of the Thracian
and Pannonian Goths. In this context, “mythic” does not simply

55
See above on the evidence for the Thracian Goths prior to the revolt of 470.
Amory, People and Identity, c. 8, has recently argued that East Roman military cul-
ture in the Balkans was such a composite that no genuine and operative sense of
Gothic identity could have survived. He is surely right to see the Balkans as a melt-
ing pot to a considerable extent, but his argument does nothing to counter the
specific evidence in Malchus, John of Antioch and Malalas that, prior to the arrival
of the Amal-led Goths in 473/4, there was already a large and distinct body of
Gothic foederati established in Thrace.
56
General account: Heather, The Goths, cc. 2–3. It sometimes goes unnoticed that
contemporary Graeco-Roman sources (not just Jordanes’ Getica) provide strong sup-
port for the operation of migratory processes of some kind from the Baltic to the
Black Seas in the third century.
57
For a recent survey of the Hunnic revolution, see Heather, The Goths, pt. 2.
   105

mean “false”, but a conceptualization of past events, which accords

them a significant role in the history of the group concerned, which
may or may not be in accord with their real significance at the time
the events occurred.58 There is some evidence, interestingly enough,
that Goths in France in the later fifth century and Goths in Cons-
tantinople in the sixth were telling some similar stories about the
ancient Gothic past, and, in particular, about how they came to find
themselves beside the Black Sea.59 These and similar stories may have
played some role in generating a sense of commonality, which became
politically active, when broader circumstances made it advantageous,
in the Balkans in the summer of 488.
Alongside this shared broader history, it is likely enough that the
new unity of the Pannonians and Thracians was sustained by shared
values, institutions, and expectations. Again, we are entering murky
waters, but a few comments are worth making. First, as we have
seen, it is probable that a shared basic social organization marked
out most of the Germanic groups operating in and around the fringes
of the Roman world between the fourth and the sixth centuries, with
a shared tripartite division of humanity into three castes: free, freed
(but permanently dependent), and slave. This was not something
specifically Gothic, but, in the Balkans of 488, would certainly have
separated Goths off from Roman society.60 They also had in com-
mon the same variant of non-Nicene, Homoean Christianity (com-
monly known as Arianism). Again the evidence for this is better from
the Italian period, but, contrary to some recent assertions, there is
no sign that this was an invention of the Italian period (see below).
The united Goths brought their religion with them from the Balkans.
In other key respects too, Gothic ways of doing things probably
differed from Roman norms: marriage practice, inheritance, and

58
For an introduction to myth in this sense, see e.g. A.D. Smith, The Ethnic
Origins of Nations (Oxford 1986) c. 1.
59
Jordanes in mid-sixth century would appear to have heard similar stories about
the deep Gothic past as those collected by the mysterious Ablabius, who was cer-
tainly working somewhere in the west (quite probably in Visigothic Gaul) in the
later fifth or very early sixth centuries: Getica 4,28–9, with Heather, Goths and Romans,
p. 328.
60
It is very striking that such a categorization of society later replaced Roman
social models among the successor states: see further P.J. Heather, “State, Lordship
and Community in the West (c. A.D. 400–600)”, Cambridge Ancient History 14 (2nd
edn., 2000) pp. 437–68, here pp. 461 ff.
106  

dispute settlement. Again, we have no evidence that there were shared

and specific Gothic ways of doing these things which the Thracians
and Pannonians might have preserved to underpin their unification
in the Balkans. Again, however, there is equally no evidence that
they did not, and, as we shall see in due course, arrangements in
Italy make it very clear that dispute settlement, at least, was carried
on in a manner which was very different from what established
norms in Roman society. Most of these items might be thought of
as generally Germanic rather than specifically Gothic, but in the
Balkans of 488 that really didn’t matter. The key context here was
self-definition against Constantinople and a possible Roman option,
not self-definition against other Germanic-speakers, so that these ele-
ments would have been sufficient for the task in hand.
These intriguing possibilities must not be accorded too much air
time. We have little (but not no) evidence for such commonalities,
one can get most of the way towards a satisfactory explanation for
the unification of the Goths, as we have seen, on the basis of sim-
ple material calculation, and the history of the fourth to the sixth
centuries does throw up examples of new groups being created where
there certainly wasn’t a shared past or common grammar of social
norms. The best instance of this is provided by the Vandal and Alan
coalition which made its way eventually to Carthage. This was born
out of shared perceptions of a need to operate in larger numbers
on a more permanent basis following the punishment inflicted upon
both groups by joint Romano-Visigothic campaigns in Spain between
416 and 418. And where the Vandals were Germanic-speakers like
the Goths, the Alans were Iranian-speaking nomads, with—originally
at least—an entirely different economy and social structure. As one
might expect, the stock-dependent Alans had a much more egali-
tarian social structure in the fourth century than agricultural Germanic
societies of the same period.61 Nonetheless, these very different groups,
when placed under sufficient duress, found it possible to work together.

61
Ammianus Marcellinus, Res Gestae 31,2,25.
   107

B. Ostrogoths in the Italian kingdom

The Ostrogothic kingdom in Italy was an effect of western imper-

ial collapse. As we have seen, Theoderic the Amal put together a
new and very substantial militarized force of Goths in the Balkans
between 473 and 488. Everything suggests that this was a larger unit
than had ever previously existed in Gothic society east of Italy or
outside of the Roman Empire.62 Nonetheless, as the history of its
relations with Constantinople makes clear, this force was not by itself
large enough to confront even half of the Roman Empire and win.
In the short term, it was too large to be easily defeated, but nev-
ertheless not strong enough to force the Constantinopolitan author-
ities into an agreement which they did not want to maintain. There
is every reason to suppose that this would also have been the case
in Italy had Theoderic found there the western half of the Empire
still substantially intact. Like Alaric before him between 408 and
411, Theoderic would have been too powerful to defeat easily, but
not powerful enough to constrain the Ravennate authorities into
accepting his domination.63 In other words, Theoderic’s ability to
carve out an independent kingdom at the heart of the old Western
Empire was the product not only of his own ability to create a new
Gothic force of unprecedented power, but also of the collapse of the
west. In the fractured post-Roman west, Theoderic’s Ostrogoths were
powerful enough to carve out a kingdom; before the fracturing, they
would not have been.
On another level, however, the Ostrogoths can be seen as symp-
tom—if not cause—as well as effect. For, like the Ostrogoths, the
other main kingdom-forming entities at the heart of the old Roman
west were new power combinations, created in similar circumstances.
The Hasding Vandal monarchs, who established themselves in North
Africa, sat at the head of a new coalition of previously separate
groupings: the Hasding Vandals, the Siling Vandals, and a number

62
Although the available figures suggest that Alaric’s newly united Visigoths may
have been similar in order of magnitude: Heather, Goths and Romans, pp. 213–4
(commenting on passages in Zosimus and Photius’ summary of Olympiodorus of
Thebes; these figures are highly problematic).
63
The fundamental problem which confronted Alaric as he sat in only tempo-
rary control of events outside the city of Rome in these years, and recognition of
which on both sides underpinned the eventual peace deal between the Romans and
Alaric’s successors Vallia and Theoderic I: Heather, Goths and Romans, c. 6.
108  

of separate groups of Alans. Like the constituent parts of the Ostrogoths,

all these groups were refugees from the ebb and flow of Hunnic
power in east-central Europe, and they were forced to unite in the
face of the power of the Roman state, whose campaigns in Spain
threatened their piecemeal destruction. The same was true of the
Visigoths. Composed of the Tervingi and Greuthungi of 376 and
survivors of Radagaisus’ attack on Italy in 405/6, these constituent
elements can again be shown to have learned to work together
because, in the face of Roman power, there was safety in numbers.64
While the creation of the Ostrogothic kingdom belongs in the cat-
egory of “effects” of Roman imperial collapse, therefore, the processes
of group-creation fit a more general pattern which can be found at
various points and places in the fifth-century west, and which was
a fundamental cause of the eventual deposition of Romulus Augustulus
in early September 476.

Settlement and Ostrogothic Identity

Up to the late 1970s an unhealthy scholarly consensus prevailed over
the question of how exactly the Ostrogoths were settled in the Italian
landscape after Theoderic’s overthrow of Odovacar. The Ostrogoths,
it was argued, were granted land—designated tertiae “thirds” in the
sources because the allocations comprised one third of the available
land—in a number of distinct clusters in the northern part of the
peninsula. This view was overturned by Walter Goffart, who suc-
cessfully showed that the technical term tertiae actually referred to
one third of the existing land tax paid to designated Goths as a mil-
itary donative on an annual basis. Goffart concluded, indeed, that
these land tax allocations comprised the entire economic reward
which Theoderic granted to his followers upon the conquest of Italy.
His further thoughts then turned to how individual Goths turned
this monetary reward into the ownership of actual landed estates.

64
Spanish campaigns and Vandals: Hydatius, Chronicon 24, ed. T. Mommsen,
MGH AA 11 (Berlin 1894). The Sueves of north-western Spain would also appear
to have been a new or renewed amalgamation, created on the march out of Quadi,
Marcomanni and other Middle Danubian groups: Wolfram, Goths, p. 387 n. 55.
On the motivations behind the unification of the Visigoths, see Heather, Goths and
Romans, pp. 315–7.
   109

This was a necessary part of the argument, because there is no doubt

that Goths did eventually come to own land in Italy.65
In 1997, Patrick Amory went further, adding a few new elements
of his own to Goffart’s vision of the economic terms of the settle-
ment, to argue that its more general conditions made it impossible
for any separate sense of Ostrogothic identity to have survived the
integration of Theoderic’s followers into the Italian landscape. First,
Amory argues, Theoderic’s following consisted overwhelmingly of
armed men—an army—who did not bring dependent families of
women and children with them into Italy. Once in Italy, therefore,
they must, for the most part, have intermarried with the indigenous
population.66 Second, Theoderic recruited widely into his army from
among indigenous Italian groups, once his kingdom was established.
Former members of Odovacar’s army, together with large numbers
of native Italo-Romans were incorporated into the military estab-
lishment of kingdom, so that, again, no continuous sense of com-
munity was maintained between the army that conquered Italy
between 489 and 493, and the army of Italy as it stood in 526.67
Third and finally (this being the intersection between Amory and
Goffart), members of the incoming army used the financial gains
they made on an annual basis from the tertiae payments to buy land
piecemeal right across Italy between 493 and the 520s. This process
facilitated their rapid assimilation into local Italian landowning
societies, from which Justinian’s war of conquest dragged them
in the 530s and 540s, forcing them to choose to be either Goths
or imperial Romans, when what they really wanted to be was Tus-
cans, Campanians, Umbrians, or whatever.68 The question of fur-
ther recruitment will be dealt with below, but, in my view, Amory’s
overall vision of the Ostrogoths’ settlement cannot stand, and, as a

65
W. Goffart, Barbarians and Romans A.D. 418–584: The Techniques of Accommodation
(Princeton NJ 1980), esp. c. 3 on Italy.
66
Amory, People and Identity, pp. 95–102 passim, esp. p. 95 “an army is not a
self-replacing breeding community”; cf. pp. 41–2, esp. n. 108, arguing that while
there were some women, there were not many, and that Procopius’ vision (Procopius,
Wars 5,1,12) of Goths coming to Italy with their wives and families is a distortion
rooted in Procopius’ ethnographic assumptions (on which see further below).
67
Amory, People and Identity, pp. 93–5.
68
This argument is spread over more pages than the other two, but see e.g.
Amory, People and Identity, p. 95 (initial financial arrangements); pp. 149–65 (longer
term consequences); c. 5 (the war and its effects).
110  

consequence, the conclusions he draws about the (non-)continuity of

Ostrogothic identity are unconvincing.
First, there is actually much evidence that Theoderic’s armed fol-
lowers brought families with them to Italy on a substantial scale.
Amory says that the presence of families on the trek is only reported
by Procopius—writing forty to fifty years after the event—and dis-
misses it as a product of the latter’s ethnographic ideology. Procopius
expected foreign powers to be tribal, so he recorded them as mov-
ing about with families as a matter of course. This highly conve-
nient blanket dismissal is very problematic. Procopius was more or
less a contemporary observer who had seen many more Goths than
Amory. The immigrant Goths’ families are also mentioned in a wider
range of sources than he supposes. In particular, they are mentioned
in Ennodius’ Panegyric to Theoderic, composed about fifteen years after
the trek and pronounced before the very people who had partici-
pated in it, and in his Life of Epiphanius.69 Detailed, if fragmentary,
records of the Amal’s force in the Balkans also record that in the
late 470s, the Goths’ forces included a “multitude of non-combatants”,
and a wagon train 2,000 strong, which seems even to have included
seed grain among its baggage. East Roman negotiators at this point
certainly assumed that any settlement would involve the granting of
good agricultural land to Theoderic’s followers, and that a lack of
indigenous inhabitants would be an advantage. In other words, they
were expecting that the Goths would do their own farming, not just
live off a surplus pushed in their direction by an indigenous popu-
lation.70 This evidence is neither ideal nor comprehensive, but, col-
lectively, makes it pretty clear that a numerous non-military and at
least partly familial element was part of Theoderic’s following. The
idea that Theoderic’s followers had to marry local Italians is thus
unconvincing.71

69
Ennodius, Panegyricus Theoderico 26–7, ed. W. Hartel, CSEL 6 (Vienna 1882);
Ennodius, Vita Epifani 118–19; cf. 111–12, ibid.
70
Malchus, Fragmenta 18,4, with Heather, Goths and Romans, pp. 244–5.
71
Amory’s prosopography offers only two possible examples (both based on name
forms and mentioned in the same letter) of Gothic men marrying Roman women:
Brandila and Procula, Patza and Regina (Cassiodorus, Variae 5,32). The prosopog-
raphy also includes one Tzitta married to Honorata, but Tzitta was the comman-
der of a Byzantine unit in the Cottian Alps in the period after the reconquest.
Similarly sparse are the prosopographical records of (again to judge by the names)
Goths married to Goths: Waduulfus and Riccifrida, Felethanc and Ranilo (both
   111

Amory’s evidence that the Goths were dispersed right across the
Italian landscape, often as urban garrisons, is no more compelling.
He argues this in particular against John Moorhead, who viewed the
dispersal of Goths into urban garrisons as an effect of mobilisation
against the Byzantine threat in the 530s, rather than the normal
course of affairs.72 In part, Amory may be right, but not completely
so. The Gothic garrison of the city of Naples, for instance, besieged
by Belisarius at the beginning of the war for Italy proper, was
specifically sent there by their king, while their families were else-
where.73 Naples was clearly not where they normally lived. More
generally, Procopius’ evidence indicates very strongly that, in the
530s, the Goths of Italy were concentrated in three regional clus-
ters: in Samnium and Picenum on the Adriatic coast and between
Ravenna and Rome, in Liguria to the north west, and in the Veneto
to the north east.74 This picture is not entirely incompatible with
some of Amory’s views, since the Goths might well have been settled
as urban garrisons within each regional clustering,75 but he would very
much dispute Procopius’ overall implication that the Goths settled
from 493 onwards were still recognisable as such in the landscape
thirty-five or so years later. He would—I take it—regard Procopius’
evidence as another element in the ethnographic “lie” running through
his work (see above), that the Goths were an alien “people” holding
Roman territory which needed to be returned to Roman control.
This is, of course, possible, but Procopius was a contemporary
observer, and Amory has no positive counter-evidence to offer in
support of his hypothesis. It is worth underlining, moreover, just how
“big” the lie would have to have been in this case. The vision of

from Ravenna papyri of the 540s and 550s) and the royal pairing of Theodahad
and Gudeliva. The real point is that the prosopography offers no help here at all.
72
Amory, People and Identity, pp. 93–4 (with notes), contra J. Moorhead, Theoderic
in Italy (Oxford 1993) p. 69 with n. 12.
73
Procopius, Wars 5,8,8–9. Note too ibid. 5,11,28–9, where the ceding of terri-
tory to the Franks allowed troops to be mobilised for the war in Italy which would
otherwise have had garrison duties. The ease with which this was done does not
suggest that the Gothic troops in question were being resettled from previously per-
manent homes in southern Gaul.
74
Evidence conveniently gathered by V. Bierbrauer, Die Ostgotischen Grab- und
Schatzfunde in Italien, Biblioteca degli studi medievali 7 (Spoleto 1975) pp. 23–39.
75
E.g. the Goths guarding the routes through the Cottian Alps into Liguria were
clearly settled in the fortresses of the region: Procopius, Wars 6,28,30–5.
112  

regionally clustered Gothic settlement is not transmitted in one self-

contained, descriptive, passage of the Wars. Rather, it is embedded
in its detailed narrative progression. After he had been holed up in
Rome for over a year (December 536 onwards), besieged by the
Gothic army under Wittigis, Belisarius received some cavalry rein-
forcements, which he used to regain the strategic initiative. They
were employed in separate groups to attack the besieging Goths’
wives and children in their settlements in Picenum. By March 538,
enough pressure had been applied to cause the relevant Gothic men-
folk to want to return home to protect their families. This in turn
prompted the ending of the siege of Rome,76 allowing Belisarius’
forces to regain control of the war’s progression. In 539, the same
tactic was applied in Liguria, causing the dispersal, at a crucial
moment, of two elements of his army: the garrisons of the Cottian
Alps and a mobile force commanded by his nephew Urais.77 It was
precisely this pressure on the Gothic settlements, indeed, which pre-
vented the raising of any relief force and ultimately caused Wittigis’
surrender in May 540.78 If Procopius’ vision of distinct and clustered
Gothic settlement is held to be false, therefore, his whole war nar-
rative must be so too. This is, of course, possible, but surely not
likely.79 I see no reason not to accept, therefore, that the Goths were
settled in regional clusters.80

76
Ibid. 6,7,28–34.
77
Ibid. 6,28,30–5.
78
For a fuller account, see Heather, Goths, pp. 264–7.
79
The biographical information in the prosopography, drawn from other sources,
adds very little to Procopius’ account, and certainly does not contradict it. The
Variae mention Goths with property interests in Picenum and the two Tuscanies
north of Rome (4,14), and Adria (1,19) and Faventia (8,27) in the vicinity of Ravenna.
It also reports the obviously exceptional cases of Tuluin (on whom see below) being
offered a royal estate at Lucullanum near Naples, which he turned down (8,25),
and Theodahad’s land-grabbing in Tuscany (4,39; 5,12; cf. Procopius, Wars 5,4,1
ff.). The papyri mention possible Goths with property interests in Ravenna (Die
nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700 8,II,16, ed. J.-O. Tjäder,
3 vols. [Lund-Stockholm 1954–82] [henceforth: P. Ital.]) and surrounding regions:
Rimini (P. Ital. 36–7, but dated 575/91), Faenza (P. Ital. 8), and Faventia (P. Ital.
30). In Samnium, we hear of Goths with property in Rieti (P. Ital. 7), and north
of Rome in the town of Nepi (P. Ital. 49). There is nothing here to contradict
Procopius’ general account, and the papyri clearly confirm the Samnium and east
coast distributions.
80
Such a distribution also made good strategic sense. So disposed, Theoderic
would have settled his army to cover the main transalpine routes into his new king-
dom, the east coast against possible Byzantine intervention, and main internal east-
west routes across the Apennines.
   113

I would also, finally, take a somewhat different view of the eco-

nomics of the Gothic settlement than does Amory. I do think Goffart
successfully showed that Cassiodorus’ tertiae were indeed one third
of existing tax revenues now earmarked for Gothic use. In my opin-
ion (and that of others), however, this money was used to pay an
annual donative to Goths of military age, which only supplemented
initial grants of land to all of Theoderic’s original followers. Goffart’s
attempts to argue away all the references to land grants are not con-
vincing. His main argument, for instance, is the entirely negative
one that we do not hear of much complaining, which there surely
would have been, had land really been transferred from Roman
landowners to Gothic settlers. Given that all our sources stem directly
or indirectly from Theoderic’s court, however, it is hard to see where
these complaints would be preserved. Goffart also failed to take
account of other possible sources of land, namely the substantial
imperial and other public lands administered by city authorities,
which could have been used to endow the in-coming Goths with-
out any need to resort to private confiscations. While accepting his
account of the tertiae, therefore, Goffart has failed, in my view, to
subvert the indications in Procopius, Cassiodorus, and Ennodius that
there was, in addition, an original tranche of land grants as well.81
Such a vision of the economic arrangements makes good sense of
Theoderic’s general political situation. By 493, he needed to reward
followers who, as we have seen, had spent the last twenty years
marching behind him from Hungary to Constantinople and back to
Italy, fighting countless engagements on the way. Rewards on a gen-
erous scale were a political necessity by this point, and it is hard to
believe that his followers would have been content with a share only
in the interest generated by Italy’s landed economy (tax revenue),
rather than a stake in its landed capital.

81
The most important for Ostrogothic Italy of the many responses to Goffart’s
work is S.J.B. Barnish “Taxation, Land and barbarian settlement in the Western
Empire”, Papers of the British School at Rome 54 (1986) pp. 170–95; cf. I.N. Wood,
“Ethnicity and the ethnogenesis of the Burgundians,” Typen der Ethnogenese unter beson-
derer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der
Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse 201.
Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien 1990)
pp. 53–69 (on the Burgundian evidence). Amory, People and Identity, p. 95 n. 45,
claims that it makes no difference to his argument whether the Goths were given
land or tax shares. If they were given land, with their families, in clusters, it cer-
tainly does.
114  

All in all, there is nothing in the detailed terms of the Ostrogothic

settlement in Italy, which leads one a priori towards the view that it
must necessarily have undermined whatever sense of group identity
already existed among Theoderic’s followers. For Amory’s vision of
a cash rich group of men buying land at random across the Italian
landscape, I would substitute one where Theoderic’s followers, often
consisting—as the sources describe them—of family groups, benefited
from organised land grants, which placed them in clusters in strate-
gically chosen corners of Italy. This initial pay-off was then supple-
mented by an annual donative to designated males of military age,
some of which income may have been used to buy further lands.
In my view, the overall effect of the settlement was to install Theoderic’s
followers as a highly privileged and therefore distinctive, group in
certain parts of the Italian landscape.

Identity and Structure in the Ostrogothic Kingdom

If there was thus nothing about the detailed terms of the settlement
immediately to undermine the corporate identity of Theoderic’s fol-
lowers, there is no doubt that many of the structures by which the
kingdom was subsequently run were thoroughly Roman in inspira-
tion and substance. Odovacar’s seizure of power in 476 had been
by nature a coup d’état rather than a conquest, and both ideologi-
cally and structurally he had wanted and largely maintained the late
imperial world as he found it.82 This heritage was in turn passed on
to Theoderic, the Roman or rather sub-Roman structure of whose
kingdom emerges very clearly from the Variae, the collection of official
correspondence penned by Cassiodorus in two periods of office: one
507–11, the other c. 523–538/9.83 Administrative organization remained
as it had been under the fifth-century emperors. The main unit of
local government was the civitas, comprising both an urban centre
and a stretch of dependent rural territory which was administered
by a council of local landowners (overseen by various imperially-

82
Malchus, Fragmenta 14, is the fundamental source for Odovacar’s constitutional
posturing.
83
P.S. Barnwell, Emperor, Prefects and Kings. The Roman West, 395–565 (London
1992) c. 15 unconvincingly argues that Cassiodorus’ career is a fabrication. A bet-
ter introduction by S.J.B. Barnish in Cassiodorus, Variae, ed. and transl. S.J.B.
Barnish, Translated Texts for Historians 12 (Liverpool 1992).
   115

appointed officials) meeting in the town, where all the relevant records
were kept. The civitates were grouped into provinces, each with their
own governor. At the centre, a number of bureaus both supervised
the broader edifice, not least that of the Praetorian Prefect, and exer-
cised such central functions as the control of revenues and dis-
bursements. The court to which all of these officials were ultimately
responsible was that of a king rather than an emperor, but other-
wise the bureaucratic skeleton of the Ostrogothic Italian kingdom
preserved Roman patterns to a remarkable degree. Most of the other
successor kingdoms did so to some extent, but the range of appoint-
ment letters preserved in books six and seven of the Variae show that
the Ostrogothic kingdom retained a highly differentiated and spe-
cialized bureaucracy where some of the other kingdoms were mov-
ing towards a smaller number of less specialized officials.
The tasks being carried by these offices, and the men undertak-
ing them also represented substantial holdovers from the genuinely
late Roman world of the mid-fifth century. The main governmen-
tal tasks addressed by these structures were the raising of taxation,
and the maintenance of law and order. The nature and even the
names of the taxes being raised directly continued late Roman sys-
tems, both in terms of the overwhelmingly important land tax, and
other lesser taxes, such as those on senators and upon mercantile
activity. Law and order, likewise, was administered through the same
hierarchy of courts as it had been under the Empire, and, through-
out his reign, Theoderic was extremely concerned to emphasise that
the law being applied in these courts was straightforwardly Roman
law, whose remit had not been disrupted by the arrival of Gothic
rule. As we shall see, there is more to this than meets the eye, but
it does seem to have been substantially true that Roman law con-
tinued to be applied, at least to the non-Gothic population of Italy,
entirely as before.84 The men undertaking these bureaucratic tasks,
likewise, were from the Roman landowning classes of Italy, as, indeed,
they always had been. These Italians did not, and had never, formed
one homogeneous, undifferentiated group. There had always been
great differences of wealth and status between them, and different

84
Good accounts of the governmental structure of the kingdom and its func-
tioning can be found in A.H.M. Jones, The Later Roman Empire 284–602. A social,
economic and administrative survey, 3 vols. (Oxford 1964) pp. 253–7, or Moorhead,
Theoderic, cc. 2–3.
116  

types of career profile to suit. These different career trajectories re-

mained, and there is no sign in fact, contrary to Momigliano’s famous
thesis, that the more blue-blooded aristocracy based in and around
Rome failed to accept Gothic rule with the degree of enthusiasm
shown by an aristocracy of service based on Ravenna.85
In spirit too, the kingdom remained thoroughly Roman. The Variae
collection is a plentiful source for ideology and propaganda, arguably
better on these matters than for points of historical substance. What
emerges very clearly is that Theoderic’s regime maintained all the
major elements of the legitimizing discourse which had provided ide-
ological justification and hence general political consent for the Roman
Empire in its heyday. The Ostrogothic kingdom, as the first letter
of the Variae collection famously maintains, was the second properly
“Roman” state in the world. This meant that it had been put in
place to further the grand, Divine design for humankind and the
universe at large, its king, like the emperor in Constantinople, directly
appointed by God. At heart, this design was to ensure that human
beings reached the full zenith of individual development where the
rational mind controlled the physical senses, thus generating a fully
civilized social order: in the technical Latin term, the state of civili-
tas. The main building blocks of this higher order were a proper
classical education, and the maintenance of Roman law, the former
guiding individuals, the latter society in general towards the teleo-
logical conclusion that the Divinity had planned for humankind. It
has sometimes been supposed that this ideological stance was a trope
of Cassiodorus alone, but this is not in my view a tenable position.
Cassiodorus was in office only for a very limited period under
Theoderic, and the ideology can be found deployed in fully devel-
oped form in a whole range of governmental programmes and stances,

85
The thesis is argued in A. Momigliano, “Cassiodorus and the Italian Culture
of His Time”, Proceedings of the British Academy 4 (1955) pp. 207–45. Although still
highly influential, it is flawed in two crucial respects. First, it fails to recognize how
close to Theoderic Symmachus and Boethius actually were up to the 520s, and,
second, is unaware of the succession crisis which hit the kingdom after c. 522 and
of the evidence that Boethius’ fall was probably caused by him having backed the
wrong contender: Theoderic’s nephew Theodahad rather than his grandson Athalaric.
On these points, see P.J. Heather, “The Historical Culture of Ostrogothic Italy”,
Teoderico il grande e i Goti d’Italia. Atti del XIII Congresso internazionale di studi sull’Alto
Medioevo (Spoleto 1993) pp. 317–53, here pp. 332–41, and S.J.B. Barnish, “Maximian,
Cassiodorus, Boethius, Theodahad: literature, philosophy and politics in Ostrogothic
Italy”, Nottingham Medieval Studies 34 (1990) pp. 16–32.
   117

a number of them known entirely independently. In short, Ostrogothic

rule presented itself ideologically as the Roman Empire continued,
and many of its structural forms added real substance to the claim.86
This being so, how were Theoderic’s followers integrated into the
functioning Roman structures of the kingdom, and is there any rea-
son to think that they preserved any continuing sense of corporate
identity?
Theoderic’s propaganda is highly suggestive of the main means
by which the king’s followers, although now, as we have seen, to
some extent dispersed across Italy preserved some sense of commu-
nity. As illustrated in the Variae, the propaganda of Theoderic’s regime
consistently plugged the line that the incoming Goths had taken the
place of the old Roman army in the structures of the state. There
was, in short, a separation of function. The Goths fought, so that
the Italo-Romans, so long as they paid their taxes, could enjoy the
benefits of peace. This idea appears at several points in the Variae,
being by nature an extension of traditional Roman propaganda
which justified taxation in terms of the need to pay for defence, and
has traditionally simply been accepted as broadly true in descriptive
terms. Separation of function thus provided a simple means by
which the immigrant Goths could slot into existing structures with
a minimum of disruption to all concerned: both Goths and Romans.87
This scholarly consensus has again been challenged by Amory, as
part of his revisionary study of identity in Ostrogothic Italy. In his
view, the propaganda of Theoderic’s regime is here deliberately mis-
leading rather than descriptive. In his view, Theoderic recruited
widely into his army from among indigenous Italian groups, once
the kingdom had been established. Former members of Odovacar’s
army, together with large numbers of native Italo-Romans were incor-
porated into the military establishment of the kingdom, and no con-
tinuous sense of community was maintained between the army that

86
The evidence is surveyed in more detail in Heather, “The Historical Culture”,
pp. 320–32. This does not mean, however, that Theoderic was ideologically sub-
servient. Looked at closely, the flattery deployed in Cassiodorus, Variae 1,1, towards
the Eastern Emperor Anastasius barely conceals a demand note, in which Theoderic’s
claimed Romanness was used to berate Anastasius for his alliance with the Franks
against the Goths: Heather, The Goths, pp. 221–30.
87
The evidence has long been recognized, but for a good modern account, see
Moorhead, Theoderic, pp. 71–5.
118  

conquered Italy between 489 and 493, and the army of Italy as it
stood in 526.88
Amory is quite right that substantial numbers of Italo-Romans—
i.e. non Goths—are found in the sources serving the Ostrogothic
regime in a variety of military capacities. What he does not explore,
however, is the categorisation of the troops involved. Some Italo-
Romans do seem to have served with Gothic field armies on cam-
paign. The famous case of the two brothers Cyprianus and Opilio,
who participated in the capture of Sirmium from the Gepids, learned
Gothic, and played a prominent role in the downfall of Boethius in
the 520s is a straightforward case in point. The career of Cyprianus’
sons seems to have followed a similar trajectory. The vast majority
of militarised Italo-Romans encountered in both the Variae and
Procopius’ Gothic War, however, consist of local defense forces. Such,
seemingly, was the nature of the forces commanded by Servatus in
Raetia and Ursus in Noricum, and the detailed narrative of the
Gothic War makes a consistent distinction between the Gothic field
army and local defence forces with primarily garrison responsibili-
ties.89 While there was clearly no absolute boundary preventing Italo-
Romans from serving on particular campaigns, we have no evidence
that this happened on any substantial scale, or that such service
entailed the individuals concerned being accepted into the field
army of the kingdom in regular fashion and qualifying for annual
donatives.
In particular, there is nothing in the Gothic War to suggest either
that large numbers of Roman landowners served with militarized
contingents of their dependents, as seems to have happened in the
Visigothic kingdom of the early sixth century,90 or that the Ostrogoths
drafted into their field armies contingents—again consisting of land-
owners and their dependents—organized corporately on the basis of

88
Amory, People and Identity, pp. 93–5.
89
Servatus: Cassiodorus, Variae 1,10. Ursus is known from dedicatory inscriptions
to a Catholic Church in Noricum: Wolfram, Goths, pp. 292–3 with n. 397. In his
prosopography, Amory categorises both as Goths because they were military men,
but the former in particular is said to have led limitanei (i.e. inferior quality troops),
and I would interpret this—after Wolfram, Goths, pp. 316–7—to mean forces that
did not form part of the Gothic field army. Amory would not recognise that such
forces existed, but see below.
90
At least at the battle of Vouillé: Gregory of Tours Historiae 2,37, ed. B. Krusch
and W. Levison, MGH SSrM 1,1 (2nd edn., Hannover 1951).
   119

civitates. The latter was certainly happening in the Frankish and Visigothic
kingdoms from the middle of the sixth century.91 In the Ostrogothic
kingdom, townsmen were generally responsible—if often only in
alliance with Gothic or East Roman troops—for defending city walls,
but these units are never recorded fighting outside their walls, and
it is unclear that these forces encompassed landowners from the sur-
rounding dependent territory (the situation later among the Franks
and Visigoths) as well as actual townsfolk. I suspect, therefore, that
rather than representative of a general trend in recruiting, Cyprian
and his brother were individual volunteers seeking favour and polit-
ical promotion by serving on a particular campaign. This—rather
than the more organized conscription which followed later—may
also have been the basis on which some Roman landowners found
themselves on the losing side at Vouillé. The sources also preserve
some contradictory accounts of what happened to the military estab-
lishment of Odovacar, with one reporting that they were simply mas-
sacred. A total massacre is surely unlikely, but one can well imagine
that key elements loyal to Odovacar, cut in half by Theoderic at a
dinner party, would have been purged for the threat they would
otherwise pose to the new regime. A former bodyguard of Aetius,
it might be recalled, was responsible for the eventual assassination
of the Emperor Valentinian III, about six months after the latter
had killed his former master.92
Even if there was no absolute barrier, therefore, there is no good
evidence that Theoderic’s propaganda was substantially misleading
in its account of a functional distinction which would necessarily
have preserved a continuing sense of corporate identity among the
Gothic immigrants. Indeed, some of the institutional arrangements
surrounding military service both explain why such a boundary existed
and how it was maintained once the settlement process had dis-
persed Theoderic’s following (in so far as it did: see above). According
to Procopius, all male Goths of military age received an annual

91
Gregory of Tours, Historiae 4,30, shows that the civitas contingents in Merovingian
Gaul consisted of local Gallo-Romans; for a survey of the evidence more generally,
see B.S. Bachrach, Merovingian Military Organisation 481–751 (Minneapolis 1972) pp.
66 ff. For Visigothic Spain, see e.g. Leges Visigothorum antiquiores 9,2, ed. K. Zeumer,
MGH Fontes iuris Germanici antiqui 5 (Hannover 1894) (within which chapter,
laws 8–9 make clear that military obligations fell on Hispano-Romans).
92
Aetius was murdered on 21 or 22 September 453, Valentinian III on 16 March
454.
120  

donative in return for which they held themselves ready for military
service.93 The sources do not make it explicit how this was admin-
istered, but the money was paid out from Ravenna, at the centre
as it were, so that it is extremely likely that some kind of register
was kept of all Goths entitled to payments. Retirement from the
army involved losing the donative, and, in the one specific case for
which we have evidence, an honourable discharge was granted in
writing to the vir sublimis Starcedius.94 Whether this represents gen-
eral practice is unclear, but a centrally organized distribution of
annual payments would certainly have involved keeping a register.95
Control of such a register represented an important lever of power
in royal hands, but also had a communal dimension. The Variae col-
lection preserves one order to the Goths of Samnium and Picenum
to assemble in Ravenna as a body to receive their donatives. This
gave Theoderic the chance, as the letter declares, to investigate the
martial vigour of each individual.96 But if this was done on an annual
basis, and the many references in the Getica to “annual gifts” sug-
gest that it may well have been, then, even when not campaigning,
there will have been an annual occasion where the Gothic soldiery
of particular regions met up with one another. This will obviously
have maintained mutual recognition and sustained senses of corpo-
rate belonging.
In the highly competitive post-Roman world of western Europe,
which saw successor kingdoms seek to establish and extend frontiers,
where previously there had been none, the premium was very much
on military service. Not surprisingly, therefore, the descendants of
Roman landowners quickly began to change career paths towards

93
Procopius, Wars 5,12,47–8.
94
Cassiodorus, Variae 5,36.
95
The need to declare to which kingdom’s register one would henceforth belong
may well explain why, on the death of Theoderic when the two kingdoms were
divided again, individual Goths about whom there was any confusion had to declare
whether they would henceforth be Visigoths or Ostrogoths: Procopius, Wars 5,13,7–8.
96
Cassiodorus, Variae 5,26–7. The letters have occasioned debate. First, they sum-
mon so-called millenarii, who, in the Visigothic kingdom, were officers commanding
one thousand men. Hence some, despite their obvious sense, have thought that they
summoned only a few Goths. See, however, W. Ensslin, Theoderich der Große (Munich
1947) pp. 195–6. Second, Goffart, Barbarians and Romans, pp. 82–8, argued that they
concerned the distribution of tax shares by which all Goths—he supposes—were
supported in Italy. Again, the letters quite specifically refer only to serving soldiers:
Barnish, “Taxation”, pp. 181–3.
   121

the military service which successor state kings so needed, and were
hence likely to reward. For Ostrogothic Italy, however, there is no
evidence that this was already happening in an organized and gen-
eral fashion. All we have are one or two individuals who did cam-
paign, and no certain evidence that even these graduated to full field
army status with an annual donative. Rather, it would seem that the
sub-Roman population was largely limited to defensive garrison duties
in its home cities, while field army service, with its annual donative,
remained a privilege of Theoderic’s original followers and their descen-
dants, the youngsters of the new generations no doubt being intro-
duced to their older peers at pre-campaign musters, and in the
donative assemblies.97
The arrangements established for the practical settlement of legal
disputes in Italy also suggest, first, that there was a real difference
between how Theoderic’s followers dealt with such matters and the
mechanisms of the native Italo-Roman population, and, second, that
these differences survived for some time after the initial settlement.
As we have seen, it was an ideological conceit of Theoderic’s regime
that it preserved the existing, divinely-established, Roman order in
all matters, a key element of which was legal affairs. Roman law,
as we have seen, was the great symbol of society functioning on a
civilized, rational basis, rather than in the barbarian mode where
“might equals right”. In practice, Theoderic was clearly aware, despite
ideologically driven statements, that Romans and Goths were now
both living under Roman law, that special legal arrangements had
to be made. The administration of justice was one of the major tasks
of all the administrative officers Theoderic appointed at different lev-
els through the kingdom. Most were “counts”—comites (sing. comes)—
but had varying competences; some controlled individual cities, some
provinces, and others particular groupings of Goths. The formulae for
appointing these counts preserved in the Variae collection consistently
make the point that cases involving just Romans were to be han-
dled by Roman officials, while the Gothic officers were to handle
inter-Gothic disputes. Cases involving both Goths and Romans were
to be handled by a pair of judges, the Gothic count and his Roman
counterpart.98 Nothing could make it clearer that Theoderic was

97
In texts such as Cassiodorus, Variae 1,38, Gothic iuvenes are sometimes picked
out by Theoderic as a specific group.
98
Cassiodorus, Variae, especially 6,22–3; 7,1–4; 9–16 etc. (all separate letters
122  

expecting that the legal integration of Italo-Romans and incoming

Goths would not be straightforward: in other words, that the latter
had their own ways of doing things. What these ways were, is unfor-
tunately not illustrated anywhere in any detail. For ideological rea-
sons, Theoderic deliberately avoided issuing the kind of lawcode
which, in the cases of other western successor states, has given us
some idea of how dispute settlement might have worked.99 It has
recently been suggested that Theoderic’s legal arrangements should
not be taken as a sign that his followers operated any particularly
“Gothic” legal mechanisms. As was rightly pointed out, the old
Roman army had its own courts and operated separately, in legal
terms, from the surrounding civilian population.100 The details of the
arrangements in Ostrogothic Italy make it clear, however, that we
are dealing with a different situation. Gothic cases were judged by
Gothic officials, for instance, whether the individual was liable for
military service or not; in other words, Goths were Goths through-
out their lives, and not just when they were part of the army.101 In
my view, therefore, it is extremely likely that Theoderic’s followers,
with a continuous history of Gothicness stretching back over their
time in the Balkans, were used to operating with wergilds, feuds, kin
groups and some of the other basic legal mechanisms which appear
in the lawcodes of other successor states.
At the very least, certain prominent Goths clearly exercised such
a degree of social patronage that the successful implementation of
decisions affecting one of their subordinates required their active
assistance. One royal lady, Theodagunda, was warned to show prompt
obedience to the king’s commands in regards to such a legal case,102
and one particularly intractable dispute was even transferred to Theo-
deric’s nephew Theodahad, perhaps because he had more influence
over the people involved than the king himself.103 The arrival of

mentioning the importance of using two judges, one Roman one Goth, for “mixed”
cases).
99
Although, as Patrick Wormald amongst others has rightly cautioned, lawcodes
can never be read as a straightforward account of actual practice. See, e.g, id., “Lex
Scripta and Verbum Regis: Legislation and Germanic Kingship, from Euric to Cnut”,
Early Medieval Kingship, ed. P.H. Sawyer and I.N. Wood (Leeds 1977) pp. 105–38.
100
Amory, People and Identity, pp. 151–65.
101
Cassiodorus, Variae 5,29.
102
Ibid. 4,37.
103
Ibid. 3,15.
   123

the Goths in Italy thus posed some serious practical legal problems,
and Theoderic took appropriate steps.
Internal political structures among Theoderic’s followers likewise
acted to preserve some continuity of old identities, and hence sepa-
ration from local Italo-Romans. An interesting case in point is pro-
vided by those Rugi who, as we have seen, joined Theoderic (under
the leadership of their prince Fredericus) only just before the trek
from the Balkans began in 487. As late as 541, they were still
identifiable as a separate component of the non-Roman population
of Italy, having refused intermarriage not only with the latter but
with any of Theoderic’s other immigrant followers. They also still
had their own sub-leader, one Eraric.104 For this to have been pos-
sible over the two generations separating their first settlement in Italy
and Procopius’ report of them, it seems extremely likely that the
Rugi must have been settled in one cluster, perhaps at some dis-
tance from other elements of Theoderic’s Ostrogoths (precisely where
cannot be recovered).105 The Rugi are obviously a particular case,
but there is no reason to think them unique as a segment of Theoderic’s
following both in having a pre-existing identity and structure, beyond
their membership of the overall group, and in being settled in a
manner which preserved it. It is quite clear, first of all, that local,
semi-autonomous political structures existed within the Ostrogoths
after the Italian settlement. The narrative of the Byzantine conquest
throws up several leaders with local powerbases, which were not
under close royal control. Early on, one Pitzas surrendered with
“half ” of the Goths settled in the region of Samnium.106 These men
seem to have been loyal to Pitzas first, and the monarchy only sec-
ond. Likewise, after Wittigis’ surrender, which initially encompassed
most of the Goths, one group in Venetia, under the leadership of
Ildebad, refused to give in.107 In similar vein, at least some of the
Goths of the Cottian Alps followed the advice of Sisigis (comman-
der of the military garrisons of the region) even when this conflicted
with royal commands.108 Some further light on this phenomenon is

104
Procopius Wars 7,2,1 ff.
105
So too Bierbrauer, Die Ostgotischen Grab- und Schatzfunde, pp. 26–7.
106
Procopius, Wars 5,15,1–2. See on this Wolfram, Goths, p. 502 n. 223: this
Pitzas was not the victor of the Gepid war.
107
Procopius, Wars 6,29,41; 7,1,25 ff.
108
Ibid. 6,28,28 ff.
124  

also shed by one of the Variae. This letter records that the Goths of
Rieti and Nursia in central Italy had decided that their choice of
local leader (labeled a prior) was to be a certain Quidila son of Sibia.
This choice had been accepted by Theoderic who died almost imme-
diately, and hence a further confirmation needed to be extracted
from the king’s grandson and successor, Athalaric.109 These Goths
thus chose their own local leader (how is not specified), but had to
consult the king and obtain his approval. If we can generalise from
this unique exchange, and there is no reason to suppose that we
cannot, this illustrates the existence of a clear balance of power in
post-settlement Gothic politics. The king surely had a veto, but local
Gothic communities had at least semi-autonomous political life around
their own leaders. The existence of leaders of this type, and of the
processes by which they were appointed, not only explains the semi-
autonomous powerbases which turn up in the war narrative, of
course, but would also, once again, have served to keep post-settle-
ment Gothic communities to some extent separate from the local
Italo-Roman societies in which they were otherwise implanted.110
Indeed, although the point should not be overstressed, some of
these post-settlement political structures were probably the direct
descendants of much older ones. This is obviously true of the Rugi,
whose sense of communal identity drew heavily on a history of asso-
ciation which went back at the very least to the independent king-
dom they established on the fringes of Noricum after the death of
Attila (453), and, in some way, beyond.111 This is the best docu-
mented example, but, again, it was not necessarily unique. Not all
pre-existing identities and leaderships were destroyed even when other
Goths joined the Amal bandwagon. After resigning his claims to

109
Cassiodorus, Variae 8,26.
110
It has been argued that the title prior in Cassiodorus, Variae 8,26, is simply an
alternative for that of comes Gothorum (“Count of the Goths”), for whom a general
formula of appointment also exists in the Variae collection (7,3): Ensslin, Theoderich,
pp. 197–8. This may be, and certainly the fact that Cassiodorus drafted a general
letter for their appointment suggests that the Gothic comes was a common enough
office.
111
For an introduction to the Rugi and other groups of the “Attilareich”, see
W. Pohl, “Die Gepiden und die Gentes an der mittleren Donau nach dem Zerfall
des Attilareiches”, Die Völker an der mittleren und unteren Donau im fünften und sechsten
Jahrhundert, ed. H. Wolfram and F. Daim, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 145 (Vienna 1980)
pp. 239–305.
   125

independence in the favour of the Amal Valamer, for instance,

Gensemund continued to lead an at least semi-independent military
force on the former’s behalf.112 And while there is no explicit evi-
dence to this effect, it seems extremely unlikely that intermediate
leadership structures among the Thracian Goths who attached them-
selves to Theoderic in 483/4 were all overturned (even though their
ruling dynasty, of course, had to be exterminated). New political
stars waxed and waned in the new circumstances created by the
Goths’ settlement in Italy, and this is only to be expected.113 None-
theless, it would clearly be wrong simply to assume that there was
no continuity from the past. Either way, the existence of local Gothic
communities with their own leaders, their own means of appointing
these leaders, and their own ties to the king, clearly allowed for the
preservation of a sense of difference among these groups, even though
the settlement had now inserted them into the middle of a whole
series of local Italo-Roman societies.
At least one further element of cultural baggage—religious per-
suasion—would also seem to have acted to preserve senses of differ-
ence between immigrants and local Italo-Romans. It has long been
assumed that, by and large, the Goths held to a non- (or rather pre-)
Nicene version of Christianity which argued that the Divine Son
should be viewed as “like” the Divine Father rather than “of the
same essence” as the Council of Nicaea had asserted in 325 and the
Council of Constantinople in 381 eventually definitively maintained.
This version of Christianity is often mistakenly called “Arianism”. In
reality, it represents an older stratum in the evolution of Christian
belief, which was transmitted, it is held, to the Goths by the fourth-
century missionary Ulfilas, complete with the associated baggage
of a Gothic translation of the Bible and Christian service books.
Although as Roman in origin as the Nicene Creed, by the sixth cen-
tury, adherence to “Homoean” (from the Greek word for “like”)
Christianity, expressed through religious books in Gothic translation,
could thus serve as another boundary to the assimilation of Goth

112
Cassiodorus, Variae 8,9, with Jordanes, Getica 48,246; the illustration works
only if the historical reinterpretation of the latter passage offered in Heather,
“Cassiodorus”, is accepted.
113
The classic illustration is Theudis, who went from his youth at Theoderic’s
court eventually to become first a dominating figure and then actually king in
Visigothic Spain: Heather, “Theoderic”, p. 157.
126  

and Italo-Roman after the establishment of the Italian kingdom. This

traditional view of the importance of religious difference in the king-
dom has, however, recently been challenged on two main grounds.
First, it has been argued, there was nothing specifically Gothic about
Homoean Christianity even by the sixth century. Rather, commu-
nities of such believers had had a continuous history in both the
Roman Balkans and Roman Italy since the fourth century. What
happened in practice was that the Homoean Church of Italy rein-
vented itself as a more specifically Gothic Church in order to attract
the patronage of Theoderic, who just happened to share the same
Creed. Second, if one actually looks at documented religious affiliations,
then a “surprisingly high” number of Catholic Goths appear. Taken
together, these two arguments reverse traditional views on the impor-
tance of different religious affiliations to a continued separation be-
tween Goths and Italo-Romans in the Ostrogothic kingdom.114 Both,
however, are deeply flawed.
To start with, while the continued existence of Homoean com-
munities in the Roman Balkans can be documented into the early
fifth century, there is no evidence for them at a later date. Likewise,
while individual Homoeans are known in Italy in the course of the
fifth century (all soldiers and military commanders of Germanic ori-
gin), no organized Homoean (“Arian”) Church with clergy and build-
ings is documented in Italy between the rise of Ambrose’s dominance
in the 380s and Theoderic’s programme of church construction from
the 490s onwards. The Homoean resurgence in early sixth-century Italy,
therefore, must be linked to the arrival of Theoderic’s Ostrogoths.115
A closer look at the prosopographical evidence also undermines
claims that the evidence throws up a “surprisingly high” number of
Catholic Goths. Table 8 of Amory’s prosopographical appendix com-
piles a list of some “46 named possible Catholics with some other
definite attested criterion of Gothic identity”. The figure is worth a
closer look, because very few of the individuals belong to the time
of Theoderic and the “Gothicness” of others is very much open to

114
Amory, People and Identity, pp. 237 ff. (the Homoean Church in Italy); pp.
476–7 (Catholic Goths = Table 8 of the prosopographical appendix). The com-
ment on the numbers (“surprisingly high”) is taken from p. 465.
115
I follow here the authoritative response to Amory’s arguments offered by
Robert Markus (“Review of P. Amory, People and Identity in Ostrogothic Italy”, Journal
of Theological Studies ns 49 [1998] pp. 414–7).
   127

question. It includes the names of nine converts from Arianism, eight

of whom, according to Amory’s own arguments, may well have con-
verted only in the later stages of the war to keep their lands; i.e.,
as far as we can tell, left to their own devices, they would have hap-
pily remained Arians.116 They can certainly not be taken as exam-
ples of Catholic Goths from the time of Theoderic. Another six
names occur in inscriptions of very uncertain date, which might have
been erected either before Theoderic came to Italy (Fravita, Mustila),
or only after Byzantine pressure made conversion politic (Guntelda,
Basilius, Guntio—all three on the one inscription—and Manifrit). A
more conservative approach, less determined to find Catholic Goths,
would not be so sure that these should be included either. Another
group of three from Ravenna (named in P. Ital. 30) are held to be
Catholics, because a Catholic banker witnessed their deed in 539.
Making these three Catholics certainly involves another substantial
assumption. Popes Boniface II and Pelagius II, together with their
fathers, Sigibuldus and Uniguldus, also appear, because the fathers
had Germanic names. We can take it that the popes were indeed
Catholic, and that their fathers were of Germanic origins, but no
source either labels them or their fathers Goths, as we might expect
in an Italian context had they been so. Definitely not Gothic, or not
Ostrogothic at least, are Bessas (a Byzantine Goth), Asbadus (a Gepid),
and Cyprianus (a Roman). The list similarly contains two eunuch
chamberlains (Seda and Senarius) who can safely be taken as not
part of the Gothic inner military elite, and a vir inluster Meribaudes
who, even according to Amory, is perhaps better attributed to the
well-known fifth-century, originally Frankish then senatorial family.
I would also count as Roman Ursus, the commander of local forces
in Raetia (see above), and his wife Ursina, while Amory himself has
various reservations about the Catholicity of another six (Simplex-
Lavinia, Erdui, Trasimundus, Gundeberga-Nonnica, Montanus, and
the princess Ostrogotho-Areagni). The “surprisingly high” number
of forty-six thus quickly reduces itself to eleven, or fourteen if we
think that using a Catholic banker in Ravenna in 539 is a good sign
of religious affiliation. Not even all of these are entirely convincing.

116
The princess Matasuentha taken to Constantinople in 540, and those appear-
ing in P. Ital. 13 and 49: Sitza, Gundila, Aderit, Ademunt-Andreas, Felithanc,
Ranilo, Eusebius-Riccitanc.
128  

To my mind, the forty-six comes down to four (or seven if we count

the three who used the Catholic banker) more or less unproblem-
atic internal Italian examples from the time of Theoderic himself.
This is obviously a much smaller number, and even if one allowed
some of the more debatable cases, the basic picture would not
change.117
While there were clearly some Catholic Goths in the Italian king-
dom, nothing suggests that there were huge numbers of them, and
it would seem clear that an organized, endowed Homoean Church
was indeed a Gothic import into Italy at the start of the sixth cen-
tury. As for more positive evidence of Gothic “Arianism”, aside from
the royal family, the vast majority of known Gothic “Arians” are in
fact clergy, twenty-seven out of thirty-three secure examples by my
count, and twenty of these are known from just two papyri.118 We
are looking, therefore, at the detritus of accidental survival, and it
is simply impossible to draw any conclusion—positive or negative—
from this kind of evidence. It is certainly going far beyond the avail-
able evidence to deny the possibility that most of the clustered Gothic
elite were or remained Homoean Christians down to the outbreak
of war in 536.119

Conclusion: Trajectories of Transformation

There is obviously much more that one would like to be able to

document about the immigrants who followed Theoderic to Italy in
489. More than enough evidence survives about the structural arrange-
ments and political conditions attendant upon the settlement, how-

117
The eleven are Ereleuva-Eusebia (Theoderic’s mother), Dumilda and her son
Theodosius (but this is another dodgy inscription dated only 477–531), Ariver, Vono-
sus (both inscriptions of uncertain date), Alico (mentioned by Ennodius), Ara (an
unlikely story in Gregory of Tours), Transmundus (an inscription securely dated to
523), Hildevara (who made a gift to a Catholic Church in 524: P. Ital. 4), Amalafridas
(half-Amal, refugee Thuringian prince), and Ranilda (Cassiodorus, Variae 10,36). The
really convincing internal Italian examples are thus Ereleuva-Eusebia, Hildevara,
Alico, and, to a lesser extent, Ranilda (see Amory, People and Identity, p. 409).
118
The names are listed at Table 7, Amory, People and Identity, p. 475. Apart
from ten members of the royal family, only six Gothic lay Arians are attested in
contemporary sources, with, much less impressively, another five mentioned in sto-
ries told by Gregory the Great. Even if one included the latter, the overall picture
would not change.
119
As does Amory, People and Identity, p. 274: “most people who were considered
Goths, like most people in Italy, were or became Catholics”.
   129

ever, to suggest that a consciousness of separate identity continued

among at least the elite core of the immigrants down to the 530s.
The economic terms of the settlement—initial land grants—made
them an obviously privileged group in the Italian landscape, and the
subsequent annual donatives paid out of Italo-Roman taxes to those
of military age not only confirmed that sense of privilege, but also
required list-keeping and distribution mechanisms, which cemented
that sense of difference in place as a major feature of the socio-polit-
ical structure of the kingdom. The immigrants’ geographical place-
ment in clustered settlements within Italy, likewise, meant that it was
possible for a sense of community to exist among various compo-
nent parts of the Ostrogoths at the local level. This certainly oper-
ated in the field of politics, where locally-placed sub-groups had their
own leaders and a continuous, semi-autonomous existence only partly
controlled from Ravenna. It may also have operated at the cultural
level, in at least two specific areas. As we have seen, practical admin-
istrative arrangements within the kingdom on the legal front clearly
imply that the immigrants had norms of dispute settlement which
differed from those operating among the Italo-Romans. On the reli-
gious front too, the revived Homoean Church was a structure cre-
ated by and serving the immigrant Goths.
Significant elements of economic and political advantage combined
with the continuation of their own political, social, and cultural lives
to keep immigrant communities separate in some important ways
from the Italo-Romans among whom they were settled. None of
these, of course, represented absolutely impenetrable barriers to assim-
ilation. Even if the number of documented cases is not as high as
recently claimed, some of Theoderic’s Gothic followers were already
Catholic during his reign, and Theoderic never attempted, as did a
succession of Vandal kings (if to different degrees), to use religious
affiliation aggressively, as a benchmark of group membership among
his immigrant followers. Assimilation also operated on a more polit-
ical level, as both the papyri and Procopius’ war narrative demon-
strate. As we have seen, some groups among the immigrants preferred
to surrender to Belisarius immediately rather than fight to preserve
their independent kingdom, and Amory’s explication of the Gundilas
papyrus is straightforwardly convincing, although there are some seri-
ous textual problems.120 Gundilas clearly had assimilated himself into

120
Amory, People and Identity, Appendix 1, pp. 321–5. The text is P. Ital. 49.
130  

the local landowning society of Nepi in Etruria, and all his changes
of allegiance during the war years, including possibly a conversion
to Catholicism, are entirely explicable in terms of his overwhelming
desire to hold on to a precious parcel of land.
Indeed, had the Byzantine conquest not intervened, such processes
of assimilation would surely have continued, eventually to undermine
the boundaries which originally worked under Theoderic to keep the
immigrant Gothic elite apart from their Italo-Roman peers. The end
result would have created one unified elite. The model of assimila-
tion which pertained in the Visigothic kingdom seems the most likely
analogy. In that case, a steady increase in conversions to Catholicism
among the Homoean Goths seems to have made possible the final
volte face of Reccesuinth’s formal conversion at the Third Council
of Toledo in 587. At the same time, intermarriage and the genera-
tion of new political alliances all focused on the integrating locus of
the royal court made originally separate Gothic and Roman landown-
ers more or less indistinguishable from one another by the seventh
century, and long years of coexistence eventually generated a legal
structure and legal norms which applied to all within the kingdom.121
There is no reason to think that, left to its own devices, similar pat-
terns of transformation would not also have eventually eroded orig-
inal distinctions between Ostrogothic and Italian landowners.
Similar patterns are already visible in the time of Theoderic, whose
court was a centre of cultural patronage for Romans, and a locus
where Romans would appeal to Gothic notables for help with difficult
legal cases, and where one set of Italo-Roman and Gothic allies
might unite to overturn the influence of another. Even in Theoderic’s
reign, court disputes did not see simple patterns of Gothic notables
on one side and Romans on the other.122 That said, it is very clear
to me that the speed of these transformations must not be overesti-
mated. For Gundilas and the Goths who surrendered to Belisarius
in 536—among whom any sense of the importance of preserving
Gothic independence had been eroded—are clearly not representa-
tive of majority opinion among the Gothic freeman elite of the 530s,
a fact which emerges very clearly from the war narrative. In the

121
For a brief account of these processes, see Heather, Goths, pp. 283 ff.
122
Boethius seems to have been an ally of Theoderic’s nephew Theodahad (see
above), and was famously brought down by a competing combination of Gothic
and Roman notables, Cyprianus and Triwila.
   131

year 540, when Wittigis surrendered, it looked as though Belisarius

had won a quick victory, almost as brilliant as his recent triumph
over the Vandals. At that point, Belisarius indicated to the Goths of
northern Italy that, if they would cease hostilities and accept a
Byzantine allegiance, they could keep their landed estates.123 If the
immigrant Goths really had evolved into local Italian landowners by
the 530s, then Belisarius would have been offering them precisely
what they wanted. Large numbers of the Gothic elite, however,
refused the offer. The two clusters of settlement north of the Po—
one around Pavia, the other around Verona—only entered into nego-
tiations at all because Belisarius tricked them into thinking that he
was taking seriously their idea of him preserving their special status
by becoming western emperor at their head. When the extent of
Belisarius’ trickery became apparent, and Byzantine preoccupation
with the Persian front in the early 540s allowed them the opportu-
nity, they not only continued to resist, but widened dramatically the
scale of their revolt. The resulting, increasingly bitter, war lasted in
full fury for another decade, with revolts for a further five years. For
all the Goths who had other ideas, therefore, there were many who
wanted to preserve an independent Gothic kingdom in Italy. The
source of this point of view was the Gothic elite, and I have argued
elsewhere that the war finally came to an end when mounting casu-
alties among this group destroyed any prospect of preserving an inde-
pendent Gothic kingdom. Large number of them were also deliberately
expatriated from Italy by the Byzantines; in some cases, their pre-
cise fate is unknown, but others were transported to the eastern
front, and this, I take it, was the fate of most.124 Within the Italian

123
Procopius, Wars 6,29,35–9.
124
Later stages of the war: Heather, Goths, pp. 267–71. Elite casualties: 1,000
were killed in the battles outside Rome (Procopius, Wars 5,18,14), large numbers
when Totila’s raiding fleet was destroyed, a further 6,000 deaths were inflicted at
Busta Gallorum (we don’t know how many of these were among the elite), and
another substantial group in Teias’ last stand. Expatriation: in 538/9, the garrisons
of Petra, Clusium, Tudra, and Auximum were transported south to Naples and
Sicily (6,11,19 ff.; 13,2 ff.; 27,31 ff.), in the 550s, 7,000 were taken to Constantinople
after Ragnaris’ death (Agathias Myrinaei, Historiarum libri quinque 2,14,7, ed. R. Key-
dell, Corpus Fontium Historiae Byzantinae 2 [Berlin 1967]), and the same fate was
suffered by Widin’s supporters in 561 (Paul the Deacon, Historia Langobardorum 2,2,
ed. L. Bethmann and G. Waitz, MGH SSrL [Hannover 1878]). Given that the
elite consisted of somewhere between 1/5th and 1/2 of c. 20,000 fighting men,
these are very significant casualty rates.
132  

kingdom, therefore, the freemen elite maintained a continuity in their

sense of Gothicness until Byzantine aggression destroyed both it
and them.
This conclusion, it should be stressed, does not represent a return
to the billiard ball view of unchanging identity among migration
period groups. Identity is always a situational construct, and can
always change if sufficient pressure—whether via positive induce-
ments or overwhelming aggression—is brought to bear. Sometimes,
however, identity has more than merely very short-term stability,
especially where it is tied into privileged access to resources of some
kind. The inner freeman elite of the Ostrogoths, in my view, rep-
resent such an example. The conquest of Italy allowed them to estab-
lish themselves in a position of great privilege in a new social and
geographical context. Not surprisingly, therefore, there were elements
of their culture which were different from the indigenous popula-
tion, and not surprisingly, some of these differences, particularly reli-
gious ones, came to symbolize the superiority which their privileges
allowed them to claim. Over time, processes of cultural assimilation
would no doubt have generated new social and political orders as
it did in Gothic Spain, but, by the time the Byzantines launched
their war of conquest, this had not yet happened. Their privileges
ensured that the Gothic elite was still highly visible in the Italian
landscape, and their sense of identity was strong enough to fuel
twenty years of resistance to Byzantine aggression.
   133

Gapt

Hulmut

Augis

Amal

Hisarnis

Ostrogotha

Hunuil

Athal

Achiulf Oduulf

Ansila Ediulf Vultuulf Ermenaric

Valavrans Hunimund

Vinitharius Thorismud

Vandalarius Beremud

Thiudimer Vidimer Veteric

Valamer
Theoderic (Vidimer) Eutharic

Amalasuentha

Athalaric Mathesuentha

Germanus Posthumus

Fig. 1. The Amal genealogy.

This page intentionally left blank
 THE ENIGMATIC FIFTH CENTURY IN HISPANIA:
SOME HISTORICAL PROBLEMS

Javier Arce

Because night is here but the barbarians have not come.

And some people arrived from the borders,
And said that there are no longer any barbarians.
And now what shall become of us without any barbarians?

(C. Cavafy, “Waiting for the barbarians”, wr. 1889, pub. 1904)

I should like to take this opportunity to analyse the situation in

Hispania in the 5th century A.D. This is a complex period, which
is not particularly clear in the limited written source material of
Antiquity, insufficiently studied, intriguing, but all the same funda-
mental for an understanding how the Roman province of Hispania
broke up and was transformed. It was also a crucial period, for it
witnessed the establishment of barbarian regna within the Iberian
Peninsula in a somewhat exceptional and specific way in compari-
son with the Western Empire in general.1

1
I would like to thank all those who took part in the Bellagio colloqium (11–15
December, 2000), directed by Prof. H.-W. Goetz, for all observations and com-
ments they made about my lecture. Some of the problems and ideas that are devel-
oped here form part of the first chapter of my next book, Esperando a los bárbaros
en Hispania (409–507 A.D.). The bibliography covering the settlement of “Barbarian”
peoples in the Roman provinces is extensive. In addition to the contributions included
in this volume, I have to make reference to W. Goffart, Barbarians and Romans A.D.
418–584. The Techniques of Accommodation (Princeton NJ 1980), to Kingdoms of the Empire.
The Integration of Barbarians in Late Antiquity, ed. W. Pohl, The Transformation of the
Roman World 1 (Leiden-New York-Köln 1997), and to Strategies of Distinction. The
Construction of Ethnic Communities, 300 –800, ed. W. Pohl with H. Reimitz, The
Transformation of the Roman World 2 (Leiden-New York-Köln 1998). Always use-
ful A.H.M. Jones, The Later Roman Empire 284–602. A social, economic and administra-
tive survey, 3 vols. (Oxford 1964). For the case of Hispania see R. Collins, Early
Medieval Spain. Unity in Diversity, 400–1000, New Studies in Medieval History (London
1983, 2nd edn. 1995) and P.S. Barnwell, Emperors, Prefects and Kings. The Roman West
395–565 (London 1992). Finally, W. Liebeschuetz, “Cities, taxation and the accom-
modation of the barbarians: the theories of Durliat and Goffart”, Kingdoms of the
Empire, pp. 135–51.
136  

In the 4th century A.D. Hispania had remained sheltered from

the great problems besetting the Empire. Diocletian’s administrative
reorganization of the Diocesis and a relatively detailed list of gov-
ernors—vicarii, praesides and consulares—are well known.2 One of the
most hotly debated questions in recent years has been whether or
not a standing army was to be found in the Diocesis in this cen-
tury. In spite of the (anachronistic) indications of the Notitia Dignitatum,
it would seem to have been demonstrated that the regular Roman
army was no longer present in the Iberian Peninsula. Indirect proof
of this is the fact that no mint was producing coinage within the
Diocesis Hispaniarum, and that its provinces relied on issues of the
Gauls, Rome or the productive mints of the Eastern Empire for
supplies of coinage;3 the hypothesis that, at least from the time of
Diocletian onwards, there was always, as a norm, a mint in those
regions of the Empire where there was an army is well known.4
Another piece of evidence for this might be the fact that, when
Constantine III’s troops entered the Iberian Peninsula in the early
5th century, there is no record of an imperial army having offered
resistance in the Diocesis; on the contrary, members of Theodosius’
family mustered a private army, recruited from among their own
tenants and peasants, which confronted the usurper’s imperial army.5

2
Cf. A. Chastagnol, “Les espagnols dans l’aristocratie gouvernamentale à l’époque
de Théodose”, id., Aspects de l’Antiquitè tardive (Rome 1994) pp. 11–42; J. Arce, “Los
gobernadores de la Diocesis Hispaniarum (ss. IV–V) y la continuidad de las estruc-
turas administrativas romanas en la Peninsula Ibérica”, Antiquité Tardive 7 (1999) pp.
73–83; id., El último siglo de la España romana (284–409 d.C.) (3rd repr., Madrid 1997)
pp. 31–63.
3
Arce, El último siglo, pp. 63 ff. with id., “La Notitia Dignitatum et l’armée
romaine dans la Diocesis Hispaniarum”, Chiron 10 (1980) pp. 539 ff.
4
M.F. Hendy, “Mint and Fiscal Administration under Diocletian, his Colleagues
and his Successors, A.D. 305–324”, Journal of Roman Studies 62 (1972) pp. 75–82.
5
Orosius, Historiarum adversum paganos libri VII 7,40,6, ed. K. Zangemeister, CSEL
5 (Vienna 1882; repr. Hildesheim 1967) [henceforth: Orosius]: servulos suos ex pro-
priis praediis colligentes ac vernaculis alentes sumptibus (“having assembled from their own
possessions and with their own resources an army of serfs and local folk”); cf. also
Zosimus, Histoire Nouvelle 6,4, ed. and transl. F. Paschoud, 4 vols. (Paris 1971–89)
[henceforth: Zosimus] and Sozomen, Kirchengeschichte 9,1, ed. J. Bidez (Berlin 1960).
On this subject cf. Arce, El último siglo, pp. 78 and 153. The importance of this
information should be emphasized as it clearly demonstrates that there was not an
official army of Honorius in Hispania to oppose the tyrannus Constantine III, and
that it was necessary to recruit troops among the “serfs and familiy folk” of the
villas belonging to the members of the Theodosian family living in Hispania (the
reference to “farming folk” is a detail supplied by Zosimus). Recruiting an army
      137

For all these reasons, if any remains of a standing army existed in

the Iberian Peninsula, it was no longer operative.6 However, Hispania
entered the mainstream sphere of events of the Western Empire in
the 5th century with Constantine III’s usurpation, Gerontius’ counter-
revolt and the proclamation of Maximus as Emperor in 411. These
episodes led to generals paying greater attention to events in the
Iberian Peninsula and to Honorius’ policy of not resigning himself
to losing the Diocesis; but, at the same time, they were to lead to
the arrival of the barbarians—the Sueves, Vandals and Alans—who
were at the time stationed in the region of Aquitaine and who crossed
into the Peninsula in October 409 as allies as a consequence of a
pact with Gerontius to contribute to the latter’s establishment in
Hispania as an usurper in opposition to Constantine III.7

implies arming it, and this was possible despite the legislation, collected in the Codex
Theodosianus, which forbade civilians from making use of weapons. Such private
armies, for which we also have evidence on other occasions in other parts of the
Empire (Ammianus Marcellinus, Res Gestae 29,5,34; 29,5,36, ed. J.C. Rolfe, 3 vols.
(Cambridge Mass. 1935–39)—Firmus in Africa under Valentinian I—and other
examples in Sidonius Apollinaris, Ep. 3,13, ed. A. Loyen, 3 vols. (Paris 1960–70);
Procopius, Bellum Gothicum 3,18–22 etc.; cf. Ch. Lécrivain, “Etudes sur le Bas-
Empire”, Mélanges d’Archéologie et d’Histoire 10 (1890) pp. 253–83, esp. pp. 267 ff.:
The private soldiers of the Late Antique Empire worried the Emperor as they might
lead to the creation of independent “kingdoms” in the provinces, and thus a law
in the Codex Theodosianus (12,14,1, ed. T. Mommsen and P.M. Meyer, 3 vols. [3rd
edn., Berlin 1962]) entrusted the care of the countryside to the landowning nobility,
although the official authorities were concerned about the risk of these turning
against the state: cf. ibid. 15,12,3 (A.D. 397). In Hispania, it is clear that this was
not the case of the armies of Didymus and Verinianus, since these were raised
against the usurper Constantine III.
6
Cf. Arce, El último siglo, p. 151; J.F. Matthews, Western Aristocracies and Imperial
Court, A.D. 364–425 (Oxford 1975) p. 320; C.E. Stevens, “Marcus, Gratian and
Constantine”, Athenaeum 35 (1957) pp. 316–47, here p. 318; E.A. Freeman, Western
Empire in the Fifth Century (London 1904) pp. 66 ff.
7
The pact is referred to in Olympiodorus of Thebes, Fragmenta 16 (eirene), ed.
and transl. R.C. Blockley, The Fragmentary Classicising Historians of the Later
Roman Empire 2 (Liverpool 1882; repr. 1983) and in Zosimus, 6,5,2 (with the
commentary of F. Paschoud) and Orosius, 7,40,9. On this see now: Ph. Wynn,
“Frigeridus, the British Tyrants and the Early Fifth Century Barbarian Invasions
of Gaul and Spain”, Athenaeum 85 (1997) pp. 69–117; J.F. Drinkwater, “The usurpers
Constantine III (407–11) and Jovinus (411–13)”, Britannia 29 (1998) pp. 269–98
(which fails to take into account either my own works on the subject or that of
Wynn) and M. Kulikowski, “Barbarians in Gaul, usurpers in Britain”, Britannia 31
(2000) pp. 325 ff.
138  

Barbarians in Tarraconensis

In the first place, I shall discuss a text that I consider to be of fun-

damental importance for an understanding of the situation in Hispania
in the early 5th century, shortly after the arrival of the Sueves,
Vandals and Alans, and once the territorial division to which authors
such as Hydatius and Orosius make reference had been implemented.
I refer to epistula 11 from Consentius to Augustine, which is to be
found among those recently discovered and published by Divjack.8
In the year 418, Consentius, who was an educated monk living
in the Balearic Islands, wrote to Augustine in Africa, explaining the
situation of the religious problems in the area of Tarraconensis. Both
these provinces, Insulae Balearum and Tarraconensis, had remained
unaffected by the division of lands carried out by the barbarians in
411 and, at that moment, were still under Roman control. The let-
ter describes how Severus of Osca (modern Huesca), a rich and pow-
erful priest, had become the leader of the Priscillianist heresy in the
region. Among other things, he had bought a large volume con-
taining magical writings. In the year prior to the writing of Consentius’
letter to Augustine, in other words in 417, Severus, who thought
that the barbarians were at some distance from the region in which
he was moving (Tarraconensis, to be more precise the Ebro valley),
wanted to take these writings—volumina—to the castellum that he had
inherited, formerly his mother’s property. During the journey, he
was attacked by a group of barbarians, who stole the books and,
believing them to be of great value, went to the nearby city of Ilerda
(actual Lleida), aiming to sell them. When they found out their con-
tents, they voluntarily handed them over to Sagittius, bishop of the
city at that time (417). In turn, the latter, who secretly belonged to
the Priscillianist sect, took advantage of the opportunity to tear out the
magical passages that seemed to be of greatest interest and sent the
volume that he considered to be most dangerous as regards its con-
tents, to the Bishop of Tarraco and kept the others in the archive
of his church.9

8
Augustinus, Epistulae ex duobus codicibus nuper in lucem prolatae, ed. J. Divjack, CSEL
98 [Augustinus, Opera] (Vienna 1981).
9
Cf. ibid. 11,2,2–3; 11,2,4; 11,2,5–8 and fig. 1.
      139

The distribution of territory in 411

Several aspects can be deduced from this account. In the first place,
the fact that the priest was aware that the barbarians were far away
from where he lived: Severus aestimans barbaros longius abscessisse. In prac-
tice, it was six years since the division of lands among the barbar-
ians in Hispania had taken place. On the one hand, Hydatius
describes, or defines, it as the consequence of the Lord’s compas-
sion following the disgraces suffered by the Roman inhabitants of
the Peninsula because of the barbarians’ massacres and destruction;10
and on the other, he states that the pact or peace was established
by the barbarian groupings themselves, who divided lands to settle
on (sorte ad inhabitandum sibi provinciarum dividunt regiones, “they divided
the regions of the provinces among themselves to settle on them”).
So, the Sueves, Vandals and Alans had spent two years wandering
around the Iberian Peninsula without finding any stable home (from
409 to 411). It is reasonable to suppose that the Sueves, Vandals
and Alans did not in the first instance enter the Iberian Peninsula
in order to settle, but rather as a result of an agreement with
Gerontius, the usurper, in order to collaborate with his plan to obtain
Hispania. Their arrival was the result of a specific pact and it is not
improbable that they intended to return to their original bases once
this mission has been fulfilled.11 This category of “wandering peo-
ples” has nothing to do with the normal process of settlement, that
is by agreement with Rome, of the barbarian peoples that took place
elsewhere at other moments.12 However, once these two years had
elapsed, the situation in Hispania had changed considerably; Gerontius
had to commit suicide, Maximus, the usurping Emperor raised to
the throne by Gerontius, had to flee in the face of the decisiveness
of the magister militum Constantius, who thus recovered a part of the
Iberian Peninsula for Honorius. This was the moment when the bar-
barians scattered throughout Hispania decided amongst themselves
to divide up the regions of the provinces. Orosius only tells us how

10
Hydatius, Chronicon 41, ed. R.W. Burgess [The Chronicle of Hydatius and the
Consularia Constantinopolitana] (Oxford 1993): Subversis [. . .] provinciis Hispaniae barbari
ad pacem ineundam domino miserante conversi.
11
Cf. note 7 and the works cited therein.
12
Cf. note 1 and the works cited therein.
140  

the distribution of the land was carried out: sorte.13 Hydatius is more
explicit as not only does he mention how it was carried out, but
also what the result of the distribution was:14 the Asding Vandals
obtained the easternmost part of Gallaecia, and the Sueves that part
closest to the Atlantic; the Alans received Carthaginensis and Lusitania,
while the Siling Vandals were given Baetica.
Two questions arise here: did the Roman authorities intervene in
this division? In other words, was there an agreement between the
barbarians and Romans regarding the splitting of lands among the
former and thus were there any established pre-conditions? Or, in
contrast, was this a unilateral decision in which the Roman author-
ities had no say? They drew lots to divide the lands (a well-known
custom among the Germanic peoples, cf. note 1), but this does not
necessarily imply that there was no Roman involvement. They could
have agreed the division with the Romans and then have drawn lots
to decide which portions would be allocated to which tribe. Yet no
text makes any reference to any type of agreement between the bar-
barians and the Romans. The barbarians considered themselves enti-
tled to carry out the division because they had been allowed to enter
the Iberian Peninsula in 409. Nevertheless, there is one significant
fact: the division respected the province of Tarraconensis (as well as
the Insulae Baleares and Tingitania),15 which implies that for the time
being Honorius’ sovereignty in these provinces was recognized, and
since it is more probable, that of Maximus, who at that juncture
was the usurper in Hispania and resided in Tarraconensis. There
was, then, total respect for Roman territorial organization, which
would have been strange if the barbarians had acted independently:
sors would decide who would obtain one region or another, but
within the established limits of the provinces. E.A. Thompson16 con-
sidered that the division was carried out without taking the Romans
into account at all, but this opinion is difficult to sustain if we bear
in mind: a) that in 409 the barbarians had entered as a result of an

13
Orosius, 7,40,10 (“according to luck” or “by lots”).
14
Hydatius, Chronicon 41 (a. 411).
15
These provinces had formed part of the Diocesis Hispaniarum since different
moments in the 4th century. On the evolution of administrative divisions in the
Iberian Peninsula from Diocletian onwards see Arce, El último siglo, pp. 31 ff.
16
Romans and Barbarians: the Decline of the Western Empire (Madison 1982) pp. 155–6;
contra Matthews, Western Aristocracies, pp. 287–8, and Arce, “Los gobernadores” pp.
79 ff.
      141

agreement with Gerontius17 and b) because in that case there would

be no explanation why the province of Tarraconensis was left
untouched for the Romans. Having reached this point, there are
three possible scenarios to explain what happened: first, that the set-
tlement was in accordance with a pre-established agreement with
Gerontius and that all that happened in 411 was the division by
means of sortes. This pact would have included an undertaking not
to settle in Tarraconensis, which would have remained in the hands
of Gerontius and Maximus; second, that the 411 agreement was
made with Fl. Constantius, Honorius’ comes et magister utriusque militiae,
who after having defeated Constantine III and deposed Maximus,
had banished him to live in the regions assigned to the barbarians.18
Constantius had retained a certain control over the Iberian Peninsula
(included in the agreement) and continued to command in Tarraco-
nensis, which thus constituted a sort of hinterland with respect to
Gallic territories. Shortly afterwards, in 418, Constantius was to give
lands between Toulouse and Aquitaine to the Goths, thereby pro-
viding an additional safeguard against the tribes settled in the Diocesis
Hispaniarum; and third, that the barbarians made the agreement with
the usurper Maximus. Once the latter was overthrown, Tarraconensis
remained in the hands of legitimate Roman forces and he had to
take refuge among his former allies (with whom he was later, in 422,
to attempt a further revolt),19 who were already de facto masters of
the rest of the Iberian Peninsula.
Of these three possibilities, I consider the last to be the most likely:
after having wandered for two years, the Sueves, Vandals and Alans
reached an agreement with their former ally, now Emperor; they
divided up the territory, leaving Tarraconensis in his hands. But the
fall of Maximus left the province in Honorius’ hands again. From
this moment onwards, Tarraconensis was to be a constant target of
conquest on the part of the barbarians.
The kings who divided up the territories of the rest of the Peninsula
were Guteric, for the Asding Vandals;20 Fredbal for the Siling Vandals21

17
Olympiodorus, Fragmenta 16.
18
Ibid.; Orosius, 7,42,5.
19
J. Martindale, The Prosopography of the Later Roman Empire 2 (Cambridge 1980):
Maximus 4 [henceforth: PLRE II].
20
Hydatius, Chronicon 71; C. Courtois, Les Vandales et l’Afrique (Paris 1955) p. 393;
PLRE II, s.v.
21
Hydatius, Chronicon 62; Courtois, Les Vandales, pp. 54 and 237 with n. 1.
142  

and Hermeric, king of the Sueves.22 Their followers cannot have

been very numerous and could not have occupied all the lands; for
that reason, Severus of Ilerda thought that they were far away from
Tarraconensis when he undertook his journey. But Consentius’ let-
ter enables us to see that one of the activities of the newcomers was
to attack and rob, even outside the frontiers of the regions that they
had been assigned for settlement purposes. An old custom that was
not only a characteristic of the barbarians, but also an endemic fea-
ture of earlier periods.23

Barbarians as book thieves

In the episode that Consentius recounts the haul was by no means

spectacular—neither solidi, tableware nor any other object of imme-
diately obvious value, but rather a number of books (volumina). These
barbarians were able to appreciate them (the books were probably
illustrated), to such an extent that they at once took them to the
nearest civitas, Ilerda in this case, in order to make enquiries about
their value and sell them. By no means were these uncouth, des-
perate highwaymen unable to appreciate the value of a book or a
manuscript and only seeking gold or other precious metals. Their
contact with Roman culture had made them aware of, understand
and appreciate the value books might have, above all some, as
unusual and intrinsically valuable objects.24 They almost certainly did
not stop to enquire about their contents. In fact, they went straight
to Ilerda and there they discovered how much they might be worth
and, when they found out that the contents were of a religious nature
(whether heretical or otherwise), they sought out the bishop and vo-
luntarily handed them over. Here we have, in my opinion, another
detail of fundamental significance: when some barbarian highway
robbers in Tarraconensis realized the religious and magical contents

22
PLRE II, s.v.
23
Cf. Brent Shaw, “Bandits in the Roman Empire”, Past and Present 105 (1984)
pp. 3–52.
24
See C. Bertelli, “The Production and Distribution of Books in Late Antiquity”,
The Sixth Century. Production, Distribution and Demand, ed. R. Hodges and W. Bowden,
The Transformation of the Roman World 3 (Leiden 1998) pp. 41–60, here p. 39:
“Books are indicators of wealth in any time, but in late Antiquity they have a par-
ticular significance in this sense”.
      143

of their loot, they understood that they were holding something that
it was advisable to return and preferable not to traffic in or attempt
to take advantage of. It is as if the taboo nature of the volumes has
been disclosed to them, or if profaning their sacred nature the vol-
umes might be dangerous, and so it was simply better to get rid of
them. For this reason, they sought out Bishop Sagittius and handed
them over to him. This fear on the part of the barbarians of having
or keeping sacred texts and their willingness to renounce the oppor-
tunity to sell them once their contents were made known, are equally
meaningful and are probably indicative of their religious beliefs
(Arians, pagans, Catholics) and, at the same time, their ingeneousness.
However, there is something even more significant in this account
by Consentius. These barbarians are not only unknown, strange high-
way robbers and outlaws. These barbarians calmly enter cities with
absolute impunity. They walk in and talk to different people in order
to dispose of their loot (although perhaps they are unlikely to have
presented it as such) and make enquiries as regards the contents of
the same. They are not unfamiliar in the city; the inhabitants are
used to their presence (one pertinent question here might concern
what sort of clothing they were wearing: the answer is probably that
they could not be distinguished from Hispano-Romans on grounds
of dress), and the ecclesiastical authorities—the bishop—receives them
as if this were the most usual and normal event, although in this
case he took advantage of them and tricked them. In the same way,
the episode is evidence for these barbarians having recognized some
ecclesiastical authority.
Nevertheless, Consentius’ text is indicative of many other details.
It demonstrates that the Roman roads of the region—the road between
Osca and Ilerda and that from Ilerda to Tarraco, along which the
cursus publicus still moved absolutely normally, were still open to
traffic.25 It also illustrates the existence of castella—such as that belong-
ing to Severus’ mother—which cannot be anything other than fortified
villae.26 Likewise, it is evidence for urban life, in which one could
find members of the clergy and bishops, buyers and sellers, who
might deal in books or anything else; urban life in places such as

25
Cf. J. Arce, “La Penisola Iberica”, Storia di Roma, vol. 3,2: L’età tardoantica. I
luoghi e le culture, ed. A. Momigliano (Turin 1993) pp. 379–404.
26
Palladius describes the villa as a praesidium (Op. Agr. 1,7) in contrast with
Collumella (Re Rust. 5,2).
144  

Osca and Ilerda, cities of ancient origin, the importance and activ-
ity of which went back to the Republican period27 and which some-
what misleadingly appear as deserted cities lying in ruins in almost
contemporary poetical works of rhetorical nature, as is the case in
Ausonius.28 Lastly, it shows that churches and episcopal complexes,
with their archives, where Sagittius stored the part of the volumes
that he wanted to send to the bishop of Tarraco, existed; it is also
indicative of the circulation of manuscripts and books among culti-
vated and educated individuals, and the persistent deep-rooted pres-
ence of Priscillianism, which took root in this region in particular, as
is demonstrated by, for example, the Council of Caesaraugusta
of 380.29
These facts, and other events described in the correspondence
between Consentius and Augustine, are not indicative of anguish and
a feeling of impending doom among the communities and societies
to be found in 5th century Hispania as result of the barbarian pres-
ence but rather, as Frend has pointed out, a way of life whose every-
day round of activities was as far as might be imagined from the
barbarianism and the apocalyptic destruction described by Hydatius
or Orosius himself, a way of life that was concerned with a wide
range of subjects, among which were heresy, theological debate and
magic.
All this was true at least in Tarraconensis, which was still Roman
or under Roman administrative control.

Coexistence

In this overall context it is easier to explain the words of Orosius

concerning the fact that this period was short-lived even though there
were interludes of raids and devastation (he was referring to the
Goths who came to fight against the barbarians in the name of
Rome), and that Hispania was treated better by the barbarians than
by the Romans during the period of occupation, to the extent that,
within a short time, the barbarians started to lay down their weapons

27
On these cities in late Antiquity see Arce, El último siglo, pp. 86 ff.
28
Ausonius, Ep. 29,50–61 (Ilerda: deiectae ruinae).
29
Concilios visigóticos e hispano-romanos, ed. J. Vives (Barcelona-Madrid 1963).
      145

and to devote themselves to the plough, treating the Hispano-Romans

as companions and friends, “so that among them one could find
Romans who, living with the barbarians, preferred to live in poverty
in freedom to anxiously paying taxation among their own people”.30
This text, despite all its apparent rhetoric by means of which Orosius
attempted to endow the barbarians with virtues and qualities prefer-
able to those of the Romans (pagan, of course), is probably a rea-
sonable reflection of the situation that the inhabitants of the provinces
of the Iberian Peninsula were in before the barbarians arrived. There
were moments of tension, but coexistence soon came to be the norm,
as Consentius’ letter shows. The 411 agreement gave the barbarians
the opportunity to own lands to settle and work, but in all proba-
bility it did not mean, if we follow Orosius’ words, the existence of
a tax, a tertia that the inhabitants of the Peninsula had to pay the
newcomers,31 and this was a noteworthy difference between the case
of Hispania and the way in which foederati were settled in other parts
of the Empire. In this respect, facing this state of affairs, it is under-
standable that many would prefer to live and work alongside the
“barbarians” instead of being dependent on the totally suffocating
Roman fiscal system. Orosius’ text also implies a consequence: that
the Roman tax-collecting administration32 ceased to exist, which does
not, however, mean that the administrative system itself had disap-
peared although it might have had other functions. We do not know
how long this situation lasted. However, it is reasonable to state that,
provided the gentes that settled in a particular region made no use
of force and did not resort to destruction and pillaging, coexistence
among them all, for the peasant, for the tenant farmer, for the city-
dweller, even for the possessor of great estates and swathes of land,
could not have been particularly difficult. The problems arose for
two reasons in particular.

30
Orosius, 7,40,7. This text is echoed in Sidonius Apollinaris.
31
See, for the tertia, the comments of Goffart and Liebeschuetz in the works
quoted in note 1.
32
For the vivid illustration of the system see A. Giardina and F. Grelle, “La
tavola di Trinitapoli: una nueva costituzione di Valentiniano I”, Mélanges de l’École
Française de Rome 95 (1983) pp. 249–303.
146  

The unequal division of lands

These gentes, who made up different regna, continually and repeat-

edly confronted each other throughout the 5th century and, indi-
rectly at least, the inhabitants suffered the consequences of these
confrontations. The second reason is that, in the same way, and also
continually and repeatedly, the official Roman army, always making
use of the Goths, converted Hispania into the scenario of its harass-
ment of the newcomers. What the chronicles record is precisely and
exclusively the story of these punitive measures against Sueves or
Vandals or Alans, whether to establish order between these tribal
groupings, to settle their differences, or to keep their pretensions
under control, and to maintain a certain political and military pres-
tige through a policy of protection and supervision. The account of
these interventions is what Hydatius’ chronicle has handed down to
us (and as a result of a process of simplistic copying is what the sub-
sequent historiography has repeated). However, the most minute
detail of the chronicle should be examined in order to reveal some-
thing other than battles and war campaigns.
It is not surprising that these gentes—the Sueves, Vandals and
Alans—should confront each other. This was not due to their bel-
licose or warlike nature (a feature that is sometimes attributed to
them), but was the result of one simple fact: the unfortunate agree-
ment of territorial division that they made with one another. If one
takes a look at a map, it can be seen that the division was utterly
senseless and unrealistic. They were certainly not to blame, because
they were unaware of the geography of the Iberian Peninsula or that
part of the Iberian Peninsula that they were going to divide up.
They knew the administrative division, but not the territorial extent
of it. If the division is considered, now it is easy to see that the
Alans received the largest area of all—Lusitania and Carthaginensis
(fig. 2), almost double the area received by the other three groups—
the Sueves, the Siling Vandals and the Asding Vandals. The Siling
Vandals were alloted Baetica, also an extensive area, whereas the
Sueves and the Asding Vandals were confined to the relatively lim-
ited area of Gallaecia. The central tableland or Meseta and Lusitania
contituted an enormous space, full of possibilities and resources, with
countless Roman cities and villae of widely varying types and agri-
cultural worth. Since the beginnings of Roman settlement under the
Republic, Baetica had traditionally been considered the most fertile
      147

region, of great geographical extent, with enormous agricultural re-

sources, so it was intensely exploited. The most outstanding cities in
the Peninsula insofar as Roman civilisation was concerned were to be
found within its confines, and this had been the case for over four
hundred years. The other group of Vandals and the Sueves were
awarded Gallaecia. Hydatius’ text is not very clear as regards which
part was allocated to whom, but Tranoy has shown satisfactorily that
the westernmost part was for the Sueves and the easternmost for the
Asding Vandals.33 So, the least extensive lands and perhaps also the
least fertile from an agricultural point of view were the lot of these
two groups.
The division was carried out sorte, by lots. Although it is clear that
they had prior knowledge of the territorial limits of the administra-
tive divisions to be allocated (four provinces were available), none of
them knew exactly what possibilities the different regions offered or
how large they were, even though this distribution took place after
they had been living in the Peninsula for two years, presumably fol-
lowing a nomadic existence or inhabiting temporary settlements.
The Sueves and Vandals received an area in which the sea and the
Cantabrian and Atlantic sea routes played a fundamental role: the
Cantabrian ones because of the contacts they provided with a part
of Tarraconensis and with areas in north-western Gaul (Armorica, Brit-
tany) and even the British Isles; and the Atlantic ones because of
the routes following the coast of modern-day Portugal towards the
south and the Straits of Gibraltar. The Alans received the area where
the capital of the Diocesis since Diocletian, Emerita, was located.
Shortly afterwards, this city was to become a much sought-after
point on the part of their neighbours, the Sueves and the Van-
dals; however, they also had at their disposal, in the easternmost part
of the province, the city of Carthago Nova, a Mediterranean port
of vital strategic importance in order to control access to both Africa
and the Balearic Islands (another of the provinces that had re-
mained outside the regions to be divided among these tribes). The
Siling Vandals in Baetica had the possibility of Africa and Mauritania
Tingitana within easy reach.
In order to evaluate quite how unequal this division actually was,
it is essential to establish the number of individuals there were in

33
Hydatius, Chronicon, ed. A. Tranoy, vol. 2, Sources Chrétiennes 219 (Paris 1974)
p. 40.
148  

each of the tribal groupings to be settled, that is to say, how many

Alans, how many Sueves or how many Vandals were supposed to
occupy or fill the various areas. This is where we are faced with the
most uncertainty and are entirely dependent on speculation and the
realm of hypothesis.
E.A. Thompson calculated that the Sueves who entered the Iberian
Peninsula numbered some 200,000.34 There could not have been
more than 300,000 Asding and Siling Vandals (it has been estimated
that some 150,000 Siling Vandals later crossed into Africa)35 and it
might be speculated that there must have been a very small num-
ber of Alans, as they were practically annihilated in a couple of bat-
tles against the Vandals, Sueves and Romans.36 Consequently, the
Alans, the fewest in numerical terms, received the largest area, and
the others did not number enough to occupy or to exploit their own
territories completely or even simply to control them.

Superficial contacts and limited cultural interaction

The problem, once again, is to establish how each of these peoples

filled or occupied their territories; and also how they adapted to the
area, how they communicated in linguistic terms, what sort of con-
trol they exercised over the local population (who were far more
numerous),37 and also what material remains they might have left,
or, to put it in another way, what sort of cultural imprint they
bequeathed.
Prima facie, it can be stated that it is very difficult to believe that
their presence transformed the countryside or existing Roman struc-
tures. Apart from the fact that their numbers were limited, a fun-
damental fact that justifies this statement must be emphasized: with
the exception of the Sueves, these peoples remained in the Iberian

34
Thompson, Romans and Barbarians, pp. 158–9, a number considered by Liebeschuetz
“several times too high” in his article in this volume, p. 64 n. 39.
35
Victor of Vita, Historia persecutionis Africanae provinciae 1,2, ed. M. Petschenig,
CSEL 7 (Bonn 1881); cf. Procopius, Bellum Vandalorum 3,18, with the numbers dis-
cussed by Liebeschuetz in this volume.
36
Hydatius, Chronicon 55 (a. 417); 59–60 (a. 418).
37
A. Balil, “Aspectos sociales del Bajo Imperio en Hispania, IV–VI”, Latomus 24
(1965) pp. 886–904, estimated the total Hispano-Roman population at ca. 6–7 mil-
lions. Cf. also G. Ripoll, “The arrival of the Visigoths in Hispania: Population prob-
lems and the process of acculturation”, Strategies of Distinction, pp. 153–87.
      149

Peninsula for only a very short time. Reference has already been
made to the total annihilation of the Alans by the other tribes. The
Asding Vandals were absorbed by the Sueves and the Siling Vandals
left the Peninsula in 429 for Africa. Archaeological evidence for the
presence of these peoples in the Iberian Peninsula is almost non-
existent, even in the case of the Sueves, who were the ones who
lasted for the longest period of time.
With this in mind, the question is the following: what sort of mate-
rial culture can we expect of them, what sort of cultural imprint?
Weapons, some items of clothing, not much else. No buildings, no
specific objects. One might expect reforms or modifications in such
buildings as villae and churches. Everyday objects must have been
the Roman ones, as well as most items of clothing and adornments.
It is very difficult to trace distinctive identifying features. If they had
created pottery for usage by a closed group, this cannot be expected
to have been anything but an imitation of pre-existing productions.
Hydatius’ text concerning the reason behind the division of lands
leaves no room for doubt: this was done so that they settle, estab-
lish a home, ad inhabitandum. It is logical to think that settlement took
place in the countryside in a concentrated fashion with relatively
small distances between settlements, that they reoccupied villae or the
various rural sites scattered across their territories—vici, pagi and var-
ious other types of settlement sites.38 This would imply either that
the Roman inhabitants accepted the presence of such new tenants
in certain parts of their residences and estates, and thus, coexistence,
or alternatively that they occupied deserted areas that had already
been abandoned, or, as a final option, that they expelled inhabitants
so as to occupy their place. However, there is nothing to prevent
us from thinking that settlement also took place in the cities. Examples
of this can be found in other parts of the Empire (in the Vita S.
Severini, for example, for Noricum), and effectively a passage in Hydatius
seems to confirm that settlement also took place in the cities and
villae alike: Spani per civitates et castella a plagis barbarorum per provincias
dominantium se subiciunt servituti,39 that is to say, “the Hispani in the

38
G. Ripoll and J. Arce, “The Transformation and End of Roman villae in the
West (IV–VIIth centuries): Problems and Perspectives, Towns and their Territories between
Late Antiquity and the Early Middle Ages, ed. G.P. Brogliolo, N. Gauthier and N. Christie,
The Transformation of the Roman World 9 (Leiden 2000) pp. 63–114.
39
Hydatius, Chronicon 41.
150  

cities and in the villae who survived the disaster were subject to servi-
tude under the barbarians who had spread far and wide through-
out the provinces”. The text seems to indicate precisely that both
in civitates and castella, the inhabitants submitted to the coexistence
imposed by the newcomers. Several passages in Procopius concern-
ing Vandal settlement in Africa are very useful in comparative terms
as regards the possible settlement habits of the barbarian peoples
in the Iberian Peninsula and show that there was no single pattern
of settlement, but rather that settlement could take on different
forms.40
Can one talk of a total occupation of the territory? It has been
seen that this is more than unlikely in view of their limited num-
bers. It is therefore not unreasonable to think that large tracts of
lands or great urban centres or many villae continued to remain in
the hands of their owners or to have had a largely Hispano-Roman
population.
Finally, in view of these circumstances, what sort of supremacy or
governing authority did these gentes have over the existing Roman
administration? There can be no other reply than people collabo-
rated and accepted a fait accompli, once the local population real-
ized they were too far away from and had been abandoned by the
central administration. In the majority of cases, the upper ranks of
the aristocracy, the potentes and possessores, continued to retain their
prestige, authority and influence.41 Both societies were condemned
to understand one another and, for some time at least, the Roman
administrative machinery and organization must have continued to
operate. The barbarians had weapons and the power to intimidate,
but for some time they had recognized the usefulness and the
supremacy of the Romans and their laws.42

40
Procopius, Bellum Vandalorum 1,5,11; 1,17,8; 2,6,9.
41
This is also the conclusion of Liebeschuetz on the Vandals in Africa: “Gens
into regnum: The Vandals” (this volume).
42
See A. Chastagnol, “Les gouverneurs de la Byzance et de Tripolitaine”, Antiquités
africaines 1 (1967) pp. 119–34, here p. 132, where it is pointed out that this was
the case in many African provinces although “only the poverty of our sources for
the period is responsible for the vagueness of our knowledge”, and see also, once
again, the conclusions of Liebeschuetz concerning the situation of Vandals in Africa
in this volume.
      151

A century, a historian

As for the year 429, there were no longer any Vandals or Alans in
Hispania. There was only a group of Sueves, mixed with a few
Asding Vandals, who were isolated in the westernmost part of the
Iberian Peninsula, in Gallaecia. It should be borne in mind that the
Alans and Vandals therefore only spent 20 years in the Iberian
Peninsula. Forty years later, the Bishop of Chaves (in Gallaecia),
Hydatius, at the end of his life, and overflowing with instructive and
moralizing apocalyptic ideas, decided to write a Chronicon of the events
that had taken place in his by then long life.43 In general, he selected
the events he knew best, that is to say the history and activity of
the Sueves. And the history of the 5th century that we know, the
only written history of the period that we have, is what Hydatius
wrote. It is a factual, specific work, aimed at showing the comings
and goings of the Roman army, the rank and file usually consisting
of Gothic allies, controlled or commanded by Roman generals or
Visigothic kings, against the peoples settled in the Iberian Peninsula.
Sometimes these forces confronted the tribes, while at others they
sought to make them harass each other, or they aimed to reconquer
enclaves that the barbarian forces had taken. Hydatius’ account is
almost totally limited to such descriptions, with occasional entries
concerning the religious situation and associated disputes. Hydatius
was not concerned about the administration or the internal situation
of the provinces; neither was he interested in the economic system,
urban history, relations with other countries, power and ruling élites,
or society in general. Moreover, he writes for his monks with moral-
istic overtones, preparing them for the Last Judgment that he thought
was to come in a few years. The history of 5th century Hispania
cannot be written on the basis of Hydatius. At the most, what can
be described, and even then only somewhat superficially, is the polit-
ical history of the Sueves, their kings, their struggles for the throne,
the royal succession, and their wish to expand their kingdom. For
that reason, today we should not repeat Hydatius’ work as if it were

43
On Hydatius, see R.W. Burgess, Hydatius, Chronicon, pp. 3–11; S. Muhlberger,
The Fifth-century Chroniclers: Prosper, Hydatius and the Gallic Chronicle of 452 (Leeds 1990)
and J. Arce, “El catastrofismo de Hydacio y los camellos de la Gallaecia”, Los últi-
mos romanos en Lusitania, ed. A. Velázquez et al. (Mérida 1995) pp. 219–29.
152  

the history of the Iberian Peninsula in the 5th century (which is

what the few historians who have taken an interest in this period
have tended to do).
However, a few details that go beyond references to expeditions
and sacking can be extracted from Hydatius’ text. In the first place,
the fact that the communication routes created by the Romans in
Hispania were still in use. The Iberian Peninsula seems to have been
accessible from Gaul via all possible routes throughout the 5th cen-
tury: across the Pyrenean passes, along the Via Herculea, which
communicated the central provinces transversally, along the Via
Augusta, which ran through the Mediterranean region and forked
towards Baetica, running as far as Cadiz, and via what subsequently
came to be known as the Via de la Plata, which crossed Lusitania
from north to south. The Roman armies moved quickly in all direc-
tions just as the army of the Sueves did; the points of reference were
always the capital cities: Emerita, Barcino, Tarraco, Carthago Nova,
Hispalis, Corduba, Caesaraugusta, Pompaelo, Ulixipo, Coninbriga,
Bracara, Legio, Lucus, Asturica. These movements were constant,
which has its implications. There can be no cursus publicus without a
maintenance system and a tax system to sustain it.44 In spite of the
attempts of the Emperor Julian (361–363) to create a cursus fiscalis,
maintenance was paid for by the inhabitants of the provinces.45 This
was a harsh burden which is still reflected in the epistula of Honorius
to the troops of Pompaelo of 422.46
Neither was maritime communication interrupted. In 450 the
Emperor Majorian prepared an extensive fleet in the region of
Carthago Nova to sail to Africa and thus attack the Vandals. The
expedition was a failure because the Vandals captured the vessels
by surprise.47 We do not know how many ships Majorian requisi-
tioned or had built, but in order to carry out a reasonable-sized
expedition the number must have been considerable. They were
probably not exclusively warships, but he must also have made use

44
Legislation on the cursus publicus is to be found in Codex Theodosianus 8,5 ff.
45
Cf. Insciptiones Latinae Selectae 755 (a law of Julian).
46
On the subject see now J. Arce, “La epistula de Honorio a las tropas de
Pompaelo: comunicaciones, ejército y moneda en Hispania (s. IV–V d.C.)”, Rutas,
ciudades y moneda en Hispania, ed. P. Centeno, M.P. García y Bellido and G. Mora,
Anejos de Archivo Español de Arqueología 20 (Madrid 1999) pp. 461–88.
47
Hydatius, Chronicon 195.
      153

of transport vessels or other similar ships. Thus, the Vandals did not
totally destroy or make use of all the coastal fleet of the Spanish
provinces in 429 in order to cross to Africa.48 Both in 429 and in
460, there was a considerable and large enough fleet of vessels avail-
able on the Mediterranean coast of Hispania. Hydatius recounts that
in 456 vessels arrived in Hispalis bringing the news of the revolt of
the Lazi against Marcian.49 The Baetis continued to be navigable
and news often arrived by sea.50 Hydatius himself was not quite so
isolated as he, with false modesty, pretended, since he received and
sent pastoral letters and was well informed about events in Rome,
Constantinople and Gaul.51 In the same way, when the king of the
Sueves fled to Bracara, chased by Theoderic in 456, he took refuge
in Portus Cale, where he had gone to embark on a vessel to escape.52
And there is more information of the use of the sea as a means of
transport or communication in the course of the 5th century, such
as when 400 Heruls landed from seven ships on the coast near
Lucus;53 they subsequently moved on, in all probability also by sea,
to Hispalis54 or when Vandals vessels, which are likely to have sailed
from Africa, attacked the place called Turonio, in litore Gallaeciae, to
take hostages.
One salient aspect throughout the Chronicon is the importance of
cities. Emerita was a clear objective for the Suevic kings and the
impression is given that they wanted to make it the capital of their
kingdom as the King not infrequently resided there. Rechila, for
example, died in the city, and his successor, Rechiarius, was pro-
claimed there.55 Cities such as Barcino and Tarraco are sometimes
a refuge of Roman generals as they withdrew after an expedition
and also played a significant role,56 as did Hispalis, which started to

48
Cf. Courtois, Les Vandales, p. 160.
49
Hydatius, Chronicon 170: orientalium naves Hispalim venientes per Marciani exercitum
caesam nuntiant. Cf. ibid., vol. 2, pp. 106 ff. and Priscus of Panium, Fragmenta 25,
ed. and transl. R.C. Blockley, The Fragmentary Classicising Historians of the Later
Roman Empire 2 (Liverpool 1982).
50
Hydatius, Chronicon 97.
51
Ibid. 29; 31; 75; 97; 112 and passim.
52
Ibid. 168.
53
Ibid. 189.
54
Ibid.
55
Ibid. 129.
56
Ibid. 69; 52; 121.
154  

be seen as a city of strategic importance.57 The conquest of these

sites by the different armies did not imply their destruction; the oppo-
site is more likely to have been the case, in view of their potential
interest as royal capitals, precisely because of their infrastructure and
suitable urban features. Bracara Augusta, with its churches and
palatium, was probably a true sedes regia.58
In some sporadic cases, Hydatius enables us to infer that local
aristocracies, noble families and local Roman authorities—defensores
civitatis, duces—were in existence throughout the period covered by
his Chronicon.59 It is difficult to evaluate the significance of the terms
referring to destruction that Hydatius makes use of to describe the
Sueves’ military activity or that of the Goths in Roman service. In
the majority of cases they are generic terms and it is impossible for
them to have had the impact that has been attributed to them pre-
cisely because of the pessimistic and apocalyptic nature of his nar-
rative style and his moralizing purpose.

Barbarian against barbarian

Hydatius’ Chronicon always transmits one constant impression: the

continued interest on the part of Rome on the Iberian Peninsula,
its constant concern for controlling and recovering it, not so much
in order to eliminate the tribes that had settled there, but so as to
gain peace treaties. This explains why on numerous occasions, the
Chronicler refers to embassies, which moved in both directions to
achieve agreements and understanding.60 However, it is inevitable
that Hydatius’ text should likewise clearly convey the increasing inter-
est of the Goths, the principal players in expeditions and embassies,
in controlling the area themselves, without taking Rome into account,
and in settling there.61 This process culminated when Euric overran
the enclaves in Tarraconensis (which had hitherto been under Roman

57
Ibid. 115; 116; 170.
58
On Bracara, Hydatius, Chronicon 167 and Sedes regiae (ann. 400 –800), ed.
G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona 2000).
59
On the subject see Arce, “Los gobernadores”, pp. 73–83.
60
Hydatius, Chronicon passim.
61
For example Hydatius, Chronicon 239.
      155

control)62 and finally under his successor, Alaric, when the author of
the Chronica Caesaraugustana in a concise, but solemn, emphatic style,
noted under the year 494: Gothi in Hispanias ingressi sunt.63 Through-
out the 5th century, the Goths had gradually formed an idea of the
territory, its cities and its resources, and little by little, as a result of
their disagreements with Rome, they became aware of the possibil-
ity of occupying it entirely on their own as a permanent home for
settlement purposes. Their coming was a result of a struggle between
barbarian peoples. It was a long process because of the alliances
with Rome, but it was not difficult since neither the Sueves, nor the
Vandals, nor the Alans (with the possible exception of the Sueves)
managed to establish a strong, unitary administrative machinery or
a capital or definitive control over the territory assigned to them.
But not even the Visigothic occupation was definitive: this was only
to come about once the regnum tolosanum has been annihilated, and
after the Ostrogothic interlude. The Visigothic presence in Tarraco-
nensis—this was the first area occupied by the incoming Visigoths
in 494—was not welcomed and met resistance on the part of some
individuals faithful to Rome, to which the province still belonged.
The successive rebellions of Burdunelus (in 497) and of Petrus (in
506) recorded in the Chronica Caesaraugustana can only be understood
in terms of resistance against the Visigoths. The former’s uprising,
once he had been let down by his supporters, was soon crushed and
he was taken to Toulouse, where he met a horrible death. The lat-
ter’s revolt was also soon put down; but on this occasion, his head
was exhibited before the population in Caesaraugusta, a clear sign
that the city was by then under the control of the Visigoths; this
must have been an exemplary punishment for the benefit of those
who might still have disagreed with the Visigothic presence.
The number of Goths who arrived in this first phase remains un-
known. However, we do know that Roman customs continued to be
practised: for in the year 504, the Chronica Caesaraugustana records that
Caesaraugustae circus spectatus est. The 5th century had been one in
which Roman traditions were still current in Hispania.

62
Isidore of Seville, Historia Gothorum 34, ed. C. Rodríguez Alonso [Las historias
de los godos, vándalos y suevos de Isidore de Sevilla. Estudio, edición crítica y traducción] (León
1975).
63
Chronica Caesaraugustana ad a. 494.
156  

Conclusions

To summarize, as regards some of the historical problems dealt with

in the preceding pages, one might draw the following conclusions.
The case of Hispania in the 5th century A.D. is unusual and
somewhat different as far as the settlements of the barbarian peo-
ples within the Roman Empire is concerned. The initial phase of
settlement took place in two stages: first, in 409, as a result of a mil-
itary agreement between, on the one hand, the Sueves, Vandals and
Alans, and, on the other, the usurper Gerontius in the Diocesis. This
first episode cannot be interpreted as a true settlement of the different
tribes mentioned, but rather as an agreement of collaboration against
the aims of the usurper Constantine III. Secondly, in 411, there was
a distribution of the territory in order to facilitate definitive settle-
ment. This distribution was carried out by means of a pact with the
usurper Maximus. Four provinces were assigned to the various gentes
by means of a system of sortes, while the rest of the Peninsula remained
in Roman hands (first Maximus’ and subsequently those of Honorius).
This distribution was very unequal as regards the extent and resources
allocated when considered alongside the number and capacity of the
different tribes, which led to disagreements and wars between different
groups, which the Roman authorities took advantage of.
These gentes would seem to have arrived as cohesive, differentiated
groups with their reges at their head. The groups that settled were
true regna, although the evidence of Hydatius allows us to deduce
that the king of the Alans seems to have possessed a sort of supremacy
and maximum authority over at least the Vandal groupings. After
the Alans had been annihilated by the Roman forces, the Vandals
and Sueves continued to follow separate processes of settlement
although one can detect a certain rivalry and hostility between them
to gain control and to maintain their supremacy.
Only the Sueves managed to establish themselves in a stable fash-
ion for a long period of time as, once the Alans has been elimi-
nated, the Vandals soon crossed into Africa, leaving the area of the
Diocesis Hispaniarum behind. The short period of time (scarcely twenty-
five years) that the two main components forming the initial settle-
ments spent in the Peninsula allows us to think that their impact on
Roman society and structures was limited and without any visible
consequences either for political administration, or for cultural mat-
ters, or as regards economic organisation. The presence of the Goths
      157

in the Roman expeditions against the tribes settled in the Peninsula

meant that the former gradually prepared themselves for their own
process of settlement in the Iberian Peninsula, an event that occurred
at the end of the 5th century.
All the available evidence, in spite of the dramatic, doomsday tone
of Hydatius’ Chronicon points to coexistence between the Barbarians
and Hispano-Romans during this period and there are no signs of
the existence of a strong or centralized or organised state among the
gentes or the various regna during the 5th century.
The various western Roman Emperors constantly tried to regain
control of the occupied provinces, although they were more con-
cerned about reaching agreements than about really recovering ter-
ritory. In addition, throughout the period, the Romans retained the
provinces of Tarraconensis, Insulae Baleares and Mauritania Tingitana.
In brief, all the available evidence would seem to indicate that the
5th century was still a Roman period in the Iberian Peninsula.

[Translated by Philip Banks]

 158
 
Fig. 1. The Iberian Peninsula in the 5th Century A.D. (from Hydatius, Chronicon, ed. A. Tranoy, vol. 2
[Paris 1974] map 2).
     
159
Fig. 2. Military campaigns of Suevi and Visigoths in the 5th Century A.D. in Spain
(from Hydatius, Chronicon, ed. A. Tranoy, vol. 2 [Paris 1974] map 1).
This page intentionally left blank
 PRO PATRIAE GENTISQVE GOTHORVM STATV
(4TH COUNCIL OF TOLEDO, CANON 75, A. 633)1

Isabel Velázquez

Introduction

The study of the relationship between gens, rex and regnum, the main
topic of the Bellagio conference, has so many historical, political,
religious, social and cultural implications that it is necessary to
acknowledge the fact that the bibliography available can not be
encompassed.2 The case of Hispania and the Visigoths is particularly
remarkable due to the interest aroused by the long history of the
Visigothic monarchy from its arrival in Hispania after the battle of
Vouillé (Vogladum) in 507. That is why trying to tackle this question,
even superficially, sparks a certain degree of caution or apprehen-
sion, not only rhetorical. Thus, I am only going to deal with a num-
ber of points focusing on the scope of these terms and their possible
manifestations in 6th–7th-century Hispania, especially in the after-
math of the 3rd Council of Toledo, whose Proceedings are the first
to include the expression gens Gothorum.
It must be remarked, again on the issue of bibliography, that
modern historiography has been conditioned by several ideological
assumptions associated with the study of Germanic peoples, as some
scholars have pointed out.3 The same constraints, and others, are

1
This article is part of the research projects CAM 06/0050/00 and TEL 1999–395.
I am very grateful to Dr Gisela Ripoll for her comments and suggestions. I also
wish to thank Dr Francisco Rodríguez-Manas for translating the text into English
and Dr Ann Christys for revising it.
2
A large and significant part of this is mentioned in the Introduction to this vol-
ume. I refer the reader to it in order not to duplicate the information; this work
only includes quotations that are indispensable for its accurate documentation and
theoretical apparatus.
3
See some recent references to the development of studies on these topics and
various references in the fundamental studies of P.J. Heather, “The Creation of the
Visigoths”, The Visigoths. From the Migration Period to the Seventh Century. An Ethnographic
Perspective, ed. id. (San Marino 1999) pp. 1–72, here pp. 43–5; W. Pohl, Le origine
etniche dell’Europa. Barbari e Romani tra antichità e medioevo (Roma 2000), with a very
162  

also present in Spanish historiography.4 In fact, they have generated

a long-standing secular tradition that has always approached Hispania’s
Visigothic period as the genesis of present-day Spain: a tradition
worth studying both from the historical-political and from the liter-
ary points of view. Together with the survival of classical Roman
tradition, since the Middle Ages and with clearly ideological and
propagandistic manner since the start of Humanism and during the
Spanish Golden Age almost up to the present, there has been a need
to value and utilise Visigothic Hispania as a historical point of ref-
erence. It is well known that on more than one occasion attempts
have been made to identify the appearance of Spain as a modern
nation with that period. Likewise, Gothic monarchs have been regarded
as the first genuine Spanish monarchs in an uninterrupted contin-
uum.5 Furthermore, the 3rd Council of Toledo (589) has even been
considered as the act of foundation of the nation. This idea of unin-
terrupted continuity has pervaded literature and historiography for
centuries, even giving birth to cycles of medieval romances, such as
those that re-create the figures of kings Witiza and Rodrigo from a
literary standpoint. It has created a zeal for political and religious
continuity in the period of the first monarchs of León and Asturias;
it has had a tremendous influence on historians such as Jiménez de
Rada in the 13th century or humanists such as Ambrosio de Morales

interesting and insightful analysis of the ideology of the origins and new paradigms
in relation also to the situation in modern Europe, in pp. 1–16. For the key bib-
liography on the development of the studies, see again the Introduction to this vol-
ume by H.-W. Goetz.
4
As P. Díaz points out in his “Visigothic Political Institutions”, The Visigoths. From
the Migration Period to the Seventh Century. An Ethnographic Perspective, ed. P.J. Heather
(San Marino 1999) pp. 321–55, here p. 321: “[. . .] the historiographic debate over
the characterization of the Visigothic Kingdom as a state or as a form of state [. . .]
a topic certainly of interest, but which teaches us more about the history of the
nineteenth and twentieth centuries, or about the different historiographic currents,
than about Visigothic history”.
5
For those who are interested, there are several statues of Gothic kings in the
gardens of the Plaza de Oriente and in the Parque del Retiro in Madrid, part of
a collection of the kings from Athaulf to Ferdinand VI, commissioned by King
Charles III in the eighteenth century. Recently, J. Fontaine has published two pho-
tographs of the statues of Liuvigild and Swinthila remarking that: “Tous ces rois
portent le costume militaire et théâtral de rois de tragédie, le même boublier, sur
le même piédestal, comme si les passions soulevées en leur temps s’étaient apaisées,
dans le seule affirmation d’une continuité de la royauté Hispanique, d’Athaulf aux
Bourbons” ( J. Fontaine, Isidore de Seville. Genèse et originalité de la culture Hispanique au
temps des Wisigoths [Turnhout 2000] p. 146 and fig. 33b).
   163

and many others including, for instance, a figure such as Saavedra

Fajardo in the 17th century, author of a Corona gótica, castellana y aus-
triaca.6 Another underlying problem of double identification, not just
between Visigothic Hispania and modern Spain, but also between the
latter and the kingdom of Castille, is sometimes apparent. It would
be very interesting for studying these issues in depth to analyse, on
the one hand, to what degree the Visigothic historical past and its
“mythicising” have moulded later periods, to ascertain what ideo-
logical, political, literary or even aesthetic purposes the use of sources
from the Visigothic era and the constant hankering for its kings, its
arts and its writers have had; to try to determine whether that past
has had any direct effect on specific events in successive periods. On
the other hand, it would also be interesting to analyse the opposite
process: how has the Visigothic era been approached and publicised?
In what way have later styles, aesthetic and literary canons, histor-
ical events and political postulates influenced, conditioned and, finally,
shaped whatever real knowledge of Visigothic Hispania we could have
arrived at?

The question of literary sources

One thing is certain: that there exists a direct relationship between

reality and the way in which it is recounted and transmitted and
that, on occasions, one exerts a powerful influence on the other and
vice versa. That is why the approach I intend to adopt in studying
the topic is, in fact, to tackle those written sources—broadly described
as literary—that “talk” about Visigothic Hispania in the 6th–7th cen-
turies, that deal with the gens Gothorum and their kings as well as the
regnum they established. I do not attempt to cover them all, only
those that I consider most relevant, which constitute our main source
of information—as far as the texts are concerned—on that period
and on how the Visigoths’ presence was then perceived; whether
they could be distinguished from other inhabitants of Hispania, whether
assimilation of one group by the other took place—and to what

6
“Gothic, Castilian and Austrian Crown”. See R. Menéndez Pidal, Los godos y
la epopeya española. “Chansons de geste” y baladas nórdicas (2nd edn., Madrid 1969), here
pp. 30–1.
164  

extent and how this was achieved. The crucial issue is, most of all,
whether the information those sources supply reflect real events or
whether, on the contrary, they convey each chronicler’s perception
of his or her personal reality. Did their writings exert any influence
on political developments or on the denouement of events? Did their
particular interpretation of facts, objective or otherwise, serve to alter
situations? And, most importantly, can we know with a reasonable
degree of certainty what they mean when they employ terms such
as gens, regnum, patria, reges; is what they describe what really hap-
pened or what they wanted to have happened?
From this premise, which stems from a more philological than
historical approach but which through reading and interpreting the
sources inevitably leads, nonetheless, to tackling the historical and
even political issues required by the topic, I believe that I address
one of the problems highlighted by H.-W. Goetz in his Introduction.
It is an issue that I regard as fundamental, that is the issue of sources,
not only because they are written in Latin (and Greek), from the
Roman standpoint, but also because “this is not only a question of
criticism of our sources”, but of the need to take into account con-
temporary authors’ perception of events.
Indeed, it is not simply a question of philological criticism of the
sources but of something more fundamental because, without an
analysis of both the form and content of the sources, treating each
text in relation to others, it is not possible to interpret accurately
the reality they convey. Moreover, the literary genres and type of
documentation to which they belong must also be taken into con-
sideration. Council Proceedings and legislation, especially the Liber
Iudicum or Lex Visigothorum for the period we are studying, are texts
of a legislative nature written in juridical language, in many cases
with a long Roman tradition and using technical terms that are
already standardised, but showing the evolution of some terms that
reflect the contemporary reality about which they are legislating. On
occasions, also, some were penned by the author himself, as we will
see in the case of the Council of Toledo, or they were written fol-
lowing the models and literary references found in the great Christian
authors such as Augustine, Jerome and Gregory the Great and
included biblical quotations. These circumstances make these sources
unique, both due to the way they were written and the particular
language and specific terms used.
   165

Turning to the authors of the chronicles7—without doubt the other

fundamental evidence for the period—, one must take into account
the fact that Chronicae such as those by John of Biclar or Isidore of
Seville are also part of a long tradition of Christian literature to
which they are indebted, a debt that is explicitly acknowleded by
John of Biclar. The chroniclers situated themselves within this tra-
dition with the aim of continuing universal history up until their
own times, although contemporary reality and their own personal
interests and circumstances led to the development of a new type of
history which focussed in practice on Hispania and her kings and
crystallised in works such as Isidore of Seville’s Historia Gothorum (or
De origine Gothorum, the title given to it by Braulius of Saragossa)
which because of its content and ideology has been categorised as
“national (or nationalist) history”.8 This work is itself linked to sev-
eral Historiae on the origins and history of the Gothic peoples that
appeared in late antiquity which shifted attention, by different means,
away from the history of the Roman Empire, starting with the Getica
of Jordanes, Cassiodorus’ Historia Gothorum (De origine actibusque Getarum),
Gregory of Tours’ Historiae or the later Historia Langobardorum of Paul
the Deacon. Thus we find a work such as Julian of Toledo’s Liber
historiae Wambae devoted entirely to this king’s rule and, in particu-
lar, to his victory over the tyrannus Paulus, to which was added a
Iudicium and an Insultatio in tyrannidem Galliae, in which the concept
of the literary genre has changed significantly. Together with this

7
For overviews of the works on Hispania: M.C. Díaz y Díaz, “Scrittori della
Penisola Iberica”, Dal Concilio di Calcedonia (451) a Beda. I Padri Latini, ed. A. Di
Berardino, Patrologia 4 (Genova 1996) pp. 61–118, and U. Domínguez del Val,
Historia de la antigua literatura latina hispano-cristiana, 4 vols. (Madrid 1998); J.N. Hillgarth,
“Historiography in Visigothic Spain”, La storiografia altomedievale, Settimane di studio
del centro italiano di studi sull’alto medioevo 17 (Spoleto 1970) pp. 261–311.
8
M. Reydellet, “Les intentions idéologiques et politiques dans la Chronique
d’Isidore de Seville”, Mélanges d’Archéologie et d’Histoire 82 (1970) pp. 363–400, here
p. 363; J. Fontaine, Isidore de Seville et la culture classique dans l’Espagne wisigothique (Paris
1959) pp. 816–7 and 867–8; id., “Conversion et culture chez les Wisigoths d’Espagne”,
La conversione al cristianesimo nell’Europa dell’alto medioevo, Settimane di studio del cen-
tro italiano di studi sull’alto medioevo 14 (1967) pp. 87–147, here pp. 117–8;
S. Teillet, Des Goths à la nation Gothique. Les origines de l’idée de nation en Occident du V e
au VII e siècle (Paris 1984) p. 463. Against, Hillgarth, “Historiography in Visigothic
Spain”, pp. 298–9. For the edition of the work cf. C. Rodríguez Alonso in Isidore
of Seville, Historia Gothorum, ed. C. Rodríguez Alonso [Las historias de los godos, ván-
dalos y suevos de Isidoro de Sevilla. Estudio, edición crítica y traducción] (León 1975) [hence-
forth: Hist. Goth.].
166  

type of works, the biographical genre is exemplified by Isidore of

Seville’s De viris illustribus and, later, by a work with the same title
by Ildefonsus of Toledo which, emulating Jerome and Gennadius’
models, replaced ancient pagan biography with catalogues of men
who achieved eminence by virtue of their faith or their role in the
history of the Church, to the point where political aims or histori-
cal circumstances rather than edifying goals or literary influences dic-
tate the selection criteria of the viri illustres included.9
In the case of many of these authors, one must also consider the
relationship between the persona of the writer and the character of
his work. These authors were powerful both culturally and politi-
cally, almost all of them members of the highest echelons of the
Church hierarchy; bishops and often prior to this abbots of monas-
teries. They all represent the period’s intellectual class; they epito-
mise cultural circles, at least those of the cultivated minority.10 They
intervene directly as well in politics and many of their names are
linked to those of rulers: Leander of Seville to Liuvigild and his sons
Hermenegild and Reccared, Isidore of Seville to Gundemar, Sisebut,
Swinthila and Sisenand, Eugenius of Toledo and Braulius of Saragossa
to Chindaswinth and Recceswinth, Julian of Toledo to Wamba.11 In
their capacity as high clergy of the Church, they participate per-
sonally in the councils and in the drawing-up of some of the most
important canons, where political affairs are settled together with the
definition of concepts of growing significance that are gradually acquir-
ing a theoretical formulation, such as gens, patria and regnum. As has

9
Among other things, his defence of Toledo’s preeminence as metropolitan seat
and urbs regia and focus of his interest. See J. Fontaine, “El de viris illustribus de
Ildefonso de Toledo: tradición y originalidad”, Anales Toledanos 3 (1971) pp. 59–96;
Ildefonsus de Toledo, De viris illustribus, ed. C. Codoñer [El “de viris illustribus” de
Ildefonso de Toledo. Estudio y edición crítica] (Salamanca 1972). On Toledo, I. Velázquez
and G. Ripoll, “Toletum, la construcción de una urbs regia”, Sedes regiae (ann. 400–800),
ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona 2000) pp. 521–78.
10
See Fontaine, Isidore de Seville et la culture classique; id., Culture et spiritualité en
Espagne du VI e au VII e siècle (London 1986); M.C. Díaz y Díaz, De Isidoro al siglo XI.
Ocho estudios sobre la vida literaria peninsular (Barcelona 1976), here pp. 57–86 and pp.
89–115; C. Codoñer, “Literatura hispano-latina tardía”, Unidad y pluralidad en el
mundo antiguo. Actas VI Congreso Español de Estudios Clásicos 1 (Madrid 1983) pp. 435–65;
id., “La literatura”, Historia de España de Ramón Menéndez Pidal, Part 3,2: España visigoda
(Madrid 1991) pp. 209–67; I. Velázquez, “Ámbitos y ambientes culturales en la
Hispania visigoda. De Martín de Braga a Isidoro de Sevilla”, Studia Ephemeridis
Augustinianum 46 (1994) pp. 328–51.
11
Teillet, Des Goths à la nation Gothique, p. 543.
   167

been pointed out, their influence is felt even in the writing of the
7th century’s legislative work, the Liber Iudicum, for it is true that
there is mutual influence between Church and monarchy; there is
a certain degree of coexistence between both, a search for a power
balance, sometimes tipping in one or the other’s favour. Both pow-
ers coexist, however, and they restrain one another; the sources indi-
cate this in different ways and they must be interpreted according
to the information that the former provide and the manner in which
it is conveyed.12
One must take into account, likewise, how the text has been put
together13 and, when it entails writing about the past, how these
authors judge and report the origins of the Goths; how, from their
own reality, they create a history of former times for which they do
not possess scientific knowledge in disciplines such as anthropology,
ethnicity and so on, nor, on occasions, detailed information about
historical events or enough knowledge about the history of languages
and customs, relying instead on data inherited from Greek and Latin
Antiquity, which they garner, interpret and recreate in the light of
their religious beliefs both concerning the genesis of the world and
its different eras and the evaluation of the canons and cultural mod-
els of the past, and in the light of their clear and undisputed pas-
toral and educational vocation. They aim to write histories of times
and places both recent and further away that are expressed in a
credible way, that convey in the present a plausible image of what
the past must have been like.
Besides, as H.-W. Goetz recalls in his Introduction, it is well known
that our vision of that world is the product of authors belonging to
the Greek and Roman worlds. It is only from that perspective, there-
fore, that we can approach the period. Not only from the sources
“written in Latin”, although the author may have originated from
a different milieu as was the case with John of Biclar—of whom
Isidore of Seville remarks that he was a Goth14—, but above all,

12
Some discussion of this point in I. Velázquez, “Impronta religiosa en el desar-
rollo jurídico de la Hispania visigoda”, ‘Ilu. Revista de ciencias de las religiones 2 (1999)
pp. 97–121.
13
For remarks on what follows cf. Pohl, Origini etniche, pp. 16–38.
14
Isidore of Seville, De viris illustribus 31, ed. C. Codoñer Merino (Salamanca
1964) [henceforth: Isidore, De vir. ill.]: Iohannes Gerundensis ecclesiae episcopus, natione
Gothus, provinciae Lusitaniae Scallabi natus.
168  

from the perspective of Christian sources, so that the “social mem-

ory” of the past, ethnic identity, the categorisation of gentes and po-
puli, the nomenclature of duces and reges—especially in the earlier
periods—are all formulated in Latin or Greek and from the per-
spective—and with a deliberate uplifting purpose—that the synthe-
sis of peoples is only possible at a spiritual level15 and that it is the
universal Catholic faith that makes peoples equal.
Last but not least, is the significance of the vocabulary employed
by the texts. We must try to determine the scope of those terms
whose meaning in one or several contexts can often be ambiguous
or different from one author to another, exhibiting some nuances or
non-exclusive usages. It is useful to consider the implication of lan-
guage and reality and how one serves the other or how the one can
be shaped by the interests of the other.

The gens Gothorum and the other gentes at the 3rd Council of Toledo

The expression gens Gothorum in the 3rd Council of Toledo desig-

nates, according to Teillet,16 a community born out of religious con-
version, or in other words, from religious unity and thus refers both
to Hispano-Romans and Goths under a single name, Gothi. This
idea, which Teillet reiterates in her work, seems to me one of the
most disputable or susceptible to scrutiny. I believe, on the contrary,
that gens Gothorum represents here a human group which, led by their
king, has converted to Catholicism. The Council is convened by
royal mandate so that the bishops under his rule (regiminis sui ) can
express their joy towards their and their people’s conversion.17 With

15
Pohl, Origini etniche, p. 27.
16
Teillet, Des Goths à la nation Gothique, p. 450. I must point out that I will
specifically refer to this book on occasions as it deals in detail with the main topic
of the Bellagio meeting and, of course, in particular with that concerning the Goths.
17
III Toledo, p. 50,10–11: Ut tam de eius conversione quam de gentis Gothorum inno-
vatione in Domino exultarent. Until the 10th Council of Toledo I follow the edition by
G. Martínez Díez and F. Rodríguez, La Colección Canónica Hispana, 5 vols. (Madrid
1966–92), here vol. 5: Concilios hispanos: segunda parte (Madrid 1992). Quotations will
be provided with no. of council, canon (= c.) if appropriate, and no. of page and
lines. For other councils besides these, the edition by J. Vives, Concilios visigóticos e
hispano-romanos (Barcelona-Madrid 1963). On the history and explanation of con-
tents: J. Orlandis and D. Ramos-Lissón, Historia de los Concilios de la España romana
y visigoda (Pamplona 1986).
   169

the exception of this and other preliminary remarks18 reference to a

gens Gothorum appears only in the king’s various interventions. In his
words, this expression seems always to refer to his people, who have
abjured with him the Arian faith. At the Council, the people are
represented by the king, the Gothic church hierarchy, bishops and
priests and the nobles among the gens: primorum Gothicae gentis, who
sign the fidei professio. These groups represent the conversion of the
entire Gothic people, to whom their king refers to as gens inclita (the
same expression is used by Leander of Seville in his homilia) et fere
omnium gentium genuina virilitate opinata, that is to say, known among
all the other peoples by their courage, a stereotype, found in con-
temporary literature that traces back the origins of one of their main
to time immemorial.19
The expression gens Gothorum, likewise, is emphasised with respect
to another specific gens: the gens Suevorum, that is another human com-
munity that inhabits the Gallaecia and has been conquered by the
Visigoths. Their presence in the Acta of the Council is the result of
Reccared’s belief that the Sueves’ conversion to Catholicism is also
his accomplishment.20 Perhaps, Reccared intends to highlight here
his recent conquest of that territory and the submission of the peo-
ple who previously ruled them. This was his father’s, Liuvigild’s,
work, however, not his.21 Reccared, on the contrary, does refer to

18
See further on, those introduced by the words of the Catholic or Gothic bish-
ops. And in the “rúbrica” preceding the professio fidei, incorporated into the Proceedings
at a later stage. On the problematic of the “rúbricas” or summaries heading sev-
eral parts of the councils and canons, see Martínez Díez and Rodríguez in La
Colección Canónica Hispana, vol. 1, pp. 249–54; vol. 3, pp. 11–2, and vol. 5, pp. 16–7.
19
It was still a “cliché” nonetheless, adopted by the sources from Tacitus’ Germania
and later Jordanes’ Getica and other works, to allude to the strength, vigour, even
ferocity of the Germanic peoples. H. Messmer, Hispania-Idee und Gotenmythos (Zürich
1960). We will see it even more exaggerated in Isidore of Seville’s Historia Gothorum,
in particular in the final Recapitulatio.
20
III Toledo, p. 58,98–101: Nec enim sola Gotorum conversio ad cumulum nostrae mer-
cedis accessit, quin immo et Suevorum gentis infinita multitudo, quam praesidio caelesti nostro
regno subiecimus. When, apparently, it was like that from the time of Teodomir and,
according to Isidore of Seville, due to the pastoral work undertaken by Martin of
Braga, see Hist. Goth. 91 and De vir. ill. 22.
21
Hist. Goth. 49 and 91; John of Biclar, Chronica 585,2, ed. T. Mommsen, MGH
AA 11 (Berlin 1894). That does not mean that, in this particular case, Reccared
did not regard gens Suevorum as the native Sueves but, probably, all the inhabitants
of the Gallaecia. As we will see later, the “ethnic” origin of the peoples, or rather
the racial differences and those based on custom, language and other traits giving
shape to a gens, must have been somewhat ambiguous and vague at the time. For
170  

other gentes in the royal Tomus: when he points out that God has
placed the burden of the kingdom upon him so that he can ensure
the welfare and prosperity of the peoples under his rule (III Toledo,
p. 54,51–53 and 56–59): Quamvis Deus omnipotens pro utilitatibus popu-
lorum regni nos culmen sunire tribuerit et moderamen gentium non paucarum
regiae nostrae curae commiserit [. . .]. Reccared is aware that the higher
his royal dignity over his “subjects”, the higher his responsibility to
safeguard the welfare of the peoples God has entrusted to him: [. . .]
quanto subditorum gloria regali extollimur, tanto providi esse debemus in his
quae ad Deum sunt, vel nostram spem augere vel gentibus a Deo nobis creditis
consulere.
These words clearly show the king’s political project. The definition
of the regnum Gothorum22 under his rule does include all the commu-
nities and peoples ( populi et gentes)23 inhabiting Hispania (and Gallia
Narbonensis). His rule, which Reccared and his predecessors achieved
through settlement and conquest, is thus legitimised and turned into
a divine mandate by his uprooting of the Arian heresy and conver-
sion to Catholicism. This is his political programme and this is what
he intends to achieve. It is also the wish of the Catholic church, as
Leander of Seville’s homily shows at the conclusion of the council,
when all the royal interventions are over. A vibrant, emotional and
perfect piece of oratory that can be described as lyrical meditation,
overflowing with joy in face of the people’s faith (de eorum nunc gaudea-
mus credulitate) and referring to the conversion of the new communi-
ties as a significant gain (lucrum) for the Church.24 The term gens

Reccared, who is not the instigator of this conversion, what matters is that this
region of Hispania is under his control and authority and that her kings, in their
case of Suevic extraction, no longer rule there. Both Reccared’s—and earlier his
father Liuvigild’s—perception must have been the conquest of an entire territory
with all its gentes where the regnum Suevorum had been previously established.
22
In the restricted sense of the definition found in Isidore of Seville’s Etymologiae
9,3,1, ed. J. Oroz Reta and M.A. Marcos Casquero (Madrid 1982) [henceforth:
Etym.]: Regnum a regibus dictum. Nam sicut reges a regendo vocati, ita regnum a regibus.
23
Akin also to the classical definitions of gens and populus later formulated by
Isidore, including the difference between this term and plebs: Etym. 9,2,1: Gens est
multitudo ab uno principio orta, sive ab alia natione secundum propriam collectionem distincta
[. . .] Gentes autem appellata propter generationes familiarum, id est a gignendo, sicut natio a
nascendo; ibid. 9,4,5: Populus est humanae multitudinis coetus, iuris consensu et concordi com-
munione sociatus. Populus autem eo distat a plebibus, quod populus universi cives sunt, connu-
meratis senioribus civitatis.
24
J. Fontaine, “La homilía de San Leandro ante el III Concilio de Toledo:
Temática y forma”, Actas del XIV Centenario del Concilio III de Toledo, 589–1989 (Toledo
   171

appears several times in this Homily, although never associated with

the Gothi, to refer to the several peoples united in the faith—some-
times also called populi—resorting to biblical quotations that do not
identify any specific peoples (although they make indirect reference).
This is so because the ordo naturalis requires all men, who stem from
a single creature (ex uno homine trahunt originem), to love each other.
The whole Council calls for rejoicing because Christ has united all
peoples in a single church (Domus mea domus orationis vocabitur omnibus
gentibus). At the end of the homily Leander expresses a desideratum,
his own dream of witnessing with his own eyes that spiritual union
in the earthly kingdom (III Toledo, p. 159,253–257):
Superest autem ut unanimiter unum omnes regnum effecti tam pro stabilitate regni
terreni quam felicitate regni caelestis Deum precibus adeamus, ut regnum et gens,
quae Christus glorificavit in terris, glorificetur ab illum non solum in terris, sed
etiam in caelis. Amen.
But the king’s political project and the Church’s desire do not attain
unanimity at the Council. It can be said that both Reccared and
Leander’s interventions are polarised towards their respective inter-
ests. The bishop of Seville constantly evokes the unity of the faith
and joy at the recent conversion of the peoples whose roughness he
had very recently lamented (quorum asperitate quondam gemebamus). He
expresses, as well, his desire for the union of all peoples on earth,
as we have seen. Reccared, on his part, presents himself as a true
imperator, now, as well as rex, he will be known by the imperial title
of princeps (as will all his successors)25 and will put his signature to
the Proceedings of the Council using the name of Flavius Reccaredus
rex. His own conception of royal power is clearly enunciated. If he
is not the author of his own speeches, at least they must have been
written by someone very close to him and under his supervision. I
do not believe though, as has sometimes been claimed, that they

1991) pp. 249–70. An insightful analysis of Leander’s Homily compared to King

Reccared’s and some bishops’ interventions at the 3rd Council, based on study of
the different linguistic expressions, can be found in J. Mellado, “Veladas discrep-
ancias y pugna por el poder en el III Concilio de Toledo”, Boletín de la Real Academia
de Córdoba 141 (2001) (forthcoming).
25
During the Council described as gloriossisimus fidelissimus princeps, piissimus, fidelissimus
rex, catholicus, orthodoxus, gloriosissimus dominus noster, apart from vestra sanctitas, vestra
beatitudo. See M.C. Díaz y Díaz, “Titulaciones regias en la monarquía visigoda”,
Revista Portuguesa de Historia 16 (1978) pp. 133–41; id., “Más sobre epítetos regios
en la Hispania visigótica”, Studi medievali 19 (1978) pp. 318–33.
172  

were written by Leander.26 Although some of Gregory the Great’s

ideas about royalty may have reached Reccared via Leander, they
had no direct influence on the king nor is Leander behind the
monarch’s words. He is the rex Gothorum who now exerts his rule
over Hispania, Gallia and Gallaecia and he makes public profession of
faith together with his very noble gens,27 represented by the nobility
and the bishops and clergy before the hierarchy of the Hispano-
Roman Catholic church. He leaves no doubt though that he him-
self has achieved this religious conversion and that it is he who rules
over all the peoples and subjects within his regnum. As Díaz y Díaz
remarks, the king commands the assembly. With a surprising turn
of phrase (non incognitum reor esse vobis), he starts by saying that he
thinks the bishops are aware that they have been convened by the
Council to re-establish ecclesiastical discipline (ordo canonicus) and he
warns them, albeit curtly, by means of the expression admoneo pariter
et exhorto, to busy themselves with this task now that his uprooting
of heresy has made it possible. “The king seems unwilling to toler-
ate the idea that the bishops claim as their own achievement the
final liquidation of Arianism” among the Goths.
But the gens Gothorum, to which the king belongs, does not lump
together Goths and Hispano-Romans. In Reccared’s words, they are
clearly distinguished from each other. In fact two expressions that
seem to be set against each other encapsulate the ultimate purpose
of the Council. The Goths, protagonists of the conversion, of whom
it is said that they are ex haerese Ariana conversos, are counterposed to
the Catholic bishops (Hispano-Roman on the whole, although there
may have been some of Gothic extraction, such as Masona): Unus
episcoporum catholicorum ad episcopos et religiosos vel maiores natu ex haerese
Ariana conversos eiusmodi allocutiae exorsus est dicens [. . .].
Church and Gothic nobility are explicitly mentioned in relation
to their professions of faith:
III Toledo, p. 180,330–331: Tunc episcopi omnes una cum clericis suis pri-
moresque gentis Goticae pari consensione dixerunt [. . .].

26
On the royal interventions, M.C. Díaz y Díaz’s analysis is decisive (“Los dis-
cursos del rey Recaredo: El Tomus”, Actas del XIV Centenario del Concilio III de Toledo,
589–1989 [Toledo 1991] pp. 223–36).
27
To which he adds the Sueves; he uses has nobilissimas gentes to refer to both,
although the latter will no longer have any kind of specific representation.
   173

III Toledo, p. 202,593–594: Post confessionem igitur et subscriptionem omnium

episcoporum et totius gentis Goticae seniorum [. . .].28
One can raise the objection that the Hispani or Romani are not men-
tioned, something that might have been expected,29 in contrast with
the Gothi or Suevi, but the context30 is exclusively that of the con-
version of the last two peoples, in which the Hispano-Romans played
no role and where the opposing terms are catholicus and Arianus. We
have seen, however, that the term gens Gothorum is stressed in con-
trast to other gentes in king Reccared’s interventions.31
Hispania, however, does appear on numerous occasions, mostly in
its singular form, rather than the designation inherited from Roman
antiquity, provincia (later dioecesis) Hispaniarum. Once again, Teillet
emphasises this use of the singular, both in these Council proceed-
ings and in John of Biclar’s Chronica to refer to the territorial unity
of the regnum.32 Its use, however, is neither systematic—it appears in
the plural three times—nor exclusive. We must recall that at the
Council bishops ab totius Hispaniae vel Gallia have gathered. It is obvi-
ous that totius Hispaniae refers to the whole Peninsula. What should
it be called, however, if, on the one hand, that was its former name
and, on the other, the political and administrative situation that
justified its use in the plural had changed? I do not mean that an
awareness of territorial unity did not exist, besides being geograph-
ically undeniable, but that we must not overstate the use of the sin-
gular Hispania as indicative of, or as a proof of, such unity.33 On the

28
The latter are Reccared’s own words. Omnium refers here to all and each of
the members of the Arian Church (as in the above quotation), whereas totius implies
all the members of the gens of the Goths, represented by their primores.
29
Romanus is the term used to designate the Hispano-Romans and, earlier, the
Gallo-Romans, in the Leges Visigothorum. On the other hand, this term has under-
gone a change concerning the reality that designated historically as, in this period,
an author such as John of Biclar will use it to name Byzantine soldiers, from the
Eastern Roman Empire, enemies at the same time (milites, hostes). Besides, Hispani
(the ancient Iberi, according to Isidore) would not include all the gentes of Hispania,
populated by diverse other gentes, as this author indicates: Etym. 9,2,107–114.
30
That is, the verbal context and the extralinguistic context.
31
Which remain, on the contrary, deliberately undifferentiated in Leander’s homily.
32
Teillet, Des Goths à la nation Gothique, p. 451, although drawing attention to the
fact that the use was already well known by other authors such as Jordanes, Getica
44,229, ed. T. Mommsen, MGH AA 5,1 (Berlin 1882): universa Hispania; 44,230:
tota Hispania; 32,166: in suis finibus, id est Hispaniae solo, quoted by Teillet in p. 326.
33
In fact, this argument concerning the use of the singular, attaching to it more
importance than may be justified, is countered by the uses in the plural in later
texts, such as the 4th Council of Toledo.
174  

other hand, allusions to Gallaecia and Gallia as two distinct parts of

Hispania34 continue to be made; the first had been recently incorpo-
rated into the regnum whereas the second had been part of that reg-
num since the beginning, although belonging to a somewhat different
geographical environment and to a territory with a different history.
Nonetheless, it is not clear whether the expression totius Hispaniae
designates an entity that replaces, or is equivalent to, the plural
Hispaniae, in the sense of incorporating the diversity of areas, provinces
and subdivisions of which the concept already existed, or a usage
adopted by earlier literary sources or, even, of an attempt to re-
affirm a territorial unity35 that had not in fact been achieved, due
to the inconvenient fact of the Byzantine presence on the south-east
coast of the peninsula and in some key towns.36
It is obvious—it already seemed so much earlier and it was, fur-
thermore, one of Liuvigild’s clear political decisions—that the goal
was territorial unity, which would be achieved only after the expul-
sion of the Byzantines by Swinthila in 625. But the fact that the
expression used is Hispania (in the singular or the plural) at the 3rd
Council of Toledo does not mean that gens Gothorum refers to all its
inhabitants, that is, to the Hispani, as well as the Gothi. It also seems
clear that it is the gens Gothorum, or at least their nobility, who have
become the ruling class in Hispania. Although the constituent ele-
ments of power are not yet clearly defined, as they were to be at
the 4th Council of Toledo, as we shall see, these groups take the
leading role in the 3rd Council’s solemn ceremony: the king, the
nobility and the clergy. With “almost complete” territorial unity
achieved by Liuvigild, and religious unity secured by Reccared—the

34
And, in my opinion, maintained some idiosyncracies for a long time, especially
Gallaecia, at least from the ecclesiastical point of view. Also, the feeling of being an
“appendix” of Hispania experienced by some sectors of Gallia Narbonensis, which, as
J. Fontaine explains (Isidore de Seville. Genèse, p. 371), leads an author like Isidore of
Seville not to devote much attention to it and treat it as one of Hispania’s borders,
always threatened by the Franks, can be behind the frequent revolts and usurpa-
tion attempts—even segregation—which broke out in this area from this period,
and culminated in Paulus’ rebellion against Wamba.
35
See John of Biclar, Chronica 372,2 (after Liuva’s demise): Hispania omnis Galliaque
Narbonensis in regno et potestate Liuvigildi.
36
On this see, in general, M. Vallejo Girvés, Bizancio y la España tardoantigua (ss.
V–VIII): Un capítulo de historia mediterránea (Alcalá de Henares 1993). Specifically,
G. Ripoll, “Acerca de la supuesta frontera entre el regnum Visigothorum y la Hispania
bizantina”, Pyrenae 27 (1996) pp. 251–67.
   175

latter a precondition for the levelling-out of the differences between

the populi (and gentes) dwelling in the regnum—the foundations are
being laid of a new society, or, if you wish, of a new natio.

A writer’s perspective: John of Biclar

It is the writers, in their capacity as intellectuals and leading mem-

bers of the Church, and of the nobility, who contribute decisively
to define and present a new image of that new society and, most
probably, also to build it. John of Biclar’s Chronica—of which Teillet
rightly says that it must be read in the light of the Proceedings of
the 3rd Council of Toledo—presents an accurate image of that new
society and one can detect in the work the basis of the mentality
that is being forged. Although I believe that here, as in the afore-
mentioned Council, gens Gothica (or Gothorum) cannot be identified
with the totality of the Gothi and Hispani or, for that matter, that
provincia Gothorum37—i.e. the territory under their rule—can be equated
with Spain “in the modern and almost national sense of the word”.38
The Chronica is written, at first sight, as a sequel to that of Victor
of Tunnuna (Huc usque Victor Tunnunensis [. . .] nos quae consecuta sunt
adicere curavimus) and is characteristic in form and content-based of
the genre of universal and Christian history,39 but follows a unique
dating system based on the years of emperors’ reigns and, starting
with the third heading, with the equivalent Gothic regnal year using
the formula: anno IV Iustini Imp. Qui est Liuvigildi regis secundus annus.
The focus of attention is, however, the Goths’ conversion to Catholicism

37
This expression is used in John of Biclar, Chronica 569,4, when talking for the
first time about Liuvigild to highlight that he succeeded in taking it back to its orig-
inal limits, previously reduced due to revolts: quae iam pro rebellione diversorum fuerat
diminuta, mirabiliter ad pristinos terminos revocat. The term provincia designates some ter-
rae, of very different sizes and history in each case: Hispania, Gallia, Italia, but also
Sabaria, Cantabria, Orospeda, etc. See J. Campos, Juan de Biclar. Obispo de Gerona, su
vida y su obra (Madrid 1960) p. 158.
38
Teillet, Des Goths à la nation Gothique, p. 440. It seems to me that this sense is
a bit “stretched” and, most of all, in contradiction with what is stated immediately
after: “On condition of bearing always in mind that the ‘province of the Goths’ is
that of the king of the Goths and his regnum, the idea of nation is inextricably linked
to that of regnum”. But that is precisely the question; what prevails is the idea of
territory dominated by the rex and his gens.
39
Hillgarth, “Historiography in Visigothic Spain”; Teillet, Des Goths à la nation
Gothique, pp. 421–49.
176  

as well as the celebration of the Council. One must bear in mind

that it was written shortly afterwards, around 590/91, but at a time
when the early effects of such a gathering were already being felt.40
The passage Chron. 587,5 notes, as can be inferred from Reccared’s
intervention at the Council, that the process of conversion was already
under way or completed, thanks to the king, during his first year on
the throne: sacerdotes sectae Arianae sapienti colloquio aggressus ratione potius
quam imperio converti ad catholicam fidem facit gentemque omnium Gothorum
et Suevorum ad unitatem et pacem revocat Christianae ecclesiae. The expres-
sions sapienti colloquio and ratione potius quam imperio show the phase of
deliberation and negotiation and of the theological debates that must
have taken place between Arians and Catholics culminating in the
official event of 589:41 debates that were already under way during
Liuvigild’s rule, such as those described in Vitas Sanctorum Patrum
Emeretensium between Masona and Sunna, part of a policy of reli-
gious unification under the Arian faith carried out, albeit unsuc-
cessfully, by Liuvigild.42 Here the achievement of having converted
the Sueves to Catholicism is attributed to Reccared.
The long passage devoted to the 3rd Council of Toledo—much
longer than would be expected in a work of this kind—and the news
about the Visigoths reveal that the author’s genuine interest, with-
out neglecting his aim of writing a Christian universal chronicle, is
to chronicle the reign of Liuvigild, the monarch who has ensured
territorial unity43 or helped the provincia Gothorum consolidate its size-
able domains, as well as that of Reccared, whose main accomplish-
ment has been the religious unification of the kingdom that reaches
its zenith in the Council. Although the work starts by relating the
events of Justin’s reign, the amount of information about the Visigoths
gradually increases. Moreover, the first time they are mentioned is
in the context of Liuvigild’s rule, when Athanagild’s death in Hispania
and Liuva’s accession to the throne (Chron. 568,3) are recorded; the
second reference is linked to this; Liuva involves his brother Liuvigild

40
Actually, the last point of the chronicle records Argimundo’s seditious attempt.
41
Note, however, the term revocat to describe the peoples, led with a clear sense
of mandate (imperium).
42
John of Biclar, Chronica 580,2; Vitas Sanctorum Patrum Emeritensium 5,5,7–9, ed.
A. Maya Sánchez, Corpus Christianorum Series Latina 116 (Turnhout 1992).
43
As indicated earlier, not yet achieved due to the Byzantine occupation; but
this fact is played down or, rather, avoided because the work states that Liuvigild
completely defeats them.
   177

in the running of the kingdom and the latter, on his part, marries
Athanagild’s widow and undertakes his first conquests, re-establish-
ing the original boundaries of the vanquished territory. The work
is, therefore, the product of his period and his personal circum-
stances. The author intends to pursue the literary model set by the
chronicles, but his work resembles a Historia Gothorum, in this case of
these two kings. It is not simply a matter of the Empire’s no longer
being the centre of the world, nor that it is considered a regnum along
with other regna44—in fact an enemy due to its presence in the
Peninsula—but that now the local, living history that the author
wishes to recount is that of the Hispanic community, of the regnum
Gothorum to which the author belongs, not just as a mere native but
also as a member of the Gothic nobility and prominent member of
the Catholic church.
It seems likely to me that this work could have been written at
the request, explicit or otherwise, of Reccared himself. Who was in
a better position than the author from Biclar to relate the deeds of
his king and his father? In fact, his vision of Liuvigild is so positive
that not even the terms in which he describes Liuvigild’s attempt to
convert Catholics to Arianism are excessively harsh: novello errore, seduc-
tionem, a seduction to which many Catholics ( plurimi nostrorum, the
only personal reference) succumb. Nothing is said, however, on his
personal exile or that suffered by others who, like him, refused to
become Arians.
John of Biclar shows Reccared as a new Constantine or Marcianus,
whose merits he surpasses in achieving the uprooting of heresy, and
he compares the Council of Toledo to those of Nicae and Chalcedon;
he even places it before them. It is this religious unity that under-
lies the definition of the gens Gothorum, but it runs in parallel with a
designation that stresses ethnic origin, linked to the past and the
origins of that people which expresses, in my view, an awareness
of belonging to the same group and of being part of the ruling
class. One might think that the author—let us not forget his Gothic

44
Perhaps, the dating system maintains the subordination of dates through quid
est just to record the initial aim of writing a universal history with such subordi-
nation not being actually conceptual. Note the strange formulation of 583: anno ergo
I Mauricii imperatoris, Liuvigildi regis XV annus, and in particular of 587, with a restric-
tion to the princeps and the mention of feliciter in the second sentence: anno V Mauricii
principis Romanorum qui est Reccaredi regis primus feliciter annus.
178  

origins—when speaking about gens Gothica, is referring to all the inhab-

itants of Hispania (more properly, of the areas under the rule of the
Goths: Hispania, Gallia et Gallaecia), basically Gothi and Hispani, as
Teillet claims. It seems to me, however, that, as in the Acta of the
Council, it refers only to the group that was under consideration at
that moment, in contradistinction to the Hispani. Only secondarily
and in their capacity as inhabitants and subjects of the territories
and of the king could they be thought of as such. Perhaps the expres-
sion gentemque omnium Gothorum et Suevorum ad unitatem et pacem revocat
christianae ecclesiae suggests indirectly that the union of all the com-
munity is in the mind—and the wishes—of John of Biclar.45 Com-
parisons with other usages of gens or with other expressions found
in the work do not make it possible, however, to state this categori-
cally. The same happens in the case of the singular use of Hispania
as I have previously argued.
On one hand, the gens Gothorum (or Gothica) is mentioned, as in
the Council, as the people who, together with the gens Suevorum, leads
Reccared to Catholicism. This excludes the Hispani.46 On the other,
many other communities are mentioned which could comprise pop-
ulation groups, such as the Langobards or Franks, but the passages
where the latter are referred to appear in a context of war. This
involves talking about armies made up of, or at least led by, mem-
bers of those gentes; on occasions, even troops belonging to the impe-
rial army with no identification whatsoever with elements of the
latter, as in Chron. 575: Iustinianus [. . .] habens secum gentes fortissimas
quae barbaro sermone Herinan nuncupatur. Or, as in the case of the gens
Suanorum, reference is made to their rex, captured alive by the Romans,
together with his treasure and children and taken to Constantinople,
thus turning this province (with all its inhabitants, of course) into a
Roman dominion; the Suevi are shown in a similar light when Liuvigild
subdues them.
But it is actually this passage dealing with the Sueves where we
can find one of the key elements of the work, which anticipates the

45
As I have previously commented, I also regard it as artificial to draw from
omnium an attempt to reflect the entire population, as well as tota Hispania. In the
first case, note also the use of omnis with sacerdotes in the passage of John of Biclar,
Chronica 590,1: Reccaredum ordinem conversionis suae et omnium sacerdotum vel gentis Gothicae
professionem tomo scriptam manu sua episcopis corrigens [. . .] decrevit.
46
Consider, though, the various nuances I have used in the discussion.
   179

formulation of the 4th Council of Toledo that we will study later:

Audeca regno privat Suevorum gentem, thesaurum et patriam in suam redigit
potestatem et Gothorum provinciam facit. The bases of the power of the
new state are established here: rex, gens et patria.47 Here patria seems
to refer to the territory where one dwells, where one comes from,
but also to all the territory, which is ruled and constitutes the reg-
num. When Liuvigild establishes his authority over this territory, it
includes all its inhabitants, but I do not believe that the gens Suevorum
deliberately includes the Gallaeci, except as a consequence of the mere
and indisputable presence of this group within the area the king now
turns into the Gothorum provincia. We can not dispute the fact, nonethe-
less, that this formulation can lead to identifying the population as
a whole as Gothi, as Teillet claims,48 however inchoate this process
might be. There are some indirect references in the work, however,
to these population groups from Hispania, always linked to the resis-
tance and struggle against Liuvigild’s conquest, a fact that must be
borne in mind. Thus, when speaking about Cordoba (Chron. 572,2),
he indicates that, once the enemies (caesis hostibus), that is the Byzantines,
have been vanquished, multasque urbes et castella interfecta rusticorum mul-
titudine in Gothorum dominium revocat. Of the “province” of Orospeda
and its civitates and castella, he comments (Chron. 577,2) that: non multo
post inibi rustici rebellantes a Gothis opprimuntur. And when he points out
that Liuvigild has managed to quell all the attempted usurpations of
his throne and to defeat Spain’s enemies (Chron. 578,4: pervasoribus
Hispaniae superatis), I do not think that this refers exclusively to the
Byzantines (= Romani ), but non-specifically to any group that has
rebelled against him such as for instance in Chron. 574,2: Cantabria
provinciae pervasores interfecit. It is even said that Liuvigild seeks tran-
quillity among the people and founds Reccopolis, a town to whose
inhabitants he grants special privileges (Chron. 578,4): requiem propriam
cum plebe resedit, civitatem [. . .] condidit quae Recopolis nuncupatur [. . .]
privilegia populo novae urbis instituit. These uses, scarce and unclear, of

47
And also the treasure, clear symbol of the monarchy which has also been men-
tioned when tackling the Sueves.
48
Teillet, Des Goths à la nation Gothique, p. 436, by virtue of a process according
to which in all the countries the new nation is made up mostly of natives but, polit-
ically, the state is assimilated into the invaders so that it loses its own name, fol-
lowing F. Lot, “La formation de la nation française”, Revue des deux mondes 15 mai
et 1er juin (1950) pp. 256–78; 418–35, here p. 260.
180  

rusticus, populus and plebs in different contexts, indicate the presence

of the common people, possibly both Goths and Hispano-Romans
(and perhaps other minorities), but still beyond the concept and
definition of gens Gothorum.
On the other hand, the opposition between Gothi and Romani
(= Byzantine troops) is usually associated with warfare, which I believe
has a different connotation susceptible to nuance. As I have remarked,
according to Teillet, the Hispano-Romans are never mentioned, but
subsumed, “for lack of a better description” under the generic term
Gothi; the author also claims, however, that the Romani are the
Byzantines and their Catholic supporters in Spain, tacitly branded
as “rebels”, in the expression pro rebellione diversorum. These diversi could
be the Byzantines and Hispano-Romans. There is some truth in this.
It is reasonable to think that the army had military contingents made
up of Hispano-Romans and not just common soldiers, but we also
know of military chiefs, in particular the dux Claudius, of Roman
extraction; he is much commended in the sources as the man who
put down the revolts of Mérida and Septimania. It seems, however,
that the army was mainly controlled by people of Gothic extraction;
military force was, in particular, one of the Visigothic monarchs’
most effective means of exercising power. Once again, it is the Gothi
who are the dominant group, the present rulers of Spain.
In my view, the issue is that Hispano-Romans “do not count” in
this history—nor did they in Reccared’s mind during his speeches
at the 3rd Council of Toledo—except, perhaps in a secondary role.
At the very least, they are not considered to be protagonists in the
events related in the Chronica, that is the power (imperium) gained and
exerted by the Visigothic monarchy over their subject territories
( provincia Gothorum = omnis Hispania, Gallaecia, Gallia) and, above all,
the emergence of a renewed, Catholic, gens Gothorum.
One of the work’s most significant political and ideological ele-
ments is that these conquests and rule over the territories of Hispania
are implicitly shown as legitimate events, using a vocabulary remi-
niscent of juridical language. For example:
Chron. 571,3: Liuvigildus rex Asidoniam [. . .] ad Gothorum revocat iura.
Chron. 572,3: (on Cordoba and other towns) in Gothorum dominium revocat.49

49
In similar terms to speak about the conquest of the gens Suanorum by the Romans.
John of Biclar, Chronica 576,2: et provinciam eius in Romanorum dominium redigit. See, in
   181

Chron. 573,3: Sabariam [. . .] provinciam ipsam in suam redigit dicionem.

Chron. 574,2: Cantabriam [. . .] Amaiam et provinciam in suam revocat dicionem.
Chron. 572,2: (on the Aregenses mountains) et loca ipsam in suam redigit
potestatem.
The monarchy and the Goths (perhaps, most of all, their nobility)
are clearly shown as the ruling forces in the kingdom. But they are
not the only ones; the Church plays a decisive role. Religious unity
brings with it Visigothic access to the Church hierarchy50 and, as
we have seen at the Council, the important participation of the
Catholic church in the apparatus of power. It can be stated that
one of the most obvious preconditions for the union and levelling-
out of differences between Goths and Hispano-Romans, vis a vis the
ruling class, is with the institutionalisation of a single ecclesiastical
community.

Population mix

At the end of the sixth century, when John of Biclar writes his work,
society had already experienced a certain degree of assimilation and
balance. They had been co-existing for more than eighty years and
the prohibition on mixed marriages had been lifted (Lex Visigothorum
3,1,1); this contributed to population mixing, as well as highlighting
a situation that had existed, in fact, much earlier and had affected
even the kings, as in the case of Theudis, an Ostrogoth married to
a Hispanic woman (Procopius, De bello Gothorum 1,12,50).
The available documentary sources, especially the slate tablets writ-
ten in the Castilian meseta,51 show onomastically (although not con-
clusively), a mix of population among the peasants, rustici. The
Hispano-Roman aristocracy gradually started to mix with the Gothic
nobility, in addition to the crucial role played by ecclesiastical unity,
which we have already mentioned. Other distinguishing elements,

general, E.A. Thompson, Los godos en España (3rd edn., Madrid 1990); L.A. García
Moreno, Historia de España visigoda (Madrid 1989); G. Ripoll and I. Velázquez, La
Hispania visigoda. Del rey Ataúlfo a Don Rodrigo (Madrid 1995).
50
The canons show that the re-consecration of Arian churches will not be nec-
essary and that the bishops will retain their posts and seats, now as Catholics.
51
Documentos de época visigoda escritos en pizarra, ed. I. Velázquez, 2 vols. (Turnhout
2000).
182  

such as land tenure, that granted a number of privileges to the Goths

as the law shows, would not have affected the entire population, nor
have been implemented in a systematic way.52 It is clear that, for
some time, differences in customs, fashion and social organisation,
must have existed, as is shown by the remains of material culture
that have been preserved. They also prove, however, that those
differences soon began progressively to fade,53 because of the intense
process of acculturation and Romanisation, already long-established,
of the Goths, which started before their arrival in Spain and dur-
ing the period of the regnum Tolosanum.
Moreover, there are no data suggesting that, from the settlement
of Hispania and, most likely before, a “Gothic” language was used.
This would have been a clearly segregating element. Latin was used
to legislate—both for Goths and Romans, as well as for other minor-
ity groups dwelling in Hispania54—and, if the Goths gradually began
to have access to schools (a very rare phenomenon at the beginning,
most likely) and culture, it was no doubt to Latin culture and liter-
ature; this was true of John of Biclar (although he also studied Greek),
Masona, bishop of Mérida and his rival Sunna and the judges and
magistrates of the officium palatinum.55 The reference in Leander of

52
I will only recall some questions, extensively covered by modern bibliography,
without dealing with them in depth but, rather, as a backdrop to the argumenta-
tion.
53
I refer here, in general, to some of the key studies by G. Ripoll, with earlier
bibliography: La necrópolis visigoda de El Carpio del Tajo (Toledo) (Madrid 1985); La ocu-
pación visigoda a través de sus necrópolis (Hispania), Thesis Microfiche 912 (Barcelona
1991); “The arrival of the Visigoths in Hispania: Population problems and the
process of Acculturation, Strategies of Distinction. The Construction of Ethnic Communities,
300–800, ed. W. Pohl with H. Reimitz, The Transformation of the Roman World
2 (Leiden-New York-Köln 1998) pp. 153–87; Toréutica de la Bética (siglos VI–VII d.C.)
(Barcelona 1998); “Symbolic Life and Signs of Identity in Visigothic Times”, The
Visigoths. From the Migration Period to the Seventh Century. An Ethnographic Perspective, ed.
P.J. Heather (San Marino 1999) pp. 403–31.
54
Without going into the issue of the personality or territoriality of the first legal
Codes. On this, with previous bibliography, J. Alvarado Planas, El problema del ger-
manismo en el derecho español (Madrid 1997). Some remarks in I. Velázquez, “Jural
relations as an indicator of Syncretism from the Law of Inheritance to the Dum
inlicita of Chindaswinth”, The Visigoths. From the Migration Period to the Seventh Century.
An Ethnographic Perspective, ed. P.J. Heather (San Marino 1999) pp. 225–58.
55
On all these issues, with some nuanced points of view, somewhat different with
regard to the level of assimilation and levelling-out of differences between Goths
and Hispano-Romans in Hispania, see in particular the introduction of M.C. Díaz
y Díaz to Isidore of Seville, Etymologiae, ed. J. Oroz Reta and M.A. Marcos Casquero,
2 vols. (Madrid 1982) pp. 7–257, here pp. 41–3 and 50–94 [henceforth: Díaz y
Díaz, “Introducción”].
   183

Seville’s Homily at the 3rd Council of Toledo to the linguarum diver-

sitas as the cause, together with superbia, of the emergence of here-
sies, was an evocation of the biblical story of the Tower of Babel,
much used by Christian authors,56 rather than to contemporary lin-
guistic reality. It is possible that some groups still spoke a tongue of
Gothic origin,57 alongside Latin, but the truth is that they must have
been quite fluent in the language then spoken by Hispano-Roman
society. It is more likely that some lexical elements were preserved
and incorporated into Latin as neologisms, exerting influence on it
as parts of a linguistic superstratum, as a result not only of the nat-
ural expansion of onomastics, but also as terms in a common lexi-
con. Germanic loanwords adopted by Latin, however, can also have
penetrated—that is actually what happened—by different means and
during long periods of time, from the early contacts between com-
munities. It is therefore very difficult to assess each loanword accu-
rately. Military or judicial terms, such as saio, thiufadus, gardingus and
others used in legislative and documentary Latin texts, are extremely
well known, as are many terms derived from Germanic languages
that have been preserved in Romance languages. They cannot be
used, however, to show clearly the degree of dynamism and pro-
ductivity of a less specific lexical corpus, that may have penetrated
Latin during the 6th and 7th centuries, without having been previ-
ously incorporated into the Latin language in earlier periods.58
Nevertheless, a few terms may serve as points of reference, some
of them first documented and explained by Isidore of Seville in his
Etymologiae. Quite apart from their specific origin, the fact that Isidore

56
A. Ferreiro, “Linguarum diversitate: Babel and Pentecost in Leander’s Homily at
the third Council of Toledo”, Actas del XIV Centenario del Concilio de Toledo, 589–1989
(Toledo 1991) pp. 237–48. And by Leander’s brother, Etym. 9,1,1: Linguarum diver-
sitas exorta est in aedificatione turris post diluvium. Nam priusquam superbia turris ilius in diver-
sos signorum sonos humanam divideret societatem, una omnium nationum lingua fuit, quae Hebraea
vocatur [. . .] Initio autem quot gentes, tot linguae fuerunt, deinde plures gentes quam linguae;
quia ex una lingua multae gentes exortae (see Augustine, De civitate Dei 16,4, ed. J. Divjack,
CSEL 98 [Augustinus, Opera] [Vienna 1981] and Genesis 11,1–9).
57
Or several non-Latin languages, for as has comprehensively been studied, the
make-up of what can be understood as gens Gothica had a very variegated forma-
tion process and it was not homogeneous.
58
For a synthesis of the numerous studies on the lexicography of Germanic ori-
gin in Spanish, see R. Lapesa, Historia de la Lengua española (8th edn., Madrid 1985)
pp. 118–28. As for proper names, above all, J. Piel and D. Kremer, Hispanogotisches
Namenbuch (Heidelberg 1976).
184  

was the first author to provide documentary evidence for them may
indicate that certain words penetrated Hispania (or earlier Gallia, even
in the Ostrogothic period of the regnum) through the Visigoths and
were incorporated into Latin during this period. We could mention,59
for instance: blavum, as a colour for a costume (Etym. 19,28,8); guara-
nen, as a “bronze”-coloured horse, also known as aeranen. According
to Isidore, it is the cervinus (the colour of a lion) to which the com-
mon people give this name (Spanish “garañón”). Flasca (Etym. 20,6,2),60
in spite of the fact that the author believes that it is a Greek word,
meaning “covering for goblets” and also “carafe or bottle”, related
to flasco, documented in Gregory the Great, Dial. 2,18,2, is still used
in Spanish with the form “frasca” alongside “frasco” (flask, bottle).
Perhaps scaptos, as a name for a weapon (Etym. 18,8,2). On the sub-
ject of clothing, he explains that renones are typical of the Germanos
(historically accurate), but he adds that common people also call
them reptus (derived from the false etymology reptare), the latter being
a lexical innovation also of Germanic origin and first documented
here (Etym. 19,23,4). Also when referring to clothing, he mentions,
together with the already known Germanic loanword bracae, tubrucus
(Etym. 19,22,30), a kind of “leggings”.61 To those we could add other
like gannatio, found in a 7th-century tablet,62 related to words of mod-

59
Terms taken from our own unpublished study on Innovaciones léxico-semánticas en
las Etimologías de Isidoro de Sevilla (Madrid 1981). Some aspects published in
I. Velázquez, “Vigencia y alcance de los términos innovados en las Etimologías de
Isidoro de Sevilla” Actas del I Congreso Andaluz de Estudios Clásicos ( Jaén 1982) pp.
461–5; ead., “Formación de neologismos en las Etimologías de Isidoro de Sevilla”,
Actas del II Congreso Andaluz de Estudios Clásicos (Málaga 1987) pp. 167–72; ead.,
“Léxico isidoriano en las Etimologías. Problemas para su estudio”, Euphrosyne 22
(1994) pp. 235–43.
60
On the origins of this and other words, see J. Sofer, Lateinisches und Romanisches
aus den Etymologiae des Isidorus von Sevilla (Göttingen 1930) p. 132.
61
Later mentioned by Paul the Deacon, Historia Langobardorum 124,10, ed.
L. Bethmann and G. Waitz, MGH SSrL (Hannover 1878): super quas equitantes tubru-
gos. On this item of clothing, medieval iconographic documentation has been pre-
served, such as the wall paintings of the church of Santa Maria de Tarrasa or a
relief on a tomb in Briviesca (presently at the Museo de Burgos), see. C. Bernis
Madrazo, Indumentaria medieval española (Madrid 1959). Another word first documented
is (h)osa. It is the first Latin word that documents the Germanic “hosa”, but its
introduction into Latin must have taken place earlier in spite of not having been
documented earlier as it has undergone the evolution, typical in Spanish, to uesa,
an ancient word that designated a type of boot.
62
Documentos de época visigoda escritos en pizarra 102. It reads: grande gannatione.
   185

ern Spanish such as “gana, ganar, ganancia” (= gain/profit, to earn,

earnings), derived from the Gothic *waithanjan and *ganan.
But, apart from these examples and similar words and proper
names, it seems that there was no surviving Germanic (Gothic) or
other language among the population of Hispania during the period
under consideration. The chapter in Isidore’s Etymologies does not
shed any light on the diversity of languages used at the time. There
is no mention of a different language being spoken by the Goths. If
that had been the case, he would have perhaps remarked on it.63
He merely explains, when dealing with the “Roman” language, that
it is a “mixed” tongue, the one that took over in Rome. Once the
Empire had reached its apogee, the number of new customs and
men increased, thus corrupting the language with soloecismi and bar-
barismi. The issue here is the evolution of Latin as it was spoken at
a time when lexical elements derived from Germanic languages may
have penetrated it. These elements contributed—together with many
other factors—to the language spoken in late Antiquity.64
In any case, returning to the degrees of assimilation and levelling-
out previously discussed, although I believe, nonetheless, that the
mixing of the two populations and the “Romanisation” of the Visigoths
was pronounced—even before their arrival in Hispania—, this does
not imply the existence of a total identification, which, if it occurred
at all, did not do so until a later date. It seems, for instance, that
judges were of Gothic extraction, as was most of the army;65 the
Visigoths possibly retained certain privileges, especially those related
to paying taxes, and real power was always in Visigothic hands, a
point we shall discuss again later. Despite the growing equality, there-
fore, reflected in the laws—mainly due to the scarcity of differing
evidence or suggesting the existence of different treatments accorded,
for example, to the Jewish minority—and also notwithstanding the

63
In spite of the almost systematic lack of specific references to his period, it is
possible to track down contemporary data through his “linguistic material”.
64
See in particular, M. Banniard, Viva voce. Communication écrite et communication
orale du IV e au IX e siècle en Occident latin (Paris 1992).
65
I refer to these periods around the 3rd Council of Toledo. For instance, a
slate including some condiciones sacramentorum, probably from 589, mentions judges
and vicars, all bearing Gothic names (and I believe that they are of personal ori-
gin here), see Documentos de época visigoda escritos en pizarra 39. A number of scholars
consider that the percentage of Hispano-Roman elements in the army at the time
was approximately 10% or 20%.
186  

fact that the former was bolstered by religious unification, actual

conditions did not show a uniform, united and monolithic society
with a clear awareness that Hispania was the natio of the gens Gothorum,
the latter including Gothi and Hispani (or Romani ).

Fleeting joys

In fact, the joy expressed by the 3rd Council and the desire for reli-
gious unity need not have been received well by everyone, nor was
the resulting political unity, as is shown by the immediate outbreak
of attempted sedition by several members of the Gothic nobility and
clergy, following on from earlier traditions, when there was a con-
stant struggle for power by individuals and factiones; now it was led
by Argimundus ( John of Biclar, Chron. 590,3) or by the bishop Athalocus
and the noblemen Granista and Wildigernus in Septimania, supported
by the Franks;66 earlier, around 587, by the bishop Sunna, enemy of
the Catholic Masona, and Sega in which, apparently, the future king
Witeric was involved.67 The latter usurped the throne of Liuva II,
Reccared’s son, who succeeded his father in 601, in an attempt to
make the monarchy hereditary, but was deposed and murdered by
Witeric in 603. Further turmoil occurred a few years later when
Swinthila68 was faced with a revolt in the province of Narbonne,69
fomented by the Merovingian Dagobert, that ended with the ruling
king surrendering himself and his family in Saragossa and Sisenand’s
proclamation as king around March 26th 631. Sisenand postponed
the celebration of a new general council until December 5th 633,

66
Gregory of Tours, Historiae 1, ed. B. Krusch and W. Levison, MGH SSrM
1,1 (2nd edn., Hannover 1951) p. 430; John of Biclar, Chronica 589; Hist. Goth. 54;
Fredegar, Chronicon 10, ed. B. Krusch, MGH SSrM 2 (Hannover 1888); Vitas Sanctorum
Patrum Emeretensium 5,10–11.
67
John of Biclar, Chronica 588,1; Vitas Sanctorum Patrum Emeretensium 5,10–11.
68
We might think as well that his rise to the throne took place after a new
revolt, perhaps stirred by some sector of the Gothic nobility unwilling, once again,
to accept a hereditary monarchy. Sisebut died, probably poisoned, and he is suc-
ceeded by his son Reccared II, himself dead a few days later; this is clearly suspi-
cious. For a description of the historical and governmental events concerning these
kings, see García Moreno, Historia, pp. 143–58. This author also considers likely
that an attempt to topple him from the throne lies behind the news of Reccared
II’s death.
69
Other reasons for associating his son Ricimir to the throne, Fredegar, Chronicon
73; Hist. Goth. 65.
   187

when the 4th Council of Toledo took place, in the course of which
the king sought to obtain the Church’s consensus to legitimise his
power.70
In opposition to the levelling and unifying tendency, supported by
the Church and promoted by the monarchy, there was a clear ten-
dency towards disintegration—in existence up to the end of the reg-
num—within the nobility who were either envious of the royal family
or tried to prevent it from becoming hereditary, thus depriving other
groups of access to power. This political situation would ultimately
lead to the failure of Visigothic Hispania as a nation, although the
3rd Council of Toledo had laid the bases for this, and there was a
clear political will to make it a reality; useful and necessary mea-
sures to enable this process, such as the introduction of a new leg-
islative corpus under Recceswinth, had been adopted. The strength
of the other forces in the regnum, the nobility ( primati gentis Gothorum)
and the Church, would constrain, however, the scope of that unity;
the first group did so in a negative way whereas the second (albeit
not always) had a positive effect. It seems obvious, despite all my
remarks, that from 589 onwards there exists a political project to
create a state (even better, a regnum Visigothorum), with its base in
Hispania (including the Gallia Narbonensis), that would become the
nation ruled by a rex from the gens Gothorum. That gens would later
tend to define the populus, subject to the king, inhabiting the terri-
tory and the texts themselves would give an account of this process,
above all the laws. But for a long time, and occasionally in later
periods, a clear distinction would still be made between gens Gothorum,
as a community of Gothic origin, to designate the ruling class with
a “nominally” ethnic connotation, in contrast to the rest of the
groups, populi (and of the individuals) living in Hispania or the provin-
cia Gothorum.

From ethnic identity to the “discourse” on ethnicity

But who constitutes this gens Gothorum? Ethnic identity was not
uniform and, in the course of the 6th century, it was not possible
to define it in a precise manner and in the 7th century it was

70
On all this period, see García Moreno, Historia, pp. 143–58.
188  

completely diluted by the mixing of populations, faster in some social

strata than in others. Nevertheless, a nominal awareness of being
from or belonging to a gens (equated to natio) that differed from oth-
ers persisted, although external characteristics, which allowed for dis-
tinction, were few. In fact—and this is of no less importance—Isidore
of Seville points out John of Biclar’s Gothic origins (see note 14) as
well as those of the bishop of Constantinople, Iohannes, natione Cappadox
(De vir. ill. 26), in contradistinction to the rest, Roman or Hispano-
Roman,71 with the exception of his brother Leander, whose origins
are also explicitly revealed: genitus patre Severianus, Carthaginiensis provin-
cia Hispaniae. It is remarkable, even contradictory, that in a society
where those differences would be bound to be erased in theory, an
expression is repeatedly mentioned in the texts that can be inter-
preted as stressing opposing traits: gens Gothorum.
From a possible ethnic identity, based on remote origins and
differences stemming from languages, customs, races, social organi-
sation, and clothing that are gradually becoming less noticeable, we
arrive at a formulation of an ethnic “intellectual discourse”, to the
elaboration of the origin legends of the group, their bravery, mili-
tary feats, conquests and legitimate dominion over new lands. The
fact that religious differences (that characterised and separated two
“ethnic” groups) no longer exist is greatly stressed. I believe that the
key to the elaboration of this “discourse” is Díaz y Díaz’s state-
ment:72
The identification of Visigoth with Arian and Hispano-Roman with
Catholic would have been transformed into a new formula: for the
Visigoths, power (exclusively royal; with an absolute majority in the
case of the high administration and with effective groups among bish-
ops and the clergy) and, from a religious point of view, Catholicism
as the national faith.
This formula of Visigothic power will be clearly stated during the
4th Council of Toledo, as we shall see later. But between the 3rd
Council of Toledo, organised by Leander of Seville and Eutropius
of Valencia,73 and the 4th, headed by Isidore of Seville, this writer’s

71
It can be added, however, that Martin of Braga says: ex Orientis partibus navi-
gans Gallaeciam venit, referring to their origins (De vir. ill. 22).
72
Díaz y Díaz, “Introducción”, p. 23.
73
See John of Biclar, Chronica 580,1. He must have played a leading role. According
to Díaz y Díaz, “Introducción”, pp. 21 and 24, he could have been behind Reccared’s
   189

exceptional prestige has come to take centre stage and he has become
the leading figure of the period. Despite not being bishop of Toledo,
metropolitan capital and sedes regia, his relationship with Visigothic
monarchs is well known; his descriptions of the kings in the Historia
Gothorum reveal his personal contacts with them and, with the excep-
tion of Witeric, the good treatment accorded to him.74 His pastoral—
and political—and literary activities, as well as his work as a bishop,
led to his becoming a “tutor” to society and its monarchs, as Braulius
of Saragossa claimed.75 This author provides us with a synthetic
overview of everything Isidore has done for his compatriots in which
he refers to his works as a coherent whole.76 This vision must under-
line the assessment of his writings and the life of the bishop of Seville,
in spite of the fact that the two both elements differ, sometimes, in
their approach: admiration for classical Antiquity and a negative
evaluation of the Roman Empire, constrained as he was by the polit-
ical situation, his attitude of praising and extolling the new people
now ruling Hispania—the gens Gothorum—and his rex and by his per-
sonal animosity towards the Byzantines.77 His works display, how-
ever, a brilliant synthesis between past and present. Thanks to him,
classical culture survived in a society where, had it not been through
his mediation,78 it might have been shipwrecked. It is thanks to him
that we can approach the notion of culture, the daily reality of his
time, the knowledge of aspects of the spoken language and, most of
all, the political and religious situation, the historical and cultural
awareness of the past, trends in religious, spiritual, legal and politi-
cal thought.

conversion; he was then abbot of the Servitano monastery, when Reccared, as co-
regent, ruled southeastern Spain. It is possible that his rise to the position of bishop
was a royal reward.
74
Also through the correspondence with Braulius of Saragossa (Isidore of Seville,
Epp. 4 and 6). See Díaz y Díaz, “Introducción”, p. 108; Fontaine, Isidore de Seville.
Genèse, pp. 129–43.
75
Quem Deus post tot defectus Hispaniae novissimis temporibus suscitans, credo ad restau-
randa antiquorum monumenta, ne usquequaque rusticitate veterasceremus quasi quandam opposuit
destinam. See C.H. Lynch, San Braulio (Madrid 1950) pp. 356–60; Díaz y Díaz,
“Introduccion”, pp. 95–162 and 215–23.
76
According to Augustine’s version, De civitate Dei 6,2 includes Cicero’s eulogy of
Varro (Cicero, Academicorum reliquiae cum Lucullo 1,3,9, ed. O. Plasberg [Leipzig 1969,
repr. 1996]), see Díaz y Díaz, “Introducción”, p. 216.
77
See J. Fontaine, “Théorie et pratique du style chez Isidore de Seville”, Vigiliae
Christianae 14,2 (1960) pp. 65–101.
78
And without any other, of course: Velázquez, “Ámbitos y ambientes culturales
en la Hispania visigoda”, here pp. 328–30.
190  

In this context we must also include his role as historian, as it is

crucial to teach past and contemporary history in order to make
sense of the period when one lives.79 But his writings also have a
political value and far-reaching significance. Taking over from his
brother Leander, he toiled for many years from his episcopal see to
consolidate the difficult balance between Church, Visigothic mon-
archy and nobility and contribute—under the premise of unity of
the faith—to the unification of the regnum. At the same time, he con-
tributed unequivocally towards strengthening the image of the reges
and the gens Gothorum. His historical works, the Chronica and the
Historia Gothorum, have a clear purpose: to recount and praise the
antiquity and courage of the people, to write the history of their
kings, both their merits and blemishes, from a Christian perspective
and to foster the union of that gens with Hispania, Isidore’s patria.
The Chronica shows a lineal succession of the earth’s kingdoms—
as opposed to the usual synchronicity of data—through the six ages
in which the world is divided, with an approach not only religious
but also political.80 The Roman Empire has a “finite” history like
others, and as a counterweight to this the Goths have managed to
play a decisive role, carrying out the translatio regni from Byzantium
to Toledo. However, there is no contradiction between this Chro-
nica, or the Historia Gothorum, and the Etymologiae in the author’s
representation of Rome.81 The writings convey an ideal of Romanitas,
or even better of the old eternal Rome—not the Roman Empire,
but Byzantium—, of whose Roman character the Goths are now
heirs.
Teillet’s remark82 is accurate: that the first ruler to be accorded
the title rex Gothorum in the Chronicle is Theoderic II, thus equating
the regnum Gothorum with the arrival of the Visigoths in Spain.83 I

79
On the educational value of history, Fontaine, Isidore de Seville, Genèse, pp. 217–8.
80
However, the religious view guides the work and a certain sense of optimism
because the unity of the Catholic church will lead to the plenitude of time. For
the manifestation of certain anti-Jewish views and the opposition to heresies, includ-
ing the millenarian views prevalent at the time, see Díaz y Díaz, “Introducción”,
pp. 137–9.
81
Against, Teillet, Des Goths à la nation Gothique, p. 464. There are, nonetheless,
some religious tensions, as I have previously indicated.
82
Ibid., p. 471.
83
It is surprising that Athanaric and the battle of Adrianopolis are not men-
tioned, when all contemporary sources record it: Teillet, Des Goths à la nation Gothique,
   191

believe that it is possible to detect in this work an interest, although

not as clearly stated as in John of Biclar, in recording the history
of the Visigothic monarchs in Hispania, albeit here “disguised” in the
shape of an universal and Christian chronicle. The simultaneity of
a rex Gothorum and regnum Gothorum, on one hand, and the latter’s
arrival in Hispania, highlighted by Teillet, evolves in the Historia
Gothorum in which the expressions regnum and reges Gothorum are men-
tioned, the verb regnare is used, applied to the earliest known mon-
archs, from Athanaric’s rule and even to the very ancient mythical
periods of these people (and kingdom) which Isidore incorporates
into his work.84 It is obvious that this work has been written ad
maiorem gloriam Gothorum. It has been designed to commend the ori-
gins of these people as well as their accomplishments so that they
can be opposed to Rome, whose mythical and epic origins are more
recent in time, for the Goths are descendants of Magog, son of
Japhet. The gens Gothorum, defined in Etym. 9,2,89, as gens fortis et
potentissima, is singled out in history because of its bravery and courage,
considered as a virtus, as well as its love of freedom.85 The ancient
gens barbara has become civilised, as opposed to other gentes which
are still shown as barbarian; it has replaced its customary laws with
laws written in Latin (Hist. Goth. 35). Once again, it cannot be claimed
that when Isidore refers to this gens he is including the Hispani. It
would be contradictory in a work conceived to deal with the origin
of the Goths, to designate all of them as Gothi, with no terminolog-
ical difference, despite the existence of a conceptual development in
relation to the Goths of Athanaric and Reccared. They are no longer
the old gens barbara, by virtue of their conversion to Catholicism (and
their coexistence with Roman culture, although this idea is not clearly
stated in the work, but is evident in Isidore’s thought).86

p. 471 n. 58. Isidore devotes a long passage to it, however, in Hist. Goth. 6–11. In
the long version it is said that, for centuries, they were ruled by duces and later by
reges. In the short one, only kings and kingdoms are referred to: per multa quippe sae-
cula et regno et regibus usi sunt.
84
There is, furthermore, a clear interest in linking them as soon as possible to
the history of Hispania. That justifies the introduction of the passage in which the
Goths lend their support to Pompeius in the civil war opposing him and Caesar,
which occurred “twelve years before the Hispanic age” (Hist. Goth. 13).
85
A custom for them, whose repression by the Romans justifies their uprising
against Valente: Hist. Goth. 8: Sed ubi viderunt se opprimi a Romanis contra consuetudinem
propriae libertatis ad rebellandum coacti sunt. See note 19.
86
There is only one passage where the formulation might prove ambiguous. The
192  

Isidore will justify the Goths’ victories over the Romans, and give
them legitimacy, by means of the same procedure used by John of
Biclar, namely the use of a juridical terminology that evokes the ius
Gothorum in contradistinction to the ius Romanorum.87 These magnificent
people have glorified themselves because their kings have subdued
(legitimately) all the territories of the old Roman province of Hispania;88
they have established their mandate there. A people whose great-
ness has reached its zenith in the conversion to the Catholic reli-
gion, understood—and this is fundamental—as the unity of God’s
church, since it is through this religious unity that peoples can be
united and this is precisely the key to the unifying conception of the
peoples of Hispania advocated and worked towards by the Church
in this period. That is why Isidore, in the midst of this process of
extolling the gens Gothorum, can afford to criticise the role of some
“non-Catholic” kings. He refines and rectifies the positive view towards
Liuvigild in John of Biclar’s writings,89 but underlines Theudis’ toler-

reason is, perhaps, that Isidore seeks, obviously, the union of the peoples dwelling
in Hispania (Hist. Goth. 52): In ipsis enim regni sui exordiis catholicam fidem adeptus totius
Gothicae gentis populos inoliti erroris labe detersa ad cultum rectae fidei revocat. It can be
assumed that, once the heresy has been stamped out (from their own gens), it leads
the peoples making up (or belonging to) the gens Gothica to practise the true faith;
this entails all of them together, the Goths themselves and all those ruled by them;
however, totius, in this occasion, could have been better connected to populus rather
than to gens.
87
Hist. Goth. 36: Theudericus Italiae rex [. . .] partem regni quam manus hostium occu-
paverat recepit, Gothorumque iuri restituit.
88
It must be stressed that, contrary to what we might think, alongside the use
of Hispania (= Spania) in singular as a territorial unit, the Roman conception of the
provinciae (according to Etym. 14,4–5) remains very much present in Isidore’s works,
besides the mention of the Gallia. There are also equivalences with regiones. As I
have mentioned already apropos John of Biclar’s Chronica, the word provincia retains
its use, not applied to the permanence of the Roman administrative divisions—
although the Church based its organization of the dioecesis on them and some over-
lapping must have still existed in some aspects—, but rather because the “reality”
of the territorial diversity, the differences not so much among peoples—or perhaps
customs—, but even their gradual incorporation into the imperium or dominium of
the Visigoths (that is, the formation of that provincia Gothorum) and, probably, their
differing types of integration and incorporation and the ecclesiastical organization
itself, must be behind those designations of regiones, provinciae or the plural uses of
Hispania.
89
In spite of the recognition of Liuvigild’s conquests, amounting to almost all
the territory of Hispania, the negative vision resulting from their lack of piety exerts
a greater influence on him (Hist. Goth. 49): Spania magna ex parte potitus, nam antea
gens Gothorum angustis finibus artabatur. Sed offuscavit in eo error impietatis gloriam tantae vir-
tutis; the same happens in De vir. ill. 31 where, when speaking about John of Biclar,
   193

ance towards the Catholic faith. This monarch even allowed the cel-
ebration of a Council in Toledo (Hist. Goth. 41). He also stresses the
kings’ moral conduct and the way he acceded to the throne. Thus,
he “almost” justifies Thorismund’s violent death (Hist. Goth. 51), or
that of Theudis (Hist. Goth. 44) and, most importantly, Witeric’s,
employing the biblical quotation “he who lives by the sword, dies
by the sword” (quia gladio operatus fuerat, gladio periit) (Hist. Goth. 58) to
describe the latter: Witeric usurped the throne (sumpta tyrannide) and
murdered Liuva II, Reccared’s son.
The focus of the work is the unity of the Catholic church, which
unites and gives cohesion to the regnum of Hispania, and the ruling
power, namely the reges of the gens Gothorum. As we know, the work’s
long version90 begins with the Laus Spaniae and concludes with a
Recapitulatio, an emotional encomium of the gens Gothorum. Both can
be considered as a reflection of Isidore’s “optimism” which has some-
times been drawn to our attention. The Laus Spaniae is Isidore’s emo-
tional ode to his land ( patria). Represented as a fecund mother, it is
the most beautiful of all lands (omnium terrarum pulcherrima). He uses
for this eulogy the panegyric literary tradition, both in verse and
prose, exemplified by the encomium of provinces.91 Rome, chief of
all the peoples, longed for and owned (concupivit et [. . .] desponderit)
this land, magnificent and fertile, “sacred and always content, mother
of princes and peoples” (sacra semper felix principum gentiumque) and,

he recounts that which the latter author had omitted in his own Chronica, namely
that he had suffered exile and persecution on the part of Liuvigild, for refusing to
abjure the Catholic faith: Hunc supradictus rex [Liuvigildus], cum ad nefandae haeresis
credulitatem compelleret, et hic omnino resisteret, exilio trusus, Barcinona relegatus, per decem annos
multas insidias et persecutiones ab Arianis perpessus est.
90
It is common belief that the short version was written at the end of Sisebut’s
rule, around 619, whereas the long one was compiled under Swinthila, around 625,
when the latter had already installed his son as co-regent to the throne. However,
the first editors, Grial, Flórez and Arévalo, considered the longer version to be the
first one whereas the second had been mutilated (apart from undergoing addenda),
as a result of a sort of damnatio memoriae by Swinthila, discredited from the 4th
Council of Toledo onwards (see the discussion by C. Rodríguez Alonso in the edi-
tion of Hist. Goth., pp. 26–49). However, I think that these aspects should be re-
examined as there are still some surprising points to discover in the differences
between both versions not sufficiently elucidated. A revision of the problem in
R. Collins, “Isidore, Maximus and the Historia Gothorum”, Historiographie im frühen
Mittelalter, ed. A. Scharer and G. Scheibelreiter (Vienna-Munich 1994) pp. 345–58.
91
See Fontaine’s analysis, Isidore de Seville. Genèse, pp. 361–77.
194  

likewise, it is the “prosperous Gothic nation” who has conquered it

and enjoys it safely until now: denuo tamen Gothorum florentissima gens
[. . .] certatim rapit et amavit fruiturque hactenus [. . .] securas).
I believe that Isidore shows two distinct realities here, now drawn
together: Spania and the gens Gothorum. Spain as a marvellous land
that has always existed and has bestowed so much wealth and men
on the world. A land conquered by Rome and then by the Goths.
One cannot fail to observe the choice of vocabulary: concupivit et
desponderit for Rome and rapit et amavit fruiturque for the gens Gothorum,
more passionate in the second case but with a certain legitimate res-
onance in the desponderit of the first, nor the complacency towards
the new situation suggested by the terms amavit and fruitur and securas
which end this eulogy. Nevertheless, the expression hactenus, “until
now”, is disturbing. Is there any evidence of millenarian foresight or
is it his conviction that Spania, Isidore’s beloved motherland, will
exist forever, before, during and after her rulers? Let us recall that,
in his chronicle, the author conceives of the kingdoms as being sub-
ject to the beginning and end of time. For Isidore, Spain is now
(happily) ruled by the Goths, this gens florentissima.
I do not believe that the placing of this encomium of Spain at
the beginning of the work is arbitrary, nor that after describing the
origins, military feats and history of the Gothic kings and people, it
ends with a recapitulatio praising them. This consists of a synthesis of
their origins, where mention is again made of the etymology of their
name, their pilgrimage from their native lands—Western Europe’s
glacial mountains—up to their arrival in Spain where they would
settle and establish their empire. Their natural virtues are extolled,
as well as their physical traits and external characteristics: they are
quick, ingenious, confident in their strength, tall, distinguished in
their demeanour, resilient, courageous (so much so that they have
defeated Rome, the grande dame of all nations, and all the other
groups they have faced in battle); they are freedom-loving, noble in
the art of war, skilful in handling weapons, and have mastered the
art of sailing of which they had no previous experience, having been
recently acquired during Sisebut’s rule, so that the “Roman soldier”
(here again we can see Isidore’s anti-Byzantine bias) has become their
servant, seeing how they are served by all peoples, including Spain
herself: servit illis Romanus miles quibus servire tot gentes et ipsam Spaniam
videt. Fontaine emphasises the role of the Laus Spaniae in the syn-
thesis of Roman and Hispanic traditions and there is, indeed, a pro-
   195

nounced “Roman and Hispanic patriotism” at the same time.92 A

political mentality, detectable in the rest of the work, in which the
“ancient loyalty to the Roman tradition does not exclude the inte-
gration of all those who were previously barbarian”. Isidore helps to
foster the great hope of “co-operation between Goths and Hispano-
Romans in a unified Spain after more than a century of hostility
and destruction”.
In my view, however, the Recapitulatio is the counterpoint to real-
ity. Isidore tries to join the two realities, Spania and the gens Gothorum,
but it is clear that they are not the same and that the latter does
not include the Hispani, nor any other group. The author has cre-
ated an “ethnic” myth; he has not stinted in his praise and offers
an image that helps to consolidate the prestige of this gens. The last
sentence uttered is crucial and clear. Just as so many other peoples,
Spain also serves them. He has not only elaborated an “ethnic” dis-
course (mythical, stereotypical and without a demonstrable ethnic
basis at that moment), conveying an image of who the Goths are
and what they look like—basically, perfect and excellent—, but he
also describes and accepts compliantly those ruling his land and hold-
ing power over it. This exaltation and the joy manifested years before
at the 3rd Council of Toledo by the Catholic bishops probably sought
with determination to achieve the total unity of the peoples of
Hispania; but not everyone would have seen it in the same way.
Reference has already been made to the revolts that broke out imme-
diately against the monarchy; one must also take into account the
fact that among the Hispano-Romans there was neither unanimity
nor general satisfaction with the new status quo. It is even possible
that an attitude such as that of the extremely important bishop of
Seville was not rightly understood at the time, giving the impression
of an inconstant and “collaborationist” attitude, as is demonstrated
by the marked contrast between the flattering passage on Swinthila
in the Historia Gothorum (long version) and the condemnation of the
same king at the 4th Council of Toledo, over which he presided.93

92
As I have indicated earlier, I see no contradiction between the “Roman”
approach of the Etymologiae and this, as Teillet states.
93
This would justify perhaps the absence of a more detailed biography written
by his contemporaries, considering the prestige and admiration he enjoyed due to
his literary and encyclopaedic endeavours. Perhaps his functions as a bishop were
not always regarded in a positive way, see Díaz y Díaz, “Introducción”, p. 95.
196  

The 4th Council of Toledo: The definition of power

This Council has been described as constitutional94 because it for-

mulated for the first time a clear definition of what could be under-
stood politically as Visigothic power, based on the interrelationship
between gens, patria and rex, and what a king’s qualities and duties
ought to be. Teillet is right95 when she states that this theoretical
formulation is probably attributable to Isidore of Seville. The writer
presents the relationship between gens, patria and regnum by means of
loyalty to the king, which is his political ideology formulated through-
out his historical works and in his Sententiae, a work that can be con-
sidered as an authentic “spiritual testament”.96 The Council was the
climax of his pastoral work, of his attempts to forge unity, already
sought by his brother Leander, his last task as “tutor” of the monar-
chy based in Toledo.97
The Council comprises seventy-four canons devoted to ecclesias-
tical matters, many of them aimed at achieving the unification of
the Hispanic rite in the churches of Spain and Gallia. I think that
it is worth highlighting here the almost systematic way in which
Spania and Gallia are treated as the two components of the regnum
now led by Sisenand;98 a regnum whose territorial unity had been
achieved under his predecessor Swinthila. This question does not
seem negligible to me, considering the weight of Teillet’s arguments,

94
Orlandis and Ramos-Lissón, Historia de los Concilios, pp. 261–98.
95
The analysis of this Council and its connection with Isidore of Seville in ead.,
Des Goths à la nation Gothique, pp. 503–36.
96
Fontaine, Isidore de Seville. Genèse, pp. 235–50.
97
Ibid., pp. 129–43.
98
IV Toledo, c. 2: per omnem Spaniam atque Galliam; c. 3: totius Spaniae et Galliae
synodos (is there an acknowledgement that Hispania is made up of several areas and
now its entire territory is comprised?); c. 13: sed pari modo Gallia Spaniaque celebret;
c. 14: ut per omnes ecclesias Spaniae vel Galliae. When reference is made to Hispania,
without mentioning Gallia, this is seemingly due to the description of aspects only
pertaining to the first one, where certain conditions apply. Here, however, the use
is usually in the plural, or in the singular if it is determining another word: c. 5:
in Spaniis; c. 6: in Spaniis, but also: de hac Spaniae diversitate; c. 9: per multarum loca ter-
rarum regionesque Spaniae; c. 10: per Spanias, or even plural in c. 12: in quibus quoque
Spaniarum ecclesiis laudes. Besides, in c. 41, apropos the type of tonsure required of
clergymen and lectors, there is an explicit reference to different uses in Gallaecia
and formerly common among heretics in Hispania; now it is time to change because
both the tonsure and the cloak must be the same there: non sicut hucusque in Gallaeciae
partibus [. . .] ritus enim iste in Spanias haereticorum fuit [. . .] et una sit tonsura vel habitus
sicut totius Spaniae est unus.
   197

as I have indicated above, on the use of the term Spania in the sin-
gular to denote the concept of Spain and the abandonment of the
Roman concept of provincia Hispaniarum. In my view, the reiteration
of Spania and Gallia at this Council has a notable ecclesiastical pur-
pose given that its context is the need to unify the ecclesiastical rites
of both areas, and, on the other hand, a political interest, in that it
corresponds exactly to the territories ruled by the Visigothic regnum.99
Besides, in later conciliar texts, alongside the single designation of
Hispania, the duality Hispania (and its provinces) and Gallia, or even
Gallaecia as part of the regnum is still maintained, especially when ref-
erence is made to bishops, their presence at councils or the general
scope of prescriptions set in them.100 It is common to find that Hispania
is preceded by the adjective totus or omnis, alongside Gallia and, as I
remarked earlier, possibly reflecting the plurality of provinces into
which Roman Hispania had been divided which still retained some
validity for the distribution of power and the apportioning of eccle-
siastical territories. Furthermore, doubtless due to the weight of lit-
erary tradition, the term is still occasionally used in the plural, with
a lexical distribution as synonym for tota Spania, as one can see at
the 18th Council of Toledo, the last council whose proceedings have
been preserved.101
Nevertheless, although this is basically a question of lexical (and
morphological) uses, it does not lack a certain degree of complex-
ity, along the line of the argument employed by Teillet that I have

99
In fact, the beginning of the text shows a certain syntactic ambiguity that
could be deliberate: Dum studio amoris Christi ac diligentia religiosissimi Sisenandi regis
Spaniae atque Galliae sacerdotes apud Toletanam urbem in nomine Domini convenissemus. The
placing of the determining expression Spaniae atque Galliae between Sisenandi regis and
sacerdotes is significant, although it is rightly qualifying sacerdotes. To the extreme that
J. Vives’ Spanish translation, Concilios, p. 186, reads thus: “Sisenand, king of Spain
and the Gaul”.
100
XII Toledo, p. 383: per totos Hispaniae fines [. . .] totius Hispaniae duces; c. 6,
p. 394: placuit omnibus pontificis Spaniae et Galliae; XIV Toledo, c. 1, p. 442: cunctorum
Spanorum praesulum totam Hispaniam vel Galliam synodale edictum; c. 2, p. 442: clara omnis
populos Hispaniae implet; c. 4, p. 433: provincias regni and legatos Hispaniae; c. 5, p. 444:
omnes Hispaniae adgregati; and especially at the 13th Council of Toledo, apropos the
taxes levied from the peoples of the kingdom in Erwig’s Lex in confirmatione concilii,
p. 436: in provinciam Galliae vel Galliciae atque in omnes provincias Hispaniae.
101
XVII Toledo, p. 522: plerique Spaniarum Galliarum pontifices convenissem; c. 2:
decernimus ut ita a totius Spaniae et Galliarum pontificibus custodiatur; c. 3: per totius Spaniae
et Galliarum ecclesias eadem sollemnitas celebretur; c. 6: per universas Spaniae et Galliarum
provincias.
198  

partly discussed. I am referring to the fact that, alongside a clear

identification between Hispania and the regnum Gothorum and the Iberian
Peninsula as a whole (including Gallia, although I think that the lat-
ter was always considered as an additional inheritance; see note 34),
the different uses in the texts suggest that we are not facing a sin-
gle and monolithic reality. Especially because what is indisputable,
I believe, is the deliberate use of Spania or Hispania in the singular
as a unity on a number of occasions and, most of all, by certain
authors. Such is the case in Isidore of Seville and later undoubtedly
in Julian of Toledo’s Historia Wambae, where Hispania is the regnum
and the unified motherland (or, as Teillet claims, the official name
of the Visigothic nation, that is the Spanish nation),102 in marked
contradistinction to Francia and, even, to Gallia, against which an
Insultatio is directed. Moreover, the use—very rare in the texts we
have seen so far—of Hispani (three times) as opposed to Gothi (once).103
The author refers to an exercitus Hispaniae, Hispanorum, rather then
Gothorum. But it is precisely this deliberate and propagandistic use,
clearly perceived in the work, which reveals, perhaps, a determined
attempt to stress unity around the king in the face of the grave ten-
sions at this period between the Toledo monarchy’s centralising pol-
icy and the segregationist (and secessionist in this case)104 tendencies
of many regions of the kingdom, spurred by the latter’s internal frag-
mentation, infighting and ambitions, if not in favour of independence
at least with the intention of controlling their territories directly, on
the part of the powerful nobility.105
Returning to the 4th Council, it is the last canon, canon 75, which
is most important from a political point of view and justifies the con-
vening of the Council.106 In it, we read that they have been sum-

102
Teillet, Des Goths à la nation Gothique, pp. 628–32. In the modern sense of
nation. Notice the different assessment made by the author with regard to the texts
of the work and the councils mentioned immediately before it.
103
I will deal with this work later.
104
I. Velázquez, “Wamba y Paulo: dos personalidades enfrentadas y una rebe-
lión”, Espacio, tiempo y forma, Serie II, Historia Antigua 2 (1989) pp. 213–22.
105
L.A. García Moreno, El fin del reino visigodo de Toledo (Madrid 1975); id., Historia,
pp. 170–90; Ripoll and Velázquez, La Hispania visigoda, pp. 37–41.
106
I do not think it fortuitous that this Council, convened by Sisenand to legit-
imise his coup d’etat against Swinthila, tackles religious issues, with the exception
of this canon 75, which mainly addresses the religious unity. The Church, and per-
haps Isidore of Seville himself, was used, or manipulated, by the king who needed
a religious legitimation for his actions as a tyrannus, in spite of the fact that the
   199

moned by the king to draw up the pontificale decretum (domno nostro

Sisenando regi, cuius devotio nos ad hoc decretum salutiferum convocavit). This
passage, together with the opening words of the Council, set out the
key issue to be considered: the need to legitimise the monarchy by
religious means and its institutionalisation as an elective monarchy
linked to the Church. Thus, from the outset Sisenand is described
as religiosissimus rex and this juxtaposed to the gathering of sacerdotes
(an increasingly frequent label applied to bishops, with doubtless non-
clerical resonance) motivated by the studio amoris Christi and the dili-
gentia religiosissimi Sisenandi. In this passage in canon 75 a Gloria in
honour of God is pronounced and prayers are said for the piissimo
et amatori Christi regi whose devotio has brought them together.
Canon 75 is written (perhaps with Isidore of Seville as main insti-
gator?) pro robore nostrorum regum et stabilitate gentis Gothorum, that is “for
the strengthening of the kings and the stability of the gens Gothorum”.
To attain this double and inseparable security they begin by show-
ing that reneging on the oath of fealty to the king is an impious
act, characteristic of people who do not fear God’s judgement. This
oath is regarded as sacred: fidem sacramento promissam regibus suis; they
do not hesitate to include phrases such as Nolite tangere christos meos
(Psalm. 104,15; II Reg. 1,4) or Quis inquit extendit manum suam in Christum
Domini et innocens erit? (I Reg. 26,9). Phrases that, were it not because
of the lack of clear evidence that the royal unction took place before
that of Wamba, would lead us to think that it was already current,
or was about to be introduced at the time of the Council.107 Those
attending the Council have no qualms about branding as a sacri-
legium renouncing the oath of fealty, as the transgression is not only
against kings but also against God, in whose name the oath was
sworn ( pollicetur ipsa promissio).108 Thus, to avoid divine wrath worship

image projected by the Acta is that of an imploring and humiliated king: ingressus
primum coram sacerdotibus Dei humo prostratus cum lacrymis et gemitibus pro se interveniendum
Deo postulavit. For its part, the Church cashed in on the moment to legislate on the
religious unity, true bastion of the unity and peace of the kingdom, and sought to
reconcile spirits in a veritable attempt to instil some stability into the state of affairs.
107
A. Barbero, “El pensamiento político visigodo y las primeras unciones regias
en la Europa medieval”, La sociedad visigoda y su entorno histórico, ed. J. Faci (Barcelona
1992) pp. 1–77, here pp. 25–7; Orlandis and Ramos-Lissón, Historia de los Concilios,
pp. 293–4.
108
This explains the fact that divine wrath has destroyed many earthly kingdoms,
per impietatem fidei; this is still, perhaps, a tacit justification of how the Goths have
been able to vanquish Rome.
200  

of God and loyalty to the king must be steadfast during one’s life-
time. Next comes an enumeration of forbidden acts, a slippery slope
of mental attitudes, words and deeds that lead to treason: [. . .]
infidelitatis subtilitas impia, subodola mentis perfidia, periurii nefas, coniura-
tionum nefanda molimina; nullus apud nos praesumtione regnum arripiat; nul-
lus excitet mutuas seditiones civium; nemo meditetur interitus regum. Such
attitudes could well be applied to the gens Gothorum themselves, given
the situation they have just experienced, namely the usurpation of
Sisenand’s power and, very possibly, other attempts directed against
him by nobles still supporting Swinthila; this fact could explain the
delay in convening the Council.109 One of the key formulations, the
elective nature of the monarchy, follows immediately:
sed defuncto in pace principe primatus totius gentis cum sacerdotibus successorem
regni concilio communi constituant, ut unitatis concordia a nobis retinetur, nullum
patriae gentisque discidium per vim atque ambitum oriatur.
The elements that shape and sustain power are being defined here:
rex, the nobility of the gens and the Church. The aristocracy and the
bishops choose the king. This system will ensure social harmony and
will prevent disunity of patria and gens as a result of ambition and
violence. This unity between patria and gens is first mentioned here
in a clear manner; their unity will be sealed henceforwards, although
with several changes. Their mutual association is reminiscent, espe-
cially in its full formulation patriae gentisque Gothorum, of the senatus
populusque Romanus. This written formulation, now binding by virtue
of its inclusion in the Council Proceedings, fills a legal vacuum around
monarchy and the structure of government with the consensus of
the three active forces at the time:

• The king and the monarchy retained their power over all the oth-
ers and their rule over the regnum by means of the oath of fealty
sworn by the others. But there was a legal caveat: he must be
chosen and could not, therefore, try to make the monarchy hered-
itary, ignoring the nobility’s interests. His royal nature was not
based on nobility of blood (or not on blood alone) but on his
deeds as a monarch worthy of that name, endowed with the main

109
Orlandis and Ramos-Lissón, Historia de los Concilios, pp. 293–4; Díaz y Díaz,
“Introducción”, pp. 35–6. Specifically, Iudila’s uprising in the Baetica and part of
the Lusitania. Thompson, Los godos en España, pp. 201–3.
   201

virtues: pietas and iustitia. These legal constraints reveal Isidore of

Seville’s ideological stand, enunciated in his works:
Recte igitur faciendo regis nomen tenetur, peccando amittitur. Vnde at apud veteres
tale erat proverbium: “Rex eris, si recte facias: si non facias, non eris”. Regiae
virtutes praecipuae duae: iustitia et pietas. Plus autem in regibus laudatur pietas;
nam iustitia per se severa est (Etym. 9,3,4–5).110
Moreover, what is tacitly being justified here, with specific reference
to recent events concerning Swinthila and Sisenand (and possibly
Iudila) is the alternative to the sacred and inviolable nature of the
monarchy and the oath of fealty, namely tyranny, usurpation of royal
power by a monarch who is not worthy of that title; that is why
Swinthila is expressly condemned, although it is alleged that he him-
self abdicated through horror at his own crimes and renounced his
royal privileges: qui scelera propria metuens se ipsum regno privavit et potes-
tatis fascibus exuit.

• One of the main aspirations of the nobility is gradually being

accomodated while, at the same time, recognising its prominent
role within the kingdom. Against the reiterated attempts to make
the monarchy hereditary and thus allowing a single noble family
exclusive access to power, election from among the primati—so
that any of them could be selected, thus legitimising once again
Sisenand’s rise to power—is unequivocally stated. Elements of
Hispano-Roman society, possibly stemming from the late-Roman
aristocracy itself, in possession of land and wealth, would by now
have joined this nobility, although their origins would be diluted
and no reference to the name Hispani or Romani would be made.
• The Church takes on a decisive and determined role in the selec-
tion of the monarch. Thus, the Catholic church, whose main rep-
resentatives are both Hispano-Roman and Visigothic,111 becomes

110
The title of king is deserved when one’s conduct is right and no evil deed is
ever committed, hence the ancient proverb: “You shall become king if your con-
duct is right, otherwise you shall not”. Royal virtues are mainly two: justice and
piety. Piety in kings, however, is praised more strongly, as justice is inherently severe,
cf. Barbero, “El pensamiento”, p. 23. See Sentent. 3,48,7: Reges a recte agendo vocati
sunt, ideoque recte faciendo regis nomen tenetur peccando amittitur. See M. Reydellet, La ro-
yauté dans la littérature latine de Sidoine Apollinaire à Isidore de Séville (Roma 1981) pp.
515–9.
111
The proper names used in the Acta, reveal the gradual increase of bishops of
Gothic extraction. As Díaz y Díaz recalls, “Introducción”, pp. 41 and 60, although
202  

the third element of power. Although, at first, it must have advo-

cated a hereditary monarchy, rather than an elective one, in an
attempt to achieve stability, here it sidles with the latter, in this
consensus of all the parties in search of equilibrium and peace.
By contrast, kings will intervene more and more in the appoint-
ment of bishops.112

To achieve social harmony and the unity of the regnum, in my view

clearly defined here as patriae gentisque, the well-known sententia is
drawn up; it is to be pronounced three times, with significant changes
according to the religious occasion, of mentioning each member of
the Holy Trinity, to prevent any “among us” or “among the peo-
ples of Spain” from betraying the motherland, the Goths or the king;
those who make such an attempt, and their accomplices, will be
guilty of anathema and expelled from the Church:
Quicumque igitur a nobis uel totius Spaniae populis qualibet coniu-
ratione vel studio sacramentum fidei suae, quod pro patriae gentisque
Gothorum statu vel conservatione regiae salutis pollicitus est, temeraverint,

names do not infallibly identify the origins, in Hispania a phenomenon occurred

that was different from other places: the prestige of Roman names compelled some
Goths to adopt them, in particular those of Iohannes and Stephanus; see G. Kampers,
Personengeschichtliche Studien zum Westgotenreich in Spanien (Münster 1979) pp. 160–1.
However, it is possible that “the ethnic prestige” could have led to the opposite
case (D. Claude, Adel, Kirche und Königtum im Westgotenreich [Sigmaringen 1971] pp.
111 ff.; García Moreno, Historia, p. 228). But I believe one must proceed with
extreme caution on this issue because, simultaneously, the opposite case also took
place. Perhaps this increase of Gothic names had a higher incidence from the sec-
ond third of the seventh century, when sources start to drop, as we will see later,
the mention of the gens from the determining name Gothorum and, in particular,
since the transformation of the palace nobility under Chindaswinth. Nonetheless—
and with all due reservation—I think that it is indispensable to carry out not only
a prosopographic study, but also a comparative one that analyses the use of proper
names also in other locations.
112
Already under Reccared. The royal election is presented as one of the possi-
ble means to gain a bishopric at the 2nd Council of Barcelona in 599: aut per sacra
regalia aut per consensionem cleri vel plebis (the latter to a lesser extent) vel per electionem
assessionemque pontificum, see Díaz y Díaz, “Introducción”, pp. 38–9. Isidore himself
considered the royal appointment as normal when speaking about a bishop’s appoint-
ment for the Tarraconense, around 632 (see Braulius, Epp. 6 and 5), in spite of the
formulation of canon 19 at this 4th Council of Toledo favouring the imposition of
the traditional rules: sed nec ille deinceps sacerdos erit, quem nec clerus nec populus propriae
civitatis elegit, vel auctoritas metropolitani vel provincialium sacerdotum assensio exquisivit. The
most frequent royal election during the seventh century can be behind the increase
in the number of Gothic bishops.
   203

aut regum nece attractaverit aut potestate regni exuerit, aut praesumptione tyran-
nica regni fastigium usurpaverit, anathema sit in conspectu Dei Patris et
angelorum atque ab ecclesia catholica, quam periurio profanaverit, efficiatur extra-
neus et ab omni coetu113 Christianorum alienus114 cum omnibus impietatis suae
sociis, quia oportet ut una poena teneat obnoxios quo similis error invenerit impli-
catos.
• In the second case the reason for breaking the oath will be uttered:
tractatu vel studio sacramentum fidei suae [. . .] violaverit, and the
anathema will be: in conspectu Christi et apostolorum eius.
• In the third, the main variation is: meditatione vel studio sacra-
mentum fidei suae pro patria salute gentisque Gothorum statu vel
incolomitate regie potestatis pollicitus est violaverit, and the anath-
ema will be: sit in conspectu Spiritus Sancti et martyrum Christi.

I believe that the distinction between nobis and totius (or cunctis in the
second and third formulae) Spaniae populis sheds light on the reality
of the situation. On the one hand are the “peoples” (= populi ) of
Hispania and on the other is the regnum of Hispania herself, ruled by
a king (rex), constituting the common patria, but also regarded as the
motherland of the gens Gothorum, that is, the human group to which
the king belongs and whose ruling class exercises power, although it
is nowadays made up of an aristocracy that includes Gothic and
Hispano-Roman elements.115 The thin line separating the notion of
a motherland common to all from a “patrimonialist” interpretation
of a motherland that a people, or human group, has taken over—
to whom it belongs by law and which is ruled by them—is evident
in this formulation. In contrast to the first two expressions pro patriae
gentisque Gothorum statu, which show a unity of concept and refer to
the “stability of the motherland and the Gothic people”, the third

113
Consortio is the second formula and communione the third.
114
Henceforth the text varies significantly, although all convey the same sense.
In the second formula: et damnatus in futuro dei iudicio habeatur cum comparticibus suis,
quia dignum est qui talibus sociantur, ipsi eius damnationis eorum participatione obnoxii tenean-
tur. In the third: neque partem iustorum habeat sed cum diabolo et angelis eius aeternis sup-
pliciis condemnetur una cum eis qui eadem coniuratione nituntur, ut par poena perditionis constringat
quos in pernicie societas copulat.
115
García Moreno, Historia, pp. 223–8; 317–9. It is obvious that, as the sources
show, there already were mixed marriages from Theudis’ rule onwards, as we
remarked, and, most especially, since Liuvigild’s law. But that does not mean that
the gens Gothorum was identified with a *gens hispanorum, giving up their prestige of
birth and the “ethnic” model advocated as a power group or ruling group.
204  

shows this subtle distinction of ideas: “for the wellbeing of the moth-
erland and the stability of the Goths” already inseparable as ele-
ments shaping the state, but two independent realities. It is this
“motherland” which is thus presented as differentiated by means of
a simple syntactic procedure, identifying Hispania with patria, under-
stood as an ideological and territorial unity, and gens Gothorum with
her rulers; it is this separation, which, as I remarked, is not fortu-
itous, that can be seen in the layout of the Laus Spaniae and the
Recapitulatio (a Laus Gothorum as it is entitled in some manuscripts) at
the beginning and the end respectively of the Historia Gothorum by
Isidore of Seville.
What, in my view, is defined at the 4th Council of Toledo is
power, rather than the nation. “Ethnic discourse” reaches its zenith
here. It is not a question of an ethnic group, of an ancient race,
but of a group that retains the prestige of its origins and its virtues
as a group, mythologised through literature and re-created using
clichés that glorify their magnificence; although this group has grad-
ually incorporated members of the other majority group in the land
where they live, the Hispano-Romans. The Goths as a group pre-
serve the distinctiveness of their origins, even after generations on
Spanish soil,—now definitely in their hands—as a mark of superi-
ority and their leading freemen, the nobility with its factiones, the
high functionaries of the state, in sum the most “distinguished” men
in this society, exercise power, control the king and exert the weight
of their gens (now practically a class) at this and other councils.
I believe that the political formulation of the 4th Council is moti-
vated, even forced, by the contemporary situation, that is the polit-
ical disgrace of Swinthila and the rise to power of Sisenand, as well
as by the pressure exerted by the factiones of the nobility. It also high-
lights what will be even more noticeable from now on: the tension
between monarchy and nobility (and the Church) and between the
different factions of this nobility. Royal families will measure their
strength to use against that of the other powers of the regnum, and
the Acta of the Council and legislation will be a good reflection of
the balance of power at any one time. But it is obvious that such
formulation is the crucial step to achieve the much-longed-for unity
of all of Spain’s peoples. Isidore, who certainly considered it very
difficult to achieve and has worked towards this end, as we have
seen, succeeds only partially in fulfilling his ambitions although his
contribution to this unity has been decisive. The impact of events,
   205

the evolution of society and political circumstances will combine in

successive years to highlight the increasing existence of forces in
favour of integration that would effectively lead to an identification
of Hispania (Spain) with the regnum and of the gens and all her inhab-
itants, overcoming that division between Gothi and Hispani, under the
rule of a single rex. But there are also forces that favour segregation,
and tend to split that unity, to a great extent due to personal and
collective ambitions to seize the throne and control the regnum.

Patria et gens: Attempts of unification

This threat of fragmentation and the tensions between the different

noble groups to which I have referred compel us to insist upon and
define with greater precision in successive councils those who are
eligible to accede to the throne, seek the protection of the families
and the fideles regis, secure awards and honours for the latter, the
ones who protect the king. At the 5th and 6th Councils of Toledo,
celebrated under Chintila’s rule, attention is again drawn to these
issues; seditious attempts are condemned and the question of who
can accede to the throne is settled: those who have been elected and
are of Gothic extraction. At the 6th, other requirements are added:
they must be of honourable conduct and it is explicitly stated that
no one who has been tonsured or whose scalp has been shaved can
become king, nor those of lowly or foreign origin, doubtless indi-
cating fears that could be regarded as well-founded and genuine
restrictions on gaining power:116
V Toledo, c. 3, p. 282,81–4: [. . .] quem nec electio omnium provehit nec
Goticae gentis nobilitas ad hunc honoris apicem trahit [. . .].
VI Toledo, c. 17, pp. 326–7,341–345: [. . .] Rege vero defunctus nullo tyran-
nica praesumptione regnum assumat, nullus sub religionis habitus detonsus aut

116
Higher and perhaps more controlled than that of the Gothic origin as such.
Perhaps here the Goticae gentis nobilitas refers to the “nobility” of status or class, char-
acteristic of the gens Gothica (= ruling class of Gothic extraction, but also already very
mixed). It seems that the emphasis is more on a class nobility, of righteous cus-
toms, superior to all other persons, only nominally identified as Gothic, as opposed,
for example, to foreigners, Franks, Byzantines, etc. See M. de Jong, “Adding insult
to injury: Julian of Toledo and his Historiae Wambae”, The Visigoths. From the Migration
Period to the Seventh Century. An Ethnographic Perspective, ed. P.J. Heather (San Marino
1999) pp. 373–89, esp. p. 383.
206  

turpiter decalvatus aut servilem originem trahens vel extraneae gentis homo, nisi
genere Gothus et moribus dignus, provehatur ad apicem regni.
However, it must be pointed out that at the 5th Council reference
is made to the gens and the patria in a generic way, without speci-
fying the Goths. I believe that it indicates, for the first time, the fact
that the sources are beginning to show that this gens lack a distinct
ethnic origin but, rather, that it comprises the people (= human group)
dwelling in the regnum. It is a ruling117 issued by the king enjoining
all provinces to celebrate the practices of the new litanies. The begin-
ning of the text is revealing: “As it is the duty of every good prince
to safeguard the welfare of the motherland and his people [. . .]”
(Quum boni principis cura omni nitantur vigilantia providere patriae gentisque
suae comoda). The expression patriae gentisque, inherited from the pre-
vious Council, appears now without the appelative Gothorum.118 It is
clear that Chintila is not only legislating, or better sanctioning, a
proposal issued by the Church, for all the population but, rather,
that he means all of them. He commands all the provinces in his
kingdom ( per omnes regni nostri provincias) to celebrate the litanies and
rules that all the prominent men in the kingdom (tam obtimatum quam
comitum, iudicum etiam ceterorumque ordinum praecipua), are obliged to
defend the motherland ( patriae nostrae), and to ensure that everyone,
without distinction based on class, sex or age, can fulfil them. The
patria clearly reveals, on its part, the two facets of the Roman con-
cept, the concrete territorial aspect, and the abstract notion of polit-
ical unity. At the 6th Council, canon 14, when dealing with the
commendations for the fideles regis, the king’s obligation to protect
those loyal to his predecessor, without depriving them of their priv-
ileges and possessions, so that they can be treated according to their
individual value and in a way deemed necessary for the mother-
land (p. 320,227) is acknowledged: sicut eos prospexerit necessarios esse
patriae.
As is well-known, the threats and fears of sedition culminate in
the elderly Chindaswinth’s coup against the young Tulga and the

117
Although not explicitly formulated, I think that, rather than a sentence, it
takes the form of a lex in confirmatione concilii, see Velázquez, “Impronta religiosa”,
pp. 109–11. Not included in the edition by G. Martínez Díez and F. Rodríguez,
as it was not transmitted by the Hispana. Text in Concilios visigóticos e hispano-romanos,
pp. 231–2.
118
The possessive suae is, nonetheless, symptomatic.
   207

extremely harsh law (Lex Visigothorum 2,1,8) against those who, hav-
ing conspired against the king, have gone into exile, within or out-
side Spain (ad adversam gentem vel extraneam partem). The punishment is
the death penalty or, at the very least, blinding, and all the mag-
nates, and bishops and the entire people are made to swear alle-
giance to the king. The so-called “purges” of Chindaswinth during
which, according to some historians, the number of Visigothic aris-
tocratic families could have been halved, led to the replacement of
the nobility of aristocratic origin by the primati palatii, that is a new
political and administrative oligarchy, based on loyalty to the king,
the power of the army and the establishment of the state adminis-
tration. That does not mean that the new mighty men were not
overwhelmingly noble, many of Gothic extraction but, rather, that
the concept of new power class is a response to other premises. The
new situation also has an effect on the king’s interventionism in the
Church, now undergoing a period of some weakness. The 7th Council
of Toledo takes place almost nine years after the previous one (in
646), with few bishops and with one fundamental purpose, to sanc-
tion the aforementioned law. It is clear that the morbus Gothorum is a
widespread affliction. The poor attendance of bishops at the Council,
especially from Tarraconense and Septimania, evinces those areas’
reluctance to participate and the opposition of considerable segments
within the nobility and the Church to the measures against the nobil-
ity adopted by Chindaswinth.
At this Council, again apropos the attempts to usurp the throne
and assassinate the king, the already customary formula that defines
the shape of the kingdom’s power—against whose security no action
must be taken—is repeated. Expressions such as the following include
the formula of canon 75 from the 4th Council of Toledo and, more
precisely, the third proviso which, as we showed, highlighted the
subtle difference in the co-ordination of elements, linking, on the
one hand, patria and gens and, on the other, both and rex:
VII Toledo, c. 1, p. 341,46–47: quod gentem Gothorum vel patriam aut regem.
VII Toledo, c. 1, p. 341,52: genti Gothorum vel patriae aut principi.
VII Toledo, c. 1, p. 343,72–73: contra gentem Gothorum vel patriam seu regem.
A different syntactic combination that, in spite of appearing to be a
simple stylistic variation at the aforementioned 4th Council vis a vis
patriae gentisque Gothorum of the first two formulas used, may imply a
different political concept, given that patria takes on a more generic
208  

sense; we could say that the latter involves the motherland common
to all, whereas gens Gothorum refers to the ruling class, still using the
“ethnic” description as a symbol of prestige, albeit deprived of that
sense of exclusive (and patrimonialist) identification of patria Gothorum
with Hispania, complemented by other uses of patria, not linked to
gens. And both—and that was the case already at the 4th Council—
were distinct from the king. A similar formulation, with a different
use and syntactic arrangement but of the same type, appears again
at the 8th Council of Toledo, convened by Recceswinth in 653, two
months after the death of his father Chindaswinth,119 when the col-
lective oath to the king is evoked in the royal Tomus (p. 375,112–116):
ut cuiusquumque ordinis vel honoris persona in necem regiam excidiumque Gothorum
gentis ad patriae detecta fuisset [. . .]. The Council takes up this issue
without using the exact wording, referring to gens without the adjec-
tive Gothorum. It mentions, in a positive sense if you wish, the oaths
of allegiance to the king, and the obligation to fulfil them but ignor-
ing the magnitude of the penalty imposed (c. 2, p. 403,418–423):
pro regiae potestatis salute vel contutatione gentis et patriae [. . .].
As is well known, Recceswinth has convened this Council to sub-
mit to the bishops’ consideration the possibility of annulling or allay-
ing the terrible effects of his father’s law. In addition, succession to
the throne, albeit peaceful and preceded by a protracted period of
co-regency,120 has been hereditary and this mode of succession clashed
with the interests and wishes of the nobility and, possibly, with those
of large sectors within the Church—to the extent that the procedure
governing succession to the throne, based on selection by consensus
among the leading men of the palace and the bishops, either in the
urbs regia or wherever the predecessor has died to preclude seces-
sionist moves, is reiterated (c. 10, pp. 428–9,689–692): in urbe regia
aut in loco ubi princeps decesserit cum pontificum maiorumque palatii omnimodo
eligantur adsensu, non forinsecus aut conspiratione paucorum aut rusticorum
plebium seditioso tumultu. The delicate issue is also mooted of the king’s

119
Celebrated at the praetorian Church of the Saint Apostles, a definite symbol
of the aulic power and the close relationship between monarchy and Church, under
the leadership of Eugenius II of Toledo. On the seats of the councils, see Velázquez
and Ripoll, “Toletum: la construcción de una urbs regia”.
120
And instigated by the Church itself, in particular by Braulius of Saragossa
and Eutropius who, together with Celsus, a high dignitary from the Tarraconense,
sent a petition (suggerendum) to Chindaswinth himself so that he, given his advanced
age, could tie his son to the throne, Braulius, Ep. 37.
   209

ownership of the patrimony acquired by his father through violent

and unjust methods, which ought to be restored. Although a com-
promise solution is reached, a difference between the patrimony of
the crown and the personal patrimony of the monarch is established.
The Council’s crucial decisions are a reflection of the tense situa-
tion generated by his father’s policies. It is likely that pressure from
the aristocracy, Froia’s attempt to usurp the throne, and even the
actions of individuals such as Fructuosus of Braga, still abbot of
Dumio, who sends a missive to the king and the bishops calling for
clemency),121 force Recceswinth to address the assembly in a man-
ner far removed from his father’s authoritarianism, theoretically
endorsing the Council’s measures. He is keen to re-establish certain
links with the nobility, punished by his father, and reconcile the
different sectors, which are gradually becoming more powerful and
now not only attending the Council but also signing the proceed-
ings as fully lawful members of the Council. The king presents a
Tomus in which he presents an image of himself as a Christian prince
(an image also endorsed by the assembly itself ), copying imperial
models; his speech can be considered as a short treatise on the
monarchy. The latter takes on juridical legitimacy through deeds, by
law rather than through the person of the king. The king must be
obeyed by virtue of his role, not his person: non personae sed potentiae
suae haec deberi non ambigit. Regem etenim iura faciunt non persona. These
words are redolent of Isidore’s doctrines, as we have already seen.122
The Tomus also shows the role as legislative chamber and judicial
tribunal that Recceswinth envisages for the Council assembly, which
he requests to amend the legal texts as they deem appropriate. This
and the references by Braulius of Saragossa (Epist. 39–41) reveal
the close links between the creation of the Liber and the Council’s

121
Epistulae Wisigothicae 19, ed. J. Gil, Miscellanea Wisigothica, Publicaciones de la
Universidad de Sevilla (Sevilla 1972). I believe that this occurred on the occasion
of the announcement of the Council’s celebration, see Velázquez, “Impronta reli-
giosa”, p. 112. Doubtless, during Recceswinth’s rule a number of ecclesiastical figures
and writers had a crucial role in the political scene; they shaped, in my view, the
political views of the king in some measure.
122
And which are complementary to those regarding the laws (Etym. 2,10,3) and
would crystallise in the Titulus 2 of book 1 of the Liber Iudicum, ed. K. Zeumer,
MGH LL nationum Germanicarum 1 (Hannover 1902) [henceforth: Lex Visigothorum],
see Velázquez, “Impronta religiosa”, pp. 104–5. On the king’s speech, based on
that of Emperor Justinian (Corippus, Eloge de l’empereur Justin II 2,2, ed. S. Antès
[Paris 1981]), see the elucidating analysis by Teillet, Des Goths à la nation Gothique,
pp. 537–71.
210  

prescriptions and the central role played by some members of the

Church hierarchy. The promulgation of the Liber Iudicum, begun
under Chindaswinth and completed under Recceswinth, constitutes
not only the complete renovation and strengthening of one of the
main instruments of power, the legislative body, but also the most
effective means and the most far-reaching political gamble under-
taken by the monarchy of Toledo in an effort to achieve the com-
plete unification of the regnum under royal auspices. This instrument
would be renovated and widened by successive kings, in particular
by Erwig and Egica, who would also utilise several councils to pro-
cure the Church’s intervention in the drafting or supervision of laws
and the promulgation of their leges in confirmatione concilii. The leg-
islative role played by the assembly will be maintained at later coun-
cils. Thus, Erwig asks those gathered at the 12th Council of Toledo
(681) to correct unfair laws and beseeches the bishops and Spain’s
higher dignitaries (religiosi provinciarum rectores et clarissimorum ordinis totius
Spaniae duces)123 to draft the new laws clearly (evidentium sententiarum
titulis [. . .] conscribite).
At the 8th Council, the tension between the king’s interests, on
the one hand, and those of the nobility and the Church on the
other, is palpable, although the Church does not abandon its role
as mediator. Recceswinth’s apparent respect for conciliar norms is
less noticeable when it comes to promulgating the law Lex Visigothorum
2,1,6, which he alters in his favour and traces the issue of acquisi-
tion of the royal patrimony back to Chintila’s rule, as well as side-
stepping the grave allegations against his father. What is more, he
does not convene any new councils. Only three sparsely-attended
provincial assemblies would take place, although these return insis-
tently to the issues pertaining to the usurpation of the throne, the
safety of the fideles regis and the patria, specifically on military mat-
ters. The 10th Council of Toledo (656) refers to the decrees and

123
Erwig sanctions a group of twenty-eight laws against Jews, replaces Wamba’s
military law (Lex Visigothorum 4,2,8) with another (Lex Visigothorum 9,2,9). At the 13th
Council (a. 683) he issues a lex in confirmatione concilii, incorporated into the Liber (Lex
Visigothorum 12,1,1). Egica, to a lesser extent, will use the 16th Council (a. 693) to
undertake his legal reforms, sanctioning the Lex Visigothorum 12,2,18 against the Jews
and ratifying the Lex Visigothorum 2,5,19 on the bad use of the so-called habeas cor-
pus, issued at the 13th Council. See Velázquez, “Impronta religiosa”, pp. 117–21;
Orlandis and Ramos-Lissón, Historia de los Concilios, pp. 415–29; 494–6.
   211

laws issued to preclude anyone from acting “against the princes’ wel-
fare, and that of the people and the motherland”.124 At the Council
of Mérida (666), in canon 3, it is stated that everyone must con-
tribute to the “king’s prosperity and that of his loyal men from
among the gens, and of the motherland!: [. . .] rege fideliumque suorum
gentis aut patriae debeant prosperitatem,125 so that it stipulates (instituit) that
one must pray every time the king goes with his army on campaign
against the enemies, “for his wellbeing, that of his supporters and
of the army”: pro eius suorumque fidelium atque exercitus sui salute.
At the councils celebrated under Recceswinth, as before during
the 5th and 6th, references to patria and gens, without mentioning
Gothorum, which are also found in later councils, with a few excep-
tions and the kind of formulations used, following the trilogy regnum,
gens, patria, albeit with slight alterations, reveal, I believe, that aware-
ness of a new political order is taking shape. In fact, the role of rex
has been exemplified most of all at this Council by the king and
the assembly. The gens is no longer defined by virtue of their eth-
nicity but by their political adherence and seems to encompass the
entire population in most of the phraseology, although its main use
in contexts dealing with the usurpation of power (where, on occa-
sions, reference is also made to plebs or rustici ) suggests that they
actually have in mind the higher classes of that gens (= natio), that
is the old and new aristocracies of the officium palatinum as well as
the ecclesiastical hierarchies, the only groups that, whether or not
they drag the populi or the rustici with them, can jeopardise the sta-
bility of the motherland, the gens and the life of the king himself.
The patria, on its part, retains its territorial sense but is formally
linked—although with shades of meaning, as I have pointed out—
to the gens in a juridical conceptual unity.

124
With a variation: principum gentisque aut patriam that could be compared to the
option of gens sua adopted before the aforementioned 5th Council (see note 118).
Does that non-explicit consciousness of the rex and his gens, apart from the mother-
land, still exist?
125
I think these fideles regis are the entourage of primati who have sworn loyalty
to the king, not the subjects in general, as Vives interprets it, cf. Concilios visigóticos
e hispano-romanos, p. 327.
212  

Illusions and disillusions of the Church

The disillusion that the Church must have experienced when faced
with the loss of vitality and frequency of councils, due to Recceswinth’s
action perhaps because of the suspect result of the 8th Council of
Toledo, and also of the tense situation at the time can be perceived
at the 9th Council of Toledo, led by Wamba, where it is stated that
the “light of the councils has been extinguished” by circumstances
and the passage of time. The new king is presented as the religiosus
princeps. For posterity, the image of this king would always be that
drawn by Julian of Toledo in his Historia Wambae, a work conceived
as an exemplum, almost a vita, belonging to the genre of biblical evo-
cation and of a panegyric nature, as Teillet has accurately analysed.126
This religiosus princeps has all the Christian virtues and is in opposi-
tion, like are the absolutes good and evil, to the tyrannus Paulus who
has tried to take the throne, or perhaps even to secede from it and
has proclaimed himself rex of the Pars Orientalis.127 Above all, Wamba
is the king anointed by God. He has been sacrally anointed by the
bishop Quiricus in Toledo and has refused to become king until
being designated by this ritual. Now the monarchy has acquired a
sacral nature, its power emanates from God, the king is chosen by
divine will, as shown by the unction, although he has also been cho-
sen by all the nation ( gens et patria, following the model of an impe-
rial monarchy and by law, such as that of Recceswinth), and his
rule is sought by the people, even foreseen in prophecy:
Hist. Wamb. 2: Adfuit enim in diebus nostris clarissimus Wamba princeps, quem
digne principari Dominus voluit, quem sacerdotalis unctio declaravit, quem totius
gentis et patriae communio elegit, quem populorum amabilitas exquisivit, qui ante
regni fastigium multorum revelationibus celeberrime praedicitur regnaturus.
Julian of Toledo’s work is a passionate hymn not just to the king
but, I believe, most of all to the exercitus Hispaniae, and the mother-
land herself, Hispania, as opposed to the gentes externae, Francia, even
Gallia (the old Narbonensis) which, albeit part of the regnum, is here
denigrated and insulted by the author. It can be said that it is a
“nationalist” work, or that it exhibits a pronounced adherence to

126
Teillet, Des Goths à la nation Gothique, pp. 585–636. But see also the extremely
interesting contributions of De Jong, “Adding insult to injury”.
127
Velázquez, “Wamba y Paulo”.
   213

“Hispanic nationalism”. It seems clear that here the concept of gens

et patria already comprises all the population; in fact, as I have
remarked, it deals mainly with Hispani and the exercitus Hispaniae as
opposed to Francia. Now, references to Gothi and gens Gothorum are
few and quite symptomatic. In chapter 2 he claims that all voices,
united, want Wamba, and no-one else from the gens Gothorum, to
become king,128 when all acclaim Wamba. The other context is that
of the struggle between armies (Hist. Wamb. 9; 10; 12; 13), where
Gothi appear opposed to Franci or to Vascones, and identification with
which the author is in complete accord when referring to the army
as “our men”, it is the army of the Goths, the army of Spain who
are “our men” (Hist. Wamb. 9; 10; 12; 13; 18). The third context is
the emphatic and literary evocation of the “ethnic myth”: the allu-
sion to the gens and their courage, to the famosissima virtus of the
Goths, remarked upon by the tyrannus Paulus himself (Hist. Wamb. 16;
17). In fact, the work is an ode to the gens and the patria against the
enemy. God leads Hispania’s armies to victory, with their anointed
king at their head. The entire work is a model of contrasts between
a “good and bad ritual” of royal anointing and, in particular, a dis-
play of the gens of Hispania as a new biblical Israel, as a people cho-
sen by God.129
However, we must stress that the initial enthusiasm of the Church
manifested both by the 11th Council and the Historia Wambae, must
have given way to disillusion as the king tried to control the clergy
and prevent the accumulation of property in ecclesiastical hands. In
a parallel development, the “protofeudalisation” of the state and the
weakness of the central monarchy must have been evident. Wamba’s
loss of power can hardly be more suspicious: his grave illness, the
public penitence that made his return to power impossible and the
accession of Erwig to the throne constitute the “most convincing
proof of the failure of Wamba’s centralist and, to a certain extent,

128
This may be related to the restriction that always prevailed whereby kings
could only be Gothic, apart from some partial exceptions such as Theudis (Ostrogothic),
or Erwig (Byzantine), see Díaz, “Visigothic Political Institutions”, pp. 339–40.
129
De Jong, “Adding insult to injury”, p. 378. A ritual conceived by the great
expert on liturgical rites. And a “good ritual” that will occur later with Wamba’s
public penance that forced him to abdicate. Barbero, “El pensamiento”. On the
possible dating of Julian’s work, in relation to the Insultatio and the Iudicium, besides
De Jong, see Y. García López, “La cronología de la Historia Wambae”, Anuario de
Estudios medievales 23 (1993) pp. 121–39.
214  

anti-noble policies”.130 Growing disillusionment is already reflected

by the 12th Council of Toledo (681) under Erwig, presided over by
Julian, the friend of the king and perhaps an accomplice in his con-
spiracy, who, to put it no stronger than this, did not see the influence
he had enjoyed under Wamba curtailed during the new king’s rule.
I have signalled this Council’s importance apropos legislative activ-
ity and the consultative nature and power enjoyed by the Council
Assembly. It is obvious that Erwig seeks to confirm the legitimacy
of his accession to the throne and tries to secure the support of the
Church while simultaneously granting privileges to the nobility to
ensure the stability of the kingdom. Now, and increasingly under
later kings, we see the need to gain the support of the fideles regis
and the Church in the face of the growing feudalisation of the State
and aristocratic power over groups and military clientelae in a pseudo-
lord-vassal system. It is significant that, in an almost inversely pro-
portional way, as the state is gradually more fragmented and the
monarchy weakened, the texts of the councils reveal an ever more
refined notion of gens et patria allied to the king.
Erwig addresses the 12th Council in his Tomus saying that the
acceptance of the Council decree provides “enormous health to the
peoples and the kingdom of the gens”: magna salus populis gentisque nos-
trae regno conquiritur.131 In canon 1 (p. 387), the assembly stresses the
need to obey Erwig and do whatever is necessary to ensure his safety
and that of the nation and that renders benefit to the motherland:
enitendum quidquid eius saluti proficit quidquid genti vel utilitatibus patriae suae
consuleat. And it is the populi who, at the 13th Council of Toledo
(683), seem to pay taxes ( populis fiscalibus), taxes that must be neither
too onerous nor suspended arbitrarily so as not to damage the sta-
tus of the gens: iudicium est quippe salutare in populis quando sic commissa
reguntur, ut nec incauta exactio populus gravet nec indiscreta remissio statum
gentis faciat deperire. The term gens is difficult to define at this point
as it can refer to the populi as a whole, understood as a political and
juridical entity, rather than to specific groups, but always bearing in

130
García Moreno, Historia, p. 176; Díaz, “Visigothic Political Institutions”, pp.
338–47; A. Barbero and M. Vigil, La formación del feudalismo en la Península Ibérica
(Barcelona 1978).
131
These populi and the habitatores of villulis vel territoriis sive vicis are part of those
drafted for the army who have been jeopardised by Wamba’s “military law”; Erwig
himself is also compelled to draft them (see note 123).
   215

mind that gens means the free people (and not the foreigner) and is
not independent of the nobility—based on class, not ethnicity—but
with the idea, however, that this entails adherence to a certain lin-
eage, based on social status. Thus, Erwig regards as “horrible and
despicable” (votis nostris horribile et animis execrabile) the idea that freed
men and serfs and those of noble status can be equated as this would
demean the status of the gens: [. . .] quum nobilitate conditio libertorum vel
servorum etiam adaequata gentis nostrae statum degenerat. For this reason
and with the exception of the servi fiscales, serfs and freed men are
banned from holding an officium palatinum. This ban is further elab-
orated in canon 6 (p. 482).
Obviously, this has nothing to do with a Gothic ancestry132 but,
rather, with the differences between freemen, serfs and freedmen. In
fact, Erwig puts more emphasis on populus than on gens. Thus, he
notes (p. 450) that before accepting the crown, he sought to bind
himself by means of an oath not to deny justice to the peoples
entrusted to him: ut iustitiam commissis populis non negarem. When he
submits to the Council the choice between obeying Wamba’s oath
(that is to protect the royal family) or safeguarding his people’s wel-
fare, the dichotomy is set between the proles regia and the populus and
its interests (negotia); between the fides unius domus regum and the promises
made to the plebs; between the amor privatus and the generalis patriae
et gentis affectio, in an exact equivalence between gens (et patria) and
populus. It is the same populus for whom his successor, Egica, implores
God’s assistance, as well as for himself, at the 14th Council of Toledo
(693): et mihi [. . .] et cuncto populo regiminis mei respectio divina semper opitu-
letur. And, although at this Council, because of the dismal situation,
sentences of anathema are repeated once again against the man who
tries to harm the rex, the gens Gothorum and the patria—the same for-
mulas uttered at the 4th Council of Toledo, and first repeated at
the end of Julian of Toledo’s Iudicium—, Egica invokes it, however,
in a lex in confirmatione concilii, without mentioning the Goths: ut si
quisque contra regiam potestatem gentem ac patriam nostram agere conatus exti-
terit [. . .], with a nostram qualifying patriam, rather than gentem, which
must not be overlooked.

132
Also, Erwig is the only king, as far as we know, with a different ethnic ori-
gin (apart from the Ostrogoth Theudis [see note 128]); he was a Byzantine, Ardabasto’s
nephew, expelled from Constantinople and a refugee in Spain. Erwig became “an
adopted Goth” by marrying one of Chindaswinth’s nieces.
216  

Conclusion: The Spanish nation, a still unfulfilled dream

Almost sixty years have elapsed since the theoretical formulation of

the 4th Council of Toledo. It is likely that now, all the monarchs
of Toledo have a clearer notion that the regnum Gothorum is Hispania
and the latter is the nation they rule, with all her peoples. The gens
that constitutes these peoples is far removed from the gens denoting
a lineage or an origin. And the equivalence gens = populus is mani-
fest, although those who still count and hold power are the nobles
and the Church, that is the prominent men of the kingdom. But it
is their power, or that of some of them, that endangers the unity of
the motherland and the stability of the whole nation (= gens), hence
the obsessive use of that formula rex et patria ac gens. An idea of unity
and nation championed by Toledo’s centralist monarchy that failed
to prosper because the fragmenting forces of sectors of the nobility
and dissident groups, sometimes enjoying more popular support than
others, regularly strove to overturn it.
This conception of a Hispania as the motherland of a new gens
Gothorum (= Catholic people, governed by their new lords, the Goths)
and ruled by a rex from that gens, would gradually be interpreted as
the motherland of a gens Gothorum retaining the prestige of a lineage
and maintaining a “social memory”, but absorbing the clergy and
the Hispano-Roman nobility from amongst the ruling class, whilst
the population will remain indifferentiated (apart from serfs, freed-
men and minorities like the Jews). A gens that intellectuals, writers
(and at the same time clerics) present as the heir of Roman culture
after having made the mater Spania their regnum, an ancient, coura-
geous and strong gens, superior to Rome and no longer barbara because
they have embraced the Catholic faith and have become the new
people of the Bible, united in the faith to the other gentes.
That creation of a new Hispania (= regnum Gothorum), the secular
motherland of the natives of the Iberian Peninsula, where a gens
already made up of gothi et hispani populi133 live, under the single and
centralised rule of a powerful rex, a religiosus princeps, blessed by God
and the Church, was a dream that Leander of Seville timidly dared

133
Where the Jews constitute a minority increasingly subject to persecution, a
number of kings excepted, and where Gallia is an annexed province, an area of
conflicts which will end up being denigrated.
   217

to formulate at the 3rd Council of Toledo and his brother Isidore

thought capable of being realised, joining their praises of Spain and
her rulers. A dream that royal power tried to define and put within
political boundaries in the Council regulations and in legislation,
gradually abandoning the “ethnic discourse”, increasingly emptier of
content, although at the same time formulating the genuine identity
of the gens and the motherland through its ruling class.
Hispania (or if one prefers, Spain as a nation) was Chindaswinth’s
true political project and, most of all, Recceswinth’s; both tried to
issue laws for the new society of the period. A dream that could
have become true had it not been for the tensions among nobles,
the lust for independence of local aristocracies, the increasingly ambi-
tious aspirations of very powerful figures with clienteles and “almost”
their own subjects and the gradual weakness of a monarchy with its
seat in Toledo.
A dream eagerly pursued, with ever more zeal as it grew increas-
ingly impossible to fulfil, by the latter kings, from Wamba and
Toledo’s Church itself with Julian at its head. And a dream, to sum
up, that failed despite having preoccupied many minds. The con-
cept of Spain as a nation was born, but this was not the time to
make the dream come true.

[Translated by Francisca Rodríguez-Manas]

This page intentionally left blank
 THE TRANSFORMATION OF HISPANIA AFTER 711

Ann Christys

At first sight, it seems rather eccentric to include a study of the Arab

conquest of Hispania in a discussion of the relationship between gens
and regnum in the development of the Germanic states that succeeded
the Roman Empire. After all, Hispania had emerged from the up-
heavals of the Migration period as a Visigothic kingdom. Surely the
conquest of 711 was something completely different? In his re-cent
account of Muslim Spain and Portugal, Hugh Kennedy described it as
“the logical and necessary extension of the conquest of North Africa”
which “had been achieved by an alliance of Arabs and Berbers in
the name of Islam”.1 Thus the peninsula was transformed almost
overnight—or so it seems—from the Visigothic successor state to the
Roman empire into part of a new, Islamic empire. The conquest of
711 was indeed the start of a new era in the peninsula. For the cen-
turies to come, most of her history would be written in Arabic. It
is not surprising that al-Andalus looks very different from the world
of the Visigoths. Yet the Arabic histories are not our sole source for
the conquest. Roger Collins based his account of The Arab Conquest
of Spain2 on a Latin chronicle, known as the Chronica Muzarabica, or
the “Chronicle of 754”, because it ends in that year,3 rejecting the
Arabic evidence as fiction. The “Chronicle of 754” should not be
read merely as a narrative of the conquest period untainted by later
legends. It lends itself to the methodological approach that has been
adopted by historians of the Transformation of the Roman World.
Indeed the chronicler uses the very same vocabulary of gens and
regna. Such a reading can be extended to the Arabic histories, which
are also concerned, in their very different way, with the ethnogen-
esis of the people of al-Andalus. My aim in this paper is to present

1
H. Kennedy, Muslim Spain and Portugal (London 1996) p. 5.
2
R. Collins, The Arab Conquest of Spain, 710–797 (Oxford 1989).
3
Chronica Muzarabica, ed. J. Gil, Corpus Scriptorum Muzarabicorum 1 (Madrid
1973) pp. 15–54; transl. K.B. Wolf, Conquerors and Chroniclers of Early Medieval Spain
(Liverpool 1990) pp. 111–58 [henceforth: transl. Wolf ].
220  

both the “Chronicle of 754” and a ninth-century Arabic history of

al-Andalus, by 'Abd al-Malik ibn Óabìb,4 from this perspective. I
hope to bring the conquest back into the mainstream of early medieval
European history and to throw some light on the view of the con-
quest which historians continue to draw from the Arabic sources.5
Both the Latin and Arabic sources agree that the invaders came
from North Africa and were Berbers and Arabs. The conquest began
with a raid or a series of raids across the straits of Gibraltar. The
most significant of these was led by ˇàriq ibn Ziyàd, who was shortly
afterwards followed by Mùsà ibn Nusayr, the governor of North
Africa appointed by the caliph in Damascus. Within four years, the
new arrivals seem to have controlled the southeast of the peninsula,
advanced as far north as Zaragoza and the Picos de Europa and
pushed on over the Pyrenees. They took Narbonne, and raided
Aquitaine and Provence until—at least according to the Frankish
histories—Charles Martel brought this expansion to an end with his
victory between Tours and Poitiers, in 733 or 734—the battle which
Gibbon represented as the decisive clash of the two empires of
Christendom and Islam.
This much is fairly uncontroversial, but the details are obscure
and open to debate. Historians writing in Arabic, often many cen-
turies after the event, gave an implausibly elaborate account of the
period. They attributed the conquest of the peninsula to the actions
of a certain count Julian, who betrayed Hispania to the Arabs because
the last Visigothic king, Rodrigo, had raped Julian’s daughter, en-
trusted to his care. Conquered Hispania yielded fabulous treasures.
In Toledo, Mùsà found “gold and silver and emeralds and rubies
and other things God alone knows”, and the crowns of his Visigothic
predecessors.6 Here Mùsà opened the miraculous House of the Bolts.
The story goes that each Visigothic ruler had to add another bolt
to the door of this house, but he was forbidden to see what was
inside. When Rodrigo entered the chamber against the advice of his
ministers, his downfall was inevitable. Intense rivalry developed
between ˇàriq and Mùsà, as each sought recognition from the caliph
in Damascus as the conqueror of al-Andalus. Central to this claim

4
Ibn Óabìb, Kitàb al-Ta"rìkh, ed. J. Aguadé (Madrid 1991).
5
See e.g. Kennedy, Muslim Spain and Portugal, pp. 10–3 and 18–29.
6
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 140.
    221

was that both men said they had discovered the Table of Solomon
in Toledo. In a story with many parallels, Mùsà presented the Table
to the caliph, but one of its legs was missing. ˇàriq’s triumphant
production of the missing leg contributed to Mùsà’s disgrace. Some
of these stories made their way into Latin histories, although prob-
ably not until the eleventh century or later. They are not in the
“Chronicle of 754”, and attempts to substantiate the legends from
the chronicle, for instance by linking Julian with a North African
bishop Urbanus who accompanied Mùsà to Spain, are unconvinc-
ing.7 These stories are usually left out of the narrative of the con-
quest in favour of episodes which are more plausible, although equally
unsubstantiated, but they are part of the history of al-Andalus; the
origins and possible meaning of some of them will be considered
later. First, though, I will introduce the relatively sober account of
the “Chronicle of 754”.
The “Chronicle of 754”, which ends in that year, covers almost
the whole of the period which modern historians have labelled “the
period of the governors”, from 711 to the establishment of 'Abd al-
Ra˙màn, the first Umayyad ruler of al-Andalus c. 756. This is prob-
ably a coincidence, as this periodisation is to some extent an artifact
of the creation by later Umayyad propagandists of foundation myths
about 'Abd al-Ra˙màn’s arrival in the peninsula. The names of the
governors are preserved in three slightly different versions in the
“Chronicle of 754”, the “History” of Ibn Óabìb and the “Prophetic
Chronicle”, written in the Asturias in the 880s.8 The concordance
between these three apparently independent sources suggests a cer-
tain core of fact, although the chronology of the conquest remains
imprecise.
The “Chronicle of 754” survives in two versions. The earliest is
incomplete and consists of six folia, now divided between Madrid
and London,9 which may date from the middle of the ninth century.10

7
Chronica Muzarabica, p. 35; Collins, The Arab Conquest, p. 36.
8
See the Appendix; I have not been able to consult G.V. Sumner, “The chronol-
ogy of the governors of al-Andalus to the accession of 'Abd al-Ra˙màn I”, Medieval
Studies 48 (1986) pp. 422–69.
9
B.M. Egerton 1934; Biblioteca de la Academia de la Historia 81.
10
M.C. Díaz y Díaz, “La transmisión textual del Biclarense”, Analecta Sacra Tarra-
conensia 35 (1962) pp. 57–78, here p. 71.
222  

Another version was copied into a thirteenth-century codex now in

Madrid.11 This codex is a compendium of histories which stretch
from Eusebius to the chronicle of Ricardo Pictaviense continued to
1244 and includes the sixth-century “Chronicle” of John of Biclar,
and Isidore’s “Chronicle” and “History of the Goths, Sueves and
Vandals”. The “Chronicle of 754” was also copied into at least two
other codices of histories. One, from Alcobaça, disappeared some-
time after 1622,12 but a fourteenth-century manuscript from Navarre
survives in Paris.13 The vagaries of manuscript survival have left only
one apparently complete version of the “Chronicle of 754”, which
make the value of the chronicle as evidence for the first half of the
eighth century difficult to interpret. But the surviving version con-
cludes with a paragraph in which the chronicler calculated the age
of the world to the point where his history stops, in “the tenth year
of the emperor Constantine, the fourth year of 'Abd Allah the
Amiralmuminim, the seventh year of Yusuf in the land of Spain and
the one hundred and thirty-sixth of the Arabs”. This allows us to
read the “Chronicle of 754” as being based on an eighth-century
original, although both the language and content of that original
may have been altered in transmission.
The “Chronicle of 754” begins with the accession of Heraclius in
Era 649 (611; the accepted date is 610). It is structured by Byzantine
imperial accessions and the author included material on the Byzan-
tines, the Islamic conquest of Syria and the early caliphs. In general,
the chronicler’s concern was not with these events as such. He used
them to establish a chronological framework, on which he hung a
history of the Visigothic kings, using his eastern material to turn the
history of Hispania into a universal chronicle. Apart from a few dis-
crepancies, he was successful in tying in Byzantine imperial regnal
years with Hispanic Era dates, the regnal years of the caliphs, the
years of the Hegira, and an annus mundi dating which began with
the creation of the world in 5200 B.C.14 For the period up to 711,

11
Biblioteca de la Universidad Complutense 116–Z-46 (Villa-Amil catalogue no.
134).
12
C. Cardelle de Hartmann, “The textual transmission of the Mozarabic chron-
icle of 754”, Early Medieval Europe 8 (1999) pp. 13–29; J.E. López Pereira, Estudio
Crítico sobre la Crónica Mozárabe de 754 (Zaragoza 1980) p. 8.
13
Library of the Arsenal 982.
14
Chronica Muzarabica, p. 35.
    223

the Byzantine, Visigothic and Islamic histories run in parallel, with

one chapter given to each in turn. After this date, the chronicler
concentrated exclusively on the history of the peninsula, although the
framework of Byzantine, hegira and annus mundi dates was maintained.
The “Chronicle of 754” was written in the tradition of provi-
dential history, which the author had inherited from Eusebius.15 This
was history with a message, the triumph of the empire of God, which
for Eusebius was co-extensive with the Roman empire. After the fall
of Rome, western historians preferred to give the role previously
played by the empire to the empire’s barbarian successors. At first,
it was not easy to envisage the Visigothic kingdom of Hispania as
God’s chosen instrument, since the persistence of the Arian heresy
made the Visigothic kings suspect in the eyes of the catholic clerics
who wrote the chronicles. It was only after the conversion of Reccared
to Catholicism in 589 that John of Biclar and Isidore were able to
write convincing providential history in the peninsula. One version
of Isidore’s “History of the Goths, Sueves and Vandals”, which ends
in the reign of Suithila,16 climaxed with the triumph of the Gothic
monarchy, the expulsion of the Byzantines, and Reccared as the new
Constantine. Hispania was firmly integrated into a universal history
of the Christian empire.
The dramatic events of 711 could be envisaged as part of the
same story. The “Chronicle of 754” was written as a continuation
of the “Chronicle of John of Biclar” and included excerpts from
Isidore’s “History of the Goths”. The chronicler showed how the
peninsula flourished under a succession of Visigothic kings, starting
with the pious and victorious Sisenand (612–621), until the kingdom
was destroyed by the ambition of the usurper Rodrigo, who “rebel-
liously seized the kingdom at the instigation of the senate”. At the
Transductine mountains, Rodrigo faced the invading armies of “Arabs
and Moors [. . .] and in that battle the entire army of the Goths,
which had come to him fraudulently and in rivalry out of ambition
for the kingship, fled and he was slain. Thus Roderic [Rodrigo]
wretchedly lost not only his rule but his homeland, his rivals also

15
J.N. Hillgarth, “Historiography in Visigothic Spain”, La storiografia altomedievale,
Settimane di studio del centro italiano di studi sull’alto medioevo (Spoleto 1970)
pp. 261–311.
16
R. Collins, “Isidore, Maximus and the Historia Gothorum”, Historiographie im frühen
Mittelalter, ed. A. Scharer and G. Scheibelreiter (Munich 1994) pp. 345–58.
224  

being killed [. . .],”17 and—lamented the chronicler—Hispania was

ruined:
Who can recount such perils? Who can enumerate such grievous dis-
aster? Even if every limb were transformed into a tongue, it would be
beyond human nature to express the ruin of Hispania and its many
and great evils. But let me summarize everything on one brief page.
Leaving aside all of the innumerable disasters from the time of Adam
up to the present which this cruel, unclean world has brought to count-
less regions and cities—that which, historically, the city of Troy sus-
tained when it fell; that which Jerusalem suffered, as foretold by the
eloquence of the prophets; that which Babylon bore, according to the
eloquence of the scriptures; that which Rome went through, martyri-
ally graced with the nobility of the apostles—all this and more Hispania,
once so delightful and now rendered so miserable, endured as much
to its honour as to its disgrace.18
The new masters of Hispania were clearly recognized as a gens,
labelled at different points in the chronicle either as Arabs or as
Saracens. They were usually, but not always, distinguished from the
Mauri, the Moors [or Berbers] of North Africa, who took part in
the conquest, but not in the formation of the regnum, as we shall see.
According to the “Chronicle of 754”, these same Arabs and Saracens
had brought about the downfall of Heraclius as a punishment for
his hubris after defeating the Persians.19 The chronicler dated the
first “year of the Arabs” from the date of their rebellion against
Byzantium “through the trickery of Mu˙ammad”,20 which he gave
as Spanish Era 656 (618), well before the date traditionally accepted
as the start of Mu˙ammad’s rise to power. The chronicler under-
lined that this was the beginning of a new kingdom by also dating
the foundation of their capital in Damascus to the same year.21 Yet
the chronicler did not link the conquest of Hispania with Mu˙ammad,
whose name is not mentioned in this section of the chronicle. There
is no sense of the arrival of a new religion. The chronicler proba-
bly knew something about Islam. A second eighth-century Hispanic
chronicle survives, probably written from the same eastern sources

17
Chronica Muzarabica, p. 31; transl. Wolf, p. 131.
18
Ibid., p. 33; transl. Wolf, pp. 132–3.
19
Ibid., p. 17; transl. Wolf, p. 113.
20
Ibid., p. 18; transl. Wolf, pp. 113–4.
21
Ibid., p. 19.
    225

as the “Chronicle of 754”,22 although it includes only one brief pas-

sage on Hispania after 711. In this chronicle, Mu˙ammad is described
as a prince “whom they honour and reverence to such an extent
that in all their sacraments and letters they affirm him to be an
apostle of their god”.23 Yet anti-Muslim polemic does not seem to
have been written in the peninsula until the middle of the ninth cen-
tury, when Eulogius characterized Islam as a heretical Christian sect.24
The lament for the destruction of Hispania reads like a much shorter
version of Gildas’ “Ruin of Britain”, written two centuries earlier—
a diatribe in which the failings of the conquered were of far greater
significance to the author than the nature of the scourge sent to
punish them.25 For the author of the “Chronicle of 754”, it was
Visigothic disunity which had occasioned the ruin of Spain, and
which would be the main reason why the invaders were able to
establish themselves. For Spain, said the chronicler, “was greatly
afflicted not only by the enemy but by domestic strife”.26 Rodrigo’s
enemies defected to the incomers. The chronicler described how
“after forcing his way up to Toledo, [Mùsà] imposed on the adja-
cent regions an evil and fraudulent peace. He decapitated on a
scaffold those noble lords who still remained, arresting them in their
flight from Toledo with the help of Oppa, king Egica’s son. With
Oppa’s support, he killed them all with the sword”.27 The treachery
of the Visigothic royal house was a recurring theme in both the Latin
and the Arabic accounts of the ruin of Spain. There is no sugges-
tion that apostasy to Islam played any part in this. Mùsà was accom-
panied throughout Spain by Urbanus “that most noble man of the
African region, reared under the doctrine of the Catholic faith”.28

22
Chronica Byzantia-Arabica, ed. J. Gil, Corpus Scriptorum Muzarabicorum 1 (Madrid
1973) pp. 7–14; for a comparison between the two chronicles see A. Christys,
Christians in al-Andalus 711–c. 1000 (Richmond, Surrey 2002) pp. 28–51.
23
Chronica Byzantia-Arabica, p. 9.
24
Eulogius, Apologeticus martyrum, ed. J. Gil, Corpus Scriptorum Muzarabicorum
2 (Madrid 1973) pp. 475–95, here p. 487; K.B. Wolf, “Christian views of Islam in
early medieval Spain”, Medieval Christian Perceptions of Islam: A Book of Essays, ed. J.V.
Tolan (New York-London 1996) pp. 85–108.
25
Gildas, The Ruin of Britain and other works, ed. and transl. M. Winterbottom
(London-Chichester 1978).
26
Chronica Muzarabica, p. 31; transl. Wolf, p. 132.
27
Ibid., pp. 31–2; transl. Wolf, pp. 132–3.
28
Ibid., p. 35; transl. Wolf, p. 134.
226  

The Arabs, or Saracens, and Berbers seemed merely to be moving

into a vacuum created by internal disorder.
The compiler of the “Chronicle of 754” used exactly the same
terms gens and regnum to describe the new regnum of the Arabs/Saracens
as he used for the regna of Byzantium and Visigothic Hispania. This
will be clearer from the following quotations from the chronicler’s
brief accounts of the Islamic conquests of Byzantium and Hispania.
Thus, in the fourteenth “year of the Arabs”, Amer ['Umar, 634–44]
gubernacula prioris suscipiens omnia populo decem per annos rigidus mansit in
regno.29 His successor 'Uthman [644–652/3] sue gentis administrationem
suscepit annos XII.30 During the reign of 'Uthman, the Arabs extended
their rule to North Africa, where Sarracenorum sociabit regimini et dicioni
subiecit. Here they faced a new enemy, but “the battle line of the
Moors [Maurorum] immediately turned in flight and all the nobil-
ity of Africa [. . .] was wiped out entirely”.31 When Walìd (705–15)
took up the sceptra regni Sarracenorum he ordered Mùsà ibn Nusayr,
the governor of North Africa, to attack the regnum Gothorum.32 Mùsà
in turn delegated the attack to ˇàriq ibn Ziyàd, who was described
in some of the Arabic histories as a Berber, although the “Chronicle
of 754” made no reference to his origins. Hearing of the invasion,
Rodrigo, “directed armies against the Arabs and the Moors sent by
Mùsà, that is against Tàriq ibn Ziyàd and the others”.33 After Rodrigo’s
defeat, Mùsà himself arrived in the peninsula, imposed a “fraudu-
lent peace” and established a regnum efferum based on Cordoba.34 As
we shall see, the chronicler makes several references to the taxation
and redistribution of land by the conquerors, to whom he attaches
ethnic labels. One of the governors, al-Sam˙ (c. 718–21) “divided
by lot among his allies the booty, arms and whatever else in the
way of plunder the gens Arabica in Spain had not yet divided”.35
There are few if any references to “strategies of distinction”36
between the inhabitants of the peninsula and the incomers. The

29
Ibid., p. 19.
30
Ibid., p. 21.
31
Ibid., pp. 24–5.
32
Ibid., p. 31.
33
Ibid.; transl. Wolf, p. 131.
34
Ibid., p. 32.
35
Ibid., p. 37; transl. Wolf, p. 138.
36
W. Pohl, “Telling the difference: Signs of ethnic identity”, Strategies of Distinction.
The construction of Ethnic Communities, 300–800, ed. id. and H. Reimitz, The Trans-
formation of the Roman World 2 (Leiden-Boston-Köln 1998) pp. 17–70.
    227

chronicler did not think it worth mentioning whether the Arabs and
Saracens spoke a different language from the inhabitants of the penin-
sula, or whether they knew Arabic. The references to interpreters,
found in later accounts of embassies between al-Andalus and Latin
and Greek-speaking rulers, are totally lacking from the “Chronicle
of 754” (and indeed from the Arabic accounts). As we have seen,
the religion of the conquerors is not mentioned. Similarly, the chron-
icler showed a total lack of interest in vestimentary markers.37 The
notion that the inhabitants of the peninsula would have instantly rec-
ognized the Arabs as different first appeared in the “History” of Ibn
Óabìb.38 It came in one of the most famous stories of the conquest—
the story of the House of Bolts in Toledo, which we have already
met. This may originally have been an Egyptian folk tale, and was
transmitted to Ibn Óabìb by the Egyptian scholar al-Layth ibn Sa'd
(d. 791).39 The following, shorter version was in-cluded in the “History
of the Conquest” by Ibn al-Qù†ìya (d. 977):
It is said that the Visigoth kings had a palace at Tulaytula [Toledo]
in which was a sepulchre containing the Four Evangelists, on which
they swore their oaths. The palace was revered and never opened. When
a king died his name was inscribed there. When Roderick [Rodrigo]
came to the throne, he put the crown on his head himself, which gave
great offence to the Christians; then he opened the palace and the
sepulchre, despite the attempts of the Christians to prevent him. Inside
they found effigies of the Arabs, bows slung over their shoulders and
turbans on their heads. At the bottom of the plinths it was written:
‘When this palace is opened and these images are brought out, a peo-
ple in their likeness will come to al-Andalus and conquer it.’140
Although the conquerors were portrayed as a scourge sent by God
to punish the sinners of Hispania, a careful reading of the “Chronicle
of 754” undermines the chronicler’s message. The identity of the

37
Pohl, “Telling the difference”, p. 42.
38
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 140.
39
M.A. Makki, “Egipto y los origines de la historiografía árabe-española”, Revista
del Instituto Egipcio de Estudios Islámicos 5 (1957) pp. 157–248, here pp. 178–9; trans-
lated by M. Kennedy in The Formation of al-Andalus, vol. 2: Language, Religion, Culture
and Sciences, ed. M. Fierro and J. Samsó (Aldershot 1998) pp. 173–234.
40
Ibn al-Qù†ìya, Historia de la conquista de España de Abenalcotía de Córdoba, ed. P.
Gayangos, E. Saavedra and F. Codera (Madrid 1868); repr. with Spanish transl.
by J. Ribera y Tarragó (Madrid 1926) p. 92; this passage transl. in Christians and
Moors in Spain 3, ed. C.P. Melville and A. Ubaydli (Warminster 1992) p. 3.
228  

conquerors had little impact on the form taken by their regnum, which
was constituted on the outlines of the Visigothic kingdom. Adminis-
trative and military boundaries may have been preserved and tax-
collecting mechanisms utilized, although the evidence is slight.41
Mùsà made directly for Toledo, and if Cordoba, rather than Toledo
soon became the Islamic capital, it may have been a response to
the difficulty in ensuring the loyalty of the former Visigothic capi-
tal, which is reflected in repeated accounts of revolts in later Arabic
histories, rather than an indication that the new regime was delib-
erating reshaping Hispania. After the brief period of turmoil brought
by ˇàriq and Mùsà, the chronicler implies that, for a time at least,
the stability of the kingdom was restored. There is no hint of the
rivalry between ˇàriq and Mùsà which was one of the main themes
of the conquest stories in Arabic, although the “Chronicle of 754”
described how Mùsà was recalled to Damascus by al-Walìd, who
was angry with Mùsà in spite of the prisoners and the vast quanti-
ties of booty he brought with him, and ordered his public humilia-
tion.42 Indeed, the chronicler shows one of the early governors
identifying with his Visigothic predecessors in the most extreme way
possible. Mùsà’s son 'Abd al-Aziz married Rodrigo’s widow, Egilona
and “tried to throw off the Arab yoke from his neck and retain the
conquered kingdom of Iberia for himself ”.43 'Abd al-Aziz was mur-
dered by a rival Arab faction; the chronicler represented the succes-
sion of al-Óurr as legitimate: in regno Esperie per principalia iussa succedit.
Al-Óurr also pursued a conciliatory line towards the indigenous pop-
ulation. He “restored to the Christians the small estates that had ori-
gin- ally been confiscated for the sake of peace” and “punished the
Moors [. . .] on account of the treasure they had hidden”.44 A decade
later, Yahya (c. 727–30) risked antagonizing the Islamic regime’s sup-
porters even further when “with bitter deceit [. . .] [he] stirred up
the Saracens and Moors of Spain by confiscating property that they
were holding for the sake of peace and restoring many things to the
Christians”.45 It is perhaps surprising to read that in references to

41
J. Vallvé Bermejo, “España en el siglo VIII: ejército y sociedad”, Al-Andalus
43 (1978) pp. 51–112.
42
Chronica Muzarabica, pp. 33–5.
43
Ibid., p. 37; transl. Wolf, p. 136.
44
Ibid., p. 36; transl. Wolf, p. 137.
45
Ibid., p. 39; transl. Wolf, p. 141.
    229

conflict between the rulers of Hispania and Francia, the chronicler

took a pro-Hispanic rather than a pro-Christian line. Eudo’s defeat
of the governor al-Sam˙ at Toulouse c. 721 was reported as only a
temporary setback rather than a cause for Christian rejoicing. The
defeat did not hinder the peaceful succession of Muslim governors
and the “Chronicle of 754” reported that Eudo was himself later
defeated.46 Al-Sam˙’s successor, Ambasa, was shown defending the
old Visigothic borders against the Franks, and although he imposed
heavy taxation on the Christians “he triumphed in Spain with the
highest honour”.47 The impression given by the chronicler, despite
his laments, is one of continuity rather than collapse.
Yet there was one obvious difference between Visigothic and Islamic
Hispania. In the Arabic tradition, the new regime established by
Mùsà may have begun as a part of an Islamic empire based on
Damascus. The Arabic histories make repeated reference to voyages
to Damascus to seek authorization and arbitration at the caliphal
court. Even Hispanic Christians might resort to Damascus. Ibn al-
Qù†ìya told how one of his ancestors, Sara, the granddaughter of
Witiza, the penultimate Visigothic ruler, appealed to Hishàm (724–43)
against her uncle, who had stolen land belonging to her young broth-
ers, in contravention of the treaty which Sara’s father had made
with al-Walìd at the time of the conquest.48 Many of these stories
are late and implausible. The survival of the “Chronicle of 754”
means that the picture in Hispania seems clearer than that for the
Islamic conquests in the east, where the long period between the
events and their recording means that the degree of central control
of events may be a literary creation and can be disputed.49 The
“Chronicle of 754” also sees the new conquest as sponsored by
Damascus. Yet the Chronicle‘s account of the relationship between
Cordoba and Damascus is both equivocal and changing. The chron-
icler notes that the early governors were appointed “by order of the
prince”50 (presumably the caliph in Damascus) or by the governor

46
Ibid., pp. 41–3.
47
Ibid., p. 39; transl. Wolf, p. 140.
48
Ibn al-Qù†ìya, Historia de la conquista, pp. 4–6.
49
F.M. Donner, “Centralised authority and military autonomy in the early Islamic
conquests”, The Byzantine and Early Islamic Near East, vol. 3: States, Resources and Armies,
ed. A. Cameron (Princeton NJ 1995) pp. 337–60.
50
Chronica Muzarabica, p. 37.
230  

of Africa “whose privilege it was to confer power in Spain with the

consent of the prince”.51 After the defeat of al-Sam˙ at Toulouse,
a temporary commander was elected until “Anbasah, by order of
the prince, came to serve as governor”.52 The last of the governors
whose appointment was reported in this way was Hudjifah, who gov-
erned briefly and unsuccessfully c. 728. Later rulers of Hispania seem
to have been independent of Damascus. 'Abd al-Malik, who in 734
“was sent by the order of the prince to be governor of Spain” was
replaced by 'Uqbah three years later, but returned to power with-
out the intervention of Damascus c. 740: consensu omnium [. . .]. eli-
gitur Arabum in regno.53 'Abd al-Malik acted on his own authority in
attempting, without success, to deny access to Spain to the armies
of Balj whom the caliph had sent to suppress a Berber revolt in
North Africa. This precipitated a civil war between the orientalia under
Balj and the occidentalia, the Arabs already established in the penin-
sula. This period was the subject of a separate epitoma written by the
author of the “Chronicle of 754” but now unfortunately lost. As the
peninsula emerged from chaos, Thalaba was “appointed with unan-
imous support” and his successor Yusuf “was wonderfully acclaimed
by all the senate of the palace as king of the land”. This was how
the Visigothic succession had been determined, at least in theory.
The list of the governors in the “Prophetic Chronicle”54 suggests that
the caliph’s control lasted a little longer. It names Tawaba as the
last of the governors appointed by the caliph, and describes his suc-
cessor Yusuf, who was overthrown by 'Abd al-Ra˙màn I, as an
independent king (and also, incidentally, as an Umayyad). The list
of the rulers of al-Andalus in Ibn Habìb’s list, entitled “On the gov-
ernors of al-Andalus from the day of the conquest to the year 265
after the Hegira” (878–9) makes little distinction between the Umay-
yads and their predecessors.55 All three sources suggest that within a
few years of the conquest, Islamic Hispania may have been in effect
independent from Damascus, rather than the farthest outpost of
empire. Islamic scholars from al-Dawùdi (d. 1011) onwards were to
argue about the legal basis of the conquest and whether the caliph

51
Ibid., p. 40; transl. Wolf, p. 142.
52
Ibid., p. 38.
53
Ibid., p. 45.
54
Chroniques Asturiènnes ( fin IXe siècle), ed. and transl. Y. Bonnaz (Paris 1987) p. 8.
55
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 434.
    231

had ever received the fifth of the booty to which he was entitled.56
The scanty surviving coinage may bear this out, although it is difficult
to interpret. Coins in the form of Byzantine solidi were struck with
the letters SPN as the place of minting and no reference to Damascus.
After 716, coins with bilingual Latin and Arabic inscriptions and the
name al-Andalus appeared. This independence may find an echo in
the story preserved in the Arabic tradition that al-Walìd had con-
sidered abandoning Hispania.
Wolf has already shown how literary models influenced the way
in which the “Chronicle of 754” “reported the Muslim invasion and
settlement more or less matter-of-factly as a change of regime in
Spain”. This was a consequence of “the way in which the chroni-
clers tapped into earlier works of Iberian history”.57 Above all, it was
the influence of Isidore that determined how the new gens and reg-
num would be described. Isidore was a protagonist of the “Chronicle
of 754” as well as providing the material for the period to 621 and
a powerful literary model. One strand of the chronicle is an account
of the Visigothic church councils, listing one council for each reign.
The chronicler noted that Isidore defeated a heretical “Acephalite”
bishop at the council of Seville of 619,58 and was present at the
fourth council of Toledo in 633.59 The language of the chronicle is
characterized by Isidore’s brevitas.60 Wolf argued that this allowed the
chronicler to skirt round the problems caused to a writer of provi-
dential Christian history by the continuing existence of the Arab
kingdom and the cooperation between the native population and the
conquerors: “by describing the Muslim caliphs as reges and their gen-
erals as duces, he dresses a foreign people in terms familiar to a
reader well versed in the language of Latin history”.61 But Isidore’s
influence went much further than providing the tools to allow the
chronicler to skate over the facts. Isidore, working from classical

56
H. Abdul Wahab and F. Dachraoui, “Le régime foncier en Sicile aux IXème–
X siècles: edition et traduction d’un chapitre du ‘Kitab al-amwal’ de al-Dawudi”,
ème

Etudes d’orientalisme dédiées à la mémoire de Lévi-Provençal 2 (Paris 1962) pp. 401–55, here
p. 428.
57
K.B. Wolf, Conquerors and Chroniclers of Early Medieval Spain (Liverpool 1990) p. xvi.
58
Chronica Muzarabica, p. 20.
59
Ibid.
60
Collins, The Arab Conquest, pp. 38–9.
61
Wolf, Conquerors and Chroniclers, p. 41.
232  

models, had created the template from which the description of a

new gens and regnum could be created.
The author of the “Chronicle of 754” held preconceived ideas
about the ethnicity of the conquerors of 711. The origins of the
newcomers were far away and long ago. Thus it is not surprising
that the chronicler gives no indication that the new rulers of Hispania
were involved in any process of ethnogenesis, in the moulding of
natives and conquerors into a new gens, such as had taken place in
the Migration period. Hispanic scholars were aware of the existence
of Arabs and Saracens even before they began to have an impact
on the Byzantine empire. In Book IX of the “Etymologies”, Isidore
identified the Saracens as the descendants of Ishmael; they were
called Saracens either because they claimed descent from Sara,
Abraham’s wife, or because of their Syrian origin.62 The Sarraceni
were the Arabs of Syria and the north of the Arabian peninsula.
Arab is given as an alternative name for the Sabaeans from Arabia,
whose genealogy stretched back to Noah’s son Ham.63 Isidore placed
them in the south of the Arabian peninsula, in Yemen. Disputes
between the Arabs of North and South Arabia were a constant fea-
ture of Islamic history, including the conquest period in Hispania,
although the way these groups identified themselves depended on
the current circumstances. As Collins noted,64 the civil strife between
the different groups of the conquerors in Hispania in the 740s ought
to have made the chronicler more attentive to Isidore’s distinction
between Arab and Saracen, although in fact he ignored it. It is pos-
sible that the chronicler had dealt with this question in his epitoma
on the civil wars, but it is equally likely that this was a tacit recog-
nition that the etymology and genealogy of the conquerors were of
little importance.
This may be because characterizing the regnum was more impor-
tant than identifying the new gens. In the Laus Spaniae—o sacra sem-
perque felix principium gentiumque mater—which begins the “History of
the Goths”, adapted from two works of Cyprian in praise of vir-
ginity, Isidore had raised up Spain as “the pride and ornament of

62
Isidore of Seville, Etymologiarum sive originum libri XX 9,2,6; 2,57, ed. W.M.
Lindsay, 2 vols. (Oxford 1911).
63
Ibid., 9,2,2; 9,2,14; 9,2,18.
64
Collins, The Arab Conquest, p. 62.
    233

the world, the most illustrious part of the earth, in which the Getic
people are gloriously prolific, rejoicing much and flourishing greatly”.65
The glories of Spain, entrusted once to the Romans, were now the
property of the Goths: “although this same Romulean power, ini-
tially victorious, betrothed you to itself, now it is the most flourishing
people of the Goths, who in their turn, after many victories, have
eagerly seized you and loved you; they enjoy you up to the present
time amidst royal emblems and great wealth, secure in the good for-
tune of empire”. Isidore provided a genealogy and history for the
Goths that made them worthy of this great prize, culminating in
Suithila, “the first to obtain the monarchy of the entire kingdom of
Spain”.66 Isidore’s focus was geographical, and he was able to talk
about Hispania and “the peoples of Hispania” without meaning any
particular ethnic group.67 Goths and Romans were equal citizens of
the regnum Gothorum. The regnum was under the control of elected, not
hereditary rulers, whose origin, as in the case of Erwig, could be
outside the peninsula. Both John of Biclar and Isidore saw Franks
as enemies and exaggerated Gothic victories over them, not because
of ethnic differences, but because Hispania was under threat. This
was the line that the author of the “Chronicle of 754” took on the
relationship between Hispania and Francia after 711. After the con-
quest, the term Goth disappeared from the chronicler’s vocabulary,
and the natives of Hispania were distinguished only as Christians
from the incoming Arabs, Saracens and Moors, whose rulers took
over from the Visigoths, upholding the glory of Hispania.
In contrast, the chronicler’s treatment of the Moors (or Berbers,
although the chronicler never used this term) showed them as a sep-
arate gens who had little or no role to play in the continuity of the
Hispanic regnum. The ethnic identity of these invaders was well known
in Hispania. John of Biclar had identified them as the opponents of

65
Isidore of Seville, Historia Gothorum, Wandalorum, Sueborum, ed. T. Mommsen,
MGH AA 11 (Berlin 1894) pp. 241–303, here p. 267; transl. Wolf, Conquerors and
Chroniclers, p. 80.
66
Isidore of Seville, Historia Gothorum, p. 292.
67
Id., Etymologiarum 9,2,110; D. Claude, “Remarks about relations between Visigoths
and Hispano-Romans in the seventh century”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transformation
of the Roman World 2 (Leiden-New York-Köln 1998) pp. 117–30.
234  

the Byzantines in North Africa.68 There was a long history of these

peoples from across the straits of Gibraltar raiding the peninsula,
going back to the Roman period. The “Chronicle of 754” implied
that ˇàriq’s arrival was the culmination of a process of constant
harassment of Hispania from North Africa, for the Moors “had
long been devastating the province assigned to them”.69 Recent schol-
arship on the army of the conquest characterizes it as consisting
largely of Berbers. They were almost certainly a mixture of converts
to Islam, pagans practicing their own local religions and even Chris-
tians.70 An eleventh-century source mentions a group called the
afàriqa among the conquerors, and this may be a reference to the
Romanised indigenous population of North Africa.71 With the pos-
sible exception of ˇàriq, none of these men achieved prominence in
the government of Islamic Hispania. Yet neither was the identity of
the Moors subsumed into that of the successors of the Visigoths.
Individual Mauri continued to play a role. The chronicler gave a
detailed account of the rebellion of “one of the people of the Moors
named Munnuza”72 against the governor 'Abd al-Ra˙màn al-Ghafiqì
c. 731, linking this with the oppression of the Berbers of North Africa
by their Arab rulers. When, in the 740s, the Berbers revolted, the
Mauri of al-Andalus “gathered for war, wanting to subject 'Abd al-
Malik to themselves, cross over the sea in ships, and offer his con-
quered kingdom (regnum eius adsumptum) to their allies on the other
side of the sea (transmarinis sodalibus)”.73 There was no suggestion that
this gens was to be identified with the people of Hispania.
The establishment of the Arabs in Hispania was of course more
than a literary construct, a reworking of Isidore’s “History of the
Goths”. There are enough similarities between the “Chronicle of
754” and the Arabic histories of al-Andalus to confirm the outline
of the chronicler’s account. It is true that the Arabic chronicles were
all written at least a century after the event, and survive in manu-

68
John of Biclar, Chronica 569; 570; 571; 578, ed. T. Mommsen, MGH AA 11
(Berlin 1894) pp. 207–20.
69
Chronica Muzarabica, p. 31.
70
M. Brett and E. Fentress, The Berbers (Oxford 1996; repr. 1998) p. 92.
71
J. Hopkins, Medieval Government in Barbary (London 1958) pp. 62–6; E. Savage,
A Gateway to Hell, A Gateway to Paradise: the North African Response to the Arab Conquest
(Princeton NJ 1997) p. 102.
72
Chronica Muzarabica, p. 41.
73
Ibid., p. 46; transl. Wolf, p. 149.
    235

scripts which are later still. The quotations which later historians
make of earlier sources, now lost, resemble a game of Chinese whis-
pers, in which some members in the chain may have transmitted
their version of the message accurately (although some did not), but
the original message has been lost and cannot be reconstructed.
Collins’ reliance on the “Chronicle of 754” as his principal source
for the history of the conquest seems justified. But his sweeping dis-
missal of the Arabic histories should be overturned in favour of a
critical reading of this evidence. In the concluding section of this
paper I will use the “History” of Ibn Óabìb to make connections
between the Arabic traditions of the conquest into the account pro-
vided by the “Chronicle of 754”.
The “History” of Ibn Óabìb is a universal history from the Creation.
It contains the earliest surviving account of the conquest in Arabic
written in al-Andalus, which is, however, little more than a list of
governors and few anecdotes. It survives in one manuscript dated
1295–6, which seems to be a copy of a version of the work com-
piled by Yusuf ibn Yahya al-Maghami, who died in 901 and was a
pupil of Ibn Óabìb.74 Ibn Óabìb spent three years in Egypt, where
he was a pupil of 'Abd al-Óakam, whose son wrote a “History of
the Conquest of the Maghreb and al-Andalus”, and where he may
have heard for the first time the fabulous tales about the locked
chamber in Toledo and the Table of Solomon. Ibn Óabìb also trav-
elled to Medina, and seems to have copied part of his history of al-
Andalus from the Medinan scholar al-Waqidi, via his pupil Ibrahim
ibn al-Mundhir al-Hizami (d. 850). The work of al-Waqidi survives
only in fragments cited in other sources; the thirteenth-century his-
torian Ibn 'Idhàri gave him as the authority for the agreement
between ˇàriq and count Julian, the battle between ˇàriq and
Rodrigo, the arrival of Mùsà and the marriage between 'Abd al-
Aziz, the son of Mùsà, with Egilona, here called the daughter, rather
than the wife of Rodrigo. What Ibn Óabìb, or his pupil al-Maghami,
took from al-Waqidi, is a brief and sober account of the rulers of
al-Andalus, which is in general agreement with the account given
by the “Chronicle of 754”:

74
Ibn al-Fara∂ì, Ta"rikh 'ulama" al-Andalus, ed. F. Codera and J. Ribera y Tarragó,
vol. 2, Bibliotheca Arabico-Hispana (Madrid 1892) no. 201; Ibn Óabìb, Kitàb al-
Ta "rìkh, p. 75.
236  

'Abd al-Malik told us from Ibrahim ibn al-Mundhir al-Hizami, from

al-Waqidi: ˇàriq remained in al-Andalus for one year from the day
of the conquest until he left, then after him Mùsà ibn Nusayr [reigned]
for two years and one month, then 'Abd al-Aziz, the son of Mùsà for
two years, then al-Óurr ibn 'Abd al-Ra˙màn al-Thaqafi for two years
and eight months [. . .]. Then Yusuf ibn 'Abd al-Rà˙màn ruled for
six years and nine months. Then 'Abd al-Ra˙màn ibn Mu'àwiya ibn
Hishàm ibn 'Abd al-Malik ibn Marwan ruled, may God be wholly
pleased with him! He was the one who entered al-Andalus [i.e. the
first of the Umayyads to rule Hispania]. Yusuf marched towards him
and met him at the garrison of Cordoba and 'Abd al-Ra˙màn ibn
Mu'àwiya defeated him and took possession of the kingdom and entered
the palace from that time on and was acknowledged as sovereign
[. . .].75
Thus it seems that the history of the conquest had been forgotten
in al-Andalus and had to be reimported from outside the peninsula.
The prominence which Ibn Óabìb gave in his “History” to the
stories which he had heard in Egypt led the prominent nineteenth-
century Arabist Dozy to dismiss the “History” as about as much use
to historians as “The 1001 Nights”—which is indeed where many
of the stories of the conquest ended up. Ibn Óabìb’s “History” was
excluded from the canon of Arabic chronicles of al-Andalus, and it
is one of the few sources that have not been translated. Yet Ibn
Óabìb’s evidence is preferable to some of the later sources upon which
modern historians rely. Most of the Arabic histories are the work of
scholars who were active after the establishment of the Umayyad
caliphate in 929. 'Abd al-Ra˙màn III and his son al-Óakam II are
said to have sponsored the writing of history and genealogy, which
became largely a series of panegyrics of the Umayyads.76 The difficulty
in viewing the history of al-Andalus from any other perspective has
been compounded by nostalgia for the glories of Umayyad Cordoba
which was generated in the tiny taifa kingdoms which emerged from
the collapse of the caliphate, and is still alive in modern Arabic cul-
ture. The first author to be dignified with the title “the historian”
by the Umayyad propagandist was A˙mad al-Ràzì (d. 955), who,
according to his son, was the first to collect and evaluate the histo-
ries of the conquest. Al-Ràzì’s work on the conquest does not sur-

75
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 434.
76
G. Martínez-Gros, L’idéologie omeyyade: La construction de la légitimité du califat de
Cordue (Xe–XIe siècles) (Madrid 1992).
    237

vive, although he was cited by later authors such as Ibn Óayyàn (d.
1076) and Ibn 'Idhàri, a thirteenth-century historian from the Maghreb.
Some of these works survive, and they were also collated by a
Moroccan historian active in the seventeenth century, al-Maqqari,
in a compendium which is still the source most commonly cited for
the history of al-Andalus.77 Al-Ràzì may have some connection with
the work bearing his name that was translated into Portuguese by
order of king Dinis (1279–1325).78 Only two short passages of this
version survive in a seventeenth-century transcription.79 One or more
Castilian translations were made shortly afterwards, but the bulk of
the work is a gazetteer of al-Andalus. In the three extant Castilian
versions, the section on the conquest is unusable because it was
turned into a historical novel by Pedro de Corral c. 1430. It is
difficult to see why al-Ràzì is always considered a more reliable
source than Ibn Óabìb.
The ethnic origins of the participants in the Arabic histories are
often mentioned. Yet one may draw rather fewer conclusions from
the Arabic sources than from the “Chronicle of 754”, since all these
accounts are biased by later rivalries between Arabs and Berbers, and
by the mesh of false genealogies which these engendered. Ibn Habìb
described ˇàriq’s first army force as consisting of 1,700 Berbers,
closely followed by another 10,000 Berbers with a mere 60 Arabs.
He recounted a prophecy given to ˇàriq that a Berber would con-
quer al-Andalus.80 By the time that Ibn Óabìb was writing, the promi-
nence given to the Berbers in this version of the conquest may have
been unpalatable. Ibn Óabìb’s stories had been thoroughly islami-
cized, and ˇarìq’s conquest of Hispania was represented as a Holy
War, and by implication, an Arab project. The relationship between
the religion for which the Holy War was being pursued and its ori-
gins in the Arab world was controversial throughout the Islamic
world. Particularly vexed was the question of the status of non-Arabs,
particularly of those who had converted to Islam and become clients

77
Al-Maqqari, Analectes sur l’histoire et la littérature des Arabes d’Espagne par al-Makkari,
ed. R. Dozy, G. Dugat, L. Krehl and W. Wright, 2 vols. (Leiden 1855/61; repr.
Amsterdam 1967).
78
Crónica del Moro Rasis, ed. D. Catalán, D. de Andrés and M.S. de Andrés
(Madrid 1974).
79
Ibid., c. 179.
80
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 136.
238  

of rulers and noble families. Among these men was ˇàriq, the client
of Mùsà. Ibn Óabìb’s history reflected these concerns, which I have
explored elsewhere.81 He also quoted a passage, probably taken from
an eastern source, which is ostensibly a conversation between a gov-
ernor of Kufa at the end of the eighth century and a scholar about
the famous scholars of the Islam.82 One by one, they were identified
as clients rather than men of pure Arab origin. The emir’s rage at
this unwelcome information was appeased only by the naming of
two genuine Arabs. The Berbers in al-Andalus were similarly a vic-
tim of their non-Arab origins. Ibn Óabìb’s pupil seems to have added
to the “History” a prophecy that Cordoba would be destroyed by
the Berbers.83
Berbers from North Africa would continue to be recruited as mer-
cenaries at times of internal strife in al-Andalus, such as the strug-
gles between the brothers of 'Abd al-Ra˙màn I and his nephew
al-Óakam I. Particularly galling to the Arabic historians was the role
of the Berbers in the rise of al-Manßur at the end of the tenth cen-
tury and the collapse of the caliphate. Berbers continued to desta-
bilize the Andalusian polity right through to the arrival of the
Almohads and Almoravids. No wonder then, that the role of the
Berbers in the conquest was played down and that of Mùsà ibn
Nusayr, a man with an impeccable Arab background whose father
had been a member of the royal guard of the caliph Mu'àwiya,84
became central. As the politics and culture of al-Andalus became
more sophisticated and emulated that of Baghdad—perhaps not
before the ninth century and reaching its zenith in the mid-tenth
century—all things Arab were praised and the Berbers denigrated.
Genealogists such as Ibn Óazm,85 writing in the eleventh century,
were anxious to give their subjects prestigious Arab genealogies and
an ancestor who came to al-Andalus at the time of the conquest. Ibn
Óabìb mentioned members of the “followers”—noble Arab families
whose immediate ancestors had been companions of Mu˙ammad—

81
A. Christys, “The History of 'Abd al-Malik ibn Óabìb and ethnogenesis in al-
Andalus”, Power and the Construction of Communities, ed. W. Pohl and H. Reimitz (forth-
coming).
82
Ibn Óabìb, Kitàb al-Ta"rìkh, pp. 173–4.
83
Ibid., p. 151.
84
Ibid., p. 138.
85
Ibn Óazm, Jamharat ansab al-'arab, ed. 'A.M. Harun (Cairo 1971).
    239

who came to al-Andalus, although he was not sure how many of

them had participated in the conquest.86 Ibn Idhàrì, writing in the
thirteenth century, noted that Arabs from North Africa came to al-
Andalus with two of the early governors, al-Óurr in 716 and al-
Sam˙ in 718,87 and Abu al-Khattar brought thirty members of his
tribe, the Kalbis, in c. 742. Arab origins were even invented for the
Berbers by Islamic scholars, who divided them into two groups, the
Butr and the Baranis, and gave them two origins, one in Palestine,
and the other in Yemen. There are very few references to these ori-
gins in the first centuries after the conquest.88 Such distortions have
also bedevilled discussions of the fate of the Berbers of al-Andalus
in the Umayyad period, especially the highly controversial subject of
Berber settlement in al-Andalus. There is little reference to Andalusis
who spoke Berber, and their languages left little trace on Romance,
in contrast to Arabic. By the eleventh century it is difficult to find
traces of the Berbers who arrived with the conquest of 711. These
recreations of history made the conquest look more Arab than it in
fact was.
Working from a different historiographical perspective to that of
scholars of the first, “Germanic”, transformation of Hispania, it is
impossible to answer all the “leading questions” posed to the con-
tributors to this volume. Analysis of any socio-economic or consti-
tutional developments which may have taken place in the first three
centuries after the conquest are too dependent on late and contra-
dictory sources usefully to be debated here. Not until the tenth cen-
tury did the rulers of al-Andalus actively promote an official line on
the origins of the Umayyad state and only for this period do we
have the evidence that Roman, or rather Byzantine, ritual played
an important role in this—if we are to believe our sources.89 Both
the Latin and the Arabic evidence considered in this paper show

86
Ibn Óabìb, Kitàb al-Ta"rìkh, p. 138.
87
Ibn Idhàrì, Al-Bayàn al-Mughrib, ed. R. Dozy, E. Lévi-Provençal and G. Colin,
2 vols. (Leiden 1948/51) p. 25.
88
Brett and Fentress, The Berbers, p. 92; M. Brett, “The Islamization of Morocco
from the Arabs to the Almoravids”, Journal of Moroccan Studies 2 (1992) pp. 57–71;
repr. id., Ibn Khaldun and the Medieval Maghreb (Aldershot 1999); H. de Felipe, “Leyendas
árabes sobre el orígen de los beréberes”, Al-Qantara 11/2 (1990) pp. 379–98.
89
M. Barceló, “El Califa patente: el ceremonial omeya de Córdoba o la esceni-
ficación del poder”, Estructuras y formas del poder en la historia, ed. R. Pastor, I. Kienie-
wicz, E. García de Enterría et al. (Salamanca 1991) pp. 51–71.
240  

the transformation of Hispania after 711 as an episode in which

questions concerning the relationship between gens and regnum may
be addressed. The author of the “Chronicle of 754” did indeed write
up the conquest period as though it were another “barbarian” migra-
tion, although he was perhaps too influenced by his reading of Isidore.
Isidore’s Hispania, which had absorbed the Goths and Hispano-
Romans, seemed to absorb the new gentes of 711 almost seamlessly
into the beloved regnum. The Arabic histories’ rewriting of the events
of the conquest to suit the Arabicized world of the ninth century
and after provide an alternative ethnogenesis which may be seen as
a distorting mirror image of a similar process of state formation.
Here the focus was more on gens than regnum. The stories about the
rivalry between the Berber ˇàriq and the Arab Mùsà, and the
invented Arabs who took part in the conquest, are part of a refor-
mulation which explained how a largely Berber army of the con-
quest became the Andalusis of Hispania.
    241

APPENDIX: The Governors of al-Andalus

Chronicle of 754 Ibn Habib Prophetic chronicle

Musa Era 749 Mùsà ibn Nusayr Musa 1 year 3 months

2 years 1 month
'Abd al-Aziz 3 years 'Abd al-Azìz ibn Mùsà 'Abd al-Aziz 2 years
2 years 6 months
Ayub 1 month Ayub 1 month
Al-Hurr 3 years Al-Hurr ibn 'Abd al- Al-Hurr 2 years
Ra˙màn al-Thaqafì 10 months
2 years 8 months
Al-Samh Al-Sam˙ ibn Malik al- Al-Samh 2 years
Khawlanì 2 years 9 months
9 months
'Abd al-Rahman 'Abd al-Rahman ibn
1 month Abdullah 1 month
Anbasah 4 1/2 years 'Anbasa ibn Sa˙ìm al- 'Udra 1 month
Kalbì 4 years 5 months
Yahya 3 years Yahya ibn Salàma Yahya ibn Salama
2 years 6 months 2 years 6 months
Odiffa 6 months Hadhìfa ibn al-A˙ùsß Hudhaifa ibn al-Ahwas
al-'Abasì 1 year 6 months
Uthman 4 months Uthmàn ibn Abi Nasca al- Uthman ibn Abi Nisa
Khathcàmì 5 months 4 months
Haytham 1 year Al-Haytham ibn 'Abìd Al-Haytham ibn 'Ubayd
4 months 10 months
Muhammad al-Ashjai Muhammad ibn Ashjai
1 month
'Abd al-Rahman 3 years 'Abd al-Rahman ibn 'Abd al-Rahman ibn
'Abd Allah 2 years 'Abd Allah 1 year
8 months 10 months
'Abd al-Malik almost 'Abd al-Malik ibn Qatßan 'Abd al-Malik ibn Qatan
4 years al-Fihrì 4 years 2 years
Uqba 'Uqba ibn al-Hajjàj al 'Uqba ibn al-Hajaj
Salùlì 5 years 2 months 4 year 5 months
'Abd al-Malik 1 year
1 month
Balj ibn Bishr al-Qaysì
11 months
Thcalaba ibn Salàma al- Tawaba ibn Salama
'Àmalì 5 months 1 year 2 months
Abu al-Khattar Hisham Abu al-Khatàr 2 years
ibn Dirar al-Kalbi
2 years
Thalabah Abu Thawàba al-
Jadhamì
Yusuf 6 years 9 months Yùsuf ibn 'Abd al- Yusuf 11 years
Era 785 Ra˙màn
[to 138 AH]
This page intentionally left blank
GENTES, KINGS AND KINGDOMS—THE EMERGENCE OF
STATES: THE KINGDOM OF THE GIBICHUNGS

Ian N. Wood

The successor states of Western Europe and North Africa are one
of the distinguishing features of the post-Roman period, possibly the
most distinctive feature of all, yet each state was created in different
circumstances, even if a number of them can be seen as beginning
their independent existence in the 470s. Equally, each state ended
in different circumstances, some already in the 530s, others in much
later transformations. Just as important, each was written up at
different times and for different reasons—those that collapsed early
were less likely to be represented to suit the needs of later genera-
tions, except as foils for their victorious conquerors. This question
of the dating of our sources is a point of particular significance when
one considers the literary accounts of the creation of each gens that
constitute their ethnogenesis. In thinking comparatively about state
formation, it is as well to keep chronology firmly in mind. The
Burgundians of the fourth century can usefully be set alongside other
peoples mentioned in the pages of Ammianus Marcellinus: those of
the fifth need to be compared with the peoples who established them-
selves in Gaul at the same time: their demise in the sixth century
naturally suggests comparison with the Vandals and Ostrogoths.
Unlike these last two peoples, however, the Burgundian kingdom
had a Nachleben in Frankish Burgundy, which has no parallels in
Byzantine Africa or sixth-century Italy, but may bring to mind other
groups subjected to the Franks in the Carolingian period, notably
the Lombards and Saxons.
Although the Burgundians already appear in the writings of Pliny1
and Ptolemy,2 and are attested in the Byzantine historian Zosimus’
account of events in the third century, notably as opponents of the
Emperor Probus,3 they first receive significant attention in the fourth

1
Pliny, Historia Naturalis 4,99, ed. G. Winkler and R. König (Munich 1988).
2
Ptolemy, Geographia 2,11,8, ed. C. Müller (Paris 1883).
3
Zosimus, Histoire Nouvelle 1,68, ed. F. Paschoud, vol. 1 (Paris 1971).
244  . 

century, when Ammianus Marcellinus recounts Valentinian I’s use

of them against Macrianus, king of the Alamans.4 In setting out this
narrative Ammianus makes a number of observations: first, that the
Burgundians were related to the Romans:5 second that they had long
been enemies of the Alamans, with whom they quarrelled over bor-
ders and access to salt-mines:6 and finally that they were ruled over
by a pair of leaders—a general, the hendinos, who could be deposed
in the event of ill-fortune, and a priest-king, the sinistus, whose power
was for life.7 That they were enemies of the Alamans is not unlikely,
nor is the significance of frontiers and salt in causing conflict. The
other two statements are more problematic. The dual rulership over
the Burgundians is often cited in discussions of sacral kingship,8 and
Ammianus himself makes comparison with the Egyptians: ut solent
Aegyptii casus eius modi suis assignare rectoribus. The use of the present
tense solent in the comparison with Egypt, however, suggests that
Ammianus was drawing his information from an old source, for the
description of Burgundian kingship is scarcely compatible with what
is known about Egyptian attitudes to their rulers in the Late Roman
period.9 Even if the comparison with Egypt were ever valid, there
is no reason to think that Ammianus’ observation held true for the
Burgundians of the fourth century. Certainly the information can-
not be used as the basis for understanding the political structures of
the fourth-century Burgundians, particularly when there is no other
evidence to suggest the existence of such a set-up. That the Burgundians
were ruled by kings is as near certain as can be, but one can deduce
no more from Ammianus’ comments. What he has to say about the
relationship of the Burgundians to the Romans is perhaps more help-
ful. Clearly this cannot be founded on biological reality, but the
Romans are known to have referred to their allies as being related
to them by blood.10 The likelihood is that the stated relationship

4
Ammianus Marcellinus, Res Gestae 28,5,9; 30,7,11, ed. J.C. Rolfe (Cambridge
Mass. 1935–39).
5
Ibid. 28,5,11.
6
Ibid.
7
Ibid. 28,5,14.
8
See most recently J. Favrod, Histoire politique du royaume burgonde (443–534)
(Lausanne 1997) pp. 43–4.
9
I.N. Wood, “Kings, Kingdoms and Consent”, Early Medieval Kingship, ed. P.H.
Sawyer and I.N. Wood (Leeds 1977) pp. 6–29, here p. 27.
10
I.N. Wood, “Defining the Franks: Frankish Origins in Early Medieval Histo-
   245

between Burgundians and Romans originated in imperial diplomacy.

The statement, nevertheless, constitutes the earliest written reference
to the ethnogenesis of the Burgundians. Moreover, although it is an
ethnogenesis which is likely to have been created by the Romans
rather than Burgundians, it may have had some impact on the rul-
ing elite of the Burgundians themselves, and indeed Ammianus explic-
itly claims that it was the Burgundians who knew of their Roman
descent (subolem se esse Romanam Burgundii sciunt).11
A generation later Orosius offered another comment on the peo-
ple. For him 80,000 Burgundians, “a new enemy with a new name”,
descended on the Rhine, but following the subjugation of Germania
by Tiberius and Drusus they were settled in burgi, whence they gained
their name.12 Orosius thus provides the people with a second ethno-
genesis, and like that of Ammianus it is a Roman one, depending
this time on a bogus Latin etymology. He adds the interesting, but
puzzling information, that by his own day, that is the second decade
of the fifth century, the Burgundians were Catholic.13 The accuracy
of this need not concern us, but it will be necessary to consider the
extent to which the Arianism of Gundobad and some of those sur-
rounding him helped differentiate the Burgundians from the Gallo-
Romans.14

riography”, Concepts of National Identity in the Middle Ages, ed. S. Forde, L. Johnson
and A.V. Murray, Leeds Texts and Monographs. New Series 14 (Leeds 1995) pp.
47–57, here p. 51.
11
Favrod, Histoire politique du royaume burgonde, p. 42 rightly notes that this is a
Burgundian claim, but it is likely, given the parallels noted in I.N. Wood, “Ethnicity
and the ethnogenesis of the Burgundians”, Typen der Ethnogenese unter besonderer
Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der Öster-
reichischen Akademie der Wissenschaften, philosophisch-historische Klasse 201.
Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien 1990)
pp. 53–69, here pp. 57–8, that the connection between the Romans and Burgundians
originated in Roman ideology.
12
Orosius, Historiarum adversum paganos libri VII 7,32,11–12, ed. K. Zangemeister,
CSEL 5 (Vienna 1882; repr. Hildesheim 1967). See Wood, “Ethnicity and the
ethnogenesis of the Burgundians”, pp. 56–7.
13
Orosius, Historiarum adversum paganos 7,32,13.
14
For the question of the religion of the Burgundians, Wood, “Ethnicity and the
ethnogenesis of the Burgundians”, pp. 58–60. The argument has been challenged
by Favrod, Histoire politique du royaume burgonde, but his criticism fails to take full
account of the evidence for Catholicism among the Burgundians. If one argues that
Orosius is incorrect, one has to explain away the presence of Catholics among the
Burgundians before the conversion of Sigismund: if one argues that the Burgundians
were originally converted to Catholicism, one has to explain the presence of Arians
at Gundobad’s court. One possibility is that the distinction between the beliefs of
246  . 

By the time that Orosius was writing Burgundians were to be

found on the Rhine.15 They had probably taken advantage of the
chaos of the years 406–7 to move westwards.16 According to Olym-
piodorus their phylarch Guntiarius (= Gundichar) was instrumental
in the appointment of the usurper Jovian in 41117—if so he was not
to be the last Burgundian to elevate a Roman emperor. The activ-
ities of Burgundians within the Empire, however, were sharply cur-
tailed in the mid 430s when they were defeated by Aetius.18 The
fullest statement of this is by Prosper, who comments: “Aetius wore
down in war Gundichar, king of the Burgundians, who was living
in Gaul, and granted him the peace for which he was begging, which
he did not enjoy for long, since the Huns destroyed him and his
people utterly.”19 Some, however, survived, and Aetius was later to
settle a group of Burgundians in Sapaudia,20 which lay largely to the
north of Lake Geneva, in the Roman province of Maxima Sequa-
norum.21 How sizeable this group was, and whether it included all
the Burgundians who survived the cataclysm on the Rhine is un-
clear. What is certain is that Aetius’ policy of resettling the Bur-
gundians paid off, since they subsequently fought for him at the
Battle of the Catalaunian or Mauriac Plains—a battle of some impor-
tance both for Aetius and for the Burgundians themselves, as we will
see.22 The settlement out of which the Burgundian kingdom devel-

the Burgundians and the Catholic episcopate only really became clear after the set-
tlement in Sapaudia, and that it was only after the settlement that individuals could
clearly be classified as Arian or Catholic.
15
Prosper of Aquitaine, Epitoma Chronicon s.a. 413, ed. T. Mommsen, MGH AA
9 (Berlin 1892; repr. Munich 1981).
16
The sole evidence for this move, however, is Jerome, Lettres 123, ed. J. Labourt,
vol. 7 (Paris 1961), discussed by Favrod, Histoire politique du royaume burgonde, pp.
45–6.
17
Olympiodorus, Fragmenta 17, ed. and transl. R.C. Blockley, The Fragmentary
Classicising Historians of the Later Roman Empire 2 (Liverpool 1982; repr. 1983).
18
The Chronicle of Hydatius and the Consularia Constantinopolitana 108 (= 436), ed.
R.W. Burgess (Oxford 1993); Chronicorum a. CCCCLII 118 (= 436), ed. T. Mommsen,
MGH AA 9 (Berlin 1892); Chronicorum a. DXI 596 (= 436?), ed. T. Mommsen,
MGH AA 9 (Berlin 1892).
19
Prosper of Aquitaine, Epitoma Chronicon 1322 (= 435).
20
Chronicorum a. CCCCLII 128 (= 443). The chronology of the Chronicle is, how-
ever, totally unreliable: see I.N. Wood, “The fall of the western empire and the
end of Roman Britain”, Britannia 18 (1987) pp. 251–62, here pp. 253–6.
21
The fullest recent discussion of Sapaudia is to be found in Favrod, Histoire poli-
tique du royaume burgonde, pp. 100–17.
22
Jordanes, De origine actibusque Getarum 36,191, ed. F. Giunta and A. Grillone,
Fonti per la storia d’Italia 117 (Rome 1991) [henceforth: Jordanes, Getica].
   247

oped was, therefore, entirely dependent on the policy and actions of

a Roman general.
In later tradition the destruction of the Rhineland Burgundians
was an epic moment. The story is at the heart of numerous epics
and legends, which have prompted the much-debated question Gab
es ein Burgunderreich bei Worms? That the destruction of a Burgundian
kingdom took place, and that the memory of its end was transformed
into legend is not in question. There is, however, no indication that
any significant memory of the event was preserved by those Bur-
gundians who were subsequently settled in Sapaudia.23 Indeed if the
memory had been carried by the Gibichungs and their followers,
it would surely have presented the destruction of the Rhineland king-
dom as a prelude to another settlement, and not as a final cata-
clysm. If the memory of the destruction of the kingdom of Gundichar
were to be ascribed to Burgundians, one might speculate as to whether
that memory was preserved by Burgundians who never transferred
to Sapaudia, and who perhaps descended from groups opposed to
the Gibichungs. Such a possibility is raised by the existence in the
late sixth and seventh centuries of the family of the Burgundofarones,
who are associated with the region round Melun, in Frankish Neustria,
rather than any part of the Burgundian kingdom.24 One might equally
hypothesise that the story of the destruction of the Rhineland king-
dom was elaborated by neighbouring peoples, like the Franks, who
had no wish to remember Burgundian survival. On the other hand
it is far from clear that the ethnicity of the legendary Rhineland

23
P. Amory, “Names, ethnic identity, and community in fifth- and sixth-century
Burgundy”, Viator 25 (1994) pp. 1–30, here p. 13, rightly comments on the lack of
interest shown in the historical background of the Burgundians by Avitus, Gundobad
or Sigismund. W. Goffart, “Conspicuously absent: martial heroism in the Histories
of Gregory of Tours and its likes”, The World of Gregory of Tours, ed. K. Mitchell
and I.N. Wood (Leiden 2002) pp. 365–93, to my mind rightly denies that there
was an unwritten Burgundian tradition about the fall of Gundichar’s kingdom.
24
Jonas of Bobbio, Vita Columbani 1,26; 2,7, ed. B. Krusch, MGH SSrG 37
(Hannover 1905). But see also Hildegar of Meaux, Vita Faronis, ed. J. Mabillon,
Acta Sanctorum Ordinis sancti Benedicti (Mâcon 1936) pp. 606–25, for an account
of the family set in the context of Gibichung history. The Burgundofarones or
Faronids are extensively studied by R. Le Jan, Famille et pouvoir dans le monde franc
(VIIe–IXe siècle) (Paris 1995) pp. 388–95, where the majority of the family is placed
firmly in the region of Meaux, with only one group having links with the “confins
de l’Austrasie et de la Bourgogne”. See also ead., “Convents, violence, and com-
petition for power in seventh-century Francia”, Topographies of Power in the Early Middle
Ages, ed. M. de Jong, F. Theuws and C. van Rhijn, The Transformation of the
Roman World 6 (Leiden 2001) pp. 243–69, here pp. 250–2, 254.
248  . 

kingdom was a matter of much interest. In the earliest of the sto-

ries concerned with Gundichar, the ninth- or tenth-century Waltharius,
which admittedly does not conclude with the kingdom’s destruction,
the king is portrayed as a Frank, as indeed is Haganus (Hagen)!25
The kingdom of Worms was a kingdom enshrined in legend: its rela-
tionship to the historical Burgundians was not a matter of crucial
importance. At all costs, there is nothing to suggest that the group
settled in Sapaudia looked back to the kingdom on the Rhine, and
its destruction, as providing a past which was central to their own
self identification.
Yet Gundobad, in a law preserved in the Liber Constitutionum issued
by his son, did look back to his ancestors, Gibich, Godomar, Gislahar,
Gundahar (= Gundichar), as well as to his father (Gundioc) and uncle
(Chilperic I), without noting any break in Burgundian history.26 The
list takes us back generations before the settlement in Sapaudia, per-
haps even to the time of the wars against Macrianus, and might be
thought to suggest a pattern of hereditary kingship which sorts ill
with the hendinos and sinistus of Ammianus.27 Equally important, the
list reveals that Gundichar was not forgotten by the Burgundians of
Sapaudia, although the events at the end of his reign seem to have
been. This list of ancestors, together with a reference to the Battle
of the Mauriac Plains,28 is the nearest that the Burgundian rulers
come to recording the past of their people, in contrast to Theo-
doric, who seems to have had Cassiodorus compile a history of the
Ostrogoths.29 The ancestral list of the Liber Constitutionum is short, but
it is nevertheless interesting: it is the only surviving Burgundian or

25
The poem may have originated in the Strasbourg region. Fragments of an ear-
lier Anglo-Saxon poem, Waldere, may come from a version of the same story:
Ekkehard, Waltharius of Gaeraldus, ed. A.K. Bate (Reading 1978) pp. 1–8. For the
full legend of the destruction of Gundichar’s kingdom one has to wait until the
thirteenth-century Nibelungenlied.
26
Liber Constitutionum 3, ed. L.R. von Salis, MGH LL nationum Germanicarum
2,1 (Hannover 1892); for the dating of the code, see I.N. Wood, “Disputes in late
fifth- and early sixth-century Gaul: some problems”, The Settlement of Disputes in Early
Medieval Europe, ed. W. Davies and P. Fouracre (Cambridge 1986) pp. 7–22, here
p. 10.
27
By contrast Favrod, Histoire politique du royaume burgonde, p. 46, sees this as an
indication that Gundichar was the descendent of a hendinos.
28
Liber Constitutionum 17,1.
29
See most recently A. Søby Christensen, “Cassiodorus and the making of a his-
tory of the Goths”, Analecta Romana 26 (1999) pp. 173–7.
   249

Gibichung kinglist, and, like the list in the later Lombard code of
Rothari,30 it is enshrined within a lawcode. It might be said to link
Burgundian kingship and lawgiving: on the other hand, unlike the
list provided by Rothari, that given in the Liber Constitutionum is no
part of a prologue, but is provided simply in the context of a state-
ment about royal freedmen and freedwomen: those who were already
freed under the listed ancestors of the legislator were to retain their
status. The list is not, therefore, associated with the process of leg-
islation. An equally significant sense of the past is to be found in
the statement in the Liber Constitutionum that the Battle of the Mauriac
Plains provides the terminus earlier than which lawsuits among the
Burgundians were to be dismissed.31 Once again, however, history
merely provides a limit to legal claims.
Thus, to recapitulate what can be said about the notions which
the Burgundians of the kingdom of Sapaudia held of their past, they
seem not to be concerned with the kingdom of Gundichar on the
Rhine. Even though Gundichar was an ancestor of the Gibichung
rulers of Sapaudia and the Rhône-Saône kingdom that grew out of
it, and even though the family provided the sole ruling dynasty for
the Burgundian kingdom, there is no evidence that any memory of
the Rhine catastrophe stuck to the Gibichungs. The Burgundians
of the kingdom of Sapaudia, despite the various traditions they could
have drawn on, are not presented in any early source, other than
the Chronicle of 452, as being the descendents of the Rhineland
Burgundians. Unlike the Goths, Lombards or Anglo-Saxons, who are
given seamless histories stretching from their migrations through to
the establishment of their respective kingdoms, early sources do not
set the Burgundians of Sapaudia into a history of past migration.
Even those eighth- and ninth-century saint’s Lives which do present
a history of Burgundian migration make no mention of the Rhineland
catastrophe or indeed of the subsequent transfer to Sapaudia, which
is left unrelated.32 As for the Burgundians of the late fifth and sixth
centuries, it looks as if they did not make much of their history prior
to their settlement in the province of Maxima Sequanorum. Although

30
Edictus Rothari prol., ed. F. Bluhme, MGH LL 4 (Hannover 1868).
31
Liber Constitutionum 17,1.
32
Passio Sigismundi 1, ed. B. Krusch, MGH SSrM 2 (Hannover 1888); Hildegar,
Vita Faronis 8.
250  . 

there had been a gens Burgundionum for several centuries, the settle-
ment in Sapaudia had, it seems, effectively created a new people—
and that settlement was effected by Aetius and not by any named
Gibichung. The earliest historical event with which the gens Burgundionum
is associated in the documentation produced in the Sapaudian king-
dom is the battle of the Mauriac Plains. Their royal dynasty stretched
back to a past on the Rhine, but that past, or at least its catastrophic
end, seems not to have been remembered in Gibichung circles.
The settlement in Sapaudia provided the kernel from which the
kingdom of the Burgundians was created, even though the subse-
quent establishment of that kingdom is hard to trace. The Continuatio
Havniensis Prosperi records that the Burgundians who were expanding
throughout Gaul were driven back by the Gepids—a statement which,
if correct, raises questions as to where this confrontation took place.33
Easier to place in what is known of Burgundian activity is the involve-
ment of Gundioc and Chilperic I, alongside the Visigoths, in the
attack on the Sueves in Galicia in 456.34 According to the Continuatio
Havniensis Prosperi, the death of the Suevic leader was followed by an
agreement between Gundioc and his Visigothic counterpart, Theodo-
ric, over Burgundian settlement.35 Here for the first time a formative
role in his people’s development is attributed to a member of the Gibi-
chung family. Marius of Avenches ascribes to the same period an
agreement made between Burgundians and senators to divide Gaul.36
The arrangement was almost instantly ended by the new emperor
Majorian.37

33
Continuatio Havniensis Prosperi 574 (= 455), ed. T. Mommsen, MGH AA 9 (Berlin
1892; repr. Munich 1981): At Geppidos Burgundiones intra Galliam diffusi repelluntur. The
episode would be easier to understand were the subject and object to be reversed,
which is how the passage is mistranslated by S. Muhlberger, “The Copenhagen
continuation of Prosper: a translation”, Florilegium 6 (1984) pp. 71–95, here p. 80:
“But [sic] the Burgundians spread through Gaul and drove back the Gepids”. The
translation forms an appendix to the same author’s “Heroic kings and unruly gen-
erals: the ‘Copenhagen’ continuation of Prosper reconsidered”, Florilegium 6 (1984)
pp. 50–70.
34
Jordanes, Getica 44,231.
35
Continuatio Haviensis Prosperi 583 (= 457).
36
Marius of Avenches, Chronique, s.a. 456, ed. J. Favrod (Lausanne 1991) [hence-
forth: Marius of Avenches]. This is usually held to be the same as the episode in
Fredegar, Chronicon 2,46, ed. B. Krusch, MGH SSrM 2 (Hannover 1888), where it
is dated to 372.
37
Sidonius Apollinaris, carm. 5, ll. 564–71, ed. A. Loyen, 3 vols. (Paris 1960–70);
J. Harries, Sidonius Apollinaris and the Fall of Rome (Oxford 1994) pp. 85–6.
   251

Despite Majorian’s stance we soon meet members of the Burgundian

royal family as Roman officials. Gundioc appears as magister militum
( per Gallias?) in a papal letter relating to a disputed election at Die
in c. 462/3,38 while a Chilperic—probably Gundioc’s brother rather
than his son of the same name39—was active as magister militum ( per
Gallias?) and patricius in the Geneva region almost a decade later.40
He (or perhaps his nephew) receives a reasonable press from Sidonius
Apollinaris in 472,41 and although two of the Roman rhetor’s uncles
got into hot water with the Burgundian official in 474, Sidonius
noted the extent to which the magister militum could be assuaged by
his, almost certainly Catholic and perhaps Roman, wife.42
Opinions about the Burgundians and their position in the Empire
at this time were, in fact, divided. Chilperic’s rule in Geneva was
denounced by the ascetic Lupicinus as tyranny.43 Sidonius himself
objected to having Burgundians as neighbours.44 On the other hand
a friend of Sidonius, Arvandus, was happy to envisage a division of
Gaul between Goths and Burgundians in c. 469. His ideas led to his
being tried for treason and convicted.45 Yet, to judge by the events
of the year 456 mentioned by Marius of Avenches he was not alone
in thinking the way he did—and the rapid collapse of the Western
Empire was effectively to justify his interpretation of the situation.
As for the two relatives of Sidonius who fell foul of Chilperic in
474, Simplicius and Apollinaris, their mistake was to throw in their

38
Epistolae Arelatenses Genuinae 19, ed. W. Gundlach, MGH EE 3 (Berlin 1892).
39
The problem of distinguishing Chilperic I and Chilperic II is usefully discussed
by Favrod, Histoire politique du royaume burgonde, pp. 149–51. Like Favrod I regard
the crucial text as being Liber Constitutionum 3, which makes it quite clear that
Chilperic I was a ruler of significance. Without this text, however, one would
attribute to Chilperic II most of the actions (those of 456 apart) currently attrib-
uted to Chilperic I.
40
Vita patrum Iurensium 2,10 (= 92), ed. F. Martine, Sources Chrétiennes 142 (Paris
1968).
41
Sidonius, ep. 6,12,3.
42
Sidonius, epp. 5,6; 5,7. The image of the new Lucumon swayed by a new
Tanaquil may suggest that Chilperic’s wife was Roman. These letters are read in
a very different way by D. Frye, “Gundobad, the Leges Burgundionum, and the strug-
gle for sovereignty in Burgundy”, Classica et Medievalia 41 (1990) pp. 203–4, seeing
them as highly critical. Favrod, Histoire politique du royaume burgonde, p. 50 n. 180,
reads Sidonius, ep. 6,12,3, as indicating that Chilperic was an Arian married to a
Catholic wife. I find this interpretation unlikely, as discussed below.
43
Vita patrum Iurensium 2,10 (= 92–5).
44
Sidonius, carm. 12.
45
Sidonius, ep. 1,7. See Harries, Sidonius Apollinaris and the Fall of Rome, pp. 159–66.
252  . 

lot with the new emperor, Julius Nepos.46 What made this such
a crime was the simple fact that the previous emperor, Glycerius,
was the appointee of Chilperic’s nephew, or perhaps brother, Gun-
dobad.47 Unlike his other relatives Gundobad had gone to Italy,
where he became a protegé of the magister militum praesentalis Ricimer,
to whom he was related—although it is not clear whether he was
the latter’s nephew or son-in-law. Either he or Ricimer was the killer
of Anthemius.48 When Ricimer died, Gundobad took his place. He
received the title of patricius from Olybrius,49 but was nevertheless
personally responsible for the emperor’s death and for the appoint-
ment of Glycerius.50 The arrival of an eastern appointee, Julius Nepos,
seems to have prompted him to abandon Italian politics, and to
return to Gaul. In such a context the support of Sidonius’ uncles
for Nepos was scarcely sensible.
It is useful to take stock of the position of the Burgundians and
their Gibichung leaders in the 470s, directly before the emergence
of a Burgundian kingdom. The people, or a proportion of what
remained of it after conflict with the Huns, had been settled in
Sapaudia—by Aetius, as part of a policy which included the settle-
ment of Alans in Valence and in Gallia Ulterior.51 Comparison can
also be made with the earlier settlement of the Visigoths in Toulouse.52
Unlike the Alans and the Visigoths, however, the leaders of the
Burgundians had taken up mainstream jobs within the Roman Empire,
with three of them acting as magistri militum of one sort or another.
Whereas Euric was steadily expanding his authority as king of the
Visigoths in the 470s, Gundioc and his relatives were very much
acting as legitimist Roman figures. If they fell out with the Empire
in 474, it was because the imposition of a new emperor left them

46
Sidonius, ep. 5,6.
47
John of Antioch, Fragmenta 209,1–2, ed. C. Müller, Fragmenta Historicorum
Graecorum, vols. 4 and 5 (Paris 1868/70), transl. C.D. Gordon, The Age of Attila
(Ann Arbor 1960) pp. 122–3.
48
Chronicorum a. DXI 650 (= 472).
49
Fasti Vindobonenses Priores 608 (= 472), ed. T. Mommsen, MGH AA 9.
50
John of Antioch, Fragmenta 209,2.
51
Chronicorum a. CCCCLII 124 (= 440); 127 (= 442).
52
E.A. Thompson, “The settlement of the barbarians in southern Gaul”, Journal
of Roman Studies 46 (1956) pp. 65–75, reprinted in id., Romans and Barbarians: the
Decline of the Western Empire (Madison 1982) pp. 23–37; for an alternative reading
and date, I.N. Wood, “The barbarian invasions and first settlements”, Cambridge
Ancient History 13 (2nd edn., 1998) pp. 516–37.
   253

stranded: comparison here might be made with Aegidius, Majorian’s

magister militum, who was forced to plough his own furrow after
Ricimer’s appointment of Libius Severus as emperor in 461.53 Aegidius
makes an interesting comparison, in that like him the Burgundian
rulers never ceased to be imperial legitimists, despite the events of
474 and the years that followed. The issue was not whether one was
subject to the Empire or not, but rather whether there was a legit-
imate emperor to whom one was subject.
Exactly what happened between 474 and 494 is unclear—and it
is therefore uncertain when one might place the beginning of the
Burgundian kingdom, as distinct from that of the presence of the
Burgundians in Gaul and the appointment of members of the Gibi-
chung family to official positions within the Roman hierarchy. Al-
though one may also ask when the Visigoths ceased to be servants
of the Roman state, in their case it is easier to trace the growth of
an independent power, notably under Euric.54 In 474 Chilperic was
still magister militum and patricius. According to Ennodius of Pavia,
writing early in the sixth century, by 494 Gundobad could be described
as a Burgundian king. He was based in Lyons, while one of his
brothers, Godegisel, had a separate but subordinate court in Geneva.55
As we will see the development of such royal courts did not neces-
sarily imply that the Burgundian rulers regarded themselves as inde-
pendent of the Empire even after 476. Between 476 and 494 Chilperic
vanished, as did two brothers of Gundobad.56 The third brother,
Godegisel, was subsequently to challenge Gundobad’s power: in 500
he allied with Clovis, and the two drove Gundobad south. The lat-
ter was, however, able to come to terms with the Frank, to whom
he momentarily became tributary. Clovis retired north, leaving Gun-
dobad free to deal with his brother, which he did in the siege of
Vienne. Having reestablished himself, the Burgundian king, accord-
ing to Gregory of Tours, issued leges mitiores.57 What these were is
uncertain: they certainly cannot be equated with the Liber Constitutionum

53
Harries, Sidonius Apollinaris and the Fall of Rome, pp. 96–7.
54
K.F. Stroheker, Eurich, König der Westgoten (Stuttgart 1937).
55
Ennodius, Vita Epifani 174, ed. F. Vogel, MGH AA 7 (Berlin 1885).
56
The fates of Chilperic II and Godomar are uncertain, despite Gregory of
Tours, Historiae 2,28, ed. B. Krusch and W. Levison, MGH SSrM 1,1 (2nd edn.,
Hannover 1951).
57
Marius of Avenches, s.a. 500; Gregory of Tours, Historiae 2,33.
254  . 

itself, although they may be enshrined within that code.58 That

Gundobad was concerned with the well-being of all his subjects is
also implied by his Pax et Abundantia coinage,59 whose inscription is
referred to in a dedication homily of Avitus of Vienne.60
Gundobad ruled as rex, although it is worth noting that the more
precise title rex Burgundionum is not often associated with him61—a
point which may, or may not, reflect the nature of our sources,
rather than the nature of his power. On the other hand, and more
important, he was very definitely concerned with Roman titles, and
not just while he was active as Ricimer’s successor in Italy, nor only
for himself. Admittedly he had his son, Sigismund, elevated in reg-
num, at Carouge, outside Geneva.62 Since this elevation seems to
have taken place shortly after 500, it is probable that Sigismund was
intruded into Godegisel’s place, but it is equally clear that he remained
subordinate to his father.63 This “Germanic” office, however, was
soon complemented by a Roman one, for Sigismund also held the
title of patricius before the end of 515.64 This was a title that both
Gundobad and Chilperic had used.65 Further, Sigismund started to

58
There is an assumption that they should be identified with Liber Constitutionum
42–88; Frye, “Gundobad, the Leges Burgundionum, and the struggle for sovereignty
in Burgundy”, p. 209. Perhaps significantly Liber Constitutionum 42 is dated 501 and
54 (on ordeal) 502.
59
P. Grierson, M. Blackburn and L. Travaini, Medieval European Coinage 1 (Cam-
bridge 1986) p. 76.
60
Avitus of Vienne, hom. 24, ed. R. Peiper, MGH AA 6,2 (München 1883).
61
The use of the title in von Salis’ reconstruction of the Prima Constitutio of the
Liber Constitutionum is questionable. See also H. Wolfram, Lateinische Königs- und
Fürstentitel bis zum Ende des 8. Jahrhunderts, Intitulatio 1 (Graz 1967) pp. 87–9. Gundobad
does use the title rex Burgundionum in Liber Constitutionum, constitutio extravagans 19, while
it is associated with Sigismund in constitutio extravagans 20; Sigismund also appears
as rex in the capitula to the council of Epaon: Les canons des conciles mérovingiens (VI e–
VIIe siècles), vol. 1, ed. J. Gaudemet and B. Basdevant, Sources Chrétiennes 353
(Paris 1989). The word rex (but not rex Burgundionum) is used for both Gundobad
and Sigismund in the addresses of Avitus of Vienne’s letters, but it is not clear
whether these are original. Cassiodorus does, however, use the title rex Burgundionum
(Cassiodorus, Variae 1,46; 3,2, ed. T. Mommsen, MGH AA 12 [Berlin 1894], and
transl. S.J.B. Barnish, Translated Texts for Historians 12 [Liverpool 1992]), though
this obviously reflects Ostrogothic requirements. See Amory, “Names, ethnic iden-
tity, and community in fifth- and sixth-century Burgundy”, p. 13.
62
Fredegar, Chronicon 3,33.
63
See Avitus of Vienne, Letters and Selected Prose, ed. D. Shanzer and I.N. Wood
(Liverpool 2002) pp. 20–2.
64
Avitus, ep. 9; Vita Abbatum Acaunensium absque epitaphiis 3, ed. B. Krusch, MGH
SSrM 7 (Hannover 1920).
65
Gundobad: Fasti Vindobonenses Priores 608 (= 472); Chilperic: Vita patrum Iurensium,
2,10 (= 92).
   255

negotiate with the emperor Anastasius for his father’s title of magis-
ter militum—something which Theodoric the Ostrogoth seems to have
tried to block, but which the Burgundian did obtain, apparently after
his father’s death in 516.66 Meanwhile the demise of Gundobad led
to the full elevation of Sigismund as king.67
Sigismund, like his father, was intent on being seen as a Roman
official as much as a Germanic king.68 In this the two might be com-
pared with Theodoric the Ostrogoth—although the latter’s claim to
rule Italy and to the ornamenta palatii of the West depended on con-
quest, in a way that Gundobad and Sigismund’s claims to office in
Gaul did not.69 Gundobad had been a legitimate Roman office-holder
in the West before 476, and his son was his successor. At the same
time, of course, the ornamenta palatii are a reminder that Italy had
been the seat of the western emperor, while Gaul had only been in
the hands of imperial officials, at least in the last days of the Empire.
Clovis, too, had his eye on imperial approval, but even the accla-
mations at Tours came nowhere near to placing him within the old
hierarchy which Gundobad and Sigismund seem to have hankered
after.70
The attitudes of Gundobad and Sigismund towards their Roman
offices is best expressed in a series of letters written to the Emperor
Anastasius on the latter’s behalf by Avitus of Vienne. In the first, a
fragmentary letter of c. 515, Sigismund refers to his father as gentis
regem, sed militem vestrum, the king of his people but the soldier of the
emperor.71 This doublet was picked up again in a marginally later
letter where Sigismund was angling for his father’s title of magister
militum. Here he states, cum gentem nostram videamur regere, non aliud nos
quam milites vestros credimus, “while we are seen to rule our people, we
believe ourselves to be nothing other than your soldiers”.72 The word

66
Avitus, epp. 93; 94. See also ep. 47.
67
Marius of Avenches, s.a. 516.
68
Amory, “Names, ethnic identity, and community in fifth- and sixth-century
Burgundy”, p. 8: “From the instant that a Burgundian first appears on the histor-
ical stage, we seem to be viewing a man steeped in the formal hierarchy and cer-
emony of Mediterranean politics”; p. 13: “the Burgundian royal family appear in
all the contemporary sources as quintessential late antique magnates.”
69
A.H.M. Jones, “The constitutional position of Odoacer and Theoderic”, Journal
of Roman Studies 52 (1962) pp. 126–30.
70
On this see M. McCormick, Eternal Victory. Triumphal rulership in late antiquity,
Byzantium, and the early medieval West (Cambridge 1986) pp. 335–7.
71
Avitus, ep. 46A.
72
Ibid. 93.
256  . 

miles no doubt involved an allusion to Gundobad’s office of magister

militum.
But while the king is the emperor’s soldier, what do these letters
say about the king’s people? Sigismund goes so far as to claim that
vester quidem est populus meus, “my people is yours”.73 Elsewhere he
sees his people as subiecti of the emperor, and claims that they illus-
trate the extent of the imperial respublica: ornat quippe imperii vestri ampli-
tudinem longinquitas subiectorum et diffusionem reipublicae vestrae adserit.74 In
official correspondence the Burgundian kingdom was thus an exten-
sion of the Empire, as were its inhabitants. What is less clear is what
Avitus intended to imply by the words gens and populus. A precise
reading of the words would seem to suggest that both words refer
to the Germanic followers of the rulers. In the law code, the Liber
Constitutionum, in those cases where it is possible to distinguish between
Romans and Burgundians, populus noster refers exclusively to the
latter. On the other hand, there are laws, quod in populo nostro debeat
custodire, which were admittedly issued at an assembly of Burgundians,
but which unquestionably include legislation which is not directed
solely at the king’s Germanic subjects.75 This Burgundian assembly,
which is recorded in only some manuscripts of the law code, more-
over, took place in very specific circumstances. It clearly followed a
defeat, apparently at the hands of the Franks, and is probably to be
placed after the retirement of Sigismund to Agaune. In the circum-
stances it is best seen as a conventus of army leaders under the aegis
of Godomar.76 In this context the phrase conventus Burgundionum reflects
not so much a continuity of Burgundian Selbstbewußtsein as the real-
ities of leadership in a period of acute military crisis. In general, the
implication would seem to be that in official Gibichung documents
the phrase populus meus or noster was normally a synonym for the gens
Burgundionum, and that it meant the Germanic followers of the king,

73
Ibid. The phrase is taken over in the title of an article by G. Scheibelreiter,
“Vester est populus meus. Byzantinische Reichsideologie und germanisches Selbst-
verständnis”, Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna
1989) pp. 203–20; the passage in question is discussed on pp. 206–8. See also the
commentary on these letters in Avitus of Vienne, Letters and Selected Prose, pp. 141–53.
74
Avitus, ep. 78.
75
Liber Constitutionum, constitutio extravagans 21.
76
For the date of the assembly, K. Binding, Das burgundisch-romanische Königreich
(Leipzig 1868) pp. 260–2; Favrod, Histoire politique du royaume burgonde, p. 451.
   257

but that on occasion it might be convenient for its meaning to seem

open-ended. Sigismund would surely not have wanted Anastasius to
think that he himself had no jurisdiction over the Roman popula-
tion of his kingdom. Had the emperor consulted the patriarch of
Constantinople, who received a letter written in Avitus’ own name,
he would have found the bishop describing Sigismund as domnus meus
filius vester, my lord and your son.77 The Romans of the Burgundian
kingdom certainly knew that they were subject to Gibichung lords,
but the same letter of Avitus also recognises Anastasius as princeps.78
Avitus’ letters are an important insight into official Gibichung views
of the populus over whom the Burgundian kings ruled. Equally impor-
tant is the evidence of the law-code. Unfortunately later tradition
ascribing the Liber Constitutionum to Gundobad, and a consequent tam-
pering with the preface or Prima Constitutio, has muddied the waters
considerably. Suffice it to say that the preface is dated to 29 March,
in the second year of a king’s reign, and that it was issued in Lyons.
Since Sigismund issued other legislation on 29 March 517, the sec-
ond year of his reign, at his Easter court in Lyons,79 it is highly
likely that the Liber Constitutionum was issued at the same time. Of
course the code included earlier legislation, and we do not, in any
case, have it in its pristine form. The actual production of a code,
however, must be ascribed to Sigismund, and set in the context of
the king’s Easter court.
It is equally important to note that this is not a law of the Bur-
gundians, a Lex Burgundionum, but a book of constitutions, a Liber
Constitutionum, even though one clause specifically refers to acts of the
legislator’s named ancestors.80 It regulates relations between Burgun-
dians, and between Burgundians and Romans, and it deals also with
other barbarians and incomers.81 It does refer the Romans back to
“Roman laws just as has been established by our predecessors”,82
perhaps to the Codex Theodosianus, or to the enigmatic Lex Romana
Burgundionum, assuming it to have been an official compilation, which

77
Avitus, ep. 9.
78
See also ibid. 48.
79
Liber Constitutionum 52.
80
Ibid. 3.
81
P. Amory, “The meaning and purpose of ethnic terminology in the Burgundian
laws”, Early Medieval Europe 2 (1993) pp. 1–28, here pp. 10–2.
82
Liber Constitutionum, prima constitutio 8.
258  . 

is by no means certain. Overall, however, the Liber Constitutionum is

not law simply for the Burgundians. Of course the Pactus legis Salicae
is not concerned solely with Clovis’ Franks,83 yet there is a distinc-
tion to be drawn between the Frankish and Burgundian codes. The
Liber Constitutionum is prefaced with a long Prima Constitutio which dis-
cusses the implementation of the law in some detail, not least the
responsibilities of Burgundian and Roman judges,84 and it is followed
by a list of the names of the, admittedly Burgundian, comites who
signed it. It also contains within it edicts identifiable by their dating
clauses. The Pactus legis Salicae has no such prefatory information
and no list of witnessing officials. It is possible to trace within the
Frankish code some clauses which must have originated as edicts,85
but in general it belongs far less obviously to a Roman tradition of
law-giving than does the Burgundian book of constitutions. Sigismund’s
legislation is yet another indication of the problem of categorising
the rulership of the Gibichungs or indeed their kingdom as pre-
dominantly Germanic.
It is important to look at kings and the perceptions they have of
their followers, but it is also useful to try to come a little closer to
an understanding of the make-up of the Germanic population of the
Gibichung kingdom, and here the law-code is of particular use. With
regard to the Burgundians as a gens, we have traced them essentially
to their settlement in Sapaudia in the 440s, and have mentioned the
rather fragmentary evidence relating to attempts at expansion, with
and without the agreement of Roman senators, in the ensuing years.
It is necessary, however, to touch briefly on the thorny problem of
the Burgundian settlement, for which the Liber Constitutionum is our
chief, but not sole, source. That is not to say that the much-debated
issue of tax versus land requires detailed reconsideration here.86 Both

83
I.N. Wood, The Merovingian Kingdoms, 450 –751 (London-New York 1994)
p. 110.
84
Liber Constitutionum, prima constitutio 3; 5; 8; 11; 13.
85
Wood, The Merovingian Kingdoms, p. 110.
86
W. Goffart, Barbarians and Romans A.D. 418–584. The Techniques of Accommodation
(Princeton NJ 1980); W. Liebeschuetz, “Cities, taxation and the accommodation of
the barbarians: the theories of Durliat and Goffart”, Kingdoms of the Empire. The
Integration of Barbarians in Late Antiquity, ed. W. Pohl, The Transformation of the
Roman World 1 (Leiden-New York-Köln 1997) pp. 135–51; J. Durliat, “Cité, impôt
et intégration des barbares”, ibid. pp. 153–79; H. Wolfram, “Neglected evidence
on the accommodation of the barbarians in Gaul”, ibid. pp. 181–3; Wood, “Ethnicity
and the ethnogenesis of the Burgundians”, Appendix: the settlement of the Burgundians,
   259

tax and land are mentioned explicitly in the evidence for the earli-
est stages of the settlement of the Burgundians. That land was involved
seems to be implied by the Chronicle of 452’s statement that Sapaudia
Burgundionum reliquiis datur cum indigenis dividenda,87 and more obviously
by Marius’ announcement for the year 457 that Eo anno Burgundiones
partem Galliae occupaverunt terrasque cum Galliis senatoribus diviserunt. Yet
Fredegar, in a passage related, clearly erroneously, to 373, links
Burgundian settlement to the reduction of taxation on the Romans,
which could well imply that on occasion tax grants were the basis
of barbarian accommodation.88 On the other hand Sidonius seems
to refer to the billeting of Burgundians.89 In other words there was
no single mechanism for the accommodation of the followers of the
Gibichungs, and any attempt to squeeze a single model from the
evidence is doomed to failure. Moreover, that there were a number
of different stages in the settlement of the Burgundians is quite clear
from the evidence, and it is worth pausing on this issue.
Leaving aside the evidence for the initial settlement in Sapaudia,
and the temporary agreement with senators in 457, the legal evi-
dence shows very clearly that settlement was an on-going process.90
The Liber Constitutionum, in an unfortunately dateless clause, talks of
“our people” ( populus noster) who received land and slaves from the
legislator and his predecessors (which suggests that the clause was
issued by Gundobad, or just possibly Sigismund), and announces that
these men should not receive one third of the slaves or two parts
of the land where they had been given hospitalitas.91 It is unlikely that
this clause refers to allocations of revenue from tax, because of the
mention of slaves, but it might imply either billeting or the parti-
tion of property: it certainly implies illegal annexation of slaves and
land: it also indicates previous grants by kings of both these com-
modities. The implication would seem to be that close followers of

pp. 65–9; for a judicious survey, Favrod, Histoire politique du royaume burgonde, pp.
189–206.
87
Chronicorum a. CCCCLII 128 (= 443). Compare Continuatio Havniensis Prosperi 583
(= 457).
88
Fredegar, Chronicon 2,46; Goffart, Barbarians and Romans, p. 107.
89
Sidonius, carm. 12.
90
Goffart, Barbarians and Romans, pp. 127–161; D. Boyson, “Romano-Burgundian
society in the Age of Gundobad: some legal, archaeological and historical evidence”,
Nottingham Medieval Studies 32 (1988) pp. 91–118, here pp. 98–100, provides a sum-
mary.
91
Liber Constitutionum 54,1.
260  . 

a king might be treated in one way, and others in another. The pic-
ture is further complicated by a later law, the very one issued by
the conventus Burgundionum, which states that the present arrangement
for those Burgundians coming from elsewhere (qui infra venerunt) is
that they should receive half the land and half the slaves.92 Who
these Burgundians are is uncertain, but they may have been forced
out of previous arrangements when parts of the kingdom were lost
to the Franks, Ostrogoths or even the Alamans. Regardless of which
method of accommodation was being used at any one time, what is
clear is that the settlement of the Burgundian gens was not a single
event, but a process, and that different groups received land, slaves
or hospitalitas in different proportions at different periods.
The Liber Constitutionum has more light to shed on the nature of
the gens, not least because of its use of such terms as Burgundio/nes.
It is necessary to state from the start that a very high percentage of
the laws deal with Romans and Burgundians alike: indeed such
phrases as si quis Burgundio aut Romanus abound.93 On the other hand
there are a significant number of laws which are addressed to Bur-
gundians alone, and of those a good percentage deal with settlement
and inheritance. It is fairly clear that a major concern of the laws
issued by the Gibichungs was the settlement of their military fol-
lowing.94 In addition there are two eye-catching laws concerned with
preventing undue Burgundian influence in Roman law-suits.95 There
is, it seems, a binary opposition of Roman and Burgundian in the
law-book, which the legislators were both setting out rhetorically and
at the same time trying largely to override.
This opposition could also be expressed in terms of Roman and
barbarian, for, interestingly, the word barbarus is frequently used as
a synonym for Burgundian, and just as one finds phrases like Burgundio
aut Romanus, so too one finds tam barbarus quam Romanus.96 Indeed in
only one instance does the term barbarus appear to mean something

92
Liber Constitutionum, constitutio extravagans 21,12.
93
E.g. Liber Constitutionum 9; Frye, “Gundobad, the Leges Burgundionum, and the
struggle for sovereignty in Burgundy”, p. 206, notes that 12 of the first 41 titles
proclaim the equality of Burgundians and Romans.
94
Amory, “The meaning and purpose of ethnic terminology in the Burgundian
laws”, p. 26.
95
E.g. Liber Constitutionum 22; 55.
96
E.g. ibid. 8,1. The words also appear as variants in the various manuscripts:
see ibid. 47,1.
   261

other than Burgundian: in clause 79, de praescriptione temporum (on the

setting of time limits relating to the possession of land) Gundobad
legislates on the situation where “anyone of our people has invited
a person belonging to a barbarian nation to live with him” (si quis
in populo nostro barbarae nationis personam, ut in re sua consisteret, invitas-
set).97 Here, the normal meaning of the term barbarus as Burgundian
appears to be excluded by the phrase in populo nostro, which as we
have seen almost without exception implies Burgundian, leaving one
to conclude that barbarus in this one clause means non-Burgundian.
That there were Germanic incomers who wished to settle in the
Gibichung kingdom, is made absolutely clear in the final ordinance
of the Liber Constitutionum, where Goths who have been captives among
the Franks, and indeed any other outsider (quicumque persona de alia
regione), are given permission to settle.98 The barbari are, thus, usually
Burgundians, but could also be Germanic people of non-Burgundian
origin. That the same word is used for both groups implies a fluidity,
and indicates that the Burgundians of Sapaudia were essentially, but
not exclusively, the Germanic followers of the Gibichungs. Burgundio
and barbarus are very nearly synonymous: and as we have already
seen, so too are Burgundio and populus noster.
The evidence of the code, in its clauses on settlement in particu-
lar, and also of Avitus is an important reminder that, however much
Gundioc, Chilperic I, Gundobad and Sigismund wanted to be Roman
officials, and can indeed be portrayed as such, they had behind them
a military following, which needed to be accommodated. The two
aspects of rule were not, in any case, separate: the chief Roman
office held by the Gibichung rulers was that of master of the sol-
diers, magister militum, and, as we have seen, Gundobad and Sigismund
openly conceived of themselves as soldiers of the emperor. Their
military following carried with it women and children. Although the
term “Burgundian” in the Liber Constitutionum and in other sources is
unquestionably employed within identifiable rhetorical strategies,99
there were people who could be so classified, and who had to be settled.

97
Ibid. 79,1: the translation is that of K.F. Drew, The Burgundian Code (Philadelphia
1949) p. 75. That the lawgiver in this instance is Gundobad is clear from the date
at the end of the clause.
98
Liber Constitutionum, constitutio extravagans 21,4,6.
99
Amory, “The meaning and purpose of ethnic terminology in the Burgundian
laws”.
262  . 

That the Burgundians could be regarded as a military unit up

until the days of the settlement in Sapaudia is clear enough. If we
compare that settlement with other settlements of the same period,
it is likely that it had a military purpose.100 The Burgundians cer-
tainly continued to act as a warrior people—although it must be
said that they tended to be successful as raiders, for instance against
Italy101 and Brioude in the Auvergne,102 and when they were acting
together with other peoples—the Goths in the 450s103 and the Franks
in 508:104 otherwise they did less well. Their one major victory was
against the army of a single, and young, Frankish king, Chlodomer,
at Vézeronce.105 They collapsed totally before the onslaught of two
adult sons of Clovis ten years later.106 There may have been other
wars which can no longer be distinguished in the historical record:
Avitus wrote to Sigismund in the context of campaigns which have
not been securely identified.107 This dossier of military activity leaves
open the question as to how powerful a military force the Burgundians
were after the 430s, but it sheds some light on the Gibichung claim
to be the soldiers of the emperor.
Juxtaposition of this last claim with what can be deduced of the
gens Burgundionum suggests that the Gibichungs thought of themselves
as being Roman officials who happened to have access to Burgundian
troops, rather than as being primarily kings of the Burgundians or
of a Burgundian kingdom. The former impression is dominant in
the letters of Avitus, which provide much of our information, where
first Gundobad and then Sigismund are seen through Roman eyes,
surrounded by Romans.108 Of course they knew they had to pay
attention to the needs of their Burgundian subjects and to govern
them, and they had Burgundian comites to do so. It is these men,

100
Thompson, “The settlement of the barbarians in southern Gaul”.
101
Ennodius, Vita Epifani 168–77.
102
Gregory of Tours, Liber de Virtutibus sancti Juliani 7–8, ed. B. Krusch, MGH
SSrM 1,2 (Hannover 1885).
103
Jordanes, Getica 44,231; Continuatio Havniensis Prosperi 583 (= 457).
104
Chronicorum a. DXI 689 (= 508).
105
Marius of Avenches, s.a. 524; Gregory of Tours, Historiae 3,6.
106
Marius of Avenches, s.a. 534; Gregory of Tours, Historiae 3,11. Marius’ asser-
tion that Theudebert was present may be undermined by Gregory. See I.N. Wood,
“Clermont and Burgundy: 511–534”, Nottingham Medieval Studies 32 (1988) pp. 119–25.
107
Avitus, epp. 91; 92.
108
Amory, “Names, ethnic identity, and community in fifth- and sixth-century
Burgundy”, pp. 19–22.
   263

and not their Roman counterparts, who attest the Prima Constitutio
of the Book of Constitutions. The problematic Prima Constitutio apart,
however, there are only two moments in the Liber Constitutionum where
the ruler is designated rex Burgundionum,109 and a third, already dis-
cussed, where legislation comes specifically out of an assembly of the
Burgundians110—all three pieces of legislation, one might note, not
being in the original compilation, but in the Constitutiones Extravagantes.
Otherwise one only finds the phrase rex Burgundionum used as an
official title in Cassiodorus’ Variae, a view from outside.111 The Liber
Constitutionum, moreover, is, as we have seen, not concerned solely
with Burgundians, but with all Sigismund’s subjects—and in dealing
with Burgundian-Roman relations it was concerned with the issue
of equality.112
There is, therefore, no indication that the Burgundian kings wished
to perpetuate a policy of apartheid between their Germanic follow-
ers and their Roman subjects, despite the fact that the former raised
some very particular legal problems, concerned above all with set-
tlement, landholding and inheritance, which required specific legis-
lation. The fact that there were issues of settlement, as well as social
and legal differences between the Romans and Burgundians, meant
the Gibichungs could never iron out the distinction between the two
peoples entirely, but the laws suggest that the official policy of the
rulers was to treat both peoples as equally as possible.
Even in the area of religion, which is sometimes seen as crucial
to the distinction between barbarians and Romans,113 there appear
to have been no hard and fast lines in the Burgundian kingdom, at
least as portrayed in the evidence from the fifth and early sixth cen-
turies. First, Orosius and other early fifth-century writers saw the
Burgundians as Catholics.114 By the later fifth century it is clear that
Gundobad and those around him were largely Arian, although there
were still Catholics in the royal family, particularly amongst the

109
Liber Constitutionum, constitutiones extravagantes 19; 20.
110
Ibid. 21.
111
Cassiodorus, Variae 1,46; 3,2. Also Ennodius, Vita Epifani 174.
112
Frye, “Gundobad, the Leges Burgundionum, and the struggle for sovereignty in
Burgundy”, p. 206; Boyson, “Romano-Burgundian society in the Age of Gundobad:
some legal, archaeological and historical evidence”, p. 92.
113
P. Brown, The World of Late Antiquity (London 1971) pp. 124–5.
114
Wood, “Ethnicity and the ethnogenesis of the Burgundians”, pp. 58–9.
264  . 

women.115 Further, although Gundobad had an Arian priesthood,

which he was never quite willing to abandon,116 he discussed theo-
logical matters with Catholics and was open to persuasion—unlike
his Vandal counterparts117 or, later, the Visigoth Leovigild.118 He
even seems to have condoned a religious Dialogus written by Avitus
of Vienne, in which the bishop demolished the king’s arguments.119
In addition, Gundobad put up no opposition to the conversion to
Catholicism of his son, Sigismund, which took place apparently shortly
after 500, as a result of a journey to Rome.120 Sigismund himself
was not prevented by his father from founding major Catholic
churches,121 and it is even possible that Gundobad himself endowed
Catholic foundations.122 All in all, despite the fact that Gundobad
was Arian, there is nothing to suggest that religion acted as a dis-
tinguishing mark between Burgundian and Roman, or that it was
exploited to keep the two peoples separate.
It is worth returning to set alongside these observations on royal
power and religion the complete absence of the cultivation by Bur-
gundians of an ethnogenetic tradition, either comparable to the stories
told by Ammianus Marcellinus or Orosius, or similar to the tale of
Scandinavian origins of the Burgundians put about in the eighth
century.123 During the period of the Burgundian kingdom, there

115
Wood, “Ethnicity and the ethnogenesis of the Burgundians”, pp. 59–60. See
also the reservations of Favrod, Histoire politique du royaume burgonde, pp. 50–4. Favrod,
however, fails to deal with the question of named Catholic Burgundians of the fifth
century, while I do not find his reading of Sidonius, ep. 6,12,3 (p. 50 n. 180) con-
vincing. It is open to question as to whether Sidonius would have made much of
Chilperic’s enjoyment of Patiens’ feasts, if the Burgundian had been Arian.
116
Gregory of Tours, Historiae 2,34.
117
Apart from the evidence of Victor of Vita, Historia persecutionis Africanae provin-
ciae, ed. C. Halm, MGH AA 3,1 (Berlin 1879) pp. 1–58, there is also Ferrandus,
Vita Fulgentii, PL 65, cols. 117–50.
118
See especially Vitas Sanctorum Patrum Emeretensium, ed. A. Maya Sánchez, Corpus
Christianorum Series Latina 116 (Turnhout 1992).
119
See Avitus of Vienne, Letters and Selected Prose, pp. 187–93.
120
Avitus, ep. 8. See the commentary in Avitus of Vienne, Letters and Selected Prose,
pp. 220–4.
121
Especially St. Maurice d’Agaune: see B.H. Rosenwein, “Perennial prayer at
Agaune”, Monks and Nuns, Saints and Outcasts, ed. S. Farmer and B.H. Rosenwein
(Ithaca 2000) pp. 37–56.
122
Avitus, hom. 24; C. Perrat and A. Audin, “Alcimi Ecdicii viennensis episcopi
homilia dicta in dedicatione superioris basilicae”, Studi in onore di Aristide Calderini e
Roberto Paribeni 2 (Milan 1957) pp. 433–51.
123
Passio Sigismundi 1; Wood, “Ethnicity and the ethnogenesis of the Burgundians”,
   265

appears to have been no legend of Burgundian ethnogenesis. Of

course, one should be careful about this silence: it may reflect a gap
in our source material. It may, on the other hand, be Theodoric in
Italy who was unusual in sponsoring the composition of a literary
ethnogenesis for his people in the early sixth century. We know of
nothing comparable for the Franks, Vandals or Visigoths from the
same period—and even in the seventh century Isidore resorted only
to finding Biblical and Scythian origins for his Goths.124 Nevertheless,
whether or not it is the Ostrogoths rather than the Burgundians who
are the exception in the early sixth century, the absence of any
ethnogenetic tradition in the evidence for the court-culture of Gundobad
and Sigismund is worth noting.
Legislation and the letters of Avitus of Vienne are views from the
centre. It is possible that the Burgundians themselves took a different
view of their relations with the Romans. Discovering the opinion of
the Burgundian “man in the street” is, however, a difficult matter,
although the laws do provide some possible insights. The fact that
Burgundians were only too delighted to exploit differences between
Romans125 shows that they were extremely happy to seize opportu-
nities to benefit themselves. So too, their cavalier attitude towards
the Roman practice of oath-taking, which prompted the reinstitu-
tion of the traditional Burgundian custom of trial by battle,126 indi-
cates that there were plenty of incomers who did not take Roman
legal methods seriously. One may wonder whether the majority of
Burgundians remained unintegrated, as has been the case of many
immigrant groups down the centuries. One might also ask whether
such groups felt alienated from the increasingly Romanised Burgun-
dians of the royal court. Such an interpretation was argued by
Edward Thompson in the case of the Visigoths.127 One might well

p. 56. For the context in which the Passio was written, B.H. Rosenwein, “One site,
many meanings: Saint-Maurice d’Agaune as a place of power in the early Middle
Ages”, Topographies of Power in the Early Middle Ages, ed. M. de Jong, F. Theuws and
C. van Rhijn, The Transformation of the Roman World 6 (Leiden 2001) pp.
271–90, here p. 285. The fullest ethnogenesis of the Burgundians comes in Hildegar’s
Vita Faronis.
124
Isidore of Seville, Historia Gothorum, Wandalorum, Sueborum 1–3; recapitulatio 66,
ed. T. Mommsen, MGH AA 11 (Berlin 1894); Søby Christensen, “Cassiodorus and
the making of a history of the Goths”, p. 176.
125
Liber Constitutionum 22; 51,1–2.
126
Ibid. 45.
127
E.A. Thompson, “The Visigoths from Fritigern to Euric”, Historia 12 (1963)
266  . 

hypothesise that the adoption of a highly Romanised, not to say

extravagant,128 lifestyle at court alienated the rank and file Burgundian
settlers, and, indeed, helped weaken the kingdom. Such a hypothe-
sis, although ultimately unprovable, is plausible: that Sigismund, after
retiring to the monastery of Agaune, was supposedly betrayed to the
Franks by a Burgundian called Trapsta129 is one possible indication
that not all members of the gens were happy with developments at
court. So too, the conventus Burgundionum of the constitutiones extrava-
gantes may indicate that Godomar was having to conciliate Burgundian
warriors after Sigismund’s fall. On the other hand the Passio which
contains the information about Sigismund’s betrayal comes from the
eighth century, and may not be trustworthy as evidence. Even if it
is, Trapsta’s betrayal of his king may reflect a personal grudge rather
than any class sentiment. The best that can be said is that our view
of the development of the Burgundians as a gens is one formulated
at the centre of Gibichung power. The majority of Burgundians may
have had different views, rather more closely allied to the needs of
daily life.
In evaluating the largely official evidence for the Burgundian king-
dom it is difficult not to conclude with Justin Favrod, that “Le roy-
aume ne semble exister que pour l’étranger. Jamais personne ne
considère cette unité politique comme une patrie ou comme un lieu
d’origine. En outre, la Burgondie semble exister surtout par ses rois
[. . .]”130 To this one should add that the kings themselves had an
extraordinarily Roman view of their positions. The kingdom of the
Gibichungs does not appear to have been regarded by its rulers as
a Burgundian kingdom, even if Theodoric, Cassiodorus and Ennodius
saw it in that way. As already mentioned Aegidius is as good a par-
allel for Gundobad as any other contemporary.
As Favrod also remarked, “Le royaume burgonde ne reçut donc
pas de nom avant sa disparition.”131 It was in the Nachleben of the
Gibichung kingdom, as the basis for one of the Merovingian Teilreiche,

pp. 105–26, reprinted in id., Romans and Barbarians. The decline of the Western Empire
(Madison 1982) pp. 38–57.
128
See, for example, Avitus, ep. 86, with the commentary in Avitus of Vienne,
Letters and Selected Prose, pp. 279–84.
129
Passio Sigismundi 9.
130
Favrod, Histoire politique du royaume burgonde, p. 134.
131
Ibid., p. 136.
   267

that the phrase regnum Burgundionum and the related geographical term
Burgundia began to stick, and it was only in the Carolingian world
that the Burgundians finally received an ethnogenesis equivalent to
that of the Goths and the Lombards, bringing them from the Baltic
to the Rhône.132 By this time, however, the Burgundians were some-
thing other than the followers of the Gibichungs, although increas-
ingly they were those supposedly subject to the Lex Gundobada—a
title given to the Liber Constitutionum perhaps for no better reason
than the fact that Gundobad had been responsible for the most dis-
tinctive of all the laws of the code: that which substituted trial by
battle for the oath.133
Leaving aside the development of the Burgundians and the notion
of Burgundy or a Burgundian kingdom in the generations after the
fall of the Gibichung kingdom in 534, it is worth returning to empha-
sise the crucial lessons of the evidence for the previous century of
Burgundian history. What the Burgundian example does is draw
attention to a case for which the evidence, like that for the Ostrogothic
kingdom of Italy, comes largely from the early sixth century. The
picture in both instances is extremely Roman—although Theodoric,
unlike Gundobad or Sigismund, showed an interest in discovering a
past, or creating an ethnogenesis, for his people, and this suggests a
different awareness of the Germanic gens present in each regime.
The fact of this distinction between the Burgundians and the
Ostrogoths—and indeed the fact that the evidence for the Vandals
is different from either—suggests that we should not conclude that
all peoples in the late fifth and early sixth centuries were like the
Burgundians, or that all ruling dynasties in the period were like the
Gibichungs. On the other hand the evidence for the Burgundians
does give some indication of quite how close a barbarian dynasty
and its gentile following might come to integration into the Roman
world—at least in the fifth century: the sixth century is, of course,
another matter.
One indication of the difference between the late fifth and early
sixth century on the one hand and the late sixth on the other can
be seen in a simple comparison between the image of the Gibichung

132
Wood, “Ethnicity and the ethnogenesis of the Burgundians”, p. 56.
133
Wood, “Disputes in late fifth- and early sixth-century Gaul: some problems”,
p. 16.
268  . 

kingdom proposed here, largely on the basis of evidence from the

period 443–534, and that set out by Gregory of Tours. The rulers
of the Burgundians are introduced by the later historian as the
descendents of the (Visigothic) persecutor Athanaric,134 which is un-
likely to have been the case. In the bishop of Tours’ account their
early history was dominated by fraternal dissension, which led to the
killing of Chilperic, the exile of Chrotechildis, and her eventual mar-
riage to Clovis. Fraternal conflict is picked up again with the out-
break of hostilities between Gundobad and Godegisel.135 Gregory’s
picture of heresy and fratricide is momentarily deepened by a long
excursus on the religious relations of Gundobad and Avitus of Vienne,
taken largely from the latter’s own works.136 A bloodthirsty narrative
line is subsequently resumed with Sigismund’s murder of his son,
on the prompting of his second wife, and his griefstricken retirement
to his foundation of Agaune.137 The collapse of the king prompts
Chrotechildis to pursue her long-harboured vengeance for the mur-
der of her own father, Chilperic II, against the descendants of Gun-
dobad. This results first in the death of Sigismund, whose body is
thrown down a well in symbolic retaliation for the supposed death
of Chilperic, and then in that of Chrotechildis’ son Chlodomer on
the field of Vezéronce.138 The story only ends ten years later, with
the destruction of the Burgundian kingdom at the hands of the
Frankish queen’s other sons, Chlothar and Childebert.139 Factually,
much—although by no means all—of this account can be supported
from elsewhere, but the tone is completely at odds with that pre-
sented in sources from the early sixth century. Arianism is given an
importance by Gregory of Tours that it seems not to have had—
on the other hand from Gregory’s point of view it provides a use-
ful point of comparison for his interpretation of the conversion of
Clovis.140 So too, although fraternal conflict and murder probably
was a significant factor in the reign of Gundobad, Avitus of Vienne

134
Gregory of Tours, Historiae 2,28.
135
Ibid. 2,32–3.
136
Ibid. 2,34.
137
Ibid. 3,5.
138
Ibid. 3,6.
139
Ibid. 3,11. For the problems of Gregory’s narrative here, Wood, “Clermont
and Burgundy: 511–534”.
140
I.N. Wood, “Gregory of Tours and Clovis”, Revue Belge de Philologie et d’Histoire
63 (1985) pp. 249–72.
   269

gives a very different impression of this from that to be found in

Gregory of Tours,141 while the theme of Chrotechildis’ vengeance
seems to be a later gloss.142 Equally important, the Roman tone
which suffuses the evidence from either side of 500 is absent in
Gregory’s presentation. The Burgundians in the bishop of Tours’
narrative are a foil for his image of Clovis—which is also less Roman
than it could have been, although the Franks could scarcely have
pretended to have the Roman credentials which the Gibichungs cer-
tainly could boast.
A search for comparisons in the evidence from the late fifth and
early sixth centuries suggests a spectrum of kingdoms, which includes
not just the Burgundian kingdom of the Gibichungs, but also the
Roman kingdom of Aegidius—and which stretches at the other, less
Roman, extreme to the earliest English foundations: Gregory of Tours
may not have been wrong to have entitled Syagrius Romanorum rex.143
Gundobad in particular, but also Sigismund, needs to be considered
not just alongside the kings of the Vandals, Visigoths and Ostrogoths,
but also alongside Odoacer, Ricimer, Aetius and Stilicho. It would
not have taken radically different circumstances for the Gibichungs
and their followers to have remained little more than magistri mili-
tum with their federate armies—only hindsight requires that they are
understood as the founders and rulers of a Germanic kingdom. They
deserve to be categorised as a half-way house—as a type of feder-
ate group which was made irrelevant by the collapse of the Western
Empire and the rise of the Franks, but which suggests that there
were other possibilities in the period from 476 to 534 than those
offered by the Germanic kingdoms which won out.

141
Avitus of Vienne, Letters and Selected Prose, ep. 5, pp. 208–12.
142
Wood, The Merovingian Kingdoms, p. 43.
143
Gregory of Tours, Historiae 2,27.
This page intentionally left blank
 THE RELATIONSHIP BETWEEN FRANKISH GENS
AND REGNUM: A PROPOSAL BASED ON THE
ARCHAEOLOGICAL EVIDENCE

Michael Schmauder

How can an archaeologist, with the help of his specific source mate-
rial, succeed in making statements regarding the concepts of gens and
regnum, and how far is it possible for him to take a position in the
question of the relationship between the two? Both questions will be
answered at the beginning of this article. A prerequisite for doing
so is to consider the state and progress of historical research on both
these concepts, which are still subjects of discussion. Only in this
way may the knowledge and the considerations regarding this field
of research, as formulated on the basis of archaeological material,
be given its due historical relevance. Here, the old problem of inter-
disciplinary research comes to the fore: once the field of specific
archaeological methodology, like, for instance, typology and chronol-
ogy, is abandoned and questions of a sociological-cultural nature are
touched upon a link with historical terminology is unavoidable.
First of all, it should be made clear how far archaeological research
can take a position regarding the problem of the relation between
gens and regnum for the Franks at all. This, naturally, raises the ques-
tion of the nature of the available evidence. For the whole of the
Migration Period and the early Middle Ages it consists primarily of
funerary material. Excavated settlements—and especially published
excavation reports of these—are, at present, much rarer. Only small
excavations at Merovingian centres have thus far taken place, and
their subsequent destruction by later building activity does not give
much reason for optimism concerning future finds in these places.1

1
For a summary, see: P. Périn, “Paris, merowingische Metropole”, Die Franken—
Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung
im Reiss-Museum Mannheim 1 (Mainz 1996) pp. 121–8; H. Ament, “Die Franken in
den Römerstädten der Rheinzone”, ibid., pp. 129–37; H.-P. Kuhnen, “Zwischen
Reichs- und Stadtgeschichte—Trier in Spätantike und Frühmittelalter”, ibid., pp.
138–44; B. Päffgen and S. Ristow, “Die Römerstadt Köln zur Merowingerzeit”,
272  

Consequently, archaeological research of the Migration Period and

the early Middle Ages will need to base its results predominantly on
funerary finds. This raises—even before discussing the aspects to
which archaeology is related—the evaluation of the specific prob-
lems posed by grave-goods.

A question of interpretation? The sources for early medieval archaeology

The formulation of statements based on archaeology depends on the

state of the sources, the quality of which is primarily based on the
preservation of the archaeological finds.2 Several factors influence
archaeological finds and findings, and change their previous appear-
ance. The find itself reflects, at best, a part of a former “living” real-
ity. This is especially applicable to grave-goods. Hardly any type of
artefact is subject to a stronger “distortion” than that from a funer-
ary context.3 All the known data are the result of a selection process
carried out by the contemporary society and by the deceased, on
background of his past social position within a (complex) cultural
context. Hence, the valuation of grave-goods cannot be uniform.
Recently, Horst Wolfgang Böhme was prompted, by the exhibition,
“Gräber—Spiegel des Lebens”, to state that grave-goods essentially

ibid., pp. 145–59; M. Grünewald, “Worms zwischen Burgund und Saliern”, ibid.,
pp. 160–2.
2
For the interpretative possibilities presented by cases where finds are recovered
under favourable circumstances, see: O. Doppelfeld, “Das fränkische Knabengrab
unter dem Chor des Kölner Domes”, Germania 42 (1964) pp. 156–88; R. Marti,
“Das Grab eines wohlhabenden Alamannen in Altdorf UR, Pfarrkirche St. Martin”,
Jahrbuch der Schweizerischen Gesellschaft für Ur- und Frühgeschichte 78 (1995) pp. 83–130;
H. Amrein, A. Rast-Eicher and R. Windler with a contribution by E. Langenegger,
“Neue Untersuchungen zum Frauengrab des 7. Jahrhunderts in der reformierten
Kirche von Bülach (Kanton Zürich)”, Zeitschrift für Schweizerische Archäologie und Kunst-
geschichte 56 (1999) pp. 73–114. See also the section on the ship-burial on the Fallward
near Wremen.
3
Cf. most recently: A. Miron, “Die babylonische Verwirrung. Überlegungen zur
Terminologie der Spätlatène-Chronologie”, Studien zur Archäologie der Kelten, Römer
und Germanen in Mittel- und Westeuropa. Alfred Haffner zum 60. Geburtstag gewidmet, ed.
A. Müller-Karpe et al., Internationale Archäologie, Studia honoraria 4 (Rahden
Westf. 1998) pp. 429–38; id., “Zentrale Fragestellung—Das Grab als kulturhistorische
Quelle”, Kelten, Germanen, Römer im Mittelgebirgsraum zwischen Luxemburg und Thüringen,
ed. A. Haffner and S. von Schnurbein, Kolloquien zur Vor- und Frühgeschichte 5
(Bonn 2000) pp. 257–61.
   () 273

represent “a distorted mirror”4 and that “the valuation of grave-

goods as evidence for living culture [. . .] should always be made
against a background of the then binding traditions of grave-goods,
functioning as a filter in the selection of these”.5 But, how should
this filter be defined? On this Ursula Koch has remarked: “undoubt-
edly grave-furnishings were seen as serving as a worthy passage to,
and a suitable life in, the here-after”.6 Jürgen Hannig’s remarks in
his ground-breaking article “Ars donandi. Zur Ökonomie des Schenkens
im früheren Mittelalter” point in the same direction.7 He writes:
“[. . .] grave-goods should be considered less as documentation of a
living person’s rank and social standing than as means of securing
the deceased’s position in the distribution processes in the here-after”.8
In clear contrast to Hannig’s views is Hagen Keller’s opinion: “Going
from the finds, one ought to conjure up the image of how an expen-
sively prepared chamber-grave, with the deceased and his rich outfit
for the here-after, was ‘publicly’—that is, not only with his direct
neighbours, but also with present relatives—‘sent on his way’, in
order to understand how the precedence of the deceased, and par-
ticularly of those ‘left behind’, could be demonstrated and estab-
lished in the community.”9 Recently, Sebastian Brather has expressed

4
“eher ein[en] Zerrspiegel”: H.W. Böhme, “Franken und Romanen im Spiegel
spätrömischer Grabfunde im nördlichen Gallien”, Die Franken und die Alemannen
bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände zum Re-
allexikon der Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 31–58,
here p. 31.
5
“die Auswertung der jeweiligen Grabausstattungen als Zeugnis für die lebende
Kultur [. . .] in allen Fällen vor dem Hintergrund der gerade verbindlichen Beigaben-
sitte, die wie ein Filter bei der Beigabenauswahl wirkt, vorzunehmen ist”: ibid., n. 1.
6
“Zweifellos sollte die Grabausstattung einem würdigen Auftritt und angemesse-
nen Leben im Jenseits dienen”: U. Koch, “Stätten der Totenruhe—Grabformen und
Bestattungssitten der Franken”, Die Franken—Wegbereiter Europas 2, pp. 723–37, here
p. 728.
7
J. Hannig, “Ars donandi. Zur Ökonomie des Schenkens im früheren Mittelalter”,
Armut, Liebe, Ehre, ed. R. van Dülmen, Studien zur historischen Kulturforschung 1
(Frankfurt a. M. 1986) pp. 11–37; pp. 275–8.
8
“[. . .] die Grabbeigaben sind wohl weniger als Dokumentation des ehemaligen
Ranges und der sozialen Stellung im Diesseits anzusehen denn als Mittel, dem
Toten seine Stellung in den jenseitigen Distributionsprozessen zu bewahren”: ibid.,
p. 30.
9
“Man muß es sich—an den Funden orientiert—bildlich vorstellen, wie ein auf-
wendig vorbereitetes Kammergrab mit dem Toten und mitsamt seiner reichen Aus-
stattung für das Jenseits ‘vor allen’—wohl nicht nur vor der Nachbarschaft, sondern
auch vor herbeigeeilten Verwandten—‘beschickt’ wurde, um zu ahnen, wie der
Vorrang des Toten und wohl mehr noch der ‘Hinterbliebenen’ in der Gemeinschaft
274  

yet another opinion. According to him, the furnishings of graves, if

they are really an “image of the deceased’s role in life”10 (which,
according to Brather, can be evidenced only in individual cases),
may be considered rather as “a mirror of social prestige, and hence
of the social identity, of individuals and groups.”11 However, Frans
Theuws and Henk A. Hiddink have pointed out that “during the
interpretation [. . .] the specific meaning of the rites was as little
noted as the form chosen in such a case”.12 This consideration fits
in well with the hope, recently expressed by Hagen Keller, that it
will become possible to look at the “localisation” (Verortung) of “struc-
tural changes”, which from the middle of the fifth century led to
the creation of the Frankish realm and to a process of consolidation
for other Germanic gentes, rather than concentrating on “the typo-
logical predecessors and previous phases—burials, weapons as grave-
goods, provisions of material goods, royal graves in Germania, and
graves of foederati in Gaul”.13 In what follows, this will be discussed
as it relates to the case of the Franks.

demonstriert und gefestigt werden konnte”: H. Keller, “Strukturveränderungen in

der westgermanischen Welt am Vorabend der fränkischen Großreichsbildung. Fragen,
Suchbilder, Hypothesen”, Die Franken und die Alemannen bis zur “Schlacht bei Zülpich”
(496/97), ed. D. Geuenich. Ergänzungsbände zum Reallexikon der Germanischen
Altertumskunde 19 (Berlin-New York 1998) pp. 580–607, here p. 599.
10
“als Abbild der Rolle des Toten im Leben”: S. Brather, “Ethnische Identitäten
als Konstrukte der frühgeschichtlichen Archäologie”, Germania 78 (2000) pp. 139–77,
here p. 168. Regarding the valuation of grave-goods, this idea accords with the title
chosen by J. Biel for a grave from the Hallstatt-period at Hochdorf: J. Biel, “Die
Ausstattung des Toten. Reichtum im Grabe—Spiegel der Macht”, Der Keltenfürst von
Hochdorf. Methoden und Ergebnisse der Landesarchäologie (Stuttgart 1985) pp. 79–105.
11
“eher ein Spiegel des Sozialprestiges und damit der sozialen Identität von
Individuen und Gruppen”: Brather, “Ethnische Identitäten”, p. 168.
12
“bei der Interpretation [. . .] die spezifische Bedeutung der Riten ebensowenig
Beachtung [fand] wie die Form, die in einem solchen Fall gewählt wurde”: F. Theuws
and H.A. Hiddink, “Der Kontakt zu Rom”, Die Franken—Wegbereiter Europas 1, pp.
66–80, here p. 70.
13
“die typologischen Vorläufer und Vorstufen—Körperbestattung, Waffenbeigabe,
Sachgüterausstattung, Fürstengräber in Germanien, Foederatengräber in Gallien”:
Keller, “Strukturveränderungen in der westgermanischen Welt”, p. 593. Keller here
asks something of early Medieval archaeology that has, previously, been formulated
even clearer by Reinhard Wenskus: “No matter how important, before anything
else, answering typological questions may be, to historians the morphology as
researched by prehistorians [sic!] often seems unrewarding and uninteresting. Striv-
ing for methodological purity can, thus, lead in directions that no longer have any-
thing to do with historical questions. It may render archaeology uninteresting for
historians, it becomes sterile to them.” (R. Wenskus, “Randbemerkungen zum
Verhältnis von Historie und Archäologie, insbesondere mittelalterlicher Geschichte
und Mittelalterarchäologie”, Geschichtswissenschaft und Archäologie. Untersuchungen zur
   () 275

Gens and regnum: the possibilities of archaeology

In his book on the relationship between Alamans and Franks, Frank

Siegmund has listed those criteria that influence the furnishings of
a grave.14 In his opinion, we can consider a grave as the smallest
unit of information. These criteria, therefore, enable us to shed light
on wider relationships, like that between gens and regnum, which is
the emphasis here.
According to Siegmund, it is of special importance when work-
ing with funerary material primarily to consider aspects of gender
and of chronological phasing, which are of greater significance than
regional perspectives and ethnic predetermination. It is only after
working on these questions that one can proceed to those concern-
ing “vertical social structure”.
Against the background of Siegmund’s ideas, the following ques-
tions can now be formulated regarding the relationships between
Frankish gens and regnum, which will be discussed in what follows:

1. Can a gens, like the Franks, be identified by archaeology and, if

so, at which point in time and on the basis of which criteria?
2. Where geographically do the Franks appear as a gens, and how
does their further territorial development proceed?
3. What type of social structure is evidenced by the actions of the
Franks as a gens, and how does this change with the founding
and development of the Frankish regnum?
4. From an archaeological perspective, what is the position of the
Frankish gens and regnum over the course of the centuries?

The historical point of departure: a diffuse image

Research agrees that the name of the Franks derives from a Germanic
language-root (Old High German Franchon, Old English Francan, Old
Norse Frakkar).15 Generally, its meaning is now translated as “wild,

Siedlungs-, Wirtschafts- und Kirchengeschichte, ed. H. Jankuhn and R. Wenskus [Sigmaringen

1979] pp. 637–57, here p. 645).
14
F. Siegmund, Alemannen und Franken, Ergänzungsbände zum Reallexikon der
Germanischen Altertumskunde 23 (Berlin-New York 2000) pp. 77–80.
15
H. Ament, H.H. Anton, H. Beck, A. Quak, F. Rexroth, K. Schäferdiek, H. Steuer,
D. Strauch and N. Voorwinden, “Franken”, Reallexikon der Germanischen Altertumskunde
276  

courageous, fierce”, as suggested by late antique and early medieval

sources.16 It has been established that the name Franci appears for
the first time in 291 A.D., in a panegyric by Mamertinus on Emperor
Maximianus.17 From the description, it may be gathered that already
at this point in time the Franks were a gens, for explicit mention
is made of a rex.18 Other sources mentioning the Franci in third-
century contexts date from the fourth century or later.19 Probably the
Franks, like a whole series of other historical gentes (Bructeri, Chamavi,
Ampsivarii, Salians, Tubantes, Angrivarii, Falchovarii and Chattuarii[?]),
were a gens initially settled in the area of the northern Netherlands,
Westphalia and Lower Saxony. The occurrence of pirate raids20 make
it seem likely that the Frankish territory was located near the coast.21
Given their mention in Roman sources, the gentes of the Bructeri,
Chamavi, Ampsivarii and Chattuarii (who apparently disappeared in
the course of the fourth century to the profit of the Franks), pre-
sumably, did not live far from the borders of the Roman Empire.22
Julian, writing in the fourth century, mentions a people called the
Salians, who by modern scholars have been considered (a part of )
one of the Frankish tribes.23 The location of the Salian settlement is

9 (2nd edn., 1995) pp. 373–461, here “I. Philologisches § 1. Namenkundliches” by

H. Beck, pp. 373–4.
16
Ibid., p. 374. See further: R. Wenskus, Stammesbildung und Verfassung. Das Werden
der frühmittelalterlichen gentes (Köln-Graz 1961) pp. 513–4.
17
Panegyrici Latini 11,7, p. 376; 8,17, p. 382, ed. J. Herrmann, Griechische und
lateinische Quellen zur Frühgeschichte Mitteleuropas bis zur Mitte des 1. Jahrtausends u. Z.,
vol. 3: Von Tacitus bis Ausonius, Schriften und Quellen der Alten Welt 37,3 (Berlin
1991).
18
Ibid., p. 376: Francos ad petendam pacem cum rege venientes.
19
M. Springer, “Geschichtsbilder, Urteile und Vorurteile. Franken und Sachsen
in den Vorstellungen unserer Zeit und in der Vergangenheit”, 799 —Kunst und Kultur
der Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn. Beiträge zum Katalog der
Ausstellung Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz 1999) pp.
224–32.
20
Ibid.
21
The imprecise evidence for the settlement area of the Franks, during the third
and the beginning of the fourth centuries, may be associated with their origins in
the north-eastern coastal region.
22
Ch. Reichmann, “Frühe Franken in Germanien”, Die Franken—Wegbereiter Europas
1, pp. 55–65.
23
Cf. the critique by M. Springer on the meaning of Salian: M. Springer, “Salier
und Salisches Recht—Beobachtungen zu den Wörtern Salii und Salicus”, Die Franken—
Wegbereiter Europas 1, pp. 485–7; id., “Gab es ein Volk der Salier?”, Nomen et gens.
Zur historischen Aussagekraft frühmittelalterlicher Personennamen, ed. D. Geuenich, W. Hau-
brichs and J. Jarnut, Ergänzungsbände zum Reallexikon der Germanischen Altertums-
kunde 16 (Berlin-New York 1997) pp. 58–83.
   () 277

unclear, though the reference in Zosimus (3,6), that they were dri-
ven away by the Saxons, may indicate that they should be sited in
eastern Westphalia and western Lower Saxony.24 On the fourth-cen-
tury Tabula Peuteringiana, the name FRANCIA already appears east of
the Rhine in the vicinity of Colonia Ulpia Traiana (Xanten).25

Can the Frankish gens be identified archaeologically?

The archaeological point of departure—the third and fourth centuries
in the Netherlands, Westphalia and Lower Saxony

The inconclusive historical data, regarding in which area the Franci

settled, still provide the starting-point, also in archaeology, for deter-
mining and attributing finds and findings as connected to the Frankish
gens.26 Archaeologically it is impossible to differentiate the various
gentes in the area of the Netherlands, Westphalia and Lower Saxony
mentioned in the historical sources. Following Rafael von Uslar, they
are generally clustered together as the “Rhine-Weser Germanic finds
group”.27 In this respect, maps constructed on the basis of historical

24
See the various places the name occurs on the maps by Ch. Reichmann
(Reichmann, “Frühe Franken”, p. 57 fig. 39) and H.W. Böhme (H.W. Böhme,
“Franken oder Sachsen? Beiträge zur Siedlungs- und Bevölkerungsgeschichte in
Westfalen vom 4.–7. Jahrhundert”, Studien zur Sachsenforschung 12, ed. H.-J. Häßler
[Oldenburg 1999] pp. 43–73 fig. 1).
25
For coins with the legend Francia, see: M. Schmauder, “Verzögerte Landnahme?
Die Dacia Traiana und die sogenannten decumates agri”, Integration und Herrschaft.
Ethnische Identitäten und soziale Organisation im Frühmittelalter, ed. M. Diesenberger and
W. Pohl (Vienna 2002).
26
Böhme, “Franken oder Sachsen?”, pp. 45–6 fig. 1; id., “Ethnos und Religion
der Bewohner Westfalens. Methodische und historische Problematik”, 799—Kunst
und Kultur der Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn. Beiträge zum
Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz
1999) pp. 237–45, here p. 238 fig. 1. On the Saxon settlement in England, H. Härke
has recently rightly made the following critical remark: “The identification of the
tribes concerned, and (where present) their kingdoms or settlement areas, nearly
always depends on the ethnic interpretation of the archaeological material in com-
bination with the written sources—a method that was so common in archaeologi-
cal research of early history that, also after 1945, it has hardly been critically
discussed even once. On the other hand, for proto-history it has, since the end of
the war at the latest, been emphatically abandoned.” (H. Härke, “Sächsische Ethnizität
und archäologische Deutung im frühmittelalterlichen England”, Studien zur Sachsenfor-
schung 12, ed. H.-J. Häßler [Oldenburg 1999] pp. 109–22, here p. 110).
27
R. von Uslar, Westgermanische Bodenfunde des ersten bis dritten Jahrhunderts nach Christus
aus Mittel- und Westdeutschland, Germanische Denkmäler der Frühzeit 3 (Berlin 1938).
This group of finds has still not been reevaluated.
278  

sources, showing settlement areas of gentes presumed to be “con-

cealed” by the Franks, offer a representation for which there is no
archaeological proof. The reason for that lies in an exceptionally
rare use of grave-goods, further limited by the custom of cremating
the dead. On the continuity of the “Rhine-Weser Germanic finds
group”, Hermann Ament states: “The Rhine-Weser Germanic form
complex of the later imperial period [second to third century A.D.,
author’s note] continues, without any noteworthy break, a finds com-
plex of the same name from the early imperial period, as can be
gathered from the many examples of settlement continuity in essen-
tially the same areas and from the constant and organic develop-
ment of objects”.28 Even using the groupings of a gens of Franks and
a gens of Saxons provided by the historical sources, it is impossible
to divide the material culture of the fourth and fifth centuries in
Westphalia. A similar point has recently been made by Christoph
Grünewald: “I think that our investigation around Westphalia has
shown that a clear definition of ‘Frankish’ and ‘Saxon’ is not possi-
ble. Neither in grave-goods or grave-forms, nor in the organisation
of cemeteries or in the alignment of individual graves are clear cri-
teria available”.29 Archaeology clearly demonstrates that those Germans,
who in the fourth and fifth centuries lived and were buried in north-
ern Gaul as foederati or—to use a neutral term—as settlers, originally
came from an area centred around the northern Netherlands, West-
phalia and the Elbe-Weser triangle in Lower Saxony.30 These con-

28
“Der jüngerkaiserzeitliche rheinwesergerm. Formenkreis setzt ohne nennenswerten
Bruch eine gleichnamige Fundgruppe der Ält. RKZ fort, wie an der mehrfach nach-
weisbaren Siedlungskonstanz, dem im wesentlichen gleichen Verbreitungsgebiet und
der gleichsam organischen Weiterentwicklung der Sachformen abzulesen ist”: Ament,
Anton, Beck, Quak, Rexroth, Schäferdiek, Steuer, Strauch and Voorwinden, “Franken”,
here p. 388.
29
“Ich denke, unser Rundblick durch Westfalen hat gezeigt, daß eine klare
Definition von Fränkischem und Sächsischem nicht gelingt. Weder die Grabbeigaben
noch die Grabformen, weder die Anlage der Gräberfelder noch die Ausrichtung
der Einzelgräber geben klare Kriterien an die Hand”: Ch. Grünewald, “Neues zu
Sachsen und Franken in Westfalen”, Studien zur Sachsenforschung 12, ed. H.-J. Häßler
(Oldenburg 1999) pp. 83–108, here p. 107.
30
See, for instance: H.W. Böhme, Germanische Grabfunde des 4. bis 5. Jahrhunderts.
Zwischen unterer Elbe und Loire. Studien zur Chronologie und Bevölkerungsgeschichte, Münchener
Beiträge zur Vor- und Frühgeschichte 19 (München 1974); id., “Söldner und Siedler
im spätantiken Nordgallien”, Die Franken—Wegbereiter Europas 1, pp. 91–101, esp. fig.
68 and 75; id., “Sächsische Söldner im römischen Heer. Das Land zwischen Ems
und Niederelbe während des 4. und 5. Jahrhunderts”, Über allen Fronten. Nordwest-
deutschland zwischen Augustus und Karl dem Großen, ed. M. Fansa (Oldenburg 1999)
   () 279

nections can be proven not only on basis of the well-known late Roman
items of clothing (though the cultural significance of these is doubted),31
but also, among other things, by the distribution of hand-made

pp. 49–73. Further, a considerable proportion of funerary finds indicates that a

substantial migration from central Germany took place from the middle of the fifth
century onwards. (Ch. Reichmann, “Krefeld [2352/003]”, Bonner Jahrbücher 197
[1997] pp. 314–6; H.W. Böhme, “Les Thuringiens dans le Nord du royaume franc”,
Revue Archéologique de Picardie 3/4 [1988] pp. 57–69).
31
Cf. the critique by Guy Halsall: G. Halsall, “The origins of the Reihengräberzivilisation:
forty years on”, Fifth-century Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton
(Cambridge 1992) pp. 196–207; id., Early Medieval Cemeteries: An archaeological intro-
duction to burial in the Post-Roman West (Glasgow 1995); id., “Archaeology and the late
Roman frontier in northern Gaul: The so-called ‘Föderatengräber’ reconsidered”,
Grenze und Differenz im frühen Mittelalter, ed. W. Pohl and H. Reimitz, Forschungen
zur Geschichte des Mittelalters 1 (Wien 2000) pp. 167–80. Some remarks should
briefly be made here; though these cannot replace a more detailed examination of
the arguments raised by G. Halsall. The idea that graves containing late Roman
dress elements can be regarded as Bagaud burials raises various problems. Firstly,
from an archaeological perspective, the fundamental question that ought to be
addressed is why the members of the Bagaud movement would use a different bur-
ial ritual? It is also difficult to explain why similar archaeological phenomena occur
in other border regions of the Roman Empire, including areas where there is no
evidence for a Bagaud movement, but rather for the settlement and stationing
of barbarian foederati. For the area on the Lower Danube, the latest research by
K. von der Lohe, for the cemeteries at Piatra Freçatei, shows that already after one
generation a newly arrived people, on the basis of its particular manner of grave
construction, can archaeologically only be recognized as “non-Romans”. For his-
torical research, in particuler the fact that there is evidence for Bagauds in the third
and fifth centuries, but not for the fourth, will cause problems (S. Szadeczky-Kardoss,
“Bagaudae”, Realencyclopädie der classischen Altertumswissenschaft, Supplement 11 [1968] pp.
346–54; R. Van Dam, Leadership and Community in Late Antique Gaul [Berkeley 1985]).
It is not clear to the author why Halsall presupposes an Alamannic origin of this
phenomenon. As far as the author is aware, there are no precedents of this in mod-
ern research. J. Kleemann has recently worked on the emergence of the custom of
interment burials. His conclusions clearly point in another direction ( J. Kleemann,
“Zum Aufkommen der Körperbestattung in Niedersachsen”, Studien zur Sachsenforschung
13, ed. H.-J. Häßler [Oldenburg 1999] pp. 253–62). Also, mention ought to be
made of the newly discovered burial of a body in Leer (Lower Saxony), which dates
from the second century. The deceased had been buried “lying on his right side
with the legs drawn up in a slightly squatting position” (R. Bärenfänger, “Fundmel-
dungen: 6. Leer FstNr. 2710/5:10”, Emdener Jahrbuch 77 [1997] p. 269). It is inter-
esting that the evidence for interment burials in Germanic settlements so far has
had no wider consequences for the discussion (W. Haarnagel, Die Grabung Fed-der-
sen Wierde. Methode, Hausbau, Siedlungs- und Wirtschaftsformen sowie Sozialstrukturen [Wiesbaden
1979] pp. 230–8). See now: M.D. Schön, “Gräber und Siedlungen bei Ottendorf-
Westerwörden, Landkreis Cuxhaven”, Probleme der Küstenforschung im südlichen Nordseegebiet
26 (1999) pp. 123–208, esp. pp. 149–61. Halsall’s arguments on the origins of
specific shapes of fibulas, like the Tutulusfibel (trumpet brooch) and Stützarmfibel (sup-
porting-arm brooch), are dubious. The supposition that the presence of these fibulas
were the only difference from the female graves generally acknowledged as late
280  

pottery in Belgium, the Netherlands and northern Germany.32 With

the inclusion of the north German coastal region, the area that is
usually classified as typically “Saxon” is incorporated too.33 From a
cultural perspective, it is important that the finds, recovered from
settlements situated in a wide strip along the Roman border, belong

Roman, in which necklaces, earrings, bracelets and hair pins are present, limits
the ethnic determination claims unnecessarily (on the Roman customs of grave-
goods, see the thorough study of the cemetery at Kaiseraugst: M. Martin, Das
spätrömisch-frühmittelalterliche Gräberfeld von Kaiseraugst, Kt. Aargau, 2 vols., Basler Beiträge
zur Ur- und Frühgeschichte 5 [Derendingen 1976/91], esp. vol. 2, pp. 293–307).
In fact, the question of the provenance of fibula-types is less fundamental than the
use of particular fibulas in relation to a specific costume (it is beyond doubt that
several Germanic fibula-types were derived from Roman examples. The Tutulusfibel
can, however, be traced back to Barbaricum. Cf.: V. Brieske, Schmuck und Trachten-
bestandteile des Gräberfeldes von Liebenau, Kr. Nienburg/Weser, Studien zur Sachsenfor-
schung 5,6 [Oldenburg 2001], esp. pp. 15–9). Unless found in northern Gaul at a
future date, these fibula-types are undisputedly associated with the west Germanic
peplos-dress with cloak (M. Martin, “Fibel und Fibeltracht. § 41. W-germ. F. und
F.-Tracht. A. Diskontinuität der F.-Tracht zur RGZ und MZ”, Reallexikon der
Germanischen Altertumskunde 8 [2nd edn., 1994] pp. 549–56; id., “Tradition und Wandel
der fibelgeschmückten frühmittelalterlichen Frauenkleidung”, Jahrbuch des Römisch-
Germanischen Zentralmuseums Mainz 38,2 [1991] pp. 629–80). Further, in numerous
graves, which have been interpreted as those of foederati, there are a variety of ele-
ments that point to the Germanic lands. Most recently, Claus von Carnap-Bornheim
has made an exemplary study of this for the grave of the chef militaire of Vermand
(C. von Carnap-Bornheim, “Kaiserzeitliche germanische Traditionen im Fundgut
des Grabes des ‘chef militaire’ in Vermand und im Childerich-Grab in Tournai”,
Germanen beiderseits des spätantiken Limes, ed. Th. Fischer, G. Precht and J. Tejral, Spisy
Archeologického Ústavu Av ’R Brno 14 [Köln-Brno 1999] pp. 47–61). Lastly,
Halsall’s argument (Halsall, “The origins of the Reihengräberzivilisation”, p. 200) that
the inclusion of an obolus (a coin placed in the mouth) did not differ between
Germans and Romans, is still open to debate. M. Martin has researched the late
antique/Frankish cemetery at Krefeld-Gellep from this perspective, and has been
able to show that—in the fourth and fifth centuries and later—the inclusion of an
obolus was there, as elsewhere, a typical Germanic custom of Roman origin (Martin,
Das spätrömisch-frühmittelalterliche Gräberfeld, esp. vol. 2, pp. 165–9).
32
P. de Paepe and L. van Impe, “Historical Context and Provenancing of Late
Roman Hand-Made Pottery from Belgium, the Netherlands and Germany. First
Report”, Archaeologie in Vlanderen 1 (1991) pp. 145–80; Y. Hollevoet and J.-P. van
Roeyen, “Germanic settlers at Sint-Gillis-Waas? (Prov. East-Flanders)”, Archaeologie
in Vlandern 2 (1992) pp. 209–21; Theuws and Hiddink, “Der Kontakt zu Rom”,
p. 78.
33
From a historical perspective, this is no real surprise: in late antique sources,
more than once incursiones of Franks and Saxons undertaken together are mentioned.
( Julian, Or. 1,34d, ed. W.C. Wright [The works of the Emperor Julian] [New York
1930] pp. 88–91; Ammianus Marcellinus, Res Gestae—Römische Geschichte 27,8,5, ed.
W. Seyfarth, 4 vols. [Berlin 1970] pp. 76–7; Zosimus, Neue Geschichte 3,6–7, ed.
O. Veh and S. Rebenich, Bibliothek der griechischen Literatur 31 [Stuttgart 1990]
pp. 124–5). Cf.: B. Gutmann, Studien zur römischen Außenpolitik in der Spätantike (364–395
n. Chr.) (Bonn 1991) pp. 44–5.
   () 281

to a limes area that was intensively drawn into the Roman sphere
of influence. This is demonstrated, especially, by the large amount
of Roman small coinage,34 but also by the adoption of Roman tech-
niques of ceramics production,35 as well as the presence of Roman
glass and coarseware ceramics.36
It ought, therefore, to be borne in mind that, on the basis of the
archaeological finds from the fourth and the first half of the fifth
centuries, no identification as an individual gens (as an ethnically
identifiable group) is possible for the Franks, though they are, already
at that time, mentioned in historical sources.

Evidence for Frankish settlement activity in Northern Gaul?

Germans between the Rhine and the Loire

With his book on Merovingian Bügelfibeln (bow fibulas), Alexander

Koch is the latest to deal with the question of Germanic settlers
in the Dutch, Belgian, west German and north French regions.37
Koch’s discussions are largely based on research done by Hermann
Ament.38

34
D.G. Wigg, Münzumlauf in Nordgallien um die Mitte des 4. Jahrhunderts n. Chr.,
Studien zu Fundmünzen der Antike 8 (Berlin 1991); F. Berger, Untersuchungen zu
römerzeitlichen Münzfunden in Nordwestdeutschland, Studien zu Fundmünzen der Antike
9 (Berlin 1992).
35
G. Mildenberger, “Terra Nigra aus Nordhessen”, Fundberichte aus Hessen 12
(1972) pp. 104–26; M. Erdrich, “Terra Nigra-Fußschalen wie Chenet 342 oder
Gellep 273: eine salisch-fränkische Keramikgattung”, Germania 76 (1998) pp. 875–83;
A.D. Verlinde and M. Erdrich, “Eine germanische Siedlung der späten Kaiserzeit
mit umwehrter Anlage und umfangreicher Eisenindustrie in Heeten Overijssel, Nie-
derlande”, Germania 76 (1998) pp. 693–719.
36
Most recently, Michael Erdrich has stated that “the Roman evidence found
in the Germanic settlements and burials outside the borders of the Empire can
hardly be considered as the result of cross-border exchange of goods, but should
rather be regarded as the effect of Roman decisions in foreign politics and security
at times of far reaching domestic and military crises.” (M. Erdrich, “Wirtschaftsbezie-
hungen zwischen der Germania inferior und dem germanischen Vorland—ein
Wunschbild”, Germania inferior. Besiedlung, Gesellschaft und Wirtschaft an der Grenze der
römisch-germanischen Welt, ed. T. Grünewald with H.-J. Schalles, Ergänzungsbände
zum Reallexikon der Germanischen Altertumskunde 28 [Berlin-New York 2001]
pp. 306–35, here p. 335).
37
A. Koch, “Zur Problematik der sog. fränkischen Landnahme”, id., Bügelfibeln
der Merowingerzeit im westlichen Frankenreich, Römisch-Germanisches Zentralmuseum
Mainz, Monographien 41,2 (Mainz 1998) pp. 565–80.
38
The problem of the different interpretation of the archaeological material posed
282  

The early phase of Germanic settlement, in the area described,

has been thoroughly discussed archaeologically by Horst Wolfgang
Böhme, on the basis of funerary material.39 The conclusions drawn
by Böhme have been widely recognised by research. According to
these findings, there is, in conformity with the cemetery material
from the fourth and fifth centuries, evidence for the presence of a
substantial number of Germani, or—as Böhme even more carefully
expresses it—of “non-Roman people”40 in northern Gaul. From the
late fourth century onwards, an increasing number of cemeteries
appears (Furfooz, Vermand, Vireux-Molhain, Vron etc.)41 and a
greater number of Germanic burials emerge at the edges of exist-
ing late Roman cemeteries. These are predominantly recognisable
by the presence of weapons in male graves and by the widely prac-
tised custom of including grave-goods.
The newly founded burial grounds were, until the third quarter
of the fifth century, generally used for two generations. Böhme assumes

by components of French early medieval archaeological research can not be dis-

cussed here. On this subject, see Koch, “Zur Problematik der sog. fränkischen Land-
nahme”, pp. 567–9.
39
The settlements of the laeti, mentioned as early as the third century (K. Böhner,
“Zur historischen Interpretation der sogenannten Laetengräber”, Jahrbuch des Römisch-
Germanischen Zentralmuseums Mainz 10 [1963] pp. 139–67; R. Günther, “Laeti, foederati
und Gentilen in Nord- und Nordwestgallien im Zusammenhang mit der sogenann-
ten Laetenzivilisation”, Zeitschrift für Archäologie 5 [1971] pp. 39–59; id., “Einige
neue Untersuchungen zu den Laeten und Gentilen in Gallien im 4. Jahrhundert
und zu ihrer historischen Bedeutung”, Klio 69 [1977] pp. 311–21) are becoming
increasingly comprehensible through archaeological research ( J. Henning, “Germanische
Siedlungen im spätrömischen Nordgallien. Neue Ergebnisse archäologischer For-
schungen”, Zeitschrift für Archäologie 23 [1989] pp. 211–7; P. van Ossel, Établissements
ruraux de l’Antiquité tardive dans le nord de la Gaule, Gallia, Suppl. 51 [Paris 1992];
Theuws and Hiddink, “Kontakt zu Rom”, esp. pp. 75–80). Possible evidence for
the presence of Germanic settlers, as early as “the first to second third of the fourth
century” (K.H. Lenz, Siedlungen der Römischen Kaiserzeit auf der Aldenhovener Platte [Köln-
Bonn 1999] p. 89), was provided by the archaeological analysis of K.H. Lenz for
the area of the Aldenhovener Platte c. 20 km north-east of Aachen. The label
“Germanic” for this settlement is, however, exclusively based on the wood con-
struction-type.
40
Böhme, “Söldner und Siedler”, p. 92.
41
Furfooz: A. Dasnoy, “La nécropole de Furfooz”, Annales de la Société Archéologique
de Namur 55 (1969) pp. 121–94; Böhme, Germanische Grabfunde, pp. 288–90, Kat.-Nr.
120, pl. 88–90; Vermand: ibid., pp. 330–4, Kat.-Nr. 192, pl. 136–142 (with lit.);
La Picardie, berceau de la France. Clovis et les derniers Romains. 1500ème anniversaire de la
bataille de Soissons 486–1986 (Amiens 1986) pp. 70–2, Kat.-Nr. 24; Vireux-Molhain:
Le cimitière et la fortification du Bas-Empire de Vireux-Molhain, Dép. Ardennes, ed. J.-P.
Lemant (Mainz 1985); Vron: C. Seillier, “Les tombes de transition du cimetière
germanique de Vron (Somme)”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz
36 (1989) pp. 599–634.
   () 283

that the Germanic settlement “obviously took place in the shape of

small family units of closed ethnicity, whereby gentile identity was
preserved”.42 On basis of the archaeological material, this statement
is hard to verify, which also led Böhme to make the following obser-
vation: “Unfortunately, it is hard to decide in single cases exactly
which tribes west of the Rhine once belonged to the ‘imperial
Germans’ (Reichsgermanen) in northern Gaul”.43 Neither are develop-
ments in attire, for which notable variations are apparent in Gaul
during the mid- and second half of the fifth century, necessarily evi-
dence of different ethnicities.44 The distribution of the “simple belt-
fittings with two decorated pins, a buckle with reinforced fittings, a
lancet-shaped tongue and three to five holes”,45 which are mainly
found in Germania II and Belgica II, but nowhere in the area
between Loire and Somme, may indicate different ethnicities. On
the subject of the distribution of this type of belt-fitting, Heike Aouni
has rightly critically remarked that “the belt-type A of the simple
belt-fittings with buckles of Haillot variant A type [. . .] are products
of various workshops, supplying different groups, and consequently
these types have no a priori power of ethnic expression”.46

42
“offensichtlich als kleine geschlossene ethnische Einheiten im Familienverband
erfolgte, wobei gentile Identität gewahrt blieb”: Böhme, “Söldner und Siedler”,
p. 97. Unfortunately, there is still not sufficient anthropological research to confirm
Böhme’s—probably correct—assumption.
43
“Leider ist im Einzelfall nur sehr schwer zu entscheiden, welchen rechts-
rheinischen Stämmen im einzelnen die nordgallischen ‘Reichsgermanen’ einst ange-
hörten”: ibid., p. 97.
44
Böhme draws a parallel between this finding and the settlement areas of Salian
and Rhine Franks, as known from written sources (ibid., p. 99).
45
Ibid.
46
“[. . .], daß Gurttyp a der einfachen Garnituren mit Schnallen vom Typ Haillot
Variante a [. . .] Produkte verschiedener Fabrikationsstätten mit unterschiedlichen
Belieferungsströmen sind, so daß diese Typen nicht a priori ethnische Aussagekraft
besitzen”: H. Aouni, “Das spätantik-frühmittelalterliche Gräberfeld von Jülich—die
‘einfachen’ Gürtelgarnituren”, Archäologie des Frankenreiches—Neueste Forschungen. Internationales
Kolloquium anläßlich der Ausstellung “Die Franken—Les Francs” Berlin, 10. bis 12. Oktober
1997, ed. W. Menghin, Acta Praehistorica et Archaeologica 30 (1998) pp. 19–37, here
p. 36. Frank Siegmund evaluates, in a similarly critical manner, the ethnic inter-
pretation on the basis of the spread of find-types (Siegmund, Alemannen und Franken,
p. 84). From the context of his study on the Goldgriffspatha (sword with gold mounted
grip) of the Merovingian period, it becomes clear that H.W. Böhme is convinced
that the spread of certain dress elements and weaponry primarily represent the mar-
ket for certain production centres. In the same spirit Böhme writes: “The inhabit-
ants of northern Gaul and its immediately neighbouring northern and eastern border
zones—especially those of Germanic warrior bands who had become foederati—were,
284  

There is no doubt that among the numerous Germani who set-

tled in northern Gaul during the fourth and fifth centuries, there
was a substantial number of those Germani who in the historical
sources are indicated as Franks. To identify them in the archaeo-
logical finds-material with real certainty is, however, impossible.
Only in the second half of the fifth century does a process of con-
solidation begin to appear in the material culture that, more clearly
than before, demonstrates a unification of burial customs and of the
range of grave-goods. Still, it should be emphasised here, that behind
the political term “Merovingian realm” one cannot archaeologically
recognise a closed ethnicity during the fifth and sixth centuries. In
all regions that fell under the rulership of the Merovingian dynasty,
we find burials of various ethnicities in greater or lesser numbers.47

The south of France: a gradual southward expansion

According to the archaeological sources, the conquest of southern

Gaul was far less intensive than that of the northern Gaulish regions.
The research undertaken over the last two decades gives a well-

between the end of the fourth century and the middle of the fifth, exclusively sup-
plied with the products of specialised workshops in northern Gaul.” (H.W. Böhme,
“Der Frankenkönig Childerich zwischen Attila und Aëtius. Zu den Goldgriffspathen
der Merowingerzeit”, Festschrift für Otto-Herman Frey, ed. C. Dobiat, Marburger Studien
zur Vor- und Frühgeschichte 16 [Hitzeroth 1994] pp. 69–110, here p. 102).
47
For Gaul, see for instance: M. Kazanski, “À propos de quelques types de
fibules ansées de l’époque des Grandes Invasions trouvées en Gaule”, Archéologie
Médiévale 14 (1984) pp. 7–27; id., “La diffusion de la mode danubienne en Gaule
(fin du IVe siècle—début du VIe siècle): essai d’interpretation historique”, Antiquités
Nationales 21 (1989) pp. 59–73; id., “Les barbares ‘orientaux’ dans l’armée romaine
en Gaule”, Antiquités Nationales 29 (1997) pp. 201–17; H.W. Böhme, “Zur Bedeutung
des spätrömischen Militärdienstes für die Stammesbildung der Bajuwaren”, Die Baju-
waren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-
Salzburg 1988) pp. 23–37; id., “Les Thuringiens dans le Nord”; id., “Studien zu
Gallien in der Spätantike”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 33
(1986) pp. 844–8; C. Seillier, “Céramique de type anglo-saxon du cimetière de
Waben (Pas-de-Calais)”, Antiquités Nationales 21 (1989) pp. 83–9; A. Wieczorek, “Die
Ausbreitung der fränkischen Herrschaft in den Rheinlanden vor und seit Chlodwig
I.”, Die Franken—Wegbereiter Europas 1, pp. 240–60, esp. fig. 172; A. Koch, “Fremde
Fibeln im Frankenreich. Ein Beitrag zur Frage nichtfränkischer germanischer Ethnien
in Nordgallien”, Archäologie des Frankenreiches, pp. 69–89; W. Ebel-Zepezauer, “Ostger-
manische Blechfibeln des 5. und 6. Jahrhunderts zwischen Rhein und Garonne”,
Archäologisches Korrespondenzblatt 29 (1998) pp. 297–303; V. Bierbrauer, “Les Wisigoths
dans le royaume franc”, Antiquités Nationales 29 (1997) pp. 167–200.
   () 285

differentiated image of this process. Cemeteries like that of L’Isle-

Jourdain (situated between Bordeaux and Toulouse) testify to the
incorporation of this region into the Frankish realm, already at the
beginning of the sixth century.48 The necropolis, in use between about
510 and 550, contains a range of grave-goods that can be compared
without hesitation to those in the north-east of France. Burials like,
for instance, 322 and 333 can be regarded as characteristic here.
Tomb 322 was of a male and included, among other things, a fran-
cisca, a Schilddornschnalle (buckle with shield-shaped section attaching
the buckle to the tongue), knives, a Knickwandtopf (biconical clay pot)
and various coins.49 The female grave 333 contained, among other
finds, a pair of Fünfknopf fibulas (brooch with five terminals at the
headplate) and a pair of Scheibenfibeln (disc fibulas) (that is, a com-
plete set of fibulas), a rock-crystal with setting, amber beads, a small
knife, a coin pendant as well as a Knickwandtopf with spout.50 The
research results from this site confirm in exemplary fashion the work
of Françoise Stutz, who has studied the Frankish finds of middle and
southern France.51 According to her, there is a significant density of
Frankish finds in the entire area, which points to a Frankish presence
from the sixth century.

Different developments: the west and east Frankish regions (selected aspects)

The baptism of Clovis in the year 508 marks a decisive step in the
acculturation of Frankish life with the dominant “Roman cultural
models”.52 This change is especially clear when comparing Clovis’
burial to that of his father, Childeric.53 While Childeric was interred

48
J.-L. Boudartchouk, “La nécropole francque de Ictium à L’Isle-Jourdain (Gers,
Midi-Pyrénées, France)”, Archäologie des Frankenreiches, pp. 126–36.
49
Ibid., p. 133 fig. 5.
50
Ibid., pp. 133–5 fig. 6.
51
F. Stutz, Les objets mérovingiens de type septentrional dans le Sud de la Gaule, Mémoire
de Maîtrise à l’Université Paris 1 (Paris 1993); ead., Les objets mérovingiens de type
septentrional dans la moitié sud de la Gaule. Du corpus à l’histoire: position des problèmes,
Mémoire de DEA à l’Université (Paris 1994). See further: E. James, The Merovingian
archaeology of south-west Gaul, 2 vols., British Archaeological Reports, Supplementary
Series 25 (Oxford 1977).
52
See, with older literature: A. Dierkens, “Die Taufe Chlodwigs”, Die Franken—
Wegbereiter Europas 1, pp. 183–91; M. Rouche, “Die Bedeutung der Taufe Chlodwigs”,
ibid., pp. 192–9.
53
M. Müller-Wille, “Königtum und Adel im Spiegel der Grabfunde”, Die Franken—
286  

in a burial mound surrounded by horse burials, in accordance with

the funerary practices for the highest social ranks east of the Rhine,54
Clovis chose a burial ad sanctos.55 The royal example was immedi-
ately copied, and after the turn of the sixth century noble burials
can be found in the churches at Chartres, Cologne (St Severin),
Rouen, Tournai and Xanten. Further, a substantial number of new
churches was constructed, which from the very beginning were used
as burial places for the Merovingian noble class.56 This development,
which is characteristic of sixth-century western Francia, only reaches
the eastern Frankish lands around 600. From the second third of
the seventh century onwards, foundations of churches with burials
inside the churches or in their immediate neighbourhood also occur
in the Baiuvarian lands. East of the Lech, according to Böhme, buri-
als in churches only appear from the eighth century onwards.57
The reduction of grave-goods follows a similar chronological and
geographical development to that of high status burials in or near
churches. While most of these practices in the south-western Frankish
regions came to an end as early as “the end of the sixth century”,58
in the east Frankish lands (depending on the region) they continued
far into the eighth century—albeit in increasingly reduced form. The
early abandonment of the inclusion of grave-goods in the south-west
of the Frankish realm is causally connected to the dominance of the
“Roman cultural model”.59 In the area between the Seine, the Marne
and the Meuse, this Roman influence, and hence Frankish accul-
turation with the “Roman cultural model”, is most notable in the

Wegbereiter Europas 1, pp. 206–21; id., Zwei religiöse Welten: Bestattungen der fränkischen
Könige Childerich und Chlodwig, Akademie der Wissenschaften und der Literatur, Abhand-
lungen der Geistes- und Sozialwissenschaftlichen Klasse, Jahrgang 1998,1 (Stuttgart
1998) pp. 3–45.
54
R. Brulet, “Tournai und der Bestattungsplatz um Saint-Brice”, Die Franken—
Wegbereiter Europas 1, pp. 163–70; Böhme, “Frankenkönig Childerich”, pp. 69–74.
55
K.-H. Krüger, Königsgrabkirchen der Franken, Angelsachsen und Langobarden bis zur
Mitte des 8. Jahrhunderts, Münstersche Mittelalter-Schriften 4 (München 1971) pp.
40–54.
56
See the distribution-maps by Böhme (H.W. Böhme, “Adelsgräber im Franken-
reich. Archäologische Zeugnisse zur Herausbildung einer Herrenschicht unter den mero-
wingischen Königen”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 40 (1993)
pp. 397–534, here p. 520 fig. 99).
57
Ibid., p. 519.
58
V. Bierbrauer, “Romanen im fränkischen Siedelgebiet”, Die Franken—Wegbereiter
Europas 1, pp. 110–20, here p. 113.
59
Ibid.
   () 287

absence of weapons as grave-goods from around 600 onwards. From

an archaeological point of view, this, too, blurs the borders between
Franks and Romans in the west Frankish lands around this time. It
is, therefore, hardly possible to draw conclusions regarding the
relation between gens and regnum on the basis of the archaeological
material.

Some remarks on the social stratification of the archaeological material

An example of the difficulty in using specific historical terminology

for certain archaeological find-complexes is the often-discussed term
“nobility”. The question whether a nobility existed in the early
Frankish period has been, and still is, intensively discussed among
historians. Arguments based on archaeological evidence play an im-
portant part in this discussion. Historians who have contributed sig-
nificantly to this discussion have been Franz Irsigler,60 Heike Grahn-
Hoek,61 Thomas Zotz,62 Hans-Werner Goetz,63 Gabriele von Olberg64
and most recently Margarete Weidemann.65 The results of the research
regarding the use of the term “nobility” emphasise the need to dis-
tinguish between indicating a social group privileged by lordship,
descent, status and property, or using the term for a group with a
clear and precise juridical definition. Hans-Werner Goetz has demon-
strated that the former usually seems unproblematic, since nobility
can be equated with the Latin nobilitas. In contrast, M. Weidemann

60
F. Irsigler, Untersuchungen zur Geschichte des frühfränkischen Adels, Rheinisches Archiv
70 (Bonn 1969).
61
H. Grahn-Hoek, Die fränkische Oberschicht im 6. Jahrhundert. Studien zu ihrer recht-
lichen und politischen Stellung, Vorträge und Forschungen Sonderband 21 (Sigmaringen
1976). See further: K. Schreiner, “Adel oder Oberschicht? Bemerkungen zur sozio-
logischen Schichtung der fränkischen Gesellschaft im 6. Jahrhundert”, Vierteljahresschrift
für Sozial- und Wirtschaftsgeschichte 68 (1981) pp. 225–31.
62
T. Zotz, “Adel, Oberschicht, Freie. Zur Terminologie der frühmittelalterlichen
Sozialgeschichte”, Zeitschrift für die Geschichte des Oberrheins 125 (1977) pp. 2–20.
63
H.-W. Goetz, “‘Nobilis’. Der Adel im Selbstverständnis der Karolingerzeit”,
Vierteljahresschrift für Sozial- und Wirtschaftsgeschichte 70 (1983) pp. 153–91; id., Moderne
Mediävistik. Stand und Perspektiven der Mittelalterforschung (Darmstadt 1999).
64
G. von Olberg, Die Bezeichungen für soziale Stände, Schichten und Gruppen in den Leges
Barbarorum, Arbeiten zur Frühmittelalterforschung 11 (Münster 1991).
65
M. Weidemann, “Adel im Merowingerreich. Untersuchung zu seiner Rechtsstel-
lung”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 40 (1993) pp. 535–55.
See also: K.F. Werner, Naissance de la Noblesse (Paris 1998).
288  

has recently stated the following on the topic of the nobility as a

juridically-defined class in the Merovingian period: “The Frankish
nobility rather developed—as one may sum it up—on the basis of
royal privileges, in which single persons were given special privileges
in both public and private law, to which the free class had no
birth-right.”66
From the archaeological evidence, it is impossible to contribute to
the debate on the legal status of the nobility.67 It is, however, on
basis of the burial types and the grave-goods, possible archaeologi-
cally to recognise a stratum of graves that may be associated with
the nobilitas referred to in the written sources.68 This has been demon-
strated most impressively by Hayo Vierck69 giving the example of
funerary material from one grave at Lehmbeck in Westphalia, and
by Anke Burzler70 on basis of the material from Alamannia and
Bavaria. New finds from Westphalia and Lower Saxony now, for
the first time, provide evidence for an “early nobility” in the area
from which the Franks originated.

66
Weidemann, “Adel im Merowingerreich”, p. 554.
67
There is, of course, much to suggest that burial within a church results from
a privileged juridical position (as also H.W. Böhme states: “There is no need for
further proof for the fact that the burial of influential, prominent and rich ladies
and gentlemen in their own, separate necropolises, respectively in or near churches
[. . .] could only take place due to a special, privileged juridical position” [Böhme,
“Adelsgräber im Frankenreich”, p. 525]), though this cannot be shown archaeo-
logically.
68
The basis for such an interpretation have been established first and foremost
by: R. Christlein, “Besitzabstufungen zur Merowingerzeit im Spiegel reicher Grabfunde
aus West- und Süddeutschland”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz
20 (1973) pp. 147–80; H. Ament, Fränkische Adelsgräber von Flonheim in Rheinhessen,
Germanische Denkmäler der Völkerwanderungszeit B 5 (Berlin 1970) pp. 130–63.
On the basis of this research, H.W. Böhme rigorously uses the concept of nobil-
ity: Böhme, “Adelsgräber im Frankenreich”.
69
H. Vierck, “Ein westfälisches ‘Adelsgrab’ des 8. Jahrhunderts n. Chr. Zum
archäologischen Nachweis der frühkarolingischen und altsächsischen Oberschichten”,
Studien zur Sachsenforschung 2, ed. H.-J. Häßler (Hildesheim 1980) pp. 457–88. The
author finds it remarkable how little the ideas, and the promising methodological
hypotheses, developed within early medieval archaeological research have achieved
or been adopted.
70
A. Burzler, Archäologische Beiträge zum Nobilifizierungsprozeß in der jüngeren Merowingerzeit
(unpublished Ph.D. München 1991).
   () 289

Early nobility? On the sociological structure in the Dutch-Westphalian-Lower

Saxon region during the fourth and fifth centuries

The finds recovered from funerary contexts, discussed in the fol-

lowing, clearly show how far the cultural-historical eloquence of
archaeological find(ing)s and their valuation depend on the circum-
stances of their recovery. In 1994, a tomb, cut into the clay below
the groundwater-level, was discovered near the Lower Saxon North
Sea coast on the Fallward Wremen in the country of Wursten. Due
to these fortunate circumstances, all the wooden elements of this bur-
ial had been preserved. The burial pit was covered with round poles
and planks. Sixty oak boards and “a roof-like construction of rounded
and squared wooden elements”71 together formed a boat-shaped struc-
ture. At its western extremity, shaped like a ship’s stern, was placed
a “65 cm high, throne-like seat made of a hollowed-out alder tree-
trunk”.72 To the richly carved throne belongs a footstool, decorated
with a depiction of a deer-hunt, the function of which is indicated
by a runic inscription reading skamella (footstool). A wooden vessel
with a bird-shaped lid is a unique piece, though its function and
meaning are unclear. Further, a bowl made of mountain maple was
found, measuring 60 cm in diameter and decorated with carvings.
Other finds include a small wooden table extravagantly constructed
and with turned feet, as well as clay vessels, receptacles in wood
(among others, a small cabinet) and metal (a late Roman kettle) as
well as a well-preserved late antique belt-fitting. Under less favourable
circumstances, little more than the unspectacular clay vessels, the
late Roman kettle and the belt-fitting would have been preserved
from this assembly.
One may ask how this burial would have been interpreted, had
it not been preserved in such excellent circumstances. It seems clear
that the grave would have been labelled as “soldier/foederatus”, a
type on which H.W. Böhme has worked.73 This label indicates war-
riors of barbarian provenance, who for a certain period would have

71
M.D. Schön, Der Thron aus der Marsch. Ausgrabungen an der Fallward bei Wremen im
Landkreis Cuxhaven, Begleitheft zu Ausstellungen/Museum Burg Bederkesa, Landkreis
Cuxhaven 1 (Bremerhaven 1995), p. 19.
72
Ibid., p. 20.
73
Böhme, Germanische Grabfunde; id., “Söldner und Siedler”; id., “Franken und
Romanen”; id., “Sächsische Söldner”.
290  

done military duties in the Roman Empire before returning to their

place of origin with Roman dress-elements, badges of rank and other
objects.74 In 1982, H. Steuer described this type of burials as fol-
lows: “Considered together, these graves of the foederati-civilisation
are not especially rich when one estimates its absolute value, for gold
is lacking [. . .]. The absence of gold among the grave-goods hardly
make it possible to categorise even the richer of these foederati buri-
als as belonging to ‘the highest, socially and politically leading part
of the population.’ Had it been members of the peak of society
buried in these graves, then one would at least have found some
golden Zwiebelknopffibeln [cruciform brooch with onion-shaped termi-
nals]! Moreover, finds of contemporary treasures are known that are
substantially more valuable and do contain gold.”75 The complex
found at Wremen fundamentally changes this image. In particular,
it has become clear that an elevated social position is not solely man-
ifested by the presence of gold objects. In the light of the results of
new research, Steuer’s statement must therefore be revised. The foot-
stool (suppedaneum) combined with the throne,76 as well as the con-
struction of the grave and its rich fittings with the beautifully worked
wooden receptacles, leave no doubt that it was a person of high
social standing who was buried on the Fallward, and whose chosen
form of representation also included the late Roman military belt
(cingulum militiae). It is of particular significance that the cemetery at
Wremen is not far from the terp Feddersen Wierde.77 The recog-

74
See, for instance, the find at Lengerich: M. Schmauder, “Der Verwahrfund
von Lengerich, Ldkr. Emsland: Spiegel innerrömischer Kämpfe?”, Die Kunde 50
(1999) pp. 91–118.
75
“Insgesamt betrachtet sind diese Gräber der Foederaten-Zivilisation nicht beson-
ders reich, wenn man den absoluten Wert abschätzt, denn Gold fehlt [. . .]. Der
Mangel an Gold unter den Beigaben lässt es kaum zu, daß man bei den auch
reicheren Bestattungen dieses Foederaten-Horizontes von der ‘obersten, sozial und
politisch führenden Bevölkerungsschicht’ sprechen kann. Wären in den Gräbern die
Spitzen der Gesellschaft bestattet, dann müßte man zumindest einige goldene Zwie-
belknopffibeln gefunden haben! Zumal gleichzeitig Schatzfunde bekannt sind, die
wesentlich wertvoller sind und Gold enthalten”: H. Steuer, Frühgeschichtliche Sozialstruk-
turen in Mitteleuropa, Abhandlungen der Akademie der Wissenschaften Göttingen,
philologisch-historische Klasse 3,128 (Göttingen 1982) p. 346.
76
For such thrones in the form of amulets, see: H. Drescher and K. Hauck,
“Götterthrone des heidnischen Nordens”, Frühmittelalterliche Studien 16 (1982) pp.
237–301.
77
See, with older literature: P. Schmid, “Feddersen Wierde”, Reallexikon der Germa-
nischen Altertumskunde 8 (2nd edn., 1994) pp. 249–66.
   () 291

nisable development of a “lordly court” at Feddersen Wierde dur-

ing the third century, and its existence with continuous habitation
until the middle of the fifth century, ought henceforth to be inter-
preted against the backdrop of similar burials, like the one on the
Fallward. That is why Karl-Heinz Willroth sees the Fallward burial
as “a key to the understanding of the settlement-structure on the
Feddersen Wierde”.78
Decorative metalwork of a display shield found in the settlement
of Dortmund-Oespel79 also point to the existence of complex social
differentiation. The nearest parallel for this piece has been found in
the sacrificial complex at Illerup in Denmark.80 But also a burial like
the one in Vermand (dép. Aisne), discovered as early as 1885 but
still not completely published, indicates—according to the latest ideas
presented by C. von Carnap-Bornheim—how far social differentiation
must have progressed outside the borders of the Roman Empire in
order to enable burials of this type in northern Gaul.81
It is therefore no exaggeration to suppose an elaborate social
differentiation also for the area in which the “early Franks” are to
be located, as is generally acknowledged in the contemporary sources.
This assessment is particularly important in view of the established
opinion in current scholarship, stating that the process of social
differentiation only started on a large scale after the so-called “set-
tlement-period”. As demonstrated, individual aspects show how the
origins of the nobilitas, who is so evident in the fourth and the first

78
K.-H. Willroth, “Siedlungen und Gräber als Spiegel der Stammesbildung.
Gedanken zur Abgrenzung germanischer Stämme in der ausgehenden vorrömischen
Eisenzeit in Norddeutschland und Südskandinavien”, Studien zur Archäologie des Ostse-
eraumes, Festschrift für Michael Müller-Wille (Neumünster 1999) pp. 359–71, here p. 359.
See the comparable case of Stepping Mølle, Haderslev Amt: P. Ethelberg, “The
Chieftain’s Farms of the Over Jerstal Group”, Journal of Danish Archaeology 11 (1992/
93) pp. 111–35. The special significance that the artisans of Feddersen Wierde seem
to deserve in the context of the development of a ruling class suggests a differ-
entiated social structure also in the case of a comparable site defined by craftsmen,
like the settlement of Heeten (Verlinde and Erdrich, “Germanische Siedlung”, esp.
p. 718).
79
H. Brink-Kloke, A. von Bohlen, M. Doll, E. Lietz and C. Poniecki, “Ein
(kleines) germanisches Dorf—Die jüngerkaiserzeitliche Siedlung von Dortmund-
Oespel”, Millionen Jahre Geschichte. Fundort Nordrhein-Westfalen, ed. H.G. Horn, H. Hellen-
kemper, G. Isenberg and H. Koschik (Mainz 2000) pp. 343–8, esp. pp. 343–4.
80
C. von Carnap-Bornheim and J. Ilkjær, Illerup Ådal 5: Die Prachtausrüstungen,
Jutland Archaeological Society Publications 25,5 (Aarhus 1996) pp. 279–89.
81
von Carnap-Bornheim, “Kaiserzeitliche germanische Traditionen”.
292  

half of the fifth centuries through the richly equipped foederati-type

graves, may date back into the third century. The developments of
the fourth and the first half of the fifth centuries lead, from the sec-
ond half of the fifth century onwards, to a type of burials charac-
terised by Goldgriffspatha82 (here, a continuation of a specific group of
people is not intended). With the emergence of separated burials in
private churches (Eigenkirchen) and/or arranged in rows in cemeter-
ies this particular social stratum—undoubtedly differentiated in time
and region—becomes discernible in the whole of the Frankish realm.

The problem of ethnic interpretation

The problem regarding the evaluation of archaeological finds in rela-

tion to a possible ethnic labelling has been a remarkably controver-
sial discussion for some time. Most recently, especially Sebastian
Brather83 and Frank Siegmund84 have been working on this ques-
tion from an archaeological point of view.85
Siegmund states succinctly “that each science tends to follow a
specific understanding of ethnos”,86 from which it follows that “the
criteria for ethnos-research [should] first and foremost [be] developed
and interpreted within the discipline itself ”.87 The present author agrees
that this is a fundamental prerequisite for working on the question
of whether or not archaeological finds can express ethnicity.
Different from Siegmund, the research and interpretations of
Brather, which will be the subject in the following, are founded on
results in the conceptual development of ethnology. According to
Brather, it is particularly difficult, if not completely impossible, to
formulate ethnic identity on the basis of archaeological material.
Consequently, he states that: “Since ‘ethnicity’ on the basis of its

82
D. Quast, Die merowingerzeitlichen Grabfunde aus Gültlingen (Stadt Wildberg, Kreis
Calw), Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg
52 (Stuttgart 1993) pp. 20–6; Böhme, “Frankenkönig Childerich”; W. Menghin,
“Schwerter des Goldgriffspathahorizonts im Museum für Vor- und Frühgeschichte,
Berlin”, Acta Praehistorica et Archaeologica 26/27 (1994/95) pp. 140–91.
83
Brather, “Ethnische Identitäten”.
84
Siegmund, Alemannen und Franken.
85
Further discussion of these facts is unfortunately impossible here.
86
Siegmund, Alemannen und Franken, p. 74.
87
Ibid.
   () 293

flexible, group-specific character can—or, by the same token, can-

not—be linked to more or less every object, it poses substantial and
long unresolved methodological problems for its vague identification
in archaeological material”.88 Two examples chosen by Brather ought
to be repeated here, in order to show the points on which the pre-
sent author wishes to present a methodological critique of Brather’s
line of argumentation and instead favours a different interpretation
of the archaeological finds.
The disappearance of rich graves in south-western Germany (the
Alamannic area) at the beginning of the sixth century, can, on basis
of the archaeological evidence, be understood as part of a break that
can be traced also in other spheres.89 In this period, especially some
types of fibulas, which were typical of rich female graves in this
region, appear also in neighbouring regions and in areas further
away. Archaeologically, these finds can only be explained by the
“moving out” of the upper class from south-western Germany, for
there is no evidence for a “transfer” of the Alamannic settlement
area as a whole.90 The historical events that lead to this “moving
out” cannot be determined on the basis of the archaeological finds.
Only after an additional analysis of the historical sources can it
be decided if there are parallels between archaeological findings
and historical traditions. The methodology chosen by Brather, of

88
Brather, “Ethnische Identitäten”, p. 175.
89
The destruction of the settlement on the “Runde Berg”, which possessed the
function of a central place in this period (cf. in general with further literature:
H. Bernhard et al., Der Runde Berg bei Urach, Führer zu archäologischen Denkmälern
in Baden-Württemberg 14 [Stuttgart 1991]), reveals that the disappearance of richly
furnished graves cannot have been solely the result of an alteration in burial cus-
toms, but was rather related to decisive repercussions caused by a change in the
settlement structure as a whole. See the composition of the Alamannic cemeteries
that cease to exist around 500, in Ursula Koch, “Der Beginn fränkischer Besied-
lung im Rhein-Neckar-Raum. Gräber von Mannheim-Vogelstang und Mannheim-
Straßenheim”, Mannheimer Geschichtsblätter NF 7 (2000) pp. 57–106, esp. p. 57 n. 3.
90
Neither were Alamannic type cemeteries laid out in these regions, nor do we
find evidence for the construction of the hill-top settlements that are so character-
istic of the Alamannic settlement region. On the latter, see: H. Steuer, “Höhensied-
lungen des 4. und 5. Jahrhunderts in Südwestdeutschland”, Freiburger Forschungen zum
ersten Jahrtausend in Südwestdeutschland 1 (Sigmaringen 1990) pp. 139–205; id., “Herr-
schaft von der Höhe. Vom mobilen Söldnertrupp zur Residenz auf repräsentativen
Bergkuppen”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart 1997)
pp. 149–62. On the other hand, a large number of Alamannic cemeteries was
abandoned in the early sixth century after a brief period of use (H. Ament, Das
alamannische Gräberfeld von Eschborn [Main-Taunus-Kreis], Materialien zur Vor- und Früh-
geschichte in Hessen 14 [Wiesbaden 1992] p. 50).
294  

interpreting complex archaeological data on the basis of a historical

source, is highly questionable from this point of view. For Brather
formulates the following thesis: “The inclusion of the Alamannic
region into the Frankish realm did furthermore not lead to an exo-
dus of the entire old ruling class. Rather, it enabled a not insub-
stantial part of the (loyal) Alamannic elite to reach important positions.
Agathias, for instance, writes about the brothers Leuthari and Butilin
[. . .] who—Alamannic by birth, but with great influence among the
Franks—served as Frankish army leaders”.91 The example from
Agathias given by Brather is problematic for several reasons.92 Firstly,
it cannot explain the archaeological fact of the absence of rich graves
in south-western (Alamannic) Germany; secondly, facts concerning
the behaviour of two individual members of the Alamannic ruling
class is unacceptably generalised here.93 Moreover, the instance quoted
from Agathias relates to the period of the rule of the Frankish King
Theudebert (534–547), and hence to an incident dating from the
beginning of the second third of the sixth century at the earliest.
Brather builds his argumentation specifically on this—for Agathias
clearly exceptional—case,94 and develops the idea that the main pur-
pose of the Alamannic ruling class was to “be able to preserve, or

91
“Die Einbeziehung des alemannischen Raums in das Frankenreich hat bei
weitem nicht zum Exodus der gesamten alten Führungsschicht geführt, sondern im
Gegenteil einem nicht unbeträchtlichen Teil der (loyalen) alemannischen Elite weiter-
hin wichtigste Positionen ermöglicht. So berichtet beispielsweise Agathias von den
Brüdern Leuthari und Butilin [. . .], die—Alemannen der Abstammung nach, aber
mit großem Einfluß bei den Franken—wohl als fränkische Heerführer dienten.”:
Brather, “Ethnische Identitäten”, p. 167.
92
Agathias Myrinaei, Historiarum libri quinque A (1) 6,2, transl. C. Dirlmeier and
G. Gottlieb, Quellen zur Geschichte der Alamannen 2, Schriften der Heidelberger Akademie
der Wissenschaften, Kommission für Alamannische Altertumskunde 3 (Heidelberg-
Sigmaringen 1978) p. 80.
93
Still in discussion is whether Leuthari and Butilin were Alamannic dukes by
descent, or if they achieved this position only after their service as Frankish army
leaders (D. Geuenich, “Die Alamannen unter fränkischer Herrschaft”, Die Alamannen.
Ausstellungskatalaog Stuttgart, Zürich, Augsburg [Stuttgart 1997] pp. 204–8, esp. p. 204
n. 2).
94
It ought, here, to be emphasised, that further written evidence for a consid-
erable flight of Alamans from the Alamannic settlement region does exist. Making
a mention of this would have ensured a more objective valuation of the historical
sources. (Cassiodorus, Variae 2,41, ed. T. Mommsen, MGH AA 12 [Berlin 1894]
p. 73; Ennodius, Panegyricus Theoderico 15, ed. F. Vogel, MGH AA 7 [Berlin 1885]
p. 212.) See also the reference in Agathias (Agathias Myrinaei, Historiarum libri quinque
A [1] 6,4), that the Alamans were dependent on Theoderic the Great, and had to
pay him tribute.
   () 295

win back, their hitherto existing positions of power, also under the
changing power-relations.”95 In other words, it was processes respec-
tively of social levelling and assimilation that, according to Brather,
determined the relationship between societies of the Migration Period
and of the early Middle Ages. It was, according to Brather, these
processes that led to an exchange between social elites, which is ren-
dered most visible in the female dress jewellery. However, processes
of social levelling hardly explain either the absence of rich graves in
the Alamannic settlement region, or the fact that members of the
Frankish ruling class clearly very consciously settled near existing
Alamannic settlements and buried their dead in Alamannic ceme-
teries. These new settlers are not only distinguished by their different
dress, but—and this is of special importance—especially by their lav-
ish burial-customs.96 It is not “social continuities”97 that are visible
here but a cultural break, for which the departure of the Alamannic
ruling class is but one aspect. The archaeological situation in south-
western Germany in the period after 500 can also usefully be com-
pared with the historical sources (as has been done by Ursula Koch
and others),98 linking the absence of rich graves in the Alamannic

95
Brather, “Ethnische Identitäten”, p. 167.
96
U. Koch, “Ethnische Vielfalt im Südwesten. Beobachtungen in merowinger-
zeitlichen Gräberfeldern an Neckar und Donau”, Die Alamannen. Ausstellungskatalaog
Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp. 219–32, esp. pp. 224–5. There is much
evidence for the permanent presence of Frankish groups in Alamannia in the early
sixth century. Cf. for instance: M. Martin, Das fränkische Gräberfeld von Basel-Bernering
(Basel 1976), esp. pp. 146–55; G. Fingerlin, “Franken am Kaiserstuhl”, Archäologische
Nachrichten aus Baden 44 (1990) pp. 7–15; id., “Ein früher Stützpunkt fränkischer
Herrschaft am Oberrhein. Neue merowingerzeitliche Grabfunde aus Bad Krotzingen,
Kreis Breisgau-Hochschwarzwald”, Archäologische Ausgrabungen in Baden-Württemberg
(1998) pp. 200–3; E. Pohl, “Der Neuburger Stadtberg und sein Umfeld am Über-
gang von der Antike zum Mittelalter (4. bis 7. Jhdt.) aus archäologischer Sicht”,
Neuburg an der Donau. Archäologie rund um den Stadtberg, ed. K.H. Rieder and A. Tillmann
(Neuburg 1993) pp. 109–32, esp. pp. 124–5; M. Trier, “Die frühmittelalterliche
Besiedlung des Lechtals im Spiegel der archäologischen Überlieferung”, Der Landkreis
Augsburg, vol. 2: Vor- und Frühgeschichte, Archäologie einer Landschaft, ed. W. Pötzl and
O. Schneider (Augsburg 1996) pp. 267–322, esp. pp. 279–84; B. Grosskopf-Theune,
“Die Kontrolle der Verkehrswege. Ein Schlüssel zur fränkischen Herrschaftssicherung”,
Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp. 237–42.
97
Brather, “Ethnische Identitäten”, p. 167.
98
U. Koch, “Fernbeziehungen im Spiegel merowingerzeitlicher Grabfunde—Wer
waren die Kontaktpersonen?”, Archäologisches Nachrichtenblatt 3 (1998) pp. 107–17, esp.
pp. 110–1; ead., “Besiegt, beraubt, vertrieben. Die Folgen der Niederlagen von
496/497 und 506”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart
1997) pp. 191–201; ead., “Ethnische Vielfalt im Südwesten”.
296  

lands and the archaeologically proven “transfer” of the Alamannic

ruling class.99
Undoubtedly, there would have been isolated cases of members
of the Alamannic ruling class trying to secure their social status by
integrating with the Frankish structure of rulership. However, no
cases of this type can be evidenced archaeologically by reference to
a demonstrable, unbroken tradition of Alamannic high-status graves
within Alamannia itself.100 Though, even if (as Brather proposes) the
archaeological material may be regarded as evidence for the survival
of social continuity/identities—which, as argued above, is highly
doubtful—it does not signify that the archaeological material auto-
matically loses its ethnic value. After all, the conservation of ethnic
independence under foreign political rulership was not only possible,
but to a certain extent also common.101
The second example, which will be discussed here, belongs in the
Frankish realm. Brather discusses a particular section from Alexander
Koch’s research on the “Merovingian Bügelfibeln in the western
Frankish realm”. The citation from Koch’s work at the beginning
of Brather’s discussion essentially accords with the view of the pre-
sent author, as presented above: a radical change of dress only
occurred in extreme situations during the Migration Period and the
early Middle Ages. The following quote from Koch in Brather’s
work, however, falsifies Koch’s original statement. The half sentence
quoted by Brather is in Koch’s work preceded by the following: “In
terms of material culture, ethnic identity can [in general only be
outwardly articulated by means of dress]”.102 Since Koch’s intention
is to restrict himself only to a valuation of objects, as stated in the

99
The near complete absence of burials of the Alamannic elite in the former
Alamannic settlement area does not make sense if a large segment of this elite
proved themselves loyal to the Franks in their role as new rulers.
100
It does not do justice to neither the archaeological finds nor the historical
sources to see Alamannic elite loyalty to the Franks as the rule rather than the
exception (Brather, “Ethnische Identitäten”, p. 167).
101
Slavic groups, for instance, that had settled in the region of the Altmark, led a
culturally independent existence also under Carolingian dominance. See M. Schmauder,
“Überlegungen zur östlichen Grenze des karolingischen Reiches unter Karl dem
Großen”, Grenze und Differenz im frühen Mittelalter, ed. W. Pohl and H. Reimitz, For-
schungen zur Geschichte des Mittelalters 1 (Wien 2000) pp. 79–81.
102
“Im dinglichen Bereich kann ethnische Identität praktisch nur durch Tracht nach
außen artikuliert werden”: Koch, Bügelfibeln der Merowingerzeit, p. 535; the italics indicate
the part cited by Brather.
   () 297

first half of his sentence, he is in no way lapsing into the “national-

romantic notions of the nineteenth century”103 that Brather accuses
him of. Koch’s insights are therefore not “from a current point of
view, flawed from the outset”,104 as Brather asserts.105 Further on,
Brather refers to the spread of the “proto- or early Merovingian
fibula shapes of Groß-Umstadt and Nieder-Florstadt”.106 In the dis-
tribution of these types of fibulas, Brather does not see evidence for
the mobility of individuals or groups of people belonging to a speci-
fic ethnos, but, here too, interprets the distribution of fibulas as a
confirmation of “processes of social levelling”.107 However, in the eyes
of the present author, the association of these types of fibulas with
specifically Bohemian ceramics indicates that these fibulas are to be
connected with a particular ethnic group. Further, this group—and
this seems essential—is also clearly recognisable from its particular
funerary practices.108
At this point, it ought to be emphasised that, in the opinion of
the present author, maps showing the spread of individual types of
finds are generally only of a very limited use in an ethnic interpre-
tation of archaeological finds. This is, in this connection, undoubt-
edly due to a fundamental methodological problem.109 However, it

103
Brather, “Ethnische Identitäten”, p. 168.
104
Ibid.
105
That Koch is aware of this problem is evident also in other instances. He,
for instance, differentiates and speaks of a “specific ethnic character as on a par
with dress” (Koch, Bügelfibeln der Merowingerzeit, p. 562).
106
Ibid., p. 19. For these fibula-types, see also: H.W. Böhme, “Eine elbgerman-
ische Bügelfibel des 5. Jahrhunderts aus Limetz-Villez (Yvelines, Frankreich)”, Archäoe-
logisches Korrespondenzblatt 19 (1989) pp. 397–406; D. Quast, “Vom Einzelgrab zum
Friedhof. Beginn der Reihengräbersitte im 5. Jahrhundert”, Die Alamannen. Ausstel-
lungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp. 171–90, esp. fig. 176.
107
Brather, “Ethnische Identitäten”, p. 168.
108
With the aid of specific pottery types, but especially on the basis of particu-
lar forms in the cult of the dead, H. Ament has succeeded in finding evidence for
an immigration from the Bohemian Danube-region within the cemetery at Eschborn
(Main-Taunus-Kreis); (H. Ament, Eschborn, Main-Taunus-Kreis, Grabfunde des 5. Jahrhunderts:
ein alamannisches Gräberfeld an der Wende vom Altertum zum Mittelalter, Archäologische
Denkmäler in Hessen 41 [Wiesbaden 1984] p. 50). See further: B. Svoboda, “Zum
Verhältnis frühgeschichtlicher Funde des 4. und 5. Jahrhunderts aus Bayern und
Böhmen”, Bayerische Vorgeschichtsblätter 28 (1963) pp. 97–116; T. Springer, “Germanen-
funde der Völkerwanderungszeit in Nordbayern. Bemerkungen zur Keramik vom
Typ Friedenhain-PÓre“t’ovice”, Archäologisches Korrespondenzblatt 15 (1985) pp. 235–43.
109
Specific spreads in distribution on distribution-maps may be caused by sev-
eral different reasons. In the previous, mention has already been made of the
298  

also applies to a sociological interpretation of these maps, in the

manner presented by Brather as if virtually an alternative to an eth-
nic interpretation. In order to create a foundation for an ethnical
interpretation of archaeological finds, one must, in the view of the
present author, necessarily consider the context in which the finds
have been recovered. It is only on the basis of a (possibly) complex
gathering of the archaeological material that the determining crite-
ria necessary for establishing the presence of an ethnos may be possible.
Siegmund has subsumed these criteria under the heading “customs
and habits”.110
Two aspects ought, once again, to be emphasised: firstly, partic-
ularly when dealing with an ethnic interpretation, it is necessary to
strictly separate archaeological and historical sources (that also applies
to the use of particular terms referred to in the sources in relation
to archaeological finds).111 Secondly, any statement regarding ethnos
is only valid when a wide spectrum of criteria is employed. In order
to illustrate the latter point, two examples from the Frankish realm
are given in order to demonstrate how difficult, and by no means
always successful, an ethnic interpretation of archaeological finds
can be.112

simple late antique belt fitting of the type Halliot, the distribution of which is pre-
sumably connected to the market of a place of production, of the type of work-
shop environment that can often be found in the early Middle Ages; see for instance:
K. Banghard, “Eine frühmittelalterliche Gürtelgarnitur und ihre Motivgeschichte”,
Archäologische Nachrichten aus Baden 59 (1998) pp. 24–35. Further explanations may
be cited, which as a whole will demonstrate that the distribution-maps are certainly
not primarily concerned with ethnic significance.
110
Siegmund, Alemannen und Franken, p. 85.
111
A good example of this is the use of the word “francisca”. For this see: W.
Pohl, “Telling the difference: Signs of ethnic identity”, Strategies of Distinction. The
Construction of Ethnic Communities, 300 –800, ed. id. with H. Reimitz, The Trans-
formation of the Roman World 2 (Leiden-Boston-Köln 1998) pp. 17–69, esp. pp.
32–7 (it should be emphasised that the archaeological results, as far as the fre-
quency of axes in respectively Alamannic and Frankish graves is concerned, is even
more straightforward than Pohl indicates. According to Siegmund [F. Siegmund,
“Kleidung und Bewaffnung der Männer im östlichen Frankenreich”, Die Franken—
Wegbereiter Europas 1, pp. 691–706, esp. p. 705] the proportion in the sixth century
is 26% against 6%, and not “half the percentage in Alamannic graves” [Pohl,
“Telling the difference”, p. 33]).
112
For examples from outside the Frankish sphere of influence, on the basis of
which ethnical processes can be shown, see the Bavarian cemetery at Altenerding.
(V. Bierbrauer, “Das Reihengräberfeld von Altenerding in Oberbayern und die
bajuwarische Ethnogenese—Eine Problemskizze”, Zeitschrift für Archäologie des Mittelalters
13 [1985] pp. 7–25).
   () 299

The importance of an ethnic interpretation of the archaeological

material is, according to the present author, essential for a cultural
analysis. Even the frequent failure to answer these questions, as we
have seen for the early Frankish period, presents important obser-
vations, which are of great potential as cultural evidence.

Valuation and validity of the criteria.

An example: the cemetery at Zeuzleben113

The cemetery at Zeuzleben, situated in present-day Franken, repre-

sents a concise example of the difficulties that may occur when attach-
ing ethnic labels to cemeteries. This cemetery came into use around
520–530, and was abandoned around 600. Particularly remarkable
was the large and varied number of animal burials (mostly horses,
but also dogs). Further, it was noteworthy that the male and female
graves were organised in different manners within the family-group
burials, and that the characteristic grave-types were of the so-called
Pfostenkammer form (chamber grave). A. Rettner has rightly interpreted
these finds in relation to the Thuringian area, where clear parallels
can be found.
The study of the grave-goods and dress-elements, however, brought
out completely different connections. Western influences on dress
dominate both in the male and the female graves from the time of
the earliest interments, including forms that are generally only found
west of the Rhine. This finding was confirmed in the study of the
ceramic grave-goods: a handled pot of the Mayener form Alzey 27,
coarseware jars, flat dishes, but also buckets with handle attachment,
bronze wash-basins and especially the lidded cups, clearly indicate
the lands west of the Rhine. Middle German influences can be recog-
nised only in the designs of the hand-made pottery belonging to the
first period of use (520/30–550/60). During the second period
(550/60–570/80), these have disappeared in favour of copies of
western models. Still, two categories of grave-goods remained rela-
tively constant at least during the first period. Firstly, the textiles,
tools and amulets found in female graves; that are, objects not only

113
A. Rettner, “Thüringisches und Fränkisches in Zeuzleben”, Archäologie des Fran-
kenreiches, pp. 113–25.
300  

of a functional character but also belonging to the realm of religion.

Secondly, the male weapons; though during the last phase of the
cemetery (570/80–c. 600) the sax appears, which at this time was
widespread throughout the west Frankish region.114 Interestingly, the
position of the deposition of riding equipment in the male graves
remained similarly largely untouched by western influences. As opposed
to the custom west of the Rhine, where riding equipment as a rule
was deposited near the feet, at Zeuzleben it was placed near the
head of the deceased.
Summarising, Rettner writes: “It is amazing that changes happen
most quickly in dress and the vessel types used as grave-goods, that
is, in spheres where one may expect the most reliable information
on the people. With some little delay weaponry changes too. In the
course of this process, shapes are changed easier than production
methods—if we think back to the ceramics for a moment—and faster
than customs of deposition. The tradition of construction, the post-
construction method, was maintained unaltered the longest, as was
the custom of funerary meals with rich gifts of food, and in partic-
ular the old manner of burials in groups with divisions by gender”.115

Franks and Romans? The cemetery at Dieue-sur-Meuse

The cemetery at Dieue-sur-Meuse, in western Lotharingia, has over

the past years been an interesting starting-point for discussions regard-
ing the question of the use of a burial-ground by both Franks and
Romans.116

114
These finds clearly show that the question of armament is obviously not to
be solved solely from a pragmatic point of view, in the sense of a change in fighting
technique, as Brather has postulated (Brather, “Ethnische Identitäten”, p. 170).
115
“Erstaunlich dabei ist, daß sich der Wandel am schnellsten innerhalb der
Tracht und bei der Gefäßbeigabe vollzieht, also in Bereichen, wo man eigentlich
die zuverlässigsten Aufschlüsse zum Volkstum erwartet. Mit etwas Verzögerung
ändert sich auch die Bewaffnung. Formen werden im Laufe dieser Prozesse schneller
aufgegeben als Herstellungsweisen—denken wir an die Keramik zurück—, auch
schneller als Gewohnheiten und Deponierung. Am längsten hielt man an einer
Bautradition fest, der Pfostenbauweise, ferner am Totenmahl mit reichen Speisebei-
gaben und vor allem an den alten Gruppenarealen mit der Geschlechtertrennung”:
Rettner, “Thüringisches und Fränkisches in Zeuzleben”, p. 124.
116
J. Guillaume, “Les nécropoles mérovingiennes de Dieue/Meuse (France)”, Acta
Praehistorica et Archaeologica 5/6 (1976/77) pp. 301–9.
   () 301

On the basis of an analysis of the cemetery, Hermann Ament in

1976 reached the conclusion that both groups were represented
there.117 Central to his argument were the characteristics of dress,
as well as the forms and manner of burial. According to Ament,
characteristic Merovingian belt-forms appear on the northern and
eastern sides of the cemetery, whereas these forms have not been
found in the western part. The graves found in the latter area are,
additionally, distinguished from the others by their lesser size, the
higher frequency of so-called Plattengräber and especially by their pro-
nounced poverty of grave-goods.
Ament sums up the archaeological results as follows: “There is no
way around the ethnic interpretation of this situation. In the graves
of the western group, we find the burials of the descendants of those
provincials who, as early as the fifth century, were buried in the south-
ern part of the cemetery, in other words, Romanic burials of the
Merovingian period. In the graves organised and equipped as “Reihen”-
graves, on the other hand, the remains of newly arrived Germanic,
in this case Frankish, elements of the population may be identified”.118
Renewed work on the cemetery at Dieue-sur-Meuse and the inter-
pretation proposed for it by Ament has most recently been carried

117
H. Ament, “Das Gräberfeld von Dieue-sur-Meuse, ein Bestattungsplatz von
Franken und Romanen”, Acta Praehistorica et Archaeologica 7–8 (1976/77) pp. 301–9.
For the question of the archaeological evidence for Romance speakers in the Frankish
realm, see: K. Böhner, “Romanen und Franken im Trierer Land nach dem Zeugnis
der archäologischen Quellen”, Siedlung, Sprache und Bevölkerungsstruktur im Frankenreich,
ed. F. Petri, Wege der Forschungen 49 (Darmstadt 1973) pp. 346–82; H. Ament,
“Franken und Romanen im Merowingerreich als archäolgisches Forschungsproblem”,
Bonner Jahrbücher 178 (1978) pp. 377–94; id., “Romanen an Rhein und Mosel im
frühen Mittelalter”, Bonner Jahrbücher 192 (1992) pp. 261–71; V. Bierbrauer, “Romanen
im fränkischen Siedelgebiet”, Die Franken—Wegbereiter Europas 1, pp. 110–20. A. Rettner
has recently made an in-depth study of the continuity of Romanic groups in Raetia
after the fifth century. His observations, which are still to appear in print, are based
on a combination of various archaeological observations on the basis of which he
successfully provides proof for a “remnant Romanitas”; (most recently as a lecture
“Romanen in Südbayern” during the yearly meeting of the West- und Süddeutscher
Verband für Altertumsforschung e. V. in Trier [ June 2001]).
118
“An der ethnischen Interpretation dieses Sachverhaltes führt kein Weg vor-
bei. In den Gräbern der Westgruppe fassen wir die Bestattungen der Nachfahren
jener Provinzialen, die sich bereits im 5. Jahrhundert im Süden des Friedhofsareals
beisetzen ließen, fassen wir also romanische Bestattungen der Merowingerzeit.
Hingegen wird man in den nach Reihengräberart angelegten und ausgestatteten
Gräbern die Hinterlassenschaft eines neu hinzugekommen germanischen, hier also
wohl fränkischen Bevölkerungselementes erblicken dürfen”: Ament, “Das Gräberfeld
von Dieue-sur-Meuse”, p. 305.
302  

out by Uwe Fiedler.119 On the basis of a mapping of the various

male dress belt-types, a continuity of use of the cemetery by Franks
can be observed from the south towards the north-east. A similar
image emerges in the case of the spread of various types of weapons.
Lance-points and arrow-heads, as well as franciscas and axes are
predominantly found towards the south, whereas saxes appear mainly
towards the north, but also in the western part of the cemetery. The
latter are obviously part of a newer type of military equipment.
According to Fiedler, a mapping of Bügelfibeln and smaller fibulas also
shows that the oldest Merovingian period graves adjoin the late
Roman area in the south. The mapping of various types of ceram-
ics, among which the Knickwandtöpfe play an important part, shows
a similar development. Also in this case it is clear that the late (high)
forms of Knickwandtöpfe appear towards the north and the (north-)
west. The occurrence of graves containing stone structures or mul-
tiple burials also correspond to this pattern. Both phenomena like-
wise occur towards the north and north-west. It is particularly striking
that the wearing of pendants made out of deer-antler discs, bone
beads, vertebra and animal teeth, so typical of Germanic female
dress, are also found in the western group that Ament classifies as
Roman.
Fiedler’s observations indicate, contrary to Ament’s ideas, that the
cemetery at Dieue-sur-Meuse should be regarded as a Frankish bur-
ial-ground. The character of the cemetery is distinguished by a pro-
gressive reduction of grave-goods, accompanied by a simultaneous
adoption of Romanic burial customs. The analysis clearly demon-
strates how exceptionally difficult an ethnic interpretation can be—
not just for a single find, but also for a complete cemetery. There
is no doubt that the last word on the cemetery at Dieue-sur-Meuse
has not yet been spoken.

119
U. Fiedler, “Zu Hermann Aments These eines romanischen Bestattungsareals
auf dem merowingerzeitlichen Gräberfeld von Dieue-sur-Meuse”, Archäologie des
Frankenreiches, pp. 244–52.
   () 303

Acculturation as set entity

For aspects of acculturation in the case of Visigoths, Ostrogoths and

Lombards, the work of Volker Bierbrauer is still fundamental.120
Research conducted over the last two decades clearly shows that
processes of acculturation are the defining characteristic of the
Migration Period and the early Middle Ages.121 A similar, dynamic
process is evidenced also in the entire area dominated by the Franks
(in the political sense of the word), where the acculturation of the
Germani with “Romanic cultural models” especially is clearly visi-
ble. M. Martin has undertaken a thorough study of the effects of
late Roman influences on Germanic fibulas.122 Martin has shown
that as early as around 400, the Germanic women living in north-
ern Gaul abandoned the previously customary peplos, in favour of a
tunic worn without fibulas, and in this manner followed a practice
of dress customary in Roman circles. For the sixth and seventh cen-
turies, M. Schulze and H. Vierck have worked on the pronounced
influence of the Byzantine display-dress style on the Frankish female
costume, which is most evident in the wearing of a single fibula and
in the costly use of pearls as decoration on clothing.123 Both exam-
ples ought to be understood as indications of the fact that changes
in dress are universal elements in archaeology. The dynamics of
change in archaeological finds usual for this period, especially in
clothing, is to be emphasised particularly for the Migration Period
and the early Middle Ages. Despite archaeological research often

120
V. Bierbrauer, “Frühgeschichtliche Akkulturationsprozesse in den germani-
schen Staaten am Mittelmeer (Westgoten, Ostgoten, Langobarden) aus der Sicht
des Archäologen”, Atti del VI ° congresso internazionale di studi sull’alto medioevo (Spoleto
1980) pp. 89–105.
121
Early medieval England provides an example of acculturation that is only
archaeologically evident at a late date. According to the latest work by H. Härke,
it is not before the seventh and eighth centuries that an acculturation of the large
Roman segment of the population with the Germanic (Anglo-Saxon) immigrants
took place (Härke, “Sächsische Ethnizität”, esp. pp. 117–21).
122
Martin, “Tradition und Wandel”; id., “Fibel und Fibeltracht”.
123
M. Schulze, “Einflüsse byzantinischer Prunkgewänder auf die fränkische
Frauentracht”, Archäologisches Korrespondenzblatt 6 (1976) pp. 149–61; H. Vierck, “La
‘chemise de Sainte-Bathilde’ à Chelles et l’influence byzantine sur l’art de la cour
mérovingien au VIIe siècle”, Actes du Colloque International d’Archéologie, Rouen 1975
(Rouen 1978) pp. 521–70; id., “Imitatio imperii und interpretatio Germanica vor
der Wikingerzeit”, Les Pays du Nord et Byzance, ed. R. Zeitler, Acta Universitatis
Upsaliensis N.S. 19 (Uppsala 1981) pp. 64–113.
304  

giving the impression of being mainly concerned with constant images,

this is essentially not the case. Archaeologists are rather concerned
primarily with the recognition of those forms, within a heterogeneous
assemblage, that can be established as continuous over a longer
period of time. Constants are at best certain basic forms in the
archaeologically relevant finds-material that—with a few exceptions—
nevertheless are subject to constant typological change. The search
for stable factors in this context is represented by the attempt at
finding a sequence within these developments, on which basis a vari-
ety of questions may be discussed. Sebastian Brather quite rightly
states that: “Ethnic identities are generated by historical processes.
An ethnic identity is, therefore, not a durable or even an eternal,
unchanging entity—even if it vehemently and continuously claims to
be exactly that. In decisive situations—like moments of crisis, states
of threshold, or phases of transition—it may even change rapidly.
This is especially true for individuals”.124 Nevertheless, it is, in the
opinion of the present author, still a rule that dress was supposed
to be worn for “a life-time”. It was only under special circumstances
that it was demonstrably altered or abandoned.125 Similarly, during
the Migration Period and the early Middle Ages it is not automat-
ically the case that “a nobleman, a Gaul by birth for instance, [could]
be a Frank when close to the king and an Alaman as a duke”.126
Rather, cases of this type remain notable exceptions.

124
“Ethnische Identitäten entstehen durch historische Prozesse. Die ethnische
Identität kann deshalb keine dauerhafte oder gar ewige, unveränderliche Substanz
besitzen—auch wenn sie dies stets vehement behauptet. Sie vermag in entschei-
denden Situationen—Krisenmomenten, Schwellenzuständen oder Übergangsphasen—
sogar rasch zu wechseln. Dies gilt besonders für Individuen”: Brather, “Ethnische
Identitäten”, p. 161.
125
A classic example of this is Priskos’s description of the Greek merchant from
Viminacium, who gave up his previous existence in order to join the court of Atilla
and apparently adapted completely to the “Scythian” manner of dress. (Priscus of
Panium, Fragmenta 8, transl. E. Doblhofer, Byzantinische Geschichtsschreiber 4 [Byzan-
tinische Diplomaten und östliche Barbaren. Aus den Excerpta de legationibus des Konstantin Porphyro-
genetos ausgewählte Abschnitte des Priskos und Menander Protector] [Graz-Wien-Köln 1955]
pp. 42–3.)
126
“[ein] Adliger konnte beispielsweise von Geburt ein Gallier, in Königsnähe
ein Franke und als Herzog ein Alemanne sein”: Brather, “Ethnische Identitäten”,
pp. 167–8 (with reference to: P.J. Geary, “Ethnic identity as a situational construct
in the early middle ages”, Mitteilungen der Anthropologischen Gesellschaft in Wien 113
[1983] pp. 15–36).
   () 305

Conclusion

From an ethnic point of view, it is for the fourth and the first half
of the fifth centuries extraordinarily difficult to define that which
may be called Frankish, and in the end the image remains vague.
Nevertheless, a notable increase of Germanic finds and findings in
northern Gaul is demonstrable, which may be set in relation to the
Frankish expansion as recounted in the historical sources. In the
course of the sixth century, characteristic aspects become evident—
especially in burial-customs, grave-goods and the shapes of various
objects—that it is possible to call “Frankish”. Through the archae-
ological material (both in finds and findings) the Franks can be dis-
cerned particularly clearly in the areas of the sixth-century expansion
(southern France, south and middle Germany). The systematic set-
tlement of newly conquered regions points to a central organisa-
tion.127 The “archaeologically recognisable Franks” may, therefore,
on the basis of the archaeological material be ascribable to the “gens
of the Franks”.128 This, however, does not represent a political
definition of the Frankish gens.129 The regnum Francorum or—expressed
differently—the “realm of the Merovingians” from an archaeologi-
cal point of view does not represent an ethnic unity, but was from
its very beginnings a melting pot of people and groups of various
origin. On the basis of the archaeological evidence it is possible to
prove that belonging to the Frankish realm in a political sense did
not automatically mean an abandonment of individual (ethnic) iden-
tity (as for instance with the Alamans or the Romans). That this,
over the course of time, nevertheless did happen in the majority of
cases is shown by the manifestations of acculturation found every-
where. At the same time, a highly differentiated social stratigraphy
can be recognised or reconstructed archaeologically across the cen-
turies, on basis of the clear differentiation within the contents of the

127
Cf. note 98. See also V. Babucke, “Nach Osten bis an den Lech. Zur alaman-
nischen Besiedlung der westlichen Raetia secunda”, Die Alamannen. Ausstellungskatalaog
Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp. 249–60, esp. 251–4.
128
Undoubtedly, also those Franks, who in the west Frankish realm abandoned
grave-goods at an early date, belonged to the gens of the Franks.
129
M. Schmauder, “Die Oberschichtgräber und Verwahrfunde Südosteuropas und
das Childerichgrab von Tournai. Anmerkungen zu den spätantiken Randkulturen”,
Archäologie des Frankenreiches, pp. 55–68, esp. pp. 62–6.
306  

graves. For southern Germany this material has been studied in detail
over the past three decades. By making use of these developments
also archaeological research-results will allow ever deeper insights
into the inner structure and construction of the gens Francorum and the
regnum Francorum.

[Translated by Carine van Rhijn and Inge Lyse Hansen]

 GENS, KINGS AND KINGDOMS: THE FRANKS

Hans-Werner Goetz

With regard to the relationship between gens and regnum, the Franks
are a most irksome example: whereas we may easily observe far-
reaching political changes under Clovis’s reign, it is extremely difficult
to define the Franks as a “people”, to estimate the role of the Frankish
people during the establishment of a kingdom, or to describe the rela-
tion between gens and regnum under Merovingian rulership. Current
surveys of Merovingian history, such as those by Eugen Ewig, Rein-
hold Kaiser, Patrick Geary, or the latest Mannheim catalogue on
the Franks, do not touch on this question explicitly.1 Ian Wood alone
seems to offer a suggestion: “Clovis had transformed the Franks from
being an essentially northern people to one which was influential
in the wider politics of Gaul and the Mediterranean”.2 However,
this statement equally evades the question of how the Franks were
“transformed”. Consequently, in spite of our comparatively rich
knowledge of Frankish history, if we wish to proceed further with
our enquiries we cannot rely on straightforward research, but must
search for and examine indications of a less explicit nature. In the
following I shall attempt first to briefly outline the initial situation
of a gens Francorum and then to survey the establishment and the
development of a Frankish kingdom and the situation under the
Merovingian kings, as far as the relationship between the Frankish
people and the Frankish kingdom is concerned. Against this back-
ground I can try and suggest some summarizing observations and
conclusions with regard to changes within and brought about by the
Frankish kingdom. It is self-evident that in a brief essay like this it
is neither possible nor necessary to present a historical survey of

1
In the recent San Marino volume on “an ethnographic perspective” concern-
ing Franks and Alamanni there is a long final discussion which raises many topics,
but omits saying anything about ethnogenesis and the establishment of the king-
dom! Cf. Franks and Alamanni in the Merovingian Period. An Ethnographic Perspective, ed.
I.N. Wood, Studies in Historical Archaeoethnology 3 (San Marino 1998).
2
I.N. Wood, The Merovingian kingdoms, 450–751 (London-New York 1994) p. 49.
308 - 

Frankish history and Frankish institutions, which can easily be reread

in a large number of excellent books.3

I. The People: the “Early Franks” before Clovis?

It is obvious that “Franks” existed before Clovis’s establishment of

a huge “Frankish” realm, though it is difficult to say who these
Franks were or where they came from.4 Franci, a term handed down
in Roman sources since the end of the third century5 and thence-
forth mentioned again and again, was obviously a collective term
for a number of peoples living in the Lower Rhine region and the
Roman provinces in northern Gaul until this period. There is, how-
ever, no sufficient evidence of there having been any greater inva-

3
Cf. E. Zöllner, Geschichte der Franken bis zur Mitte des sechsten Jahrhunderts (München
1970); K.F. Werner, Les origines (avant l’an mil ), Histoire de France 1 (Paris 1984);
P. Périn and L.-C. Feffer, Les Francs, 2 vols. (Paris 1987); W. Bleiber, Das Frankenreich
der Merowinger (Wien-Köln-Graz 1988); E. Ewig, Die Merowinger und das Frankenreich,
Urban 392 (Stuttgart 1988); P.J. Geary, Before France and Germany. The Creation and
Transformation of the Merovingian World (New York-Oxford 1988); E. James, The Franks
(London 1988); Wood, The Merovingian kingdoms; R. Schneider, Das Frankenreich,
Oldenbourg Grundriß der Geschichte 5 (3rd edn., München 1995); R. Kaiser, Die
Franken: Roms Erben und Wegbereiter Europas?, Historisches Seminar N.F. 10 (Idstein
1997); id., Das römische Erbe und das Merowingerreich, Enzyklopädie deutscher Geschichte
26 (2nd edn., München 1997) (bibliography!); H. Schutz, The Germanic Realms in Pre-
Carolingian Central Europe, 400–750 (New York 2000) pp. 137–254. Further E. Ewig,
Spätantikes und fränkisches Gallien. Gesammelte Schriften (1952–1973), 2 vols., ed. H. Atsma,
Beihefte der Francia 3 (Zürich-München 1976/77); Die Franken—Wegbereiter Europas.
Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim, 2 vols. (Mainz 1996) (exhibition catalogue), and the article “Franken” in
the Reallexikon der Germanischen Altertumskunde 9 (2nd edn., 1995) pp. 373–461 by H. Beck
(Onomastics, pp. 373–4), A. Quak (Language, pp. 374–81), N. Voorwinden (Litera-
ture, pp. 381–7), H. Ament (Archaeology, pp. 387–414), H.H. Anton (History,
pp. 414–35), D. Strauch (Law and Constitution, pp. 435–44), K. Schäferdiek
(Christianization, pp. 444–7), F. Rexroth (Culture, pp. 447–61) and H. Steuer (Every-
day life, p. 461).
4
The most exhaustive representation of the early Frankish history is still Zöllner,
Geschichte der Franken, a book, however, which is almost exclusively based on writ-
ten sources. Cf. now W. Pohl, Die Germanen, Enzyklopädie deutscher Geschichte 57
(München 2000) pp. 33–7; for recent discussions see ibid., pp. 107–14.
5
Franci were mentioned for the first time with reference to the year 257/58, but
in a source (Epitome de Caesaribus 33,3, ed. F. Pichelmayer [Leipzig 1961]) written
a century later! The first contemporary allusion to the Franci was in the Panegyricus
for Constantius Chlorus (293/306): Panegyrici latini 8,21,1, ed. R.A.B. Mynors (Oxford
1964) p. 229, and was written in a manner suggesting that by that time the term
had already become completely common.
   () 309

sion or migration of larger groups into this area: These peoples seem
to have gradually “become” “Franks” (or simply been named so by
the Roman sources). They were one of those large gentile units (like
the Alamanni or the Saxones) that more or less suddenly emerged
in the course of the third century, assimilating and replacing in their
respective area a multitude of small “tribes” that were formerly men-
tioned in the sources. The reasons, however, for this shift of nam-
ing or for the regrouping of these peoples remain completely unknown.
In the case of the Franks, according to predominant opinion, these
were the Chamavi, Bructeri, Chattuarii, Ampsivarii and, possibly a
little later, the Tubantes, Usipites, Tencteri and Chasuarii, whilst the
integration of the Chatti is disputed. In virtue of archaeological evi-
dence and the Merovingian myth of a sea god being their ancestor,
a myth handed down by Fredegar, some researchers favour the claim
that the Franks descended from the ancient Germanic group that
lived in the North Sea region (Nordseegermanen), and therefore include
the Chauci among the Frankish peoples. (According to others, how-
ever, the Chauci were more related to the Saxons.) In any case,
these names do not disappear completely or immediately, but were
increasingly replaced by the the term Franci. Though the meaning
of the term is questionable,6 it is most probably Germanic (or, accord-
ing to some opinions, possibly though less probably, Celtic) in ori-
gin and therefore not a neologism of the Romans. However, it is
widely considered to be a term given to the Franks by others (Fremd-
bezeichnung). Franci may therefore have been the name of one of those
small peoples living on either side of the Roman border, which grad-
ually came to be used by the Romans as a general term to denote
the whole band of barbarians in the Lower Rhine region. What-
ever its origins, this designation was adopted by the Roman sources.
Thus, at least in the eyes of the Romans, the Franks were regarded
from the beginning as some kind of unit that came close to being
a (barbarian) people,7 though we have no knowledge of the criteria
according to which this unification came about (ethnic? political?

6
For the different suggestions, see H. Beck, “Onomastics”, Ament et al., “Franken”,
pp. 373–4.
7
For the Roman image of the barbarian, cf. A. Chauvot, “Images positives,
images négatives des Barbares dans les sources latines à la fin du Ve siècle et au
début du VIe siècle après J.-C.”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis,
l’événement, ed. M. Rouche (Paris 1997) pp. 3–14.
310 - 

cultural? religious?). It is, however, significant that the Roman sources

show evidence of an ethnic view: they regarded the Franks as being
held together by some ethnic bond which allowed the different groups
to be classified together under one name, whilst at the same time
distinguishing them from other comparable groups like the Saxons
or Alamans. Accordingly, the term Francus is linked to descent: As
Ammianus Marcellinus reports, the magister militum (and later usurper)
Silvanus emphasized his Frankish origin.8
In spite of this ethnic concept, these Franks cannot really be con-
sidered to have been either a united federal association of tribes
(Stammesverband ) or even a unified people,9 neither from the archae-
ological nor from the historical perspective, nor do we know any-
thing about their self-perception as a gens. It is, however, significant
that Ammianus Marcellinus testified to the existence of Frankish
kings,10 that he distinguished several Frankish peoples (like the famous
“Franks, that is those [Franks] who are normally called Salians”),11
and that several Frankish kings ruled at the same time: It is appar-
ent that even the Roman sources never saw the Franci as a unified
community acting as a whole. The term normally signified no more
than certain groups of Frankish origin12 which invaded Roman terri-
tory or served as warriors in the Roman army. Throughout “the”
Franks of the Roman sources were just single groups (or small parts

8
Ammianus Marcellinus, Res Gestae 15,5,16, ad a. 355, ed. W. Seyfarth, Bib-
liotheca Teubneriana (Leipzig 1978) vol. 1, p. 49 [ henceforth: Amm. Marc.]. Silva-
nus’ father Bonitus was a Frank, but he stood up for Constantine (ibid. 15,5,33,
p. 53), which proves that descent and political actions could diverge.
9
Cf. H.H. Anton, “History”, Ament et al., “Franken”, pp. 414–5. A similar state-
ment has recently been made concerning the Saxons by M. Becher, “Non enim habent
regem idem Antiqui Saxones . . . Verfassung und Ethnogenese in Sachsen während des
8. Jahrhunderts”, Studien zur Sachsenforschung 12, ed. H.-J. Häßler (Oldenburg 1999)
pp. 1–31.
10
Cf. Amm. Marc. 16,3, vol. 1, p. 72, concerning Julian’s peacemaking with the
reges Francorum; cf. also ibid., 31,10,6, vol. 2, p. 183, for Mallobaudes.
11
Amm. Marc. 17,8,3, vol. 1, p. 117. If the Salians, as M. Springer, “Gab es
ein Volk der Salier?”, Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher
Personennamen, ed. D. Geuenich, W. Haubrichs and J. Jarnut, Ergänzungsbände zum
Reallexikon der Germanischen Altertumskunde 16 (Berlin-New York 1997) pp.
58–83, has argued convincingly, actually never were a tribe or a people, it is nev-
ertheless evident—and this alone is decisive in our context—that Ammianus believed
them to be one. A second Frankish people were the Attuarii (Amm. Marc. 20,10,2,
ad a. 356, vol. 1, p. 206).
12
Already one of our earliest testimonies, Panegyricus 1 on Constantine I from
310, speaks of diuersis Francorum gentibus; cf. Panegyrici latini 6,5,3, p. 189.
   () 311

of “the” Franks) under different leaders. Contrary to Roman per-

ception, the Franks, like other groups of the Migration period, were
not an ethnic unit, but an amalgamation of different peoples and
groups. Unlike other groups, however, they were not yet politically
united under kings during this period (probably due to the fact that
there was no migration). Accordingly, we cannot discern a Heerkö-
nigtum with functions comparable to that of other Germanic peoples.13
Historically the relationship of these Franks with the Roman Empire
could be divided into at least three categories:14 On the one hand,
as mentioned above, in the later third and fourth centuries, “they”
(again that meant single groups) invaded Gaul repeatedly as pirates,
troops or raiders.15 On the other hand, “they” (that is some of them)
fought on the Roman side and served in the Roman army, or held
offices far from their native country at the imperial court:16 Several
Frankish leaders, such as Merobaudes, Richomer, Bauto or Arbogast,
were magistri militum.17 It seems that at least in some cases they attained
a double function: as Frankish “princes” and as Roman generals,18
and some of them, such as Magnentius (350) or Silvanus (355), even

13
On the Frankish warlords, cf. now J. Laing, Warriors of the Dark Ages (Stroud
2000) pp. 77–87.
14
Cf. F. Beisel, Studien zu den fränkisch-römischen Beziehungen. Von ihren Anfängen bis zum
Ausgang des 6. Jahrhunderts, Wissenschaftliche Schriften im Wissenschaftlichen Verlag
Dr. Schulz-Kirchner, Reihe 9: Geschichtswissenschaftliche Beiträge 105 (Idstein
1987) pp. 11–63.
15
Cf. Amm. Marc. 17,2,1, ad a. 357, vol. 1, p. 105 ( Julian broke into a forma-
tion of 600 lightly armoured Franks); 21,5,3, ad a. 360/61, p. 222 (invasions of the
Alamanni und Franks); 27,8,5, ad a. 369/70, vol. 2, p. 47 (raids of the Franks and
Saxons). Cf. Orosius, Historiarum adversum paganos libri VII 7,25,3, ed. K. Zangemeister,
CSEL 5 (Vienna 1882; repr. Hildesheim 1967) p. 488, following Eutropius, Brevia-
rium ab urbe condita 9,21, p. 134, ed. F.L. Müller (Stuttgart 1995) (Franci et Saxones
afflicted the northern coast); 7,35,12, p. 529 (collected armed forces—uires—of the
Gauls and Franks); 7,40,3, p. 550; Gregory of Tours, Historiae 2,9, ad a. 388, 389,
407/11, 413/15, 419/20, 428/31, 436, 440, ed. B. Krusch and W. Levison, MGH
SSrM 1,1 (2nd edn., Hannover 1951) pp. 52 ff. For this and further evidence, cf.
I. Runde, “Die Franken und Alemannen vor 500. Ein chronologischer Überblick”,
Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich,
Ergänzungs-bände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-
New York 1998) pp. 656–82.
16
Cf. Amm. Marc. 15,5,11, ad a. 355, vol. 1, p. 48: Malarich collected the
Franks many of whom exerted a great influence at the court.
17
Cf. L. Cracco Ruggini, “Les généraux francs aux IVe–Ve siècles et leurs grou-
pes aristocratiques”, Clovis 1, pp. 673–88.
18
It is significant that Ammianus, by calling the Frank Mallobaudes domesticorum
comitem regemque Francorum, attributed to him a Roman-Frankish double position
(Amm. Marc. 31,10,6, vol. 2, p. 183).
312 - 

usurped the imperial “throne”. Finally, “Franks” settled (quietly) with

their wives and children (and hence not merely as soldiers) in nor-
thern Gaul. Franks, therefore, were enemies as well as associates
of the Romans;19 Roman-Frankish relations were characterized by
military confrontation as well as peaceful settlement and service in
the Roman army. It is true that we may distinguish two phases
of Frankish-Roman relations.20 The first phase, lasting till the sec-
ond half of the fourth century, was characterized by confrontation
and raids which increasingly gave way to a (more or less) peaceful
settlement on Roman soil from the end of the third century on-
wards, a process that, according to funeral findings, was intensified
from the second half of the fourth century. Naturally, however, these
“phases” were overlapping, and it might be more appropriate to
conceive raiding and settling as simultaneous occurences carried out
by different groups respectively. Nevertheless we may assume that
settlements increased and were, at least in some cases, carried through
with Roman “permission”. Actually, our oldest evidence, the Panegyrici
on Constantius and Constantine, praise these emperors for having
settled Franks on—deserted—Roman territory, thus making them
peasants or peasant warriors in Roman service.21 Additionally, there
is Ammianus’ famous account of the year 358 when Julian settled
“Salian” Franks, curiously enough, in the same area of Toxandria
which they had illegally invaded and from where he had ostensibly
driven them away immediately before making that treaty.22 Seven
decades later, probably in 435 or 436, “Rhenish” Franks were allowed
to settle on the left banks of the river Rhine.23 It has been fervently
discussed whether these settlements were based on federate treaties
( foedera; Föderatenverhältnisse) or if the Franks only submitted themselves
as dediticii. However, such a controversy loses its impact if we acknowl-

19
According to the strange account in Gregory of Tours, Historiae 2,12, pp. 61–2,
the Franks even made the Roman governor Aegidius their king; cf. J. Jarnut, “Gregor
von Tours, Frankengeschichte II,12: Franci Egidium sibi regem adsciscunt. Faktum
oder Sage?”, Ethnogenese und Überlieferung. Angewandte Methoden der Frühmittelalterforschung,
ed. K. Brunner and B. Merta, Veröffentlichungen des Instituts für Österreichische
Geschichtsforschung 31 (Wien-München 1994) pp. 129–34.
20
Cf. Kaiser, Die Franken, pp. 38–9.
21
Panegyrici latini 8,9,4, p. 221; 8,21,1, p. 229; 6,5,3, p. 189; 6,6,2, pp. 189–90.
22
Amm. Marc. 17,8,3, vol. 1, p. 117.
23
Cf. Jordanes, De origine actibusque Getarum 34,176, p. 104; 36,191, p. 106, ed.
F. Giunta and A. Grillone, Fonti per la storia d’Italia (Rom 1991); cf. Ewig, Die
Merowinger, p. 119.
   () 313

edge, along with the latest research, that there never was a clearly
defined “federate system”, but that the respective conditions were
negotiated separately for each case.24 As well as these foederati, there
were also the laeti, barbarian warriors, who were settled on Roman
territory by Roman authorities. Furthermore, it can be assumed that
many Germanic soldiers who had served in the Roman army in
Gaul settled there after their retirement.
As far as political structures are concerned, the Franks, by the
middle of the fifth century, seem to have been governed by several
kinglets who were each assigned to certain centres (such as Childeric
and, later on, his son Clovis to Tournai). There is, however, no indi-
cation as to whether these small kingdoms were connected with
(smaller) ethnic groups within the Frankish people or whether the
(former) gentile units had survived to become regional (political) fed-
erations.25 Nor is there any clear indication of a kind of Frankish
“overlord”, parallel to the Anglo-Saxon bretwalda, among these kinglets,
though Childeric’s famous tomb already displays all the signs of a
powerful prince.26 It may also be significant that Gregory of Tours,
when looking back at the early history of the Franks, although he
is mistaken here, states that there was no mention of Frankish king-
doms in the earlier tradition,27 which shows us that he did not
acknowledge the Frankish kinglets as (real) kings. And even one of
the few things about the early Franks that historians used to agree
upon, namely the distinction between two larger groups, the “Salian”
and the Rhenish Franks (later on called Ribuarii ), who each had their
proper laws, was not a political distinction, and has now become
even more doubtful after Matthias Springer’s recent argumenta-
tion that neither Salii 28 nor Ribuarii 29 were names of peoples at all;

24
Cf. the articles by Gerhard Wirth, Peter J. Heather and Walter Pohl in: King-
doms of the Empire. The Integration of Barbarians in Late Antiquity, ed. W. Pohl, The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997).
25
Thus Kaiser, Das römische Erbe, p. 60. For the early kingdoms, see F. Staab,
“Les royaumes francs au Ve siècle”, Clovis 1, pp. 539–66.
26
For Childeric’s tomb, see now P. Périn and M. Kazanski, “Das Grab Childe-
richs I.”, Die Franken—Wegbereiter Europas 1, pp. 173–82; R. Brulet, “La tombe de
Childéric et la topographie funéraire de Tournai à la fin du Ve siècle”, Clovis 1,
pp. 59–78.
27
Gregory of Tours, Historiae 2,9, p. 57.
28
Cf. Springer, “Gab es ein Volk der Salier?”.
29
M. Springer, “Riparii —Ribuarier—Rheinfranken nebst einigen Bemerkungen
zum Geographen von Ravenna”, Die Franken und die Alemannen, pp. 201–69.
314 - 

according to Springer, the Ribuarii (a comparatively late, possibly

political term) did not mean “Rhenish Franks”, but the region of
the colony of Cologne: Thus it seems that there never were two
“peoples” of Salian and Rhenish Franks, but that there were numer-
ous groups under different kings and leaders.
For further historical suggestions, we are dependent on the results
of other disciplines. From the linguistic point of view, at least with
regard to a later period, it is difficult to identify the Franks by any
one specific language: “The” Franks did not speak one language,
but rather several dialects. Furthermore, the border-line of the Ger-
manic consonant shift (Lautverschiebung) ran right across the area of
Frankish settlement.30 (Nevertheless, linguists continue to use the
term “Frankish”, although it is difficult to say nowadays who the
Franks really were.) Moreover, the designation of language as a cri-
terion for defining a people has become an increasingly weak argu-
ment31 (even according to early medieval standards, the same language
could be used by several gentes):32 “Ethnic” communities were not
identical with linguistic groups. It is true that Otfrid of Weißenburg
referred to a Frankish language, that is, a “proper tongue” of the
Franks,33 though not until the ninth century, and not only is this an
isolated, but also an ambiguous piece of evidence, for by frenkisk, a
term parallel to the much discussed theodisk, Otfrid probably meant

30
Cf. A. Quak, “Language”, Ament et al., “Franken”, pp. 374–81.
31
Cf. E. Neusz, “Sprachraumbildung am Niederrhein und die Franken. Anmer-
kungen zu Verfahren der Sprachgeschichtsschreibung”, Die Franken und die Alemannen,
pp. 156–92, esp. pp. 165–6: Our opinions of “tribal languages” (Stammessprachen) are
not confirmed empirically; actually, our knowledge of ethnogenetical processes strongly
opposes this view.
32
Cf. H.-W. Goetz, “Gentes. Zur zeitgenössischen Terminologie und Wahrneh-
mung ostfränkischer Ethnogenese im 9. Jahrhundert”, Mitteilungen des Instituts für Öster-
reichische Geschichtsforschung 108 (2000) pp. 85–116; id., “Gentes et linguae. Völker und
Sprachen im Ostfränkisch-deutschen Reich in der Wahrnehmung der Zeitgenossen”,
Theodisca. Beiträge zur althochdeutschen und altniederdeutschen Sprache und Literatur in der Kultur
des frühen Mittelalters, ed. W. Haubrichs, E. Hellgardt, R. Hildebrandt, S. Müller and
K. Ridder, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
22 (Berlin-New York 2000) pp. 290–312; W. Pohl, “Telling the difference: Signs
of ethnic identity”, Strategies of Distinction. The Construction of Ethnic Communities, 300–800,
ed. id. with H. Reimitz, The Transformation of the Roman World 2 (Leiden-
Boston-Köln 1998) pp. 17–69.
33
Otfrid of Weißenburg, Evangelienbuch 1,1,31 ff., ed. O. Erdmann (Halle 1882;
repr. 6th edn. Tübingen 1973) p. 12: The Franks (Fráncon) spoke in their own tongue
(in sína zungun).
   () 315

no more than common, colloquial speech.34 Moreoever, it has become

widely accepted nowadays that the (later) language boundary is not
identical with the Frankish settlement area. Whereas earlier research
assumed a “Germanization” of northern France from the fifth cen-
tury on, followed later by a re-Romanization,35 this view has been
challenged by the observation that Germanic place-names were sel-
dom used for any lengthy period of time; we should, therefore, not
overestimate the relevance of Frankish settlement.36 Consequently,
the traditional opinion has been abandoned in favour of the assump-
tion of a large bilingual language belt which included both Romance
and Frankish linguistic enclaves (some, as in the Moselle region, with
a duration that sometimes reached far into the Middle Ages).37 Many
loan-words on both sides are the result of a long-standing bilingual
situation and extensive communication between Romans and Franks.
The same seems to be true for place-names: not every site end-
ing with -heim or -court is indisputably “Frankish” as was assumed by
former research; and it is still to be seen whether it will ever be pos-
sible to detect specific “Frankish” attitudes and traits in the giving
of personal names (a research project that has been pursued for some

34
Cf. Neusz, “Sprachraumbildung am Niederrhein”, pp. 168–9.
35
Cf. F. Petri, Die fränkische Landnahme und die Entstehung der germanisch-romanischen
Sprachgrenze in der interdisziplinären Diskussion, Erträge der Forschung 70 (Darmstadt
1977).
36
Cf. W. Haubrichs, “Fränkische Lehnwörter, Ortsnamen und Personennamen
im Nordosten der Gallia. Die ‘Germania submersa’ als Quelle der Sprach- und
Siedlungsgeschichte”, Die Franken und die Alemannen, pp. 102–29 (with an exhaustive
bibliography).
37
Cf. W. Haubrichs, “Germania submersa. Zu Fragen der Quantität und Dauer
germanischer Siedlungsinseln im romanischen Lothringen und Südbelgien”, Verborum
amor. Studien zur Geschichte und Kunst der deutschen Sprache. Festschrift Stefan Sonderegger,
ed. H. Burger, A.M. Haas and P. von Matt (Berlin-New York 1992) pp. 633–66;
W. Kleiber, “Mosella Romana. Hydronomie, Toponymie und Reliktwortdistribu-
tion”, Die Franken und die Alemannen, pp. 130–55. Regarding linguistic aspects of the
continuity of (Roman) language, settlement, population and culture in the Moselle,
Central and Upper Rhine and Black Forest regions, see W. Kleiber and M. Pfister,
Aspekte und Probleme der römisch-germanischen Kontinuität. Sprachkontinuität an Mosel, Mittel-
und Oberrhein sowie im Schwarzwald (Stuttgart 1992). The continuity of the Roman
language is interpreted as an indication of a continuity of settlement. For the vari-
ety (and inexplicability) of language barriers, see W. Haubrichs, “Über die allmäh-
liche Verfertigung von Sprachgrenzen. Das Beispiel der Kontaktzonen von Germania
und Romania”, Grenzen und Grenzregionen. Frontières et régions frontalières. Borders and
Border Regions, ed. W. Haubrichs and R. Schneider, Veröffentlichungen der Kom-
mission für Saarländische Landesgeschichte und Volksforschung 22 (Saarbrücken
1993) pp. 99–129.
316 - 

years and is being continued by an interdisciplinary group of schol-

ars).38 No doubt there were differences between a “Germanic” and
a “Romanized” population,39 but these became less and less dis-
cernible (probably an indication of new ethnogenetical processes
within the Frankish empire). Thus, the population in northern Gaul
was (and increasingly became) a “mixed civilization”.40 (This is very
different from Britain which after the period of migration apparently
became almost completely “Anglo-Saxon”.)41 Yet the great number
of Romance place-names have to be regarded as an indication of
the widespread continuity of Roman settlement.
On these grounds, it is almost impossible to discern distinct cen-
tres or border-lines of Frankish settlement in northern Gaul, though,
from the third century, the main areas were with certainty situated
in the northern parts of the former Roman provinces Germania and
Belgica, around the rivers Lower Rhine, Waal and Meuse. From
here, Frankish settlement spread gradually, stretching as is generally
assumed, as far as the Scheldt (Escaut), Somme and Seine by the
end of the fifth century.42 By comparison, before Clovis’s reign there
had been no considerable Frankish settlement in the regions between
the rivers Seine and Loire, and there never was any considerable
settlement in southern France (south of the Loire).43

38
Cf. J. Jarnut, “Nomen et gens: Political and linguistic aspects of personal names
between the third and the eighth century—presenting an interdisciplinary project
from a historical perspective”, Strategies of Distinction. The Construction of Ethnic Communities,
300–800, ed. W. Pohl with H. Reimitz, The Transformation of the Roman World
2 (Leiden-Boston-Köln 1998) pp. 113–6; Nomen et gens; D. Geuenich and W. Kette-
mann, “Das Pilotprojekt zur gens Alamannorum. Erste Erfahrungen mit einem Teilpro-
jekt von ‘Nomen et gens’”, ibid., pp. 279–303. Cf. also D. Geuenich, “Personennamen
als Personengruppennamen”, Proceedings of the Thirteenth International Congress of Onomastic
Sciences 1, ed. K. Rymut (Wroc∑aw 1981) pp. 437–45.
39
Nevertheless it seems that Beisel, Studien zu den fränkisch-römischen Beziehungen,
p. 49 exaggerates an “ethnic” contrast between “Germans” and “Romans”.
40
Cf. H.W. Böhme, Germanische Grabfunde des 4. bis 5. Jahrhunderts. Zwischen unterer
Elbe und Loire. Studien zur Chronologie und Bevölkerungsgeschichte, Münchener Beiträge zur
Vor- und Frühgeschichte 19 (München 1974).
41
Cf. H. Härke, “Sächsische Ethnizität und archäologische Deutung im früh-
mittelalterlichen England”, Studien zur Sachsenforschung 12, pp. 109–22.
42
There may have been a caesura between Childeric and Clovis, since settlement
south of the Somme is more or less restricted to the time after 486; cf. P. Périn,
“La Progression des Francs en Gaule du Nord au Ve siècle. Histoire et archéolo-
gie”, Die Franken und die Alemannen, pp. 59–81; L. Verslype, “L’occupation méro-
vingienne aux confins de l’Austrasie et de la Neustrie septentrionales et l’image
archéologique des aristocraties”, Clovis 1, pp. 567–605.
43
Cf. G. Halsall, “Social identities and social relationships in early Merovingian
Gaul”, Franks and Alamanni, pp. 141–65; 165–75.
   () 317

From the archaeological point of view,44 there is no sign of a “con-

quest” or rapid settlement, but, according to current opinions, we
have to assume a gradual expansion over a period of several cen-
turies: Frankish settlement in northern Gaul was not the result of
“migration” but of gradual infiltration, partly by invasion, partly
officially supported, or at least tolerated.45 We have to reckon, how-
ever, with great regional differences.46 Nevertheless, the decision as
to what is “Frankish” in northern Gaul has become less clear now-
adays than it used to be.47 Former assumptions that the so-called
Reihengräberfelder (tombs in a line, with the interment of whole corpses,
Körperbestattungen, and with grave goods) were Frankish have become
doubtful because such burial rites were not restricted to northern
France. On the other hand, not all burial grounds in northern Gaul
were non-Roman!48 On the contrary, we can observe that the “other”
type of burial was adopted by both Roman and “Germanic” minori-
ties.49 And what seemed to be Roman graves could in fact have been
Christian ones.50 Thus, though the Reihengräberkultur was presumably
“Germanic” in its origin, it did not necessarily remain so for any
length of time, particularly not in regions with a mixed civilization.51
Accordingly, it has become almost impossible to define “Frankish”

44
Cf. the contribution of Michael Schmauder in this volume.
45
For this expansion, see now Staab, “Les royaumes francs”.
46
Cf. G. Halsall, Settlement and Social Organization. The Merovingian Region of Metz
(Cambridge 1995), who, in the region of Metz, observes changes in the fourth cen-
tury in the context of the Alamannic invasions and, again, important social changes
at about 600.
47
Archaeologists provide evidence of “alien” elements and settlements in northern
Gaul since the middle of the fourth century, but, of course, they cannot prove that
these were Franks. Cf. H.W. Böhme, “Söldner und Siedler im spätantiken Nordgal-
lien”, Die Franken—Wegbereiter Europas 1, pp. 91–101.
48
Cf. H.W. Böhme, “Franken und Romanen im Spiegel spätrömischer Grab-
funde im nördlichen Gallien”, Die Franken und die Alemannen, pp. 31–58. So far, we
have knowledge of up to 200 cemeteries of the fourth and fifth centuries in north-
ern Gaul. Similarly, the end of this “funeral civilization” in the seventh century can
neither be attributed to ethnological reasons nor to Christianization, but it was
due to a changing ecclesiastical organization (with funerals near the churches); cf.
U. Koch, “Stätten der Totenruhe—Grabformen und Bestattungssitten der Franken”,
Die Franken—Wegbereiter Europas 2, pp. 723–37.
49
Cf. H. Ament, “Franken und Romanen im Merowingerreich als archäolo-
gisches Forschungsproblem”, Bonner Jahrbücher 178 (1978) pp. 377–94; B.K. Young,
“Pratiques funéraires et mentalités païennes”, Clovis 1, pp. 15–42.
50
Cf. P. Périn, “A propos de publications étrangères récentes concernant le peu-
plement en Gaule à l’époque mérovingienne: La ‘question franque’”, Francia 8 (1980)
pp. 537–52.
51
According to G. Halsall, “The origins of the Reihengräberzivilisation: forty years
318 - 

from an anthropological point of view,52 though there may have been

differences between Franks and Romans,53 nor can we clearly define
certain types of habits, jewellery, costume, or weapons as being un-
questionably “Frankish” (particularly since, later on, “Frankish” women
even in the Belgica dressed themselves in the Roman style).54 Simi-
larly, it seems hardly possible to detect a specific Frankish culture,
so that some archaeologists, such as Guy Halsall, tend to doubt the
ethnic relevance of archaeological findings at all:55 The continuation
of material culture does not necessarily mean a common ethnic
identity (and vice versa).56 Considering these “mingled cultures” in
Gaul, it seems, once more, that the Franks were (or at least developed
into) a political rather than an ethnic or cultural phenomenon.57

on”, Fifth-century Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge
1992) pp. 196–207, gravegoods do not represent “Germanic” or “pagan” (or “Ger-
manic pagan”) customs.
52
Cf. M. Kunter and U. Wittwer-Backofen, “Die Franken—Anthropologische
Bevölkerungsrekonstruktionen im östlichen Siedlungsgebiet”, Die Franken—Wegbereiter
Europas 2, pp. 653–61.
53
Cf. L. Buchet, “Die Landnahme der Franken in Gallien aus der Sicht der An-
thropologen”, Die Franken—Wegbereiter Europas 2, pp. 662–7.
54
Cf. G. Zeller, “Tracht der Frauen”, Die Franken—Wegbereiter Europas 2, pp.
672–83. Cf. Schmauder this volume, p. 303.
55
Cf. Halsall, “Social identities”; id., “The origins of the Reihengräberzivilisation”.
Nowadays, it has become even more difficult to distinguish between the Franks and
other Germanic peoples, such as the Saxons or Alamanni, by their archaeological
heritage although, according to the written sources, these peoples were clearly dis-
tinct. Cf. Ch. Grünewald, “Neues zu Sachsen und Franken in Westfalen”, Studien
zur Sachsenforschung 12, pp. 83–108; H.W. Böhme, “Franken oder Sachsen? Beiträge
zur Siedlungs- und Bevölkerungsgeschichte in Westfalen vom 4.–7. Jahrhundert”,
ibid., pp. 43–73.
56
Thus H. Ament, “Die Ethnogenese der Germanen aus der Sicht der Vor- und
Frühgeschichte”, Ethnogenese europäischer Völker. Aus der Sicht der Anthropologie und Vor-
und Frühgeschichte, ed. W. Bernhard and A. Kandler-Pálsson (Stuttgart-New York
1986) pp. 247–56. An opposite view is now again taken by F. Siegmund, Alemannen
und Franken, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
23 (Berlin-New York 2000), who after discussing the archaeological problem of eth-
nicity (pp. 55 ff.), defends an ethnical perspective as long as it does not adapt eth-
nic concepts from other disciplines (p. 81). Nevertheless, maps of distribution of
certain objects cannot prove ethnic distinctions, and it seems to be completely
arbitrary when Siegmund (pp. 253 ff.) interprets the archaeological “cultural pat-
terns” as ethnic units which he identifies with being Franks and Alamans (pp. 305
ff.) on the ground of the (delusive) assumption that archaeological evidence is true
and objective (pp. 351 ff.).
57
Thus Périn, “A propos de publications étrangères récentes”, p. 552.
   () 319

All in all, we can conclude that there were Franks before Clovis’s
reign, but these Franks were far from being discernible as a confined
“ethnic” community. Whereas late Roman sources (like Ammianus
Marcellinus) perceived the Franks as an ethnic group, the (admit-
tedly few) contemporary sources of the late fifth and early sixth cen-
tury tended to avoid the term. It may well be that the mixture of
“Roman” and “Germanic” civilization in northern Gaul stimulated
an ethnogenetical process that had begun long before the establish-
ment of a Frankish kingdom. Whatever it was that made these Franks
appear as a unit in the fourth and fifth centuries, they obviously did
not form a political union either. Nevertheless, there was a political
order before Clovis, and in spite of Gregory of Tours complaining
that he could not find any kings in the early history of the Franks,
there were kings who, to some extent at least, were related to each
other. From this we may conclude that there were even certain forms
of royal “dynasties”, stirpes regiae (among which the Merovingians
were one). It is only by their later success and their historiographical
tradition that their line of succession is the only one which is fairly
well-known. Gregory, however, was obviously searching for one king
of the (united) Franks.

II. The Kingdom: the Achievement of Clovis

The reign of Clovis, therefore, may not have changed much, at least
not immediately, with regard to Frankish-Roman civilization. Politically,
however, the amazingly rapid development from being a kinglet of
Tournai to becoming the ruler of the largest and, beside Theoderic’s
kingdom of the Ostrogoths, most important kingdom within some
20 years must be considered as a decisive caesura at least politically.
Under Clovis, the Franks, for the first time in their history, became
a political unit ruled by one king. No doubt, there were some “fore-
runners” or developments that supported a unification of Gaul, but,
characteristically enough, they originated mainly on the “Roman”
side: One may go back as far as the so-called “Gallic ‘separate
realm’” (Gallisches Sonderreich) of the third century from Postumus to
Tetricus (260–274), or to Diocletian’s system of tetrarchy with a
decentralized imperial court in Trier. Moreover, Clovis’s reign had
been prepared by his father’s, Childeric’s, federal policy with Aegi-
dius, the Roman governor, in the central parts of Gaul (around
320 - 

Soissons and Paris). The conquest of this last Roman resort in Gaul
under Aegidius and his son Syagrius no doubt has to be regarded
as an important milestone in the course of Clovis’s rise. But, of
course, there were also the other “Germanic” kingdoms (of the Visi-
goths in Aquitania, the Burgundians in the Rhone area or the
Ostrogoths in Italy) that might have served as models.
Compared to other realms, the establishment of the huge Frankish
kingdom was a comparatively late “foundation”, but it was achieved
under one single king within one generation (disregarding later expan-
sions under Clovis’s successors), and it may be added here that, con-
trary to the other “Germanic” kingdoms, the Frankish realm never
perished (which does not necessarily imply, however, that its “suc-
cess” derived from its being different). Clovis himself as ruler was
the central figure of this realm. His reign most probably stimulated
a great deal of changes. First and foremost, the area of domination
grew to an extent that made it larger than any other Germanic king-
dom of that period. Moreover, Clovis managed to eliminate the other
Frankish kings (some of whom, at least, were his relatives) one after
the other until, probably towards the end of his reign, he finally
even became king of the “Rhenish Franks” after having murdered
their king Sigibert and his son Chloderic. The conquest of the remain-
ing regions of the Roman Empire in Gaul, the province of Syagrius,
made him leader and governor of a district not only inhabited by
a population that was still predominantly Roman, but also of a region
with a functioning Roman administration and with Roman law and
habits. After (probably) two campaigns and victories against the
Alamanni (who, as it seems, were not united either under one king
before that time),58 Clovis’s realm expanded into the Upper Rhine
region, and by his notorious victory at Vouillé (506) over the Visi-
goths he conquered a complete, huge kingdom which made him
ruler of the whole of south-west Gaul south of the Loire (later on
called Aquitania), a region that had been under Visigothic rulership,
but remained Roman in faith, population and administrative struc-
tures. It is significant that Clovis (as is generally assumed) transferred
his “capital”, that is, his preferred sedes regia, first from Tournai to
Soissons and, later on (about 508) further to Paris,59 whereas, after

58
Cf. D. Geuenich, “Chlodwigs Alemannenschlacht(en) und Taufe”, Die Franken
und die Alemannen, pp. 423–37.
59
For Frankish (and other) residences cf. E. Ewig, “Résidence et capitale pen-
   () 321

the conquest of the Visigothic realm, he did not move the “capital”
further again to Tours, a town near the border between the two
former kingdoms which, noted for the cult of Saint-Martin, he undoubt-
edly honoured as a significant sacred place: The “core” of Clovis’s
Frankish realm was obviously not the regions of Frankish origin, but
the Roman parts with their vast demesnes (around Paris) which
remained the economic basis of all the Frankish and, later on, also
of the French kings.
In the end, Clovis’s “Frankish” kingdom (whatever it may have
been called at that time), from its substance and population, was far
from being “Frankish” any more, but was an amalgamation of sev-
eral (former) kingdoms including the small Frankish kingdoms of the
north, an important Roman province, the huge Visigothic kingdom
of southern Gaul and the territories of the Alamanni. Clovis him-
self obviously secured his position on the basis of several public func-
tions (though we do not know the title he assumed for himself ): He
was king of the Franks as well as successor of the Roman province
of Aegidius and Syagrius, ruling as a sovereign, but also in the name
of the Emperor who, granting him the honour of an “honorary con-
sulship” in 508, acknowledged his rule.60 In addition he may well
have perceived himself as king of the Visigoths since this kingdom
was divided under his sons and was thus treated as a separate realm
(a method that was continued later on with Burgundy and Provence,
the former realms of the Burgundians and the remaining Visigoths).
Thus, Clovis’s kingdom comprised several kingdoms (Vielreichestaat) and
was composed of even more peoples (Vielvölkerstaat) among which the
Franks were only one (and nowhere near the majority, probably not
even in the north). We do not know the percentage of Franks
in the population,61 but from the fact that the Romanic language
not only survived, but, at least in the end, dominated we may con-
clude that the Franks always remained a minority, even in northern
Gaul. The majority of the population was Roman or, rather, Ro-
manized, having grown together, through a long process, primarily

dant le haut Moyen Age”, id., Spätantikes und Fränkisches Gallien 1, pp. 362–408, par-
ticularly pp. 383 ff.; A. Dierkens and P. Périn, “Les sedes regiae mérovingiennes entre
Seine et Rhin”, Sedes regiae (ann. 400–800), ed. G. Ripoll and J.M. Gurt with A. Cha-
varría (Barcelona 2000) pp. 267–304.
60
Cf. R. Mathisen, “Clovis, Anastase et Grégoire de Tours: consul, patrice et
roi”, Clovis 1, pp. 395–407.
61
Estimates range from 2 to 25%. The percentage decreased, however, from the
north farther to the south as well as to the west.
322 - 

from “Romans” (from various provinces) and Celtic Gauls. Moreover,

we may well ask how “Romanized” the Franks already were even
under or before Clovis’s reign! If the Franks before Clovis can be
called a “people”, even though they did not form a recognizable
unit, the Frankish kingdom after Clovis was by no means an ethnic
(or cultural) unit, but a kingdom made up of many peoples and of
many kingdoms which in themselves consisted of mixed ethnic groups
(Völkergemisch). This, however, was common in the “Germanic” world
during and after the Migration period.
As a preliminary resume we may maintain that, though there was
a Frankish people before Clovis (or at least a people who were called
“Franks”), it is impossible to decide which bonds united them. There
were kings, but no united kingdom. It was Clovis who founded the
Frankish kingdom, but he did not establish the Frankish people
which somehow survived (and expanded) under his reign. This seems
a very important distinction: “Kingdom” and “people”, though ob-
viously related to each other, were distinct institutions. The word-
ing Francorum gens et regnum in Gregory of Tours62 indicates a distinction
as well as a relation between these two elements. The situation,
however, is extremely complicated: Though a Frankish people had
existed before Clovis’s reign, it was not until Clovis’s conquests that
the Franks were politically united. Once they were united, however,
they ruled a kingdom that incorporated many peoples, in a realm
in which they, themselves, were a minority. In the final analysis, we
cannot grasp the “ethnicity” of the Franks. On the one hand the
establishment of the Frankish kingdom can be regarded as an impor-
tant factor in this process. On the other hand it created a frame-
work which, when regarding the Franks as a “people”, makes it very
probable (although it remains to be investigated) that there was also
a development of the “Frankish people”, that is, a change of its
ethnicity under the new conditions of a Frankish kingdom.

III. The Structure and Development of the Frankish Kingdom and People

1. Merowingian Kingship: Under this premise, we can now pursue the

elements and structure of the “Frankish” kingdom, considering particu-

62
For example, Gregory of Tours, Historiae 5 prol., p. 193.
   () 323

larly the relationship between the Frankish people and “its” king-
dom: the political and social institutions, the relationship between
Franks and Romans as well as between Franks and other “Germanic”
peoples within the realm, the personal relationships and the Franks’
concepts of themselves. This vast new kingdom called for newly
adapted structures. The political unification had not yet established
an integrated realm. On the contrary, one may assume that the king
himself, even if he was not the only factor, was, at least, the most
important one in the process of integration:63 It was not the Franks as
a whole, but the Merovingian dynasty that formed that notorious
“nucleus of tradition” (Traditionskern) which played such an impor-
tant part in Wenskus’s and Wolfram’s theory of ethnogenesis (though
recently the earlier opinions relating to a Germanic Königsheil have
rightly been increasingly dismissed). The Merovingians monopolized
political power and “sanctified” themselves by way of sacred cus-
toms (as “long-haired kings”) and through attitudes and myths (like
the descent from a sea-beast).64 However, they perceived themselves,
or were perceived, as “kings of the Franks” in a regnum Francorum,
in spite of their rule over many peoples. And yet, due to the lack
of original charters and the vast amount of falsifications, we cannot
tell exactly from what time on these denominations were used. In
royal charters, the title rex Francorum is handed down for the first
time in a copy of a charter of Theudebert II from 59665 and in
two original charters of Chlothar II, one issued between 584 and
628,66 the other dating from 625.67 Later on, the title was extremely

63
Kingship can be regarded as an, or as the most important element of any
ethnogenetical process, and it is not by chance that H. Wolfram, “Typen der
Ethnogenese. Ein Versuch”, Die Franken und die Alemannen, pp. 608–27, recently
defined different “types” of ethnogenesis primarily according to the monarchic or
non-monarchic organisation of these peoples.
64
These features should not be confused with a Sakralkönigtum which is more or
less a construct of modern research; cf. A.C. Murray, “Fredegar, Merovech, and
‘Sacral Kingship’”, After Rome’s Fall. Narrators and Sources of Early Medieval History.
Essays presented to Walter Goffart, ed. id. (Toronto-Buffalo-London 1998) pp. 121–51.
65
D. Merov. 25, ed. T. Kölzer, MGH DD regum Francorum e stirpe Merovingica
(Hannover 2001) vol. 1, p. 69. Cf. now also R. Schneider, “König und Königsherrschaft
bei den Franken”, Von Sacerdotium und Regnum. Geist und weltliche Gewalt im frühen und
hohen Mittelalter. Festschrift für Egon Boshof zum 65. Geburtstag, ed. F.-R. Erkens and H.
Wolff (Köln-Weimar-Wien 2002) pp. 11–26, here p. 15, who nevertheless believes
that the title was used from the reign of Clovis.
66
D. Merov. 22, p. 63.
67
Ibid., 28, p. 76.
324 - 

common, but it is impossible to say when it was used for the first
time. The term regnum Francorum, however, was used very rarely: in
a falsification dating from 639/4268 and in an undisputed charter of
Childeric II from 670.69 In historiographical sources, however, the
regnum Francorum was the kingdom of the Frankish king(s). Thus, it
was the kings (and their court) that guaranteed not only political
unification, but also warranted that this unified realm was perceived
as a Frankish one. In fact, there was hardly any unity and uniformity
(or conformity) inside this realm except through the king and per-
haps the church (and even the church was an “amalgamation” of
single bishops, each being responsible for his see,70 and not infre-
quently quarrelling with their respective comes, as illustrated by Gregory
of Tours’ famous quarrel with Count Leudastus of Tours).71
Yet, if we are accustomed to speaking about the Frankish king-
dom of the Merovingians and, later on, the Carolingians, it seems
necessary to emphasize a crucial point here. Although Clovis no
doubt established a united realm and expanded it by huge conquests,
his reign as a single king without colleagues was exceptional. It is
well-known that the four sons of Clovis divided the paternal king-
dom among themselves, and obviously this turned out to be the
usual procedure throughout Merovingian history:72 Under Clovis’s
successors, in the sixth and seventh centuries, there were only four
more or less brief periods when the whole Frankish realm was equally
united again under one king (Chlothar I, 558–561; Chlothar II,
613–623; Dagobert I, 629–639; Childeric II, 673–675): Although
the Frankish kingdom was ruled either by one king (though this was
seldom) or at least (and throughout) by one royal family, it was not
really a political unity, even if the notion of one Frankish kingdom
may still have been maintained. Curiously enough, the “divisions”,
which possibly confirm the opinion of the kingdom being the prop-

68
Ibid., 73, p. 188.
69
Ibid., 108, p. 280.
70
For the Merovingian bishops, see G. Scheibelreiter, Der Bischof in merowingischer
Zeit, Veröffentlichungen des Instituts für Österreichische Geschichtsforschung 27 (Wien-
Köln-Graz 1983), and B. Basdevant-Gaudemet, “L’évêque, d’après la législation de
quelques conciles mérovingiens”, Clovis 1, pp. 471–94.
71
Cf. Gregory of Tours, Historiae 5,48–49, pp. 257–63; 6,32, pp. 302–4.
72
Cf. E. Ewig, “Die fränkischen Teilungen und Teilreiche (511–613)”, id., Spätantikes
und fränkisches Gallien 1, pp. 114–71.
   () 325

erty of the Merovingian kings, are in themselves proof of such a

notion. In reality, however, there is no doubt a discrepancy between
the divisions and the concept of the existence of a single regnum
Francorum. In Gregory of Tours, we can sense that strange relation-
ship (obviously not felt as a contradiction by Gregory himself ) between
this ideal and the sovereignty of the single parts: For Gregory, all
Merovingian kings were legitimate (though not all of them were good
rulers). The Merovingians themselves claimed to be heirs by way of
their birthright, thereby still treating the kingdom as a whole, and
there were, in fact, numerous forms of royal inheritance patterns.73
In practice, the Frankish kings never achieved complete integration
of the different parts (and peoples) of their kingdom.

2. Disunity: If the Frankish kingdom, therefore, even as a political

factor was a unit only in a limited, dynastic sense, internally it seemed
to be even more heterogeneous, and its parts increasingly disinte-
grated. The vast expansions under Clovis (Aquitania, Alemannia) and
his successors (Burgundy, Thuringia, Bavaria, and, for a short time,
northern Italy) integrated not only external peoples (such as the
Alamans and the Bavarians), but also former kingdoms (such as the
kingdoms of the Visigoths, Burgundians and Thuringians) as well as
former Roman territories: The Frankish kingdom remained, as it
seems, a disorganized amalgamation of several (older) kingdoms, and
it is well known that Gaul can easily be divided into a “Germanic”
northern and a Romanized southern part (though this distinction,
too, seems somewhat superficial); at the same time we may observe
a comparable, though completely different distinction between the
western and eastern parts of the Frankish territory (later on called
Niustria, or Neustria, and Austria, or Austrasia).74 When Clovis’s sons
not only divided the paternal realm, but actually split up the two
central parts of the kingdom, the Roman “Niustria” and the Gothic
kingdom (later on called “Aquitania”), this action could indicate that
they (and their successors) may never have aspired to achieve com-
plete political integration.

73
Cf. I.N. Wood, “Kings, Kingdoms and Consent”, Early Medieval Kingship, ed.
P.H. Sawyer and I.N. Wood (Leeds 1977) pp. 6–29.
74
This is confirmed by the archaeological evidence; cf. Schmauder this volume,
pp. 285–7.
326 - 

Moreover, the numerous reports in Gregory of Tours and later

chronicles about quarrels and wars between the Merovingian kings
and about mutual invasions seriously contradict the impression of
(political) unity. According to these chronicles, the Merovingians
seemed to have fought one another throughout most of Merovin-
gian history, and it is not by chance that Ian Wood, when writing
on this subject, titles his chapter on the sixth-century kingdom “Stabi-
lity in Disunity”75 (though this can also be seen as typically con-
temporary and not as particularly “Merovingian”). Yet the struggles
between the Merovingians were struggles between kings, not between
peoples (or even between “Romanized” and “Germanic” parts):76
According to Gregory, it was “the” Franks fighting on each side.
We may doubt, therefore, whether the Frankish kingdoms after Clovis
were ever a political, let alone a “gentile” (national) unit. We may
even ask if contemporary historians have used Clovis’s realm too
exclusively as a model when forming their judgments of later Frank-
ish kingdoms.
Although the manner of dividing the realm changed during the
next centuries, and accordingly the portions of the Frankish king-
doms did not remain the same, the habit of dividing the kingdom
was maintained throughout these times. From the reign of the sons
of Chlothar I, these parts tended to become independent kingdoms
in their own right. It is obvious, therefore, that, under the later
Merovingians, the four great parts of the realm, Niustria, Austrasia,
Burgundia and Aquitania, became more and more important as
structural factors.77 (Later on, Niustria, Austrasia, and Burgundia had
a maior domus of its own.) In Merovingian times the new large regions
vaguely corresponded to the boundaries of former realms, as in
Aquitania or Burgundy. We must, however, admit that the parts
which were to become independent kingdoms, did not correlate with
the original boundaries, indicating that in the meantime fundamen-

75
Wood, The Merovingian kingdoms, p. 88.
76
Thus Ewig, “Die fränkischen Teilungen”, p. 171.
77
Cf. E. Ewig, “Die fränkischen Teilreiche im 7. Jahrhundert (613–714)”, id.,
Spätantikes und fränkisches Gallien 1, pp. 172–230; for the later development: id.,
“Descriptio Franciae”, ibid., pp. 274–322. For the growth of these regna, cf. K.F.
Werner, “Völker und Regna”, Beiträge zur mittelalterlichen Reichs- und Nationsbildung in
Deutschland und Frankreich, ed. C. Brühl and B. Schneidmüller, Historische Zeitschrift
Beiheft N.F. 24 (München 1997) pp. 15–44, particularly pp. 18 ff.
   () 327

tal changes must have occurred. By Carolingian times, these bonds

had completely lost their impact.
Furthermore, the Frankish kingdom was not a legal unit, either.
Whereas the Lex Salica encompassed laws concerning Franks and
Romans, the different “Germanic” peoples lived, as it seems, under
their own proper laws (although all of these were codified later,
mainly in the seventh and eighth century, like the Leges Alamannorum
and the Lex Baiuvariorum), and this situation continued under the
Carolingian kings. These non-Frankish “Germanic” peoples inside
the realm still preserved a specific legal system of their own though
their laws were obviously codified under Merovingian influence. As
is well known, there were even two Frankish laws, the Lex Salica and
the Lex Ribuaria. Actually, there was not just one Salian law, as we
know of at least eight different versions of the Lex Salica. Similarly,
the duces, established as an element of Frankish (royal) administration
(and therefore integration), developed, in some regions at least, into
influential regional powers, which were based upon and tied to cer-
tain “gentile” structures, subkingdoms or traditional territories (such
as Aquitania, Burgundy, Provence, Thuringia, Alemannia or Bavaria).

3. Integration: Thus, on the one hand, the Frankish kingdom was

never more than an amalgamation of different parts. On the other
hand, however, we cannot neglect the impression that the contem-
porary notion of a Frankish kingdom (or, later on, of Frankish king-
doms) increased constantly. Obviously, there had been a process
of—more or less successful—integration. What, we may therefore
ask, were the key factors of this process? In spite of the undoubt-
edly important political role of the king, it could not be the “mon-
archy” as such, since the periods characterized by the reign of a sole
monarch turned out to be the rare exceptions to the rule in the
Merovingian as well as in the Carolingian period. Taking this into
consideration, we should not neglect the possibility of a growing
homogeneity of a population that initially had had distinctly different
origins, traditions, and (social) structures. The relationship between
“Romans” and “Franks” (and other peoples) is a decisive factor here.
In spite of an actual “mixed civilization”—a state that in many parts
of northern Gaul might have been achieved long before Clovis—,
and despite the legal differentiation between these groups, with the
Romans being entitled to only half the wergeld of a Frank, there
seems to have been not only a symbiosis, but even an increasing
328 - 

“syncretism”78 of ethnic groups and a growing similarity between the

peoples and cultures. This can be observed in the fact that the same
public functions (for example the offices of comites or bishops) were
open to both groups almost from the beginning though in differing
ratios.79 The Gallo-Roman senatorial aristocracy emphatically sup-
ported the Frankish state, particularly in southern France. The his-
tories of Gregory of Tours at the end of the sixth century provide
one of the best examples to prove this statement. This holds true
even though Gregory did not apply the term “Roman” to the “Ro-
man” population of Gaul, but from his perspective distinguished,
in a much more precise manner, between the different regions or
territories of southern Gaul (and their respective population),80 where-
as the Franks were seen more as a whole. Even the social structures
did not seem so different as they may appear at first sight (and dis-
tinctions become even fewer when we take into account the “Ro-
manization” of the Germanic war-leaders long before Clovis’s time).
There was some kind of Germanic aristocracy long before Clovis
(and, in spite of a long controversy, it was not dissolved under Clo-
vis’s reign, but survived).81 With regard to the ornate grave goods,
archaeologists speak of “aristocratic graves” (Adelsgräber) or even of
“hereditary aristocratic graves” (adlige Erbgrabanlagen).82 It has been
claimed that, at least from the point of view of the Gallo-Roman

78
Cf. F. Rexroth, “Culture”, Ament et al., “Franken”, p. 452.
79
According to M. Weidemann, Kulturgeschichte der Merowingerzeit nach den Werken
Gregors von Tours (Mainz 1982) vol. 1, pp. 30 and 64, the majority of comites (27)
mentioned in Gregory of Tours bore Roman names, only 12 had Germanic names.
Among the duces, however, 19 had Germanic and only 11 Roman names.
80
Gregory distinguishes only twice between a Frank (Francus genere) in contrast to
someone from Arles (Historiae 10,2, p. 482) or from Clermont (Historiae 4,40, p. 173:
Arvernus). It is also significant that not only the names of “peoples” were territo-
rialized, but, for example in southern Gaul, territorial names were “gentilized” (such
as Turonici or Biturgici ); cf. W. Pohl, “Zur Bedeutung ethnischer Unterscheidungen
in der frühen Karolingerzeit”, Studien zur Sachsenforschung 12, pp. 193–298, here p. 201.
81
Cf. F. Irsigler, Untersuchungen zur Geschichte des frühfränkischen Adels, Rheinisches
Archiv 70 (Bonn 1969); cf. Schmauder this volume. It need not concern us here
whether this political “elite”, whose existence as such is not disputed, may already
be called an “aristocracy” or rather an upper class (Oberschicht); cf. the controversial
statement of H. Grahn-Hoek, Die fränkische Oberschicht im 6. Jahrhundert. Studien zu
ihrer rechtlichen und politischen Stellung, Vorträge und Forschungen Sonderband 21 (Sigma-
ringen 1976).
82
Cf. H. Ament, “Archaeology”, id., “Franken”, pp. 400–1. For separate burial
grounds of a Germanic leading class, cf. H.W. Böhme, “Adelsgräber im Frankenreich.
Archäologische Zeugnisse zur Herausbildung einer Herrenschicht unter den mero-
   () 329

senatorial aristocracy, there was still a difference between the Roman

aristocracy and the Frankish leaders in Gregory’s time, and that
Gregory had a clear notion of these differences.83 It is significant,
though, that authors, such as Gregory of Tours and Fredegar, appar-
ently did not perceive or certainly did not emphasize an (ethnic)
contrast between the “Roman” and the “Germanic” population within
the Frankish kingdom.84
In his contemporary letter of congratulation on Clovis’s baptism,85
Avitus, the metropolitan bishop of Vienne, still sensed an “ethnic”
difference between Romans and Franks addressing the latter as ves-
tra gens!86 A century and a half later, such distinctions and similar
evidence of “ethnic” antagonism seem to have lost their impact, with
both groups becoming more and more united, a fact which is also
reflected in the process of namegiving: Whereas Roman and Ger-
manic persons’ names (anthroponyms) are regarded to have initially
been an indicator of the “ethnic” origin of men and women, from
the seventh century onwards it became increasingly difficult to infer
one’s descent in this way since Romans adopted Germanic names
(and vice versa).87 (It would be interesting to investigate this sym-
biosis further by research on changes in the naming of the Franks.)88
Another element of integration was the administration of the king-
dom. Although there were differences between the northern and
southern parts of the realm, these sectors seem to have been fairly
homogeneous as far as government and administration are concerned.

wingischen Königen”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 40 (1993)

pp. 397–534.
83
Cf. Irsigler, Untersuchungen zur Geschichte des frühfränkischen Adels, pp. 84 ff.
84
Cf. H.-W. Goetz, “Die germanisch-romanische (Kultur-)Synthese in der Wahrneh-
mung der merowingischen Geschichtsschreibung”, Akkulturation. Probleme einer german-
isch-romanischen Kultursynthese in Spätantike und frühem Mittelalter, ed. D. Hägermann,
W. Haubrichs and J. Jarnut (forthcoming).
85
See n. 97 below.
86
When later on, however, Clovis, in his letter to the bishops from 507/511
(MGH Capit. 1, no. 1, ed. A. Boretius [Hannover 1883] pp. 1–2), speaks of popu-
lus noster, he may have been thinking of the Franks, but it is more probable that
he referred to all his people as a political body, presumably including the Romans.
87
Cf. H. Ebling, Prosopographie der Amtsträger des Merowingerreiches. Von Chlothar II.
(613) bis Karl Martell (714), Beihefte der Francia 2 (München 1974); id., J. Jarnut
and G. Kampers, “Nomen et gens. Untersuchungen zu den Führungsschichten des
Franken-, Langobarden- und Westgotenreiches im 6. und 7. Jahrhundert”, Francia
8 (1980) pp. 687–745.
88
This is one of the tasks of the interdisciplinary workgroup “Nomen et gens”.
Cf. n. 38 above.
330 - 

If we may, for example, at all assume different origins for the offices
of a Germanic grafio and a Gallo-Roman comes,89 it seems to be impos-
sible to perceive differences between these offices in Merovingian
times.90 Thus, on the whole, though there may have been both
“Germanic” and/or Roman roots for the single administrative offices,
the strongest impression that we get from the sources is their struc-
tural unity, so that the administration of the Frankish kingdom must
be considered an important element of integration. It should not,
however, be forgotten that the tendency towards a certain indepen-
dence of individual officials, particularly the duces of the seventh and
eighth centuries, favoured the disintegration of certain parts of the
realm later on. On the whole, “Clovis’s kingdom from the begin-
ning experienced a much more thorough mixture of Frankish and
Roman traditions”, as Patrick Geary concludes,91 and what Walter
Pohl infers rightly from a comparison of the Germanic states is true
for the Frankish kingdom as well: “a clear distinction between Roman
and Germanic origins [. . .] would not help to understand a process
in which there was a continuum of solutions to problems that were
common to ‘Romans’ and ‘barbarians’, who were becoming harder
and harder to distinguish. The states were both Roman and bar-
barian, and so, in a sense, were most of their leading members.”92
The strongest factor of integration, however, was the “Catholic”
church because, after the baptism of Clovis—the long discussion
about the exact date need not concern us here—,93 Romans and
Franks began to experience a common religious unity. The epochal
importance of this event has always been acknowledged and em-
phasized,94 though Christianizing the Franks actually turned out to

89
For such a difference, cf. still T. Bauer, “Graf/Grafio (Historisches)”, Real-
lexikon der Germanischen Altertumskunde 12 (2nd edn., 1999) pp. 532–55, particularly
p. 540.
90
Cf. strictly in this view: A.C. Murray, “The Position of the Grafio in the Cons-
titutional History of Merovingian Gaul”, Speculum 61 (1986) pp. 787–805.
91
Geary, Before France and Germany, p. 89.
92
W. Pohl, “The Barbarian Successor States”, The Transformation of the Roman
World A.D. 400–900, ed. L. Webster and M. Brown (London 1997) pp. 33–47, here
p. 45.
93
For Clovis’s baptism cf. now A. Dierkens, “Die Taufe Chlodwigs”, Die Franken—
Wegbereiter Europas 1, pp. 183–91. For the liturgy of baptism, see V. Saxer, “Les
rites du baptême de Clovis dans le cadre de la pratique paléochrétienne”, Clovis 1,
pp. 229–41; for its political consequences, see A. Angenendt, “Le parrainage dans
le haut Moyen Âge. Du rituel liturgique au cérémonial politique”, ibid., pp. 243–54.
94
Cf. Geuenich, “Chlodwigs Alemannenschlacht(en) und Taufe”; M. Rouche,
   () 331

be a long process, and it was only gradually that stricter measures

against heathens were taken and began to bear fruit. Again, the bap-
tism of Clovis had been prepared not only by his wife Chrodechilde,
as Gregory wants us to believe,95 but also by his Roman-Catholic
environment, including the bishops, and we may recall the fact that
Bishop Remigius of Rheims congratulated Clovis on his succession
to the “throne” long before his conversion.96 Once Clovis and his
followers were converted, Christianization promoted a symbiosis of
Romans and Franks, there being, at least officially, no “religious
dualism” between “Catholics” and “Arians” in the Frankish king-
dom, as existed in the other “Germanic” realms of that epoch, but
a confessional unity. The letter of congratulation written by Bishop
Avitus of Vienne after Clovis’s conversion97 reveals that religious
unity at least facilitated royal government. According to Avitus, Clovis
had not really gained a new function—as stated in Remigius’s let-
ter on the occasion of Clovis’s succession to the “throne”, there had
been, or should have been, cooperation between the king and the
bishops even before his conversion—, but a (religious) strengthening
of his kingship: Now Clovis would not only reign in saeculo, but also
in caelo; and his rigor armorum would be strengthened by his belief
because God had chosen the Franks to be “his people”, thus giving
them the task to Christianize those foreign peoples who were still
living in ignorance (the externi populi paganorum). In 511, Clovis conse-
quently took over responsibility for the church by commanding the
bishops of his entire kingdom to gather for the council of Orleans,
and the bishops themselves conceded that the king’s consent granted
a higher authority to the sentences of the priests.98 The Frankish
king was always “lord” of the church and of the bishops, who were
not to be consecrated without his consent (the royal “decree of

“Die Bedeutung der Taufe Chlodwigs”, Die Franken—Wegbereiter Europas 1, pp. 192–9;
B. Fauvarque, “Le baptême de Clovis, ouverture du millénaire des saints”, Clovis
1, pp. 271–86; F. Monfrin, “La conversion du roi et des siens”, ibid., pp. 289–320.
95
Cf. C. Nolte, Conversio und Christianitas. Frauen in der Christianisierung vom 5. bis 8.
Jahrhundert, Monographien zur Geschichte des Mittelalters 41 (Stuttgart 1995) pp.
72–86.
96
Epistolae Austrasicae 2, ed. W. Gundlach, MGH EE 3 (Berlin 1892) p. 113.
97
Avitus of Vienne, Epistolae 46, ed. R. Peiper, MGH AA 6,2 (München 1883)
pp. 75–6.
98
Letter of the bishops to Clovis: MGH Conc. 1, ed. F. Maassen (Hannover
1893) pp. 2–3. For the council of Orleans, see J. Heuclin, “Le concile d’Orléans
de 511, un premier concordat?”, Clovis 1, pp. 435–50.
332 - 

consecration”).99 According to Gregory of Tours, he alone, in his

defence of one of his colleagues, who was accused before King
Chilperic, dared to criticize the king. But at the same time the author
admits that his warning was a moral one because he had no legal
right to officially disagree with the king.100 Thus, as in this case, we
find at least some evidence in Gregory’s statement that the bishops
of the whole Frankish realm stuck together over certain issues, but
far more often we get the impression that they were primarily loyal
to their respective king and consequently antagonistic towards “the
church” of other, hostile Merovingian kingdoms.

4. Roman elements/“Romanization”: We must not neglect the fact (as

has been done in the past by German historians, but is now being
increasingly emphasized again) that, although the Franks were the
leading group in the Frankish kingdom, not only the Christian reli-
gion, but also decisive parts of the government had been adopted
(with adaptations) from the Romans. The (later) kings conferred upon
themselves the title of rex Francorum, but at the same time they adopted
Roman epithets (such as gloriosissimus etc.). Accordingly, Clovis’s reign
was widely based on Roman elements:101 a Roman administration,102
Roman (or Gaulish) administrational territories (civitates and pagi ),103
imperial demesnes, the fisc, the monetary system, charters, written
laws and last but not least the continuity of Roman legal practi-
ces.104 Finally, it should not be forgotten either that the church, too,

99
Cf. D. Claude, “Die Bestellung der Bischöfe im merowingischen Reiche”,
Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Kanonistische Abteilung 49 (1963) pp. 1–75;
C. Servatius, “‘Per ordinationem principis ordinetur’. Zum Modus der Bischofs-
ernennung im Edikt Chlothars II. vom Jahre 614”, Zeitschrift für Kirchengeschichte 84
(1973) pp. 1–29.
100
Gregory of Tours, Historiae 5,18, pp. 219–20, for the case of Bishop Prae-
textatus of Rouen.
101
Cf. K.F. Werner, “Die ‘Franken’. Staat oder Volk?”, Die Franken und die Ale-
mannen, pp. 95–101; E. Ewig, “Das Fortleben römischer Institutionen in Gallien und
Germanien”, id., Spätantikes und fränkisches Gallien 1, pp. 409–34; I.N. Wood, “Die
Franken und ihr Erbe—‘Translatio Imperii’”, Die Franken—Wegbereiter Europas 1, pp.
358–64.
102
Cf. U. Nonn, “Zur Verwaltungsorganisation in der nördlichen Galloromania”,
Die Franken und die Alemannen, pp. 82–94.
103
Cf. S.T. Loseby, “Gregory’s cities: Urban functions in sixth-century Gaul”,
Franks and Alamanni, pp. 239–70; 270–84.
104
For this aspect, cf. now S. Esders, Römische Rechtstradition und merowingisches
Königtum. Zum Rechtscharakter politischer Herrschaft in Burgund im 6. und 7. Jahrhundert,
   () 333

was Roman in its character and administration (and contrary to for-

mer beliefs, as an element of continuity, it probably did not even
prompt a complete topographical change in the towns).105 In the
sixth century, the vast majority of the bishops were still Roman,106
and it was only in the course of time that Frankish noblemen became
bishops. Contrary to former opinions, therefore, the political system
of the Franks (as of all the “Germanic” kingdoms) seemed decisively
characterized by political continuity. Clovis, undoubtedly, was leader
of his people, but at the same time this was a federate people that
had already been integrated into the Roman Empire for a long time,
and he governed his realm by measures which he took over from
his Roman predecessors. This continuity derived from the integra-
tion of the barbarians into the Empire; in fact, it was only this
assimilation that enabled the barbarian leaders to rule parts of it.
The continuity of Roman elements, however, did not mean that
there was no change. On the contrary, it was a characteristic fea-
ture that those Roman elements on which the Frankish government
was based were transformed gradually, though we can only get a
sketchy impression of these changes through a comparison with later
situations. While the legal tradition, for example, followed the Roman
model, it was not possible to adapt the “Germanic” legal tradition
completely to this model; it was not even possible to translate all
the legal expressions into Latin so that the Lex Salica abounds with

Veröffentlichungen des Max-Planck-Instituts für Geschichte 134 (Göttingen 1997),

who shows, with regard to the Praeceptum of Chlothar II of 613, that the surviv-
ing Roman legal traditions in Burgundy not only determined human life, but were
acknowledged by the kings. For the Roman origin of the Pactus legis Salicae, see
É. Magnou-Nortier, “Remarques sur la genèse du Pactus legis Salicae et sur le privilège
d’immunité (IVe–VIIe siècles)”, Clovis 1, pp. 495–538.
105
Cf. N. Gauthier, “La topographie chrétienne entre idéologie et pragmatisme”,
The Idea and Ideal of the Town Between Late Antiquity and the Early Middle Ages, ed. G.P.
Brogiolo and B. Ward-Perkins, The Transformation of the Roman World 4 (Leiden-
Boston-Köln 1999) pp. 195–209. The same is valid for funerals and burial grounds
until the seventh century, except for the funerary basilicas over the graves of mar-
tyrs; it was not until the eighth century that cemeteries were moved from the periph-
ery and graves were regrouped around the churches; cf. A. Dierkens and P. Périn,
“Death and burial in Gaul and Germania, 4th-8th century”, The Transformation of
the Roman World A.D. 400–900, ed. L. Webster and M. Brown (London 1997) pp.
79–95, here pp. 94–5.
106
For the “image” of the bishops of Aquitania, as reflected in their funeral
epitaphs, see M. Heinzelmann, Bischofsherrschaft in Gallien. Zur Kontinuität römischer
Führungsschichten vom 4. bis zum 7. Jahrhundert. Soziale, prosopographische und bildungs-
geschichtliche Aspekte, Beihefte der Francia 5 (Zürich-München 1976).
334 - 

Germanic terms, the so-called Malberg glosses.107 Again, also in the

field of government, law, and administration, we find a mixture of
Roman and “Germanic” elements which soon developed into an
integrated whole. “Germanic” elements may have been stronger in
the “inner circle” of the “royal court”, and the Franks formed the
core of the royal army. As Patrick Geary puts it bluntly: the army
was Frankish, the culture was Roman(ized).108 On the whole, how-
ever, the Franks were as much (or even more) “Romanized” than
the Romans were “Germanized”, or, as Karl Ferdinand Werner
remarks: “Not the Romans who kept their Church and their law,
but the Franks experienced since 500 a revolution of their way of
life: the submission under the Church and the princeps-dominus, to
whom they owed obsequium in their militia.”109 Therefore, if it was
mainly the Franks who were subject to changes in the Frankish king-
dom, this should have affected the concept of the gens Francorum.

5. Concept and Development of the Frankish “people”: As a further approach

we should, therefore, investigate the role and meaning of the Frank-
ish people inside this symbiotic kingdom.110 “‘The Franks’ is a term
that has no absolute meaning, but one which changes according to
time or circumstance.”111 In Merovingian times, no doubt, in spite

107
Cf. R. Schmidt-Wiegand, Stammesrecht und Volkssprache. Ausgewählte Aufsätze zu
den Leges barbarorum (Weinheim 1991); ead., “Rechtsvorstellungen bei den Franken
und Alemannen vor 500”, Die Franken und die Alemannen, pp. 545–57, again tries to
trace back certain legal practices of the Franks to the time before the first codification
of the Lex Salica.
108
Geary, Before France and Germany, p. 93.
109
Werner, “Die ‘Franken’”, p. 100.
110
Some enlightening remarks on a “gentile” consciousness of the Franks as a
gens are made by I. Haselbach, Aufstieg und Herrschaft der Karlinger in der Darstellung der
sogenannten Annales Mettenses priores. Ein Beitrag zur Geschichte der politischen Ideen im Reiche
Karls des Großen, Historische Studien 412 (Lübeck-Hamburg 1970) pp. 133–7, in
regard to the early Carolingian Annales Mettenses priores. Cf. now H.-W. Goetz, “Zur
Wandlung des Frankennamens im Frühmittelalter”, Integration und Herrschaft. Ethnische
Identitäten und soziale Organisation im Frühmittelalter, ed. M. Diesenberger and W. Pohl
(Vienna 2002) pp. 133–50); id., “Gens. Terminology and Perception of the Germanic
Peoples from Late Antiquity to the Early Middle Ages”, The Construction of Communities
in the Early Middle Ages, ed. R. Corradini, M. Diesenberger and H. Reimitz (forth-
coming).
111
Thus James, The Franks, p. 9. For the discrepancy between the modern and
early medieval understanding of a gens and the changing of peoples’ names, cf. M.
Springer, “Geschichtsbilder, Urteile und Vorurteile. Franken und Sachsen in den
Vorstellungen unserer Zeit und in der Vergangenheit”, 799—Kunst und Kultur der
Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn. Beiträge zum Katalog der
Ausstellung Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz 1999) pp.
224–32.
   () 335

of the widespread “Romanization” of government, language, society,

learning and culture, when one considers the role of the Franks in
this newly established, vast kingdom, they were (and remained) the
politically leading people in this realm, to a degree that for the
authors of our chronicles it was “the” Franks who acted and made
decisions in all important political questions. It is mainly in this con-
text that “Franks” were mentioned in our sources (whereas only in
a few cases was a man defined as being a Frank by birth).112 Although,
according to corresponding formulas, a consensus fidelium as the right
of an aristocracy to participate in the king’s policy may have been
a development of Carolingian rather than of Merovingian times,113
there can be no doubt about the participation of “Frankish” freemen
(or rather the leading groups among them—who again may not have
necessarily been “Franks” in the narrower ethnic sense). According
to the chronicles, it was the Franci who elected the king or met in
assemblies or courts, who concluded peace treaties or contracts114 or
who went to war: The Franks were considered as the politically
decisive group in the kingdom, together with the king. Fredegar,
when speaking of certain events, has a liking for the phrase “the
king and his Franks”.115 Accordingly, the Franks were distinguished
from the peoples outside the Frankish realm—Gregory mentions wars
and peace treaties, interventions, legations and weddings between
“Franks” and members of “foreign” peoples—, whereas they were
hardly ever distinguished from other peoples inside this realm. For
Gregory116 and Fredegar,117 the Franks were distinguishable from Romans,

112
The terms Francus genere or Francus natione are only once used in Gregory of
Tours’ Histories (Historiae 10,2, p. 482), but three times in Fredegar’s chronicle
(Chronicon 4,18, p. 128; 4,24, p. 130; 4,34, p. 133, ed. B. Krusch, MGH SSrM 2
[Hannover 1888]).
113
Cf. J. Hannig, Consensus fidelium. Frühfeudale Interpretationen des Verhältnisses von
Königtum und Adel am Beispiel des Frankenreiches, Monographien zur Geschichte des
Mittelalters 27 (Stuttgart 1982).
114
According to Fredegar, Chronicon 3,19, p. 100, for example, Gundobad, the
king of the Burgundians, made peace with “the Franks”.
115
Cf. for example Fredegar, Chronicon 2,58, p. 83: Chlodovei regis et Francis; 3,16,
p. 99; 3,30, p. 103; 4,71, p. 156; 3,21, p. 101: cum Francis meis.
116
Goths: Gregory of Tours, Historiae 2,7, p. 50; 2,18, p. 65; 10,31, p. 531; Romans:
2,9, p. 53; 2,18, p. 65; 2,19, p. 65; 10,31, p. 526; Alamanni: 2,9, p. 56; Burgundians:
2,9, p. 56; 2,23, p. 69; 3,6, p. 103; Saxons: 2,19, p. 65; 4,14, pp. 145–6; 4,16, pp.
149–50; Thuringians: 3,7, pp. 103 ff.; Bretons: 4,4, p. 137.
117
Goths: Fredegar, Chronicon 2,58, p. 82, in Alaric’s war against Clovis; 3,12,
p. 98: war of the Romani et Franci against the Goths; Alamanni: 3,21, p. 101: the Alamans
did not find a people ( gens), that would have helped them against the Franks;
336 - 

Goths, Burgundians, Alamans, Saxons, Thuringians and Bretons.

All this may indicate that the Franks were regarded as the lead-
ing class inside a kingdom which was accordingly considered to
be and was named “Frankish”: Franci corresponded with the regnum
of the Merovingians, or with one part of the divided kingdom.118
When Chlothar and Childebert made war against the Burgundians,
the Franks who were subject to Theudebert (Franci vero, qui ad eum
aspiciebant), compelled their king to join his brothers.119 “Franks” lived
in each of the divided kingdoms which all remained and were per-
ceived as regna Francorum. Consequently, Franci increasingly became
an expression used for the whole population of the Frankish realm.
The exercitus Francorum was collected from the whole kingdom.120
Therefore, “gens, the terminology of space and royal territory were
never approximately identical.” The gens Francorum was at the same
time “gentile” and “supragentile” and “made other gentes appear to
be their part and their counterpart at the same time”.121
Regarding the terminology and the usage of Franci and Francia,
therefore, it seems that in the end, that is already by the seventh

Burgundians: 3,35–36, p. 104; Saxons: 2,6, p. 46; 2,45, p. 68; 3,51, p. 107; Bretons:
4,11, p. 127.
118
The ambiguity of the term “Frankish” is now strongly emphasized by Pohl,
“Zur Bedeutung ethnischer Unterscheidungen”, esp. pp. 199 ff. Cf. also Goetz, “Zur
Wandlung des Frankennamens”. The evidence for the use of Franci and Francia
already in E. Ewig, “Volkstum und Volksbewußtsein im Frankenreich des 7. Jahr-
hunderts”, id., Spätantikes und fränkisches Gallien 1, pp. 231–73, esp. pp. 259–70. For
the later usage of political terms, see id., “Beobachtungen zur politisch-geographi-
schen Terminologie des Fränkischen Großreiches und der Teilreiche des 9. Jahrhun-
derts”, ibid., pp. 323–61. For Gregory of Tours, cf. also E. James, “Gregory of
Tours and the Franks”, After Rome’s Fall. Narrators and Sources of Early Medieval History.
Essays presented to Walter Goffart, ed. A.C. Murray (Toronto-Buffalo-London 1998) pp.
51–66.
119
Gregory of Tours, Historiae 3,11, p. 107. Cf. ibid., 4,51, p. 188: The Franks
who once had obeyed the older Childebert now sent legations to Sigibert.
120
Cf. for example Fredegar, Chronicon 4,73, p. 158, regarding a campaign against
Spain. According to another report, however, King Dagobert levied troops from all
over the regnum Burgundiae for a campaign against the Basks. The leaders here were
clearly distinguished according to their descent: eight were ex genere Francorum, one
was Roman, one Burgundian, one Saxon.
121
Thus Pohl, “Zur Bedeutung ethnischer Unterscheidungen”, p. 205. H. Klein-
schmidt, “The Geuissae and Bede: On the Innovations of Bede’s Concept of the
Gens”, The Community, the Family and the Saint. Patterns of Power in Early Medieval Europe,
Selected Proceedings of the International Medieval Congress. University of Leeds, 4–7 July 1994,
10 –13 July 1995, ed. J. Hill and M. Swan, International Medieval Research 4
(Turnhout 1998) pp. 77–102, claims that the political concept of a gens was a sec-
ondary, post-migrational one.
   () 337

century, “Frankish” had lost much of its former ethnic connotation

and referred to legal, cultural and not least to political dimensions,
namely to the very kingdom of the Franks:122 Franci expanded and
developed from the people(s) on the northern margins of the former
Roman Empire to the inhabitants of the Frankish kingdom of the
Merovingians,123 though it was not until the ninth century that these
meanings became prevalent.124 Due to the growth of the Merovin-
gian realm, its name-giving people, the Franks, gained a new mean-
ing. This may be considered as an indication of a shift from a more
or less ethnic to a political association. It may also be a sign of
successful integration of peoples into the Frankish realm or, in this
respect, even of a new phase in the ethnogenesis of these peoples
under “Frankish” influence.125 If the last assumption is true, by look-
ing at the terminology of the Franci, we can detect here not only a
decisive shift of the usage of terms, but also of the everchanging
ethnogenetical process of the formation and re-formation of “the”
Franks.
It is not by chance that a territorial term Francia was, sometimes
but not frequently, used parallel to Franci from which it was derived.
Like Franci, it had a double meaning, referring to the region of the
Franks (as opposed, for example, to Burgundy or Aquitania) as well
as to the whole kingdom. In both respects, it is evidence of an increa-
sing territorial understanding, an interpretation that is supported by
frequent expressions such as fines, confinia, limes, termini, terra, regio or
partes Francorum. However, whereas in Merovingian times the “Franks”
were most frequently associated with the whole kingdom, in the
course of the ninth century they tended to become more and more
restricted to the “Frankish” territories inside the realm, namely to
Francia around the Ile-de-France in the west (that is, a territory which

122
Cf. P.J. Fouracre, “The Nature of Frankish Political Institutions in the Sev-
enth Century”, Franks and Alamanni, pp. 285–301; 301–16. Ibid., p. 297: “but what
being a Frank amounted to in the seventh century, apart from claiming certain
legal privileges, we cannot tell. The term ‘Frankish’ has a much wider application
as the adjective derived from the kingdom of the Franks, the inhabitants of which
were, of course, mostly non-Franks. It is not, therefore, a term of ethnic designa-
tion, but covers the plurality of laws and customs we have been discussing [. . .]”
123
Ibid., p. 298.
124
For the evidence from chosen authors, see Goetz, “Zur Wandlung des Fran-
kennamens”.
125
Cf. H.J. Hummer, “Franks and Alamanni: a discontinuous ethnogenesis”, Franks
and Alamanni, pp. 9–21; 21–32, here p. 32.
338 - 

was originally Roman[!], but the central region of the Merovingian

kingdom) and, in the east, to Franconia around the river Main region
(a territory that had originally been partly Alamannic, partly Thu-
ringian or even Slavic!): By then, as it seems, the term had adopted
a completely altered meaning; it had been totally separated from its
“ethnic” origins and assumed a predominantly territorial connota-
tion! Thus, in the course of the four centuries from 500 to 900, the
terms Franci and Francia had changed their meanings twice: first from
“people” to kingdom, and afterwards to regional territories.

6. “Gentile” or political consciousness? Finally, one of the most difficult

problems of gaining some insight into Frankish “ethnicity” is the
question of the existence of a conscious ethnic identity (Stammesbe-
wußtsein) as an important element of ethnogenesis, of its possible
changes after the establishment of the Frankish kingdom and of its
relation to the structural development as well as to the terminology
dealt with above. This aspect has been emphatically reiterated by
Reinhard Wenskus and Herwig Wolfram, but questioned recently by
Patrick Amory with regard to the Ostrogoths.126 Since we cannot
presume or discover anything like an ethnic identity before Clovis
(or Gregory) beyond the term Franci, we consequently are not able
to point to explicit changes, but have to be content with searching
for indications of a consciousness of “ethnic” identity in the course
of Frankish history and—this is the decisive point—of its relation to
a consciousness of belonging to the same “nation” or rather the same
“kingdom”. Although the Franks were perceived as a gens through-
out the period considered here127 (though we have to admit that gens
Francorum is fairly seldom found in the sources),128 the terminology

126
P. Amory, People and Identity in Ostrogothic Italy, 489–554, Cambridge Studies
in Medieval Life and Thought (Cambridge 1997).
127
Cf. for example Gregory of Tours, Historiae 2,12, p. 61; 6,2, p. 266; Fredegar,
Chronicon 2,6, p. 46 (nulla gens was able to overwhelm the Franks); 3,9, pp. 94–5;
3,11, p. 96: Modo est gens Francorum tuae disciplinae perdomita (the Franks had sub-
mitted themselves under Aegidius’ reign). Cf. Goetz, “Zur Wandlung des Franken-
namens”.
128
Gregory and Fredegar do not use the term gens Francorum too frequently (see
note 62); more often, Fredegar’s successors speak of a gens Francorum. Later on, cf.
also Annales regni Francorum ad a. 787, p. 76; ad a. 811, p. 134, ed. F. Kurze, MGH
SSrG 6 (Hannover 1895); Annales Fuldenses ad a. 841, p. 32; ad a. 873, p. 79; ad
a. 884, p. 101, ed. F. Kurze, MGH SSrG 7 (Hannover 1891). According to the
wording, the authors sometimes use just the term gens after having related an action
   () 339

favours a political rather than an ethnic consciousness since the term,

as has already been emphasized above, is much more frequently
used with reference to the Frankish kingdom than to the people.
Nevertheless, there are at least a few indications of a “gentile con-
sciousness”. According to the famous (later) prologue to the Pactus
legis Salicae probably dating from the late seventh or early eighth cen-
tury, the Franks and their magnates ( proceres) had come together to
warrant the peace by preserving the laws.129 This may be under-
stood as proof of the existence of a Frankish sense of identity and
a consciousness of their eminent status among (and above) “the other
peoples” (ceteris gentibus). It is true, that in this (legal) context, the
Franks were no more than just one people among others in the king-
dom, but, in the concept they had of themselves, they excelled all
the others by their strength (brachium fortitudinis) and their legal author-
ity (auctoritas legalis).
It is the so-called Origines gentium which are usually considered as
the clearest indicator of a “gentile” consciousness.130 Thus, the famous
Trojan myth, as it was passed on by Fredegar (with Priam as the first
king of the Franks and, after several divisions of this people, with
Francio as the first and namegiving king of a branch that wandered
from Phrygia to Europe, namely to the region between the rivers
Rhine and Danube), is an important factor indeed. It is significant,
however, that whilst it must be granted that this legend was handed
down at a fairly late period, there must have been forerunners reach-
ing back as far as the Late Roman times of Ammianus Marcellinus.
The latter reports that the inhabitants of the region near the mouth
of the river Rhine (he assumed that they were Celts) may have been
indigenous to that area, but according to some writers they had
come there as dispersed troops after the destruction of Troy.131 It is

of “the Franks” immediately before. Cf., for example, Gregory of Tours, Historiae 2,9,
pp. 56–8; 8,16, p. 383 (in gente sua); Fredegar, Chronicon 3,11, p. 96; 3,22, p. 102;
4,38, p. 139; Regino of Prüm, Chronicon ad a. 880, ed. F. Kurze, MGH SSrG 50
(Hannover 1890) p. 116.
129
Lex Salica prol., ed. K.A. Eckhardt, MGH LL nationum Germanicarum 4,2
(Hannover 1969) pp. 2 ff. Cf. R. Schmidt-Wiegand, “‘Gens Francorum inclita’. Zu
Gestalt und Inhalt des längeren Prologes der Lex Salica”, Festschrift Adolf Hofmeister
zum 70. Geburtstag, ed. U. Scheil (Halle 1955) pp. 233–50.
130
Cf. H. Wolfram, “Origo et religio. Ethnic traditions and literature in early medie-
val texts”, Early Medieval Europe 3 (1994) pp. 19–38.
131
Amm. Marc. 15,9,3–5, vol. 1, p. 61.
340 - 

even more significant that there were several myths concerning the
origin of the Franks—Gregory of Tours, for example, favoured their
arrival from Pannonia132—and, even more remarkable, we find two
different versions in Fredegar himself 133 which later on were both
adapted and somehow “combined” by the Liber historiae Francorum:134
The Franks were obviously trying to acquire a conscious identity
derived from a concept of common descent and origin. They did
not, however, have only one unique belief in a common origin, but
were still searching for it. Moreover, the Origines gentium were obvi-
ously more or less scholarly constructions. What is even more inter-
esting: nothing could display more clearly than this Trojan myth
how far from being “ethnic” in our sense “ethnic thinking” was in
those times: The myth proves the importance of migrations as well
as of remote origins for the early medieval mind, the importance of
kings (from the very beginning), the enormous relevance of belief in
a common descent (Herkunftsbewußtsein) and the emphasis on the equal-
ity of the Franks with the Romans, both stemming from Troy—
according to Fredegar, who explicitly refers to Vergil,135 from the
abolishment of Roman subjugation under Pompeius up to the pre-
sent day, nulla gens could overwhelm the Franks.136 At the same time,
however, it shows the complete irrelevance of a “Germanic” de-
scent.137 It is no less significant that all these Frankish Origines gentis

132
Gregory of Tours, Historiae 2,9, pp. 52–8, esp. p. 57: Tradunt enim multi, eos-
dem de Pannonia fuisse degressus.
133
Fredegar, Chronicon 2,4–6, pp. 45–6; 3,2, p. 93.
134
Liber historiae Francorum 1 and 4, ed. B. Krusch, MGH SSrM 2 (Hannover
1888) pp. 241–4: Under the sons of Priam (Marcomir) and Antenor (Sunno), the
Franks went from Pannonia to Sicambria and further to the cities of Germania in
the Lower Rhine region where they elected Faramund, the son of Marcomir, as
their (long-haired) king. For these different versions, see now E. Ewig, “Troiamythos
und fränkische Frühgeschichte”, Die Franken und die Alemannen, pp. 1–30, and H.H.
Anton, “Troja-Herkunft, origo gentis und frühe Verfaßtheit der Franken in der gal-
lisch-fränkischen Tradition des 5. bis 8. Jahrhunderts”, Mitteilungen des Instituts für
Österreichische Geschichtsforschung 108 (2000) pp. 1–30; cf. also J. Barlow, “Gregory of
Tours and the Myth of the Trojan Origins of the Franks”, Frühmittelalterliche Studien
29 (1995) pp. 86–95. For the Liber historiae Francorum, see R.A. Gerberding, The Rise
of the Carolingians and the Liber Historiae Francorum, Oxford Historical Monographs
(Oxford 1987) pp. 11–30.
135
Fredegar, Chronicon 3,2, p. 93.
136
Ibid., 2,6, p. 46.
137
Cf. also I.N. Wood, “Defining the Franks: Frankish Origins in Early Medieval
Historiography”, Concepts of National Identity in the Middle Ages, ed. S. Forde, L. Johnson
and A.V. Murray, Leeds Texts and Monographs. New Series 14 (Leeds 1995)
   () 341

presuppose a strong relationship with the origins of kingship, as it

is expressively said in the Liber historiae Francorum: Principium regum
Francorum eorumque origine vel gentium illarum ac gesta proferamus.138 The
consciousness of an “ethnic” identity was at the same time a con-
sciousness of a “royal” identity linking the Franks and their “long-
haired kings”.139 Again, it is not easy, and would contradict early
medieval thinking, to distinguish between the “people” and the kings
(or kingdoms) of the Franks.
When asking, therefore, what may be called “Frankish” in Frankish
“national” identity and in Frankish political consciousness (Reichs-
bewußtsein), one finds that this phenomenon tends to be much more
a collective identity of (all) the gentes in the Frankish kingdom, under
the rulership of a name-giving gens, rather than a collective identity
of this gens. It increasingly included all the realm’s peoples, as far as
their populus, the politically influential classes, were concerned. This
identity, however, concentrated on the royal dynasty. In some sources,
at least, such as the early Carolingian Annales Mettenses priores, rex, gens
and terra formed an indivisible unity.140

IV. Conclusion

When summarizing the elements that characterize the relationship

between the Frankish “people” and the Frankish kingdom, and sug-
gesting some answers to the leading questions of this volume, we
may tentatively come to the conclusion that the Frankish ethno-
genesis was a long and continuous process of almost imperceptible
developments, but that its progression was also subject to occasional
caesuras (for example Clovis’s reign). Before Clovis, there had been
kings, but neither great kingdoms nor a clearly definable, homogeneous

pp. 47–57, here pp. 50 ff. Whereas the former research emphasized the origo’s func-
tion of equalizing Franks and Romans, Wood regards it as an instrument to ex-
plain and strengthen the existing ties between these groups in Gaul by believing in
the theory of a common descent.
138
Liber historiae Francorum 1, p. 241.
139
Due to Gregory of Tours’ inability to discern the early kings of the Franks
who were initially governed by duces (cf. note 27 above), in his opinion the Frank-
ish kingship was bound to the “long-haired” kings, the Merovingians, from the
beginning.
140
Thus Haselbach, Aufstieg und Herrschaft, p. 137.
342 - 

people. Nevertheless there had been a gens Francorum before Clovis

which had already been ruled by kings, though completely un-
united in a political sense. It seems comparatively clear, therefore,
that, on the one hand, the formation of the Frankish kingdom
was not the achievement of an “established” gens, but of its Mero-
vingian kings, that is, of Clovis and his successors. In the rela-
tionship of gens and regnum in the process of the development of a
Frankish kingdom the central impact was a political one. “Kingdom”,
therefore, need not be conceived in ethnic terms141 whereas, after
Clovis, the “Franks” were perceived as a political group. On the
other hand, although the formation of the Frankish kingdom brought
about enormous changes, it did not immediately constitute a new
“people”. At least, we lack sufficient indications that the symbiosis
of Romans and Franks created a new “people”, or that it “trans-
formed” the Frankish people (or the various peoples under Frankish
rule) into a new gens (at least not in the centuries considered here).
But it obviously altered the character and gradually changed the
meaning of being a Frank which increasingly denoted the members
of the Frankish kingdom disregarding their “ethnic” origin (and it is
almost exclusively this connotation which enables us to grasp these
changes). Clovis, therefore, did not “create” the Frankish people, but
he laid the foundation for a new, and different, ethnogenesis, as it
was only through the expansion and unification of a “Frankish” king-
dom that the gens (and therefore also that ethnogenesis) acquired a
new dimension which was mainly political. After Clovis, there was
a united kingdom (notwithstanding the partitions), with a consider-
able amount of integration, which survived the partitioning, although
its borders included far more than the (former) Frankish people. The
question, therefore, is neither how Clovis’s reign changed the Frankish
kingdom nor if it was only under his reign that the Frankish peo-
ple came into existence, but, rather: what part this people played in
the new realm, and, above all, what effect this political alteration
had on the Frankish people who henceforth formed only (one) part
of this kingdom. It remains extremely difficult to answer this ques-
tion. It seems, however, that by royal power, with the support of
other integrative forces (such as the church, Roman heritage, and a
uniform administration), the many peoples in this realm, and not

141
Thus Beisel, Studien zu den fränkisch-römischen Beziehungen, p. 49.
   () 343

least the Germanic and the Roman population, “grew together”

within a comparatively short period of about one century into one
integrative and syncretistic (political) body, in a process of consi-
derable “ethnic syncretism”. In addition they all increasingly con-
sidered themselves as being “Franks”, a term that shifted in its
significance from the members of a gens to the members or popula-
tion of the realm. The kings obviously had a most decisive role in
establishing this realm. Whereas personally the Roman—or “Greek”—
Emperor had no part in it at all, it seems to have been crucial for
the new kingdom that a considerable proportion of the Roman
structures survived. The Frankish kings’ power was absolutely equal
to that of the emperors, whereas Rome increasingly turned into a
Christian myth.142 We may also recall that kingship as such, as well
as the Merovingian dynasty as stirps regia, was completely undisputed
until the eighth century. And yet at the same time, and significantly
enough, through the constant divisions of this realm, under the
Merovingians as well as under the Carolingians, “Frankish” unity
began to dissolve. This created first, new regna (or, in some cases,
given certain modifications, re-established older ones). Similarly, by
and by, this process created new peoples within the realm or, rather,
within the new kingdoms which at least in some of their aspects
bore some resemblance to the original realms and peoples.
To come back to Ian Wood’s statement that has been quoted
above, we may completely agree, therefore, that through Clovis, the
Franks became an influential people in the wider world of politics,
but possibly this is not the most important perspective. Franks, rather,
became the leading people in a new realm. The formation of a new
gens, however, was a long-term development. As long as the Frankish
kingdom(s) existed, they always remained kingdoms of several peo-
ples (under Frankish rulership), which never turned into a real “eth-
nic” group. They did, however, soon become a political unit whose
population as a whole perceived itself as being “Frankish”, thus cre-
ating a new, political as well as territorial, meaning of this term. It
was not until centuries later that they became “nations”. Before such
a process could come to an end, the Frankish kingdom was divided
continuously, creating different conditions respectively: first, in Mero-
vingian times, into Neustria (with Burgundy) and Austrasia, and,

142
Ibid., p. 205.
344 - 

afterwards, into the West and East Frankish kingdoms, Italy and
Burgundy. And yet, even the perception of the (old) Franks develop-
ing into a new kingdom, may pertain purely to modern thinking
whereas medieval authors made no sharp distinction between ethnic
and political categories. It is probable that a significant shift can be
ascribed to the fact that Francus adopted a double meaning: it could
mean a Frank by birth or a member of the Frankish kingdom. In
the east, the Franks were still only one gens within this realm (in
Franconia) and the inhabitants of the East Frankish kingdom (or even
of all the Frankish kingdoms). In this sense, however, being a “Frank”
had become something completely different because the peoples in
the Frankish kingdoms (east and west) gradually all became Franks
although, at the the same time, they retained (or revived, or devel-
oped) an ethnic identity that was also politically and territorially
motivated. This should be considered a significant factor in the trans-
formation of the Roman world.
 THE BRITONS: FROM ROMANS TO BARBARIANS

Alex Woolf

If the theme of this volume is to explore the development of gentes

into regna then the history of the Britons in Late Antiquity has, in
some senses, no place in it. Perhaps the best known fact about the
Britons, from the days of Gildas (mid sixth century) to those of
Llywelyn the Last († 1282), was that they were never able to unite
and, through constant civil strife, allowed the Island of Britain to be
eaten up by the English. Nevertheless an investigation of a portion
of the former Roman West which lost its sovereignty to local mag-
nates rather than to barbarian warlords may well prove instructive
to those who might otherwise be tempted to see the establishment
of Germanic regna as an inevitable consequence of the collapse of
imperial control. Perhaps the most vexing question of the Late Antique
transition on the Continent is why no Gallic or Hispanic polities
ruled by native dynasties emerged. In the fifth century the activities
of Didymus and Verinianus in Spain,1 of Ecdicius among the Arverni2
and of the shadowy Syagrius amongst the Suessiones,3 demonstrate
that autochthonous military-political action amongst the native aris-
tocracy of the West was not impossible but in each of these cases
there were no long term polities or dynasties established.4 It is prob-
ably alongside such fifth century figures that we should place the
British leader Ambrosius Aurelianus, the vir modestus whose parents
“had worn the purple and were slain in it”.5 The commonality of
experience in the various provinces of the Continental Empire is in

1
Orosius, Historiarum adversum paganos libri VII 7,40, ed. K. Zangemeister, CSEL
5 (Vienna 1882; repr. Hildesheim 1967) [henceforth: Orosius].
2
Sidonius Apollinaris, Poems and Letters: Epistulae 3,3,3 ff., ed. W.B. Anderson
(Cambridge Mass. 1936).
3
Gregory of Tours, Zehn Bücher Geschichten 2,27, ed. R. Buchner (Berlin 1955).
4
There is also the suggestion that independent native resistance against the Goths
took place in Tarraconensis c. 470 in Isidore of Seville, History of the Kings of the
Goths 34, transl. K.B. Wolf [Conquerors and Chroniclers of Early Medieval Spain] (Liverpool
1990).
5
Gildas, Ruin of Britain 25,3, ed. and transl. M. Winterbottom (London-Chichester
1978).
346  

sharp contrast to the very different experiences of the British terri-

tories in which native non-imperial resistance appears to have led
to the establishment of enduring regna.
Difficulties arise, however, when we attempt to understand the
processes by which the British regna were established. Following the
departure of the usurper Constantine for the Continent in 406/7
the narrative history of fifth-century Britain is fragmentary and obscure
and can only be pieced together with the greatest difficulty. In an
attempt to summarise and elucidate this narrative it is perhaps best
to go through our source materials one by one.

Olympiodorus of Thebes (preserved in Sozomen and Zosimus)

Olympiodorus of Thebes was born in or around the year 380, he

was an imperial official from 407 until at least 425. His History sur-
vives only in fragments but it is the source for two later works, the
last part of Book IX of Sozomen’s Ecclesiastical History and the sec-
ond half of Book V and what remains of Book VI of Zosimus’ New
History. The former was composed in the mid fifth century and the
latter probably in the early sixth.6
The most famous passage from this body of evidence to refer to
Britain is the so called “Honorian Rescript”. In the midst of a pas-
sage describing the measures taken against the Goth Alaric in Italy
by the Emperor Honorius, Zosimus, following Olympiodorus, writes
“Honorius sending letters to the poleis of Brettania [sic] ordering them
to defend themselves”.7 It has been pointed out that Olympiodorus
regularly calls Bruttium in southern Italy Brettia and suggested that
in the context of the chapter this is what is meant here, the expan-
sion to Brettania being a slip on the part of Zosimus or a copyist.8
E.A. Thompson argues, against this interpretation, that Bruttium is
actually not much nearer Liguria and Ravenna (where our sources

6
See J.F. Matthews, “Olympiodorus of Thebes and the history of the west (A.D.
407–425)”, Journal of Roman Studies 60 (1970) pp. 79–97, and P. Bartholomew, “Fifth-
century facts”, Britannia 13 (1982) pp. 261–70. For editions see Sozomen, Kirchengeschichte,
ed. J. Bidez (Berlin 1960) and Zosimus, Historia Nova: the decline of Rome, ed. J.J.
Buchanan and H.T. Davis (San Antonio 1967).
7
Zosimus, Historia Nova 6,10,2.
8
Bartholomew, “Fifth-century facts”, p. 262.
   347

place Alaric and Honorius, the protagonists in this passage) than

Britain is, and that Alaric had made no effort to march south at
this stage so the cities there were not under threat. Why, Thompson
also asks, should Honorius address the civitates individually and not
the provincial governor? He also cites Gildas as an independent cor-
roborator of the reading “Britain”. We should perhaps reserve judge-
ment on this issue, which is not proven either way beyond reasonable
doubt, unless further evidence is produced. The letters were sent,
wherever they went, in the early summer of 410.9
The next passage in Zosimus we should turn to is 6,5,3, in which
he writes “The barbarians above the Rhine, assaulting everything at
their pleasure, reduced both the inhabitants of Britain and some of
the Celtic peoples [i.e. the population of Celtica, the north-western
region of Gaul] to defecting from the Roman rule and living their
own lives disassociated from the Roman law. The Britons, therefore,
taking up arms and fighting on their own behalf, freed the poleis
from the barbarians who were pressing upon them: and the whole
of Armorica and other provinces of Gaul, imitating the Britons, freed
themselves in the same way, expelling the Roman officials and estab-
lishing a sovereign constitution on their own authority.” Worried by
the fact that the Britons were fighting against barbarians and the
Armoricans against Romans (Constantine III’s regime), Bartholomew
suggested that the Greek word translated “pressing upon”, epikeimenon,
epikeimenon, should be read “billeted upon”.10 Thompson refuted this
by pointing out that Zosimus used this word at least twenty-eight
times and always either impersonally as “situated upon (a river, a
roof etc.)” or personally meaning “menacing”, “threatening” and so
forth. To summarise then, several civitates in Britain were simulta-
neously oppressed by barbarians and their defence by the local cit-
izens was so impressive that it inspired the Armoricans en masse, to
confront the regime that dominated Gaul and Spain. The British
event was clearly not a local skirmish with a couple of boatloads of
pirates. These events are dated to 409 or 410.

9
E.A. Thompson, “Zosimus 6.10.2 and the letters of Honorius”, Classical Quarterly
32 (1982) pp. 445–62.
10
Bartholomew, “Fifth-century facts”.
348  

The Chronicle of 452

The Chronicle of 452 is a difficult text. For most of the last century
it was regarded as genuine and broadly accurate, if a little shaky in
places, but in 1978 Molly Miller suggested that it had been seriously
re-edited in the Carolingian period and that the second reference to
Britain (§ 126) was an interpolation.11 Four main factors effected her
thesis. Firstly the re-editing of the early part of Jerome’s Chronicle,
which precedes it in all the manuscripts, secondly the usage of the
term Britanniae which she considered anachronistic, thirdly the appar-
ent coincidence of the date of § 126 with Bede’s date for the Adventus
Saxonum (always assumed to be calculated from Gildas), and lastly
the fact that the entry apparently runs across two annals. Muhlberger
and Wood have defended the veracity of the entry.12
Muhlberger points out that the earlier sections of Jerome had orig-
inally followed Eusebius’ model of parallel fila relating contemporary
events in different kingdoms. After the universal imperium of Rome
emerged only one filum was necessary and from A.D. 70 the chron-
icles take on a single thread. According to Muhlberger it was the
parallel columns that annoyed the Carolingian scribes and these
which were removed and that editing in the latter part of the text
was unnecessary and minimal. Similarly Jerome frequently spread
his entries over two years, particular if they concerned processes
rather than tight events.
Wood questions whether the use of Britanniae, probably technically
anachronistic in the fifth century, really casts doubt on the authen-
ticity of the Chronicle since Prosper and Constantius, both undisputed
fifth-century sources, also use it, as does Saint Patrick. This, Wood
argues, is Gallo-Roman literary usage not civil service jargon.13
Similarly Miller relates the date of § 126 with reference to a four-
year Olympiad scheme, which cannot be as precise as she claims

11
M. Miller, “The last British entry in the ‘Gallic Chronicles’”, Britannia 9 (1978)
pp. 315–8.
12
S. Muhlberger, “The Gallic Chronicle for 452 and its Authority for British
Events”, Britannia 14 (1983) pp. 23–33. Muhlberger’s views are re-iterated and
expanded in his volume The Fifth-century Chroniclers: Prosper, Hydatius and the Gallic
Chronicle of 452 (Leeds 1990) and in I.N. Wood, “The Fall of the Western Empire
and the End of Roman Britain”, Britannia 18 (1987) pp. 251–62.
13
Wood, “The Fall of the Western Empire”, p. 253.
   349

and in any case has been shown to be totally unreliable within the
Chronicle of 452 (and which in the earliest, British Museum, manu-
script may have been added by a later hand).14 Mommsen himself
pointed this out and declared a preference for the regnal dates which
Muhlberger has shown to be accurate from the death of Honorius.
In regnal terms § 126 is located at 441–2 and not 445–6. This ear-
lier date is confirmed by the Chronicle of 511 which records the same
British event as taking place under the year 440, probably indicat-
ing that we should date the event to 440/1.15
Exact dating within the Chronicle of 452 is not perfect until after
446, events often appearing up to four years too late if they take
place outside of Gaul. The author was probably based in the south
of Gaul and may have been, or been connected to, Faustus abbot
of Lérins (c. 433–60), later bishop of Riez, himself a Briton.16 It can
be assumed with some certainty that the event cannot be predated
by much, if at all, as many have argued, for the piece was written
in 452 and no chronological errors occur after 446. Almost certainly
the event can be secured to within the previous decade. The entry
reads: “The Britains [sic], having up to this time suffered various
defeats and catastrophes, were reduced to Saxon rule.” One other
entry, § 62, refers to Britain; dated to 408–9 it reads: “The Britains
devastated by an incursion of the Saxons”. So early in the Chronicle
its date cannot alone be trusted but since it is linked with the Suevi
and Vandal invasions of Gaul and Spain and with Constantine’s
usurpation it is usual to identify this event with the coming of the
barbarians whom Zosimus and, one assumes, Olympiodorus have
oppressing the poleis of Britain in 409. Whatever the truth of this
equation it is unlikely to be more than five years out.
In 1988 Jones and Casey produced a lengthy paper in which they
claimed to explain all the major chronological lapses of the Chronicle
and to prove that all the interpolated material was incidental to the
annals themselves.17 They then attempted to claim a chronological

14
M.E. Jones and P.J. Casey, “The Gallic Chronicle Restored: A Chronology
for the Anglo-Saxon Invasions and the End of Roman Britain”, Britannia 19 (1988)
pp. 367–98.
15
Wood, “The Fall of the Western Empire”, pp. 255–6.
16
I.N. Wood, “Continuity of calamity?: the constraints of literary models”, Fifth-
century Gaul: a crisis of identity, ed. J.F. Drinkwater and H. Elton (Cambridge 1992)
pp. 9–18, here p. 14.
17
See note 14.
350  

precision which seems a little too optimistic. They conclude their

thesis with a fatal and erroneous assumption: “the Saxon adventus
and settlement were clearly earlier than the dates presented by Bede
or inferred from Gildas. The British defeat appears surprisingly com-
pressed in time, catastrophic [. . .]”. What they have forgotten is that
their chronicler wrote all this only in 452, and that the long period
covered by the narrative of Gildas leaves plenty of opportunity for
a British recovery subsequent to that date, indeed De Excidio 24 seems
to describe just such a period of Saxon dominion preceding the
recovery which led to the British victory of Mons Badonicus.18
Jones’ and Casey’s confidence in the source and of their ability
to amend of scribal errors has been questioned by Burgess.19 The
two separate hands that they saw in the British Museum manuscript
are denied as is the primacy of that manuscript. Burgess’s bottom
line is that the Chronicle’s errors are real and original and not the
result of either scribal bungling or unsympathetic modern editing.20
With regard to the two “Saxons in Britain” entries Burgess suggests
that the first, being so distant in time from the author is not reli-
able and conflates the British revolt of c. 409, which he thinks of
as a loss of Britain by Rome, with the beginning of the Saxon wars
which led to the event recorded in c. 441. Because in most of the
West the loss of a province to Rome was the result of barbarian
invasion the chronicler, we are told, assumed this to be the case in
Britain. The second entry, § 126, he is convinced is correct, backed
up as it is by the Chronicle of 511 (“The Britains, deserted by the
Romans, fall under the authority of the Saxons”), and being located
only eleven years before the compilation of the Chronicle of 452, clearly
within living memory. Burgess, however, avoids the certitude of Jones
and Casey and prefers to state, wisely in my view, that in 452 it
was believed, in southern Gaul, that Britain had fallen to the Saxons
about ten years previously.

18
Gildas, Ruin of Britain 25.
19
R.W. Burgess, “The Dark Ages Return to Fifth-Century Britain: the ‘restored’
Gallic Chronicle exploded”, Britannia 21 (1990) pp. 185–95.
20
Ibid., p. 188.
   351

Gildas

We must now examine the value of Gildas’ De Excidio Britanniae as

a witness to fifth-century events and to the extent of Saxon military
activity in the fifth century. The text of De Excidio appears to carry
no clear dating information and we do not know when it was com-
posed. One statement is clear, however, that Gildas was born in the
same year as the siege of Mons Badonicus which was almost the
last battle in the war between the Saxons and the Britons which he
records.21 Wherever one wishes to date either his authorship, or
Badonicus, it is an inescapable conclusion that his parents’ genera-
tion was at its prime at the time of the closing stages of the conflict.
His reliability as a source thus rests upon the length of the war
before Mons Badonicus. If it was five years he will have grown up
surrounded by eye witnesses who will have been able to give him
all sorts of information (perhaps not technical strategic details but
those are not our concern) if the war went on for more than fifty
years then we might become very sceptical of his account of the
early stages. What we must not do, however, is to use his poor
knowledge of much earlier events, the Roman Walls, Boudica and
even Magnus Maximus to devalue his knowledge of the more recent
past. Gildas could easily have been reliably informed of events stretch-
ing back twenty or thirty years before Badon and possibly have had
quite a good idea of what things were like before that. Gildas is not
a precise historian, he makes no claims to be, but for the thirty or
forty years before his birth he has every right to be relied upon,
partly because he is claiming no precision.
Unfortunately, without anchor dates we cannot tell how long the
war was. Gildas clearly says, however, that “the fire of vengeance,
justly kindled by former crimes, spread from sea to sea, fed by the
hands of our enemies in the East, and did not cease until, destroy-
ing the neighbouring towns and lands, it reached the other side of
the island and dipped its red and savage tongue in the western sea”.22
Although much of the way in which he describes the brutality of
the Saxon onslaught is obviously rhetorical he clearly states twice in
this sentence that they overran the country from coast to coast and

21
Ibid.
22
Gildas, Ruin of Britain 24.
352  

elsewhere, with reference to fleeing bishops and overseas refugees it

is apparent he is not describing a local phenomenon.

Discussion of the sources

Admittedly the source material for the Saxon invasions of Britain is

spread pretty thinly, but it is consistent. Each of our sources views
the Saxon invasions as national calamities. It is true that we are not
talking about a coup d’état of the sort executed by the Goths in Italy,
but it is clear that most people in most parts of the country must
have felt threatened by them. They were no mere raids by a cou-
ple of ship loads. Gildas’ account of the first Saxons arriving in
“three keels” appear to be folkloric and, in any case, in the next
sentence he says that they were immediately followed by “a larger
company of [Germania’s] wolfish offspring”.23 An apt Continental
parallel might be the Suevi who entered Gaul and Spain in 406.
Unlike the Goths in Italy or Clovis’ Franks they did not seize the
imperial administration but rampaged about the provinces living from
plunder.
The evidence of these sources is evidence for warfare and should
not be confused with the separate issue of Germanic settlement or
Germanicization of indigenous communities—these are three very
distinct processes. One wonders what definite archaeological evidence
survives of the Suevic campaigns in Spain, or even the far more set-
tled Goths in Aquitaine. These points have been raised by Sims-
Williams, who also stresses the distinction between conquest and
settlement and argues in support of a literal reading of the Gallic
Chronicles.24
The appeal to Aëtius, “thrice consul”, cited in § 20 of De Excidio,
is often used to date Gildas’ narrative and it formed the basis of
Bede’s computation, how does it fit in with the evidence presented
here? Gildas claims that the Saxons were invited in only after Aëtius
failed to send help but clearly, if the Saxons were already in the

23
Ibid. 23. See P. Sims-Williams, “Gildas and the Anglo-Saxons”, Cambridge
Medieval Celtic Studies 6 (1983) pp. 1–30, for a discussion of the folkloric nature of
the “three keels” episode.
24
Ibid.
   353

dominant position by 441, as the Gallic Chronicles appear to imply,

then there is a fault in Gildas’ narrative if it demands that the Saxons
did not arrive until after 446 (the year of Aëtius’ third consulship).
Earlier it was argued that it was his lack of any attempt at preci-
sion that guaranteed Gildas’ usefulness. At this point, in his only
attempt to quote from a recent historical document, he slips up.
Aëtius became consul for the third time in 446 and was murdered
in 454, so the letter, addressed to “Aëtius, thrice consul” must date
to this period, when the Britains ought to be “under the authority
of the Saxons”.
Patrick Sims-Williams has suggested a solution. Gildas is quoting
a real document from c. 450, but his contextualising of it is not cor-
rect.25 In the last century both Skene26 and Thurneysen27 had held
that the nameless barbarians of the letter must have been the Saxons,
on the basis of the Gallic Chronicles, and that Gildas had misplaced
the letter. Sims-Williams restated that position, arguing that the
description of the Britons being caught between the barbarians and
the sea fits much better with conflict with the Saxons than with the
northern barbarians. For our current purpose we should note that
the letter to Aëtius is addressed gemitus Britannorum, or possibly Britan-
niarum; the “sigh of the Britons [or Britains].” A national appeal in the
face of a national threat that required the desperate remedy of sur-
rendering national independence, which is what Aëtius’ expedition
would have meant if it had come off.
Taking all the sources together we are left with the impression of
a national war c. 410, a Saxon conquest in c. 441 and a war of
resistance (which the archaeological evidence might suggest extended
to c. 520), with an appeal to the Magister Militias per Gallias in the
middle of it.

25
P. Sims-Williams, “The settlement of England in Bede and the Chronicle”, Anglo-
Saxon England 12 (1983) pp. 1–41.
26
W.F. Skene, The Four Ancient Books of Wales containing the Cymric poems attributed
to the bards of the sixth century, vol. 1 (Edinburgh 1868) pp. 35–7.
27
R. Thurneysen, “Wann sind die Germanen nach England gekommen?”, Englische
Studien 22 (1896) pp. 163–79.
354  

Tentative historical narrative

An attempt to pull all the surviving narratives together makes most

sense if we recall that Gildas believed Magnus Maximus to be the
last Emperor to rule in Britain.28 Keeping this in mind we are in a
position to accept that much of his “post-Roman” history may actu-
ally have occurred between 388 and 409. We know that Gildas used
Orosius whose narrative continued up to the time of Constan-
tine III and included an account of the raising up and murder of
the British usurpers Marcus and Gratianus in 405–6,29 and it is pos-
sible that he summarises it in his own account of the raising up and
as quickly casting down of usurpers he describes as the first post-
Roman stage of British history.30 Since we have seen from our Conti-
nental sources that there were Saxons in Britain as early as c. 409
when the Britons rebelled against Constantine it seems that we could
most easily reconcile them with Gildas’ narrative by presuming that
the tyrant who invited the Saxons in was in fact Constan-tine III.
Gildas refers to this tyrant (not identified with Vortigern until Bede’s
day) as infausto tyranno31—ill-fated tyrant—which suggests that he knew
that this ruler was to come to a sticky end.32
Such a short chronology would also help to solve the problem of
Ambrosius Aurelianus. We are told that his parents had worn the
purple and been slain in the previous anarchy and that he was a
Roman. Both Marcus and Gratianus had assumed the imperial title
and are thus candidates for his paternity.33 Ambrosius may then be
taken as the leader of the uprising of 409 and the liberator of the
British poleis. A long period of warfare may be presumed to have
followed in which the Saxons had obtained the upper hand by the
440s, prompting the appeal to Aëtius, with some kind of peace finally
emerging in the first quarter of the sixth century. Unfortunately, a
short chronology of this sort, putting Ambrosius into the first quar-

28
Gildas, Ruin of Britain 15,2.
29
Orosius 7,40.
30
Gildas, Ruin of Britain 21,4.
31
Ibid. 23,4.
32
Cf. the account of Constantine’s demise in Orosius 7,42.
33
Marcus may be the better candidate for two reasons. Firstly we are told that
he was a Roman whilst Gratian was a Briton and Gildas describes Ambrosius as
a Roman and secondly because the name Marcus in late Antiquity was frequently
paired with Aurelius.
   355

ter of the fifth century, leaves us with even less evidence for the
development of political structures between his day and that of Gildas
writing, it is presumed, in the mid sixth century.

The “sub-Roman” period

Moving from fifth-century events to the processes of kingdom for-

mation we are forced to re-focus our attention on the sixth century
which appears to be a very different world from the Roman West
of the late fourth and early fifth centuries. The world Gildas wrote
in is one illuminated almost entirely by his own text, by epigraphic
evidence and by the preservation of contemporary elements in much
later texts (be it poetry or charter material), the veracity of which
is hotly debated by scholars. The leap from Roman Britain, often
presumed to be well understood, to this “Celtic Heroic Age” has
proved to be a gymnastic feat which has defeated the most tena-
cious of scholars. One question that has perhaps not been investi-
gated as deeply as it should is whether or not the concept of a
homogenous “sub-Roman” culture is truly valid. Within the Roman
period there is an oft-noted, though somewhat simplistic dichotomy
drawn between a “civil”, “lowland” zone and a “military”, “high-
land” zone, the latter largely in the North and West and the for-
mer in the South and East. Its continuity into the post-Roman period,
when a lazier dichotomy between the British and English zones of
the island seems more natural has rarely been investigated.

Cultural division in “sub-Roman” Britain

This “highland” and “lowland” zone division is, however, recognis-

able in the post-Roman period, even within the British controlled
parts of the island. About two hundred and forty-two inscribed stones
survive from western Britain which have been dated, on palaeo-
graphic grounds, to between A.D. 400 and 700 (or, less cautiously,
between 450–650). Twenty-eight of these are bilingual inscriptions
in Primitive Irish and Latin and the rest simply in Latin.34 These

34
M.A. Handley, The Inscribed Charters of Early Medieval Wales (unpublished M.Phil.
dissertation, University of Cambridge 1995) pp. 1–2.
356  

monuments are, perhaps, the most characteristic landscape feature

surviving from Late Antiquity in this region. The period during which
these inscribed stones were being erected coincides fairly exactly with
the period in which so called “pagan” Anglo-Saxon burials of var-
ious sorts were taking place in the east of the country.
If one plots the distribution of inscribed stones against the dis-
tribution of “pagan” Anglo-Saxon burials one is left with a broad
band of territory in which neither class of site is found, running
north from the stretch of coast between Dartmoor and the Hamp-
shire Avon,35 comprising east and north Devon, most of Dorset and
Somerset, most of Gwent and a large part of Glamorgan, western
Gloucestershire, Herefordshire, much of Powys, Shropshire, most of
Worcestershire, Cheshire and most of Staffordshire and Derbyshire
as well as most of the north of England outside of East Yorkshire.
In addition to this broad band one might, tentatively, add the area
immediately north and north-east of London.36 The extreme south
of Scotland contains a few inscribed stones with one or two outliers
in the far north of England.37 There are of course a few stones and
burials, such as the Maiden Castle warrior-burial, which lie outside
the main distribution.38
While a large part of what had been the “lowland” or “civil” zone
seems to have fallen under the political control of Germanic-speaking
groups with their distinctive burial practices, the boundary between
those parts of this zone which remained in British hands and the
“highland” or “military” zone seems to have been marked by the
transition between the regions in which the inscribed stone tradition
took root and that in which it did not.
Interestingly, this division seems to be reflected in our one liter-
ary source for the period, the De Excidio Britanniae of Gildas. De Excidio

35
There is, however, a small group of late stones at Lady St. Mary’s, Wareham
(see D. Hinton, “The Inscribed Stones in Lady St Mary Church, Wareham”,
Proceedings of the Dorset Natural History and Archaeological Society 114 [1992] p. 260, and
B.A.E. Yorke, Wessex in the Early Middle Ages [London 1995] pp. 69–72), to which
we shall return later.
36
K.R. Dark, Civitas to Kingdom (Leicester 1994) pp. 86–9.
37
C. Thomas, “The Early Christian Inscriptions of Southern Scotland”, Glasgow
Archaeological Journal 17 (1992) pp. 1–10; K.R. Dark, “A sub-Roman defence of
Hadrian’s Wall?”, Britannia 22 (1992) pp. 111–20.
38
A. Meaney, A Gazetteer of Early Anglo-Saxon Burial Sites (London 1964) p. 81
n. 602.
   357

takes the form of a letter addressed by Gildas to the leaders of British

society, possibly modelled on the speech attributed to the deacon
Stephen at his trial before the Sanhedrin in Acts 7.39 This letter attrib-
utes the decline of British national fortunes to its recipients’ poor
leadership and sinful personal lives. It almost certainly dates from
the mid-sixth century.40 Gildas’ lament is addressed to the secular
and ecclesiastical leaders of the Britons, but the only individuals
amongst his audience whom he names are five kings. The kings
named are, in order, Constantinus, Aurelius Caninus, Vortiporix,
Cuneglasus and Maglocunus.41 Various attempts have been made to
identify these kings both with figures recorded in later Welsh pedi-
grees and legends and with territories. An element of consensus has
emerged and the most scholarly summary of this remains David
Dumville’s “Gildas and Maelgwn”.42
Constantinus is specifically stated to be “the usurping whelp of
the foul lioness of Damnonia”.43 “Damnonia”, here, is usually taken
to be the Dumnonian civitas or its successor state, and this Constantine
may be the Custennin Gorneu (“of Cornwall”) of later Welsh tra-
dition44 and perhaps the Custennin who appears in various pedigrees
to have been the son of one Cynfawr,45 who in turn may be the
“Quonomorus” mentioned in Wrmonoc’s ninth-century Life of Paul
Aurelian and in the, possibly seventh-century, Vita Samsonis.46
Aurelius Caninus, the second king, bears an entirely Latin name
but Jackson has argued that the “caninus” part is a Latinizing pun
on a British name formed of the element cuno- “hound”.47 Converting
this to Latin, Jackson argued, allowed Gildas to play on the nega-
tive associations that dog imagery has in the Bible, in contrast to the
positive associations of hound imagery in traditional Celtic society.

39
R. Gardner, “Gildas’ New Testament Models”, Cambrian Medieval Celtic Studies,
30 (1995) pp. 1–12.
40
D.N. Dumville, “Gildas and Maelgwn: problems of dating”, Gildas: New Approaches,
ed. id. and M. Lapidge (Woodbridge 1984) pp. 51–60.
41
Gildas, Ruin of Britain 28–36, discussed by K.H. Jackson, “Varia II: Gildas and
the names of the British Princes”, Cambridge Medieval Celtic Studies 3 (1982) pp. 30–40.
42
See note 40.
43
Translation from Jackson, “Varia II”, p. 30.
44
Trioedd Ynys Prydain, ed. R. Bromwich (Cardiff 1978) pp. 314–5.
45
Ibid., pp. 355–60.
46
Both vitae are edited in Vitae Sanctorum Britanniae et Genealogiae, ed. A.W. Wade-
Evans (Cardiff 1944).
47
Jackson, “Varia II”, p. 31.
358  

Jackson followed, albeit cautiously, by Dumville, identified the under-

lying British name as Cunignos, which would give Middle Welsh
Cynin.48 This choice is informed largely by the presence of i as the
second vowel in Gildas’ form, but as this spelling is constrained by
the requirements of the Latin then an alternative “Cynan” is also
possible. A Cynin and a Cynan both appear in approximately the
correct place in the pedigrees of ninth century kings of Powys. Cynin
appears in the pedigree of Hesselis in Harley Genealogy 23 [hence-
forth: HG], and Cynan in the pedigree of Cynan ap Cadell in
HG 27.49
The extent of Powys has varied greatly over the last fourteen-hun-
dred years. The modern county, created in 1974, bears little rela-
tionship to any of its medieval incarnations. The kingdom disappeared
in the ninth century, conquered first by Mercia50 before being absorbed
by Gwynedd, as Mercia herself succumbed to the Great Army later
in the century.51 Powys reappeared as a kingdom in the final years
of the eleventh century when Cadwgan ap Bleddyn divided Gwynedd
with his cousin Gruffudd ap Cynan, but this kingdom bore little
resemblance to its predecessors.52 Locating Powys, or at least the
dynasty of the Cadelling, in Late Antiquity is no less fraught, but it
seems likely to have been centred upon the valley of the Dee.53
Vorteporix is the easiest of the kings to locate. Gildas explicitly
calls him tyrant of the Demetae (the people of Dyfed),54 and not
only does someone of this name appear in the Demetian pedigree
in about the right generation,55 but a memorial stone in primitive
Irish and Latin commemorating a “Voteporix” survives at Castell
Dwyran, near Carmarthen.56

48
Dumville, “Gildas and Maelgwn”, pp. 56–7.
49
Early Welsh Genealogical Tracts, ed. P.C. Bartrum (Cardiff 1966).
50
Annales Cambriae s.a. 822, ed. J. Morris (Chichester 1980).
51
W. Davies, Patterns of Power in Early Wales (Oxford 1990) pp. 34–5.
52
Brut y Tywysogion or the Chronicle of the Princes. Peniarth MS 20 Version s.a. 1097–99,
ed. T. Jones (Cardiff 1952); Brut y Tywysogion or the Chronicle of the Princes. Red Book
of Hergest Version s.a. 1097–99, ed. T. Jones (Cardiff 1955).
53
The full arguments for this location would be too digressive at this stage. I
hope to publish them elsewhere.
54
Gildas, Ruin of Britain 31.
55
Early Welsh Genealogical Tracts, p. 4.
56
V.E. Nash-Williams, Early Christian Monuments of Wales (Cardiff 1950) p. 107
n. 138.
   359

The fourth king addressed by Gildas is the most problematic. The

name Cuneglasus would give the Middle Welsh “Cynlas”.57 According
to HG 3 there was a Cynlas who was a first cousin of Maglocunus58
(see below), but, since he would seem to be a dynastic rival of
Maglocunus who is himself addressed by Gildas in terms which sug-
gest that he has overthrown rivals (one explicitly his father’s brother),59
it is hard to imagine the two ruling together.60 Dumville has sug-
gested that Cynlas’ descendants may have originally ruled a king-
dom in north-east Wales which was clearly separate from Gwynedd
(Maglocunus’ kingdom), based on Rhos (the area between the Conwy
and the Clwyd), and that they falsified their pedigree to create a
descent from the progenitor of the Gwynedd dynasty (Maglocunus’
grandfather Einion Yrth) when they seized the kingship of Gwynedd
in the late eighth century.61 Charles Thomas has suggested that
Cuneglasus, since he appears between Vortiporix and Maglocunus in
Gildas’ text, must have been located between them geographically.
He suggests that Ceredigion, on the west coast of Wales, might be
the appropriate location for him.62 The existence of a sixth-century
polity coterminous with eighth-century Ceredigion seems unlikely,
however, given that the southern part of the kingdom seems, on the
basis of a very high density of bilingual Irish and Latin inscriptions,
to have been part of whatever unit the “Irish” zone of western Dyfed
fell into, while north of the river Aeron, which bisects Ceredigion,
Irish inscriptions are completely absent.63
The last of Gildas’ kings, Maglocunus, has been identified since
the early middle ages with the Venedotian dynast Maelgwn the Tall
and not even the sceptical Professor Dumville seems inclined to chal-
lenge this identification. Gildas refers to Maglocunus as “Dragon of
the Island”64 and later traditions firmly link him with Anglesey and
its immediate environs.65

57
Jackson, “Varia II”, pp. 32–3.
58
Early Welsh Genealogical Tracts, p. 10.
59
Gildas, Ruin of Britain 33,4.
60
Dumville, “Gildas and Maelgwn”, p. 58.
61
Ibid., pp. 58–9.
62
C. Thomas, Christianity in Roman Britain to A.D. 500 (London 1981) p. 251.
63
C. Thomas, And Shall these Mute Stones Speak (Cardiff 1994) pp. 89–112; Dark,
Civitas to Kingdom, p. 83.
64
Gildas, Ruin of Britain 33.
65
Trioedd Ynys Prydain, pp. 437–41.
360  

The purpose of this digression on the identification of Gildas’ kings

has been to make the point that the rulers he addresses occupy those
territories which are characterised archaeologically by the presence
of Class I inscribed stones. Within southern Britain the only region
with a concentration of stones whose ruler he does not appear to
address is the tiny land-locked realm of Brycheiniog. At the same
time Gildas addresses by name no rulers of any of the other regions
which seem to have remained under British control during the sixth
century but which do not have significant concentrations of Class I
stones; that is to say, central and south-eastern Wales and western
England.
Apart from concentrations of stones around Whithorn and Kirk-
madrine, which may bear witness to direct contacts with Gaul and
stand somewhat outside the Class I tradition, there are very few
Class I stones in the North.66 The most northerly Class I stone is
the Catstane at Edinburgh Airport and the most southerly at Chester-
holm (Vindolanda) on the Stanegate, but all the others lie between
the Lammermuir/Lowther watershed and the line of Hadrian’s Wall.
Since Vindolanda was part of the Wall complex it is probably safe
to conclude that here too the distribution of Class I stones coincides
with the “military” zone. One point worth noting is the absence of
any Class I stones from Lancashire, the Pennines or the Lake District,
probably all regions controlled by Britons into the seventh century
and traditionally viewed by Romanists as part of the “military” or
“highland” zone.
The implications of this analysis for the present study should be
to alert us to the fact that there were two kinds of Britons in Late
Antiquity. Most of the territory conquered by Germanic-speaking
groups prior to the ninth century was occupied at the time of con-
quest by the “lowland”, “civil” British, who did not erect inscribed
stone memorials and who did not possess kings who were either
addressed by Gildas or whose deeds were remembered in Welsh
poetry. One must be aware that it is not simply the fact of their
disappearance that led them to be forgotten, for the kings of the
Old North were remembered in Wales and, indeed, were the ideal
models of heroic kingship.67 This may suggest that kings, and by

66
C. Thomas, Whithorn’s Christian beginnings (Whithorn 1992) and cf. note 35.
67
K.H. Jackson, “The Britons of Southern Scotland”, Antiquity 29 (1955) pp. 77–
   361

extension regna, analogous to those of Wales, Dumnonia and the Old

North, did not exist in the “lowland” zone. Indeed the different kind
of relationship between Angles and Britons in Bernicia, with more
emphasis on British ideas and culture influencing the incomers, pos-
tulated, on the basis of archaeological and geographical evidence, by
modern scholars, may owe something to the different nature of the
Britons encountered in this region.68
The lowland British have proved notoriously difficult to iden-
tify in the archaeological record. The main problem in identifying
and quantifying their material remains is that while the bulk of the
population seems, in terms of agrarian and domestic practice, to
have continued to enjoy fourth-century life-styles, coinage and mass-
produced ceramics, the cornerstones of Romano-British chronology,
ceased to circulate in Britain, or at least to be produced.69 The low-
land British are best known from cemetery evidence. The dominant
mortuary rite continued late Roman practice and comprised extended
inhumation accompanied by few or no grave goods.70 There is lit-
tle evidence for social stratification within these cemeteries even
though some of them are very large; Cannington, in Somerset, for
example is thought to contain some two thousand burials, only a
quarter of which have been excavated, dating from the second to
the eighth century.71 Further east similar cemeteries are also known,
though they rarely seem to have stayed in use so late. At Queensford
Farm, near Dorchester-on-Thames, a cemetery, also containing about

88; L. Alcock, “Gwyr y Gogledd: an archaeological appraisal”, Archaeologia Cambrensis

132 (1983) pp. 1–18.
68
For this mixed culture see P.H. Sawyer, From Roman Britain to Norman England
(London 1978) pp. 88–91; L. Alcock, Economy, Society and Warfare Among the Britons
and Saxons (Cardiff 1987) pp. 264–6; R. Miket, “A re-statement of evidence for
Bernician Anglo-Saxon burials”, Anglo-Saxon Cemeteries 1979: the fourth Anglo-Saxon
Symposium at Oxford, ed. P.A. Rahtz, T. Dickinson and L. Watt (Oxford 1980) pp.
289–305.
69
Ken Dark (Civitas to Kingdom, pp. 200–6) suggests that the coinage in circula-
tion in the first quarter of the fifth century would have been sufficient to maintain
a partially monetarised economy for several generations thereafter. This is an unortho-
dox view, but if proved right would mean that many occupation sites dated to the
fourth or early fifth centuries on the basis of coinage might well have continued
into a much later period.
70
P.A. Rahtz, “Celtic Society in Somerset, A.D. 400–700”, Bulletin of the Board
of Celtic Studies 30 (1982) pp. 176–200, here p. 187.
71
Ibid.
362  

two thousand graves, was in use from the late fourth to, at least,
the mid-sixth centuries, and perhaps even into the seventh.72
The other major archaeological phenomenon associated with the
lowland British is the re-occupation of Iron Age hill-forts;73 most
famously South Cadbury74 and Cadbury-Congresbury.75 Many of
these sites had remained in some kind of use through the Roman
period, but this was mainly of a ritual nature.76 In the fifth and sixth
centuries refortification took place and domestic occupation was re-
established. At the larger sites, including the two just mentioned,
imported Mediterranean pottery has been recovered. The re-emer-
gence of hill-fort settlement coincides with the disappearance of evi-
dence for occupation on villa sites and it is fairly safe to assume that
the former replaced the latter as centres of elite residence, albeit for
a more restricted elite.77 It is not clear, however, that the adoption
of such sites should be seen as a conscious militarization of the elite
rather than as part of a simple desire for greater personal security
and the appropriation of dominant places within the landscape.78
One should certainly be cautious of crying “continuity” from the
Iron Age; reoccupation after several hundred years of abandonment
as residence sites may reflect conscious archaism but this is not the
same as continuity.79 Whilst the villa, an Italianate country house,
symbolised its owners’ links to the affluent, yet remote, society in
which the god-like emperor and the imperial court existed, the hill-
fort served as a cruder reminder of exactly what the sources of social
power were.
What can be said with some conviction is that urban life certainly
came to an end at some point in the fifth century: although some
occupation of town sites may have continued this occupation was

72
R.A. Chambers, “The Late- and Sub-Roman Cemetery at Queensford Farm,
Dorchester-on-Thames, Oxon.”, Oxoniensia 52 (1987) pp. 35–69.
73
A.S. Esmonde Cleary, The Ending of Roman Britain (London 1989) p. 179.
74
L. Alcock, Cadbury Castle, Somerset: the Early Medieval Archaeology (Cardiff 1995).
75
P.A. Rahtz, Cadbury-Congresbury 1968–73: a Late/Post Roman Hilltop Settlement in
Somerset (Oxford 1992).
76
P.A. Rahtz and L. Watt, “The End of Roman Temples in the West of Britain”,
The End of Roman Britain, ed. P.J. Casey, British Archaeological Reports, British series
71 (Oxford 1979) pp. 183–210.
77
Esmonde Cleary, Ending of Roman Britain, p. 158.
78
Cf. Dark, Civitas to Kingdom, pp. 40–44.
79
Ibid.
   363

not urban in character. At Wroxeter80 and Verulamium large tim-

ber structures do seem to have been built in the mid- to later fifth
century, but these seem to have been the dwellings of high status
individuals,81 perhaps using the enclosed urban area much as some
of their contemporaries used the ancient hill-fort ramparts. With the
disappearance of towns and coinage, craft specialisation and mass
production also came to an end.
Circumstantial evidence for the character of “sub-Roman” British
material culture can be found in Ireland. As long ago as 1950 Hugh
Hencken drew attention to the Romano-British character of many
of the finds from Lagore crannog.82 This artificial island was the
chief residence of the kings of Southern Brega and its initial phase of
construction has been dated through documentary references, radio-
carbon and the presence of E-ware to the mid-seventh century.
Lloyd Laing reviewed the evidence for the “Romanization of Ire-
land” in 1985. He argued that while some of the influences may
date back to the fourth century, others probably reflect continued
links into the post-Roman period. The pursuit of late Romano-British
influences upon Irish metalworking styles, particularly with reference
to penanular brooches, has been taken up by Ragnall Ó Floinn.83
He argues convincingly for a transmission of first artefacts and then
skills from the middle and lower Severn region to Mag Brega and
thence a secondary spread into the rest of the Gaelic World. This
Irish evidence suggests that forms of material culture which cannot
be demonstrated to have continued in production in Britain beyond
the beginning of the fifth century, due to the absence of datable
associations, may indeed have continued in production into the sixth
and seventh centuries.
Turning to documentary sources there is very little that survives
relating to the lowland zone. A potentially useful, yet very complex,

80
R. White and P. Barker, Wroxeter: the Life and Death of a Roman City (Stroud
1998).
81
Esmonde Cleary, Ending of Roman Britain, p. 153.
82
H. Hencken, “Lagore: a royal residence of the seventh to tenth centuries”, Pro-
ceedings of the Royal Irish Academy (C) 42 (1950) pp. 1–76, here pp. 15–6. Cf. E. Rynne,
“An important seventh-century sword from Co. Roscommon”, Journal of the Old
Athlone Society 2 (1985) pp. 148–51.
83
R.Ó Floinn, “Artefacts in Context: personal ornament in early medieval
Ireland—a case study” (paper presented at the Society of Medieval Archaeology
Conference Expressions of Cultural Identity, April 1997 Glasgow).
364  

group of sources are the collections of “charters” from south-east

Wales, principally those preserved in the Liber Landavensis, which
record land grants to various churches between the mid-sixth and
the late eleventh centuries.84 The text is a complicated one and
Wendy Davies has identified at least six separate stages of copying
of the early charter material after its original composition, which, in
turn, may well have taken place at as many as nine separate cen-
tres.85 The absolute veracity of the material from the early charters
is therefore doubly in doubt. Firstly, the final collection was made
as part of a political campaign by Bishop Urban of Llandaff to estab-
lish the superiority of his see in the mid-twelfth century, and thus
the motivation to produce fraudulent documents, or to amend exist-
ing ones, was certainly there.86 Secondly, six redactions allow plenty
of opportunity for scribal error or innocent, but inauthentic, clarifi-
cation to alter the text considerably. This said, there is still enough
evidence within the extant texts, such as archaic spelling of names,
Latin usage appropriate to Late Antiquity and so forth, to suggest
that real documents, perhaps even a handful dating to the sixth cen-
tury, do lie behind the twelfth-century text. While some scholars87
would be more cautious than Davies in identifying precisely what is
archaic most would agree that the witness borne by these texts is,
nevertheless, valuable.
For the present purpose, we need not concern ourselves with ques-
tions of the absolute accuracy of individual land grants, but need
only look at some of the general distinguishing features which mark
out the material purporting to come from Late Antiquity from that
belonging to later periods. First of all we should note that before
the mid- to late seventh century there was no centralised kingdom,
but instead the territory of the old Silurian civitas was shared between
“a number of minute kingdoms” in which “[k]ingship was some-
times shared between sons, and very occasionally cousins, [. . .] but

84
W. Davies, The Llandaff Charters (Aberystwyth 1979).
85
Ibid., pp. 23–8.
86
J.R. Davies, “Liber Landavensis: Its Date and the Identity of its Editor”, Cambrian
Medieval Celtic Studies 35 (1998) pp. 1–12.
87
E.g. P. Sims-Williams, “Review of W. Davies, The Llandaff Charters”, Journal of
Ecclesiastical History 33 (1982) pp. 124–9, and Dark, Civitas to Kingdom, pp. 140–8. It
should be noted that Sims-Williams (pers. comm.) claims that his original review
was not intended to be as negative as Dark and others seem to have believed.
   365

these do not appear to have been associated with any exclusively

defined territories”.88 Another feature of the early period, defined by
Davies as pre-740, is the large size of the land-grants, these being
“of the order of 500–1,000 acres”.89 From c. 760 to the eleventh
century there were only six grants larger than two-hundred and fifty
acres. Up to the early eighth century estates were termed agri, after
about 720 the term villae and its Welsh equivalent trefi are used.90
Davies also points out that whilst many of the villae can be traced
in the medieval and modern settlement patterns the agri have left
no trace on the landscape. Davies has some further interesting com-
ments to make:
As I have indicated above, there were nearly always several kings. The
political organisation engendered was of an unusual type, for there was
little to suggest that these kings had any governmental function, any
responsibility for law and order, or any role in arbitration. The kings
appear to have treated the kingship as property, and shared it just as
property would have been shared. [. . .] On the other hand, in the
seventh and eighth centuries there were local assemblies, meetings of
elders (seniores) and “better men” (meliores) with distinct regional iden-
tities, existing apart from the king.91
Society in south-eastern Wales seems to have changed considerably
in the course of the eighth to eleventh centuries, and the new social
formation resembles that apparent in the later medieval law-codes.
As well as the changes noted above, kin groups begin to emerge,
consenting to the alienation of land, and taking part in decision mak-
ing, but only from the tenth century.92 Davies speculated on why it
was that only kings granted land in the earlier period and came to
no firm conclusions.93 Given the lack of evidence that these early
donors had any of the attributes of kings, however, we might ask
whether or not their elevation to the kingship might not have been
part of the editorial process. Either through conscious fraud, churches
not wishing to rely upon the donation of long dead commoners with

88
W. Davies, “Land and Power in Early Medieval Wales”, Past and Present 81
(1978) pp. 3–23, here pp. 8–9.
89
Ibid., p. 11.
90
Ibid.
91
Ibid., pp. 13–4.
92
Ibid., p. 9.
93
Ibid., p. 16.
366  

no surviving kin to legitimise their tenure, or through genuine con-

fusion given the problematic terminology for rulership in Welsh tra-
dition94 early donors may have been granted the title king by
subsequent redactors of the material; the essential continuity of king-
ship making any subsequent king the heir, in some sense, of all pre-
vious kings. Certainly these “kings” do not sound like Gildas’ western
tyrants.95
If we free ourselves from the terminology of kingship what we
seem to be observing in the Llandaff material is a tenurial system
remarkably similar to that which existed in the later Roman West.96
This system seems to survive almost intact into the early eighth cen-
tury, and, thereafter, begins to be assimilated into the kin-based
tenurial system of the more barbaric “highland” zone.97
We might hypothesise, then, that, for all its limitations, the Llandaff
material may give us a clue as to how lowland British society was
organised. Returning to Gildas we should perhaps note that his
address to the secular rulers of Britain begins; Reges habet Britannia,
sed tyrannos; judices habet, sed impios—“kings Britain has, but tyrants;
judges she has, but wicked ones”.98 As Paul Schaffner has shown,
the term iudex can, in a Late Antique context, mean simply a ruler,
or it can mean some sort of magistrate elected from amongst the
seniores of a region.99 The ambiguity of the text makes it impossible
to tell whether Gildas intends his reges and iudices to be taken as two
terms for the same individuals or not. Given the distinctions we have
identified between “highland” and “lowland” society, however, it is
perfectly possible that he is using this term to castigate the leading
members of “lowland” society, too numerous and ephemeral in their
greatness to warrant singling out.
Dumville, inspired by Fanning’s discussion of Aegidius and Syagrius
in Gaul, has recently raised the spectre of an imperium (or more than
one) in sixth-century Britain.100 Might there have been a supreme

94
Davies, Patterns of Power, pp. 10–5.
95
Or should we be looking to the Burgundians for parallels?
96
A.H.M. Jones, The Decline of the Ancient World (London 1966) pp. 302–4.
97
For which see T.M. Charles-Edwards, Early Irish and Welsh Kinship (Oxford
1993) pp. 274–304.
98
Gildas, Ruin of Britain 27.
99
P. Schaffner, “Britain’s Iudices”, Gildas: New Approaches, ed. M. Lapidge and
D.N. Dumville (Woodbridge 1984) pp. 151–5.
100
D.N. Dumville, “The idea of government in sub-Roman Britain”, After Empire.
   367

overlord based in the lowland British community? It is, at first glance,

hard to see why such a ruler did not emerge. Britain had had emper-
ors before, up until Constantine III’s regime was expelled in 409,
and Gildas’ “proud tyrant”101 may have been one such in the suc-
ceeding generation (if, indeed, he is not Constantine III, as argued
above). The evidence for the existence of such an individual after
the middle of the fifth century, however, is entirely lacking, and the
acephalous nature of political society in lowland Britain may well
have contributed to the success of Germanic conquest and cultural
survival. The absence of convincing historical, archaeological, or even
literary evidence for any kind of British Empire suggests that low-
land Britain remained relatively fragmented in terms of formal polit-
ical structure. The broadly co-extensive frontiers of some Romano-
British civitates and Anglo-Saxon kingdoms of the seventh and eighth
centuries (e.g. Icenia and East Anglia, Parisia and Deira or Essex
and Trinovantia) may well suggest that the civitates, or large blocks
within them, retained some sort of identity, and perhaps groups of
seniores might have ruled, or tyrannised, their republics corporately
in much the same way as the principales described by Salvian dom-
inated the civitates of southern Gaul in his day.102
The social origins of the rulers of the “lowland” British have been
debated of late. Edward Thompson, and latterly Ken Dark, have
both argued for a peasants’ revolt being instrumental in the ending
of Roman Britain.103 The latter even suggested that rural proletarian
revolution was the force which led Christianity to supersede pagan-
ism as the dominant religion. Dark’s equation of agrarian revolu-
tion with Christian fundamentalism (non pagani sed pagani?), seems
bizarre and is based on inappropriate parallels with militant monas-
ticism in contemporary Egypt. The motivation in seeking such an
explanation for the radical transformation of material conditions

Towards an Ethnology of Europe’s Barbarians, ed. G. Ausenda, Studies in Historical

Archaeoethnology (Woodbridge 1995) pp. 177–216, here p. 183; S. Fanning,
“Emperors and Empires in fifth-century Gaul”, Fifth-century Gaul: a crisis of identity?,
ed. J.F. Drinkwater and H. Elton (Cambridge 1992) pp. 288–97.
101
Gildas, Ruin of Britain 23.
102
M. Maas, “Ethnicity, Orthodoxy and Community in Salvian of Marseilles”,
Fifth-century Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge
1992) pp. 275–84.
103
E.A. Thompson, Saint Germanus of Auxerre and the End of Roman Britain (Woodbridge
1984) pp. 34–8; Dark, Civitas to Kingdom, pp. 55–64.
368  

seems to lie in the modern commentators’ incomprehension as to

why “civilised” people would choose to give up town life. Such
incomprehension, however, does not manifest itself in the minds of
economic historians and archaeologists who are prepared to recog-
nise that phenomena such as urbanism are parts of complex inter-
regional systems. Esmonde Cleary has presented a perfectly acceptable
account of the collapse of the material fabric of late Roman Britain,104
and this has been neatly summarised by John Blair:
The most obvious fact of the early fifth century is that towns and vil-
las became irrelevancies, so that to look for “continuity” in the modes
of life which served them is to chase a will-o-the-wisp. Complex gov-
ernment, bureaucracy, coinage and long-distance trade simply could
not survive the secession of Britain from the Empire. “Civilisation”,
manifested in urbanism, specialised industry, plumbing, central heat-
ing, mosaics and all that was distinctively Roman, depended on these
things, though occasional Roman structures succumbed to the obliv-
ion slightly later than others.105
There is no need to appeal to social revolution to explain the col-
lapse of the infrastructure of early fifth-century Britain and, since
the peasant character of the bacaudae rebels in contemporary Gaul,
the paradigm which Thompson invoked, has been seriously ques-
tioned, such appeals should best be abandoned.106
Up to this point, a contrast has been emphasised between the
“highland” zone with its kings, inscribed stones and kin-based soci-
ety on the one hand and the “lowland” zone with its civil (if no
longer civilised) government and highly stratified society, preserving
much of what was fundamental to Roman self-perceptions, (despite
lacking the ability to maintain the material fabric which usually
encased them). Such a distinction must be made, the more so because
generations of scholars (most recently Dark) have sought to charac-
terise British society in the fifth, sixth and seventh centuries as an
homogenous entity typified by the images presented in the gwarchanau
of the Gododdin, despite the fact that the Uotadinii, whose heroes are

104
Esmonde Cleary, Ending of Roman Britain, pp. 131–61.
105
J. Blair, Anglo-Saxon Oxfordshire (Oxford-Stroud 1994) p. 3.
106
See for new thoughts on the bacaudae R. Van Dam, Leadership and Community
in Late Antique Gaul (Berkeley 1985) pp. 25–56, and J.F. Drinkwater, “The Bacaudae
of Fifth-Century Gaul”, Fifth-century Gaul: a crisis of identity?, ed. id. and H. Elton
(Cambridge 1992) pp. 208–17.
   369

celebrated in the Gododdin, lived in Lothian, beyond the “highland”

zone, and may well have been regarded as Picti by the Brittones of
the South.107 This said, however, it would be wrong to fall into the
trap of creating a very sharp division between a Romanized zone
and a “Celtic” zone.
Before moving on to investigate the evidence for Romanizing traits
in “highland” culture we should briefly note the evidence for a native
“Celtic” renaissance amongst the lowland Britons. This is extremely
slight and confined to personal names. Six personal names of British
origin attributed to sixth-century individuals are recorded in the
ninth-century Anglo-Saxon Chronicle: Cerdic, Cynric, Ceawlin, Conmægl,
Condidan and Farinmægl. Of these six names five appear to be of
British-Celtic origin and the sixth, Condidan, may be, or may, alter-
natively, be a rather peculiar development from Latin Constantinus.108
Two of these names appear with what might be contemporary or
near contemporary spelling (and apparently, it should be noted, using
British orthography).109 Cerdic, Cynric and Ceawlin seem to have
been modified during oral transmission.110 Unless Condidan is a form
of Constantinus there are no Roman names here. This is not con-
clusive evidence (the sample is far too small), but it does suggest a
community as aware of their British heritage as of their Roman.
More than twenty-five years ago John Ward discussed the re-
appearance of British names in fifth-century Britain, using as his
source material Welsh pedigrees and the ninth-century Historia
Brittonum.111 He pointed to the fact that the genealogical tradition

107
For the latest edition of the Gododdin, a collection of verses preserved in
medieval Wales which purport to be original elegies composed for warriors from
southern Scotland in the decades around 600 see The Gododdin of Aneirin: Text and
Context from Dark Age North Britain, ed. J.T. Koch (Cardiff 1997). Koch’s edition is
accompanied by extensive introductory material and a full bibliography.
108
K.H. Jackson discusses the name but does not commit himself in his Language
and History in Early Britain (Edinburgh 1953) p. 677.
109
Ibid., but cf. P. Sims-Williams, “Dating the Transition to Neo-Brittonic”, Britain
400–600: Language and History, ed. A. Bammesberger and A. Wollmann (Heidelberg
1990) pp. 217–61, here p. 245 with n. 104. Conmægl and Farinmægl are the names
in question, reproduced here in their most conservative forms from the “C” text
of the Anglo-Saxon Chronicle.
110
R. Coates, “On some controversy surrounding Gewissae/Gewissei, Cerdic and
Ceawlin”, Nomina 13 (1990) pp. 1–11; D. Parsons, “British *Caratîcos, Old English
Cerdic”, Cambrian Medieval Celtic Studies 33 (1997) pp. 1–8.
111
J. Ward, “Vortigern and the End of Roman Britain”, Britannia 3 (1972) pp.
277–89.
370  

frequently recorded pedigrees which began with Roman names and

went on, lower down, to use British names, noting, on the basis of
dead-reckoning, that the generation with British names but Roman-
named fathers appeared to have been born in the mid- to late fourth
century. His conclusion was that a “nativist” ideology had come into
being at that point amongst the British elite as a result of dissatis-
faction with the protection the Empire afforded Britain. In the light
of Dumville’s devastating critique of the value of central- and late-
medieval literary material for the elucidating of the sub-Roman
period,112 Ward’s findings must be viewed with the greatest of scep-
ticism, but his original question, “why did native names come back
into fashion?”, deserves consideration.
On the other hand, the “highland” zone was not entirely devoid
of Romanizing, or Romancing, traits. The inscribed stones do, after
all, almost all bear inscriptions in Latin. Some also bear inscriptions
in Irish but only one, at Tywyn,113 and that very late in the sequence,
bears an inscription in British (or rather Old Welsh by this stage).
“Moreover,” as Thomas Charles-Edwards writes, “the character of
the Latin used in the inscriptions shows that it was a spoken lan-
guage, not merely a language of the quill and chisel.”114 Charles-
Edwards goes on to argue that the epigraphic evidence can be used
to show how long Latin, or rather Romance, was “used in a wide
variety of styles and registers”,115 in contrast to its later medieval use
as a formal scholarly argot:
It is characteristic of the spoken Latin of Late Antiquity that the quan-
titative distinction between long and short vowels, to which differences
in the point of articulation were ancillary, disappeared. Instead, differences
in the point of articulation, formerly secondary, were now crucial. So
vivo gives Italian vivo but bibo gives Italian bevo, because a Latin short
i was more open than a long i and hence i > e, but i > i.116 The way
in which this worked itself out varied between the individual Romance
languages, but the direction of change was fundamentally the same.

112
D.N. Dumville, “Sub-Roman Britain: History and Legend”, History 62 (1977)
pp. 173–92.
113
V.E. Nash-Williams, Early Christian Monuments of Wales (Cardiff 1950), no. 287.
114
T.M. Charles-Edwards, “Language and society among the insular Celts, A.D.
400–1000”, The Celtic World, ed. M. Green (London 1995) pp. 703–36, here p. 704.
115
Ibid., p. 716.
116
Here Charles-Edwards cites V. Väänänen, Introduction au Latin vulgaire (Paris
1963) at §§ 42–6.
   371

At a rather earlier period the diphthongs, ae, oe, had become simple
vowels.117 Among the changes to the consonants is the disappearance
of final -m (very early) and -s (late and only in some areas of Romance,
including British Latin).118 The consequences for morphology were far-
reaching: there was no distinction between, for example, Petrus, Petrum,
Petro (all > Petro or Pedro). All these changes are attested in the British
Latin inscriptions of the fifth and sixth centuries:
VASSO for vassus, ADQUAE for atque119
CONGERIES for congerie120
CIVE for civis, CONSOBRINO for consobrinus121
MULTITUDINEM for multitudine122
[Charles-Edwards]123

A further example of Romance interference in the epigraphers’ Latin

is the indiscriminate use of second declension genitive singular end-
ing, without regard to syntax.124 Professor Charles-Edwards goes on
to contrast these kinds of flaws in the Latin of the Class I stones
with the types of mistakes made by medieval authors whose native
tongue was not Romance and who learned their Latin from gram-
mar-books, pointing out that while they may make mistakes in their
use of Latin cases, “many of those responsible for the texts of the
inscriptions were unaware of any case system at all. They had not
learnt their Latin from grammar” he goes on, and “Latin was, there-
fore, in the time of Voteporix [the mid-sixth century], a spoken lan-
guage alongside Welsh”.125
A much more widely discussed area of evidence for the existence
of Insular Romance lies in the relationship observable between Latin,
on the one hand, and Irish and Welsh on the other. The peculiar
orthography adopted to write these two languages, probably in the
sixth or seventh centuries, almost certainly reflects the way written
Latin was pronounced in Britain in this period.126 Thus, because

117
Ibid., § 59.
118
Ibid., §§ 127–9; Charles-Edwards also cites C. Smith, “Vulgar Latin in Roman
Britain: epigraphic and other evidence”, Aufstieg und Niedergang der römischen Welt II
29,2, ed. H. Temporini and W. Haase (Berlin 1983) pp. 893–948, here pp. 925–6.
119
Nash-Williams, Early Christian Monuments, no. 33.
120
Ibid., no. 101.
121
Ibid., no. 103.
122
Ibid., no. 78.
123
Charles-Edwards, “Language and society”, p. 716.
124
Ibid.
125
Ibid., p. 717.
126
See for example, J. Stevenson, “The beginning of literacy in Ireland”, Proceedings
372  

Latin d had undergone spirantization to become /ä/ in Insular

Romance, in certain medial and terminal positions, d was used to
render the phoneme /ä/ in medial and terminal positions in Old
Welsh and Old Irish where “original /d/” had never existed; e.g.
Middle Welsh mynyd from British *monijo-.127 The last twenty years
have seen a great deal of work carried out in this area,128 prefigured
by two seminal articles by David Greene,129 although we have yet
to see an accessible volume explaining the full implications of the
evidence for Insular Romance to non-linguists.
That the evidence does not simply indicate poor scholarly Latin
is apparent in the discussions produced by the various scholars whose
work has been cited. Were more evidence called for, we might indi-
cate Gildas’ De Excidio, written in fine literary language, and obvi-
ously intended to be understood by a secular and a lay audience.130
The one concession that Gildas may be making to Romance speak-
ers is his spelling of the name of the Magister Militum Aëtius, which
he renders “Agitius”, presumably expecting anyone reading the text
aloud to spirantize the g and render the name /Ahidius/ thus repro-
ducing an approximation of the value of the cluster aë, which is
extremely rare and might, perhaps, have been mistaken for the com-

of the Royal Irish Academy (C) 89 (1989) pp. 127–65; J.T. Koch, “When Was Welsh
Literature First Written Down?”, Studia Celtica 20/21 (1986) pp. 43–66; P. Sims-
Williams, “The Emergence of Old Welsh, Cornish and Breton orthography, 600–800:
the Evidence of Archaic Old Welsh”, Bulletin of the Board of Celtic Studies 38 (1991)
pp. 20–86. The subject is reviewed by Damian McManus in “Linguarum Diversitas:
Latin and the Vernaculars in Early Medieval Britain”, Peritia 3 (1984) pp. 151–88.
127
Sims-Williams, “Dating the Transition”, p. 223 n. 22. Spirantization also hap-
pened within the Insular Celtic languages and so d standing for /ä/ sometimes rep-
resents a development from an original /d/ and sometimes does not.
128
Most significantly A.S. Gratwick, “Latinitas Britannica: Was British Latin Archaic?”,
Latin and the Vernacular Languages in Early Medieval Britain, ed. N. Brooks (Leicester
1982) pp. 1–79; D.E. Evans, “Language Contact in pre-Roman and Roman Britain”,
Aufstieg und Niedergang des Römischen Welt II 29,2, ed. H. Temporini and W. Haase
(Berlin 1983) pp. 948–87; Smith, “Vulgar Latin”; D. McManus, “A Chronology of
Latin loan-words in early Irish”, Ériu 34 (1983) pp. 21–71; id., “Linguarum Diversitas”;
A. Harvey, “The Significance of Cothraige”, Ériu 36 (1985) pp. 1–9; P. Russell,
“Recent Work on British Latin”, Cambridge Medieval Celtic Studies 9 (1985) pp. 19–29.
129
D. Greene, “The Making of Insular Celtic”, Proceedings of the Second International
Congress of Celtic Studies (Cardiff 1966) pp. 121–36; id., “Some Linguistic Evidence
Relating to the British Church”, Christianity in Roman Britain, ed. M.W. Barley and
R.P.C. Hanson (Leicester 1968) pp. 75–86.
130
M. Lapidge, “Gildas’ Education and the Latin culture of sub-Roman Britain”,
Gildas: New Approaches, ed. id. and D.N. Dumville (Woodbridge 1984) pp. 27–50;
N. Wright, “Gildas’ prose style and its origins”, ibid., pp. 107–28.
   373

mon diphthong ae. The implication of this is that formal educated

Latin, at least as a written standard, existed alongside Insular Romance
and this implies that the Romance-speaking community was rela-
tively large, even in the West. We are not to imagine that Latin
usage was an affectation of a tiny number of kings and clerics but
that Insular Romance was the normal language of intercourse for a
significant proportion of the population. Perhaps we should imagine
a linguistic divide similar to that apparent in twelfth-century England
with Insular Romance playing the part of Old French.

Welsh Ethnogenesis

When considering the establishment of barbarian kingdoms on the

Continent one finds oneself musing on the survival of Germanic lan-
guage and the rate at which it was replaced by Romance as the
medium for elite discourse in the Western provinces. In Britain, pa-
radoxically, the decline of Romance and the adoption of British
Celtic by the elites is the parallel phenomena. In the present context
we might also consider that this transition is in some way related to
the establishment of a recognisably British gens separate from Gens
Romanorum. Thomas Charles-Edwards evades this thorny problem,
pointing out merely that by the ninth century the neo-Brittonic lan-
guages (principally Welsh) were certainly the languages of the elites
in the British World.131 Evidence for the exact timing of the death
of Insular Romance has not, so far, been identified, but this need
not mean that this evidence does not exist. One place in which we
might look for this evidence is in the literary and linguistic forms
that supplanted the Latin and Insular Romance traditions, and the
most likely location of those traditions is in early Welsh verse.
Historia Brittonum, after describing the taking of the fortress of
Bamburgh by the apical figure of the Bernician dynasty, Ida,132 goes
on to name five poets who were at that time “simultaneously famed
in British verse” (simul uno tempore in poemate Brittanico claruerunt): Tal-
haearn, Aneirin, Taliesin, Bluchbard and Cian.133 It is presumed from

131
Charles-Edwards, “Language and society”, p. 718.
132
Historia Brittonum 61, ed. J. Morris (Chichester 1980) p. 78.
133
Ibid. 62, p. 78.
374  

the context that these poets were northerners and that the time in
which they were famed was the time of Ida’s campaigns in and
around Bamburgh.134 Of the poetry of Talhaearn, Bluchbard and
Cian nothing now survives, but poetry ascribed to Aneirin and Tali-
esin is still extant. The veracity of that ascription is problematic.
The poems survive only in late medieval manuscripts and some mod-
ernising of the text has certainly taken place. While the poems are
clearly older than the manuscripts it is less clear how much older
they need be. A recent bold attempt at reconstructing the text of
Canu Aneirin (the Gododdin) has been undertaken by John Koch135 and
whilst his discussion of the historical context has come in for some
criticism136 there seems to be a willingness amongst most scholars to
believe that some kind of sixth- or seventh-century composition lies
at the core of this work.137 The supposed work of Taliesin has not
generated so much discussion in recent years and we are still very
much dependent upon the standard edition.138 Sir Ifor Williams
identified twelve poems, out of the myriad ascribed to Taliesin by
later medieval compilers, which he believed could be genuine sixth-
or seventh-century pieces. Dumville has contested the possibility that
three of these poems (I, II and VII) could be that old, but on con-
tentious historical grounds rather than on the basis of linguistic or
literary analysis.139
In a sense, the absolute dating and authenticity of the surviving
poetry, the hengerdd, of these earliest poets, the cynfeirdd, is not of
paramount importance to the argument that follows. What is signifi-
cant is that it was believed that the earliest poets in the Welsh tra-
dition were northerners and that those poems which have some claim
to being their work concern the north and were almost certainly
composed there. Only one of this group of poems seems to have been

134
But see now P. Sims-Williams, “The Death of Urien”, Cambrian Medieval Celtic
Studies 32 (1996) pp. 25–56.
135
The Gododdin of Aneirin.
136
E.g. O.J. Padel, “A New Study of the Gododdin”, Cambrian Medieval Celtic
Studies 35 (1998) pp. 45–56.
137
See the various papers collected in Early Welsh Poetry: Studies in the Book of
Aneirin, ed. B.F. Roberts (Aberystwyth 1988).
138
The Poems of Taliesin, ed. I. Williams (Dublin 1987). In fact this is a transla-
tion of the Welsh-language Canu Taliesin first published in 1960.
139
D.N. Dumville, “Early Welsh Poetry: Problems of Historicity”, Early Welsh
Poetry: Studies in the Book of Aneirin, pp. 1–16.
   375

composed for a southern king; the panegyric to Cynan Garwyn of

Powys. Cynan lived in the generation before 613 if we are to trust
the annalistic references to his son’s death at the battle of Chester
in that year.140 After this, the two earliest poems dealing with specifi-
cally Welsh subjects, in the geographical sense, appear to be Moliant
Cadwallon, not composed before 632, and Marwnad Cynddylan, per-
haps a decade or so later.141 From this time forth the Old North,
as it became known, provided the heroic ideal and model for sub-
sequent praise-poetry and englynion.142
Gildas describes, in unflattering terms, poets at the court of Mag-
locunus,143 and it has always been assumed that these poets sang
the king’s praises in British.144 In the light of the foregoing discus-
sion of Insular Romance it is perfectly possible, however, that they
recited Latin panegyrics. Indeed Patrick Sims-Williams has demon-
strated the influence of Latin panegyric diction and form on early
Welsh and, less certainly, on early Irish praise poetry, a phenome-
non which almost certainly requires composition and recitation of
Latin poetry to have taken place in sixth-century Britain.145 If court
poetry in the south was composed in Latin, for a largely Insular
Romance speaking audience, it could explain why, when the switch
to the vernacular as the language of court took place, northern mod-
els, and perhaps even northern bards were called upon. It seems
incredible that no verses survive in praise of Maglocunus, even if
only in fragmentary form or in citation in other works, if such poetry,
in the style of that ascribed to Taliesin and Aneirin, had existed.
The poetry of twelfth- and thirteenth-century Venedotian court poets
contain no allusions to verse celebrating sixth-century kings of Gwy-
nedd, yet are self-consciously modelled upon the verses ascribed to
Taliesin, suggesting that the sixth-century Venedotian poetry did not

140
Annales Cambriae s.a. 613; Annals of Ulster (to A.D. 1131) 613,3, ed. S. Mac Airt
and G. Mac Niocaill (Dublin 1983).
141
M. Haycock, “Metrical models for the poems in the Book of Taliesin”, Early
Welsh Poetry: Studies in the Book of Aneirin, pp. 155–78, here p. 171.
142
J. Rowland, Early Welsh Saga Poetry: a Study and Edition of the Englynion (Woodbridge
1990).
143
Gildas, Ruin of Britain 34,6; 35,3.
144
Charles-Edwards, “Language and Society”, p. 734.
145
P. Sims-Williams, “Gildas and vernacular poetry”, Gildas: New Approaches, ed.
M. Lapidge and D.N. Dumville (Woodbridge 1984) pp. 169–90.
376  

survive in their day. Maglocunus’ position in De Excidio and his pro-

minent role in subsequent Welsh tradition, and as the apical figure
of the Venedotian dynasty down to the end of the eighth century,146
make him the most famous of all Welsh kings before Rhodri Mawr
(† 878). Had praise poetry composed for him in Welsh existed we
might have expected some hint of it to survive.
In broad terms, then, the switch from Insular Romance as the pre-
ferred language of a Romanizing elite to the language of the country,
Cymraeg,147 will have taken place in the period between the emergence
of the earliest vernacular praise poetry and its adoption throughout
the British-speaking world. In absolute terms we should probably think
of a transition starting in the North c. 550,148 taking root in Wales
before the middle of the seventh century and reaching Dumnonia
by the early eighth at the very latest.149
Seen in this chronological context, the impetus for the abandon-
ment of Insular Romance in the West can be seen to coincide with
the conquest of the majority of the eastern, lowland, Britons by the
Anglo-Saxons in the same time-period (c. 550–700). In these terms
the long survival of Insular Romance in the West, for some two
hundred years after the disappearance of effective Roman power and
centralised government, can be re-assessed. The western and north-
ern kingdoms maintained their Romanizing character so long as they
lay adjacent to a large contiguous zone in which Romance culture
throve. To some extent, the kings and their courts must have seen
themselves as peripheral to and dependent upon, in cultural terms

146
Dumville, “Gildas and Maelgwn”, p. 58.
147
Charles-Edwards, “Language and Society”, pp. 710–5.
148
Dumville, “Early Welsh Poetry”; J.T. Koch, “The Cynfeirdd poetry and the
language of the sixth century”, Early Welsh Poetry: Studies in the Book of Aneirin, pp.
17–42.
149
The latter date is reached on the assumption of the authenticity of poem 21
in the Black Book of Carmarthen, an elegy to Geraint fil Erbin, a king of Dumnonia
(Dyfnaint). This may be the Geraint to whom Aldhelm wrote and against whom Ine
fought (Yorke, Wessex in the Early Middle Ages, p. 179 and 65 respectively). Alternatively,
given the poem’s account of Geraint’s death at Llongborth (Langport, on the river
Parret in Somerset), he may have been the leader of the force against whom
Cenwealh of the Gewisse fought (Anglo-Saxon Chronicle: Two of the Saxon Chronicles
Parallel 658, ed. C. Plummer [Oxford 1899] when he drove the Britons to the
Parret. This latter identification would fit better with the pedigrees, which make
Geraint the grandson of Custennin son of Cynfawr, who may, as we have seen,
been a contemporary of Gildas (Early Welsh Genealogical Tracts, p. 45).
   377

at least, the lowlands. This should not surprise us. Despite the absence,
as yet, of a rich material culture emanating from amongst the low-
land British, their land was far more productive and the density of
affluent, educated aristocrats must have been much higher. Watching
this world collapse before them, and finding the new rulers of the
lowlands uninspiring, the kings of the West must have turned towards
their own countrymen and to the language of local people for cul-
tural comfort.
This reorganisation of identity will not have happened overnight.
At the time it may well have seemed almost imperceptible. For the
most part people capable of speaking both languages well will have
simply begun to switch their preference from the one to the other.
After a generation or two children would have grown up without
sufficient exposure to Insular Romance to perpetuate it. An inter-
esting question to consider is whether this switch had any effect on
the character of the Welsh language.
Kenneth Jackson, in 1953, thought that most of the borrowings
from Latin into British (between 500 and 900 items) had occurred
prior to A.D. 400.150 Subsequent scholarship in his own and adjacent
fields has, however, shown that he was imposing needless constraints
upon the data.151 Recognition that British Latin (Insular Romance)
developed, at least in part, in tandem with British Celtic,152 has at
once freed us from such constraints and left us somewhat at a loss
for chronological markers in the development of either British or
Insular Romance historical phonology. Going beyond lexical bor-
rowings, David Greene noted “a number of vague parallels between
Vulgar Latin [Romance] and British (but not, significantly, Irish) [. . .]
and there may be some connection [. . .]. If so, it would be an ‘areal’
development, not a matter of Vulgar Latin loan words”.153 Unfor-
tunately no systematic research has been carried out in this area.154

150
Jackson, Language and History, pp. 76–121.
151
Sims-Williams, “Dating the Transition”, pp. 223–36.
152
E.g. Koch, “The Cynfeirdd poetry”, pp. 23–5.
153
Greene’s original comments were in “The Making of Insular Celtic” but they
are summarised here by Sims-Williams in “Dating the Transition”, at pp. 218–9.
154
The revised position outlined in this paragraph would seem to allow for Peter
Schrijver’s claims for British phonological interference in North Sea Germanic, (out-
lined in his “The Celtic Contribution to the Development of the North Sea Germanic
Vowel System”, North-Western European language evolution 35 [1999] pp. 3–47), to be
derived from either British Celtic or British Latin.
378  

If we compare the development of British and Irish, however, we

can note some very basic features that may arise out of a super-
strate situation. The neo-Brittonic transition, during which British
was transformed into Old Welsh, Cornish, Cumbric and Breton,
comprised a number of stages; lenition (the softening of medial and
terminal consonants), apocope (the loss of final syllables), syncope of
medial vowels, spirantization and the development of a new quan-
tity system, in that order.155 Sims-Williams would modify this slightly,
dividing “lenition” into two stages “first spirantization” of b, d, g and
m, and then “voicing” of p, t and k.156 Similar, though not in all
cases identical, changes occurred in the transition from Primitive
Irish (the language of the bulk of ogham inscriptions) to Old Irish
(the language of the earliest manuscripts). The transformations in
Irish probably occurred at a slightly later date than the comparable
ones in British, although the two sequences seem to have over-
lapped.157
Despite the similarity of the phonological transformation of the
two languages, however, the morpho-syntactical structures failed to
keep pace with one another. In Welsh the neo-Brittonic transition
is characterised by the complete collapse of the case system and a
massive simplification of the verbal system. This has been ascribed
to the effects of apocope,158 but since apocope occurred also in Irish
which retained much of its grammatical complexity this seems
insufficient explanation in itself.159 One possibility, though at present
unprovable, is that this was, in part at least, the result of a massive
switch to Old Welsh by preferential Romance speakers who, as the
social elite, exerted influence on subsequent development of the lan-
guage in disproportion to their numbers.
Turning to the archaeological record it is interesting to note that
the switch from Insular Romance to Welsh as the language of the
elite coincides with the falling off of imported ceramics from the
Mediterranean and western Gaul.160 Ewen Campbell has suggested

155
Jackson, Language and History, pp. 695–6.
156
Sims-Williams, “Dating the Transition”, pp. 220–1.
157
Ibid., pp. 230–6.
158
Jackson, Language and History, p. 691.
159
Cf. D.E. Evans, “Insular Celtic and the Emergence of Welsh”, Britain 400–600:
Language and History, ed. A. Bammesberger and A. Wollmann (Heidelberg 1990) pp.
149–77.
160
A. Lane, “Trade, Gifts and Cultural Exchange in Dark Age Western Scotland”,
   379

that it is the ending of this exchange which causes various trans-

formations in the archaeological record of western Britain, such as
the abandonment of central fortified sites in Cornwall, and perhaps
Wales, and widespread changes in ecclesiastical art and architec-
ture.161 An alternative scenario might locate agency with the Britons,
since in the Gaelic World these overseas contacts continued and even
expanded, suggesting that there was no problem with supply.162 If
supply was not in issue then it is perhaps demand that we should
consider.
The adoption of a vernacular identity may also, as noted when dis-
cussing the Llandaff material, represent a shift in ideology. Distinctive
sociolects, and even more so language differentiation on the basis of
class, bespeaks a highly stratified society in which the aristocracy are
primarily exploitative. This is probably one reason why the bulk of
the lowland British population are so hard to identify in the archae-
ological record; their land-lords did not leave them with much to
leave us. The social transformation in the West, c. 550–650, can be
seen as the replacement of an exploitative consuming aristocracy,
which had developed under the aegis of the Empire, by a series of
localised, kin-based, redistributive chieftaincies, much like those which
were simultaneously transforming themselves into stratified societies
in Anglo-Saxon England.

Conclusion

Thomas Charles-Edwards has pointed out that the position of Irish

in Dyfed may well have been analogous to that of Insular Romance
throughout the highland zone.163 He points to the fact that the Class
I stones are mostly in Latin, but that a few, principally in western
Dyfed, southern Ceredigion and Brycheiniog, were also in Primitive
Irish. This is in stark contrast to the absence of British inscriptions.
He infers from this, and from the failure of Irish to survive as the

Scotland in Dark Age Britain, ed. B.E. Crawford (St Andrews 1994) pp. 103–15, here
p. 107.
161
E. Campbell, “Trade in the Dark Age West: a Peripheral Activity”, Scotland
in Dark Age Britain, ed. B.E. Crawford (St Andrews 1996) pp. 79–93, here pp. 86–7.
162
Ibid., p. 86.
163
Charles-Edwards, “Language and Society”, p. 704.
380  

language of south-west Wales, despite the retention of a royal dynasty

of Irish origin, that the position of Irish was similar to that of Latin
or Romance. This linguistic history of the Demetian kingdom is of
course typical of the histories of the barbarian kingdoms of Late
Antiquity: a military aristocracy takes over the government of a
region and whilst maintaining their separate identity and language
for a generation or two they gradually assimilated with the local
population, adopting their language and belief systems.164 The same
thing happened with the Franks in most of Gaul, the Goths in Spain
and the Langobards in Italy. In much of western and northern Britain
the native Roman aristocracies seem to have undergone a similar
process so that in many respects, material, sociological and linguis-
tic, the ethnogenesis of the Britons and the establishment of British
regna can be seen as a descent in barbarism.

164
For a fuller discussion of the Demetian kingdom see Thomas, Mute Stones, pp.
51–112, and Charles-Edwards, “Language and Society”, pp. 717–36.
 ANGLO-SAXON GENTES AND REGNA

Barbara Yorke

Problems and sources

A century ago the question of the circumstances surrounding the

emergence of Anglo-Saxon kingdoms appeared to be quite adequately
answered by Gildas’s account of the adventus Saxonum with some addi-
tions by Bede.1 According to a combined account from these writ-
ers, after Britain’s ties with Rome had been severed, Vortigern and
his council hired Saxons who were attacking the east coast to fight
for them, but the Saxon forces eventually grew strong enough to
seize power for themselves. Bede located these events in c. 450, and
claimed that the first kingdom founded as an immediate result of
the coup was that of Kent. At the beginning of the twentieth cen-
tury this appeared to be a believable, straightforward account, and
one with possible parallels in other areas of the former Roman Empire
in the west. Unfortunately not one aspect of it has emerged unscathed
in more recent historiography,2 and it is no longer generally seen as
an adequate explanation for the changes that occurred in eastern
Britain during the fifth and sixth centuries, let alone for the emer-
gence of Anglo-Saxon kingdoms. Even though it has been generally
accepted that Gildas’s account must have some validity, uncertainty
about when and where he wrote, and over whether events he de-
scribed took place in one part of the country consecutively or con-
temporaneously in different parts of the country, not to mention the
problems in interpreting his highly rhetorical style and exhortatory
reasons for writing, have all tended to cast doubt over the wisdom

1
Gildas, The Ruin of Britain and other works, ed. and transl. M. Winterbottom
(London-Chichester 1978) pp. 97–8; Bede, Historia Ecclesiastica gentis Anglorum 1,15,
ed. B. Colgrave and R.A.B. Mynors [Bede’s Ecclesiastical History of the English People]
(Oxford 1969) pp. 49–53.
2
P. Sims-Williams, “Gildas and the Anglo-Saxons”, Cambridge Medieval Celtic Studies
6 (1983) pp. 1–30; id., “The settlement of England in Bede and the Chronicle”, Anglo-
Saxon England 12 (1983) pp. 1–41.
382  

of placing too much reliance on his relatively brief narrative passage

concerning the Saxon adventus.3 Bede stands suspected of promoting
Kentish and archiepiscopal interests in assigning the leadership of
events to the supposed Kentish founding fathers, Hengist and Horsa,4
or of furthering the idea of an united Christian gens Anglorum through
an account that assigned a common origin to all the Anglo-Saxon
peoples of Britain.5 With other written references to events in east-
ern Britain between c. 450 and 590 being only brief, and so throw-
ing little light on what had occurred,6 the onus for studying this
period has fallen on archaeology with claims that in the absence of
adequate written records fifth- and sixth-century Britain should be
treated as a “prehistoric” period.7 The result has been to suggest
that the immediate post-Roman history of eastern Britain followed
a very different trajectory from that of other former provinces of the
Roman Empire. In the initial planning of the symposium that gave
rise to this volume, there was some doubt over whether the Anglo-
Saxon kingdoms qualified for inclusion, so deep-rooted is the cur-
rent assumption that the severance of Britain from the empire in
c. 410 was a decisive cut-off point so that when Anglo-Saxon king-
doms finally emerged at the end of the seventh century it was in a
context that owed little or nothing to Britain’s Roman heritage.
Behind much archaeological interpretation one can trace the
influence of a model that assumes a complete systems collapse of
the former Roman Empire and its supporting structures in eastern
Britain.8 Through explicit or implicit use of this model it has been
possible to assume that a new society emerged, composed of some
Germanic incomers (the scale of immigration is still hotly debated)
and of formerly oppressed British farmers, that lived for a while

3
See, for instance, discussions within Gildas: New Approaches, ed. M. Lapidge and
D.N. Dumville (Woodbridge 1984) and N. Higham, The English Conquest. Gildas and
Britain in the Fifth Century (Manchester 1994).
4
M. Miller, “Bede’s use of Gildas”, English Historical Review 90 (1975) pp. 241–61,
here p. 254.
5
N. Brooks, Bede and the English ( Jarrow 1999) pp. 4–5.
6
D.N. Dumville, “Sub-Roman Britain: History and Legend”, History 62 (1977)
pp. 173–92.
7
C. Arnold, The Archaeology of Anglo-Saxon Kingdoms (2nd edn., London 1997).
8
C. Renfrew, “Post collapse resurgence: culture process in the Dark Ages”,
Ranking, Resource and Exchange. Aspects of the Archaeology of Early European Society, ed.
C. Renfrew and S. Shennan (Cambridge 1982) pp. 113–5.
  - 383

without hierarchies beyond those that could be found in individual

households. What the archaeological evidence seems to show is that
the bulk of the population lived in rural hamlets where the indi-
vidual farmsteads were characteristically uniform in size and form.9
Cemeteries were organised into burial plots based around these house-
holds where differential burial might reflect variations in status due
to age, gender and economic role within the farmstead unit.10 A
clear change occurs in the character of the evidence in the latter
part of the sixth century, with a much more marked demonstration
of hierarchy in both settlement sites and cemeteries,11 that seems to
coincide with the point at which the Anglo-Saxon kingdoms known
from written sources of the seventh and eighth centuries appear on
the historical horizon.12 An influential archaeological model used to
explain how kingdoms could have emerged from the starting-point
of scattered farmstead units is that of peer-polity interaction.13 It is
a model that ultimately derives from anthropological study of very
diverse groups from outside the western European world.14
Archaeology can undoubtedly reveal much about society in the
eastern half of England in the fifth and sixth centuries, but there

9
D. Powlesland, “Early Anglo-Saxon settlements, structures, form and layout”,
The Anglo-Saxons from the Migration Period to the Eighth Century: An Ethnographic Perspective,
ed. J. Hines (Woodbridge 1997) pp. 101–24; H. Hamerow, “The earliest Anglo-
Saxon kingdoms”, New Cambridge Medieval History, vol. 1: c. 500–700 (Cambridge
forthcoming).
10
H. Härke, “Early Anglo-Saxon social structure”, The Anglo-Saxons from the Migration
Period to the Eighth Century: An Ethnographic Perspective, pp. 125–60; N. Stoodley, The
Spindle and the Spear: A Critical Enquiry into the Construction and Meaning of Gender in the
Early Anglo-Saxon Inhumation Burial Rite, British Archaeological Reports, British series
288 (Oxford 1999).
11
J. Shephard, “The social identity of the individual in isolated barrows and bar-
row cemeteries in Anglo-Saxon England”, Space, Hierarchy and Society, ed. B. Burnham
and J. Kingsbury, British Archaeological Reports, International series 59 (Oxford
1979) pp. 47–79; C. Scull, “Social archaeology and Anglo-Saxon kingdom origins”,
The Making of Kingdoms, ed. T. Dickinson and D. Griffiths, Anglo-Saxon Studies in
Archaeology and History 10 (Oxford 1999) pp. 7–24.
12
For instance, the evidence of the royal genealogies is apparently consistent with
the first royal houses coming into existence in the latter half of the sixth century;
D.N. Dumville, “Kingship, genealogies and regnal lists”, Early Medieval Kingship, ed.
P.H. Sawyer and I.N. Wood (Leeds 1977) pp. 72–104, here p. 91.
13
For a clear, balanced acount of how the model might have worked in Anglo-
Saxon England see C. Scull, “Archaeology, early Anglo-Saxon society and the ori-
gins of kingdoms”, Anglo-Saxon Studies in Archaeology and History 6 (1993) pp. 65–82.
14
See, for instance, T. Earle, “Chiefdoms in archaeological and historical per-
spective”, Annual Review of Anthropology 16 (1987) pp. 279–308.
384  

are grounds for asking whether it is capable of providing the whole

story. One of the advantages of the type of comparative study that
has given rise to this volume is that it provides an opportunity to
review aspects of the emergence of Germanic kingdoms in eastern
Britain in the context of potentially comparable developments in
other areas of the western empire, and so to reopen the question of
a link between the Anglo-Saxon kingdoms and the late Roman dio-
cese of Britain. Historians have on the whole been more concerned
than archaeologists about an apparent divergence between the his-
tory of Britain and that of other former Roman provinces in the
fifth and sixth centuries, and with the concept of an apparent power
vacuum between c. 450–c. 550.15 The picture drawn from the archaeo-
logical evidence can also begin to sound dangerously like the Tacitean
society of Germanic free peasants so in vogue in the nineteenth cen-
tury, and the competition for power in the peer-polity model like a
Darwinian battle for the survival of the fittest compressed into a rel-
atively short time-span. The furnished cemeteries of the late fifth
and sixth centuries have been a natural focus of research, especially
in the last two decades, but they do not necessarily represent the
practices of the whole of society. For instance, not only are there
unfurnished burials within these cemeteries, but also entire ceme-
teries without grave-goods in the eastern half of England that seem
to have been in use during the same period, suggesting that modes
of life (and death) were not as uniform as is sometimes implied.16 It
may be that archaeology shows us only a partial picture, and it is
a moot point whether political systems always have an impact that
is discernible in the archaeological record. How much of the com-

15
E. James, “The origins of barbarian kingdoms: the continental evidence”, The
Origins of Anglo-Saxon Kingdoms, ed. S. Bassett (Leicester 1989) pp. 40–52; G. Halsall,
“The origins of Anglo-Saxon kingdoms: a Merovingianist speaks out” (unpublished
lecture, Institute of Historical Research, London 1995).
16
R.A. Chambers, “The Late- and Sub-Roman Cemetery at Queensford Farm,
Dorchester-on-Thames, Oxon.”, Oxoniensia 52 (1987) pp. 35–69; T. Williamson, The
Origins of Hertfordshire (Manchester 2000) p. 70 (Verulamium). Such unfurnished ceme-
teries are, of course, difficult to date without recourse to scientific techniques like
radio carbon dating (used at Queensford farm) and so many similar cemeteries may
have failed to be assigned to the correct period or before the twentieth century
may not even have been recorded. Similarly there is evidence for occupation dat-
ing to fifth and sixth centuries at late Roman villas and other Romano-British sites
which has produced little that is distinctively “Germanic”—see, for instance, K.R.
Dark, Britain and the End of the Roman Empire (Stroud 2000) pp. 62–9.
  - 385

plexities of Frankish history in the late fifth and sixth centuries could
have been constructed from the archaeological record alone? The
lack of detailed written accounts is something that means there will
always be uncertainty over many of the aspects surrounding the
emergence of Anglo-Saxon gentes and regna, and that some of the
questions posed by the organisers of the symposium will have to go
unanswered. The issue of some of the general assumptions about
developments in the eastern half of Britain in the fifth and sixth cen-
turies, fuelled by recent archaeological studies, will be central to the
discussions that follow.

Angles, Saxons and Jutes: gentes before and after migration to Britain

Bede’s gloss on Gildas’s description of the Anglo-Saxon adventus

described the new arrivals as coming de tribus Germaniae populis for-
tioribus, id est Saxonibus, Anglis, Iutis.17 It has been noted that Bede’s
Latin vocabulary in this passage differs in a number of respects from
that he customarily used in the Historia Ecclesiastica, leading to the
suggestion that it represents an addition he received, perhaps from
Canterbury, at a relatively late stage in the work’s composition.18
Nevertheless populi is a word that he employed, together with natio,
as a synonym for gens.19 Whether Bede composed the passage or not,
we can take it that he accepted that the Angles, Saxons and Jutes
could be identified as distinct gentes. Indeed, earlier in the same chap-
ter he wrote of the Anglorum sive Saxonum gens when adapting Gildas’s
text. The existence of peoples called Angles, Saxons and Jutes on
the North Sea littoral, in what is now northern Germany and south-
ern Denmark,20 as well as in Britain naturally led in early studies of
the fifth and sixth centuries to the expectation that there had been
a simple migration of gentes.21 However, more recent considerations

17
Bede, Hist. Eccl. 1,15, pp. 50–1.
18
E. John, “The point of Woden”, Anglo-Saxon Studies in Archaeology and History 5
(1992) pp. 127–34, here p. 129; Brooks, Bede and the English, pp. 11–2.
19
P.F. Jones, A Concordance to the Historia Ecclesiastica of Bede (Cambridge Mass.
1929) pp. 311–2; 403–4; J. Campbell, Bede’s Reges et Principes ( Jarrow 1979) pp. 3–4.
20
I.N. Wood, “Before and after migration to Britain”, The Anglo-Saxons from the
Migration Period to the Eighth Century: An Ethnographic Perspective, pp. 41–64.
21
See, for instance, J.N.L. Myres, “The Angles, the Saxons and the Jutes”,
Proceedings of the British Academy 56 (1970) pp. 1–32.
386  

have suggested that matters were not so straightforward and that, as

seems to have been the case with several other Germanic peoples
discussed in this volume, continuity of name may conceal major
changes in the constitution of a gens after migration from the home-
land and settlement in a Roman province.
Archaeological evidence from the fifth century, in the form of arte-
facts, funeral rituals and building traditions, shows without doubt
that there was migration to eastern and southern England from the
traditional homelands of the Anglo-Saxons in north Germany and
southern Scandinavia, even if the scale of migration is still a matter
of debate.22 Language change in the eastern half of England whereby
Old English dialects appear to have totally replaced Latin and Celtic
as the lingua franca also supports the idea of significant migration even
if there is also debate over the scale and social level of migration
needed to bring about such change.23 In the eastern (Anglian) dis-
tricts of England there was a predominance of people who appear
to have come from continental Angeln in modern Schleswig-Holstein,
and even some archaeological and environmental support for Bede’s
claim that the area was heavily depopulated as a result.24 But Angles
were not the only Germanic migrants into the area. Böhme’s map-
ping of brooches from the Saxon homelands shows a wide distrib-
ution throughout eastern and southern England, and that they were
not confined to the areas that were identified as “Saxon” in later
kingdom names.25 A broader span of archaeological evidence sug-

22
J. Hines, “The becoming of the English: identity, material culture and lan-
guage in early Anglo-Saxon England”, Anglo-Saxon Studies in Archaeology and History 7
(1994) pp. 49–59; C. Scull, “Approaches to material culture and social dynamics
of the migration period of eastern England”, Europe Between Late Antiquity and the
Middle Ages, ed. J. Bintliff and H. Hamerow, British Archaeological Reports, Inter-
national series 617 (Oxford 1995) pp. 71–83; Hamerow, “The earliest Anglo-Saxon
kingdoms”.
23
J. Hines, “Philology, archaeology and the adventus Saxonum vel Anglorum”, Britain
400–600: Language and History, ed. A. Bammesberger and A. Wollmann (Heidelberg
1990) pp. 17–36; M. Gelling, “Why aren’t we speaking Welsh?”, Anglo-Saxon Studies
in Archaeology and History 6 (1993) pp. 51–6; T.M. Charles-Edwards, “Language and
society among the insular Celts, A.D. 400–1000”, The Celtic World, ed. M. Green
(London 1995) pp. 703–36.
24
M. Gebühr, “Angulus desertus?”, Studien zur Sachsenforschung 11, ed. H.-J. Häß-
ler (Oldenburg 1998) pp. 43–8; Hamerow, “The earliest Anglo-Saxon kingdoms”.
25
H.W. Böhme, “Das Ende der Römerherrschaft in Britannien und die angel-
sächsische Besiedlung Englands im 5. Jahrhundert”, Jahrbuch des Römisch-Germanischen
Zentralmuseums Mainz 33 (1986) pp. 469–574.
  - 387

gests that people were migrating to Britain in the fifth century from
other Germanic areas besides those of the Angles, Saxons and Jutes.
Franks, Frisians and Norwegians are among those who appear to
have entered the country.26 Bede also seems to have been aware that
many different Germanic people may have migrated to Britain for
in a later chapter he wrote, of an individual involved in sending
missions to Germania, that in Germania plurimas noverat esse nationes, a
quibus Angli vel Saxones, qui nunc Britanniam incolunt, genus et originem duxisse
noscuntur.27 There has been much debate about problems involved in
the interpretation and validity of the list that Bede provides which
names Frisians, Rugians, Danes, Huns, Old Saxons and Bructeri.28
Whether the peoples have been correctly identified or not, what is
of central interest is that Bede appears to claim that Anglian and
Saxon identities in Britain emerged as a secondary development as
a result of various different Germanic nationes migrating to the province.
Archaeological evidence from the sixth century seems to be broadly
in agreement with this claim. Material culture and traditions of build-
ing and burial do not seem to have been translated wholesale from
any one area of Germany to eastern Britain.29 Rather former ways
of doing things were adapted and new hybridised forms emerged
that drew not only on disparate North Sea backgrounds, but also
probably on Romano-British skills and practices, though there are
differing views on the Romano-British contributions to such aspects
as the Anglo-Saxon building tradition or the quoit brooch style.30

26
Hines, “The becoming of the English”; H. Hamerow, “Migration theory and
the Anglo-Saxon ‘identity crisis’”, Migrations and Invasion in Archaeological Explanation,
ed. J. Chapman and H. Hamerow, British Archaeological Reports, International
series 664 (1997) pp. 33–44.
27
“He knew that there were very many peoples in Germany from whom the
Angles and Saxons, who now live in Britain, derive their descent and origin”; Bede,
Hist. Eccl. 5,9, pp. 476–7.
28
W. Pohl, “Ethnic names and identities in the British Isles: a comparative per-
spective”, The Anglo-Saxons from the Migration Period to the Eighth Century: An Ethnographic
Perspective, pp. 7–40, here pp. 14–6; Wood, “Before and after migration to Britain”,
pp. 44–5; 55–8.
29
P. Sørensen, “Jutes in Kent? Considerations of the problem of ethnicity in
southern Scandinavia and Kent in the migration period”, Method and Theory in
Historical Archaeology—Papers of the Medieval Brugge 1997 Conference, ed. G. de Boe and
F. Verhaeghe (Zellik 1997) pp. 165–73; C. Hills, “Did the people of Spong Hill
come from Schleswig-Holstein?”, Studien zur Sachsenforschung 11, pp. 145–54.
30
A. Marshall and G. Marshall, “Differentiation, change and continuity in Anglo-
388  

These trends are exemplified by what had emerged by the sixth cen-
tury as the dominant female dress-set for wealthier women in the
Anglian areas of eastern England (as broadly identified by the later
Anglian kingdom names). It built upon continental Anglian forms,
but with modifications and additions drawn from traditions of dress
from other Germanic areas, such as wrist-clasps which were appar-
ently otherwise only a regular female dress accessory at this time in
western Norway.31 Not all individual aspects of the Anglian dress-set
were restricted to areas within the later Anglian borders, but wrist-
clasps are largely confined to it, and the combination and arrange-
ment of items is also distinctive.32 Jutish areas also favoured their
own forms of female dress and had grave-assemblages that included
items rarely found elsewhere, including items of southern Scandinavian
and Frankish origin that influenced the development of distinctive
local forms.33 Women of the Saxon areas of southern England also
seem to have had preferences that went back to ancestral forms, but
Saxon female costume was more understated in its Germanness than
that in Anglian or Jutish areas and its brooches may have been more
indebted to Romano-British prototypes.34 It is not the case, of course,
that all female burials conformed uniformly to provincial dress-codes.
Many more localised variations in patterns of dress and wearing of
jewellery can be identified, and individual cemeteries might have

Saxon buildings”, The Archaeological Journal 150 (1994) pp. 366–402; H. Hamerow,
“The archaeology of rural settlements in early medieval Europe”, Early Medieval
Europe 3 (1994) pp. 167–79; B. Ager, “The alternative quoit brooch: an update”,
Anglo-Saxon Cemeteries: A Reappraisal, ed. E. Southworth (Stroud 1990) pp. 153–61;
G. Harrison, “Quoit brooches and the Roman-Medieval transition”, Theoretical Roman
Archaeology Conference, ed. P. Baker (Oxford 1999) pp. 108–20.
31
J. Hines, The Scandinavian Character of Anglian England in the Pre-Viking Period,
British Archaeological Reports, British series 124 (Oxford 1984); id., “The Scandinavian
character of Anglian England: an update”, The Age of Sutton Hoo, ed. M. Carver
(Woodbridge 1992) pp. 315–29.
32
S. Lucy, The Anglo-Saxon Way of Death (Stroud 2000) pp. 133–9.
33
B. Brugmann, “Britons, Angles, Saxons, Jutes and Franks”, The Anglo-Saxon
Cemetery on Mill Hill, Deal, Kent, ed. K. Parfitt and B. Brugmann, Society for Medieval
Archaeology, Monograph series 14 (London 1997) pp. 110–24; C. Behr, “The ori-
gins of kingship in early medieval Kent”, Early Medieval Europe 9 (2000) pp. 25–52;
N. Stoodley, “Female assemblages from southern Hampshire” (unpublished semi-
nar paper, Institute of Archaeology, London 2001).
34
T. Dickinson, “Early Saxon saucer brooches: a preliminary overview”, Anglo-
Saxon Studies in Archaeology and History 6 (1993) pp. 11–44; ead., “Material culture as
social expression: the case of Saxon saucer brooches with running spiral decora-
tion”, Studien zur Sachsenforschung 7, ed. H.-J. Häßler (Hildesheim 1991) pp. 39–70.
  - 389

their own distinctive burial rituals.35 Many women were buried with-
out any indications of specific dress-forms that have survived to the
present day, and it would appear that it was only certain women of
certain households who wore the more elaborate forms of Tracht
including the distinctive paired brooches on the shoulders of a pinafore-
type dress. But with these provisos we can say that there were broad
principles in variations of dress between the Anglian, Saxon and
Jutish areas through which elites of local households appear to have
been signalling some form of commonality. Similar results emerge
from the study of language. Bede noted dialectical differences between
the Anglian and Saxon areas, and arguments have been advanced
for a distinctive Jutish dialect as well.36 But Old English generally
does not seem to represent a simple importation of language from
one area of Germany, but a more complex amalgam caused by the
type of intermingling of North Sea Germanic-speaking peoples that
has already been discussed.37
The Angles, Saxons and Jutes therefore fit into a pattern observ-
able elsewhere in this volume. It seems likely that there existed gentes
with these names before settlement in Britain, but that the process
of migration with its intermingling of Germanic peoples of diverse
origin (not forgetting the encounters with Romano-British which
remain a largely unknown entity) led to the emergence of newly
defined peoples bearing these names in the eastern half of England.
Their history is one strand in a constantly evolving pattern of gentes
that characterises the early middle ages.38 By the end of the sixth
century new identities as gentes appear to have been developing in
eastern England with the appearance among the broader Anglian,
Saxon and Jutish groupings of individual kingdoms that were each
characterised as a gens.39 This development seems to have been accom-
panied by abandonment of some of the forms of costume and choice

35
G. Fisher, “Kingdom and community in early Anglo-Saxon eastern England”,
Regional Approaches to Mortuary Analysis, ed. L. Anderson Beck (New York 1995) pp.
147–66; S. Lucy, The Early Anglo-Saxon Cemeteries of Eastern Yorkshire. An Analysis and
Re-interpretation, British Archaeological Reports, British series 272 (Oxford 1998);
Stoodley, The Spindle and the Spear.
36
E. Seebold, “Was ist jütisch? Was ist kentish?” Britain 400–600: Language and
History, pp. 335–52.
37
Hines, “Philology, archaeology and the adventus”.
38
S. Reynolds, Kingdoms and Communities in Western Europe, 900–1300 (Cambridge
1984).
39
B.A.E. Yorke, “Political and ethnic identity: a case study of Anglo-Saxon
390  

of grave-furnishings that it has been argued characterised the ear-

lier gentes in favour of forms that can be described in broad terms
as more Romanised.40 In particular, the rulers of the new kingdoms
drew upon ways of signalling royal power drawn ultimately from
late Antique society, though the choice of items from different facets
and periods of the late Antique world, often combined with “Germanic”
symbols of power produced idiosyncratic assemblages of material.41
The eclecticism of the new Anglo-Saxon elite culture ultimately
opened the way for the adoption of Christianity that proved most
effective in underpinning the new regimes, including through the
recording of versions of their early history.42

The role of the king in the establishment of gens and regnum

In the kingdoms established in the eastern half of Britain in the lat-

ter part of the sixth century the written evidence suggests that the
concepts of royal house, regnum and gens were closely entwined. The
vocabulary of the Historia Ecclesiastica shows that when writing of the
twelve or so kingdoms ( provinciae or regna) of the seventh century Bede
assumed that they could be equated with gentes.43 To refine matters
further, the natural order to Bede, following conversion, was an equa-
tion of gens, regnum and bishopric, and he was particularly exercised
by the problem of gentes who did not have their own bishop. He
clearly believed that the Jutes of Wight, whom he identified as inhab-
iting a distinct provincia and as having their own royal line, were enti-
tled to their own bishopric, even after the royal house was annihilated
and the province incorporated into Wessex in 686, and he twice

practice”, Social Identity in Early Medieval Britain, ed. B. Frazer and A.J. Tyrell (London
2000) pp. 69–89.
40
H. Geake, The Use of Grave-Goods in Conversion-Period England c. 600–c. 850, British
Archaeological Reports, British series 261 (Oxford 1997).
41
W. Filmer-Sankey, “The ‘Roman Emperor’ in the Sutton Hoo Ship Burial”,
Journal of the British Archaeological Association 149 (1996) pp. 1–9; M. Archibald,
M. Brown and L. Webster, “Heirs of Rome: the shaping of Britain A.D. 400–900”,
The Transformation of the Roman World, ed. L. Webster and M. Brown (London 1997)
pp. 208–48.
42
B.A.E. Yorke, “The reception of Christianity at the Anglo-Saxon royal courts”,
St Augustine and the Conversion of England, ed. R. Gameson (Stroud 1999) pp. 152–73.
43
Campbell, Bede’s Reges, pp. 3–4; Yorke, “Political and ethnic identity”, pp.
73–6.
  - 391

remarks on the failure of the West Saxons to provide it.44 To Bede’s

mind a gens was intimately connected with its royal house, and though
he believed identity as a gens might survive the disappearance of its
ruling family, he does not seem to have envisaged a gens, in the con-
text of the seventh-century kingdoms, which had never had its own
royal house.
Now it could be argued that as a well-educated and polemical
author Bede’s terminology might reflect his own reinterpretation of
the Anglo-Saxon situation. Bede did not, of course, work in a vac-
uum. His political vocabulary can be seen to have been derived from
other late Antique or early medieval writers, with the usage of Isidore
of Seville likely to have been particularly influential.45 But that Bede’s
equation of gens and regnum reflected how such matters were per-
ceived at the Anglo-Saxon royal courts is suggested by the evidence
of genealogies and origin myths. In the latter, kings are presented
as the embodiment of their people and their ancestors represented
as leading their people to a new land and creating a new gens. One
of the fullest foundation accounts, that of Cerdic and Cynric in the
Anglo-Saxon Chronicle, records their arrival in 495 with three ships, but
in the duplication of this entry in 514 it is said to be the West Saxons
who arrived in three ships.46 Cerdic and Cynric could stand for the
West Saxons as a whole. The thinking represented here may embody
Isidore’s definition of a gens “as a multitude sprung from one prin-
ciple”, that is king and people could be envisaged as having a com-
mon descent.47 Felix, author of the eighth-century Life of St Guthlac,
may have been intending to convey something similar when he
described Penwalh, ancestor of the Mercian royal prince Guthlac,
as being de egregia stirpe Merciorum (“of the distinguished stock of the
Mercians”) and of King Æthelbald of Mercia as de inclita Merciorum

44
Bede, Hist. Eccl. 4,16, pp. 382–5; 5,23, pp. 556–61; B.A.E. Yorke, “The Jutes
of Hampshire and Wight and the origins of Wessex”, The Origins of Anglo-Saxon
Kingdoms, ed. S. Bassett (Leicester 1989) pp. 84–96.
45
Reynolds, Kingdoms and Communities, pp. 250–5.
46
Anglo-Saxon Chronicle: Two of the Saxon Chronicles Parallel 1, ed. C. Plummer, 2
vols. (Oxford 1899) pp. 14–5; K. Harrison, “Early Wessex annals in the Anglo-
Saxon Chronicle”, English Historical Review 86 (1971) pp. 527–33.
47
Isidore of Seville, Etymologiarum sive originum libri XX 9,2,1, ed. W.M. Lindsay,
2 vols. (Oxford 1911); G. Loud, “The Gens Normannorum—myth or reality?”, Proceedings
of the Battle Conference of Anglo-Norman Studies 4 (1981) pp. 104–16, here p. 109.
392  

prole (“of the celebrated stock of the Mercians”).48 The distinguished

nature of this stock was guaranteed by the descent of the kings from
a variety of gods and heroes of the northern Germanic world incor-
porated into their genealogies.49 Of course, we only know such things
because they were recorded in the literate, post-conversion period
and the presentation of genealogies and foundation myths was no
doubt shaped by men who came from the same educated Christian
milieu as Bede.50 But that need not mean that the basic claims within
them were only created at the point when they were written down.
For instance, the Jutish links and descent from Woden in the Kentish
genealogy and origin traditions could be seen as a written manifes-
tation of what archaeological and iconographic evidence suggests was
already being claimed in sixth-century Kent,51 and parallels between
the Sutton Hoo ship-burial and some of its finds with southern
Sweden may echo claims of Scandinavian descent made in the East
Anglian genealogy.52
There is increasing acceptance for the idea that as new dynasties
came to power in the sixth and early seventh centuries they sought
to buttress their parvenu status by appeals to divine and heroic
descent that can be traced in both the archaeological and early writ-
ten sources.53 In particular, they claimed that their ancestors had
founded both gens and regnum by leading an army from the Germanic
homelands. Bede’s identification of Hengist and Horsa with Gildas’s
Saxon leaders can be seen as one literary stage in the development
of such traditions. Parallels can be found with many other Germanic
kingdoms studied in this volume, and elsewhere,54 for this type of

48
Felix, Life of St Guthlac 1, pp. 72–3; 40, pp. 124–7, ed. B. Colgrave (Cambridge
1956).
49
K. Sisam, “Anglo-Saxon royal genealogies”, Proceedings of the British Academy 39
(1953) pp. 287–348; Dumville, “Kingship, genealogies”; H. Moisl, “Anglo-Saxon
genealogies and Germanic oral tradition”, Journal of Medieval History 7 (1981) pp.
215–48.
50
N. Howe, Migration and Mythmaking in Anglo-Saxon England (Yale 1989).
51
Brugmann, “Britons, Angles, Saxons”; for possible evidence for the cult of
Woden in Kent, see S.C. Hawkes, H.R.E. Davidson and C. Hawkes, “The Finglesham
man”, Antiquity 39 (1965) pp. 17–32 ; Behr, “Origins of kingship”, pp. 39–52.
52
R. Bruce-Mitford, The Sutton Hoo Ship-Burial II: Arms, Armour and Regalia (London
1978) pp. 186–220; S. Newton, The Origins of Beowulf and Pre-Viking East Anglia
(Woodbridge 1993).
53
Archibald, Brown and Webster, “Heirs of Rome”; M. Carver, Sutton Hoo: Burial
Ground of Kings (London 1998).
54
Loud, “Gens Normannorum”; S. Reynolds, “Medieval origines gentium and the
  - 393

self-conscious construction of myths of ethnogenesis, often with some

basis in reality, in order to underpin the development of regna.
This much is relatively uncontroversial. Much more difficult to
answer is what may have lain behind the construction of the late-
fifth- and early-sixth-century Anglian, Saxon and Jutish identities as
gentes as defined by Bede and seemingly supported by archaeologi-
cal evidence. Were these earlier gentes also linked with the develop-
ment of kingship, or, failing adequate evidence for that, can they be
construed in any broader sense as “political” entities? It is an issue
that has received surprisingly little exploration, and there has been
little discussion of what it meant to be Anglian, Saxon or Jutish in
sixth-century Britain or of what provided the context for the for-
mation of their identities as gentes. It has been suggested tentatively
that cult practices might be significant in forming separate Anglian
and Saxon identities,55 but that need not rule out a political iden-
tification as well, and, indeed, cults could have been utilised to sup-
port the latter’s formation.56
We can begin further discussion with the Jutes who retained a
common political identity into the time covered by Bede’s informants
whereas any earlier Anglian or Saxon confederations appear to have
been superseded in the circumstances that led to the formation of
individual Anglian and Saxon kingdoms of the seventh century. The
name of the Jutes is not preserved in a later kingdom name, but
Bede explains that the people of Kent (Cantuari), the gens of the Isle
of Wight and “also that ( gens) opposite the Isle of Wight, that part
of the kingdom of the West Saxons which is still today called Iutarum
natio” were all of Jutish origin.57 Place-names from southern Hampshire

community of the realm”, History (1983) pp. 375–90; I.N. Wood, “Ethnicity and
the ethnogenesis of the Burgundians”, Typen der Ethnogenese unter besonderer Berücksichtigung
der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 201. Veröffentlichun-
gen der Kommission für Frühmittelalterforschung 12 (Wien 1990) pp. 53–69.
55
C. Hills, “Spong Hill and the Adventus Saxonum”, Spaces of the Living and Dead:
an Archaeological Dialogue, ed. C.E. Karkov, K.M. Wickham-Crowley and B.K. Young,
American Early Medieval Studies 3 (Oxford 1999) pp. 15–26.
56
As the cult of Woden may have done in the later sixth and seventh centuries.
For arguments for cult underpinning kingdom formation in Scandinavia see L.
Hedeager, “Myth and art: a passport to political authority in Scandinavia during
the migration period”, The Making of Kingdoms, pp. 151–6.
57
Bede, Hist. Eccl. 1,15, pp. 50–1.
394  

incorporating the name “Jute” appear to confirm the latter part of

Bede’s claim.58 By the seventh century the common Jutish identity
claimed by Kent and Wight was encapsulated by a claim that their
ruling houses shared ancestors and that they led a migration of Jutish
people to their respective provinces.59 What may lie behind the links
between Kent, Wight and southern Hampshire is that these areas
controlled the shortest crossings between Britain and Francia.60 One
of the ways in which the Kentish cemeteries differ from others in
Anglo-Saxon England is through their greater inclusion of imported
material that came from Francia.61 Elite burials of Kentish type are
known from the Isle of Wight, though unfortunately mostly through
inadequately recorded excavations of the nineteenth century,62 and
recent metal-detector finds are producing more Kentish and Frankish
material from the island.63 But Francia may have been more than
just a trading partner, and Ian Wood has argued for substantial
Frankish political influence in Kent (and this might also mean in the
other Jutish areas) for much of the sixth century.64
Tracing and dating the origins of the shared Jutish identity of
these three areas is no easy matter, but the archaeological evidence
suggests that it could be older than the establishment of the royal
house of Kent. This too cannot be dated exactly. The first Kentish
king who is known to have ruled is Eormenric, who, on the evi-
dence provided by Gregory of Tours on the marriage of his son
Æthelbert to a daughter of King Charibert of the Franks, was rul-
ing by c. 580.65 The Kentish royal genealogy swiftly plunges into

58
Yorke, “Jutes”, pp. 90–1.
59
Ibid., pp. 84–96. A foundation legend is also recorded for the southern
Hampshire Jutes in the Anglo-Saxon Chronicle which has stylistic features in common
with the accounts of the supposed ancestors of the royal houses of Wight and Kent.
60
B.A.E. Yorke, “Gregory of Tours and sixth-century Anglo-Saxon England”,
The World of Gregory of Tours, ed. K. Mitchell and I.N. Wood (Leiden 2002) pp.
113–30.
61
J. Huggett, “Imported grave goods and the early Anglo-Saxon economy”,
Medieval Archaeology 32 (1988) pp. 63–96.
62
C. Arnold, Anglo-Saxon Cemeteries of the Isle of Wight (London 1982), especially
pp. 50–72, a rich female burial from Chessell Down.
63
K. Ulmschneider, “Archaeology, History and the Isle of Wight in the Middle
Saxon period”, Medieval Archaeology 43 (1999) pp. 19–44. Less Kentish and Frankish
material is known from the cemeteries of southern Hampshire, but here too metal-
detectorists are making significant new discoveries; Stoodley, “Female assemblages”.
64
I.N. Wood, The Merovingian North Sea (Alingsås 1983).
65
Gregory of Tours, Historiae 4,26, pp. 157–9; 9,26, p. 445, ed. B. Krusch and
  - 395

the world of myth and divine ancestors beyond Eormenric,66 though

that need not, of course, mean there were not kings before him,
either of his dynasty or another, for all we may be seeing here are
the limits of individual memory at the time the genealogies were
first written down, presumably in the seventh century. But it would
be possible also to interpet the Kentish genealogical material as evi-
dence for the concept of a Jutish gens existing before its individual
provinces became kingdoms. Whether comparable histories should
be inferred for the Angles and the Saxons in the sixth century is a
more complex issue to which we can return after examining another
of the key questions of this volume—the role of the Roman Empire
in the process of the formation of kingdom and gentes.

The role of the Roman Empire and the remains of its administration
in Britain

One of the ways in which the diocese of Britain differed from other
former Roman provinces studied in this volume is that its links with
the Roman Empire had been formally severed before Germanic peo-
ples began to settle within its borders. Although there have been
doubts about whether the account transmitted by Zosimus of a revolt
and expulsion of imperial officials in 409, leading to a letter from
Emperor Honoroius to the “cities” ( poleis), actually refers to Britain,67
it would appear to fit with Gildas’s account of the British civitates
being informed by Rome that they must manage their own affairs
in future and with the sudden cessation of Roman coin issues within

W. Levison, MGH SSrM 1,1 (2nd edn., Hannover 1951); N. Brooks, “The cre-
ation and structure of the kingdom of Kent”, The Origins of Anglo-Saxon Kingdoms, ed.
S. Bassett (Leicester 1989) pp. 55–74, here pp. 64–7; I.N. Wood, “The mission of
Augustine of Canterbury to the English”, Speculum 69 (1994) pp. 1–17, here pp.
10–1.
66
Brooks, “Kingdom of Kent”, pp. 58–64.
67
Zosimus, Histoire Nouvelle VI,10,2, ed. and transl. F. Paschoud, vol. 3 (Paris
1989) p. 13; J.F. Matthews, “Olympidorus of Thebes and the history of the west
(A.D. 407–425)”, Journal of Roman Studies 60 (1970) pp. 79–97; E.A. Thompson,
“Britain A.D. 406–10”, Britannia 8 (1977) pp. 303–18; id., “Fifth-century facts”,
Britannia 14 (1983) pp. 272–4; P. Bartholomew, “Fifth-century facts”, Britannia 13
(1982) pp. 261–70; I.N. Wood, “The fall of the western empire and the end of
Roman Britain”, Britannia 18 (1987) pp. 251–62. See also A. Woolf (this volume).
396  

Britain.68 As we have seen, the standard interpretation of recent years

has favoured the collapse and disappearance of Roman institutions
in eastern England, but there is evidence that points in a different
direction and which could mean, even if the Roman Empire had
no direct involvement in affairs in Britain after the beginning of the
fifth century, that there was a significant legacy to its subRoman
heirs. At the risk of appearing simplistic, one can begin by observ-
ing that the Anglo-Saxon kingdoms of the early seventh century,
especially those of the southern and eastern coastal areas, were not
dissimilar in size and geographical extent to the Roman civitates on
which power apparently devolved in c. 410. The kingdom of Kent
actually retained the name of the former Roman civitas that occu-
pied much the same area,69 while the name of the kingdom of Lindsey
derives from that of the former provincial capital of Lincoln.70 Other
east coast kingdoms appear to have had boundaries broadly corre-
sponding with those of former civitates. The kingdoms of the East
Angles, East and South Saxons seem to have occupied territories
broadly comparable to those of the civitates of the Iceni, Trinovantes
and Regni respectively. Key subdivisions of the early Anglo-Saxon
kingdoms were the regiones that would seem to be units with much
in common with the pagi, the administrative subdivisions of the civi-
tates that may often have been based on significant Iron Age settle-
ment districts. Vici, small Roman towns, were often the centres from
which pagi were administered and both name, and possibly function,
may be preserved in the wicham place-names of early Anglo-Saxon
England.71 Although in the absence of any survival of administrative
records from Roman Britain the actual extent of any pagus or vicus
territory has to remain uncertain, nevertheless evidence for continu-
ity with a later regio can sometimes be suggestive.72 The early medieval

68
Gildas, Ruin of Britain 18, p. 94; J. Kent, “The end of Roman Britain: the lit-
erary and numismatic evidence reviewed”, The End of Roman Britain, ed. P.J. Casey,
British Archaeological Reports, British series 71 (Oxford 1979) pp. 15–27.
69
Brooks, “Kingdom of Kent”, pp. 57–8; the detailed account of the Kentish
foundation legend preserved in the Historia Brittonum purports to record how the
province passed from subRoman to Saxon hands.
70
S. Bassett, “Lincoln and the Anglo-Saxon see of Lindsey”, Anglo-Saxon England
18 (1989) pp. 1–32.
71
M. Gelling, “English place-names from the compound wicham”, Medieval Archaeology
12 (1967) pp. 87–104; C.J. Balkwill, “Old English wic and the origins of the hun-
dred”, Landscape History 15 (1993) pp. 5–11.
72
B. Eagles, “The archaeological evidence for settlement in the fifth to seventh
  - 397

landscape can be seen to have evolved out of the late Roman with-
out having to resort to arguments for radical change having occurred,73
and the detection in certain areas of what would appear to be a
coincidence of the territories of major Anglo-Saxon cemeteries with
areas dependent on Roman villas may also point to some continu-
ity not only in the use of the Roman landscape,74 but also of aspects
of the organisational structures which were an integral part of it.
The major Anglo-Saxon unit of measurement for land, the hide, on
which liability for public services and royal dues was made, has also
been proposed as having its ultimate origin in the Roman period.75
If such arguments are accepted, the evidence goes beyond survival
of settlement districts to an implication of their continuing use as
taxable units throughout the fifth and sixth centuries, but to many
that would seem a big “if ”.
The accounts of Gildas and Zosimus imply that the civitates may
have become the unit around which subRoman government of Britain
was organised, a proposition which receives additional support from
the fact that civitas territories seem to have become the basis of the
British kingdoms known to Gildas in the west of Britain.76 It may
even have been the case, as Martin Millet has suggested that the
pagi served as the powerbases of local magnates who were council-
lors of the civitas.77 But there is an additional factor that has to be
taken into account when assessing how affairs in subRoman Britain

centuries A.D.”, The Medieval Landscape of Wessex, ed. M. Aston and C. Lewis (Oxford
1994) pp. 13–32, here p. 24; Williamson, Origins of Hertfordshire, pp. 66–74, though
admittedly these examples are from an inland area away from the earliest Anglo-
Saxon settlements and with very few “Germanic” finds of fifth or sixth centuries.
73
Powlesland, “Early Anglo-Saxon settlements”, pp. 101–24.
74
For example, T. Williamson, The Origins of Norfolk (Manchester 1993) pp. 58–69.
75
P.S. Barnwell, “Hlafæta, ceorl, hid and scir: Celtic, Roman or Germanic?”, Studies
in Anglo-Saxon Archaeology and History 9 (1996) pp. 58–61. The problem over issues
such as these, in the absence of documentary records, is whether there is a gen-
uine institutional continuity or one dictated by the natural geography and by the
impact of earthworks, or other structures made in Roman or earlier times, on the
natural landscape. Bede saw the hide as a distinctively “English” form of assess-
ment; see Brooks, Bede and the English, p. 10.
76
W. Davies, Wales in the Early Middle Ages (Leicester 1982) pp. 85–102; B.A.E.
Yorke, Wessex in the Early Middle Ages (London 1995) pp. 12–24; A. Woolf (this vol-
ume) suggests that in parts of lowland Britain civitates may have remained units of
government controlled by judices without making the transition to becoming king-
doms—an argument with analogies for what is proposed here for eastern England.
See also K.R. Dark, Britain and the End of the Roman Empire (Stroud 2000) pp. 144–9.
77
M. Millett, The Romanization of Britain (Cambridge 1990) pp. 149–51.
398  

may have been organised and that is the late Roman reorganisation
of the diocese of Britain into four, or possibly five, provinces each
with its own capital city and governor ( praeses or rector) responsible
to the vicarius Britannorum.78 How far this provincial organisation sur-
vived beyond 410 is, of course, unknown, but initial survival of the
provinces as the major units of political administration in fifth-cen-
tury Britain might help to explain how areas of the country came
to have such different histories in the later fifth and sixth centuries.
The western province of Britannia Prima, dependent on Cirencester,
can be seen as having taken a line of development that led it in a
quite different direction from the provinces of south and east. Ultimately
its orientation turned westwards rather than eastwards so that it came
to have more in common with other Irish Sea and western Atlantic
areas.79 The two eastern provinces of Flavia Caesariensis and Maxima
Caesariensis may also have operated as separate units which could
explain the broad Anglian and Saxon divisions in the eastern half
of England. The rough boundary that can be drawn from archaeo-
logical and written evidence between Anglian and Saxon areas
broadly corresponds with the boundary suggested for these two late
Roman provinces.80 If this explanation is acceptable—and that may
be another big “if ”—one could postulate that Anglian and Saxon
identities began to form among the Germanic settlers of two areas
that were distinct in part because they had different subRoman gov-
ernments. We need not assume that the two provinces had identi-
cal histories even though ultimately they both came to be dominated
by rulers who claimed a Germanic origin—nor that they were nec-
essarily allies bearing in mind Gildas’s analysis of civil war as one
of the major weaknesses of the British polities;81 indeed, warfare
between the two provinces might help to explain the circumstances
in which Germanic leaders with warbands became the dominant

78
P. Salway, Roman Britain (Oxford 1981) pp. 316–9; 505–38; C.A. Snyder, An
Age of Tyrants: Britain and the Britons, A.D. 400–600 (Stroud 1998) pp. 3–25.
79
See A. Woolf this volume, but the distinction he draws between developments
in the “highland” areas of western Britain and its “lowland” areas should be noted;
see also Dark, Britain and the End of the Roman Empire.
80
As suggested by Alan Vince in discussion at the 47th Sachsen Symposium in York
in 1996. For the boundaries see Salway, Roman Britain, map 7.
81
Gildas, Ruin of Britain 21, pp. 95–6; D.N. Dumville, “The idea of government
in sub-Roman Britain”, After Empire. Towards an Ethnology of Europe’s Barbarians, ed.
G. Ausenda, Studies in Historical Archaeoethnology (Woodbridge 1995) pp. 177–216.
  - 399

power. There would seem to have been a greater density of Germanic

settlement in the Anglian province and a continued link with the
North Sea homelands.82 One might speculate that it came under
Germanic control at an earlier date than the Saxon province in
which settlement by Germanic incomers was apparently initially
confined to certain districts only.83 West Saxon royal tradition that
included as their founder Cerdic, who had a “British” rather than
Saxon name, may also suggest a more complex entwinement of
British and Germanic power structures in this province.84
In the absence of adequate written records any attempt to recon-
struct political events in eastern England can be only hypothesis, but
it is one constructed in the light of developments in northern Gaul
where Ægidius or his son Syagrius may have established a kingdom
comparable in extent to one of the British provinces.85 The superbus
tyrannus, who together with his council was castigated by Gildas for
first allowing the “Saxons” into Britain, could be a candidate for the
ruler of a British province controlled through its civitates,86 while the
presence in Gaul c. 470 of a British king Riothamus with his army
and fleet is further evidence that rulers with considerable power and
resources had emerged in fifth-century Britain (though it is not known
from what part he came).87 A fleeting view of eastern Britain in the
earlier part of the fifth century comes from Constantius’s Life of St
Germanus of Auxerre which suggests that in the 420s and 430s Britain
could still appear to a visitor from Gaul as a basically Romanised
province,88 an observation which is all the more interesting as the
archaeological evidence suggests that a major transformation of many
of the essential features of Romanised Britain had already taken place
by that date.89

82
Hines, Scandinavian Character; Hills, “People of Spong Hill”, pp. 145–9.
83
Notable areas with few apparent signs of early Germanic culture are districts
around the civitas capitals of Chichester, Silchester and Verulamium; Dark, Britain
and the End of the Roman Empire, pp. 97–103.
84
R. Coates, “On some controversy surrounding Gewisae/Gewissei, Cerdic and Ceawlin”,
Nomina 13 (1989/90) pp. 1–11; Dark, Britain and the End of the Roman Empire, pp. 75–8.
85
James, “Origins of barbarian kingdoms”, pp. 46–7.
86
Gildas, Ruin of Britain 23, p. 97.
87
Dumville, “The idea of government”, pp. 181–2, and further discussion pp.
206–8.
88
Constantius of Lyon, Vie de St Germain d’Auxerre 3,12–18, pp. 144–59; 5,25–27,
pp. 170–3, ed. R. Borius, Sources Chrétiennes 112 (Paris 1965).
89
A.S. Esmonde Cleary, The Ending of Roman Britain (London 1989).
400  

It might not strain credulity too much to suggest that British rulers
could have emerged in control of the eastern provinces of late Roman
Britain, but where can we go from there? Do we envisage Anglian
and Saxon equivalents of Childeric taking control of these former
Roman provinces, though without successfully maintaining and spread-
ing their power as Childeric’s descendants were able to do? The
idea of Anglian and Saxon provinces each with their own ruler might
account for the appearance of distinct Anglian and Saxon identities
that has already been discussed, but is otherwise without any sup-
port. If we are to consider some sort of Germanic take-over of British
provincial structure it might be possible to envisage it supported by
some more devolved form of power-sharing recorded elsewhere within
the Germanic world. Bede, for instance, recorded that the Old Saxons
had no king, but satrapas who controlled different districts (Gau) within
the Old Saxon province and might temporarily unite under one
leader in times of war.90 When eastern England was briefly under
Viking control in the late ninth and early tenth centuries there appears
to have been a system of districts (based on existing Anglo-Saxon
territorial units) in which subdivisions of the army may have settled
and, presumably, were supported to some extent through tribute pay-
ments.91 Such a devolved system of power in the eastern and south-
ern districts of England in the late fifth and early sixth centuries
could explain both how Anglians and Saxons came to have an iden-
tity as distinct gentes and how Roman adminstrative subdivisions might
have continued to have significance within this period. It would not
necessarily be incompatible with interpretations of the archaeologi-
cal evidence that sees little evidence for a marked hierarchy at this
time and only temporary acquisitions of a broader powerbase by the

90
Bede, Hist. Eccl. 5,10, pp. 480–5. For further discussion of the arrangements
in Saxony see T. Reuter, Germany in the Early Middle Ages 800–1050 (Harlow 1991)
pp. 65–7, and M. Becher, “Die Sachsen im 7. und 8. Jahrhundert”, 799—Kunst
und Kultur der Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn. Beiträge zum
Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz
1999) pp. 183–94. Bede may have used satrapa in part to indicate an unfamiliar
political situation, but he was also influenced in his choice of vocabulary by an
apposite biblical text.
91
Anglo-Saxon Chronicle: Two of the Saxon Chronicles Parallel 1, pp. 100–4; C. Hart,
“The Aldewerk and minster at Shelford, Cambridgeshire”, Anglo-Saxon Studies in
Archaeology and History 8 (1995) pp. 43–68, especially pp. 54–8; L. Abrams, “Edward
the Elder’s Danelaw”, Edward the Elder 899 –924, ed. N. Higham and D. Hill
(Manchester 2001) pp. 128–43.
  - 401

more prominent households. But where it would differ from current

interpretations based on the archaeological evidence is in seeing the
kingdoms that emerged in the latter part of the sixth century as the
result (or cause) of the break up of political/military confederations
that had characterised the Anglian and Saxon provinces, rather than
the result of a complete collapse of Roman systems and a building
up from household units to district and ultimately kingdom control
through the model of peer-polity interaction.92

The circumstances that led to the formation of Anglo-Saxon regna

By the sixth century there would appear to have been distinctive

Germanic gentes (the Angles, Saxons and Jutes), but neither written
nor archaeological evidence that would confirm that these were also
regna, or contained regna within them. Instead, they may have been
controlled through distinctively Germanic forms of power-sharing
that are attested amongst other Germanic peoples either in their
homelands or in the early years of their contact with the Roman
Empire.93 Archaeological and written evidence could be in accord
in seeing the development of regna and reges as a secondary devel-
opment of the latter part of the sixth century that led to dissolution
of the former Anglian and Saxon confederations and a rejection of
the distinctive forms of dress (and presumably other distinguishing
features which are harder to trace) that had signified membership

92
A major plank of the latter argument is the existence of regiones within the
Anglo-Saxon kingdoms which have been seen as autonomous districts which became
the “building blocks” through whose combination the kingdoms were constructed.
The so-called “Tribal Hidage” list has been interpreted as providing support for
this approach. This is too big an issue to pursue fully here, but suffice it to say
that many historians are becoming increasingly wary of placing too much reliance
on the “Tribal Hidage” whose original date and circumstances of composition are
unclear. Nor is there unequivocal support from any narrative or administrative doc-
uments of the seventh or early eighth centuries for regiones having the type of
autonomous existence that has been envisaged for them; rather they appear as set-
tlement units utilised as administrative subdivisions of larger units and, unlike the
kingdoms, they are never referred to as gentes. See Yorke, “Political and ethnic iden-
tity”, pp. 82–6.
93
See Gregory of Tours, Historiae 2,9, pp. 52–8, and Hans-Werner Goetz (this
volume, pp. 310–1) for the existence of many leaders among the Franks when they
settled in Gaul in the fifth century.
402  

of the Anglian and Saxon gentes. Can we explain the circumstances

in which the development of regna took place?
Such developments are unlikely to have been monocausal, and
there is not the space to explore fully the complex interweaving of
factors which would have enabled individual families to aggregate
power and wealth into their hands to such an extent that the dis-
tance they created between themselves and their nearest competitors
was sufficiently great to allow them to establish themselves as royal
houses. Different disciplines will stress differing factors. The rise in
population that is generally felt to have taken place in the “Germanic”
areas of Britain during the sixth century, and the ability to create
and exploit agricultural surpluses and other avenues to wealth are
elements that must be given due weight.94 However, it may also be
possible to see something of the political changes that encouraged
concepts of royal power to develop.
For many of the peoples discussed in this volume it was direct
involvement with the Roman Empire that encouraged the appear-
ance of Germanic regna. The Roman Empire found it easier to deal
with one all-powerful ruler rather than several sharing control, and
after its collapse many native aristocracies too seem to have favoured
the stability of royal power, that allowed the continuation of many
facets of Roman provincial government, to the uncertainties of
fluctuating authority under traditional Germanic leadership systems.
Those living in southern and eastern Britain in the later fifth and
sixth centuries would have had the opportunity to discover some-
thing of kingship in operation in the west and north of the coun-
try.95 Anglian areas, in particular, may also have been stimulated by
the possible move towards establishment of kingdoms in the North
Sea homelands,96 and, as in the Viking age, one should not rule out
the movement of powerful leaders between the homelands and their

94
Arnold, Archaeology of Anglo-Saxon Kingdoms; Hamerow, “The earliest Anglo-Saxon
kingdoms”.
95
See A. Woolf (this volume).
96
L. Hedeager, Iron Age Societies: From Tribe to State in Northern Europe, 500 B.C. to
A.D. 700 (Oxford 1992); U. Näsman, “The ethnogenesis of the Danes and the mak-
ing of a Danish kingdom”, The Making of Kingdoms, pp. 1–10. Care must be taken
here as arguments for state-formation in Denmark and other North Sea areas have
been influenced by models of peer-polity interaction applied to Anglo-Saxon England.
  - 403

“colonies” in Britain as a possible destabilising factor. But the king-

dom which is likely to have had most influence upon eastern and
southern England, and to have acted as a model of post-Roman
royal power, is that of the Franks. The Frankish connections of Kent,
the first recorded Anglo-Saxon kingdom to emerge, are well-known,
and Ian Wood has drawn together evidence to suggest that Frankish
kings may have exercised some hegemony in southern England for
much of the sixth century.97 It may be possible to go further to sug-
gest that it was the links with Francia that lay behind the formation
of the Jutish confederation of Kent, the Isle of Wight and southern
Hampshire in order to control the English Channel.98 The Merovingian
rulers may have acted in ways analogous to the Roman Empire to
stimulate the development of kingship among gentes over which they
had some influence, but which they were not contemplating absorb-
ing into the Frankish realm. The Jutish gentes may not have had their
own kings when first drawn into the Frankish orbit, but their con-
tinuing association with the greater power might have encouraged
them to develop and would have provided a role-model for aspi-
rants to follow.
It is possible that the Merovingians played a significant role in
state-formation in other areas of southern Britain, and by so doing
provided much of the impetus that broke up the former Anglian
and Saxon confederations. Frankish involvement in the early years
of the East Anglian kingdom is attested by Bede and other sources.99
Ian Wood has suggested that Frankish hegemony may explain one
of the great mysteries of early Anglo-Saxon kingship, the overlord-
ship of southern England, as “names remembered in Canterbury as
kings acknowledged by the Merovingians and subsequently misin-
terpreted”.100 Alternatively Frankish rulers may have provided a rather
different kind of stimulus by provoking a reaction against the con-
trol they were trying to establish over the Channel which might also
have provided the circumstances in which powerful military leaders
could transform themselves into the founders of royal dynasties. One

97
Wood, Merovingian North Sea, pp. 12–8.
98
Yorke, “Gregory of Tours and sixth-century England”.
99
Bede, Hist. Eccl. 3,18, pp. 266–9; I.N. Wood, “The Franks and Sutton Hoo”,
People and Places in Northern Europe 500–1600. Essays in Honour of Peter Hayes Sawyer,
ed. I.N. Wood and N. Lund (Woodbridge 1991) pp. 1–14.
100
Wood, Merovingian North Sea, p. 14.
404  

of the main reasons Frankish rulers may have had for wishing to
control the Channel was to reduce the activities of “Saxon” raiders
attested in a variety of sources.101 The geographical origin of these
“Saxons” is not given, but it is more likely that they were the Angles
and Saxons of Britain rather than voyagers who had come directly
from the North Sea homelands.102 Frankish attempts to cleanse the
Channel may therefore have provoked a powerful response among
those whose piratical livelihoods were thus threatened. Could it be
that those first rulers of the southern English recorded by Bede, Ælle
of the South Saxons and Ceawlin of the West Saxons, were men
who attempted to organise a resistance to Frankish control of the
Channel?103 Their activities are hard to date exactly. For what it is
worth, the Anglo-Saxon Chronicle records that Ælle came to Britain in
477 and that Ceawlin began to reign in 560.104 However, as David
Dumville has shown comprehensively the West Saxon dates in the
Chronicle are the result of a revision whose effect was to make the
formation of the kingdom appear to have been earlier in date than
actually seems to have been the case and the evidence of the earli-
est West Saxon regnal-lists is that Ceawlin would have ruled either
581–8 or 571–88.105 It may well be the case that the reign of Ælle
should also be placed somewhat later so that he was the immediate
predecessor of Ceawlin.106
Ælle and Ceawlin, if this tradition of their overlordship has any
historical validity, could have been leaders of a resistance to Frankish
hegemony, or alternatively its guarantors as their successors, Æthelbert
of Kent and Raedwald of the East Angles, would appear to have

101
I.N. Wood, “The Channel from the 4th to the 7th centuries A.D.”, Maritime
Celts, Frisians and Saxons, ed. S. McGrail, Council for British Archaeology Research
Report 71 (London 1990) pp. 93–7; Yorke, “Gregory of Tours and sixth-century
England”.
102
Anglo-Saxon territories in Britain could have provided temporary bases for
North Sea fleets that might later return home; later Viking activities provide an
obvious possible analogy. For the early Anglo-Saxons as pirates see also J. Heywood,
Dark Age Naval Power (London 1991) especially pp. 54–62.
103
Bede, Hist. Eccl. 2,5, pp. 148–51.
104
Anglo-Saxon Chronicle: Two of the Saxon Chronicles Parallel 1, pp. 14–5; 18–9.
105
The uncertainty is because variant traditions give the alternative of seven or
seventeen years for his reign. D.N. Dumville, “The West Saxon Genealogical Regnal
List and the chronology of early Wessex”, Peritia 4 (1985) pp. 21–66.
106
The entry in the Chronicle for 560 links the accession of Ceawlin with that of
another Ælle, the king of Deira. 560 would in fact be a suitable date for the acces-
sion of the South Saxon Ælle.
  - 405

been. What may be of greater moment is to note that the period

in which a new confederation of southern English is said to have
come into existence, the latter part of the sixth century, is the same
period in which the first of the Anglo-Saxon kingdoms familiar to
Bede seem to have emerged. Hazy though the picture is, it appears
that something happened in the latter part of the sixth century which
led to the break up of the former Anglian and Saxon confedera-
tions, to the creation of a temporary new confederation of “south-
ern English” and ultimately to the emergence of kingdoms, apparently
based in a number of cases on former civitas units that had previ-
ously been united in the Anglian and Saxon confederations. That
something new and decisive occurred at this time is also suggested
by the rejection in the upper reaches of Anglo-Saxon society of the
distinctive forms of dress and decoration which had proclaimed them
to be members of the Anglian or Saxon confederations in favour of
new forms that signalled that they too were “heirs of Rome” like
those living in other regna of the former Roman Empire.107

Conclusion

Consideration of early Anglo-Saxon history has suggested that there

are parallels in the formation of gentes and regna with what occurred
in other former provinces of the Roman Empire when they were
settled by Germanic peoples, even if the scale and chronology appear
different. The Anglo-Saxon evidence supports the view that there
could be a gens before there was a regnum. Although aspects of inter-
pretation can be controversial, written and archaeological evidence
appears to support the existence of Anglian, Saxon and Jutish gentes
before individual Anglo-Saxon kingdoms came into existence. These
gentes bore the names of peoples in the North Sea homelands, but
should not be seen as identical with them. The archaeological evi-
dence shows that their material culture—and so presumably other
aspects of their identity as gentes—had been affected by the migra-
tion to Britain which had brought peoples from various different

107
However, Germanic elements, especially descent from the gods, were initially
also very important in supporting the new kingly regimes as discussed above; see
Yorke, “Reception of Christianity”.
406  

Germanic gentes together, as well as opening them to new experi-

ences as a result of settlement in a former Roman province, even if
their experience of Romanitas may have been less concentrated than
that of Germanic settlers in some other former areas of the Roman
Empire. The most controversial aspect of this paper is the hypothe-
sis that the Anglian, Saxon and Jutish gentes may have developed
within a territorial and administrative framework inherited from the
late Roman world and that they had, at least in a very broad sense,
a political identity. However, that political identity may not have
been expressed in the late fifth and early sixth centuries through the
development of kingship. The nature of their governmental organi-
sation is therefore very difficult to establish without the benefit of
written records describing its operation, but it can perhaps be pre-
sumed to be similar to that dimly discerned in the Germanic home-
lands, for instance, among the Old Saxons, or among some of the
other Germanic peoples such as the Franks when they first came
into contact with literate commentators.
The Anglo-Saxon kingdoms that began to develop towards the
end of the sixth century are therefore a secondary development that
grew out of the earlier gentes and, it is suggested here, a dissolution
of the political order that had supported them. In the new kingdoms
we can see a very self-conscious expression of new political identi-
ties in which gens and regnum were equated and the establishment of
both ascribed to the founder of the royal dynasty. The new politi-
cal order could draw upon both Germanic tradition, particularly that
of myth and religion, and ape the tactics of Germanic dynasties
already established in former parts of the Roman Empire which had
utilised imperial trappings to underpin their regimes. Partly through
such emulation the new Anglo-Saxon dynasties came to appreciate
the value of Christianity in underpinning state-formation, and as a
result of their new religion they, like other Germanic gentes within
the former Roman Empire, came to absorb and internalise the bib-
lical and classical traditions that explained how new gentes had been
formed.108 There may have been less of the constitutional and phys-
ical remains of the Roman world for the Anglo-Saxon kingdoms to
inherit than in some other parts of the former Roman Empire (though
this paper favours some facets surviving), but ideas could be imported.

108
Patrick J. Geary in his paper delivered at the symposium.
  - 407

When it came to the ideology and vocabulary of gens and regnum the
Anglo-Saxons of the seventh and eighth centuries were in accord
with the other Germanic gentes and regna that had grown out of the
dissolution of the Roman Empire in western Europe.
This page intentionally left blank
 GENS, REX AND REGNUM OF THE LOMBARDS*

Jörg Jarnut

1. Is there a development from a Germanic gens of the Migration Period

to a Germanic kingdom? Or does a gens (or this gens) not exist until after
the establishment of a kingdom?

The question of the relationships between gens and rex, between peo-
ple and king, or between gens and regnum, people and kingdom, can,
in the case of the Lombards, partly be easily answered, partly answered
only with the greatest difficulty. The reason for this lies, above all,
in the nature of the sources. Though there are relatively many
Lombard sources from the middle of the seventh century onwards,
information about their pre-Italian phases and the first century of
their Italian history can almost only be gathered from these later
sources. This means that only one of the eight centuries of Lombard
history, as it is transmitted via written sources, can be seen through
the eyes of Lombard (near)contemporaries. For the other seven cen-
turies (leaving apart the information of little relevance transmitted
by antique authors), we are forced to use Lombard memory—with
its mythical dimensions—of the seventh and eighth centuries as the
basis of our historical construction.1
The “beginnings” of the Lombards, then, are especially obscure.
The little information on the subject given by antique authors dur-
ing the first two centuries A.D. only testify to their existence, to their
name, to their settlement in the area of the Lower Elbe and to a
few political-military actions. However, Velleius Paterculus in the first
half of the first century A.D. refers to them as a gens.2 One thing all

* Given the comparative character of this book, the main questions will be dealt
with in the given order in this article.
1
The most important Lombard narrative sources are: Origo gentis Langobardorum,
ed. G. Waitz, MGH SSrL (Hannover 1878), Paul the Deacon, Historia Langobardorum,
ed. L. Bethmann and G. Waitz, ibid., and the Origo gentis Langobardorum codicis Gothani,
ed. G. Waitz, ibid.
2
Velleius Paterculus, Historiarum libri duo 2,106, ed. W.S. Watt, Bibliotheca
Teubneriana (Leipzig 1988) p. 73.
410  

these sources have in common is that they neither tell us about the
“constitution” (Verfassung) of the Lombard gens, nor if they were led
by a king.
The seventh and eighth centuries, the period in which the most
significant later sources were written, was a period in which Lombard
history was largely determined by the kings at Pavia and by the
élites dependent on them. If the circumstances described in these
sources are projected back onto a largely obscure past, the danger
is of creating a false impression that can as easily disorient as inform
us about the relationships between gens, rex and regnum.
It is, therefore, all the more remarkable that both the Origo gentis
Langobardorum, the oldest Lombard narrative source dating from the
second half of the seventh century, and the (partly) dependant his-
tory of the Lombards written by Paul the Deacon, agree that the
Lombards from the time of their mythical Scandinavian origins were
led by principes or duces.3
In 643, King Rothari began his law-book by listing his predeces-
sors, a record allegedly based on the statements of old men. It was
presented by him as comprising the names of all the kings ex quo in
gente nostra Langobardorum reges nominati coeperunt esse. This list—undoubt-
edly serving as establishing the king’s legitimacy as legislator—indi-
cates, on the one hand, that according to seventh-century Lombard
memory the time of the kings was preceded by a kingless period,
and, on the other, that the list of seventeen names, beginning with
Agilmund, from the Guging family, and ending with the Harud
Rothari, should be considered as an official catalogue.4 In the eyes
of Rothari and these elders, it listed all the Lombard kings. The
most important thing that this catalogue shows, however, is that in
the middle of the seventh century the Lombards considered their
gens, rather than their kingship, as the more ancient entity, and that
they, thus, considered their regnum as a creation of their gens.
It is not surprising that the Origo gentis Langobardorum, which is
closely linked to the Edict in its transmission, presents relationships
in a similar way: after the Scandinavian, kingless phase of Lombard

3
Origo gentis Langobardorum 1, p. 2: principes; Paul the Deacon, Historia Langobardorum
1,7, p. 52; 1,14, p. 54: duces.
4
Edictus Rothari Prol., ed. F. Bluhme, MGH LL 4 (Hannover 1868) p. 2.
   411

history, the Lombards made the Guging Agilmund, allegedly a son

of Agio, king.5 Here, too, gens exists before regnum.
Paul the Deacon, who, among other sources, used the Origo for
his Historia Langobardorum, copies this account. He adds, however, that
the Lombards no longer wished to be ruled by duces as before, but
that, after the example of other gentes, they desired a king and acted
accordingly; an observation to which we will return later.6
The Historia Langobardorum from the Gotha Codex, composed shortly
after 800 and hence the most recent of our four main sources, also
recounts how the Lombards chose Agilmund as their first king.7 In
this source it is made fairly clear that this must have happened in
the area where they settled near the Elbe. Hence, from there the
Lombards under Agilmund must have advanced upstream along the
Elbe as far as Bohemia.8 Combining this with the somewhat vague
information provided by the Origo and Paul the Deacon, the well-
founded hypothesis has been put forward that at the end of the
fourth century Agilmund was made king, of those Lombards who
separated from the others at the Lower Elbe and went south under
his leadership.9 A surprising, but unprovable, temporal specification
of this assumption can be found in the Renaissance copy of the
chronicle by Prosper Tiro, where it is maintained that Agilmund
became king in 389.10
As a provisional answer to our initial question it ought to be borne
in mind that the main sources from the seventh and eighth centuries

5
Origo gentis Langobardorum 2, p. 3.
6
Paul the Deacon, Historia Langobardorum 1,14, p. 54. Cf. below [part 2, § 3].
7
Origo gentis Langobardorum codicis Gothani 2, p. 8.
8
Ibid.: Postquam de eadem ripa, ut supra dictum est, Langobardi exierunt, sic Scatenauge
Albiae fluvi ripa primis novam habitationem posuerunt; sic deinde certantes Saxoniae patria attigerunt,
locus ubi Patesbruna cognominantur; ubi sicut nostri antiqui patres longo tempore asserunt habitasse,
et in multis partibus bella et pericula generarunt. Ibique primis regem levaverunt nomine Agelmund.
Cum ipso de hoc loco in antea patrias ad suam partem expugnare coeperunt; unde in Beovinidis
aciem et clauses seu tuba clangencium ad suam proprietatem perduxerunt; unde usque hodie prae-
sentem diem Wachoni regi eorum domus et habitatio apparet signa.
9
See especially H. Fröhlich, Studien zur langobardischen Thronfolge. Von den Anfängen
bis zur Eroberung des italienischen Reiches durch Karl den Großen (774), 2 vols. (Doct. diss.
Tübingen 1980) vol. 1, esp. p. 32; J. Jarnut, “Zur Frühgeschichte der Langobarden”,
Studi Medievali 3a ser. 24/1 (1983) pp. 1–16, esp. pp. 13–5.
10
Interpolationes chronicis Prosperi insertae saec. XV, ed. T. Mommsen, MGH AA 9,1
(Berlin 1882) p. 498. The reference to this late and hardly used source in R.
Schneider, Königswahl und Königserhebung im Frühmittelalter. Untersuchungen zur Herrschafts-
nachfolge bei den Langobarden und Merowingern, Monographien zur Geschichte des
Mittelalters 3 (Stuttgart 1972) p. 10 n. 27.
412  

maintain that the previously kingless gens Langobardorum, presumably

at the end of the fourth century, had consciously decided to intro-
duce kingship. Accordingly, the gens must already have existed long
before the regnum. Whether this version of events has any relation to
reality is, admittedly, by no means certain. The Old English poem
Widsith, possibly composed in the late seventh century, mentions—
apart from the rulers Agilmund, Audoin and Alboin attested to in
Lombard tradition11—no less than six kings of the Lombards (or
Lombard tribes) none of whom are recorded in Lombard tradition
itself.12 Following a plausible hypothesis formulated by Reinhard
Wenskus and the editor of Widsith, Kemp Malone, these six kings
may have ruled over the Lombards who stayed behind by the Lower
Elbe.13 Nonetheless, it cannot be completely excluded that at least
some of these reigned before Agilmund, and that Lombard kingship,
therefore, already existed before the end of the fourth century. We
must, therefore, resign ourselves to the fact that we know little with
any certainty about Lombard political arrangements before their (par-
tial) departure from the Lower Elbe. Still, the possibility exists that
contrary to later Lombard sources they were ruled by kings already
then. The saga of Gambara, Ibor and Agio, however, seems to ren-
der it more plausible that the Lombards were constituted as a gens—
without kingship necessarily being part of this concept at the beginning
of the process. Formulating this idea it ought, though, once more
to be emphasised that it is in the main based on a saga written
down in the seventh century, which purports to record events from
the first century B.C.

2. What sorts of changes and conditions lead to, or represent the development
towards, the establishment of a Germanic kingdom?

According to the later sources discussed above, the migration from

Scandinavia of the proto-Lombards, who at that time still went by
the name of Winniler, took place under the leadership of the duces

11
Widsith 70; 117, ed. K. Malone (Copenhagen 1962) pp. 25–6.
12
Ibid., 23, p. 23; 32, p. 24; 48, p. 24; 49, p. 24; 117, p. 26. For the identification
of the kings, see Malone’s “Glossary of Proper Names”, pp. 126 ff.
13
Cf. R. Wenskus, Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen
gentes (Köln-Graz 1961) pp. 489–90; Widsith, esp. pp. 108; 162–5.
   413

or principes Ibor and Agio and their mother Gambara.14 The new
patriae of Scoringa and Mauringa, which are only mentioned by Paul
the Deacon, as well as the area on the Lower Elbe in Golaida or
Scatenauge, which according to the other three main narrative sources
the Lombards settled, are said to have been conquered by these duces
or principes.15
It is, however, worth noting that all three sources—be it with vary-
ing degrees of clarity—link the separation of a part of the Lombards
on the Lower Elbe to the introduction of the institution of kingship
and—corresponding also with Rothari’s list of kings—to the eleva-
tion of the first rex Langobardorum, Agilmund.
The three main narrative sources show, more or less clearly, that
Agilmund was held to have led his people in the migration from the
Lower Elbe. This means that he was made king, by the part of the
people that separated from the others, in order to lead the immi-
nent conquests. This would make Agilmund a typical military king
(Heerkönig) according to Schlesinger’s definition.16 Paul the Deacon’s
remark mentioned previously, that the Lombards no longer wished
to be led by duces but rather raised him to the kingship ad instar cete-
rarum gentium,17 may be interpreted as a suggestion that the Lombards
only began to follow the “success-model” of kingship during this
period of expansion. If, however, the Lombards were already led by
kings on the Lower Elbe then Agilmund’s election may indicate that
those Lombards who migrated wished to be considered equal in rank
to those who stayed behind in Golaida.
How stable the institution of kingship was even under Agilmund
may be gathered from the fact that the office was not abolished

14
Cf. references cited above.
15
Origo gentis Langobardorum 2, p. 3; Paul the Deacon, Historia Langobardorum 1,7–13,
pp. 52–4; Origo gentis Langobardorum codicis Gothani 1–2, pp. 7–8. On the “migrations”
of the Lombards, see most recently Jarnut, “Zur Frühgeschichte der Langobarden”;
id., “I Longobardi nell’epoca precedente all’occupazione dell’Italia”, Langobardia, ed.
S. Gasparri and P. Cammarosano (Udine 1990) pp. 3–34 and—in part very different—
from an archaeological viewpoint, M. Menke, “Archeologia longobarda tra la bassa
Elba e l’Ungheria”, ibid., pp. 35–106; N. Christie, The Lombards, The Peoples of
Europe (Oxford UK-Cambridge Mass. 1995) pp. 1–20.
16
W. Schlesinger, “Über germanisches Heerkönigtum”, Das Königtum. Seine geisti-
gen und rechtlichen Grundlagen, ed. T. Mayer, Vorträge und Forschungen 3 (Lindau-
Konstanz 1956, 4th edn. 1973) pp. 105–41. See, further, Schneider, Königswahl und
Königserhebung, pp. 8–9; Fröhlich, Studien zur langobardischen Thronfolge 1, pp. 31–5, and
J. Jarnut, Geschichte der Langobarden, Urban 339 (Stuttgart 1982) pp. 28–9.
17
Paul the Deacon, Historia Langobardorum 1,14, p. 54.
414  

when Agilmund, at the beginning of the fifth century, suffered a cat-

astrophic defeat against the Huns and died in battle.18 Various expla-
nations may be given for this. For instance, in the seventh and eighth
centuries Agilmund was presented as the son of the first and origi-
nal leader, Agio, thereby constructing a wholly impossible dynastic
continuity.19 Further, he appears to have belonged to a family of the
upper nobility that traced its ancestry, or at least its prominent posi-
tion, back to Wodan. Agilmund was, moreover, ex genere Gugingus and,
even during the lifetime of Paul the Deacon, it was common knowl-
edge that this prosapia (family) was considered generosior (more noble)
than other comparable families.20
However, more decisive for the survival of the institution of king-
ship, after the catastrophe against the Huns, was the fact that Agilmund
had a ready successor in the person of Laiamicho, whom he had
promoted and who was appointed king. This successor had all the
necessary military qualifications and, above all, had already won a
victory against the Huns that could erase the shame that his people
had suffered by their devastating defeat.21
Two factors contributed fundamentally to further stabilise the
Lombard kingship. On the one hand, with the third king Leth and
his descendants a dynasty was established, which ruled for over 100
years during the fifth and sixth centuries and produced a further
seven kings; on the other, there were several notable rulers among
the first kings—like Godeoc who conquered Rugiland, Tato who
defeated the Herules and Wacho who won large parts of Pannonia
for the Lombards. However, the king who out-did and out-shone all
of them was Alboin, who led his people to Italy in 568.22

18
Ibid., 1,16, pp. 55–6. Cf. also Jarnut, Geschichte der Langobarden, pp. 17–8.
19
Origo gentis Langobardorum 2, p. 3; Paul the Deacon, Historia Langobardorum 1,14,
p. 54. Cf. Fröhlich, Studien zur langobardischen Thronfolge 1, pp. 31–2; Schneider,
Königswahl und Königserhebung im Frühmittelalter, pp. 9–10.
20
Edictus Rothari Prol., p. 2; Paul the Deacon, Historia Langobardorum 1,14, p. 54.
Cf. S. Gasparri, La cultura tradizionale dei Longobardi. Struttura tribale e resistenze pagane
(Spoleto 1983), esp. pp. 23–5; 31–2; id., “Kingship rituals and ideology in Lombard
Italy”, Rituals of Power. From Late Antiquity to the Early Middle Ages, ed. F. Theuws and
J.L. Nelson, The Transformation of the Roman World 8 (Leiden-Boston-Köln 2000)
pp. 95–114, esp. pp. 101–2.
21
Cf. Schneider, Königswahl und Königserhebung, pp. 10–2; Fröhlich, Studien zur lan-
gobardischen Thronfolge 1, pp. 35–40; Jarnut, Geschichte der Langobarden, p. 18.
22
Fröhlich, Studien zur langobardischen Thronfolge 1, pp. 41–67; Jarnut, Geschichte der
Langobarden, pp. 19–26; 33–5.
   415

Soon after, though, the kingship plunged into its deepest crisis.
After the assassination of the second Italian king, Clef, in 574, the
duces were unable to decide upon a successor and therefore had to
take over the leadership themselves, of the many small subgroups
into which their gens had split. The Byzantines succeeded in winning
several of them over to their side by considerable payments of money;
some they even managed to recruit as high-ranking officers for the
Roman army. Simultaneously, the emperor entered into a treaty of
aggression with the Franks who had been threatened by Lombard
assaults.23
Less than a decade after the abolishment of their kingship, the
Lombards stood facing the void. The subversion of their gens by
Roman money, combined with the concurrent military threat from
the Imperium Romanum and the superior Franks, raised the fear that
the conquering people, so successful since 568, in Italy might come
to an end that would be no less dramatic than that of the Gepids,
whom they had recently destroyed. In this situation so threatening
to their existence, the duces agreed to elevate one of their own, Clef ’s
son Authari, to the kingship24 and, in this way, to restore continu-
ity to an institution that in the past had lent a certain internal unity
to their gens.
It is important to realise how singular it is, in the history of the
migration-period kingdoms, that the leadership of a gens—in a time
of deep crisis—consciously decided to revive this institution in order
that its unity, capacity to act, even its people’s survival could be
guaranteed. The choice of Authari, the son of Clef, may be inter-
preted as an additional personal measure designed to safeguard the
continuity of the institution of kingship by bringing about a dynas-
tic revival.

23
See, for instance, L.M. Hartmann, Geschichte Italiens im Mittelalter, vol. 2,1: Römer
und Langobarden bis zur Teilung Italiens (Leipzig 1900) pp. 44–50; P. Delogu, “Il regno
longobardo”, Storia d’Italia, vol. 1: Longobardi e Bizantini, ed. id., A. Guillou and G.
Ortalli (Torino 1980) pp. 17–24; Jarnut, Geschichte der Langobarden, pp. 37–9; W.
Pohl, “The Empire and the Lombards: treaties and negotiations in the sixth cen-
tury”, Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. id., The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997) pp. 75–134,
esp. pp. 99–102.
24
Origo gentis Langobardorum 6, p. 5; Paul the Deacon, Historia Langobardorum 3,16,
pp. 100–1.
416  

In short, it can be established that the Lombards probably intro-

duced kingship first and foremost in order to acquire a particularly
efficient political-military system of organisation for the conquest and
maintenance of new patriae. At the same time, their wish to estab-
lish an equality of position with other peoples, and possibly also with
their fellow tribesmen who remained behind on the Lower Elbe, cer-
tainly also played a role.

3. What is the role of a gentile identity (Stammesbewusstsein)

for the establishment of a regnum?

When reading the eighth-century sources, it is always the Lombards,

the gens Langobardorum or the Langobardi, who elected their king—who
consequently became a rex Langobardorum. In other words, it is pre-
sented as the collective action of a community with a shared sense
of identity that founded and at the same time legitimised Lombard
kingship.
According to the sources, all the kings from Rothari onwards pre-
sented themselves as reges gentis Langobardorum in their legislation.25 In
his edicts the king was especially concerned to confirm the tribal,
and allegedly ancient, tradition of Lombard law, the codification of
which would be of general benefit to his gens. The legitimisation of
this law was guaranteed above all by the consilium and by the con-
sensus of the iudices and the warriors, but also per gairethinx secundum
ritus gentis nostrae.26
There is a peculiar contrast between, on the one hand, the tribe-
based strategies of legitimisation used by the kings as law-makers,
and, on the other, the absolute royal title that they conferred upon
themselves in their charters. So far, this contrast has never been sat-
isfactorily explained.
The rex gentis Langobardorum was, therefore, acclaimed by his gens
as their ruler and leader, and in this position was of rank fully equal
to that of the other reges, whether they ruled over the Herules, the
Gepids, the Goths or the Franks. As rex Langobardorum he was the
central focal point of his gens and the embodiment of its tribal con-

25
Edictus Rothari Prol., p. 1.
26
Ibid. Prol., pp. 1–2; 386, pp. 89–90.
   417

sciousness. Evidence for this may be found in the sources of the

sixth to the eighth centuries, though it can only be assumed for the
fourth and the fifth centuries, in which kingship was most probably
established.

4. What sorts of changes in the “constitution” (Verfassung) of a people and

a kingdom (such as central organs of power, local power structures, or links
between the two) are linked to the establishment of a kingdom?
How do socio-economic developments contribute to this process?

Strictly speaking, we can only answer these questions more or less

hypothetically given the state of the sources, as explained in the pre-
vious. We simply do not know the “constitutional” situation of the
Lombards before the election of their supposedly first king, Agilmund.
Nor are these relations much clearer for the period before the con-
quest of Italy in 568. Better insights are gained only with Rothari’s
Edict, with the initiation of a tradition of charter writing, and with
the broader representation of Lombard history given by Paul the
Deacon from the seventh century onwards. However, a vague hint
in the Origo—stating that the Lombards after having left Golaida,
that is, the Lower Elbe, possiderunt aldonus Anthaib, Bainaib seu et
Burgundaib 27—permits the well-founded suspicion that under their first
kings they forced the inhabitants of the conquered territories in east-
ern middle Europe into the status of half-free aldii. This would sug-
gest that they were capable of subjecting these conquered people to
physical, mainly agricultural, labour, while they themselves could lead
free lives as warriors. If this interpretation is correct, the introduc-
tion of kingship may have transformed the gens Langobardorum from
an at one time mainly agricultural society to a military one. This
interpretation is supported by the continuous wars characterising
Lombard history during the sixth and seventh centuries.
Prepared, in a sense, by the conquests of Rugiland and large parts
of Pannonia, which still preserved significant vestiges of Roman infra-
structure, the Lombards in 568 penetrated, in the course of their
conquest of Italy, into a world that to a large extent still bore
the stamp of Late Antiquity due to Gothic rule and to Justinian’s

27
Origo gentis Langobardorum 2, p. 3.
418  

restoration policies. The scope of this article permits the discussion

of only a few aspects of the Lombard “constitution” as it was trans-
formed by the extreme Late Antique character of their newly set-
tled territory. From the perspective of gens, rex and regnum it ought
to be stressed that in Italy, despite all the barbarising regressions of
the preceding three centuries, there still existed a clear concept of
the state and at least partially functioning state-structures that would
have encouraged the development of Lombard kingship and the
transformation of its regnum into a state.
Against this background, the Italian regnum Langobardorum devel-
oped by means of increasing centralisation, hierarchization and insti-
tutionalisation—tendencies which all strengthened its kingship.28 It
may, thus, be assumed, and this is in part confirmed by the sources,
that these developments if not initiated then were at least encour-
aged by the kings. The emergence of Pavia as the royal capital of
the regnum Langobardorum would have been decisive in this context, a
situation that became firmly established from the 620s onwards. With
its sacrum palatium, its magnificent churches and monasteries, and as
a meeting place for the popular assembly, Pavia became both the
factual and the ideological focus of the kingdom.29 In similar ways,
Benevento and Spoleto developed into centres with the duces of the
duchies as their lords.30
The royal sacrum palatium itself became a centre within the centre.
Here, the king exerted his power over all who belonged to his gens.
From there, he could also (to a greater or lesser extent) maintain
his claim to power against the duces of individual civitates, even though
the kings—like Agilulf or Liutprand—at times needed to struggle
with them for decades for precedence.

28
On the development of the Italian regnum, see most recently, for instance,
Delogu, “Il regno longobardo”, esp. pp. 39–61; 83–6; 101–6; Jarnut, Geschichte der
Langobarden, esp. pp. 50–2; S. Gasparri, “Il regno Longobardo in Italia”, Langobardia,
ed. id. and P. Cammarosano (Udine 1990) pp. 237–305; D. Harrison, The Early
State and the Towns. Forms of Integration in Lombard Italy, A.D. 568–774, Lund Studies
in International History 29 (Lund 1993) passim, esp. pp. 223–6.
29
For the role of Pavia, see especially Harrison, The Early State and the Towns,
s.v. Pavia; G.P. Brogiolo, “Capitali e residenze regie nell’Italia longobarda”, Sedes
regiae (ann. 400–800), ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona
2000) pp. 135–62, esp. pp. 144–51 (with further references).
30
For an initial overview, cf. Harrison, The Early State and the Towns, ss. vv.
Spoleto, Benevento.
   419

Of importance to the realm, and especially for the kingship, were

the royal domains organised by or possibly even first created in 584,
which were led by gastalds, actionarii and other lower administrators.
In the gastalds, the kings had ministers who were more dependent
on them than the duces. Since Rothari’s Edict, and especially since
Liutprand, there are numerous examples of kings trying to establish
these, and other office-holders, in a hierarchical organisation at the
head of which stood the all-ruling and controlling king.31
Besides the centralising and hierarchical features favouring the
kingship, another factor contributing to strengthening its position was
the development of new fields of activity through the use of the
Italian leges. King Rothari expressly emphasised that he had let old
Lombard law be first recorded in written form in the interest pre-
dominantly of those with little power. The results of this undertak-
ing were acclaimed in a splendidly staged event at the royal palace
of Pavia by the great men and warriors of his people per gairethinx
secundum ritus gentis nostrae. The king, who saw his work as a divine
mission, could then convincingly present himself as the law-giver and
highest judge acknowledged by the people. None of his duces or other
magnates held a position that even approached his, and he was
therefore able to impressively demonstrate his God-given uniqueness.
It is hardly surprising that it was the most powerful Lombard kings,
from Grimoald through to Liutprand and Aistulf, who sought to pre-
sent themselves as great law-givers led by God, thus confirming their
unique position within their people.32
From the sixth century onwards, the kings succeeded in strength-
ening their position to a possibly even greater extent by laying claim
to the right to decide on religious matters in the realm. With this
we touch upon one of the most controversial chapters in the history
of the Lombards: their Christianisation and consequently their Arianism

31
Paul the Deacon, Historia Langobardorum 3,16, p. 101. On the royal domains
and the office-holders, see, for instance, Jarnut, Geschichte der Langobarden, esp. pp.
47–52; Gasparri, “Il regno longobardo in Italia”, pp. 254–62; 274–7; Harrison, The
Early State and the Towns, esp. pp. 98–157.
32
Cf. Delogu, “Il regno longobardo”, pp. 55–61; Harrison, The Early State and
the Towns, esp. pp. 105–9; 183–7; id., “Political rhetoric and political ideology in
Lombard Italy”, Strategies of Distinction. The Construction of Ethnic Communities, 300–800,
ed. W. Pohl with H. Reimitz, The Transformation of the Roman World 2 (Leiden-
New York-Köln 1998) pp. 241–54, esp. pp. 242–5.
420  

or Catholicism.33 It is impossible to treat this subject in all its detail

here, and it must suffice to note that the Lombards, even before the
conquest of Italy (certainly at least from the time of their migration
from the Lower Elbe), had been exposed to Christian influences,
and, from at least the sixth century, had partly converted to Christian-
ity, be it in an Arian or a Catholic form. The majority of the
gens, however, remained faithful to its traditional religion centred on
Wodan. In any case, the emissaries of King Audoin could in the
late 540’s maintain to Emperor Justinian that the Lombards had
been Catholic from the beginning;34 whereas King Authari, on the
other hand, forbade the Lombards to have their children baptised
as Catholics.35 However, his successor, Agilulf,—influenced by his
Catholic wife Theodelinde—did exactly that when baptising his son
Adaloald in 603.36 With this act the complete Catholicisation of a
people, which up to then had been religiously fragmented, was ini-
tiated. The entire seventh century is characterised by the success of
the Catholic Church in silencing, with royal help, all its pagan and
confessional opponents in the Lombard realm; a process we see from
the baptism of Adaloald to the prohibition of Arianism by King
Aripert I (653–662)37 and to the defeat of the Three-Chapters-Schism
brought about by the efforts of King Cunincpert in the year 698.38
At the end of the seventh century, the Catholic Church in this way
had risen to become the state Church, a development that even
Arian kings, like Rothari and Grimoald, could not or would not
change. Simultaneously, daily life became increasingly Christianised

33
A lot of progress has been made on isolated questions regarding this chapter
of Lombard history, but still the rather hypothetical and sometimes speculative
research of G.P. Bognetti remains standard. Its essence has been gathered in L’età
longobarda, 4 vols. (Milano 1966–68). Of particular significance is his monumental
“S. Maria foris portas di Castelseprio e la storia religiosa dei Longobardi”, first
published in 1948 and reprinted in vol. 2 of L’età longobarda, pp. 12–683. See,
further, Delogu, “Il regno longobardo”, esp. pp. 30–3; 40–2; 96–101; Jarnut, Geschichte
der Langobarden, esp. pp. 53–4; 66–71; 125–7; Harrison, The Early State and the Towns,
esp. pp. 73–82; 107–9; 152–3; 171–83; 204–12.
34
Procopius, Gotenkriege 3,34, ed. O. Veh [Prokop, Werke] (München 1966)
p. 662.
35
Gregorii I Papae Registrum Epistolarum 1,17 (a. 591), ed. P. Ewald, MGH EE 1
(Berlin 1887) p. 23.
36
Paul the Deacon, Historia Langobardorum 4,27, p. 125.
37
Carmen de synodo Ticinensi, ed. L. Bethmann, MGH SSrL (Hannover 1878)
p. 190, ll. 3–5.
38
Ibid., p. 190, l. 21–p. 191, l. 15.
   421

and could therefore indirectly, that is, via the churches, which were
in his power, be shaped and controlled by the kings.
It is noteworthy that from the second half of the seventh century,
the kingship increasingly expanded its monopoly, which it had had
since the days of Rothari, to mint coins.39 The result was that by
the reign of King Desiderius no coins were produced apart from
those originating under royal control. From the end of the seventh
century, the king’s likeness and his name on the coinage gave him
a presence in everyday Lombard life that could not be attained via
any other institution.40
The economic and social differentiation that divided the Lombards
into poor and rich, for which there is much evidence in the seventh
century and which rapidly accelerated in the eighth, influenced daily
life even more. This had important consequences for their fitness for
military service according to the requirements still set in the time of
King Rothari: the process of social differentiation had apparently led
to a situation in which fewer Lombards met the criteria for being
allowed to bear weapons. The Kings Liutprand and Aistulf, there-
fore, sought to resolve the military effects of that differentiation-
process—and thereby to influence the process as a whole—by regulating
the responsibilities of the impoverished exercitales towards the king.41
The king further secured a dominant position within his gens by
having greater power than any other Lombard to determine who
belonged to the gens and who did not. He had, for instance, a priv-
ileged position in the liberation of slaves, by having the power to
determine if a freedman could become a member of the gens
Langobardorum.42 Further, it was principally the kings who decided
whether a warrior from another people was to be permitted a

39
Edictus Rothari 242, p. 60.
40
Cf. E. Bernareggi, Moneta Langobardorum (Milano 1983, English transl. Lugano
1989) passim; Harrison, The Early State and the Towns, pp. 119–23; J. Jarnut,
“Münzbilder als Zeugnisse langobardischer Herrschaftsvorstellungen”, Iconologia sacra.
Festschrift für K. Hauck zum 75. Geburtstag, ed. H. Keller and N. Staubach, Arbeiten
zur Frühmittelalterforschung 23 (Berlin-New York 1994) pp. 283–90; A. Rovelli,
“Some considerations on the coinage of Lombard and Carolingian Italy”, The Long
Eighth Century, ed. I.L. Hansen and C. Wickham, The Transformation of the Roman
World 11 (Leiden-Boston-Köln 2000) pp. 195–223.
41
Leges Liutprandi 83, ed. F. Bluhme, MGH LL 4 (Hannover 1968) pp. 140–1;
Leges Ahistulfi 2,3, ed. F. Bluhme, ibid., p. 196. Cf. also Delogu, “Il regno longo-
bardo”, pp. 128–9; 133–44; 169; Jarnut, Geschichte der Langobarden, pp. 97–101.
42
Edictus Rothari 224, pp. 54–5; Leges Liutprandi 55, p. 129.
422  

similar role in the gens Langobardorum. Indeed, already King Tato’s

victory over the Herules in 508 had led to the admission of war-
riors from various other gentes into the Lombard army, by which
they had become Lombards.43 Alboin’s bloody victory over the Gepids
in 567 had a similar effect.44 In 568, during his conquest of Italy,
the latter also succeeded in integrating Saxons, Gepids, Bulgars,
Sarmatians, Pannonians, Sueves and Noricans into the Lombard
army, by which it was decisively strengthened.45 Along similar lines,
King Grimoald in the 660s took a rather large group of wandering
Bulgars into his realm and settled them in the vicinity of Isernia.46
Since the reign of Rothari the kings had maintained that groups
wandering into the realm placed themselves under royal protection
and thus, legally, became Lombards.47
Surveying the changes in the “constitution”, which can be traced
in the sources from the establishment of Lombard kingship, we may
conclude that throughout its development the kingship was evidently
able to intensify and institutionalise its dominion and to open up
new areas of activity. In doing so, it naturally also reacted to the
significant social and economical changes, taking place after the estab-
lishment of Lombard rule in Italy, and used those, at least partially,
for its own ends.

5. What is the role of kings in this development?

Since the changes in Lombard society and rulership, discussed in

the previous section, would ultimately be incomprehensible if not
placed within the context of the development of kingship, some rep-
etition is inevitable in this section. This is also the reason that we
will first return to the development of Pavia as the capital of the
regnum, to the self-representation of the kings—from Rothari onwards—
as divinely commissioned legislators and supreme judges, and to the
key position that the kings, from the seventh century onwards, played
in the religious development of their people.

43
Paul the Deacon, Historia Langobardorum 1,20, p. 59.
44
Ibid., 1,27, pp. 69–70.
45
Ibid., 2,26, p. 87.
46
Ibid., 5,29, p. 154.
47
Edictus Rothari 367, p. 85.
   423

The paramount importance of the kings and of kingship for the

development of the gens Langobardorum in Italy was, from the reign
of Agilulf (590–616), inseparably connected to their claim of being
rex Dei gratia.48 This claim, repeated continuously in, for instance,
laws and charters, of having received their royal office by divine
grace, elevated the kings and their activities far above the other
members of their gens. At the same time, it strengthened the posi-
tion of the filicissima gens Langobardorum as, thus, being led by God.
It is remarkable that in a gens, which until the end of the seventh
century was as religiously diverse as the Lombards, this concept of
divine grace could be formulated by an Arian already by the end
of the sixth century. The concept was consolidated in the seventh
century under his Agilolfing Catholic successors, who made Catholicism
the state religion. In the prologue to his new laws, King Liutprand
could finally, in 717, formulate the by now indissoluble bonds between
royal divine grace and the Catholicism of his people with the fol-
lowing Intitulatio: Ego in Dei omnipotentis nomine Liutprand excellentissimus
rex gentis filicissimae ac catholicae Deoque dilectae Langobardorum.49
Surveying the role of kingship in relation to the changes in the
Lombard “constitution” (Verfassung), we have first to conclude that
well-founded statements about this subject can only be made for the
period from the end of the sixth century at the earliest. From the
re-establishment of the kingship on, the kings clearly succeeded in
securing and expanding both their acquired position of leadership
as well as their traditional role as champions and supreme political
leaders of their gens. They achieved this primarily by developing new
policies; for example, in the areas of law and religion and by the
centralisation and hierarchical organisation of government. The
main reason that this was possible was that they convincingly pre-
sented their leading position, over the increasingly Christian and
Catholic gens, as bestowed on them gratia Dei. Obviously, this involved
a dialectical process between the religious changes within the gens
Langobardorum—which became increasingly rooted in its Catholic
environment—and its Catholic kings; the latter gaining from this

48
S.R. Elze, “Per la storia della corona del tesoro di Monza”, Atti del VI° con-
gresso internazionale di studi sull’alto medioevo (Spoleto 1980) pp. 393–400; cf. further
Delogu, “Il regno Longobardo”, p. 43; Jarnut, Geschichte der Langobarden, pp. 45–6;
Harrison, The Early State and the Towns, pp. 118; 189.
49
Leges Liutprandi a. 5, Prol., p. 109.
424  

religious process in so far as it considerably strengthened their own

position.

6. What part does the Roman Empire play in this process?

The influence of the Imperium Romanum on the Lombards has often

been underestimated in older research, mainly because the latter
were, according to the often-cited opinion of Velleius Paterculus, a
gens etiam Germana feritate ferocior.50 However, since this view dates from
the period of Augustus it renders its validity for the situation dur-
ing the Migration Period at least questionable.
When the Lombards in 488 occupied Rugiland, i.e. Lower Austria,
they entered for the first time a region that had been influenced by
Roman culture and civilisation. When they took possession of Pan-
nonia after 510, they acquired old Roman provinces in which—
even if they had only been under loose imperial control—(an admittedly
weakened) Roman infrastructure survived, as well as the remains of
Late Antique social and economic systems.
From the moment the Lombards penetrated these old spheres of
culture, they, like other barbarian peoples in comparable circum-
stances, developed life-styles and forms of settlement that, on the one
hand, as far as possible, preserved their old traditions, and, on the
other, enabled the necessary adaptation to their new environment.
According to the many and readily available excavation results, they
settled as self-contained military units in or around fortifications of
various kinds. Here, they also received tribute from the subject provin-
cial Roman population. Already in Pannonia, they developed a pre-
liminary version of the system of dominion over a large area that
would characterise the first phase of their conquests in Italy.51

50
Cf. above, note 2.
51
On the history of the Lombards between 488 and 568, see, most recently, Menke,
“Archeologia longobarda tra la bassa Elba e l’Ungheria”, esp. pp. 87–104; J. Jarnut,
“Die langobardische Herrschaft über Rugiland und ihre politischen Hintergründe”,
Westillyricum und Nordostitalien in der spätrömischen Zeit, ed. R. Brato≥ (Ljubljana 1996)
pp. 207–213; id., “Die Langobarden zwischen Pannonien und Italien”, Slovenija in
sosednje de≥ele med antiko in karolin“ko dobo. Zaèetki slovenske etnogeneze—Slowenien und die
Nachbarländer zwischen Antike und karolingischer Epoche. Anfänge der slowenischen Ethnogenese
1, ed. R. Brato≥ (Ljubljana 2000) pp. 73–9; id., “I Longobardi nell’epoca prece-
dente all’occupazione dell’Italia”, pp. 24–33; Christie, The Lombards, pp. 20–68.
   425

Their integration into the Byzantine state reached a high-point

during the 550s, when large Lombard military units, consisting of
several thousand men, fought against the Ostrogoths in Italy and
against the Persians in the eastern Mediterranean. Their status as
foederati forced them to develop forms of military organisation that
could bridge the differences between their archaic Germanic her-
itage and the highly complicated Byzantine military system. In this
period, the functions of dux and comes took shape based on an amal-
gam of Byzantine and “Germanic” elements. On the one hand, the
dux and comes (for which translations of “duke” and “count” are far
from adequate) were officers after Byzantine example. On the other
hand, they followed “old Germanic” traditions as leaders of their
own retinue of military units over which they exercised supreme mil-
itary and judicial power. In this arrangement the dux preceded the
comes, and both fell under the king. Already at this time, the de facto
fragmentation of power and the resulting antagonism between king
and duces must have taken shape; a situation that would come to
decide the history of the Lombards in Italy.52

In short, we may conclude that the Lombards, in an environment

whose stone buildings, streets and economic system must to their
eyes still have seemed Roman, were influenced by Late Antique cul-
tural forms even before their entrance into Italy. This was all the
more so considering that in that sphere of life (which as warriors
would have influenced them most) they were, as foederati, integrated
into the East Roman state as part of the Byzantine army.53
Compared to Pannonia how much stronger must this Roman
influence have been in Italy, core land of the Imperium? Against this
background, it would be inconceivable if it had not also influenced
the development of the central institutions of the gens Langobardorum;
that is, its kingship. Some aspects of this influence have been dis-
cussed already, a few more will be discussed in the following.
For the survival of the revived kingship after its most serious cri-
sis, the interregnum of 584, it was undoubtedly of crucial importance

52
Cf. Hartmann, Geschichte Italiens im Mittelalter 2,1, pp. 38–45.
53
Cf. G.P. Bognetti, “L’influsso delle istituzioni militari romane sulle istituzioni
longobarde del seccolo VI e la natura della fara”, id., L’età longobarda 3 (Milano
1967) pp. 1–46; S. Gasparri, I duchi Longobardi, Studi storici 109 (Roma 1978), esp.
pp. 10–2.
426  

that the duces furnished it with half of the economic potential that
had fallen to them, the substantiae, and that for this arrangement—
as for previous ones—the custom of hospitalitas played a decisive part.
However one wishes to interpret the two relevant and endlessly dis-
cussed passages in Paul the Deacon, concerning the division of land
in Italy after the death of Clef and after the end of the interregnum,
all the interpretations agree that it must be recognised that the eco-
nomic organisation, and hence the economic existence of the king-
ship and the duces, ultimately rested on Roman foundations.54
Certainly, it is no coincidence that Authari, who managed to splen-
didly fulfil the tasks given to him in 584 as integrator and defender
of his gens, added the epithet Flavius to his name.55 Thus he estab-
lished an association with Romano-Gothic traditions, which eased
the integration of non-Lombard subjects into the Lombard state. The
gradual rise to power of Pavia as the central royal city in the sev-
enth century, with its sacrum palatium as focal point, can only be
understood with due consideration of the imperial sacrum palatium and
the role of the imperial capital Constantinople/Byzantium. Further,
the position of the king as legislator and highest judge, the figure
that would dominate Lombard legal life from Rothari onwards, is
only fully comprehensible when one considers the Roman emperor,
legislator and judge, as its model. It goes without saying that it was
Roman examples that inspired the Lombard king to date the char-
ters produced in his realm by the year of his rule, and to mint coins
in his name.
It was also part of the king’s increasingly imperial style of rule
that he, like the emperor, represented himself as ruling by divine
grace, and that he, at least from the second half of the seventh cen-
tury when Catholicism became the state religion, maintained an ever
closer control of the Church.

54
Paul the Deacon, Historia Langobardorum 3,16, pp. 100–1. On this problem, see,
for instance, J. Jarnut, “Die Landnahmen der Langobarden aus historischer Sicht”,
Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters. Methodische
Grundlagendiskussion im Grenzbereich zwischen Archäologie und Geschichte 1, ed. M. Müller-
Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen 1993) pp.
173–94, esp. pp. 187–90; P. Delogu, “Longobardi e Romani: Altre congetture”,
Langobardia, ed. S. Gasparri and P. Cammarosano (Udine 1990) pp. 111–68, esp.
pp. 111–20, though especially also W. Pohl, The Empire and the Lombards, pp. 112–31.
55
Paul the Deacon, Historia Langobardorum 3,16, p. 101.
   427

The royal potential for integration, secured by imitatio imperii,

facilitated the incorporation of conquered peoples and of those
Romans who from the 720s subjected themselves voluntarily after
the religiously-based crisis of Byzantine rule. It thus led to a certain
strengthening of the regnum Langobardorum, and above all also to a
transformation of the gens Langobardorum based on a process of increas-
ing Romanisation.56
In conclusion, we may assert that in the memory of a people,
who at the time in which their historical tradition began (that is, in
the seventh century) was completely shaped by its kingship, a knowl-
edge existed that there had been another kind of rule before this
institution had come into being. In other words, we may accept that
their gens was older than their regnum. Nevertheless, royal rulership
was so decisive for the Lombards that it existed for at least 400
years and even survived the fall of their independent realm. The re-
establishment of the kingship, after its most serious crisis during the
interregnum, proves that, in the imagination and experience of its sixth-
century contemporaries, it was clearly of vital importance for the
survival of their gens, which at that point threatened to cease exist-
ing altogether. The increasing strength of the kingship after 568 was
essentially based on the fact that the kings—at the latest from the
time of the Flavian Authari—began to develop an increasingly far-
reaching imitatio imperii, which facilitated the involvement and inte-
gration of the Roman majority into the Lombard state. This, however,
also led to a change of its people, the gens Langobardorum, a process
that went largely unnoticed by contemporaries.57 The decisive con-
ditions for this lay, from the seventh century onwards, in the increas-
ing Catholicism and sacrality of the kingship that made the integration
of the members of the regnum Langobardorum into a gens catholica
Langobardorum at all possible.

[Translated by Inge Lyse Hansen and Carine van Rhijn]

56
Cf. Delogu, “Il regno longobardo”, pp. 133–7; Jarnut, Geschichte der Langobarden,
pp. 97–105.
57
On the use of the term gens as an often functional construct, see now the arti-
cles in Strategies of Distinction. The Construction of Ethnic Communities, 300–800, ed. W.
Pohl with H. Reimitz, The Transformation of the Roman World 2 (Leiden-New
York-Köln 1998), especially the contributions by W. Pohl, “Introduction: Strategies
of distinction”, pp. 1–15 and “Telling the difference: Signs of ethnic identity”, pp.
17–69.
This page intentionally left blank
 THE BAVARIANS

Matthias Hardt

The early medieval Bavarians differ from the other regna and gentes
discussed in this volume by their great regional continuity and their
lack of formal kingship. Nonetheless, ethnogenesis took place in the
Roman provinces of Rhaetia and Noricum. As early as the mid-
sixth century, contemporary with the earliest mention of the Bavarians,
a family related to the house of the Agilolfings appeared, who ruled
the Bavarians in a nearly king-like manner up to the late eighth cen-
tury, despite being increasingly dependent upon the royal leadership
of the Frankish kingdom. Ethnogenesis, a royal ruling family, and
long-standing political independence justify the incorporation of the
Bavarians in a comparison of the kingdoms of the Migration Period
and the early Middle Ages. In order to render a real comparison
possible, the questions set out by the group will be followed in the
ensuing discussion. In view of the scarce source-material and an early
historiography that was exclusively written from the perspective of
neighbouring people, however, it will not always be easy to work
within this framework.

1. Is there a development from a Germanic gens of the Migration Period to

a Germanic kingdom? Or does a gens (or this gens) not exist until after the
establishment of a kingdom?

Any understanding of the relationship between ethnogenesis and king-

ship is closely connected to an interpretation of the name of the
people and a clarification of the provenance of the Bavarians; even
in the first half of the twentieth century, these questions were con-
sidered some of the “most difficult and controversial problems of
German history”.1 Further, there are two opposing opinions as to

1
L. Schmidt, Die Westgermanen (München 1940; repr. 1970) p. 194, similarly id.,
“Zum Ursprung der Baiern”, Zeitschrift für bayerische Landesgeschichte 10 (1937) pp.
12–8, esp. p. 12.
430  

whether the gens migrated as a single body into the region south of
the Danube, or rather whether an ethnogenesis took place in the
lands later possessed by the gens.2 Already in his biography of the
holy Columbanus, Jonas of Bobbio connected the Celtic Boii, who
in his days had long disappeared, with the Baioarii.3 From the very
outset, he thus had in mind a specific way of interpreting the Bavarians’
name, which a substantial part of present-day research still rightly
retains, namely that their name is foreign4 and indicates people from
a region called *Baiahaim5 or Baia. For a long time, however, there
was no uniform notion where this supposed home-land of the Bavarians

2
For the controversy on this question and on the history of research see E. Kle-
bel, “Langobarden, Bajuwaren, Slawen”, Mitteilungen der Anthropologischen Gesellschaft
in Wien 69 (1939) pp. 41–116 [repr. id., Probleme der bayerischen Verfassungsgeschichte
(München 1957) pp. 1–89], esp. pp. 48–54; R. Wenskus, Stammesbildung und Verfas-
sung. Das Werden der frühmittelalterlichen gentes (2nd edn., Köln-Wien 1977) pp. 560–9;
K. Reindel, “Das Zeitalter der Agilolfinger (bis 788)”, Handbuch der bayerischen Geschichte 1,
ed. M. Spindler (2nd edn., München 1981) pp. 101–245, esp. pp. 101–16; H. Beck,
S. Hamann and H. Roth, “Bajuwaren”, Reallexikon der Germanischen Altertumskunde 1
(2nd edn., 1973) pp. 601–27, esp. pp. 606–7 and 611–3; M. Menke, “150 Jahre
Forschungsgeschichte zu den Anfängen des Baiernstammes”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 123–220; H.L.G. Gastroph, Herrschaft und Gesellschaft in der Lex Baiuvariorum.
Ein Beitrag zur Strukturanalyse des Agilolfingischen Stammesherzogtums vom 6. zum 8. Jahrhundert,
Miscellanea Bavarica monacensia 53. Neue Schriftenreihe des Stadtarchivs München
71 (München 1974) pp. 20–35; A. Kraus, “Die Herkunft der Bayern. Zu Neuerschei-
nungen des letzten Jahrzehnts”, Bayerisch-schwäbische Landesgeschichte an der Universität
Augsburg 1975–1977, ed. P. Fried, Veröffentlichungen der Schwäbischen Forschungs-
gemeinschaft bei der Kommission für Bayerische Landesgeschichte, Reihe 7, Augsburger
Beiträge zur Landesgeschichte Bayerisch-Schwabens 1 (Sigmaringen 1979) pp. 27–46,
esp. pp. 27–8; 30; K. Reindel, “Die Bajuwaren. Quellen, Hypothesen, Tatsachen”,
Deutsches Archiv 37 (1981) pp. 451–73, esp. pp. 454–7.
3
Jonas of Bobbio, Vita Columbani 2,8, ed. B. Krusch, MGH SSrG 37 (Hannover
1905) p. 244: ad Boias, qui nunc Baioarii vocantur, tendit. See also Kraus, “Die Herkunft
der Bayern”, pp. 31–2; H. Wolfram, 378–907: Grenzen und Räume. Geschichte Öster-
reichs vor seiner Entstehung, Österreichische Geschichte 3 (Wien 1995) p. 72.
4
H. Wolfram, Salzburg, Bayern, Österreich. Die Conversio Bagoariorum et Carantanorum
und die Quellen ihrer Zeit, Mitteilungen des Instituts für Österreichische Geschichtsfor-
schung, Ergänzungsband 31 (Wien-München 1995) pp. 23–4; See on this kind of
name creation also W.P. Schmid, “Vidivarii”, Sprach- und Kulturkontakte im Polnischen,
Specimina Philologicae Slawicae, Supplementband 23 (München 1987) pp. 349–58;
W. Hartung, Süddeutschland in der frühen Merowingerzeit. Studien zu Gesellschaft, Herrschaft,
Stammesbildung bei Alamannen und Bajuwaren, Vierteljahreshefte zur Sozial- und
Wirtschaftsgeschichte, Beiheft 73 (Wiesbaden 1983) p. 175. Hartung can, however,
not be followed in his interpretation of the name.
5
E. Schwarz, “Herkunft und Einwanderungszeit der Baiern”, Südost-Forschungen
12 (1953) pp. 21–47, esp. p. 22.
   431

might be.6 The Anonymous Ravenna Geographer, writing in the first

half of the ninth century, knew of a hilly region near the Elbe called
Baia[s],7 which bordered on the Danube and Pannonia.8 The Geo-
grapher, perhaps copying a Gothic example,9 may have described this
region as inhabited by Marcomanni, but on the whole this area can-
not, altogether, be considered as, or equated to, Bohemia.10 According
to Heinz Löwe and others, this region could also have been located
on the eastern edge of Bohemia and Moravia, north of modern Bra-
tislava in Slovakia.11
It is also often stated that the Bavarians, because of their Bohemian
origins, must have descended from the Marcomanni, who had been
settled in that region since the early Roman imperial period.12 These,

6
E. Schwarz, “Ermunduren—Thüringer, Böhmen—Baiern”, Et multum et multa.
Beiträge zur Literatur, Geschichte und Kultur der Jagd. Festgabe für Kurt Lindner zum 27.
November 1971, ed. S. Schwenk, G. Tilander and C.A. Willemsen (Berlin-New York
1971) pp. 341–9, esp. pp. 345–6.
7
G. Neumann, “Baias”, Reallexikon der Germanischen Altertumskunde 1 (2nd edn.,
1973) p. 600; J. Schnetz, “‘Baias’ und der Baiernname”, Zeitschrift für bayerische Landes-
geschichte 16 (1951) pp. 1–19.
8
Ravennas Anonymus, Cosmographia. Eine Erdbeschreibung um das Jahr 700 18,1–2,
transl. J. Schnetz, Nomina Germanica 10 (Uppsala 1951) p. 61 [henceforth: Ravennas
Anonymus, Cosmographia (transl.)]; Ravennas Anonymus, Cosmographia, ed. M. Pinder
and G. Parthey [Ravennatis Anonymi Cosmographia et Guidonis Geographica] (Berlin 1860;
repr. Aalen 1962) p. 213: [. . .] est patria quae dicitur Albis Ungani, montuosa per longum,
quasi ad Orientem multum extenditur, cuius aliqua pars Baias dicitur. About the time of the
Geographer’s compilation see M. Springer, “Riparii—Ribuarier—Rheinfranken nebst
einigen Bemerkungen zum Geographen von Ravenna”, Die Franken und die Alemannen
bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände zum Real-
lexikon der Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 201–69,
here p. 234.
9
Ravennas Anonymus, Cosmographia (transl.) 4,13, p. 58 refers to scholars of
Gothic provenance. Ravennas Anonymus, Cosmographia 4,13, p. 201: Aitanaridus et
Eldewaldus et Marcomirus Gothorum philosophi. Critically about the Gothic example of
the Geographer Springer, “Riparii”, pp. 236–45.
10
For the interpretation of Baia as Baiahaim = Bohemia, see Schnetz, “‘Baias’”,
pp. 2–6.
11
H. Löwe, “Die Herkunft der Bajuwaren”, Zeitschrift für bayerische Landesgeschichte
15 (1949) pp. 5–67, here pp. 8–19, esp. p. 19; H. Zeiß, “Von den Anfängen des
Baiernstammes”, Bayerische Vorgeschichtsblätter 13 (1936) pp. 24–40, esp. pp. 26 and 28.
12
Schmidt, “Zum Ursprung der Baiern”, pp. 16–8, presupposes a Marcomannic
settlement in Rhaetia after 451; Löwe, “Die Herkunft der Bajuwaren”, pp. 22–32;
Wenskus, Stammesbildung, p. 563. Arguments against a Marcomannic background
have been presented by E. Klebel, “Baierische Siedlungsgeschichte”, Zeitschrift für
bayerische Landesgeschichte, pp. 75–82, esp. pp. 78–9; Schwarz, “Herkunft und Einwan-
derungszeit”, pp. 23–5; E. Schwarz, “Die bairische Landnahme um Regensburg im
Spiegel der Völker- und Ortsnamen”, Beiträge zur Namenforschung 1 (1949/50) pp.
51–71, esp. p. 52.
432  

however, increasingly lost importance through to the end of the

Migration Period.13 The last certain reference to the Marcomanni
can be found in the Notitia Dignitatum for the period around 396,
which refers to the region around Vienna.14 At this time, Queen
Fritgil turned to Saint Ambrose asking for a broadening of the scope
of Christian teaching that was already available.15
Joachim Werner considered what he saw as the migration of
Germanic groups from Bohemia in the 540s A.D., as linked to the
Frankish defeat of the Thuringian realm and to the Lombard immi-
gration into Pannonia after the death of Theoderic the Great in
526.16 Jörg Jarnut, too, has assumed a migration of a rather large
Lombard group from Bohemia into the lands between the Danube
and the Alps caused by a re-orientation of Lombard politics under
King Wacho (511–539).17 It is also possible that Suebian groups

13
Schwarz, “Herkunft und Einwanderungszeit”, pp. 22–3; 39–40; id., “Das Ende
der Völkerwanderungszeit in Böhmen und die Herkunftsfrage der Baiern”, Bohemia
8 (1967) pp. 23–58, esp. pp. 38–9; H.W. Böhme, “Zur Bedeutung des spätrömi-
schen Militärdienstes”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer
and H. Dopsch (Rosenheim-Salzburg 1988) pp. 23–37, esp. p. 30, also emphasizes
the archaeological evidence for Marcomannic participation in the settlement of the
Danube region in the fifth century.
14
Notitia Dignitatum accedunt Notitia Urbis Constantinopolitanae et Latercula Prouinciarum
34, ed. O. Seeck (Berlin 1876; repr. Frankfurt a. M. 1962) p. 24: tribunus generis
Marcomannorum. See further also H. Castritius, “Die Grenzverteidigung in Rätien
und Noricum im 5. Jh. n. Chr. Ein Beitrag zum Ende der Antike”, Die Bayern und
ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz, Denkschriften der Österreichi-
schen Akademie der Wissenschaften, philosophisch-historische Klasse 179. Veröffent-
lichungen der Kommission für Frühmittelalterforschung 8 (Wien 1989) pp. 17–28,
esp. pp. 20–1; F. Lotter, “Die germanischen Stammesverbände im Umkreis des
Ostalpen-Mitteldonau-Raumes nach der literarischen Überlieferung zum Zeitalter
Severins”, ibid., pp. 29–59, esp. p. 46.
15
Paulinus, Vita Sancti Ambrosii 36, ed. M. Pellegrino (Rom 1961) p. 102. See
also Lotter, “Die germanischen Stammesverbände”, pp. 45–6; Wolfram, 378–907:
Grenzen und Räume, p. 45.
16
J. Werner, “Die Herkunft der Bajuwaren und der ‘östlich-merowingische
Reihengräberkreis’”, Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner,
Schriftenreihe zur bayerischen Landesgeschichte 62 (München 1962) pp. 229–50
[repr. Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt
1965) pp. 12–43], esp. pp. 24–43. Also, H. Wolfram, “Ethnogenesen im früh-
mittelalterlichen Donau- und Ostalpenraum (6. bis 10. Jahrhundert)”, Frühmittel-
alterliche Ethnogenese im Alpenraum, ed. H. Beumann and W. Schröder, Nationes 5
(Sigmaringen 1985) pp. 97–151, esp. pp. 105–8, comes to the conclusion, that
already in the beginning of the sixth century, Lombard groups from Bohemia may
have contributed significantly to the formation of the Bavarian tribe. Similar ideas
in Klebel, “Langobarden, Bajuwaren, Slawen”, pp. 68–9.
17
J. Jarnut, Agilolfingerstudien. Untersuchungen zur Geschichte einer adligen Familie im 6.
   433

from the land of Baia, as defined previously, found their way into
Bavaria18 as early as the fifth or early sixth century; at least, some
of them attacked Passau shortly after 476 under the leadership of
Hunimund.19
Meanwhile, however, it has become clear that Germanic groups
from Bohemia and Moravia were already stationed as Roman foe-
derati and auxiliary troops in military camps on the Danube, even
before Odoaker ordered the evacuation of Noricum and Rhaetia.20

und 7. Jahrhundert, Monographien zur Geschichte des Mittelalters 32 (Stuttgart 1986)

p. 50; Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum”,
pp. 106–7; id., Salzburg, Bayern, Österreich, p. 23. For other Lombard groups, see also
A. Schmid, “Bayern und Italien vom 7. bis zum 10. Jahrhundert”, Die transalpinen Ver-
bindungen der Bayern, Alemannen und Franken bis zum 10. Jahrhundert, ed. H. Beumann and
W. Schröder, Nationes 6 (Sigmaringen 1987) pp. 51–91, esp. pp. 55–60. G. Haupt-
feld, “Die Gentes im Vorfeld von Ostgoten und Franken im sechsten Jahrhundert”,
Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
179. Veröffentlichungen der Kommission für Frühmittelalterforschung 8 (Wien 1989)
pp. 121–34, esp. pp. 124–5, presumes Germanic retinues from Bohemia, that may
have participated in the formation of the Lombard and Bavarian tribe during the
time of Theoderic the Great.
18
A Quadian provenance of the Bavarians is advocated by Schwarz, “Herkunft
und Einwanderungszeit”, pp. 25–7. On the Danube Suebians, see Klebel, “Lango-
barden, Bajuwaren, Slawen”, pp. 33–4; Wenskus, Stammesbildung, pp. 561–2; F. Lotter,
“Zur Rolle der Donausueben in der Völkerwanderungszeit”, Mitteilungen des Instituts
für Österreichische Geschichtsforschung 76 (1968) pp. 275–98; on the Suebians from the
province of Savia, see Lotter, “Die germanischen Stammesverbände”, pp. 46–7.
19
Wolfram, 378–907: Grenzen und Räume, pp. 38–9.
20
Böhme, “Zur Bedeutung des spätrömischen Militärdienstes”, pp. 29–30. Such
an early date is at least supported by the early graves from Azlburg I and II, as
well as Straubing-Bajuwarenstraße, which have been dated respectively as early as
the fourth century and to the middle of the fifth century or shortly afterwards; see
H. Geisler, “Neue archäologische Quellen zur frühesten Geschichte der Baiern: spät-
antike und frühmittelalterliche Gräberfelder aus Straubing”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 89–100, esp. p. 91; Th. Fischer, Das bajuwarische Reihengräberfeld von Staubing.
Studien zur Frühgeschichte im bayerischen Donauraum, Kataloge der Prähistorischen Staats-
sammlung 26 (Kallmünz 1993) pp. 124–5; M. Martin, “Die Gräberfelder von
Straubing-Bajuwarenstraße und Straßkirchen—zwei erstrangige Quellen zur Geschichte
der frühen Baiern im Straubinger Land”, Frühe Baiern im Straubinger Land (Straubing
c. 1995) pp. 17–39, esp. pp. 20–1; Th. Fischer, Die Römer in Deutschland (Stuttgart
1999) pp. 174–5. On the situation in Rhaetia in the fifth century, see also K. Dietz,
U. Osterhaus, S. Rieckhoff-Pauli and K. Spindler, Regensburg zur Römerzeit (Regensburg
1979) pp. 155–66. On the end of the border army, see also Castritius, “Die Grenzver-
teidigung in Rätien”, p. 26.
434  

Grave-goods of the so-called Friedenhain-PÏre“tovice-group,21 from

Barbing-Irlmauth22 as well as from Straubing and outlying areas,23
suggest the arrival of smaller groups and individuals from southern
Bohemia in the area around Regensburg and Straubing24 as early
as the beginning of the fifth century. The Cham-Further and the Stall-
wanger Senke, to the north of these central forts of the Late Antique
Rhaetian Danube limes, were the natural connections between south-
western Bohemia and the area just north of the Alps.25 These Germanic
groups often continued to settle around these fortresses, even after
the dismantling of the border-armies and the end of Roman supre-

21
B. Svoboda, “Zum Verhältnis frühgeschichtlicher Funde des 4. und 5. Jahrhun-
derts aus Bayern und Böhmen”, Bayerische Vorgeschichtsblätter 28 (1963) pp. 97–116, esp.
p. 114; Fischer, Das bajuwarische Reihengräberfeld von Staubing, pp. 108–11; id., “Zur
Archäologie des fünften Jahrhunderts in Ostbayern”, Typen der Ethnogenese unter beson-
derer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denkschriften der
Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse 204.
Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien 1990)
pp. 101–22, esp. pp. 103–11. The Friedenhain-PÏre“tovice group in its totality clearly
belongs to the fifth century, see ibid., p. 109; related finds have meanwhile come
to light in many castella of the Danube limes, and Fischer, ibid., p. 111, supposes
that “the Germanic foederati from Bohemia constituted the majority of the Roman
troops on this part of the border during the fifth century”. Similarly id., Das baju-
warische Reihengräberfeld von Staubing, pp. 96–8; 101; 108–11; id. and H. Geisler, “Her-
kunft und Stammesbildung der Baiern aus archäologischer Sicht”, Die Bajuwaren.
Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-
Salzburg 1988) pp. 61–8, here pp. 66–7; id., Das Umland des römischen Regensburg,
Münchener Beiträge zur Vor- und Frühgeschichte 42 (München 1990) pp. 33–4;
Schwarz, “Das Ende der Völkerwanderungszeit”, p. 52, failed when he wanted to
link the group Friedenhain-PÏre“tovice to the Naristi.
22
R. Eckes and H. Zeiß, “Bairische Reihengräber des 6. Jahrhunderts bei Irlmauth,
BA. Regensburg”, Bayerische Vorgeschichtsblätter 15 (1938) pp. 44–56, esp. pp. 51–4;
U. Koch, Die Grabfunde der Merowingerzeit aus dem Donautal um Regensburg, Germanische
Denkmäler der Völkerwanderungszeit A,10 (Berlin 1968) pp. 122–4; 134.
23
Koch, Die Grabfunde der Merowingerzeit, pp. 122–3; Geisler, “Neue archäologi-
sche Quellen”, pp. 89–100; Fischer, Das bajuwarische Reihengräberfeld von Staubing, pp.
122–7; Fischer and Geisler, “Herkunft und Stammesbildung der Baiern”, pp. 63–6.
24
Heinz Löwe, “Die Herkunft der Bajuwaren”, p. 31, expected these archaeo-
logically visible immigrants to be Marcomanni, who in his view crossed the Danube
in the early sixth century. See also Kraus, “Die Herkunft der Bayern”, p. 30;
Reindel, “Die Bajuwaren”, pp. 462–4; Th. Fischer, “Römer und Germanen an der
Donau”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H.
Dopsch (Rosenheim-Salzburg 1988) pp. 39–45, esp. pp. 41–3; id., Das bajuwarische
Reihengräberfeld von Staubing, pp. 108–11.
25
Fischer, “Zur Archäologie des fünften Jahrhunderts”, pp. 110–1; id., “Der
Übergang von der Spätantike zum frühen Mittelalter in Ostbayern”, Regensburg—
Kelheim—Straubing 1, ed. S. Rieckhoff-Pauli and W. Torbrügge, Führer zu archäo-
logischen Denkmälern 5 (Stuttgart 1984) pp. 236–43, esp. pp. 238–41; id., Das
bajuwarische Reihengräberfeld von Staubing, p. 110.
   435

macy.26 Around the year 536, Suebians, who may have arrived from
the Pannonian Danube-region, were able to attack Venice from what
would later become Bavarian land.27
It is highly probable that all these people, who over a long period
of time had migrated into the area between the Danube and the
Alps, would have been described as men from Baia or *Baiahaim,
*Baiawarioz.28 They further integrated with East-29 and West-Germanic,30

26
Fischer, Das bajuwarische Reihengräberfeld von Staubing, pp. 99–100.
27
Cassiodorus, Variae 12,7,1, ed. T. Mommsen, MGH AA 12 (Berlin 1894)
p. 366: Sueborum incursione vastatis. See also Löwe, “Die Herkunft der Bajuwaren”,
p. 42. E. Schwarz, “Herkunft und Einwanderungszeit”, pp. 33–5; 44 also thinks in
terms of Danube-Suebian groups, that had, to his mind, arrived in the early sixth
century from Bojer-Einöde (which was then allegedly called by this name, though
there are no written sources to support this view), and entered Rhaetia and Noricum.
This may have been of great importance for the formation of the Bavarian peo-
ple. See also Schwarz, “Das Ende der Völkerwanderungszeit”, pp. 39–43.
28
Schnetz, “‘Baias’”, p. 9; Schwarz, “Ermunduren”, p. 346; id., “Das Ende der
Völkerwanderungszeit”, pp. 37–8; Wenskus, Stammesbildung, pp. 564–6; Beck, Hamann
and Roth, “Bajuwaren”, pp. 601–6 and esp. pp. 601–2; Fischer, “Zur Archäologie
des fünften Jahrhunderts”, p. 118; Wolfram, “Ethnogenesen im frühmittelalterlichen
Donau- und Ostalpenraum”, pp. 107–8. For a criticism of the attempt to deduce
the Bavarian name from the Salzburg Romanitas see H. Rosenfeld, “Die Völkernamen
Baiern und Böhmen, die althochdeutsche Lautverschiebung und W. Mayerthalers
These ‘Baiern = Salzburger Rätoromanen’. Völkernamen, Völkerwanderung, Stam-
mesgenese und die Namen Baiern, Bayern, Bajuwaren”, Althochdeutsch, vol. 2: Wörter
und Namen. Forschungsgeschichte, ed. R. Bergmann, H. Tiefenbach and L. Voetz, Ger-
manische Bibliothek NF, Reihe 3, Untersuchungen (Heidelberg 1987) pp. 1305–32,
esp. pp. 1321–6; P. Wiesinger, “Antik-romanische Kontinuitäten im Donauraum
von Ober- und Niederösterreich am Beispiel der Gewässer-, Berg- und Siedlungs-
namen”, Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern 1, ed. W. Pohl
and H. Wolfram, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 201. Veröffentlichungen der Kommission für
Frühmittelalterforschung 12 (Wien 1990) pp. 261–328, esp. pp. 262–4.
29
Martin, “Die Gräberfelder von Straubing-Bajuwarenstraße und Straßkirchen”,
p. 26 emphasises the fact that burials of Ostrogoth women are organised in groups
within the cemetery of Altenerding. Moreover, E. Schwarz “Herkunft und Einwander-
ungszeit”, pp. 38–9, also refers to groups of nomadic horsemen, who may have
been involved in the formation of a people. His racial argumentation, which he
still expressed on this subject in 1953 shows the continuity in research with which
he was closely connected.
30
On Sciri participation in the tribal formation, see Schwarz, “Die bairische
Landnahme um Regensburg”, pp. 69–71; id., “Herkunft und Einwanderungszeit”,
pp. 36–41, who ibid., p. 31, labels the Sciri West-Germanic from a linguistic point
of view. Later, however, in his “Das Ende der Völkerwanderungszeit”, p. 45, he
stresses the long-term East-Germanic connections of the Sciri. M. Martin, “Die
Gräberfelder von Straubing-Bajuwarenstraße und Straßkirchen”, p. 30, points to the
presence of the grave of a West-Germanic woman within the cemetery of Straß-
kirchen.
436  

and particularly Alemannic31 groups, but primarily with the provin-

cial Roman population32 and with the remaining Breonic popula-
tion33 to become the Bavarians,34 whom Jordanes mentions for the
first time in the year 551.35 They were, presumably not mentioned
in Jordanes’ main source, Cassiodorus’ History of the Goths, since the
latter in his Variae only seems to have been informed about the
inhabitants of the provinces of Rhaetia and Noricum, but never used
the name of the Bavarians.36 Their settlement taking place as a group

31
Koch, Die Grabfunde der Merowingerzeit, pp. 121; 134. M. Martin, “Die Gräberfelder
von Straubing-Bajuwarenstraße und Straßkirchen”, pp. 22–4 shows the increase in
Alemannic immigration in the region around Straubing after the new Alemannic
defeats against Clovis. Also J. Jahn, Ducatus Baiuvariorum. Das bairische Herzogtum der
Agilolfinger, Monographien zur Geschichte des Mittelalters 35 (Stuttgart 1991) pp.
5–6 stresses the significance of Alemannic participation in the process of Bavarian
ethnogenesis. See further D. Geuenich and H. Keller, “Alamannen, Alamannien,
Alamannisch im frühen Mittelalter. Möglichkeiten und Schwierigkeiten des Historikers
beim Versuch der Eingrenzung”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram
and A. Schwarcz, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 179. Veröffentlichungen der Kommission für Frühmit-
telalterforschung 8 (Wien 1989) pp. 135–57, esp. pp. 146–7.
32
K. Bosl, Bayerische Geschichte (7th edn., München 1990), pp. 38; 65 gives a great
deal of significance to the “Celto-Romans” as “original substratum”, also in the
development of the name of the Bavarians. See also Schmid, “Bayern und Italien”,
pp. 54–5; H. Dopsch, “Zum Anteil der Romanen und ihrer Kultur an der Stammes-
bildung der Bajuwaren”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. id. and
H. Dannheimer (Rosenheim-Salzburg 1988) pp. 47–54.
33
Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum”,
pp. 118–24; id., Salzburg, Bayern, Österreich, pp. 32–6; K.A. Eckhardt, Merowingerblut
II. Agilolfinger und Etichonen, Deutschrechtliches Archiv 11 (Witzenhausen 1965) pp.
103–5 points out the significance of Herulic provenance of the Breonic King Sinduald.
34
K. Reindel, “Herkunft und Stammesbildung der Bajuwaren nach den schriftli-
chen Quellen”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and
H. Dopsch (Rosenheim-Salzburg 1988) pp. 56–60; Wolfram, “Ethnogenesen im
frühmittelalterlichen Donau- und Ostalpenraum”, p. 108; P. Wiesinger, “Gotische
Lehnwörter im Bairischen. Ein Beitrag zur sprachlichen Frühgeschichte des Bairischen”,
Frühmittelalterliche Ethnogenese im Alpenraum, ed. H. Beumann and W. Schröder, Nationes
5 (Sigmaringen 1985) pp. 153–200, esp. pp. 185–6; Bosl, Bayerische Geschichte, pp.
34–7; Kraus, “Die Herkunft der Bayern”, p. 41; Wiesinger, “Antik-romanische
Kontinuitäten”, p. 301; E. Zöllner, “Das Geschlecht der Agilolfinger”, Die Anfänge
des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberösterreichischen Landes-
archivs, Ergänzungsband 2 (Linz 1978) pp. 83–110, esp. pp. 97–8; Jahn, Ducatus
Baiuvariorum, p. 6.
35
Jordanes, Getica 55,280, ed T. Mommsen, MGH AA 5,1 (Berlin 1882) p. 130:
nam regio illa Suavorum ab oriente Baibaros habet, ab occidente Francos, a meridie Burgundzones,
a septentrione Thuringos. See Kraus, “Die Herkunft der Bayern”, p. 42.
36
See also E. Schwarz, “Baiern und Naristen in Burgund”, Südostdeutsche Forschungen
2 (1937) pp. 379–82, esp. p. 382; id., “Die bairische Landnahme um Regensburg”,
p. 52; id., “Herkunft und Einwanderungszeit”, p. 28; Reindel, “Die Bajuwaren”,
   437

closed to outside influence, as assumed by many supporters of the

migration-thesis,37 is therefore out of the question.38
In the following, the question that must be asked relates to the
motivating force behind this process of ethnogenesis in Roman ter-
ritory. Usually, the Ostrogoth realm under Theoderic the Great is
attributed with playing this role.39 Yet as early as in the year 537

pp. 452–3; Jarnut, Agilolfingerstudien, pp. 45–6. The supposition of a copy from the
lost Gothic history of Cassiodorus is, however, impossible to prove with any cer-
tainty, as shown by Schmidt, “Zum Ursprung der Baiern”, p. 13; Lotter, “Die ger-
manischen Stammesverbände”, pp. 54–5; H. Wolfram, “Die Christianisierung der
Baiern”, Baiernzeit in Oberösterreich. Das Land zwischen Inn und Enns vom Ausgang der
Antike bis zum Ende des 8. Jahrhunderts, Kataloge des Oberösterreichischen Landesmuseums
97 (Linz 1977) pp. 177–88, esp. p. 178, and Rosenfeld, “Die Völkernamen Baiern
und Böhmen”, pp. 1307–8, which are based on the introduction to the Getica by
T. Mommsen, p. XXXIII. Reference to the Bavarians in the so-called “table of
nations” does not seem to predate the middle of the sixth century, though W. Gof-
fart, “The supposedly ‘Frankish’ table of nations: an edition and study”, Frühmittelalter-
liche Studien 17 (1983) pp. 98–130, esp. pp. 115–28, proposes a compilation of this
passage in the time around 520.
37
Zeiß, “Von den Anfängen des Baiernstammes”, pp. 27; 29; 39. Also, E. Schwarz,
“Das Ende der Völkerwanderungszeit”, pp. 47–8; 54, maintains, as does Kraus,
“Die Herkunft der Bayern”, pp. 33–41, their immigration as a tribe, but does not
wish to exclude a merging with other tribal remains (ibid., p. 50). Zeiß, “Von den
Anfängen des Baiernstammes”, pp. 23–6 supposes the formation of the tribal name
already in the homeland. Also, H. Mitscha-Märheim, “Die Herkunft der Baiern”,
Mitteilungen der Anthropologischen Gesellschaft in Wien 80 (1950) pp. 213–44, esp. pp.
233–4, and E. Schwarz, “Herkunft und Einwanderungszeit”, p. 32, do not exclude
the possibility of immigration and tribal formation shortly after the year 488. Mitscha-
Märheim envisages the immigration of various groups, including Sciri, from Pannonia,
led by Odoacer’s brother Hunwulf.
38
Böhme, “Zur Bedeutung des spätrömischen Militärdienstes”, p. 34; Fischer and
Geisler, “Herkunft und Stammesbildung der Baiern”, p. 63; Wenskus, Stammesbildung,
pp. 561–2; F. Prinz, “Fragen der Kontinuität zwischen Antike und Mittelalter am
Beispiel Bayerns”, Zeitschrift für bayerische Landesgeschichte 37 (1974) pp. 699–727, esp.
pp. 707–8.
39
Wolfram, 378–907: Grenzen und Räume, p. 64, thinks that the Ostrogothic realm
of Theoderic the Great constituted the context for the Bavarian tribal formation:
“hence, Theoderic succeeded [. . .] in allowing the new people of the Bavarians to
be created between Iller and Enns”. See also id., Salzburg, Bayern, Österreich, p. 22;
Lotter, “Die germanischen Stammesverbände”, pp. 56–7; H.D. Kahl, “Die Baiern
und ihre Nachbarn bis zum Tode des Herzogs Theodo (717/18)”, Die Bayern und
ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz, Denkschriften der Österreichi-
schen Akademie der Wissenschaften, philosophisch-historische Klasse 179. Veröffent-
lichungen der Kommission für Frühmittelalterforschung 8 (Wien 1989) pp. 159–225,
esp. p. 164; 217; 223, as well as Wolfram, “Ethnogenesen im frühmittelalterlichen
Donau- und Ostalpenraum”, pp. 105–8; id., “Baiern und das Frankenreich”, Die
Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988) p. 130; Wolfram, “Christianisierung”, pp. 178–80; Th.
Fischer, Die Römer in Deutschland (Stuttgart 1999) pp. 174–5; Dietz, Osterhaus,
Rieckhoff-Pauli and Spindler, Regensburg zur Römerzeit, pp. 155–67; Reindel, “Die
438  

the Ostrogoths ceded their supreme rule over the territories north
of the Alps to the kings of the Franks.40 Around 540, Frankish bish-
ops ordained priests in Aguntum, Virunum and Teurnia.41 King
Theudebert I was probably the one who added the regions between
the Danube and the Alps to Frankish rule, and who supported the

Bajuwaren”, pp. 472–3. G. Hauptfeld, “Die Gentes im Vorfeld”, p. 127 thinks there
were Thuringian incentives for the formation of the Bavarian tribe.
40
H. Zeiß, “Bemerkungen zur frühmittelalterlichen Geschichte Baierns I”, Zeitschrift
für bayerische Landesgeschichte 2 (1929) pp. 343–60, esp. pp. 343–54; R. Schneider,
“Fränkische Alpenpolitik”, Die transalpinen Verbindungen der Bayern, Alemannen und Franken
bis zum 10. Jahrhundert, ed. H. Beumann and W. Schröder, Nationes 6 (Sigmaringen
1987) pp. 23–49, esp. pp. 27–8.
41
Concilium universale Constantinopolitanum sub Justiniano habitum 2, ed. E. Schwartz,
Acta conciliorum oecumenicorum 4,2 (Straßburg 1914) p. 135: quod ante annos iam
fieri coeperat, et in tribus ecclesiis nostri concilii, id est Breconensi, Tiburniensi, et Augustana
Galliarum episcopi constituerant sacerdotes; et nisi eiusdem tunc divae memoriae Justiniani prin-
cipis iussione commotio partium nostrarum remota fuisset [. . .]; Gregorii I Papae Registrum
Epistolarum 16a, ed. P. Ewald, MGH EE 1 (Berlin 1887) p. 20, spells Beconensi. See
H. Berg, “Bischöfe und Bischofssitze im Ostalpen- und Donauraum vom 4. bis zum
8. Jahrhundert”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 179. Veröffentlichungen der Kommission für Frühmittelalterfor-
schung 8 (Wien 1989) pp. 61–108, esp. pp. 65–6 (Teurnia); pp. 82–4; see also
Wolfram, 378–907: Grenzen und Räume, p. 98; H. Wolff, “Die Kontinuität städti-
schen Lebens in den nördlichen Grenzprovinzen des römischen Reiches und das
Ende der Antike”, Die Stadt in Oberitalien und in den nordwestlichen Provinzen des Römischen
Reiches. Deutsch-Italienisches Kolloquium im italienischen Kulturinstitut Köln, ed. W. Eck and
H. Galsterer, Kölner Forschungen 4 (Mainz 1991) pp. 287–318, esp. pp. 295–6
with the supposition of a continuity of ruins in Virunum and a continuity of insti-
tutions in Aguntum. On greater continuity of Teurnia, see ibid., p. 301; H. Wolff,
“Die Kontinuität der Kirchenorganisation in Raetien und Noricum bis an die
Schwelle des 7. Jahrhunderts”, Das Christentum im bairischen Raum von den Anfängen bis
ins 11. Jahrhundert, ed. id. and E. Boshof, Passauer historische Forschungen 8 (Köln-
Wien 1994) pp. 1–17, esp. pp. 8–16, therefore rather proposes Augsburg, Säben,
a Breonic diocese in the Inntal, and Teurnia St. Peter im Holz. As Breonic dio-
ceses, Pfaffenhofen in the Inntal and Martinsbühel near Zirl are likely candidates,
though it is by no means certain that there were ever bishops working in that area.
See O. Hageneder, “Die kirchliche Organisation im Zentralalpenraum vom 6. bis
10. Jahrhundert”, Frühmittelalterliche Ethnogenese im Alpenraum, ed. H. Beumann and
W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 201–35, esp. pp. 207–9; on
Aguntum-Lavant ibid., pp. 210–2; on Teurnia ibid., p. 212; on Virunum ibid., pp.
212–3; see also ibid., pp. 216–21 on the letter of the bishops of Rhaetia and Venice
to the Byzantine Emperor, and on the localisation of the three episcopal sees in
Virunum, Teurnia und Aguntum; S. Ladstätter, Die materielle Kultur der Spätantike in
den Ostalpen. Eine Fallstudie am Beispiel der westlichen Doppelkirchenanlage auf dem Hemmaberg,
Mitteilungen der Prähistorischen Kommission der Österreichischen Akademie der
Wissenschaften 35,1 (Wien 2000) pp. 37–9. For Teurnia, see also V. Bierbrauer,
“Die germanische Aufsiedlung des östlichen und mittleren Alpengebietes im 6. und
7. Jahrhundert aus archäologischer Sicht”, Frühmittelalterliche Ethnogenese im Alpenraum,
   439

Suebian campaign against Venice.42 The defeat of the Thuringian

realm in the year 534 may have also made the inhabitants of the
later Bavarian Nordgau open to new political direction.43
Among the political steps taken towards the organization of their
dominion, the Merovingian kings installed a duchy in Regensburg.44
With the house of the Agilolfings, they established a family capable
of deploying the power necessary to bring about the integration of
the region northeast of the Alps. The centre of power of the Agilolfing
family was probably initially situated in the southwestern Visigothic
border-region and later in the Austrasian-Burgundian part of the
Frankish realm,45 though Thuringian,46 Alemannic47 and Herulic-
Lombard48 origins have also been suggested for them.49 In any case,

ed. H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 9–47, esp.
pp. 33–5.
42
Schmidt, Westgermanen, p. 201. F. Beyerle, “Süddeutschland in der politischen
Konzeption Theoderichs des Großen”, Grundfragen der alemannischen Geschichte, Vorträge
und Forschungen 1 (Sigmaringen 1955) pp. 65–81, esp. p. 80 argues, that it was
Theudebert I who “granted the Bavarians places to live when they were moving
from Pannonia via Bohemia and Moravia, after which they subjected to his ruler-
ship”. See also Werner, “Die Herkunft der Bajuwaren”, pp. 30–5; Kahl, “Die Baiern
und ihre Nachbarn”, pp. 176–7; Jarnut, Agilolfingerstudien, pp. 47–9.
43
E. Schwarz, “Herkunft und Einwanderungszeit”, pp. 41–3, who, at p. 43, writes
about an Anschluß of the Oberpfalz as the Nordgau to the Bavarians after the defeat
of the Thuringian realm.
44
H. Wolfram, “Baiern und das Frankenreich”, Die Bajuwaren. Von Severin bis
Tassilo 488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988)
pp. 130–5, esp. p. 130; Jarnut, Agilolfingerstudien, pp. 49–53; Jahn, Ducatus Baiuvariorum,
p. 9. G. Hauptfeld, “Die Gentes im Vorfeld”, p. 131 has expressed the idea that
the kings of the Franks might have entered into amicitia with the Bavarians.
45
K.F. Werner, “Bedeutende Adelsfamilien im Reich Karls des Großen”, Karl
der Große. Lebenswerk und Nachleben, vol. 1: Persönlichkeit und Geschichte, ed. H. Beumann
(Düsseldorf 1965) pp. 83–142, esp. pp. 106–15; E. Hlawitschka, “Studien zur
Genealogie und Geschichte der Merowinger und frühen Karolinger”, Rheinische Viertel-
jahrsblätter 43 (1979) pp. 1–99, esp. pp. 83–8; A. Friese, Studien zur Herrschaftsgeschichte des
fränkischen Adels. Der mainländisch-thüringische Raum vom 7. bis 11. Jahrhundert, Geschichte
und Gesellschaft. Bochumer Historische Studien 18 (Stuttgart 1979) pp. 163–7;
Jarnut, Agilolfingerstudien, pp. 12–4; 28–40; 57–68.
46
Hauptfeld, “Die Gentes im Vorfeld”, pp. 128–9; N. Wagner, “Zur Herkunft
der Agilolfinger”, Zeitschrift für bayerische Landesgeschichte 41 (1978) pp. 19–48, esp. pp.
34; 42–7. Edictus Rothari, ed. F. Bluhme, MGH LL 4 (Hannover 1868) pp. 1–90,
esp. Prol., p. 2, claims Thuringian origins for the Lombard King Agilulf: Agilulf
Turingus ex genere Anawas.
47
Hartung, Süddeutschland, p. 195.
48
Eckhardt, Merowingerblut pp. 93–105; W. Goez, “Über die Anfänge der Agilul-
finger”, Jahrbuch für fränkische Landesforschung 34/35 (1975) pp. 145–61, esp. pp. 152–
61; Zöllner, “Das Geschlecht der Agilolfinger”, pp. 87–8; 95–7; 99. But see also
Hlawitschka, “Studien zur Genealogie und Geschichte”, pp. 81–95.
49
See Kahl, “Die Baiern und ihre Nachbarn”, p. 175.
440  

earlier Burgundian50 influence may also be clearly demonstrated. This

is evident from the fact that it was always Frankish-Burgundian or
Alemannic Agilolfings who were appointed dukes of the Bavarians
when the local male line of succession was interrupted.51
Around the year 555, despite clerical protest, Chlothar I offered
the Lething Lombard princess Walderada, the widow of his deceased
great-nephew Theudebald and for a short time his own wife too, in
marriage to Garibald, uni ex suis52 and duke of the Bavarians. The
Bavarians had just been identified for the first time as such in the
written sources.53 Since Gregory of Tours does not mention them
at all, nor does he link them to the dux Garivaldus,54 the Bavarians
would have only visibly started to develop an identity as a people
after the appointment of the Agilolfings. The fact that they are not
mentioned in the well known letter from Theudebert I to Justinian
I may also point in this direction.55 It may therefore be assumed that
an integrative power was created only with the appointment by the
Frankish king of a dux, who actively endeavoured to organize the
various Germanic and Roman groups and units living in the former
province of Rhaetia.56 This effort was initiated at the former centre
of Regensburg, with all its administrative-military importance, and
the groups that were present in the area around this former legionary
camp must have played a central, and an increasingly important,
role in the process of ethnogenesis. These people came mainly from

50
E. Zöllner, “Die Herkunft der Agilulfinger”, Mitteilungen des Instituts für Öster-
reichische Geschichtsforschung 59 (1951) pp. 245–64 [repr. Zur Geschichte der Bayern, ed.
K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 107–34]; id., “Das Geschlecht
der Agilolfinger”, pp. 91–4; H. Wolfram, “Christianisierung”, p. 181 interprets the
Agilolfings as “the descendants of the Burgundian royal house that was dethroned
by the Franks”.
51
Hartung, Süddeutschland, p. 97.
52
Paul the Deacon, Historia Langobardorum 1,21, ed. L. Bethmann and G. Waitz,
MGH SSrL (Hannover 1878) p. 60.
53
Hlawitschka, “Studien zur Genealogie und Geschichte”, pp. 84–8; Schmid,
“Bayern und Italien”, pp. 58–9; Wolfram, “Baiern und das Frankenreich”, p. 130;
Hauptfeld, “Die Gentes im Vorfeld”, pp. 127–8; 131; Kahl, “Die Baiern und ihre
Nachbarn”, pp. 170–5; Jarnut, Agilolfingerstudien, pp. 50–4; Wolfram, 378–907: Grenzen
und Räume, p. 76; Jahn, Ducatus Baiuvariorum, pp. 10–1.
54
Gregory of Tours, Historiae 4,9, ed. B. Krusch and W. Levison, MGH SSrM
1,1 (2nd edn., Hannover 1951) p. 141 only gives details on the dux Garivaldus.
55
Epistolae Austrasicae 20, ed. W. Gundlach, MGH EE 3 (Berlin 1892) pp. 132–3.
56
Bosl, Bayerische Geschichte, pp. 48–9. J. Jarnut, Agilolfingerstudien, p. 47 states that
“this phase of Bavarian ethnogenesis took place under the political-military control
of the Austrasian Merovingians, Theudebert and Theudebald”.
   441

Bohemia and the northeastern Danube headland and were called

Baiuvarii—men from the land of Baiahaim.57 Jörg Jarnut has shown
that it was quite possibly a deliberate measure of the Frankish King
Theudebald (whose mother was Roman) to place Garibald as an
integrative figure at the head of the developing gens in Rhaetia, since
Garibald originated from the old noble family of the Agilolfings with
its links to the Visigoths, Suebians, Franks, Burgundians and Romance
speakers.58 The common Suebian background of all the groups
involved in the ethnogenesis of the Bavarians, which in older research
was seen as the only binding element in the shaping of the Bavarian
tribe under Agilolfing leadership, should today at least be critically
discussed.59
Independently of the question of the name of the Baiuvarii and
whether it refers to a Germanic core-group from Bohemia or indeed
from another land of Baia northeast of the Danube, the ethnogen-
esis of the Bavarians took place in the territories they later inhab-
ited. The Agilolfing duces soon adopted a king-like position,60 and
were repeatedly called reges by Paul the Deacon;61 they must have
had a decisive influence on the development of Bavarian identity
and of their own consciousness as a people.62

57
Reindel, “Die Bajuwaren”, pp. 471–3. W. Hartung, Süddeutschland, pp. 168–72,
stresses, probably too one-sidedly, the interpretation of the name in the sense of
groups living “near Bohemia” of mainly Alemannic-Juthungian origins, though this
interpretation of the name cannot be completely excluded.
58
J. Jarnut, Agilolfingerstudien, p. 56 states that “in the formation of rulership over
the developing gens Baiuvariorum, King Theudebald, around the middle of the sixth
century, created a key position for the Agilolfings, who were closely related to him
and had their origins in the Visigothic region”. See also Zöllner, “Das Geschlecht
der Agilolfinger”, p. 98.
59
Jarnut, Agilolfingerstudien, p. 51.
60
On the intitulatio of the Agilolfing dukes and its example, see H. Wolfram,
Lateinische Königs- und Fürstentitel bis zum Ende des 8. Jahrhunderts, Intitulatio 1 (Graz
1967) pp. 161–84.
61
Hauptfeld, “Die Gentes im Vorfeld”, p. 133; Wolfram, 378–907: Grenzen und
Räume, p. 94; Jahn, Ducatus Baiuvariorum, pp. 11–2; id., “Hausmeier und Herzöge.
Bemerkungen zur agilolfingisch-karolingischen Rivalität bis zum Tode Karl Martells”,
Karl Martell in seiner Zeit, ed. J. Jarnut, U. Nonn and M. Richter, Beihefte der Francia
37 (Sigmaringen 1994) pp. 317–44, esp. p. 319.
62
According to J. Jahn, Ducatus Baiuvariorum, p. 561 “the Frankish appointment
of the first Bavarian dux coincided with the end of Bavarian ethnogenesis. The polit-
ical ‘voice’ of the Bavarian gens was only acquired under the aegis of the Agilolfings.”
A. Kraus, “Die Herkunft der Bayern”, pp. 45–6, however, thinks it is almost cer-
tain that their ethnogenesis had already been completed before their settlement.
442  

2. What sorts of changes and conditions lead to

(or represent the development towards) the establishment of a Germanic kingdom?

The possibly royal,63 but certainly old noble, background of the dukes
appointed by the Frankish kings, as well as their rule over a transit
area infrastructurally and strategically important for several border-
regions and boundary-areas of the Frankish realm, furthered the
self-consciousness of the Regensburg Agilolfings. Marriage ties with
princes and princesses from Lombard royal families and the im-
plicitly sought backing of the ruling family of the Merovingians,64 as
well as later support from their southern Lombard and western
Alemannic neighbours, and after that also from the Carolingian
upstarts in the Frankish lands, all contributed to the tendency towards
the independent royal power of the Agilolfings over the Bavarians.
Although the Frankish kings repeatedly undertook military action
against this independence (as happened already in 589–591), they
nonetheless always reinstated members of the Agilolfing family in the
position of Bavarian dukes. A good example is Tassilo I, in the year
592, who most likely was a son or close relative of Garibald I.65 The
success of his almost royal rule depended in part upon the military
importance of Tassilo I’s and his duchy’s struggle against the Carinthian
Slavs near the Drau and, more importantly, against the Avars aid-
ing them.66 While a real success was only achieved in the year 592
(595 witnessed a defeat and around 610, under Garibald II, there
was a victory only after a defeat),67 wars against the eastern neigh-
bours provided the opportunity to consolidate the Bavarians’ own

63
J. Jarnut, Agilolfingerstudien, pp. 36–40 thinks that Agiulf, who was appointed
governor of the defeated Suebian realm in northwestern Spain, and king of the
Suebians shortly afterwards, was the founding father of the Agilolfings.
64
Ibid., pp. 28–32; 56–63; 87–9.
65
Wolfram, 378–907: Grenzen und Räume, p. 78.
66
On the “eastern politics” of the Agilolfings, see W. Störmer, “Die Agilolfinger
im politischen Kräftefeld vom 6. bis zum 8. Jahrhundert”, Baiernzeit in Oberösterreich.
Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts,
Kataloge des Oberösterreichischen Landesmuseums 97 (Linz 1977) pp. 1–12, esp.
pp. 2–4. On the relations between the Slavs from the Alps and the Bavarians, see
Kahl, “Die Baiern und ihre Nachbarn”, pp. 194–200, on relations between Bavarians
and Avars, see ibid., pp. 201–12; Wolfram, Salzburg, Bayern, Österreich, pp. 39–46.
67
Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum”,
pp. 126–8; id., “Baiern und das Frankenreich”, pp. 131–2; id., 378–907: Grenzen
und Räume, pp. 78–9; Ladstätter, Die materielle Kultur, p. 39; Jahn, Ducatus Baiuvariorum,
pp. 17–8.
   443

position. Around 740, the Carinthian ruler Boruth was made a

Bavarian dependent after Avar attacks.68 The expansion into the east-
ern regions settled by Slavs, and the renewed victory over the Carin-
thians in 772, enabled Tassilo III to counter the increasing influence
over the western parts of the duchy exercised by Frankish nobles
with close links to the Carolingians.69

3. What is the role of a gentile identity (Stammesbewußtsein)

for the establishment of a regnum?

An Origo gentis Baiuvariorum has not survived.70 From the twelfth

century onwards, multiple traditions emerged regarding the repatri-
ation of the Bavarians, and associated with them was the belief that
the Bavarians had an Armenian origin.71 These traditions must be
treated as contemporary learned constructions that helped raise
the historic status of the Bavarians.72 Research conducted by Erich

68
Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum”,
p. 137; Störmer, “Die Agilolfinger im politischen Kräftefeld”, p. 3.
69
F. Prinz, “Zur Herrschaftsstruktur Bayerns und Alemanniens im 8. Jahrhundert”,
Blätter für deutsche Landesgeschichte 102 (1966) pp. 11–27, esp. p. 19; id., “Nochmals
zur ‘Zweiteilung des Herzogtums der Agilolfinger’”, Blätter für deutsche Landesgeschichte
113 (1977) pp. 19–32, esp. pp. 30–2; H. Wolfram, “Das Fürstentum Tassilos III.,
Herzogs der Bayern”, Mitteilungen der Gesellschaft für Salzburger Landeskunde 108 (1968)
pp. 157–79, esp. p. 165; id., 378–907: Grenzen und Räume, p. 89; Jahn, Ducatus Baiuva-
riorum, pp. 471–3.
70
Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum”,
p. 151.
71
Das Annolied, ed. M. Roediger, MGH Deutsche Chroniken 1,2 (Berlin 1895)
pp. 63–132, esp. pp. 121–2; Das Annolied 307–317, ed. E. Nellmann (4th edn.,
Stuttgart 1996); Vita Altmanni episcopi Pataviensis 28, ed. W. Wattenbach, MGH SS
12 (Berlin 1856) p. 237: Bawari traduntur ab Armenia oriundi; Die Kaiserchronik eines
Regensburger Geistlichen 317–324, ed. E. Schröder, MGH Deutsche Chroniken 1,1
(Hannover 1892) pp. 85–6. See also E. Mayer, “Übersehene Quellen zur bay-
erischen Geschichte des 6.–8. Jahrhunderts”, Zeitschrift für bayerische Landesgeschichte 4
(1931) pp. 1–36, esp. pp. 1–13.
72
H. Zeiß, “Bemerkungen zur frühmittelalterlichen Geschichte Baierns II”, Zeitschrift
für bayerische Landesgeschichte 4 (1931) pp. 351–66, esp. pp. 351–61; Mitscha-Märheim,
“Die Herkunft der Baiern”, pp. 214; 236; 238; Klebel, “Langobarden, Bajuwaren,
Slawen”, pp. 41–5; Reindel, “Das Zeitalter der Agilulfinger”, p. 102; G.R. Spohn,
“Armenien und Herzog Naimes. Zur bayerischen Stammessage im Mittelalter und
bei Peter Harer”, Zeitschrift für bayerische Landesgeschichte 34 (1971) pp. 185–210, esp.
pp. 190–2; 194–6; Beck, Hamann and Roth, “Bajuwaren”, pp. 605–6; M. Müller,
“Die bayerische ‘Stammessage’ in der Geschichtsschreibung des Mittelalters. Eine
Untersuchung zur mittelalterlichen Frühgeschichtsforschung in Bayern”, Zeitschrift für
444  

Zöllner,73 Reinhard Wenskus,74 and Wilhelm Störmer75 has shown

that the Nibelung heroic poetry was very popular among the Bavarians.
Whether this account gained importance in the context of Bavarian
ethnogenesis cannot, however, be stated with certainty. Additionally,
as the Burgundian connections of the Agilolfings were not the only
roots of the family, it cannot be firmly argued that this account of
the destruction of the Burgundians was brought to the Danube as
an established tradition by them. Notwithstanding these negative
results, there must have been an awareness of community, which,
among other things, over the centuries led to the adoption and suc-
cess of the Old High German language.76 Whether this awareness
lent support to the dukedom of the Agilolfings must remain open,
in view of the lack of an early medieval historiography focusing on
the Bavarians.

bayerische Landesgeschichte 40 (1977) pp. 341–71, esp. pp. 342–56; 367–9; Reindel,
“Die Bajuwaren”, pp. 457–60; L. Kolmer, “Die Inschriften aus dem Grab des
Bischofs Gregorius und die Herkunft der Baiern aus Armenien”, Ostbairische Grenzmarken
28 (1986) pp. 11–21, esp. pp. 15–7; W. Störmer, “Beobachtungen zu Aussagen und
Intentionen der bayerischen Stammes-‘Sage’ des 11./12. Jahrhunderts. Fiktionen—
Sage—‘Geschichtsklitterung’”, Fälschungen im Mittelalter. Internationaler Kongreß der Monu-
menta Germaniae Historica, vol. 1: Kongreßdaten und Festvorträge. Literatur und Fälschung,
Schriften der MGH 33,1 (Stuttgart 1988) pp. 451–70, esp. pp. 465–70.
73
Zöllner, “Die Herkunft der Agilulfinger”, pp. 113–4; id., “Das Geschlecht der
Agilolfinger”, pp. 92–4.
74
R. Wenskus, “Wie die Nibelungen-Überlieferung nach Bayern kam”, Zeitschrift
für bayerische Landesgeschichte 36 (1973) pp. 393–449, esp. pp. 431–43; id., Sächsischer
Stammesadel und fränkischer Reichsadel, Abhandlungen der Akademie der Wissenschaften
in Göttingen, philosophisch-historische Klasse 3,93 (Göttingen 1976) pp. 475; 513–24.
75
W. Störmer, “Nibelungentradition als Hausüberlieferung in frühmittelalterlichen
Adelsfamilien? Beobachtungen zu Nibelungennamen im 8./9. Jahrhundert vornehm-
lich in Bayern”, Nibelungenlied und Klage. Sage und Geschichte, Struktur und Gattung. Passauer
Nibelungen-Gespräche 1985, ed. F.P. Knapp (Heidelberg 1987) pp. 1–20.
76
On Old High German and the Bavarians, see I. Reiffenstein, “Stammesbild-
ung und Sprachgeschichte. Das Beispiel der bairischen Ethnogenese”, Althochdeutsch,
vol. 2: Wörter und Namen. Forschungsgeschichte, ed. R. Bergmann, H. Tiefenbach and
L. Voetz, Germanische Bibliothek NF, Reihe 3, Untersuchungen (Heidelberg 1987)
pp. 1333–41, esp. pp. 1338–9. K. Bosl, Bayerische Geschichte, p. 38, also stresses the
role of the Franks, who may have contributed to the survival of Old High German.
W. Hartung, Süddeutschland, p. 175, however, stresses the influence of the Alemans,
whom he considers the stronger element in the population.
   445

4. What sorts of changes in the “constitution” (Verfassung) of a people and

a kingdom (such as central organs of power, local power structures, or links
between the two) are linked to the establishment of a kingdom? How do
socio-economic developments contribute to this process?

It is possible that an early law-code by the Frankish King, Theudebald,

for the emerging Bavarians was connected to the appointment of
the Agilolfings as dukes in Regensburg; at least, that is, if it really
is he who was intended in the reference to Theuderich mentioned
in the preface of the Lex Baiuvariorum,77 as Jörg Jarnut suggests.78 In
the earliest codification of this law, there was extensive borrowing
from Visigothic law-forms,79 which may be explained by the Romanic-
Aquitanian contacts of both Theudebald and the Agilolfings.80 The
later Frankish kings, Childebert II (575–595), Clothar II (584–629)
and Dagobert I, also acted as law-givers.81 The Frankish influ-
ence evident here was brought by the Agilolfing dukes, whose role
was rooted in this system of law, and they thereby united the devel-
oping Bavarian gens under one law.82 The ruling circles of several
individual groups and noble families that may already have partici-
pated in Bavarian ethnogenesis: the genealogiae of the Huosi, Fagana,
Anniona, Drozza and Hachilinga—families with possessions mainly
in western Bavaria—were especially privileged in the Lex Baiuvariorum
with a special double wergeld.83 But the law-code also integrated

77
On the problem of the date of the Lex Baiuvariorum see H. Siems, “Lex
Baiuvariorum”, Handwörterbuch zur deutschen Rechtsgeschichte 2 (1978) pp. 1887–1901,
esp. pp. 1890–1; W. Hartmann, “Das Recht”, Die Bajuwaren. Von Severin bis Tassilo
488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 266–72,
esp. p. 266; W. Störmer, “Zum Prozeß sozialer Differenzierung bei den Bayern von
der Lex Baiuvariorum bis zur Synode von Dingolfing”, Typen der Ethnogenese unter
besonderer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
201. Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien
1990) pp. 155–70, esp. pp. 157–60.
78
Jarnut, Agilolfingerstudien, pp. 54–6.
79
Siems, “Lex Baiuvariorum”, pp. 1892–3.
80
Jarnut, Agilolfingerstudien, pp. 55–6.
81
Reindel, “Das Zeitalter der Agilolfinger”, p. 244; Gastroph, Herrschaft und
Gesellschaft, pp. 58–9. Limiting the legislative activities of Dagobert I: Kahl, “Die
Baiern und ihre Nachbarn”, p. 188.
82
Gastroph, Herrschaft und Gesellschaft, p. 19.
83
Lex Baiwariorum 3,1, ed. E. von Schwind, MGH LL nationum Germanicarum
5,2 (Hannover 1926) pp. 312–3: De genealogia, qui vocantur Hosi Drazza Fagana Hahilinga
Anniona: isti sunt quasi primi post Agilolfingos, qui sunt de genere ducali. Illis enim duplum
honorem concedamus et sic duplam conpositionem accipiant.
446  

them.84 The law set out the Agilolfings, who had the privilege of a
fourfold wergeld, as hereditary leaders of the Bavarians.85 Their rule
was based in Regensburg. Under Duke Theodo, a division of the
duchy or the creation of sub-duchies or sub-realms for his sons in
Passau, Salzburg86 and Freising87 is evident. His own overlordship
based in Regensburg clearly remained unaltered.88 Yet from the time
of Odilo, Salzburg possibly represented the most important centre
of the Bavarian duchy.89

84
Störmer, “Zum Prozeß sozialer Differenzierung”, pp. 164–7; id., “Nibelungentra-
dition”, pp. 9–10. H.L.G. Gastroph, Herrschaft und Gesellschaft, pp. 105–12; 136–8,
despite being mistaken when stating that, from the point of view of wergeld,
Agilolfings and genealogies may be characterised as nobility in the service of the
Frankish kings (Dienstadel ). Similarly mistaken is also Bosl, Bayerische Geschichte, p. 39.
See further Zöllner, “Das Geschlecht der Agilolfinger”, p. 97; Rosenfeld, “Die
Völkernamen Baiern und Böhmen”, pp. 1329–30. See on the interpretation of the
names of the genealogiae and on their localisation H. Krahwinkler, “Beiträge zu
Namen und Geschichte der bayerischen Genealogiae”, Typen der Ethnogenese unter
besonderer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
201. Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien
1990) pp. 217–34; further H. Dachs, “Germanischer Uradel im frühbairischen
Donaugau”, Verhandlungen des Historischen Vereins für Oberpfalz und Regensburg 86 (1936)
pp. 179–92 [repr. Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60
(Darmstadt 1965) pp. 85–106, esp. pp. 102–4]; Klebel, “Langobarden, Bajuwaren
und Slawen”, pp. 64–6; id., “Bayern und der fränkische Adel im 8. und 9. Jahr-
hundert”, Grundfragen der alemannischen Geschichte, Vorträge und Forschungen 1 (Sig-
maringen 1955) pp. 193–208, esp. pp. 194–200; F. Prinz, “Herzog und Adel im
agilulfingischen Bayern. Herzogsgut und Konsensschenkungen vor 788”, Zeitschrift
für bayerische Landesgeschichte 25 (1962) pp. 283–311 [repr. Zur Geschichte der Bayern, ed.
K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 225–63, esp. pp. 250–2];
id., “Herrschaftsstruktur”, p. 20; Hartung, Süddeutschland, pp. 177–9; 185–6; 189–200;
W. Störmer and G. Mayr, “Herzog und Adel”, Die Bajuwaren. Von Severin bis Tassilo
488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 153–9,
esp. pp. 155–7; Kahl, “Die Baiern und ihre Nachbarn”, pp. 168–70; Jarnut, Agilol-
fingerstudien, pp. 110–6; W. Volkert, “Die Ortsnamen des Hachinger Tales”, Herrschaft,
Kirche, Kultur. Beiträge zur Geschichte des Mittelalters. Festschrift für F. Prinz zum 65. Geburtstag,
ed. G. Jenal, Monographien zur Geschichte des Mittelalters 37 (Stuttgart 1993) pp.
43–60, esp. pp. 57–60. On Huosi and Fagana see Wenskus, Sächsischer Stammesadel
und fränkischer Reichsadel, pp. 470–5; 518–23; Jahn, Ducatus Baiuvariorum, pp. 232–8.
85
Lex Baiuvariorum 3,1, p. 313: Agilolvinga vero usque ad ducem in quadruplum conpo-
nantur, quia summi principes sunt inter vos. See further Kahl, “Die Baiern und ihre Nach-
barn”, pp. 172–3; Jarnut, Agilolfingerstudien, p. 83; Jahn, Ducatus Baiuvariorum, pp. 1–3;
232–3.
86
K. Reindel, “Salzburg und die Agilolfinger”, Virgil von Salzburg. Missionar und
Gelehrter, ed. H. Dopsch and R. Juffinger (Salzburg 1985) pp. 66–74, esp. pp. 67–70.
87
Jahn, Ducatus Baiuvariorum, pp. 107–9.
88
Ibid., pp. 98–9.
89
Reindel, “Salzburg und die Agilolfinger”, pp. 70–1.
   447

The realization of Agilolfing rulership took place partly through

their entourage living in the vicinity of the ducal residence.90 At the
start of the eighth century, Arbeo’s Life of Emmeram mentions satrapes
terrae and a cohors, with whose members the duke conferred.91 In the
Life of Corbinian, there is reference to Duke Grimoald’s primates, who
were staying at the palace.92 On his journey to Rome in 715/16,
Theodo “was accompanied by other members of his people”.93 The
foundation of the monastery of Innichen was confirmed by Tassilo
III and by his optimates, followers who had apparently accompanied
him on a journey to the Lombard King, Desiderius.94 Although this
group of people is often viewed as proto-nobility,95 it may also have
been a circle of ducal followers. Counts should be regarded as addi-
tional aides of the duke. In Bavaria and its border-regions, there is
evidence for them as early as c. 680, though no concomitant ruler-
ship covering the whole of Bavaria existed before 788.96 As office-
holders, they were given ducal goods that with increasing frequency

90
K. Bosl, “Das ‘jüngere’ bayerische Stammesherzogtum der Luitpoldinger”, Zeit-
schrift für bayerische Landesgeschichte 18 (1955) pp. 144–72 [repr. Zur Geschichte der Bayern,
ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 329–63, esp. pp. 336–7];
Störmer and Mayr, “Herzog und Adel”, p. 154; Störmer, “Zum Prozeß sozialer
Differenzierung”, pp. 160–1; Jahn, Ducatus Baiuvariorum, pp. 223–6.
91
Arbeo, Vita Haimhramni 10; 32; 34, ed. B. Krusch, MGH SSrG 13 (Hannover
1920) pp. 41; 75; 76. See also Jahn, Ducatus Baiuvariorum, p 48.
92
Vita Corbiniani 26, ed. H. Glaser, F. Brunhölzl and S. Benker (München-Zürich
1983) p. 136. See also Jahn, Ducatus Baiuvariorum, p. 108.
93
Liber pontificalis, ed. L. Duchesne, Bibliothèque des Ecoles Françaises d’Athènes
et de Rome, 2e sér. (Paris 1886; repr. 1955) vol. 1, p. 398: Theodo, dux gentis
Baioariorum cum alios gentis suae ad apostoli beati Petri limina orationis voto primus de gente
eadem occurrit. See also Wolfram, 378–907: Grenzen und Räume, pp. 109–10.
94
E. Zöllner, “Der bairische Adel und die Gründung von Innichen”, Mitteilungen
des Instituts für Österreichische Geschichtsforschung 68 (1960) pp. 362–87 [repr. Zur Geschichte
der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 135–71, esp.
pp. 138–9; 145–65].
95
J. Jahn, “Bayerische “Pfalzgrafen” im 8. Jahrhundert? Studien zu den Anfängen
Herzog Tassilos (III.) und zur Praxis der fränkischen Regentschaft im agilolfingi-
schen Bayern”, Früh- und hochmittelalterlicher Adel in Schwaben und Bayern, ed. I. Eberl,
W. Hartung and J. Jahn, Regio 1 (Sigmaringendorf 1988) pp. 80–114, esp. pp. 101;
113. J. Jahn, Ducatus Baiuvariorum, pp. 251–3, supposes that a privileged Bavarian
nobility did not develop until the time of Duke Odilo.
96
Paul the Deacon, Historia Langobardorum 5,36, p. 156: cum comite Baioariorum, quem
illi gravionem dicunt, qui Bauzanum et reliqua castella regebat. W. Störmer, Früher Adel.
Studien zur politischen Führungsschicht im fränkisch-deutschen Reich vom 8. bis 11. Jahrhundert,
vol. 2, Monographien zur Geschichte des Mittelalters 6 (Stuttgart 1973) pp. 392–4;
S. Haider, “Oberösterreich im bairischen Stammesherzogtum”, Baiernzeit in Oberösterreich.
Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts,
Kataloge des Oberösterreichischen Landesmuseums 97 (Linz 1977) pp. 13–26, esp.
448  

later became hereditary, as also happened in the Frankish kingdom.97

The autonomous rights of rulership, which the counts gained from
their private possessions, meant that they became increasingly inde-
pendent at times when the duchy underwent a crisis. The southern
borders of the duchy were guarded by prepositi custodes98 and a[u]ctores
montani,99 who maintained contact with the ducal palaces. From the
time of Odilo onwards, the Bavarian dukes seem to have deliber-
ately developed a system of office-holders, whom they juridically
advanced and with whose support they could sustain their power
against the genealogiae and the developing nobility.100
The role of Christianisation in the development of the Bavarians,
and in this process of integration of the various groups from which
they were composed, can only be properly estimated with diffi-
culty. The Roman population had been Christian since Late Anti-
quity and there seems to be a marked continuity of individual
cults of saints,101 like that of St Florian and the forty martyrs of
Lauriacum,102 of St Maximilian in Bischofshofen103 or St Afra in

pp. 16–7; Störmer and Mayr, “Herzog und Adel”, pp. 154–5; Jahn, Ducatus
Baiuvariorum, pp. 259–61; Wolfram, Salzburg, Bayern, Österreich, pp. 156–8; 166.
97
Jahn, Ducatus Baiuvariorum, pp. 240–4; Wolfram, Salzburg, Bayern, Österreich,
p. 158.
98
Vita Corbiniani 23, p. 128. See Jahn, Ducatus Baiuvariorum, pp. 103–4; 245.
99
Vita Corbiniani 15, p. 110. See Jahn, Ducatus Baiuvariorum, p. 100.
100
Jahn, Ducatus Baiuvariorum, pp. 254–6.
101
V. Miloj‘iÆ, “Zur Frage der Zeitstellung des Oratoriums von Mühlthal an der
Isar”, Bayerische Vorgeschichtsblätter 28 (1963) pp. 117–38, esp. pp. 135–6.
102
K. Reindel, “Die Bistumsorganisation im Alpen-Donau-Raum in der Spätan-
tike und im Frühmittelalter”, Mitteilungen des Instituts für Österreichische Geschichtsforschung
72 (1964) pp. 277–310, esp. pp. 283–4; L. Eckhart, “Das Nach- und Weiterleben der
Römerzeit in Oberösterreich”, Baiernzeit in Oberösterreich. Das Land zwischen Inn und
Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts, Kataloge des Oberösterreichi-
schen Landesmuseums 97 (Linz 1977) pp. 27–38, esp. pp. 32–3; B. Ulm, “Patrozinien
in Spätantike und Agilolfingerzeit”, Baiernzeit in Oberösterreich. Das Land zwischen Inn
und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts, Kataloge des Oberöster-
reichischen Landesmuseums 97 (Linz 1977) pp. 189–212, esp. pp. 193–5; Wolff,
“Die Kontinuität städtischen Lebens”, p. 305. For sceptical remarks regarding con-
tinuity of the grave of the saint see id., “Die Anfänge des Christentums in Ostraetien,
Ufernoricum und Nordwestpannonien: Bemerkungen zum Regenwunder und zum
hl. Florian”, Ostbairische Grenzmarken 31 (1989) pp. 27–45, esp. pp. 35–7.
103
Hageneder, “Die kirchliche Organisation im Zentralalpenraum”, pp. 228–9;
Wolfram, 378–907: Grenzen und Räume, p. 108; H. Koller, “Die bairische Kirchen-
organisation des 8. Jahrhunderts: Ansätze, Konzepte, Verwirklichung”, Das Christentum
im bairischen Raum von den Anfängen bis ins 11. Jahrhundert, ed. E. Boshof and H. Wolff,
Passauer historische Forschungen 8 (Köln-Wien 1994) pp. 273–89, esp. pp. 284–5;
F.R. Erkens, “Die Ursprünge der Lorcher Tradition im Lichte archäologischer, histo-
   449

Augsburg.104 There is likely also an early date for the many church
dedications to St Laurentius, at least some of which had origins in
the Roman period.105 Among the Germanic groups, the number of
cults influenced by Arianism is thought to have been small,106 just
as it is possible to point to people who had hardly had contact with
Christianity. Possibly around the year 615, Eustasius of Luxeuil, likely
accompanied by the Burgundian Agilolfing, Agilus,107 went to the
Bavarians.108 He had previously worked as a missionary among the
Warasks, who lived in the vicinity of Luxeuil and of whom it was
still said in the eighth century that they once came from a district

riographischer und urkundlicher Zeugnisse”, ibid., pp. 423–59, esp. pp. 431; 440;
H. Berg, “Christentum im bayerischen Raum um 700”, Der heilige Willibald—Kloster-
bischof oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter, Eichstätter
Studien NF 30 (Regensburg 1990) pp. 69–113, esp. pp. 93–4; Jahn, Ducatus Baiuva-
riorum, pp. 64–8; 86.
104
Venantius Fortunatus, Vita Sancti Martini 4,640–643, ed. F. Leo, MGH AA
4,1 (Berlin 1881) p. 368. F. Prinz, “Augsburg im Frankenreich”, Die Ausgrabungen
in St. Ulrich und Afra in Augsburg 1961–1968, ed. J. Werner, Münchener Beiträge
zur Vor- und Frühgeschichte 23 (München 1977) pp. 375–98, esp. pp. 389–90;
V. Bierbrauer, “Alamannische Besiedlung Augsburgs und seines näheren Umlandes”,
Geschichte der Stadt Augsburg von der Römerzeit bis zur Gegenwart, ed. G. Gottlieb, W. Baer,
J. Becker, J. Bellot, K. Filser, P. Fried, W. Reinhard and B. Schimmelpfennig
(Stuttgart 1984) pp. 87–100, esp. pp. 88–91; W. Sage, “Frühes Christentum und
Kirchen aus der Zeit des Überganges”, ibid., pp. 100–12, esp. pp. 102–5; Wolff,
“Die Kontinuität städtischen Lebens”, p. 301; Wolfram, 378–907: Grenzen und Räume,
p. 103.
105
G. Diepolder, “Altbayerische Laurentiuspatrozinien”, Aus Bayerns Frühzeit. F. Wag-
ner zum 75. Geburtstag, ed. J. Werner, Schriftenreihe zur bayerischen Landesgeschichte
62 (München 1962) pp. 371–96, esp. pp. 394–6.
106
Wolfram, 378–907: Grenzen und Räume, pp. 96–7; Wiesinger, “Gotische Lehnwörter
im Bairischen”, esp. pp. 188–92; K. Schäferdiek, “Gab es eine gotisch-arianische
Mission im süddeutschen Raum?”, Zeitschrift für bayerische Landesgeschichte 45 (1982)
pp. 239–57, esp. pp. 247–50, did not find real traces of Arian mission in the
Bavarian vocabulary. He considers the arrival of Arian groups to have been rather
insignificant.
107
Zöllner, “Das Geschlecht der Agilolfinger”, p. 91.
108
Jonas of Bobbio, Vita Columbani 2,8, pp. 243–4; and, based on this, the Vita
Sadalbergae 7, ed. B. Krusch, MGH SSrM 5 (Hannover-Leipzig 1910) p. 54. See
further Wolfram, “Christianisierung”, pp. 182–3; Kahl, “Die Baiern und ihre Nach-
barn”, p. 190; Wiesinger, “Gotische Lehnwörter im Bairischen”, pp. 188–9; G. Mayr,
“Frühes Christentum in Baiern”, Die Bajuwaren. Von Severin bis Tassilo 488–788,
ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 281–6, esp.
pp. 282–3; Wolfram, 378–907: Grenzen und Räume, p. 104; Berg, “Christentum”,
pp. 73–4.
450  

called Stadevanga near the rivers Naab and Regen109 (where they
were possibly known as Naristi or Varisti as early as the time of Taci-
tus).110 They had been banished from there to Burgundia. Little is
known of other missionary activities, although they clearly did take
place. An example is the failure-ridden journey of Agrestius before
the year 626, which moreover had Iro-Frankish involvements.111
It is not until the time of Duke Theodo112 that stronger ducal
influence on the ad hoc arrangement of the church-organisation113

109
Egilbert, Vita Ermenfredi 1, ed. J. Carnandet, Acta Sanctorum Ordinis sancti
Benedicti (Paris-Rom 1867) p. 107: Temporibus igitur Clotarii regis Francorum a beatus
vir Eustasius, Luxoviensis monasterii abbas, jubente S. Columbano antecessore suo, progrediens
Warescos ad fidem Domini nostri Jesu Christi convertit; qui olim de pago, ut ferunt, qui dicitur
Stadevanga, qui situs est circa Regnum flumen, partibus orienti fuerant ejecti, quique contra
Burgundiones pugnam inierunt, sed a primo certamine terga vertentes, dehinc advenerunt, atque in
pugnam reversi victores quoque effecti in eodem pago Warescorum consederunt. Rejecting a late
eighth-century date for this source: Zeiß, “Bemerkungen I”, pp. 354–356; though,
see also Löwe, “Die Herkunft der Bajuwaren”, p. 27; Schwarz, “Baiern und Naristen
in Burgund”, pp. 380–2; id., “Die bairische Landnahme um Regensburg”, pp. 53–8;
id., “Herkunft und Einwanderungszeit”, pp. 42–4; id., “Das Ende der Völkerwan-
derungszeit”, pp. 50–3; id., “Die Naristenfrage in namenkundlicher Sicht”, Zeitschrift
für bayerische Landesgeschichte 32 (1969) pp. 397–476, esp. pp. 399–405; 454–5; 468–76;
id., “Neues und Altes zur Geschichte der Naristen”, Jahrbuch für fränkische Landesforschung
22 (1962) pp. 281–9, esp. pp. 286–9; Zöllner, “Die Herkunft der Agilulfinger”, pp.
119–21.
110
Tacitus, Germania 42,1, ed. G. Perl, Griechische und lateinische Quellen zur Frühgeschichte
Mitteleuropas bis zur Mitte des 1. Jahrtausends unserer Zeit, vol. 2, ed. J. Herrmann,
Schriften und Quellen der Alten Welt 37,2 (Berlin 1990) pp. 118–9: Iuxta Hermunduros
Naristi ac deinde Marcomani et Quadi agunt. According to H. Bengtson, “Neues zur
Geschichte der Naristen”, Historia 8 (1959) pp. 213–21, esp. pp. 220–1, the Narists,
who lived on the northern Danube in Upper Austria during the middle Empire,
have no connection with the Burgundian Warasks. Neither were they involved in
Bavarian ethnogenesis.
111
Jonas of Bobbio, Vita Columbani 2,9, pp. 246–7. K. Bosl, “Der ‘Adelsheilige’,
Idealtypus und Wirklichkeit, Gesellschaft und Kultur im merowingerzeitlichen Bayern
des 7. und 8. Jahrhunderts”, Speculum Historiale. Geschichte im Spiegel von Geschichtsschrei-
bung und Geschichtsdeutung, ed. C. Bauer, L. Boehm and M. Müller (Freiburg-München
1965) pp. 167–87, esp. pp. 169–70; F. Prinz, Frühes Mönchtum im Frankenreich. Kultur
und Gesellschaft in Gallien, den Rheinlanden und Bayern am Beispiel der monastischen Entwicklung
(4. bis 8. Jahrhundert) (2nd edn., München 1988), pp. 356–8; id., “Augsburg im
Frankenreich”, pp. 380; 386–7; Wolfram, “Christianisierung”, pp. 182–3; id., 378–907:
Grenzen und Räume, pp. 104–5; id., Salzburg, Bayern, Österreich, p. 43; Mayr, “Frühes
Christentum in Baiern”, pp. 283–4; Berg, “Christentum”, p. 76.
112
E. Klebel, “Zur Geschichte des Herzogs Theodo”, Verhandlungen des Historischen
Vereins für die Oberpfalz und Regensburg 99 (1958) pp. 165–205 [repr. Zur Geschichte der
Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 172–224].
113
Koller, “Die bairische Kirchenorganisation”, pp. 279–87; Erkens, “Die Ursprünge
der Lorcher Tradition”, pp. 448–50; Berg, “Christentum”, pp. 69–78; Jahn, Ducatus
Baiuvariorum, pp. 157–9.
   451

becomes evident. After the death of the mayor of the palace, Pippin
II, in 714, Theodo intended to further the development of the church-
organisation in his lands and to install a metropolitan see in
Regensburg; the residences of his sons at Freising, Salzburg, and
Passau were to become episcopal sees.114 With the objective of
more intensive Christian penetration of the lands over which he
ruled,115 Theodo had earlier called Erhard,116 who already worked
in Regensburg, into the region and later he sought out Rupert,117
who was a relative of his wife, Folchaid, and bishop of Worms.
He also contacted Emmeram of Poitiers118 as well as Corbinian of

114
Hageneder, “Die kirchliche Organisation im Zentralalpenraum”, pp. 222–4;
Reindel, “Bistumsorganisation”, pp. 306–7; id., “Salzburg und die Agilolfinger”, pp.
66–7; H. Schmidinger, “Das Papsttum und die bayerische Kirche—Bonifatius als
Gegenspieler Virgils”, Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and
R. Juffinger (Salzburg 1985) pp. 92–101, esp. pp. 92–3; Reindel, “Das Zeitalter der
Agilolfinger”, pp. 226–7; H. Berg, “Zur Organisation der bayerischen Kirche und
zu den bayerischen Synoden des 8. Jahrhunderts”, Typen der Ethnogenese unter beson-
derer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der
Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse 201.
Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien 1990)
pp. 181–97, esp. p. 181; Wolfram, 378–907: Grenzen und Räume, pp. 82–3; 109–10;
E. Boshof, “Agilolfingisches Herzogtum und angelsächsische Mission: Bonifatius und
die bayerische Bistumsorganisation von 739”, Ostbairische Grenzmarken 31 (1989) pp.
11–26, esp. pp. 15–6; R. Kaiser, “Bistumsgründung und Kirchenorganisation im 8.
Jahrhundert”, Der heilige Willibald—Klosterbischof oder Bistumsgründer, ed. H. Dickerhof,
E. Reiter and S. Weinfurter, Eichstätter Studien NF 30 (Regensburg 1990) pp.
29–67, esp. pp. 53–6; Jahn, Ducatus Baiuvariorum, pp. 73–5; 98–9; 120–1.
115
Mayr, “Frühes Christentum in Baiern”, pp. 284–6; Berg, “Christentum”, pp.
83–112; W. Störmer, “Die bayerische Herzogskirche”, Der heilige Willibald—Klosterbischof
oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter, Eichstätter Studien
NF 30 (Regensburg 1990) pp. 115–42, esp. pp. 116–21; Jahn, Ducatus Baiuvariorum,
pp. 31–73.
116
Mayr, “Frühes Christentum in Baiern”, p. 284.
117
H. Wolfram, “Der heilige Rupert und die antikarolingische Adelsopposition”, Mit-
teilungen des Instituts für Österreichische Geschichtsforschung 80 (1972) pp. 4–34, esp. pp. 29–
30; id., “Der Heilige Rupert in Salzburg”, Frühes Mönchtum in Salzburg, ed. E. Zwink,
Salzburg Diskussionen 4 (Salzburg 1983) pp. 81–92, esp. p. 88; id., “Die Zeit der
Agilolfinger. Rupert und Virgil”, Geschichte Salzburgs, vol. 1,1: Vorgeschichte, Altertum,
Mittelalter, ed. H. Dopsch (Salzburg 1981) pp. 121–56, esp. pp. 125–34; Reindel,
“Salzburg und die Agilolfinger”, pp. 68–9; Wolfram, 378–907: Grenzen und Räume,
pp. 105–9; Berg, “Christentum”, pp. 83–96; Jahn, Ducatus Baiuvariorum, pp. 48–69.
118
Klebel, “Zur Geschichte des Herzogs Theodo”, pp. 177–86; Wolfram, “Ethno-
genesen im frühmittelalterlichen Donau- und Ostalpenraum”, pp. 131–2; id., 378–907:
Grenzen und Räume, p. 109; Boshof, “Agilolfingisches Herzogtum”, pp. 13–4; G. Mayr,
“Zur Todeszeit des Hl. Emmeram und zur frühen Geschichte des Klosters Herren-
chiemsee. Bemerkungen zur Schenkung des Ortlaip in Helfendorf ”, Zeitschrift für
bayerische Landesgeschichte 34 (1971) pp. 358–73; Berg, “Christentum”, pp. 96–106;
Jahn, Ducatus Baiuvariorum, pp. 40–7.
452  

Melun.119 These plans for establishing a church-organisation initially

failed, in part because of the quarrels among Theodo’s sons and
grandchildren immediately after his death. Not entirely without
reason, they were afraid of growing Frankish influence on the restruc-
turing of Church matters, and of Frankish military campaigns under-
taken in reaction to the Agilolfing disputes.120 It was not until 739
that Duke Odilo,121 born into the family of Alemannic Agilolfings
and appointed by Charles Martel, could finish the organisation of
the church with the support of Boniface.122 Even so, he was not able
to elevate Regensburg fully to the status of an archbishopric, nor
was he immediately able to delimit clearly the various bishoprics.123
The newly created church-organisation was rounded off by the found-
ing of the see of Neuburg on the Danube at some uncertain date
between 741 and 760,124 as well as the founding of the bishopric of

119
Berg, “Christentum”, pp. 106–12; Störmer, “Die bayerische Herzogskirche”,
pp. 121–3; Jahn, Ducatus Baiuvariorum, pp. 69–73.
120
G. Mayr, “Neuerliche Anmerkungen zur Todeszeit des Heiligen Emmeram
und zur Kirchenpolitik Herzog Theodos”, Typen der Ethnogenese unter besonderer Berücksichti-
gung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 201. Veröffentlichun-
gen der Kommission für Frühmittelalterforschung 12 (Wien 1990) pp. 199–215,
esp. pp. 214–5; Störmer, “Die bayerische Herzogskirche”, pp. 121–3; Jahn, Ducatus
Baiuvariorum, pp. 75; 101.
121
Zöllner, “Die Herkunft der Agilulfinger”, pp. 126–33; J. Jarnut, “Studien über
Herzog Odilo (736–748)”, Mitteilungen des Instituts für Österreichische Geschichtsforschung
85 (1977) pp. 273–84, esp. pp. 278–81; Wolfram, “Baiern und das Frankenreich”,
pp. 134–5; Wolfram, 378–907: Grenzen und Räume, p. 84; Jahn, Ducatus Baiuvariorum,
pp. 123–32.
122
H. Löwe, “Bonifatius und die bayerisch-fränkische Spannung”, Jahrbuch für fränki-
sche Landesforschung 15 (1955) pp. 85–127 [repr. Zur Geschichte der Bayern, ed. K. Bosl,
Wege der Forschung 60 (Darmstadt 1965) pp. 264–328, esp. pp. 280–6]; Reindel,
“Bistumsorganisation”, pp. 307–9; Wolfram, 378–907: Grenzen und Räume, pp. 110–1;
Kaiser, “Bistumsgründung”, pp. 60–2; Störmer, “Die bayerische Herzogskirche”,
pp. 124–6; Jahn, Ducatus Baiuvariorum, pp. 139–41.
123
Schmidinger, “Das Papsttum und die bayerische Kirche”, pp. 93–5; Störmer,
“Die bayerische Herzogskirche”, pp. 126–7. J. Jahn, “Bayerische ‘Pfalzgrafen’”, pp.
86–7 and 98–9, denies that Odilo wanted to create a separate church-province.
The new dioceses from this point of view were not supported by Odilo, who rather
focused on ducal monasteries as elements of Christian penetration of the land. See
also Jahn, Ducatus Baiuvariorum, pp. 152–3; 169–71; 372; id., “Hausmeier und
Herzöge”, pp. 337–8. H. Wanderwitz, “Der Libellus Virgilii und das Verhältnis
von Herzögen und Bischöfen in Bayern”, Virgil von Salzburg. Missionar und Gelehrter,
ed. H. Dopsch and R. Juffinger (Salzburg 1985) pp. 357–61, esp. pp. 359–60, has
expressed similar opinions.
124
P. Fried, “Bischof Simpert und das Bistum Neuburg-Staffelsee”, Jahrbuch des
   453

Eichstätt between 741 and the middle of the century, which hap-
pened as a result of the handing over of the expanded Bavarian
Nordgau to the Frankish mayors of the palace.125 Cooperation with
Frankish institutions in the organisation of the Church may possibly
have been the reason for Odilo’s brief expulsion, and his subsequent
exile in 740/41 to the court of the mayors of the palace. During
this period, Odilo’s marriage with Hiltrud, daughter of Charles Martel
and sister of Pippin III and Carloman, may have commenced.126
This wedding provided further justification for the future antagonis-
tic attitude of the Agilolfings towards the Carolingians.127 However,
it appears that Odilo was reserved in his opinion of the newly founded
bishoprics. In order to intensify his power, he, as later his son Tassilo
III, concentrated on founding monasteries and cells, either doing so
himself or through cooperation with Bavarian nobles.128

5. What is the role of kings in this development?

Through skillful politics, the Agilolfing dukes managed first of all to

liberate themselves from the role of the dukes appointed by the
Merovingian king, and then to attain for many years a position equal

Vereins für Augsburger Bistumsgeschichte 12 (1978) pp. 181–5, esp. pp. 182–5. But see
also Jahn, Ducatus Baiuvariorum, pp. 404–7.
125
S. Weinfurter, “Das Bistum Willibalds im Dienste des Königs. Eichstätt im
frühen Mittelalter”, Zeitschrift für bayerische Landesgeschichte 50 (1987) pp. 3–40, esp.
pp. 14–22; A. Kraus, “Der heilige Willibald von Eichstätt: Person, Zeit, Werk”, Der
heilige Willibald—Klosterbischof oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and
S. Weinfurter, Eichstätter Studien NF 30 (Regensburg 1990) pp. 9–28, esp. pp.
20–7; H. Dickerhof and S. Weinfurter, “Summa historica”, ibid., pp. 245–61, esp.
253–61; Jahn, Ducatus Baiuvariorum, pp. 159–63.
126
Contra the supposition of this expulsion, which is only described in the Breves
Notitiae 7, ed. W. Hauthaler, Salzburger Urkundenbuch, vol. 1: Traditionskodizes (Salzburg
1910) p. 27, see Zeiß, “Bemerkungen I”, pp. 356–8. But see also Zöllner, “Die
Herkunft der Agilulfinger”, pp. 131–2; Jarnut, “Studien über Herzog Odilo”, pp.
281–4; and Wolfram, 378–907: Grenzen und Räume, p. 84. Further, see Jahn, Ducatus
Baiuvariorum, pp. 172–8; 248–9; id., “Hausmeier und Herzöge”, pp. 338–40.
127
Differently J. Jahn, Ducatus Baiuvariorum, p. 190, who considers the Frankish-
Bavarian war of 743 as the “definitive recognition of Odilos dukedom”. See also,
on the ever changing and by no means always negative relations between Agilolfings
and Carolingians, id., “Hausmeier und Herzöge”, pp. 317–8; 330–43.
128
Id., Ducatus Baiuvariorum, pp. 192–220.
454  

to that of the kings.129 To achieve this, they took advantage of peri-

ods of weakness suffered by the Frankish kings, the increasingly open
conflict with the ambitious mayors of the palace, and alliances estab-
lished with neighbouring Germanic groups.130 In particular, the mar-
riages made with members of the Merovingian family, the Lombard
royal house and Lombard dukes, as well as with the Alemannic ducal
house and sometimes with Carolingian mayors of the palace and
their female relatives, raised the status of the Agilolfings. For nearly
a century they were related to the kings of the Lombards, and for
a time to the dukes of the Alemans, but above all the family was
very influential in the Austrasian-Burgundian area, in Alsace, and in
the region of the middle Rhine and Thuringia.131
However, the ambition of becoming equal in status to royalty, and
the increasing self-consciousness of the Agilolfings, attracted the con-
stant threat of military reprisals by the Frankish kings. In the end,
the trial of Tassilo III and his subsequent monastic imprisonment
set in motion the downfall of the Agilolfings and the end of the old
Bavarian duchy. Despite the Lex Baiuvariorum, which was influenced
by the Merovingian kings to a considerable extent, and the eleva-
tion of the Agilolfings to the position of dukes, the Merovingians
and the Frankish mayors of the palace retained their control over
appointments to the duchy. Garibald I and Tassilo I (circa 591) seem
certainly to have been appointed by Frankish kings. Odilo, of the
family of Alemannic dukes,132 was appointed duke by Charles Martel.

129
Jarnut, Agilolfingerstudien, pp. 91–2.
130
Jahn, Ducatus Baiuvariorum, pp. 178–85.
131
Werner, “Bedeutende Adelsfamilien”, pp. 108–15; J. Jarnut, “Beiträge zu den
fränkisch-bayerisch-langobardischen Beziehungen im 7. und 8. Jahrhundert”, Zeitschrift
für bayerische Landesgeschichte 39 (1976) pp. 331–52, esp. pp. 341–2; Friese, Studien zur
Herrschaftsgeschichte des fränkischen Adels, pp. 163–7; Zöllner, “Das Geschlecht der
Agilolfinger”, pp. 87–8; W. Störmer, “Bayerisch-ostfränkische Beziehungen vom 7.
bis zum frühen 9. Jahrhundert”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and
A. Schwarcz, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 179. Veröffentlichungen der Kommission für Früh-
mittelalterforschung 8 (Wien 1989) pp. 227–52, esp. pp. 248–52; Jarnut, Agilolfinger-
studien, pp. 41–3; 57–78; 86–9; W. Störmer, “Das Herzogsgeschlecht der Agilolfinger”,
Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988) pp. 141–52, esp. pp. 142–8.
132
Jarnut, “Studien über Herzog Odilo”, pp. 273–8; Kahl, “Die Baiern und ihre
Nachbarn”, pp. 219–20; J. Jarnut, “Genealogie und politische Bedeutung der agilol-
fingischen Herzöge”, Mitteilungen des Instituts für Österreichische Geschichtsforschung 90
(1991) pp. 1–22, esp. pp. 2–5; 12–5.
   455

Tassilo III, too, owed his rule to the mayor of the palace, Pippin III.
The family relationship between Odilo and Tassilo III, and Pippin
III and Charlemagne, respectively, did not improve the situation for
the Agilolfings. In fact, it contributed to a hopeless situation lead-
ing to a conflict between Tassilo III and Pippin III, which in the
end made Charlemagne decide to depose the duke.133 It was not
only with the Aquitanian campaign in the year 763, during which
the breach134 between Tassilo III and the Frankish king became
evident, that the former began to behave in a royal manner. The
signs of this royal interpretation of rulership included: his marriage to
Liutpirc, daughter of the Lombard King, Desiderius;135 the amicitia
with Charlemagne established in 771/72;136 his alliance with his east-
ern Avar neighbours;137 his role at the synods of Aschheim, Dingolfing
and Neuching as lord over a Bavarian Church—or rather, ducal
Church138—as well as his relations with the papacy;139 his disposal
of a treasure, which, significantly, was taken away from him in 788;140

133
On the relationship between Tassilo III, Pippin III and Charlemagne see
M. Becher, Eid und Herrschaft. Untersuchungen zum Herrscherethos Karls des Großen (Sigma-
ringen 1993) pp. 21–74. See also H. Wolfram, “Tassilo III. und Karl der Große—
Das Ende der Agilolfinger”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dann-
heimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 160–6.
134
Becher, Eid und Herrschaft, pp. 45–51; Jahn, Ducatus Baiuvariorum, pp. 371–5.
135
Wolfram, “Das Fürstentum Tassilos III.”, p. 164; Schmid, “Bayern und Italien”,
pp. 71–2; 74.
136
Jahn, Ducatus Baiuvariorum, pp. 467–8; Jarnut, “Genealogie und politische Bedeu-
tung”, pp. 18–9.
137
Wolfram, “Das Fürstentum Tassilos III.”, pp. 171–3; id., 378–907: Grenzen und
Räume, pp. 91–2; Kahl, “Die Baiern und ihre Nachbarn”, p. 211; P. Classen, “Bayern
und die politischen Mächte im Zeitalter Karls des Großen”, Die Anfänge des Klosters
Kremsmünster, ed. S. Haider, Mitteilungen des Oberösterreichischen Landesarchivs,
Ergänzungsband 2 (Linz 1978) pp. 169–87, esp. p. 181; L. Kolmer, “Zur Kommen-
dation und Absetzung Tassilos III.”, Zeitschrift für bayerische Landesgeschichte 43 (1980)
pp. 291–327, esp. pp. 323–4.
138
W. Hartmann and H. Dopsch, “Bistümer, Synoden und Metropolitanverfas-
sung”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988) pp. 318–26, esp. pp. 319–22; Berg, “Zur Organisation
der bayerischen Kirche”, p. 191; Wolfram, 378–907: Grenzen und Räume, p. 89; Stör-
mer, “Die bayerische Herzogskirche”, pp. 131–3; Jahn, Ducatus Baiuvariorum, p. 223;
id., “Hausmeier und Herzöge”, pp. 331–2.
139
Wolfram, “Das Fürstentum Tassilos III.”, p. 165; Classen, “Bayern und die
politischen Mächte”, pp. 175–6; Jahn, Ducatus Baiuvariorum, pp. 392–4; 469–70;
Jarnut, “Genealogie und politische Bedeutung”, pp. 19–20.
140
Annales Nazariani a. 788, ed. G.H. Pertz, MGH SS 1 (Berlin 1826) pp. 43–4:
adduxerant haec omnia una cum thesauris ac familia eorum copiosa valde ad iam dictum regem.
See Becher, Eid und Herrschaft, pp. 68–9.
456  

the anthropomorphic ruler’s staff,141 which he started to use in 787;

and the building of the cathedral at Salzburg, possibly intended as
coronation church.142 This claim’s material basis, however, was not
sufficient to pursue it militarily against the Frankish king.143 Thus,
Agilolfing high-flying was followed by their downfall, with the trial
of 788, and the temporary end of Bavarian ambitions of indepen-
dence.144 Yet, the significance of Regensburg as a late Carolingian
residence and the role of the Liutpoldings and the Bavarian Liudol-
fings cannot be explained in the absence of the achievements of the
Agilolfings.145

6. What part does the Roman Empire play in this process?

In this context, the Roman Empire only had an indirect influence.

Although the ethnogenesis of the Bavarians and the organisation of
the Bavarian duchy had as their basis the former Roman provinces
of Rhaetia and Noricum and their infrastructure of a network of

141
Wolfram, “Das Fürstentum Tassilos III.”, pp. 170–1; Classen, “Bayern und die
politischen Mächte”, p. 178. This staff is, however, not related to the so-called Tassilo-
chandelier in Kremsmünster, see K. Holter, “Kunstschätze der Gründungszeit”, Die
Anfänge des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberösterreichischen
Landesarchivs, Ergänzungsband 2 (Linz 1978) pp. 111–43, esp. p. 129; Wolfram,
378–907: Grenzen und Räume, p. 91.
142
H. Vetters, “Die mittelalterlichen Dome zu Salzburg”, Frühmittelalterliche Studien
5 (1971) pp. 413–35, esp. pp. 418–26; id., “Die mittelalterlichen Dome Salzburgs”,
Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and R. Juffinger (Salzburg
1985) pp. 286–316, esp. pp. 296–313; S. Haider, “Zur Baugeschichte des Salzburger
Virgil-Domes”, Mitteilungen des Instituts für Österreichische Geschichtsforschung 80 (1972) pp.
35–47, esp. p. 47; H. Sedlmayr, “Die politische Bedeutung des Virgildomes”, Mittei-
lungen der Gesellschaft für Salzburger Landeskunde 115 (1975) pp. 145–60, esp. pp. 150–1;
Wolfram, “Die Zeit der Agilolfinger”, pp. 147–8. But see also Reindel, “Salzburg
und die Agilolfinger”, p. 71, as well as H.R. Sennhauser, “Die Salzburger Dombauten
im Rahmen der frühmittelalterlichen Baukunst Europas”, Virgil von Salzburg. Missionar
und Gelehrter, ed. H. Dopsch and R. Juffinger (Salzburg 1985) p. 326.
143
Wolfram, 378–907: Grenzen und Räume, pp. 86–93.
144
Classen, “Bayern und die politischen Mächte”, pp. 180–4; Kolmer, “Zur
Kommendation und Absetzung Tassilos III.”, pp. 291–327; Jahn, Ducatus Baiuvariorum,
pp. 540–50; Becher, Eid und Herrschaft, pp. 64–71.
145
K. Bosl, “Das bayerische Stammesherzogtum”, Zeitschrift für bayerische Landesgeschichte
25 (1962) pp. 275–82 [repr. Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung
60 (Darmstadt 1965) pp. 1–11, esp. pp. 4–5]; id., “Regensburgs politische Stellung
im frühen Mittelalter”, Zeitschrift für bayerische Landesgeschichte 34 (1971) pp. 3–14.
   457

roads, Roman continuity was nonetheless limited by the departure

of Roman military and civil personnel after Odoacer’s eviction order
in 488. In the Ostrogothic period, Ravenna and the two Rhaetias
were considered to be under the jurisdiction of the diocese of Italy,
and hence under the authority of the Ostrogothic king.146 Even so,
after the regions north of the Alps were ceded by King Witigis in
536/37, or even earlier, Italian influence was clearly diminished.147
Yet, the area later inhabited by the Bavarians maintained its signifi-
cance in the context of Frankish-Byzantine relations, at least until
the Avars occupied the Pannonian plains and Slavonic groups migrated
into Inner Noricum and Upper Austria. On the eastern border of
Noricum, before their departure to Italy, the Lombards played a key
role in the intersection of Frankish and Byzantine interests. Emperor
Justinian I was able to win over the Lombards in an alliance with
Byzantium in 547, with the consequence that Theudebald I felt it
necessary to organise the eastern border-region of his realm as the
duchy of the Agilolfings after the death of Theudebert I. In this he
may have followed an organisational form of duchies from the Late
Antique and Ostrogothic period.148 The duchy proved its usefulness
when at its eastern borders nomadic horsemen and Slavonic princes,
rather than Germanic regna based on Byzantine influence, tried to
establish an independent regime.
Roman influence on the ethnogenesis of the Bavarians and on
the rule of the Agilolfings is most clearly visible in the Romanic

146
Contra Zeiß, “Bemerkungen I”, pp. 352–4. For support of the idea that the
Alpine foothills belonged to the Ostrogothic realm, see Beyerle, “Süddeutschland”,
pp. 68–9; 76–7; Wolfram, “Ethnogenesen im frühmittelalterlichen Donau- und
Ostalpenraum”, pp. 110–1; id., 378–907: Grenzen und Räume, pp. 64–5. K. Reindel,
“Bistumsorganisation”, pp. 300–1 supposes a Gothic rulership over the mountains,
but not over the Alpine foothills. K.A. Eckhardt, Merowingerblut, p. 103, believes it
temporarily belonged to the Herulian realm. F. Lotter, “Die germanischen Stammes-
verbände”, pp. 56–7, argues that another Germanic contingent from Bohemia arrived
in conjunction with the Lombard victory over the Herulians.
147
Agathias Myrinaei, Historiarum libri quinque 1,6, ed. R. Keydell, Corpus Fontium
Historiae Byzantinae 2 (Berlin 1967) pp. 16–9; cf. Schneider, “Fränkische Alpen-
politik”, pp. 27–8.
148
J. Jahn, Ducatus Baiuvariorum, pp. 6; 8–11; 561–3, as well as id., “Hausmeier
und Herzöge”, p. 320, supposes that a duchy was created on the Danube as early
as the Ostrogothic time, while H. Wolfram, “Ethnogenesen im frühmittelalterlichen
Donau- und Ostalpenraum”, p. 112, discusses a situation of competition, of little
synchronicity, between the older, Late Antique Ostrogothic Rhaetian duchy and
the duchy created by the Franks.
458  

population, which had strong roots not limited to the lands near and
in the Alps.149 Their integration was an important condition for the
successful formation of the Bavarian people;150 their Christian faith151
formed the basis for relations with the Frankish realm, which was
organised along Roman-Christian lines, and the successful mission
of immigrant Germanic groups. Their settlement on fiscal lands was
further an important economic and financial factor in the rule of
the Agilolfings.152 The educational forms preserved within their cir-
cles may have possibly played an influential role: the dukes Theodo
and Odilo both recruited members for their chancellery from fam-

149
J. Sturm, “Romanische Personennamen in den Freisinger Traditionen”, Zeitschrift
für bayerische Landesgeschichte 18 (1955) pp. 61–80, esp. pp. 68–9; Schwarz, “Die
bairische Landnahme um Regensburg”, pp. 59; 61–4; H. Dachs, “Römerkastelle
und frühmittelalterliches Herzogs- und Königsgut an der Donau”, Aus Bayerns Frühzeit.
F. Wagner zum 75. Geburtstag, ed. J. Werner, Schriftenreihe zur bayerischen Landes-
geschichte 62 (München 1962) pp. 293–320 [repr. Zur Geschichte der Bayern, ed.
K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 44–84, esp. pp. 64–9]; id.,
“Baiern und Walchen”, Zeitschrift für bayerische Landesgeschichte 33 (1970) pp. 857–958,
esp. pp. 921–7; Reindel, “Das Zeitalter der Agilulfinger”, pp. 127–30; F. Prinz,
“Salzburg zwischen Antike und Mittelalter”, Frühmittelalterliche Studien 5 (1971) pp.
10–36, esp. pp. 18–20; id., “Fragen der Kontinuität”, pp. 710–1; 722–7; Eckhart,
“Nach- und Weiterleben”, pp. 27–38; Störmer, “Zum Prozeß sozialer Differenzierung”,
pp. 162–3; M. Pfister, “Entstehung, Verbreitung und Charakteristik des Zentral-
und Ostalpen-Romanischen vor dem 12. Jahrhundert”, Frühmittelalterliche Ethnogenese
im Alpenraum, ed. H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985)
pp. 49–95, esp. pp. 61–4; Wolfram, “Ethnogenesen im frühmittelalterlichen Donau-
und Ostalpenraum”, pp. 124–6; id., “Die Zeit der Agilolfinger”, pp. 152–4; id.,
Salzburg, Bayern, Österreich, pp. 25–6; 37–9; Reiffenstein, “Stammesbildung und Sprach-
geschichte”, p. 1340; D. Messner, “Salzburgs Romanen”, Virgil von Salzburg. Missionar
und Gelehrter, ed. H. Dopsch and R. Juffinger (Salzburg 1985) pp. 103–11. There is
hardly any archaeological evidence for those Romance speakers that remained, see
Fischer, Das bajuwarische Reihengräberfeld von Staubing, pp. 87–9.
150
Prinz, “Herrschaftsstruktur”, pp. 26–7.
151
V. Miloj‘iÆ, “Zur Frage der Zeitstellung”, pp. 128–9 and 133, thinks that vari-
ous Late Antique sacred buildings were still used in the later Bavarian regions, and
concludes on this basis that the Roman population played the role of intermediary
in the Christianisation of the Bavarians. Also H. Wolfram, 378–907: Grenzen und
Räume, p. 96, stresses the importance of Roman Christianity for the conversion of
the Bavarians. See also Jahn, Ducatus Baiuvariorum, pp. 154–5.
152
Prinz, “Herrschaftsstruktur”, pp. 22–3; Wolfram, “Das Fürstentum Tassilos
III.”, pp. 166–7; J. Jahn, Ducatus Baiuvariorum, p. 86, points out that the Romance
speakers in Salzburg knew how to win gold, Th. Fischer, Das bajuwarische Reihengräberfeld
von Staubing, pp. 82–3, has demonstrated the significance of the Roman population
for the transmission of Roman pottery technology. See also H. Geisler, “Barbing-
Kreuzhof ”, Regensburg—Kelheim—Straubing 1, ed. S. Rieckhoff-Pauli and W. Torbrügge,
Führer zu archäologischen Denkmälern 5 (Stuttgart 1984) pp. 164–73, esp. pp.
170–3. On the tribute of Romance speakers for fiscal lands see also Jahn, Ducatus
Baiuvariorum, pp. 247–8.
   459

ilies of the Salzburg Romanitas. Organisation based on literacy—a

tool of increased importance for effective rulership—owed much to
the Roman population.153
For the governing efforts of the dukes, the possibility of using the
late Roman infrastructure played a particularly significant role. This
was exemplified by their reoccupation of the legionary camp in
Regensburg (a massive and extremely strong construction of lime-
stone blocks built without mortar) as a centre of power,154 as well
as their continued use of many other Roman fortifications, roadside
stations and roads.155 That they were, indeed, used is demonstrated
by, for instance, the persecution of Saint Emmeram on his way to
Rome and his murder in Helfendorf close to a former Roman tav-
ern, which was apparently still used in the Agilolfing period and at
which site, at least in the eighth century, there was ducal property.156

153
Störmer, Früher Adel, p. 213; Bosl, Bayerische Geschichte, p. 48; Prinz, “Herr-
schaftsstruktur”, p. 26; Wolfram, “Der Heilige Rupert in Salzburg”, p. 86; Jahn,
Ducatus Baiuvariorum, pp. 29–30; 204; 246–7; 558–60.
154
Arbeo, Vita Haimhramni 4, p. 32: ad Radasponam pervenit urbem, qui ex sectis lapidibus
constructa, in metropolim huius gentis in arce decreverat. Fischer, “Zur Archäologie des fünf-
ten Jahrhunderts”, p. 103; id., “Der Übergang von der Spätantike zum frühen
Mittelalter in Ostbayern”, p. 242; id., Das bajuwarische Reihengräberfeld von Staubing,
pp. 115–8; Wolff, “Die Kontinuität städtischen Lebens”, pp. 304–5; K. Schwarz,
“Regensburg während des ersten Jahrtausends im Spiegel der Ausgrabungen in
Niedermünster”, Jahresbericht der Bayerischen Bodendenkmalpflege 13/14 (1972/73) pp.
20–98; Dietz, Osterhaus, Rieckhoff-Pauli and Spindler, Regensburg zur Römerzeit, pp.
169–72; K. Reindel, “Regensburg als Sitz der Herrscher bis zum 10. Jahrhundert”,
Regensburg—Kelheim—Straubing 1, ed. S. Rieckhoff-Pauli and W. Torbrügge, Führer
zu archäologischen Denkmälern 5 (Stuttgart 1984) pp. 243–54; A. Schmid, “Regensburg
zur Agilolfingerzeit”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer
and H. Dopsch (Rosenheim-Salzburg 1988) pp. 136–40; Jahn, Ducatus Baiuvariorum,
pp. 37–8; P. Schmid, “König—Herzog—Bischof. Regensburg und seine Pfalzen”,
Deutsche Königspfalzen. Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 4:
Pfalzen—Reichsgut—Königshöfe, ed. L. Fenske, Veröffentlichungen des Max-Planck-
Instituts für Geschichte 11,4 (Göttingen 1996) pp. 53–83, esp. pp. 53–62.
155
Dachs, “Römerkastelle”, pp. 44–84; W. Störmer, “Fernstraße und Kloster.
Zur Verkehrs- und Herrschaftsstruktur des westlichen Altbayern im frühen Mittelalter”,
Zeitschrift für bayerische Landesgeschichte 29 (1966) pp. 299–343, esp. pp. 313–32; 339–43;
J. Jahn, “Urkunde und Chronik. Ein Beitrag zur historischen Glaubwürdigkeit der
Benediktbeurer Überlieferung und zur Geschichte des agilolfingischen Bayern”,
Mitteilungen des Instituts für Österreichische Geschichtsforschung 95 (1987) pp. 1–51, esp. pp.
48–9.
156
A. Sandberger, “Römisches Straßensystem und bairische Siedlung im Osten
von München”, Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner,
Schriftenreihe zur bayerischen Landesgeschichte 62 (München 1962) pp. 287–92,
esp. pp. 290–2; A. Kraus, “Zweiteilung des Herzogtums der Agilolfinger? Die Probe
aufs Exempel”, Blätter für deutsche Landesgeschichte 112 (1976) pp. 16–29, esp. p. 26;
460  

Summary

The Agilolfing dukes of the Bavarians started their career in Regensburg

employed as office-holders of the Merovingian kings. By virtue of
their background and the various gentes from which they emerged,
as well as their close relationship even with senatorial circles of south-
western Gaul, conditions were favourable for the formation of Bavarians
as a people from both eastern and western Germanic groups and
the autochthonous Roman population of Rhaetia and Noricum. Their
rule over a fertile land, equipped with infrastructure of the Roman
period and characteristics of a border-region,157 enabled the Agilolfings,
who profitted from their eminent background, to conduct an increas-
ingly independent style of politics. They secured this position by forg-
ing bonds with their Germanic neighbours. The downfall of the
Merovingian kingship created ulterior motives for increasing their
distance from the Frankish rulers, and hence the Agilolfings were
able gradually to achieve a position equal to royalty in the late sev-
enth and the eighth centuries. However, they never completely man-
aged to shake off Frankish supremacy, and consequently Tassilo III’s
accession to power in the 770s was followed by his deposition by
Charlemagne in 788 and eventually even by his forced renunciation
of all rights during the Synod of Frankfurt in 794.158 The implica-
tions of this dismissal of Bavarian ducal power do not differ much
from the fate of Desiderius, the last king of the Lombards, who also
spent the end of his life in a monastery.159 The similarities in
Charlemagne’s behaviour in both cases once again give a sense of
the position of the Agilolfing dukes in Bavaria. They found them-
selves in a permanently ambivalent position in relation to the Frank-
ish kings: on the one hand, they found a livelihood as appointed

Hartung, Süddeutschland, p. 200; Mayr, “Neuerliche Anmerkungen zur Todeszeit des

Heiligen Emmeram”, pp. 204–5; Jahn, Ducatus Baiuvariorum, pp. 44–5.
157
Jahn, Ducatus Baiuvariorum, pp. 552–4, on natural-spatial and logistic-geographical
prerequisites for tribal formation and Agilolfing rulership.
158
Classen, “Bayern und die politischen Mächte”, p. 184; Kolmer, “Zur Kommen-
dation und Absetzung Tassilos III.”, pp. 316–7; Becher, Eid und Herrschaft, pp. 72–3;
Jahn, Ducatus Baiuvariorum, pp. 549–50.
159
Classen, “Bayern und die politischen Mächte”, p. 184; W. Laske, “Die
Mönchung Herzog Tassilos III. und das Schicksal seiner Angehörigen”, Die Anfänge
des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberösterreichischen Landes-
archivs, Ergänzungsband 2 (Linz 1978) pp. 189–97, esp. pp. 190–5.
   461

dukes and enjoyed a guarantee of their position in the Lex Baiuvariorum

that had been shaped by the Franks; on the other hand, they both
expressed self-consciousness and pursued independent policies. This
unique situation gave them freedom as well as established limits on
their power, but it could not prevent their final defeat by the king
of the Franks.

[Translated by Inge Lyse Hansen and Carine van Rhijn]

This page intentionally left blank
 AVARS AND AVAR ARCHAEOLOGY
AN INTRODUCTION

Falko Daim

Introduction

The Avar Empire is one of the most fascinating cultural-historical

phenomena of the European Early Middle Ages.1 From the year 568
until Charlemagne’s Avar wars (788–803), the riders of the steppes
dominated the Carpathian Basin and its environs (fig. 1). The early
period, up to 626, in particular, and the end of the Avar Empire
are well illuminated by numerous written sources. Countless archae-
ological finds, more than 60,000 graves, some hoards and settle-
ments, permit us—provided we are prepared to adhere to a number
of methodical rules—to reconstruct a culture which appears to have
reacted very flexibly to external influences while nevertheless main-
taining some basic traits, apparently as the indispensable “backbone”
of Avar identity, from the period of first settlement to the decline
of the Avar Empire. At the same time, the large number of finds
and the great size of many cemeteries make the Avar material highly
suitable for testing modern evaluation methods. The majority of these
are sophisticated statistical methods; however, technical and scientific
examinations of archaeological finds are becoming increasingly impor-
tant (fig. 2).
The aim of this essay is to present the English-speaking audience
with an overview of Avar archaeology. At the same time, the pur-
pose is to confront the results of archaeological research with the

1
This essay would not have been possible without advice, suggestions and help
from a number of friends and colleagues. I would especially like to thank: Birgit
Bühler (Vienna), Anton Distelberger (Vienna), Róbert Müller (Keszthely), Silvia
Müller (Vienna), Péter Somogyi (Frastanz), Béla Miklós Szoke
 (Budapest) and Tivadar
Vida (Budapest). For the translation I am indebted to Birgit Bühler and to Elizabeth
Fox (Edinburgh) for thorough proof-reading of the text. I would also like to thank
Beate Lethmayer for the lay-out of the plates and Franz Siegmeth for digitalising
them.
 464
 
Fig. 1: The Avars, Central- and South-Eastern Europe around the Year 600 (according to W. Pohl)
   () 465

political development, which has been reconstructed mainly by the

work of Walter Pohl.
This introductory work is aimed at demonstrating that the diverse
character of the different types of sources—especially the written and
the archaeological evidence—of course facilitate special research-ques-
tions and results, but that only the combined view of the many
results will give a detailed picture of the fascinating world of the
Avars: a world which, with its way of life and system of values, is
so far from ours but is nevertheless part of our European past.

Avar Archaeology—Avar Finds

1 Early Avar Period I (568–626) and II (626–650/670)

The archaeological material of the Early Avar Period (568–650) is
extremely diverse and archaeologists are trying to illuminate the par-
allel lines of development with all their connections. Although we
are still a long way from a detailed overall picture, some admirable
achievements of recent archaeology and some lucky finds allow us
to reconstruct some lines of development. For mainly methodologi-
cal reasons, we tread on somewhat firmer ground only once the
large cemeteries are established. The latter enable us to study the
cultural development of small population groups.
Maurikios gives us an impressive and almost admiring account of
Avar warriors in his Strategikon. He even recommends that certain
parts of their equipment, for instance the long coat, the tent and
the horse armour (of which not a single example has so far been
found in the archaeological material), should be adopted by the
Byzantine army. It is understandable that for a long time, Hungarian
archaeology was mainly concerned with identifying the first genera-
tion of Avar settlement archaeologically. It seems that the Avar war-
riors who entered the Carpathian Basin in 568 were all accomplished
riders and—according to Maurikios—shrewd tacticians, equipped—
in a uniform and rather Spartan way—with riding equipment and
weapons. The Avars used the composite bow, which was a techni-
cal miracle, consisting mainly of wooden parts with tendons, glued
together on the inside with strips of horn, while in the centre and
at the ends it was strengthened with lamellae made of deer antler.
The different parts made use of forces of pressure on the inside and
 466
 
Fig. 2. The most important archaeological sites of Avar culture
   () 467

Austria 43 Előszállás 92 Szegvár – Sápoldal

1 Edelstal (Hungarian: 44 Gátér 93 Székkutas
Nemesvölgy) 45 Gyód 94 Szekszárd – Bogyiszlói út
2 Frohsdorf 46 Győr – Téglavető – dűlő 95 Szekszárd –
3 Leobersdorf 47 Hajdúdorog Mocfacsárda
4 Mistelbach 48 Halimba 96 Szentes – Berekhát
5 Mödling 49 Hird 97 Szentes – Kaján
6 Münchendorf 50 Hódmezővásárhely 98 Szentes – Lapistó
7 Sommerein 51 Hortobágy – Árkus 99 Szentes – Nagyhegy
8 Wien 11 – Csokorgasse 52 Igar 100 Táp
9 Wien 23 – Liesing 53 Inota 101 Tápe
10 Zillingtal 54 Iváncsa 102 Tiszafüred – Majoros
11 Zwölfaxing 55 Jánoshida 103 Tolnanémedi
56 Jutas 104 Toponár
Czech Republic 57 Káptalantóti 105 Törökbálint
12 Dolní Dunajovice 58 Kecel 106 Üllő
59 Kecskemét 107 Ürbőpuszta
Slovakia 60 Keszthely 108 Várpalota
13 Blatnica 61 Keszthely – 109 Vasasszonyfa
14 ’ataj Fenékpuszta 110 Zalakomár
15 Cífer Pác 62 Kisköre 111 Zamárdi
16 Devínska Nová Ves 63 Kiskőrös 112 Závod
17 Holiare 64 Kiszombor
18 Komarno 65 Kölked – Feketekapu Former Yugoslavia
19 Moravský Ján 66 Környe 113 Aradac (Hungarian:
20 Nové Zámky 67 Kunbábony Aradka)
21 ”ebastovce 68 Kunhegyes – 114 Brodski Drenovac
22 ”túrovo Bánhalma 115 ’adjavica
23 Záhorska Bystrica 69 Kunmadaras (Hungarian:
24 ¥elovce 70 Kunszentmárton Csadjavica)
25 Zemianský Vrbovok 71 Lébény 116 ’elarevo
26 Zitavská Tôn 72 Mártély 117 ’oka (Hungarian:
73 Mezőberény Csóka)
Hungary 74 Mosonszentjános 118 Novi Balnovci
27 Abony 75 Nagyharsány 119 Pan‘evo
28 Alattyán 76 Nagypall 120 Zmajevac (Hungarian:
29 Bócsa 77 Ozora – Tótipuszta Vörösmart)
30 Boldog 78 Pápa
31 Bóly 79 Pécs Romania
32 Budakalász – Dunapart 80 Pilismarót 121 Band (Hungarian:
33 Budapest – Csepel 81 Pókaszepetk Mez őbánd)
34 Budapest – 82 Regöly 122 Felnac (Hungarian:
Farkasrét 83 Romonya (= Ellend) Fönlak)
35 Cibakháza 84 Solymár 123 Sînpetru German
36 Cikó 85 Szarvas (Hungarian:
37 Csákberény 86 Szebény Németszentpéter)
38 Csanytelek 87 Szeged – Átokháza 124 Sînnicolau Mare
39 Csengele 88 Szeged – Fehértó (Hungarian:
40 Deszk 89 Szeged – Kundomb Nagyszentmiklós)
41 Dunaújváros 90 Szeged – Makkoserdő 125 Z>lau (Hungarian:
42 Egerág 91 Szegvár – Oromdűlő Zilah)
468  

forces of tension on the outside, the reinforcements thus acting as

levers. We have merely a vague notion of the types of glue which
were used. In our regions, only the antler reinforcements of this
miraculous weapon remain, but bows of this type which have been
preserved complete are known from the Caucasus. It is quite prob-
able that in the Avar Empire, simple wooden bows, which usually
cannot be detected by the archaeologists, were also used in addition
to the composite bows. While in the west, the main responsibility
of fighting still lay with the infantry, all Avars were mounted. Iron
stirrups, without exception, formed part of the equipment of every
Avar warrior. They stabilize the riding position in combat, which is
particularly important when using the long lance, which was appar-
ently part of the equestrian warrior’s standard equipment. This set,
consisting of two stirrups, iron bit and lance-head, sometimes occurs
as a “votive deposit pit” (Opfergrubenfund ), without human remains
(plates 1, 2 & 3). These definitely belong to the Early Avar Period,
but at least for now, cannot be dated more precisely.2 However, here
there is evidence of a burial custom which seems to have existed
among the Huns in the first half of the fifth century, although with-
out evidence of cremation. The earliest Avar finds in the Carpathian
Basin also include various bone objects, among them belt buckles,
needle containers and a tool for undoing knots. All of these gener-
ally correspond to the fashion of the eastern steppes.3 The strange
Avar burial place at Börcs-Nagydomb has even yielded an entire
belt-set which had been carved from bone.4
There is a significant accumulation of typical artefacts from the
Early Avar Period in present-day western Hungary (former Pannonia)

2
C. Bálint, “Probleme der archäologischen Forschung zur awarischen Landnahme”,
Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters. Methodische
Grundlagendiskussion im Grenzbereich zwischen Archäologie und Geschichte 1, ed. M. Müller-
Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen 1993) pp.
195–273, esp. pp. 203–4.
3
É. Garam, “Bemerkungen zum ältesten Fundmaterial der Awarenzeit”, Typen
der Ethnogenese unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F.
Daim, Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 204. Veröffentlichungen der Kommission für Frühmittelalterfor-
schung 13 (Wien 1990) pp. 253–72; Bálint, “Awarische Landnahme”, pp. 199; 217.
4
P. Tomka, “Frühawarenzeitliche Hirten in der kleinen Tiefebene”, Ethnische und
kulturelle Verhältnisse an der mittleren Donau vom 6. bis zum 11. Jahrhundert. Symposium Nitra
6. bis 10. November 1994, ed. D. Bialeková and J. Zábojník (Bratislava 1996) pp.
141–9, esp. pp. 142 ff.
   () 469

and a somewhat lesser concentration to the east of the Tisza, while

the region between the rivers Danube and Tisza remains empty.5
The reasons for this are self-evident: Pannonia offered a certain infra-
structure and large cleared areas, which were perhaps still being cul-
tivated. It also had at least a sparse local population. In addition to
this, there was the network of Roman roads, which presumably could
still be used extensively, even if it had not been restored for more
than 200 years. In the course of their settlement process, the Avars
also settled in the formerly Gepid territory east of the Tisza. As
Herwig Wolfram once said, “To rule only makes sense if one is rul-
ing people”. As far as we know today, in the Early Avar Period,
comparatively large and wealthy settlements existed only in western
Hungary (Budakalász-Dunapart, Zamárdi). However, if we want to
draw conclusions about the location of the centres of power from
the wealth of the graves, then the focus of the Avar Empire would
appear to have been located close to the Danube, between Danube
and Tisza and just east of the Tisza. This is demonstrated in a par-
ticularly impressive manner by the distribution of the princely graves
of the Bócsa-Kunbábony group (e.g. pl. 19–22) which belong to the
middle of the seventh century and—apparently—seem to manifest a
certain turning away from the Byzantine model. They are all located
east of the Danube.6
Byzantine culture seems to have held great appeal for the Avar
rider-nomads. Byzantine authors appear to have actually made fun
of this fact. However, this does not necessarily imply that the Avars
were in the process of turning into provincial Romans. The Avars
selected carefully what they wanted to integrate into their culture.
As this involves aspects of mentality, of collective psychology and of
a set of cultural values, we still have to deal with this question at
length.
As even before they settled, the Avars had had enough time to
acquaint themselves with Byzantium, it is possible that they had
already brought some of the Byzantine means of representation with
them when they came into the Carpathian Basin. In this context,
one should consider the belts with cast silver “mask fittings”. These
were common in the Black Sea and the Caucasus region in the sixth

5
Bálint, “Awarische Landnahme”, pp. 243 ff.
6
Garam, “Ältestes Fundmaterial”, fig. 11.
470  

century, while in the Carpathian Basin only a few examples have

been found so far. (pl. 2,1–3).7 Similar objects, which may be some-
what newer but appear to be related to the cast “mask fittings” occur
in two Lombard graves, Arcisa 2 and 5, in both cases together with
strap-ends of Sadovec-Arcisa type.8 A parallel was found in an Avar
grave from Budakalász (pl. 2,4). There can be no doubt that the
Lombards got to know both the strap-ends and the fittings in Italy,
as no comparable objects have been found so far in Lombard graves
in Pannonia. Grave 9 from Kisk rös  contains a belt-set, in which
cast-fittings which are related to those from Arcisa 2, have been
combined with strap-ends of Martynovka type. These are only rarely
decorated in openwork technique—as are the mask fittings—but
instead have a linear ornament which, in Hungary, has traditionally
(but incorrectly) been called Tamga-decoration (pl. 2,5).9
The belt-sets of Aradac type belong to the first half of the sev-
enth century. These are strap-ends and mounts pressed in a mould,
decorated with a triple hemispherical dent supplemented by dot-
comma ornament (e.g. from Keszthely-Fenékpuszta: pl. 2,7).10 A
mould of this type was found in Fenlac (Fönlak).11 The distribution
of this type suggests that it is in fact derived from the products of

7
C. Bálint, “Kontakte zwischen Iran, Byzanz und der Steppe. Das Grab von
Ü‘ Tepe (Sowj. Azerbajd≥an) und der beschlagverzierte Gürtel im 6. und 7.
Jahrhundert”, Awarenforschungen 1, ed. F. Daim, Archaeologia Austriaca 1. Studien
zur Archäologie der Awaren 4 (Wien 1992) pp. 309–496. Regarding the multi-part
belt-sets in the Mediterranean region see—most recently—C. Bálint, “Byzantinisches
zur Herkunftsfrage des vielteiligen Gürtels”, Kontakte zwischen Iran, Byzanz und der
Steppe im 6.–7. Jahrhundert, ed. id., Varia Archaeologica Hungarica 10 (Budapest
2000) pp. 99–162.
8
M. Martin, “Awarische und germanische Funde in Männergräbern von Linz-
Zizlau und Környe. Ein Beitrag zur Chronologie der Awarenzeit”, A Wosinsky Mór
Múzeum Évkönyve 15 (1990) pp. 65–90, esp. pp. 66–7; S. Uenze, Die spätantiken
Befestigungen von Sadovec, Münchener Beiträge zur Vor- und Frühgeschichte 43 (München
1992) pp. 187–92 and pl. 11.
9
N. Fettich, Die Metallkunst der landnehmenden Ungarn, Archaeologia Hungarica 21
(Budapest 1937) pl. CXXVI; L.V. Pekarskaja and D. Kidd, Der Silberschatz von
Martynovka (Ukraine) aus dem 6. und 7. Jahrhundert, Monographien zur Frühgeschichte
und Mittelalterarchäologie 1 (Innsbruck 1994) pl. 31,1–3; 32,1–4; 32,6.
10
Martin, “Linz-Zizlau und Környe”, p. 67 n. 8 and distribution map, fig. 4;
J. Werner, “Nomadische Gürtel bei Persern, Byzantinern und Langobarden”, Atti
del Convegno Internazionale sul Tema: La civiltà dei Longobardi in Europa. Roma 1971 (Rome
1974) pp. 109–56, esp. pp. 127 ff. with note 48.
11
J. Hampel, Alterthümer des frühen Mittelalters in Ungarn I–III (Braunschweig 1905)
pl. 446,2; see also in future Z. Rácz, Awarische Goldschmiedegräber, Monographien zur
Frühgeschichte und Mittelalterarchäologie (forthcoming).
   () 471

Byzantine craftsmen, but seems to have been restricted to the Balkans

(Gari‘in Grad)12 and the Carpathian Basin. No examples of this type
are known so far either from Italy or the Black Sea region. Syna
Uenze has suggested—at first not altogether convincingly—that the
strap-ends and belt-fittings with triple hemispherical dent may have
been derived from the buckles of Sucidava type.13 However, the vari-
ant of a Byzantine buckle from Gyenesdiás grave 64 appears to prove
that she may in fact have been right in doing so: it is decorated
with a similar motif in openwork and therefore—like a “missing
link”—establishes a definite typological connection (pl. 25,3). However,
Gyenesdiás 64 has a post quem date of 654, which is a little late
for the fittings of Aradac type.
The belt-sets of Kunágota-Mersin type (after Bálint) correspond
much more closely to the kind of belt fashion which was common
in Byzantium.14 The complete belt-set consists of matching strap-
ends and fittings, which have been either chased individually (Kunágota:
pl. 7,1) or “pressed” in a mould. The main strap-ends have medal-
lions with Christian monograms, which, however, have been replaced
by geometric ornaments—presumably especially to suit the Avar mar-
ket. One of the most beautiful examples of such a belt-set is in fact
a set of moulds from the smith’s grave of Kunszentmárton (pl. 6,2).
Belt-sets of Kunágota-Mersin type still occur—although of somewhat
lesser quality—up to the end of the seventh century, as is demon-
strated by the grave from Ozora-Tótipuszta, for which we have a
post quem date. Particularly interesting are the so-called “fringed
mounts” from horse harness, which are one of the most character-
istic types of mount from the Early Avar Period and of the early
Middle Avar Period.15 The origin of this square or triple hemi-
spherical type of fitting, which is characterised by a set of vertical
grooves, may be traced back to the cast mask fittings. For example,
three cast horse harness-mounts from Arcisa 5 have a fringe deco-
ration. A cast mask fitting of triple hemispherical shape with fringe
decoration is preserved in the ”umen museum (Bulgaria).
Especially in western Hungary, the Germanic component plays a
vital role within Avar material. This does not only mean that Germanic

12
Werner, “Nomadische Gürtel”, p. 129 n. 48.
13
Uenze, Sadovec, p. 185, fig. 14.
14
Bálint, “Awarische Landnahme”, p. 223.
15
Ibid., p. 214.
472  

types of decoration were adopted by Avar craftsmen. Corresponding

dress customs are usually linked with the “Germanic” objects. Examples
are the three- and four-part belt-sets with silver wire inlay (Tauschierung)
(Kölked-Feketekapu A), which have already been mentioned, but
especially the so-called “toothcut” decoration. The most famous
ensemble of this type, the so-called “Jankovich-Gold”, constitutes a
sword strap set of “western type” (pl. 15,1).16 Several pieces of jew-
ellery from two graves which have been published only recently, are
on a par with the Jankovich-finds: The artfully constructed gold disc
fibula, the bracelet and the fingering from the woman’s grave 119
of Kölked-Feketekapú B,17 which also included a wooden bowl with
silver fittings (pl. 14). The woman’s grave 85 did not contain quite
as much gold and silver, however, it is almost more interesting from
a historical point of view. The woman’s belt was decorated with a
lovely shield-tongue buckle made of gilded bronze with silver inlay,
which in turn was decorated with niello, and depicts a Germanic
god with two swords, around whose shoulders a snake was coiled,
biting the warrior’s left arm: According to Attila Kiss, this is a depic-
tion of Tiuz or Tyr18 (pl. 11,1). Next to it, a stylistically matching
rectangular mount had been found (pl. 11,2). The belt terminated
in a strap-end decorated with interlace (pl. 11,5), on her head, the
lady wore gold basket earrings and a hair-pin with a decorative part
made of gold (pl. 11,3). With the help of the objects’ exact position
in the grave (pl. 12), the deceased’s dress may be reconstructed, to
a certain degree (pl. 13). Splendid finds with “toothcut” ornament
occur in the necropolis of Zamárdi, south of Lake Balaton, which
has so far only been partly published and presented to the public
in the form of exhibitions. This includes a pair of bracelets made
of solid silver, but also belt pendants, which are part of female dress
(example from Kölked-Feketekapú B, grave 85: pl. 11,5 and 12) and
belt-sets as part of male dress, where the shape of the strap-ends is

16
F. Daim et al., Das awarische Gräberfeld von Leobersdorf, Niederösterreich, 2 vols.,
Studien zur Archäologie der Awaren 3. Denkschriften der Österreichischen Akademie
der Wissenschaften, philosophisch-historische Klasse 194. Veröffentlichungen der
Kommission für Frühmittelalterforschung 10 (Wien 1987) p. 132 n. 93; T. Vida,
“Merowingische Spathagurte der Awarenzeit”, Communicationes Archaeologicae Hungariae
(2000) pp. 161–75, esp. pp. 167 ff.
17
A. Kiss, “Elozetes
 jelentés (II.) A Kölked—Feketekapui avarkori települéles és
temetok  Folia Archaeologia 39 (1988) pp. 173–94, esp. Part 2, pl. 4; 5,1–4.
 ásatasárol”,
18
Ibid., Part 1, pp. 294–303.
   () 473

closely related to those from typical early Avar sets (pl. 15,2). Recently,
Tivadar Vida has devoted particular attention to the female costume
of the Germanic population of the Early Avar Period, starting with
some finds from the necropolis of Budakalász-Dunapart, where the
reciprocal permeation of eastern-Avar, Germanic, Roman and
Byzantine traditions is evident.19 Certainly, the defeated Gepids and
the remaining Lombards must have played a much more substan-
tial role within this massive agglomeration of Germanic traditions,
but it is nevertheless becoming increasingly clear that during the
Early Avar Period, intensive contacts must also have existed between
the Carpathian Basin and the Baltic, present-day north-western
Germany and the upper Danube region.20
The interaction between the local Romanic and Germanic popu-
lation and the new lords from the East, as well as the strong impulses
from Italy and the northern pre-Alpine region can be studied in one
of the most fascinating archaeological provinces of the Carpathian
Basin, the so-called “Keszthely culture”, located at the western end
of Lake Balaton. The term “Keszthely-Kultur”, which is now com-
monly used, was defined by Ilona Kovrig and Attila Kiss.21 In 1993,
Éva Garam called for a less rigid, more flexible and discerning view
of the cultural phenomena in the Lake Balaton region and in south-
ern Hungary.22 The archaeological phase comprises characteristic
types of disc fibulas and bracelets with terminals in the shape of

19
T. Vida and A. Pásztor, “Der beschlagverzierte Gürtel der Awaren am Beispiel
des Inventars von Budakalász-Dunapart, Ungarn, Grab 696”, Reitervölker aus dem
Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp. 341–7; T. Vida, “Die
Ziergehänge der awarenzeitlichen Frauen im Karpatenbecken”, Acta Archaeologica
Hungarica Academiae Scientiarum Hungaricae 51 (1999/2000) pp. 367–77; A. Pásztór and
T. Vida, “Eine frühbyzantinische Bronzekanne aus dem awarenzeitlichen Gräberfeld
von Budakalász”, Kontakte zwischen Iran, Byzanz und der Steppe im 6.–7. Jahrhundert, ed.
C. Bálint, Varia Archaeologica Hungarica 10 (Budapest 2000) pp. 303–11.
20
M. Menke, “Zu den Fibeln der Awarenzeit aus Keszthely”, A Wosinsky Mór
Múzeum Évkönyve 15 (1990) pp. 187–214; Martin, “Linz-Zizlau und Környe”; id.,
“Zu den tauschierten Gürtelgarnituren und Gürtelteilen der Männergräber von
Kölked-Feketekapu A”, Das awarenzeitlich gepidische Gräberfeld von Kölked—Feketekapu A,
ed. A. Kiss and F. Daim, Monographien zur Frühgeschichte und Mittelalterarchäologie
2. Studien zur Archäologie der Awaren 5 (Innsbruck 1996) pp. 345–61.
21
I. Kovrig, “Megjegyzések a Keszthely-kultura kérdéséhez”, Archeologiai Értesíto
85 (1958) pp. 66–72; A. Kiss, “A Keszthely-kultura helye a pannoniai római kon-
tinuitás kérdésében”, Archeologiai Értesíto 95 (1968) pp. 93–101.
22
É. Garam, “Die awarenzeitlichen Scheibenfibeln”, Communicationes Archaeologicae
Hungariae (1993) pp. 99–143, esp. p. 131.
474  

animal heads (pl. 5) in the early period (end of the sixth and first
quarter of the seventh century), and so-called stylus-pins and over-
sized basket earrings (pl. 35,1,4) in the late period (eighth century),
but also many other objects and types, which however did not get
any attention because the researchers all focused on the “classical”
types. If we adhere to the definition established by Kovrig and Kiss,
the Keszthely-culture presently comprises 18 archaeological sites.23
Among the most important are the impressive late Roman castle
(“Keszthely-Fenékpuszta”) south of the present-day town Keszthely24
and the Dobogó, a hill north-west of the town of Keszthely, on
which an early medieval cemetery with approximately 4,000 graves
was found. Another large cemetery of the Keszthely-culture was
located in the south-western part of the town Keszthely (Gräberfeld
Keszthely-Stadt). The cemetery of Alsópáhok with around 1,500
graves was found to the west of the town. The necropolis at
Lesencetomaj-Piros kereszt seems to be an extremely promising site
for Avar archaeology. It is one of the most north-eastern cemeter-
ies of the Keszthely-culture and will presumably allow us to estab-
lish a fine chronological system for the types of objects which are
typical for the Keszthely-culture.25
The archaeological finds show clearly that the area around Keszthely
must have been one of the most important regions in the Carpathian
Basin during the Lombard and Avar Period. The extraordinary con-
centration of jewellery from Byzantium (earrings, belt-fittings from
the sixth and early seventh century) and the Adria region (dress-pins
with semi-circular head, presumably second half of the seventh cen-
tury),26 as well as fibulas from the Saxon27 and Frankish regions (exca-
vation by Róbert Müller 1999) is probably due to its location at the
cross-roads of important long-distance roads from Aquileia via Emona,

23
G. Kiss, “Funde der Awarenzeit in Wiener Museen—1. Funde aus der Umgebung
von Keszthely”, Archaeologia Austriaca 68 (1984) pp. 61–201. The most recent sum-
mary is: R. Müller, “Die Keszthely-Kultur”, Reitervölker aus dem Osten. Hunnen +
Awaren, ed. F. Daim (Halbturn 1996) pp. 265–74.
24
R. Müller, “Die Festung ‘Castellum’, Pannonia Inferior”, Reitervölker aus dem
Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp. 91–5.
25
Most recently Á.S. Peremi, “Lesencetowmaj-piroskereszt keszthely-kultúrás temeto
fülbevalói”, A Vesztprém Megyei Múzeumok Közleményei 21 (2000) pp. 41–75.
26
A. Distelberger, Das awarische Gräberfeld von Mistelbach, Niederösterreich, Studien zur
Archäologie der Awaren 6 (Innsbruck 1996) pp. 77–82.
27
Menke, “Fibeln”.
   () 475

Celeia, Valcum (Keszthely-Fenékpuszta) to Aquincum and from

Sirmium via Valcum to Carnuntum.28 Some of the pieces of jew-
ellery from the cemetery of Keszthely-Fenékpuszta-Horreum from
the sixth and early seventh century, especially the gold dress pins
with genuine pearls and the inscription “BONOSA”, the presum-
ably matching gold medallion with genuine pearls and rock crystal
inlay as well as the jewelled collar, are amongst the most precious
and most beautiful of their period (pl. 5,4).29 The silver S-fibulas,
the crescent-shaped gold earrings and the basket earrings are com-
mon in Italy and the regions north of the Alps. The disc fibulas will
be discussed below. A belt-set made of sheet silver of Aradac-Fönlak
type, as known from Linz-Zizlau and other Avar sites, is also impor-
tant in this context (pl. 2,7).30 The early phase of the Keszthely-cul-
ture is usually placed in the period between 568 and the early seventh
century. These dates are based on the assumption that the Keszthely-
culture is, so to speak, an archaeological branch of Avar culture,
which is methodologically incorrect. László Barkóczi alone has it
begin a little earlier. Does a local late Roman population stand
behind the Keszthely-culture, or should we—as most archaeologists
do—postulate immigrations to account for the unusually rich mate-
rial? As Ilona Kovrig has already observed, there is no continuous
settlement from Late Antiquity until the seventh century at the sites
of the Keszthely-culture (with the exception of Fenékpuszta itself ).
The remarkable heterogeneity of the archaeological material of the
late sixth and early seventh century from the Keszthely region is
another argument both against a purely local development as well

28
R. Müller, “Die spätrömische Festung Valcum am Plattensee”, Germanen, Hunnen
und Awaren. Schätze der Völkerwanderungszeit, ed. W. Menghin (Nürnberg 1988) pp.
270–3. The course of the roads in the region of Keszthely is, in detail, still sub-
ject to discussion. Recently for example E. Tóth, “Mosaburg und Moosburg”, Acta
Archaeologica Hungarica Academiae Scientiarum Hungaricae 51 (1999/2000) pp. 439–56,
esp. fig. 4.
29
Regarding the jewelled collar see É. Garam, “Über Halsketten, Halsschmucke
mit Anhängern und Juwelenkragen byzantinischen Ursprungs aus der Awarenzeit”,
Acta Archaeologica Hungarica Academiae Scientiarum Hungaricae 43 (1991) pp. 151–79.
30
L. Barkóczi, “Das Gräberfeld von Keszthely-Fenekpuszta aus dem 6. Jahrhundert
und die frühmittelalterlichen Bevölkerungsverhältnisse am Plattensee”, Jahrbuch des
Römisch-Germanischen Zentralmuseums Mainz 18 (1971) pp. 179–91, pl. 82. Recently:
É. Garam, “Gürtelverzierungen byzantinischen Typs im Karpatenbecken des 6.-7.
Jahrhunderts”, Acta Archaeologica Hungarica Academiae Scientiarum Hungaricae 50 (1999/2000)
pp. 379–91, esp. p. 381 and fig. 1a.
476  

as against the view that the early Keszthely-culture constitutes the

remains of a particular population group which settled there.
Recently it has been possible to discover a living Christian com-
munity behind the early Keszthely-culture. Some of the disc fibulas
are decorated with Christian images, for instance Jesus busts flanked
by angels (pl. 5,8) or equestrian saints (pl. 5,5). At least a couple of
these fibulas are filled with beeswax, which is a popular secondary
relic. As the closest parallels for the disc fibulas are found in the
coastal regions of Montenegro, Albania and Calabria, it would seem
that at least some of the examples from Pannonia were in fact “pri-
vate reliquiaries”, amulets which pious pilgrims brought with them
from their travels. We can assume that before the political turmoil
following the events of the year 626, a suitable church organisation,
including a direct link with Rome, must have also existed in Keszthely.
The iron belt-sets consisting of three or four parts appear to be
linked to a strong local cultural element of Germanic origin, although
research by Max Martin suggests that contacts with the region of
present-day southern Germany must have also played a role.31 Direct
connections between the Avar territory and Italy and/or present-day
Bavaria are also evident from the “North Italian” belt-sets and iron
belt-sets with wire inlay in the form of spiral ornaments (pl. 16,1–4).
The following “western” belt-set types occur frequently in an early
Avar context:

• Iron 3– or 4–part belt-sets with wire inlay (Layer 2 according to

Joachim Werner and Rainer Christlein): for example Környe 1,
18, 66, 77, 97; Kölked-Feketekapu A 44, 180, 227, 24932 (around
600, first quarter of the seventh century).
• “North Italian” belt-sets, cast bronze: The ensembles first pre-
sented—and as far as we know today, interpreted correctly—in
1961/62 by István Bóna, occur relatively frequently in Avar con-
texts.33 Apart from genuine belt-sets, there were also—apparently—
a number of imitations. The following belt-sets are presumably

31
Martin, “Kölked-Feketekapu A”.
32
Ibid.; Martin, “Linz-Zizlau und Környe”.
33
I. Bóna, “Beiträge zu den ethnischen Verhältnissen des 6.–7. Jahrhunderts in
Westungarn I—Norditalischer Bronze-Gürtelschmuck in frühawarischen Gräberfeldern”,
Alba Regia 2/3 (1963) pp. 49–63.
   () 477

imports: Sommerein 216;34 Zamardi 1035 und Zillingtal D 46936

(first half of the seventh century, in exceptional cases also third
quarter of the seventh century) (pl. 16,1).
• Iron strap-ends and mounts with wire inlay in the form of spiral
ornament: Kölked-Feketekapu A 341;37 Pitvaros 72;38 Zamárdi;39
Zillingtal D 1440 (second quarter of the seventh century) (pl. 16,2–4).

Today, technological aspects play an important role in Avar archae-

ology. On the one hand, one would like to find out as much as pos-
sible about “daily life”, and crafts are part of it. In addition to this,
we have discovered that technological criteria allow us to refine the
chronological system of the finds and that fine chronology is the pre-
requisite for any historical interpretation of the results of archae-
ological research. The noblemen’s belt-sets were subject to rapidly
changing fashions and all kinds of techniques were used to produce
them.
The cast silver “mask fittings” (pl. 2,1–3) are one of the earliest
types of object which appear in Avar graves. However, among the
Avars they do not occur in the form of complete belt-sets. Possible
explanations for this may be either that the period during which
they were used was already more or less over and therefore the

34
F. Daim and A. Lippert, Das awarische Gräberfeld von Sommerein am Leithagebirge,
NÖ, Studien zur Archäologie der Awaren 1. Studien zur Ur- und Frühgeschichte
des Donau- und Ostalpenraumes 2. Denkschriften der Österreichischen Akademie
der Wissenschaften, philosophisch-historische Klasse 170 (Wien 1984) pl. 108,2,4,1;
108,2,5; 109,2,4,2; 109,2,6; 109,2,8; 109,2,10.
35
K. Bakay, “Az avarkor idorendjé
 ol.
 Újabb avar temetök a Balaton Környékén
[Zur Chronologie der Awarenzeit. Neue awarenzeitliche Gräberfelder in der Umgegend
des Plattensees]”, Somogyi Múzeumok Közleményei 1 (1973) pp. 1 ff., esp. p. 13, pl. IV.
36
F. Daim, “Das awarische Gräberfeld von Zillingtal: Sechs Gräber mit west-
lichen Gegenständen”, Wissenschaftliche Arbeiten aus dem Burgenland 100 (1998) pp.
97–135, esp. pl. 18 and 19.
37
Das awarenzeitlich gepidische Gräberfeld von Kölked—Feketekapu A, pl. 70 A 341,9.
38
L. Bende, “Tausírozott díszu övgarnitúra a pitvarosi avar temet ob  ol
 [Tauschierte
Gürtelgarnitur im awarischen Gräberfeld von Pitvaros]”, A Móra Ferenc Múzeum
Évkönyve—Studia Archaeologica 6 (2000) pp. 199–217 [German summary: pp. 209–11]
esp. fig. 5 ff.
39
Exhibition Museum Kaposvár 1999. Compare also the exhibition catalogue:
The largest cemetery from the Avar period in the Carpathian basin. Selection from the restored
material of the Avar cemetery at Zamárdi (Kaposvár 1998).
40
Daim, “Sechs Gräber mit westlichen Gegenständen”, pl. 2,10.
478  

Avars merely possessed a couple of “antiquities” or, alternatively, that

the main focus of the distribution of the belt-sets with mask fittings
lay elsewhere, for instance in the Caucasus region and along the
coast of the Black Sea. In the Early Avar Period, casting technol-
ogy was used mainly for buckles and for some parts of the fittings
of Bócsa-Kunbábony type (pl. 19–22). Exceptions are some cast
strap-ends with “toothcut”-decoration, for instance from Budapest-
Farkasvölgy, typically with a well developed spout.41 The first cast
subsidiary strap-ends appear in the Middle Avar Period. A little later,
in Late Avar Period I, complete belt-sets first occur. In the Early
Avar Period, sheet-fittings were, as a rule, attached by filling the
reverse with a non-precious, soft metal alloy (lead or lead alloys) and
casting small strips of wire in with it. This production process is now
being examined in the form of a research project. In the Middle
Avar Period one appears, in the majority of cases, to have glued
sheet-metal fittings onto the belt, as neither traces of lead, nor holes
for rivets or threads, nor soldered-on attachment lugs or cast attach-
ment lugs could be detected. We do not know what kinds of glue
were used, however, we may assume that the Avars had access to
a wide range of materials, for instance tree pitch, resins, fish- or skin
glues, possibly mixed with waxes. All of these are highly effective
natural substances of quite different specifications. The production
of a composite bow alone, which is attested by several hundreds of
finds, required the knowledge of excellent glues, as has already been
mentioned.
The metalsmith’s grave from Kunszentmárton, which is at the
moment being examined closely as part of a research project, is one
of the most fascinating archaeological deposits from the Carpathian
Basin in the Avar Period. The smith’s grave from Kunszentmárton
was found in 1928. It is the burial of a man equipped with horse,
lamellae armour and sword, possibly also with bow and arrow.42 At

41
M. Nagy, “Awarenzeitliche Gräberfelder im Stadtgebiet von Budapest”, Monumenta
Avarorum Archaeologica 2 (Budapest 1998) pl. 83 A 13–15; F. Daim, “‘Byzantinische’
Gürtelgarnituren des 8. Jahrhunderts”, Die Awaren am Rand der byzantinischen Welt.
Studien zu Diplomatie, Handel und Technologietransfer im Frühmittelalter, ed. id., Monographien
zur Frühgeschichte und Mittelalterarchäologie 7 (Innsbruck 2000) pp. 77–204, esp.
fig. 90.
42
D. Csallány, A kunszentmártoni avarkori ötvössír [Goldschmiedegrab aus der Awarenzeit
von Kunszentmárton] (Szentes 1933).
   () 479

the same time, it contained the most extensive collection of metal-

smith’s tools and other accessories, as well as half-products and raw
materials, known so far from early medieval Europe.
The deceased lay on his back. On his breast there was a set of
lamellae armour, which—according to the reconstruction attempted
by Csallány—consisted of iron parts sewn onto a linen fabric, arranged
in nine rows and overlapping each other.43 A sword with a single
cutting edge, possibly with p-shaped attachment plates, had been
placed beside the body.44 It is not known whether the smith wore
a three-, four- or a multi-part belt. Two iron strap-ends (?) appear
to have decorated the belt.45 “A small horse” had been buried above
the feet of the dead warrior. Unfortunately, only two foot bones of
the animal remain and for this reason, Vierck has claimed that it
cannot be excluded that this may in fact have been a partial horse
burial. The latter are known several times from the Early Avar
Period. In these cases the horse was skinned, however, the foot bones
and the skull remained in the skin. The horse skin was then arranged
in the grave in a way which seemed natural, usually bridled and
with saddle and stirrups. In this case, however, the animal wore nei-
ther bridle nor stirrups. The smith’s grave from Kunszentmárton
contained 130 objects. Apart from a small number of personal pos-
sessions (armour, sword, knife, belt components) these are imple-
ments/tools, other devices, a few half-finished products and raw
materials. We may, tentatively, reconstruct some of the technical pro-
cedures of which the man was capable. It is extremely likely that
he produced iron objects, re-cycling old metals in the process. The
numerous moulds (pl. 6) which the man had with him were used
for serial production of sheet-metal fittings by hammering them over
such a device. This indicates a process in which a piece of sheet-
metal was placed on a softer lead plate or a firm piece of leather
and the model was put on it upside-down. The mould was then
“pressed” on top of the sheet by means of hammering forcefully.

43
H. Vierck, Awarische Schmiedegräber [working title, unpublished manuscript,
Universitätsarchiv Münster] pp. 246 ff.; regarding the lamellae armour from Hajdúdorog,
which is the second complete specimen from an early Avar grave: D. Csallány, “A
hajdúdorogi avar mellpáncél [Der avarische Brustpanzer von Hajdúdorog]”, A
Debreceni Déri Múzeum Évkönyve 45 (1958/59) pp. 17–23 [German summary: p. 23].
44
Csallány, Kunszentmárton, pp. 8–9; Vierck, Awarische Schmiedegräber, p. 231.
45
Vierck, Awarische Schmiedegräber, pp. 221–2.
480  

Some pieces of raw material and some drops of cast metal suggest
that the smith from Kunszentmárton also cast bronze objects. He
was also able to do fine soldering work. The precision scale and a
larger number of Byzantine precision weights prove that he also dealt
with precious metal, presumably he also used the process of mer-
cury gilding.
The smith from Kunszentmárton had full access to Byzantine arte-
facts and was familiar with Byzantine traditions and techniques (as
is suggested by the precision scale and sets of weights, for instance.)
However, to call him a Byzantine travelling craftsman would be
going too far, because for a Roman citizen an equestrian burial
would be unthinkable.
Pottery constitutes the largest group of finds in Avar archaeology
by far. At the same time, pottery from the Early and Middle Avar
Period is—due to long years of research by Tivadar Vida—appar-
ently the most thoroughly examined type of find in Avar archaeol-
ogy.46 Vida has succeeded in distinguishing numerous groups and
variations and has been able to show local developments and influences
at supra-regional level. It will now be crucial to examine thin-sec-
tion samples and to analyse the heavy mineral content of at least a
small selection of the pottery. The pottery of the Early Avar Period
shows eastern influences in many respects: Hand-made pots with a
funnel-shaped or square mouth, with wart- or lump-shaped decora-
tion (pl. 18,3–4). Certainly, it is not feasible to ascribe all objects
which have parallels in Eastern Europe and Asia to the first Avar
generation which settled in the Carpathian Basin. We should not
underestimate the mobility of population groups between the
Carpathian Basin and the East. The Avars’ victories against Byzantium
attracted both adventurers and settlers, the wealth of the Avar elite
during this period attracted merchants and travelling craftsmen, while
political changes in the East (we know of some of them, but of many
we have no idea) frequently brought refugees into the Carpathian
Basin. We must therefore avoid any interpretations which are overly

46
T. Vida, “Zu einigen handgeformten frühawarischen Keramiktypen und ihren
östlichen Beziehungen”, Awarenforschungen, ed. F. Daim, Archaeologia Austriaca 1.
Studien zur Archäologie der Awaren 4 (Wien 1992) pp. 517–77; id., “Das
Töpferhandwerk in der Awarenzeit”, Reitervölker aus dem Osten. Hunnen + Awaren, ed.
F. Daim (Halbturn 1996) pp. 362–4; id., Die awarenzeitliche Keramik I (6.–7. Jh.),
Varia Archaeologica Hungarica 8 (Budapest 1999).
   () 481

schematic, and should by no means assume that there was a kind

of “settlement evolution”.
Craftsmen’s traditions of Late Antique origin may also be dis-
cerned in the pottery of the period. The so-called “grey pottery” is
fine ware of a high technical standard. It was made on the fast
wheel, mainly for local trade, in small workshops which were pro-
ducing for internal trade (pl. 18,2; 5–6). Gyula Rosner has managed
to locate at least one of the workshops producing “grey pottery”, in
the vicinity of Szekszárd and has subjected it to archaeological exam-
ination.47 In addition to local production, there are some “grey” ves-
sels, which may have been brought into the Avar territory from the
Byzantine Empire, for instance the amphora from Kunbábony.
The Avar defeat at Constantinople in 626 marks a turning-point
in the history of the khaghanate. The effects are noticeable in the
archaeological record, but without written sources we would not be
able to draw any conclusions from the archaeological evidence. While
up to 626, large sums of yearly tribute entered the Avar territory,
they now stop, as is illustrated by the numismatic analysis of the
coins found in this region.48 The ransoms for Byzantine prisoners
were another important source of income for the Avars. The looted
luxury goods must also have represented considerable wealth. After
626 it was necessary to adapt from a war economy: the Avars now
earned their livelihood by means of crop husbandry, but especially
by breeding livestock. Trade may also have played a certain role,
which, however, cannot be expressed in quantitative terms.
The fact that Avar sovereignty survived the crisis is, not without
reason, ascribed to the fact that their neighbours were so weak. A
number of princely graves, which presumably belong to the middle
of the seventh century and, at the latest, to the 670s, testify to this
phase of re-organisation of the Avar Empire. The earliest of these
are presumably the “princely graves” from Tepe, Bócsa and
Kunbábony, even if the chronological placement (even in terms of
relative chronology) of the finds of Bócsa-Kunbábony type is still

47
G. Rosner, “Keramikherstellung und Handel im Karpaten-Becken in der frühen
Awarenzeit”, A Wosinsky Mór Múzeum Évkönyve 15 (1989) pp. 125–33; Vida, Keramik
I, pp. 88–96; 177–81; 206–19.
48
P. Somogyi, Byzantinische Fundmünzen der Awarenzeit, Monographien zur Früh-
geschichte und Mittelalterarchäologie 5. Studien zur Archäologie der Awaren 8
(forthcoming).
482  

subject to heated discussions. In Tepe there is apparently loot in the

form of the largest Byzantine silver plate hitherto known. It had
been cut up.49 The gold mounts of the sword scabbard from Kuná-
gota appear to be the remains of a looted casket with gold fittings
(pl. 7–8).
The grave from Kunbábony, which has already been mentioned,
was discovered by chance in 1971 and is one of the most precious
archaeological deposits to have been discovered in Hungary in the
twentieth century (pl. 19–22).50 The deceased had been buried with
two valuable belts with gold fittings, two swords, a gold pot with a
handle and several wooden vessels with gold mounts. There was also
a huge Byzantine amphora, in addition to a number of other objects.
Apparently, pieces of gold foil had been sewn onto his funeral robe.
The most striking ensemble among the dress accessories and grave
goods from Kunbábony is the belt-set with “pseudo-buckles”, deco-
rative fittings in the form of buckles which were mounted vertically
onto the belt and which cannot be opened, thereby serving purely
decorative purposes (pl. 19).
The so-called “pseudo-buckles” are decorative fittings of pure gold
in the shape of buckles, sometimes with glass- or precious stone inlay,
which decorate some of the belts not only from Kunbábony but also
from Tépe and Bócsa. In recent years, they have frequently been
described as Byzantine products, mainly because they have been
made in the “Byzantine fashion”, using an intricate “composite tech-
nique”. One argument against this view is the fact that not a sin-
gle “pseudo-buckle” is known from a Byzantine context. There is
merely an ensemble from Sirmium, where decorative techniques with
substantial Byzantine analogies, such as the arcade border and a
complicated openwork lattice ornament, occur (pl. 23).51 On the other
hand, clear Christian symbolism is avoided on the belt-set from

49
É. Garam, Katalog der awarenzeitlichen Goldgegenstände und der Fundstücke aus den
Fürstengräbern im Ungarischen Nationalmuseum, Catalogi Musei Nationalis Hungarici, Seria
Archeologica 1 (Budapest 1993) pl. 94; Reitervölker aus dem Osten. Hunnen + Awaren,
ed. F. Daim (Halbturn 1996) p. 260, fig. 22.
50
E. Tóth and A. Horváth, Kunbábony. Das Grab eines Awarenkhagan (Kecskemét
1992); F. Daim, “Kunbábony”, Reallexikon der Germanischen Altertumskunde 17 (2nd edn.,
2001) pp. 490–5.
51
I. Popovic’, Zlatni avarski pojas iz okoline Sirmijuma (Golden Avarian belt from the vicin-
ity of Sirmium), National Museum, Belgrade, Monographies 10. Archaeological Institute,
Belgrade, Monographies 32 (1997).
   () 483

Sirmium—as opposed to belt-ornaments which were definitely used

in Byzantium. From this, we may deduce that this particular belt-
set was produced by a local (Byzantine) workshop for the Avar mar-
ket.52 But where did the strange “pseudo-buckle” in fact originate?
For now, this question can only be answered hypothetically. The
pseudo-buckle-fitting can only be derived from buckles which serve
purely decorative purposes on the belt. Objects of this type are in
fact known from Eastern Europe, at least from the Volga-Kama-
region, but their position in terms of absolute chronology is not yet
clear, so that it is not yet possible to submit irrefutable evidence for
this theory. Nevertheless, for now, the available data give the impres-
sion that, with the help of pseudo-buckles and Byzantine technol-
ogy, there was a desire to create a representation culture which, on
the other hand, consciously tries to avoid a pale imitation of Byzantine
customs.
Around the middle of the seventh century, an increased inflow of
Byzantine coins can once more be observed and the most recent
princely graves from Ozora-Tótipuszta, which belong to the Middle
Avar Period, contain almost exclusively Byzantine objects (pl. 28).
We also hear once more about a Byzantine delegation being sent
to the Avars. There are two possible explanations for these obser-
vations: either the Avar Empire is once more becoming attractive
to the Byzantine Empire as a political ally and trade partner, or
there is in fact a connection between the end of the Bulgarian Empire
in the Black Sea region between 641 and 668 and the migration of
Bulgarian groups, in an attempt to escape the Chazars. The Bulgarian
Empire in the lower Danube region was established by Asparuch at
this time. This was the beginning of a rather changeable relation-
ship with the Byzantines. While they were allied closely during some
periods, during others the Byzantines attempted repeatedly to destroy
the Bulgarian Empire. Regardless of this, it definitely made sense
that, in times of changing political topography, Byzantium appar-
ently tried to establish more intensive contacts with the Avars.
We may assume that, as a result of the dramatic events at
Constantinople in 626, radical changes must have taken place in
Avar settlement, as from then on, they must have been forced to
earn their living exclusively by means of agriculture. However, the

52
Similarly: ibid., p. 85.
484  

finds do not permit us to come to any conclusions regarding this

question. There are two reasons for this: Firstly, only two Avar set-
tlements have been published to date53 and secondly, in order to
enable us to attempt a historical interpretation of this scope, we
would require a very exact chronological system as a basis and this
will be impossible even in the near future. Pottery from the Early
and Middle Avar Period can, as Tivadar Vida frankly admits, be
dated within half a century at best.54
Due to a lack of finds which are relevant in terms of fine chronol-
ogy, settlements and settlement layers are more difficult to date.
Nevertheless, they can reveal much about the way of life of the pop-
ulation. Byzantine building traditions never gained much popularity
among the Avars, a point that can be compared with the observa-
tions Priskos made about Attila and his court. Since comparative
analysis of excavation results is hindered by geology (for instance,
there is over a metre of pitch black humus in Örmenykút) and vari-
ations in excavation technique, regional differences in settlement
structure are not yet clear, although we would definitely expect to
find such. From the Early and Middle Avar settlement Dunaújváros-
Öreghegy, several rectangular, sunken huts whose roof was supported
by a pair of posts at each of the narrow sides are known (pl. 17,1).55
The entrance is usually in the south-east, the stone oven is located
in the north-eastern corner. Furthermore, there are free-standing
ovens and numerous pits within the settlement. However, what is
presumably a storage pit may be observed next to many of the huts.
Several systems of ditches permeate the settlement. These ditches
overlap some of the huts, or, in turn, are overlapped by them, so
that István Bóna was able to conclude that the Avar village had two
phases.
In Lébény near Gyor,  the hitherto largest Avar settlement, con-
sisting of two sites with a total area of 13–14 hectares, was exam-
ined as a result of the construction of the Wien-Budapest motorway.56

53
I. Bóna, VII. sz-i avar települések és Arpád-kori magyar falu Dunaújvarosban [Awarische
Siedlungen aus dem VII. Jh. und arpadenzeitliches Dorf in Dunaújvaros], Fontes Archaeologici
Hungariae (Budapest 1973); C. Bálint, Die spätawarenzeitliche Siedlung von Eperjes, Varia
Archaeologica Hungarica 4 (Budapest 1991).
54
Vida, Keramik I, pp. 15–26.
55
Bóna, Dunaújváros.
56
M. Takács, “Die awarenzeitlichen Siedlungen von Lébény”, Reitervölker aus dem
Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp. 378–82.
   () 485

The earlier settlement of Lébény-Kaszás-domb also includes a sys-

tem of ditches which encloses rectangular areas. The latter may be
interpreted as pens for livestock (pl. 17,2). The two huts correspond
more or less to those from Dunaújváros, additionally, there are
numerous pits and three wells, one of them with a stone border and
a wooden well-box. The evaluation of the early Avar settlement at
Kölked-Feketekapu, which has been excavated by Attila Kiss and his
wife Zsuzsa K. Zoffmann, will be of utmost importance.
Unfortunately, it was possible to excavate only a tiny part of the
settlement at Zillingtal. Nevertheless, the results illuminate some
important aspects of Avar archaeology, three of which will be men-
tioned here: The first concerns the process of settlement in our region.
After it had been shown that Early Avar finds, almost without excep-
tion, occur on formerly Roman sites,57 the question arose whether
the Avars made use of existing Roman ruins. In Zillingtal, a Roman
house was uncovered within a settlement from the Avar Period. Avar
post holes and small ditches situated above the Roman building do
not respect it in any way. This shows that the Avars erected their
own huts above and beside the Roman buildings according to their
own ideas.58 The reason for settling in a particular place was pre-
sumably the fact that the area, which had been cleared of wood-
land in the Roman Period could still be used agriculturally and apart
from that, there was still an existing network of roads and paths,
which must also have facilitated traffic for the Avars. One should
keep in mind that the deciduous and the mixed forest are the nat-
ural types of vegetation in the Carpathian Basin. The fact that today,
parts of Hungary appear to belong to the Eurasian steppe region,
is in fact the result of human mismanagement in the Early Modern
Period.
In Zillingtal, numerous different types of pits have been excavated.
However, none of them permitted the reconstruction of a pit house
of the Dunaújváros type. In addition to this, we found rows of
wooden posts, which definitely belong to the Avar Period. The wooden

57
Daim et al., Leobersdorf, p. 175; H. Winter, Awarenzeitliche Grab- und Streufunde
aus Ostösterreich. Ein Beitrag zur Siedlungsgeschichte, Monographien zur Frühgeschichte
und Mittelalterarchäologie 4 (Innsbruck 1997) pp. 73–7.
58
F. Daim and A. Distelberger, “Die awarische Siedlung von Zillingtal—Die
Grabungen 1994–95”, Reitervölker aus dem Osten. Hunnen + Awaren, ed. F. Daim
(Halbturn 1996) pp. 372–7.
486  

posts had been carefully worked into a square shape. Here, some
above-ground buildings supported by wooden uprights, of a type
which so far had not been observed in Avar areas of settlement,
must have existed. Three furnaces for iron working, which have been
excavated carefully and are now being evaluated, are also particu-
larly interesting. Two of them appear to have been used for ore
reduction, the others for forging.
It is obviously of interest to ascertain the type of economy used,
however, this is not yet possible. Due to the animal bones, which
are in fact the remains of grave accompaniments in the form of
meat, we are familiar with the whole range of Avar animal breed-
ing. Game, on the other hand, hardly occurs at all. It will surely be
possible, in the near future, to identify precisely the cereals culti-
vated, the other crops and the produce from gardening, from the
plant remains found in settlements, and it seems that we will be in
for some surprises. At present, ethnographical analogies play a pre-
dominant role in interpretation; however, it is vital to exercise cau-
tion in using them. We must assume that each village in fact practised
its own version of a commonly accepted way of life, in the same
way that the burial customs differ in some details between settle-
ments.59 Presumably, it was especially important to the Avars to own
cattle. However, the relative proportion between breeding livestock,
farming and gardening, between the production of goods and ser-
vices (trade) will have been different in every population group. In
the process of interpreting the cemetery and the settlement from
Zillingtal, we will attempt to create economic models, starting with
a given population size and the maximum area of land available for
cultivation.
Hungarian researchers have dealt intensively with Avar burial cus-
toms, especially with the equestrian graves (pl. 3; 25; 37; 39). At
first, this research was based on the assumption that it would be
possible to link ethnic groups with particular customs. These hopes
have been destroyed; however, most types of equestrian graves—not
surprisingly—show links with Eastern Europe. There are many different
types of equestrian grave. Usually the saddled horse—this means
with saddle, stirrups and bridle—was buried beside the dead war-
rior in a reclining position. Occasionally, partial horse burials also

59
Daim et al., Leobersdorf, pp. 165–6 with notes 14 and 15.
   () 487

occur and as a rule these are east of the Tisza (pl. 3). This means
that the horse was skinned, whereas the skull and the foot bones
remained within the horse skin. The latter was then spread out or
rolled together and placed in the grave together with saddle, stir-
rups and bridle. In exceptional cases, a combination of the two types
seems to have occurred, for instance in Szegvár-Sapoldal. This bur-
ial was interpreted from a culture-historical point of view by István
Bóna.60 The documentation of this excavation, however, is restricted
to ground plans. It seems likely that this was in fact a tunnel grave.
For this type of burial, a tunnel was dug outwards from the shaft
of the grave at an oblique angle and the deceased then placed in
it. The grave pit itself contains the horse (in Szegvár-Sapoldal also
the second horse which had been skinned; pl. 3) as well as any other
animals which had been killed. Tunnel graves occur in eastern
Hungary from the Early up to the Late Avar Period.
We know smaller and larger row-grave cemeteries of late sixth to
early seventh century date, from Keszthely and from formerly Lombard
western Hungary (Környe, Kölked-Feketekapu A). One gets the
impression that the Avars adapted their burial customs gradually to
the “western” model in the course of the settlement process. Some
early Avar burial places consist of inhumation burials, which were
located at considerable distance from each other (Kunszentmárton,
Leobersdorf, Sommerein, Zillingtal, Börcs-Nagydomb).61 In the case
of Leobersdorf, it is possible to observe the gradual transition from
single-grave burial to row-grave cemetery. In this process, not only
the distance between the graves, but also their orientation, has changed
successively from NNW-SSE to W-E.

2 Middle Avar Period (650/670–710)

If we had only princely graves from the second half of the seventh
century—the inventories from Igar and Ozora-Tótipuszta—then pre-

60
I. Bóna, “A Szegvár-sápoldali lovassír. Adatok a korai avar temetkezési szokhá-
sokhoz [Das Reitergrab von Szegvár-Sápoldal. Beiträge zu den frühawarischen
Bestattungssitten]”, Archeologiai Értesíto 106,1 (1979) pp. 3–32; id., “Studien zum
frühawarischen Reitergrab von Szegvár”, Acta Archaeologica Hungarica Academiae Scientiarum
Hungaricae 32 (1980) pp. 31–95.
61
Tomka, “Hirten”.
488  

sumably the Middle Avar phase, the second group according to Ilona
Kovrig,62 would never have been identified. This is because the
archaeological material preserved in these graves is either linked
much too closely to the Early Avar material (Igar; pl. 24) or, alter-
natively, it consists almost exclusively of Byzantine objects (Ozora-
Tótipuszta; pl. 28). Important innovations during this period were
merely the sabre, which replaces the early Avar single-edge sword,
as well as a number of different types of jewellery. The sabre scab-
bard now has different types of attachment loops, for instance “don-
key-back-shaped” (Gyenesdiás 64: pl. 26,1, Ozora-Tótipuszta) or
D-shaped loops (Igar III: pl. 24,4). However, in the large necrop-
olises, the Middle Avar group may be discerned with remarkable
clarity. Moreover, in many cases it constitutes a large proportion of
the cemetery (Mödling-An der Goldenen Stiege; Zillingtal). In the
Middle Avar Period a certain homogenisation may be observed, a
standardisation of the archaeological material which makes imported
goods, from Byzantium, Italy or southern Germany, stand out with
unusual clarity. Today, this standardisation may appear to us as a
kind of impoverishment, nevertheless it suggests the existence of a
cultural power, which becomes noticeable once more at a later point,
in the process of adapting new stimuli from the South to suit its
own taste. At the same time, we may interpret this homogenisation
of Avar culture as the result of a political process, through which
all the khagan’s “people” were united. Under this system of alle-
giance, all the small groups which were united under Avar sover-
eignty, but nevertheless frequently became (temporarily) independent
under the leadership of their warlords, were absorbed by a society
which was divided—first and foremost—into horizontal strata. The
local cultural variations are clearly not so important.
The interlace- and chain-ornament of Igar type constitute the
Middle Avar version of Germanic Animal Style or its Pannonian
variant, the “toothcut”-decoration. The beginnings of such ambitions
are related to classical Byzantine interlace and can be seen on a
mould for sheet-metal strap-ends from Kunszentmárton. In the princely
grave Igar III, the strap-ends and fittings, which are made individ-
ually by chasing and punching sheet gold, are decorated with chain

62
I. Kovrig, Das awarenzeitliche Gräberfeld von Alattyán, Archaeologia Hungarica 40
(Budapest 1963).
   () 489

ornament, whereas the gold buckle is decorated with “toothcut”-dec-

oration (pl. 24,3). With its chain-ornamented belt-set made of sheet
silver and the strap-ends with triple-hemispherical end (pl. 29,1), the
relatively rich grave Mödling 35 links the equestrian grave from
Gyenesdiás (pl. 25,4) with the finds from Igar III. Apart from that,
ensembles without or with only simple geometric decoration also
occur. They may be derived from early Avar belt-sets of Szegvár-
Sapoldal type. The Late Roman, Hunnic and Germanic roots are
no longer apparent in the Middle Avar Period; mirrors, omega
brooches and belt pendants with sieve spoons no longer occur. There
is no doubt that the belt-sets of Byzantine type, especially the set
from Ozora-Tótipuszta, as well as other Byzantine jewellery, which
is remarkably abundant in the archaeological record of the Middle
Avar Period, as well as Byzantine coins, which once again, occur
more frequently in the Carpathian Basin, signify a change in the
overall situation of the Avar elite. It seems possible that the reno-
vation of the Avar Empire has, after one or two generations, reached
a stage where it was once more possible to practise active foreign
policy. At least, there is evidence of an Avar delegation to Constan-
tinople in the year 678/679.
The princely grave III from Igar and the equestrian grave from
Gyenesdiás 64, on which remarkably thorough research has been
done, may serve as examples of archaeological complexes from the
Middle Avar Period:
The three finds from Igar (Igar I, II, III) are among the most
well-known archaeological assemblages of the Middle Avar Period.
Nandór Fettich has dealt with them repeatedly, however, a precise
examination of the original documents and the inventory records
was only carried out a few years ago by Gyula Fülöp. He also
attempted to contact the descendants of the finder and succeeded in
clarifying a number of errors.63
The third find from Igar, which we would now like to look at
more closely (pl. 24), has given its name to a type of sheet-metal
belt-fittings with chain decoration (“Igar type”), which is character-

63
G. Fülöp, “Awarenzeitliche Fürstenfunde von Igar”, Acta Archaeologica Hungarica
Academiae Scientiarum Hungaricae 40 (1988) pp. 151–90; id., “New research on finds
of Avar chieftain-burials at Igar, Hungary”, From the Baltic to the Black Sea. Studies in
medieval archaeology, ed. D. Austin and L. Alcock, One world archaeology 18 (London
1990) pp. 138–46.
490  

istic for Middle Avar Period I. In 1928, the equestrian grave was
discovered in the process of removing a hill, at a depth of 4 m. As
the finders were unable to agree on how to divide up the valuable
objects, the police heard about the find and collected the gold fittings
from the villagers. According to the landowner, the bronze and gold
objects lay in the stomach region, the sword to the left of the body,
the silver cup near the left hand, the wooden vessel near the feet,
the coffin clamps around the skeleton. The man wore two small gold
earrings decorated with small spheres and two plait fasteners, con-
sisting of gilded bronze tubelets of hexagonal section (pl. 24,1–2).
The belt-set (pl. 24,3) made of sheet gold (for information on the
buckle see below) had not—as is usually stated—been pressed in a
mould. Instead it had been given individual form by chasing and
punching, whereby the basic lay-out of the ornament had first of all
been scratched on the front with a needle.64 The set is definitely not
complete, however, apart from the strap-loop and the hole-rein-
forcements all the different types seem to be present: Buckle, main
strap-end, one double escutcheon-shaped mount, one “double-bow-
shaped” and three escutcheon-shaped fittings as well as one broad
and four narrow subsidiary strap-ends. The gold buckle with inflexible
fitting has been cast and is decorated with a fine, symmetrical “tooth-
cut”-ornament. The “chain decoration” on the sheet fittings has obvi-
ously been created in the Middle Avar period by developing the
“toothcut”-ornament further. The latter is a Pannonian variation of
the Animal Style II in the form of interlace, in the “chain decora-
tion” the interlace is usually split up into separate chain links. The
silver cast and gilded hook-fitting may also have been part of the
belt. The sabre with “star-shaped” hilt is bent considerably and was
worn by means of two D-shaped attachment loops. The latter were
framed by strips of silver. If one considers the sabre from Gyenesdiás
for comparison (pl. 26,1), the two square gold fittings with face-
shaped protuberances in the corners could have decorated the attach-
ment loops of the sabre sheath. The handle was partially covered
in gold- and silver foil and has two bronze nails with decorative
function. The mouth of the scabbard is decorated with sections of
gold foil. The function of another, quite large Byzantine buckle (pl.
24,7) is not quite clear. Fülöp thinks there may have been a second

64
Birgit Bühler (forthcoming).
   () 491

belt, to which, apart from the buckle, the cast-fitting with a hook,
a broad strap-end made of low-quality silver and some fragments of
silver sheet may have belonged.65 Another possibility is that the
buckle, together with a small bronze buckle with inflexible fitting,
was used for fastening the straps belonging to the sabre, although it
seems a little too large for this purpose. In the case of both buck-
les, the lost bronze tongues had been replaced by ones made of iron.
The slightly flaring chalice made of low quality silver has fallen apart
during restoration; the pottery vessel is made of well-tempered clay
and finished on a slow wheel. It is well rounded and has a flaring,
rounded rim.66 There is also an iron sickle, which is unfortunately
heavily corroded, so that its shape can only be described vaguely.
Sickles occur only rarely.
In Gyenesdiás, just 2 km northeast of Keszthely, an Avar ceme-
tery was excavated between 1963 and 1991. Among other impor-
tant burials, it also includes a Middle Avar equestrian grave with a
Byzantine coin (pl. 25–27).67
The grave’s orientation was precisely E-W. The man lay facing
east, in a coffin which had been placed in a large, rectangular grave
pit over 2 m deep. On the southern side, beside the coffin, a horse
had been buried with the same orientation. A goat’s skin with com-
plete skull and legs had been spread out over the horse’s head and
forelegs. The warrior was outstandingly equipped. On both sides of
his skull lay gold earrings, on one of them an oval amethyst had
been preserved (pl. 25,1). There was a gold finger-ring on each hand,
one with a round cabochon on three small columns, the other with
pyramid-shaped decoration and a small red precious stone (pl. 25,2).
The belt-set consists of a Byzantine bronze buckle with inflexible
fitting, triple hemispherical openwork and three cast attachment lugs,
a variant of the Sucidava type of buckle, and gilded sheet-bronze
fittings and strap-ends (pl. 25,4). The basic shape of the strap-ends
resembles an elongated trapezium, the lower end is decorated with
three hemispheres. The fittings each consist of four hemispheres: in

65
Fülöp, “Awarenzeitliche Fürstenfunde”, pp. 165 ff.
66
Ibid., pp. 168–9.
67
R. Müller, “Vorbericht über die Freilegung des Grabes eines hohen Militärs
aus der Mittelawarenzeit in Gyenesdiás”, Communicationes Archaeologicae Hungariae (1989)
pp. 141–64; id., “Das Gräberfeld von Gyenesdiás”, Reitervölker aus dem Osten. Hunnen
+ Awaren, ed. F. Daim (Halbturn 1996) pp. 411–6.
492  

the centre there is a truncated cone which has been turned by 90

degrees. The strap-loop is decorated with a motif resembling a St.
Andrew’s cross. The latter may well reflect the—much earlier—cross
motif on the strap-loop from Kunágota. The belt-set is completed
by two hole-reinforcements and two fragments of fittings. Near the
right lower arm of the deceased, there was an iron dagger of excep-
tional length, approximately 20 cm. To the man’s left a sabre, with
an exceptionally long handle, had been placed (pl. 26,1). The attach-
ment loops are “donkey-back-shaped”, similar to those from Ozora-
Tótipuszta. They are ornamented with gold fittings in the shape of
a square which has been rotated by 90 degrees, again with a flat
truncated cone in the centre and four heart- or face-shaped protu-
berances. These fittings correspond closely to the chased gold fittings
from Igar III, which may therefore also be considered sabre orna-
ments. At the man’s left elbow there was a biconical spindle-whorl.
These are found predominantly in women’s graves. Near the man’s
feet there was a hand-made clay vessel. In the mouth of the deceased,
Róbert Müller found a so-called “coin of Charon”, which, accord-
ing to tradition, is the dead man’s payment for the ferryman who
is supposed to take him over the river Styx. The gold coin, fresh
from the mint, belongs to the reign of Constans II, minted between
654 and 659.
The horse was saddled and bridled. Its bridle was decorated with
fittings and strap-ends made of gilded sheet-bronze (pl. 26,2). In
some ways they seem very traditional (triple-hemispherical fringed
mounts), in some ways very modern (strap-ends in the shape of an
elongated trapeze, with star-shaped ends). The horse had an iron
snaffle-bit in its mouth, which must originally have had side-bars of
unknown form made of some kind of organic material. Beside the
horse’s flanks there were round stirrups. The quiver with the arrows
had been placed to the right of the horse, maybe it was hanging
from the saddle. It was apparently decorated with tin foil, which
however, fell apart in the excavation process. It contained four arrows
with triple-winged arrow-heads, one of them with a horizontal blade.
The composite bow with bone reinforcements in the centre and at
the ends lay to the left of the horse (pl. 27). Possibly, it had already
been broken when it was placed in the grave.
In Middle Avar Period II, there is a tendency to make the fittings
more geometric, for example by placing a square boss in the cen-
tre (pl. 29,3). While in the Early and Middle Avar Period, usually
   () 493

only buckles had been cast (exceptions can be found among the
group of belt-fittings with “toothcut”-decoration), now, at the end of
the Middle Avar Period, the first cast strap-ends and fittings appear,
usually only with geometric ornament and frequently together with
sheet-metal fittings. In the cemetery of Mödling, this phase, which
is a period of transition leading to the “griffin and tendril industry”,
is particularly clear (pl. 30,1–2). It would be useful to know when
exactly these innovations were introduced to Avar belt fashion. Usually,
we place the beginning of the Late Avar Period in the years around
or just after 700, however this is based more on intuition and con-
vention than on hard facts. If one would like to move the transition
between the Middle and the Late Avar Period further into the eighth
century, one would, in turn, have to move the material from Late
Avar Period III into the ninth century in order to prevent “over-
crowding” of chronological phases in the eighth century.
The fact that three types of iron belt-sets with silver inlay, which
otherwise occur mainly in southern Germany, but which could also
have originated in Italy, are found in Avar territory, is particularly
interesting: The first type is characterised by depictions of human
faces (“Feldmoching type”), the second by striped inlay and the third
by vertical stripes and zig-zag ornament:

• Iron belt-sets with silver inlay depicting human faces (“Feldmoching

type”): Eloszállás-Bajcsihegy;
 68
Zamárdi69 (third quarter of the sev-
enth century).
• Iron belt-sets with silver inlay in the form of stripes: Alattyán
520;70 Mödling-An der Goldenen Stiege 240;71 Pókaszepetk 67;72

68
Bóna, “Ethnische Verhältnisse”, pp. 64 ff. and fig. 4.
69
Exhibition Museum Kaposvár 1999. Compare also the exhibition catalogue:
The largest cemetery from the Avar period in the Carpathian basin. Selection from the restored
material of the Avar cemetery at Zamárdi (Kaposvár 1998).
70
Kovrig, Alattyán, pl. 34,47–51.
71
Mödling—Goldene Stiege, ed. F. Daim, K. Matzner and H. Schwammenhöfer
(forthcoming).
72
Á. Sós and Á. Salamon, Cemeteries of the Early Middle Ages (6th–9th c.) at Pókaszepetk
(Budapest 1995) pl. VII 67,5–12. The distribution of belt-sets with striped inlay was
examined by U. Koch, “Das fränkische Gräberfeld von Herbolzheim, Kreis Heilbronn”,
Fundberichte aus Baden-Württemberg 7 (1982) pp. 387–474, esp. p. 461, fig. 37 and 469,
as well as R. Marti, “Das Grab eines wohlhabenden Alamannen in Altdorf UR,
Pfarrkirche St. Martin”, Jahrbuch der Schweizerischen Gesellschaft für Ur- und Frühgeschichte
78 (1995) pp. 83–130, esp. pp. 109–10.
494  

Sommerein 1673 (third quarter of the seventh century) (pl. 16,5).

• Iron belt-sets with silver-inlayed zig-zag ornament: Gyenesdiás 5,74
Záhorská Bystrica 2075 und Zalakomár 54576 (third quarter of the
seventh century).

It is possible that these three types of belt-sets entered the Carpathian

Basin during a very short phase only. On the other hand, they may
be linked with a particular workshop or with a particular trade route.
Technical studies of the objects in question will help answer this
question.
The numerous large cemeteries which were used continuously in
the Middle and Late Avar Period now permit us to observe the cul-
tural development in the Avar Empire more clearly. However, the
reason why large necropolises were established at this particular
time—the end of the Early Avar Period or the beginning of the
Middle Avar Period, is difficult to determine. But certainly, the set-
tlements—inhabited for seven to nine generations—belonging to these
large cemeteries are proof of a tendency to settle down, even if the
Avars maintained many elements of traditional pastoral farming in
the process.
One of the “provincial” necropolises, which was used from the
end of the Early Avar Period up to the end of Late Avar Period is
the cemetery at Mödling-An der Goldenen Stiege.
Mödling is located south of Vienna on a slope of the Vienna
woods. In 1967, during dredging work in preparation for building
a house, Avar graves were found and then examined by the
Bezirksmuseumsverein of Mödling, the rescue excavation service of the
Urgeschichtliche Arbeitsgemeinschaft today known as Gesellschaft für Ur- und
Frühgeschichte and in particular by Hermann und Lotte Schwammen-
höfer. By 1973, a total of 497 Avar graves had been rescued.77 There

73
Daim and Lippert, Sommerein, pl. 15–16.
74
Müller, “Gräberfeld von Gyenesdiás”, p. 412.
75
L. Kraskovská, Slovansko-avarské pohrebisko pri Záhorskej Bystrici, Fontes archeolog-
ického ústavu Slovenského národného múzea v Bratislava 1 (Bratislava 1972) fig.
13; 14,1–2.
76
For an illustration see: R. Müller, Der Häuptling von Gyenes. Awarenzeit in der
Umgebung von Keszthely, Kataloge des Niederösterreichischen Landesmuseums N.F.
256 (Wien 1990) fig. 4.
77
Exhibition catalogue Germanen, Awaren, Slawen in Niederösterreich: das 1. Jahrtausend
nach Christus, ed. H. Windl (Wien 1977); K. Matzner, “Völkerwanderung und
   () 495

were also 55 cremations from the Hallstatt Period and 40 settlement

pits predominantly from the Neolithic Period. Another interesting
detail: there is also evidence of a ditch—presumably also from the
Neolithic—enclosing the area.
The cemetery from Mödling is not an example of a really rich
necropolis. It does not contain horse graves and only two graves are
spectacular: This is the man’s grave 35, from the Middle Avar Period,
with a silver belt-set of Igar type and the woman’s grave 144 from
the Late Avar Period with a more than complete repertoire of grave
goods, including the famous coat-clasp with images of two archers.
However, for two reasons the cemetery is particularly suitable for
archaeological interpretation: First of all, it has a relatively clear
structure. The central section belongs mainly to the Middle Avar
Period and starting from there, in the Late Avar Period, further buri-
als were added to the west and east. Only the northern part of the
cemetery was apparently used more than once; here however, some
of the overlapping graves assist interpretation.
As mentioned previously, the cemetery at Mödling is particularly
suitable for examining and illustrating the chronological development
of the archaeological material, especially in the Middle Avar Period
and during the transition from Middle to Late Avar Period. The
two most interesting assemblages of Middle Avar Period I come from
grave 35 and 93 (pl. 29,1,2). In grave 35, a man had been buried
with his silver sheet belt-set, decorated with chain ornament of Igar
type. He wore earrings consisting of silver rings and hollow gold
spheres decorated with granulation. Beside his head a spindle-whorl
had been placed, in the pelvic region there were bone ornaments
from the quiver and a triple-winged iron arrow head. In grave 93
a man with his bronze sheet belt-set, decorated with “wrapped” inter-
lace had been buried. On the square fittings, the interlace disinte-
grates once more to form separate chain links, which proves that
the two types of ornament are related. A vertical fitting with dou-
ble symmetry and decorated with a blue precious stone, is particu-
larly characteristic. It was worn on the back part of the belt. The
grave also contained two knives, an iron tool for making fire with

Awarenzeit”, Mödling. Landschaft, Kultur, Wirtschaft (Mödling 1975) pp. 67–88; H.

Schwammenhöfer, “Ein awarenzeitlicher Bestattungsplatz in Mödling bei Wien”,
Antike Welt 7,2 (1976) pp. 11–8.
496  

several flint stones, two arrow heads and a vessel hand-made from
clay containing small pieces of chalk or marble and finished on a
slow wheel.
Square belt-fittings with geometric decoration, which can no longer
be traced back to Animal Style ornament, are typical for Middle Avar
Period II. Grave 79 is a good example (pl. 29,3). Two square belt-
fittings which belong to the set are decorated with bosses in the form
of truncated cones with zig-zag ornament around them. A little later,
one began to cast belt-fittings with geometric decoration. A number
of graves contain mixed assemblages: parts of the belt consist of
sheet-metal, while others have been cast, for instance grave 242,
which also includes a pot finished on a slow wheel and decorated
with distinct wave-band ornament (pl. 30,1; 33,4). In grave 100,
there were square fittings with lattice decoration which match the
main strap-end as well as a cast fitting with a hook, which no longer
occurs in Late Avar Period I. The main strap-end from grave 242
has a real spout for fastening the belt-strap, while the one from grave
135 only has a gap, to which the tip of the strap is riveted by means
of a supplementary sheet of metal (pl. 30,2). Grave 135 also con-
tained a clay vessel which had been finished on a slow wheel, but
seems more archaic: It has an almost conic shape and a flaring rim
(pl. 33,2). The early women’s graves in Mödling are not very remark-
able. Bead necklaces with colourful biconical beads and “eye beads”
are characteristic of Middle Avar Period I, while the earrings are
frequently decorated with hollow spheres made of sheet-bronze, for
instance those depicted here from grave 54, which however, cannot
compare to those splendid specimens found in the contemporary
“princely graves”.
As far as pottery is concerned, changes can be detected not only
in the fine ware (“grey pottery”) but increasingly also in the simple
household pottery, made with the help of technical devices. Some
of the pots from the cemetery Mödling-An der Goldenen Stiege
which were finished by turning them on a slow wheel can be dated
with certainty to Middle Avar Period I (pl. 33,1).78 These early ves-
sels have a gentle, S-shaped profile and are made of clay which has

78
F. Daim, “Zur nachgedrehten Keramik aus dem awarischen Gräberfeld von
Mödling—An der goldenen Stiege”, Slawische Keramik in Mitteleuropa vom 8. bis 11.
Jahrhundert, ed. ’. StaÏna, Internationale Tagungen in Mikulcice 1 (Brno 1994) pp.
29–52.
   () 497

been tempered with small chalk stones, possibly even with ground
marble. This technique was already used in Late Antiquity; it seems
to have survived locally.79 Those vessels which are characteristic of
the Late Avar Period but still occur in the later Early- and Later
Middle Ages, have been finished on a slow wheel and have an angu-
lar rim (pl. 33,6–7). They may in fact be traced back to the Middle
Avar pots mentioned above. Regarding the Late Avar Period in
Mödling see below.
Some Hungarian archaeologists have linked the development of
the Middle Avar spectrum of finds with the immigration of a Bulgarian
population group led by Kuver, who is believed to have been the
son of the Bulgarian ruler. For methodical and thematic reasons,
this theory must be refuted, even if, for demographic reasons we
must assume that there was considerable immigration in the middle
and the second half of the seventh century, presumably also from
Eastern Europe.

3 Late Avar Period

In the Early and Middle Avar Period, metal-casting technology was
well known, however it was used mainly to produce buckles. The
majority of Avar belt components was either pressed or chased. In
the Mediterranean region, both techniques were used at the same
time, however, there appear to have been regional preferences. Only
in Northern Italy do we find belt-sets which consist exclusively of
cast components (second and third quarter of the seventh century,
at least; pl. 16,1). Similar sets, as well as cast buckles with U-shaped
fittings also find their way into the Carpathian Basin.80 Among the
Early Avar belt-sets with “toothcut”-decoration, there are also some
cast specimens of high technical quality.81
Within a relatively short period—if we allow that the Late Avar
period begins with this important innovation—casting becomes the
preferred technique for the production of buckles, strap-ends and
fittings (fig. 4). At the beginning of this period, for a considerable

79
Vida, Keramik I. I would also like to thank Roman Sauer (Vienna) for infor-
mation on this topic.
80
Bóna, “Ethnische Verhältnisse”. Most recently: Daim, “Sechs Gräber mit west-
lichen Gegenständen”, with further literature.
81
Daim, “‘Byzantinische’ Gürtelgarnituren”, pp. 160 ff.
 498
 
Fig. 3: A belt from the Early Avar Period decorated with fittings (Budakalász – Dunapart, Hungary, grave 696.
According to T. Vida)
   () 499

amount of time, the juxtaposition of thin, predominantly square sheet

fittings and cast products, which are frequently decorated with a very
similar geometric ornament, may be observed. It is apparent from
many details that an effort was made to adapt to the new fashion.
For instance, thick, square sheets, which are strong enough to sup-
port attached rings and large decorative rivets, are common. We
also find strap-ends which do not, as yet, have a spout but instead
are riveted onto the belt with the help of a rectangular piece of
sheet (fig. 5). To cast a spout a three-part mould would be needed,
whereas the third part must have been either an iron wedge, which
was removed from the strap-end after casting or, alternatively, a clay
wedge, which was broken up later.
It may seem surprising that by no means all details of the cast-
ing process are known. They would definitely merit further research
and archaeological experiments. Nevertheless, some substantial work
on Avar casting technology has already been done.82
Not only the period of transition from Middle- to Late Avar Period,
but also the various Late Avar phases, are well represented in the
cemetery from Mödling, which has already been mentioned. Grave
140 is a good example, containing a belt-set consisting of more than
40 parts. The main decorative element on both the strap-ends and
the fittings is a frame with scale ornament, sometimes with addi-
tional engraved or punched decoration. However, the most impres-
sive grave from Mödling is the already famous woman’s grave 144.
The coat-fastener found in this grave is decorated with the images
of two kneeling archers. Both parts consist of thick bronze sheet,
engraved and punched, and finally mercury-gilded. The woman was
lavishly equipped with a pair of silver- and of gold-earrings, a neck-
ring made of bronze wire, a rich bead necklace, consisting mainly
of large black melon- and pumpkin seed beads, two band-like bracelets,
spiral finger-rings made of bronze wire and a Roman glass bottle.
Significantly, the clay vessel had been made by hand, which seems

82
For example H.-J. Hundt, “Textilreste aus awarischen Gräbern von Leobers-
dorf und ein Exkurs über gegossene Textilstrukturen an der Rückseite bronzener
Riemenzungen”, F. Daim et al., Das awarische Gräberfeld von Leobersdorf, Niederösterreich
2, Studien zur Archäologie der Awaren 3. Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 194. Veröffentlichungen
der Kommission für Frühmittelalterforschung 10 (Wien 1987) pp. 9–18.
 500
 
Fig. 4: Typochronology of Late Avar belt ornaments based on the
Leobersdorf cemetery
   () 501

Fig. 5: The structure of Avar main strap-ends in the eighth century

to prove that the almost industrial production of the majority of Late

Avar pots on a slow wheel (pottery finished on a slow wheel) was
not perceived as increasing their value.
Let us now return to the male means of representation, the belt-
set. Apart from the lattice ornament mentioned previously, rectan-
gular fittings decorated with the griffin-motif are characteristic of
Late Avar Period I, together with main strap-ends divided into sev-
eral panels and decorated with figural imagery, “trees of life” and
many other motifs. In Late Avar Period II, the animal-combat scene
is the most characteristic motif. Two griffins attack a stag or doe,
whereby the animals are arranged in such a way that they resem-
ble the rhythm of a large-leaved scroll with circular lobes (e.g. pl.
36,8). This type of depiction is totally foreign to the principles of
classical art. Although the animal-combat scene—griffins against
deer—like the griffin by itself, belongs to the group of motifs which
had been adopted into the Avar repertoire from images in Byzantine
art, the animal-combat scene which was so popular in Late Avar
Period II must have been created by Avar artists. It reveals that they
were familiar with the concept of multi-dimensional motifs.83 The
rectangular fittings, which are typical for Late Avar Period I, are
now replaced by hinged fittings of escutcheon-like, pentagonal or

83
F. Daim, “Der awarische Greif und die byzantinische Antike”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 273–304.
502  

more or less round shape, usually decorated with simple plant-orna-

ment. In Late Avar Period III, there are some important innova-
tions. The belt-sets tend to have a high number of parts, some consist
of more than 30 fittings. The individual, larger hinged fittings are
sometimes replaced by groups of three tongue-shaped fittings. Abstract
plant-ornament prevails, especially the circular-lobe ornament. The
griffin-motif, which was so popular in the two earlier phases, disap-
pears almost completely. However, round decorative harness-mounts
and griffin-head-shaped (eagle heads with ears; pl. 37,8) fittings, which
were fixed on straps of horse harness, are one important exception.
Instead, strap-ends with circus scenes (pl. 37,5), with medallions and
animal motifs, as well as with stylized plant and tendril ornament,
the so-called “lily decoration”, now appear.
Some important changes also take place in casting technology (fig.
4): The spout for fastening the belt is now used exclusively in con-
junction with appendages, frequently in the shape of animal heads,
which enclose the rivets for fastening the end of the strap. The rel-
atively thick strap-end is often cast in one piece, leaving a slit at the
end between the two protuberances. In many cases, this slit not only
looks as if it had been sawn, but actually shows definite tool marks
which suggest the use of a fine saw. However, it seems to have been
most common to cast the strap-ends in two parts, whereby a gap,
in which the belt may be placed and fastened with rivets, remains
between the two halves. The process of casting such a strap-end is
very simple, because no wedge is needed for the ferrule. At first
glance, the two types look strikingly similar, as both usually have
parallel sides, a semi-circular end and protuberances for riveting the
belt strap.
Surprisingly, the analysis of the cast belt-sets from Leobersdorf
revealed that the belt-sets from Late Avar Period III, especially those
with strap-ends consisting of two parts, were cast using a different
alloy than those from earlier phases, namely so-called “bell bronze”,
which contains a high proportion of lead at the expense of copper
content.84 In the earlier periods, almost exclusively pure tin bronze
was used. However, it will be necessary to verify whether we may
in fact generalise from this result.

84
P. Wobrauschek, W. Haider and C. Streli, “Röntgenfluoreszenzanalyse von
Bronzefunden aus der Awarenzeit”, F. Daim et al., Das awarische Gräberfeld von
   () 503

Belt-fittings which were cast from pure precious metal—gold or

silver—occur in the Late Avar Carpathian Basin only in exceptional
cases. The half-palmette ornament on the silver belt-fitting from
Komárno-ul. Váradiho, suggests that it may have been an import
from the Mediterranean region.85 On the other hand, in the Late
Avar Period, there are a large number of mercury-gilded belt-fittings.
The latter have been examined by Gábor Kiss.86 There seems to
have been a large variety of motifs and ornaments which in some
cases have been adopted from the Mediterranean region or even
from the Far East. Evidently, the gilded belt-sets seem to have pos-
sessed an exemplary function for the costume of the less wealthy pop-
ulation (examples from Komárno-Schiffswerft: pl. 37,5–8) However,
not all the “foreign” motifs were imitated, some occur only on the
group of gilded fittings. Characteristically, the distribution map shows
a concentration of gilded belt-decoration on the fringes of the area
of Avar settlement. As not a single (!) grave is known from the Late
Avar Period which could compare with the wealth of the princely
graves from the seventh century, we should consider that perhaps,
this displacement of rich finds to the periphery may also signify a
shift of power away from the centre. However, it would also be pos-
sible that the leading social stratum adhered to a different type of
burial rite. This may even have meant lavish burial rituals, but with-
out grave accompaniments, in analogy to customs practised in the
West and the Mediterranean region at the same time. If—as in

Leobersdorf, Niederösterreich 2, Studien zur Archäologie der Awaren 3. Denkschriften

der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
194. Veröffentlichungen der Kommission für Frühmittelalterforschung 10 (Wien
1987) pp. 45–55, esp. p. 55, fig. 8. A large number of Avar finds were examined
by László Költö using X-ray fluorescence analysis: L. Költö, “Avar kori bronztár-
gyak röntgenemissziós analízise [X-ray emission analysis of bronze objects from the
Avar age]”, Somogyi Múzeumok Közleményei 5 (1982) pp. 5–68 [English summary:
p. 66]. The following collection of essays demonstrates Hungarian researchers’ inter-
est in archaeometry: Archaeometrical Research in Hungary, ed. M. Járó and L. Költö
(Budapest 1988). X-ray fluorescence analysis was also used to examine “Avar” finds
from Bohemia and Moravia, as suggested by Nad’a Profantová: J. Frána and A.
Ma“talka, “Röntgenfluoreszenzanalyse von frühmittelalterlichen Bronzen aus Böhmen
und Mähren”, Awarenforschungen, ed. F. Daim, Archaeologia Austriaca 1. Studien zur
Archäologie der Awaren 4 (Wien 1992) pp. 779–801.
85
Daim, “‘Byzantinische’ Gürtelgarnituren”, p. 106, fig. 23.
86
G. Kiss, “A késoi  avar aranyozott övdíszek [Die spätawarischen vergoldeten
Gürtelbeschläge]”, Somogyi Múzeumok Közleményei 11 (1995) pp. 99–126 [German sum-
mary: p. 106].
504  

Leobersdorf—at the end of Late Avar Period III, the “classical” cus-
tom of grave accompaniments was given up step by step, this devel-
opment may in fact have been initiated by the Avar elite two or
three generations earlier.
The relative chronology of Late Avar belt-ornaments stands on a
firm methodological basis, as the individual elements regarding shape
and decoration correlate well with technical details. For instance,
one would never find a main strap-end decorated with the typical
animal-combat scene, “two griffins killing a stag or doe” cast in two
parts or with appendages in the shape of animal heads, presumably
because at the time when strap-ends cast in two parts and with
appendages became fashionable, animal-combat scenes were no longer
in vogue. If we assume that certain motifs, shapes and techniques
occurred in the same period, all kinds of different combinations of
the individual elements should occur.
An almost confusing variety of motifs, styles and workshop tradi-
tions is represented in the Late Avar casting industry. Provided they
are recorded and interpreted adequately, they may enable us to grasp
the numerous cultural trends in the eastern part of Central Europe.
For a long time, our perception of this problem was hindered by
the belief that in the eighth century, the Avars lived in forced iso-
lation, partly because after their defeat in 626 they were dependent
on their own resources and partly because the direct route to
Constantinople had been blocked by the foundation of the Bulgarian
empire. However, the sources suggest that the situation was in fact
quite different.
The methodological problem regarding the study of Late Avar
motifs is:

• There is hardly any comparative material from Byzantium, Italy

and western Europe north of the Alps, because dress-ornaments
are present in burials only in exceptional cases.
• Persistent difficulties concerning the fine chronology of the eastern
European archaeological material.

A possible methodological “lever” for decoding Late Avar motifs

could be:

• Comparing Avar motifs with contemporary examples as well as

earlier and later examples from the same area and adjacent regions.
   () 505

• Analysis of motif context (which other motifs frequently occur in

conjunction with the one in question?).
• Taking into consideration the style used in each case.

In this way we should be able to grasp synchronic and diachronic

motif transfer by examining one complex of motifs after another and
thus be able to detect both adaptions to personal taste as well as
changes of meaning.
The fact that our research is—inevitably—limited to metal objects
is something of a hindrance, because even the small number of bone
ornaments from the late Early- and early Middle Avar Period demon-
strate that certain motifs and types of decoration may have survived
on perishable materials and then appeared on belt-sets only much
later. We would like to give three examples for Late Avar motifs,
the griffin, the flowered scroll and the boar.
The griffin, a lion with wings and an eagle’s head, predominantly
with pointed ears and a long tail, is a motif which is thousands of
years old. It is well known from Oriental, Egyptian, Greek, Roman
and Byzantine art. As a symbol of strength, it is often linked with
power. In Roman culture in particular it also plays an important
role in a sepulchral context and also enters Christian iconography.
In Avar art it first appears in the Middle Avar Period, when Byzantine
buckles may have served as models. Nevertheless, many other pos-
sibilities exist regarding the transfer of this motif. In Late Avar Period
I and II, the griffin almost dominated the cast belts (e.g. pl. 36), so
that he has been even described as the “heraldic animal” of the
Avars. Such an interpretation, however, signifies a lack of familiar-
ity with the chronology of the archaeological material: How can
something be a heraldic animal if it was a popular motif on belt-
fittings for only 60 out of altogether 250 years of Avar presence? In
Late Avar Period II, apart from the simple griffin-fittings, on which
an individual animal is depicted in a dynamic position—ready to
pounce—there is now also the main strap-end decorated with the
characteristic animal-combat scene, where two griffins attack a stag
or doe. Interestingly enough, the group of animals is not depicted
according to the classical tradition, but in a strange, intertwined man-
ner, so that its “rhythm” resembles that of a scroll with leaves. The
latter is frequently present on the reverse of the strap-end. Here,
the Avars have depicted an originally Mediterranean motif—the
animal-combat scene—in a way which corresponds to their own
506  

artistic tradition of using a sometimes very high degree of abstrac-

tion. In Late Avar Period III, the griffin hardly ever occurs. One
exception is the equestrian grave 36 from Komárno, where the round
harness-mounts and some smaller fittings are in the shape of eagle’s
heads with ears, thus depicting griffin’s heads (pl. 37,8).
Already in the Early Avar Period, the boar’s head is found as a
decorative detail on some belt-mounts and—for example—on the
bowl rim fittings from Kölked-Feketekapu B, grave 119 (pl. 14). From
the end of the Middle Avar Period onwards, boar-shaped strap-ends
occur, which, in Late Avar Period III, seem to have replaced the
griffin as a symbol of power. Like the griffin, the boar is also a tran-
scultural symbol, however, it is not used much in Christian iconog-
raphy and if so, predominantly in a negative context.87 The boar
signifies untamed strength, wildness and vigour. It is a widespread
motif among the Romans and the Celts, among the Anglo-Saxons
and in the Germanic North it is used to decorate helmets. So far,
no connection has been identified between the Germanic boar images
of the sixth and the early seventh century and those from the Middle
and Late Avar Period, so that—for lack of good alternatives—a
Central Asian origin has been suggested for the latter.88 The boar’s
head on the rim fittings belonging to a (wooden) vessel from Kölked
is strikingly similar to some belt-fittings from the late eighth century,
for example those from Békés-Vizesbánó (eastern Hungary) and from
Holiare 778 (Slovakia),89 so that we must now assume that there was
in fact a direct link. Possibly, the boar’s head in its “canonised” form
was used on organic materials in the seventh and the first half of
the eighth century, before it was used to decorate belts. Presumably,
we can exclude the possibility that an Avar used an “antique” object,
which had been found or transmitted accidentally, as the starting
point for a new belt fashion.
The reverse of one of the strap-ends in the shape of a boar’s head,
from Edelstal (Hungarian: Nemesvölgy, present-day Burgenland,

87
For an extensive summary see: S. Müller, Awarische Eber. Ein Beitrag zur früh-
mittelalerlichen Ikonographie (unpublished Proseminararbeit at the Institut für Ur- und
Frühgeschichte der Universität Wien, 2001).
88
É. Garam, “A bocsi kes oavarkori
 lelet és kore [Der spätawarenzeitliche Fund
von Bocs und sein Kreis]”, Archeologiai Értesíto 108 (1981) pp. 34–50, esp. p. 50.
89
Ibid., fig 4,1–2; 5,1.
   () 507

Austria), grave 22, is decorated with a scroll motif with flowers. The
ornament, which in this case is not very clear, may best be seen on
the round harness-mounts from Komárno-Schiffswerft 149 (pl. 38,1).
A number of the stems, which overlap in some cases, have calices
depicted in section, on top of which there are semi-circular petals,
predominantly three in number. What distinguishes this decoration
from most others of Late Avar date is a stylistic element, namely
the treatment of the surface or space: the scrolls are spread out
loosely over the large surface, giving the impression of generosity, of
unobtrusive elegance. However, the other Avar products seem to fol-
low the principle of completely covering all the available surfaces,
similar to the western artistic tradition. Both this particular type of
flowered scroll and the same generous treatment of space are also
present in Chinese metalwork from the contemporary T’ang-dynasty.
The latter, however, was influenced strongly by neighbouring cul-
tures. There are many early medieval cultures which we do not yet
know sufficiently well in order to be able to come to a final opin-
ion. In this case, however, it is apparent that there was a connec-
tion with Avar workshops. A possible stimulus could have been the
Chinese silver cup which found its way into the Carpathian Basin
in the Middle Avar Period, together with Byzantine luxury goods,
or printed silk which was decorated with similar motifs.90 The ear-
liest Avar belt-set which is decorated with this type of flower orna-
ment comes from Kiskörös, grave IX,91 and belongs to the Middle
Avar Period (third quarter of the seventh century).
An important result of research carried out recently was the
identification of genuinely Mediterranean belt-ornaments from the
eighth century (pl. 34). The main types of multi-part belt-sets from
the late sixth and the seventh century whose origin lay within Byzantine
culture have already been identified and described some time ago.92
This is due to self-evident find circumstances, for instance the
treasures from Akalan, Mersin and the finds from the Byzantine

90
Daim, “‘Byzantinische’ Gürtelgarnituren”, pp. 130 ff.
91
G. László, Études Archeologiques sur l´histoire de la Société des Avars, Archaeologia
Hungarica 34 (Budapest 1955) p. 30, fig. 9a and pl. VIII,1. I would like to thank
Béla M. Szoke
 for drawing my attention to this piece.
92
Werner, “Nomadische Gürtel”.
508  

stronghold at Sadovec (Bulgaria), but also to some of the motifs used,

such as monograms und Christian symbols of salvation. For the
eighth century the situation is not quite so clear and a combination
of methods was needed to be able to identify Mediterranean origi-
nals in Avar archaeological material outside Avar territory.93 This is
because among the objects from over 50,000 Avar graves dating to
the eighth century, there are numerous cast belt-sets with obviously
Mediterranean motifs: circus scenes, griffins, marine spirits riding on
dolphins, imperial portraits, vine scrolls and many more, most of
which were, without doubt, produced locally.94 Apart from the sta-
tistical probability as a result of the distribution pattern observed,
this is also suggested by a number of failed castings and semi-finished
products found in Avar settlements. However, it is not yet clear from
which models the Avars had borrowed (and adapted to their own
tastes) the classical motifs. Due to the lack of suitable criteria, Byzantine
originals have—to date—not been recognised as such.
Recently, such a belt-set—from Hohenberg, Styria—has been
identified as an Italian-Byzantine product (pl. 34,1). Its method of
production differs from that of the vast majority of Avar parallels.
Moreover, two good parallels have been found recently in Bolzano,
and yet another comes from Biskupija (Croatia). A fresco in the
church S. Maria Antiqua in Rome (middle of the eighth century)
depicts the donator of the north-western side-chapel, Theodotus,
wearing a multi-part belt-set of this or similar type. As Theodotus was
supreme commander of the imperial troops in Rome, he is presum-
ably shown wearing the official dress of the dux (magister militum). The
chapel and with it also the fresco was made between 741 and 752,
which fits in quite well with the date established by means of archae-
ological criteria for the belt-sets of “Hohenberg type”: the third quar-
ter of the eighth century. Written sources from the early tenth century
refer repeatedly to precious belts as diplomatic gifts. There are good
reasons to suspect that this is valid also for the eighth and ninth
century. On the whole, decorated belts appear to have played an
important role both in the representation of high-ranking civilian

93
Daim, “‘Byzantinische’ Gürtelgarnituren”.
94
The most comprehensive work—to date—on the motifs from Late Avar belt
decoration is J. Dekan, “Herkunft und Ethnizität der gegossenen Bronzeindustrie
des VIII. Jahrhunderts”, Slovenská Archeológia 20,2 (1972) pp. 317–452, although it
does not take into consideration style and provenance.
   () 509

officials and military officers, as well as in diplomatic relations, and

certainly also in trade.
The problem was now to find a way to identify the Mediterranean
products—which doubtlessly existed—from among the thousands of
Avar buckles, strap-ends and mounts. Trials using a combination of
methods have now yielded some first results:

1. The types which are to be recorded on the map are defined so

that their main distribution lies outside the area of Avar settle-
ment, while their representatives only occur in the Carpathian
Basin in exceptional cases.
2. There should be technical differences between the Mediterranean
originals and the vast majority of parallel finds from the Carpathian
Basin (e.g. with respect to casting technology, fittings made in the
composite technique, soldering technology, granulation, the pro-
duction of beaded wire and gilding).
3. Iconographical differences may also play a role. The Avars select
from the Byzantine repertoire of motifs what appeals to them
and/or corresponds to their code of imagery. At the very least,
they do not choose motifs which have negative connotations or
are “taboo” in their culture. Half-palmette ornament, which is
very popular among the Byzantines, occurs rarely in Avar art.
Images of birds are, as a rule, not adopted and are, if necessary,
replaced by other motifs; frequently the style of the depiction has
been altered.

The higher the number of these criteria which is fulfilled by any

particular type of belt-set or fitting, the higher the probability that
a Mediterranean product has in fact been recognised.
Some of the objects identified so far may be dated with some cer-
tainty with the help of associated finds, for instance the mounts from
Vrap, Hohenberg and Brestovac. For others, it is necessary to resort
to auxiliary constructs. As the formal development of Avar strap-
ends and belt-fittings corresponds, by and large, to that of the
Mediterranean models, it should now be possible to locate chrono-
logically individual objects of uncertain date with the help of tech-
nical as well as formal details, such as the spout, appendices for the
rivets and casting in one or two parts. Additionally, strong tenden-
cies towards scrollwork and figural decoration seem to have alter-
nated with each other.
510  

Apart from the belt-set from Hohenberg (middle of the eighth

century), there is also evidence for the Mediterranean origin of a
number of finds and even entire complexes of finds. The cast gold
belt-fittings with scroll- and circular lobe-ornament from the Vrap
treasure (Albania) may be dated to the initial third of the eighth
century. At least part of this find must have originated in a gold-
smith’s workshop; the rest consists of valuable tableware, some of
which may have belonged to the Church. Buckles, strap-ends and
mounts with elongated scrolls (Bändersträuße) which occur frequently
in the Crimea and in the Volga-Kama-region, belong to the same
period. In Austria there is one specimen from Micheldorf, Upper
Austria, in Slovakia there are two, from ’ataj and Holiare.
The fragment of a main strap-end made of gilded silver from
Mikul‘ice, Southern Moravia (pl. 34,2), may be dated to the mid-
dle of the eighth century. The same goes for the main strap-end
from Aleppo, Syria (pl. 34,3), which consists of four parts. Both these
strap-ends are decorated with images of birds, whereas the Avar imi-
tations depict quadrupeds.
A hinged fitting made of gilded silver from Weiden am See,
Burgenland, was produced in the second half or the final third of
the eighth century (pl. 34,4). It was put together from more than
70 individual parts. The gold belt-set from the Brestovac treasure
(Croatia) belongs to the same period. It will surely be possible to
identify further types of Mediterranean origin in the near future.
Although the production of cast and sheet belt-ornaments in the
Mediterranean region stems from the same roots, some regions with
common forms of dress or local particularities may already be dis-
cerned, for instance in Italy, on the Balkans and in the Black Sea
region.
The distribution of the Mediterranean originals known to date
indicates a significant agglomeration in the new Slav territories,
among the Carantanians, in the vicinity of the “Slawendekanie” of
Kremsmünster, in Southern Moravia, as well as in the adjacent Avar
border region. Further concentrations may be observed in the upper
Tisza-region and in the area around Keszthely, at the western end
of Lake Balaton. Of course, the objects might have reached the
places where they were found as regular trade goods, but it is also
possible that they came with Byzantine delegations which were inter-
ested in establishing contact in this region, in places where—as was
   () 511

clearly visible from afar—centres of power were in the process of

formation. Some of them were unable to survive in the long run,
but about two generations later, the differentation of society in
southern Moravia resulted in the establishment of centres of power
in the “Great Moravian Empire”. These diplomatic missions had to
bring with them gifts, which were finely graded according to value:
silk and other objects which have not survived, as well as the belts
as part of official dress which have survived in settlement strata and
graves.
Although only a small number (at least!) of Mediterranean origi-
nals are known to have been found on Avar territory, strong
Mediterranean influences are visible in the decoration of the belt-
sets and in women’s jewellery. The burial customs, however, show
a high degree of constancy, in which local traditions, apparently,
played an important role. The equestrian grave is still as popular as
ever, and in exceptional cases, women are also buried with horses.
On the fringes of the Late Avar Empire, in the same place that
agglomerations of cast gilded bronze fittings occur, we may also
observe a significantly higher proportion of equestrian graves. It seems
as though the border region had gained greater weight within the
Avar Empire. It will not be easy to determine whether this should
be attributed to the profit from small-scale frontier traffic, to promi-
nent forms of representation pertaining to equestrian warriors, or
alternatively, to the ideological superstructure of groups of border
guards. The cemeteries in the area of the city of Komárno are par-
ticularly impressive in this respect. In these cemeteries, the propor-
tion of equestrian graves is exceptionally high. In the largest necropolis
alone, Komárno-Schiffswerft, 63 out of a total of 153 graves are
equestrian burials.95 The graves from Komárno-Schiffswerft demon-
strate in an impressive manner the variety and partly also the qual-
ity of Late Avar bronze objects.
As in the earlier periods, Late Avar pottery shows numerous
regional types and preferences.96 Generally speaking, the proportion

95
A. Trugly, “Gräberfeld aus der Zeit des awarischen Reiches bei der Schiffswerft
in Komárno I”, Slovenská Archeológia 35 (1987) pp. 251–344; id., “Gräberfeld aus der
Zeit des awarischen Reiches bei der Schiffswerft in Komárno II”, Slovenská Archeológia
41 (1993) pp. 191–307; F. Daim, “Komárno/Komárom”, Reallexikon der Germanischen
Altertumskunde 17 (2nd edn., 2001) pp. 177–9.
96
In future: T. Vida, Die Keramik der Spätawarenzeit (working title, forthcoming).
512  

of vessels produced on a slow wheel (pottery “finished on a slow

wheel”) seems to increase gradually and eventually displaces hand-
made pottery (pl. 33). The vessels which have been finished on a
slow wheel are frequently decorated with zones of wavy lines and
the rims become increasingly angular. On the base imprints of the
axial rod of a potter’s wheel may frequently be observed; in some
places there are real base-marks. As in the larger cemeteries, virtu-
ally identical vessels finished on a slow wheel frequently occur, and
we may assume that there must have been something resembling
division of labour regarding the production of pottery within the
respective settlement. However, hand-made vessels—and not the more
“modern” pots finished on a slow wheel—are often found in the
wealthier graves. This implies that those vessels produced with the
help of technical resources were not, in fact considered to be more
valuable. On the other hand, the Middle Avar vessels which were
finished on a slow wheel were the starting point of a typological
series97 which did not end with the Avar Period, but instead con-
tinued—at least on the western fringes of the Avar Empire—well
into the thirteenth century.
Apart from the pottery made by hand and on a slow wheel, fine
ware which was produced industrially on a fast wheel and some-
times even painted, is also known from the Late Avar Period.
Researchers usually refer to these characteristic, frequently pouch-
shaped vessels, which often have a spout-shaped mouth, as “yellow
pottery” (e.g. pl. 37,9).
Eperjes (eastern Hungary) is the only Late Avar settlement known
to date and Csanád Bálint has attempted to give an interpretation
of the material, which is as comprehensive as possible.98 Although
the excavated area is very small, it is possible to gain much infor-
mation about this population’s way of life by comparing the finds
with those from adjacent necropolises and by taking into consider-
ation the results of chemical and physical research. Of course it is
not yet possible to say to what extent we may generalise from the
situation at Erperjes especially if we know how much way of life
and customs can vary between settlements.
In the process of excavation, five sunken, rectangular huts, some
with ovens, were examined. There was also a free-standing oven and

97
Daim, “Nachgedrehte Keramik”.
98
Bálint, Eperjes.
   () 513

several pits. We do not know whether there were also houses which
were built entirely above ground. It should be noted that there is
also evidence for iron smelting, although it now seems that this was
quite a common activity in Avar settlements.
The vast majority of archaeological finds are in fact pottery. Here
we find a spectrum of types which by far surpasses that represented
in the cemeteries, as the grave goods have been selected consciously.
In the settlements—and therefore also in Eperjes—we find hand-
made cauldrons and “baking-bells”, as well as fine ware produced
on the potter’s wheel.
There is an interesting spectrum of animal bones: The majority
of fragments, 59, came from cattle, with 13 from pigs, 17 from
sheep/goats, 8 from horses. Apart from that, there were also deer,
roe, boar and rabbit bones.99 The Avars living at Eperjes must have
bred livestock, with a clear emphasis on cattle breeding. Apart from
that, hunting also took place, and game not only served as a sup-
plement to the everyday diet, but also yielded raw materials: For
instance, deer antlers and presumably also deer tendons were needed
to make a composite bow, which was strengthened with lamellae
made of deer antlers. We may assume that a considerable propor-
tion of farming is in fact crop husbandry, however, this cannot be
expressed in quantitative terms.
Avar cemeteries can grow to a remarkable size, especially if they
are used from the Early to the end of the Avar Period, such as
Keszthely-Dobogo, with its approximately 4,000 graves, or Zamárdi,
with around 6,000 burials. Although no real princely graves—as they
are known from the Early and Middle Avar Period (e.g. Kunbábony)—
have been found from the Late Avar Period, clear differences in the
construction of the grave, dress-ornaments and grave accompani-
ments may be observed also in this particular period. Among the
most remarkable, well furnished graves, which are therefore usually
described as “rich”, are, of course the equestrian graves, where the
male—in exceptional cases also female—deceased is buried with a
horse with its saddle and bridle. The cemeteries of Komárno (Slovakia),
which were mainly used in the Late Avar Period, have a particu-
larly high percentage of equestrian graves (pl. 37–38). The tendency
to maintain some local burial customs over several generations is

99
Ibid., p. 80.
514  

particularly interesting. For instance, only a few years ago a well-

furnished equestrian grave from Pitvaros (eastern Hungary), which
had been constructed as a tunnel grave, was published: The deceased
was put in a tunnel leading downwards obliquely, which had been
dug outward from the rectangular grave. In the grave itself, a com-
plete horse with equipment of exceptional quality had been placed
(pl. 39–41).100
The majority of Avar cemeteries terminate at the end of Late
Avar Period III. However, in some cemeteries we may observe phe-
nomena which suggest that there were radical changes in burial cus-
toms or an interruption in the supply of representational goods. In
Leobersdorf, for instance, in Late Avar Period IIIb belt-sets are no
longer placed in the grave and, although the final graves are large
and deep, they contain only few dress components and no grave
accompaniments (food).101 Similar observations were made during
another excavation campaign in Vienna 11–Csokorgasse: here how-
ever, there is no direct connection with the Avar necropolis.102 In
the most recent part of the cemetery at Zalakomár, there is a high
number of repaired and replaced belt-sets. Therefore, it seems that
this type of dress-component was no longer produced, but that one
did not (as yet) wish to adopt forms of representation which were
already in common use in Carolingian western Hungary. We must
assume that the transition from Avar to East Frankish culture took
place with variable speed in the different regions. This may partly
be attributed to the skill of local potentates and church officials, but
also depended on other organisational, economic and social factors.
Therefore, it is of utmost importance to examine this process sepa-
rately for every population group and every cemetery, and then to
compare the results. It is not much use giving statistical generalisa-
tions for larger regions or, even worse, for the whole Avar Empire.
Given the fundamental changes which are visible in the archaeo-

100
L. Bende, “A pitvarosi késo avar temeto 51—sírja (Adatok a késo avar kori
lószerszámok díszítéséhez) [Das Grab 51 im spätawarenzeitlichen Gräberfeld von
Pitvaros (Beiträge zur Verzierung der spätawarenzeitlichen Pferdegeschirre)]”, A Móra
Ferenc Múzeum Évkönyve—Studia Archaeologica 4 (1998) pp. 195–230 [German summary:
pp. 210–2].
101
Daim, Leobersdorf, p. 171.
102
E.H. Huber, “Neu entdeckte Awarengräber in Wien, Simmering”, Fundort Wien.
Berichte zur Archäologie 1 (1998) pp. 117–43, esp. fig. 2 and 5.
   () 515

logical record at the time of transition from Late Avar Period IIIa
to IIIb, and in due course up to the point where the Avar necrop-
olises are abandoned, it is difficult to imagine that these changes are
not linked to the political turmoil of the Avar wars. In this context,
it seems completely exaggerated to place the entire phase Late Avar
Period III in the ninth century, because we cannot expect that Avar
craftsmen were particularly creative after the year 800. However it
must be granted that Hungarian research, and especially the spe-
cialist for the ninth century, Béla Miklós Szoke,  is correct in criti-
cising the current chronological criteria, which are much too schematic.
The Avar wars have initiated a cultural process which certainly had
not yet been completed at the time of the first Hungarian invasion
in the year 862.
The most important archaeological deposit from the Carpathian
Basin is the gold treasure from Sînnicolau Mare (Hungarian:
Nagyszentmiklós and present-day Romanian Banat). It was found in
1799 and consisted of 23 gold vessels, which together weighed around
10 kg, and is now in the Kunsthistorisches Museum in Vienna
(pl. 42–44). Numerous archaeologists, historians, linguists and re-
searchers from many other fields have studied the Sînnicolau Mare
(Nagyszentmiklós) Treasure, often at a relatively late point in their
career. Less experienced researchers tend to avoid the topic due to
the complexity of the culture-historical problems associated with it.103
The historical identification of the treasure is still highly contro-
versial, mainly because it is difficult to establish precise dates, which
however, are an important pre-requisite for any historical interpre-
tation. When were the different parts of the treasure produced? When

103
N. Mavrodinov, Le trésor protobulgare de Nagyszentmiklós, Archaeologia Hungarica
29 (Budapest 1943); B.I. Marschak, Silberschätze des Orients. Metallkunst des 3.–13.
Jahrhunderts und ihre Kontinuität (Leipzig 1986) pp. 308–16; K. Benda, “Souèasny stav
studia zlatych nádob pokladu ze Sânnicol>ul Mare (Nagyszentmiklós) [Gegenwärtiger
Studienstand über die Goldgefässe aus Sânnicol>ul Mare (Nagyszentmiklós)]”, Slovenská
Archeológia 13,2 (1965) pp. 399–414 [German summary: pp. 412–4]; G. László and
I. Rácz, Der Goldschatz von Nagyszentmiklós (Budapest 1983); C. Bálint, Die Archäologie
der Steppe (Wien-Köln 1989); R. Göbl and A. Róna-Tas, Die Inschriften des Schatzes
von Nagy-Szentmiklós. Eine paläographische Dokumentation, Denkschriften der Österre-
ichischen Akademie der Wissenschaften, philosophisch-historische Klasse 240 (Wien
1995), with extensive bibliography at pp. 64–77; F. Daim and P. Stadler, “Der
Goldschatz von Sînnicolaul Mare (Nagyszentmiklós)”, Reitervölker aus dem Osten. Hunnen
+ Awaren, ed. F. Daim (Halbturn 1996) pp. 439–45. Currently in preparation: C.
Bálint, “Der Goldschatz von Nagyszentmiklós”, Varia Archaeologica Hungarica
(Budapest forthcoming c. 2003).
516  

was the most recent object made? Is there evidence on the most
recent vessels which suggests that they were in use for a long period
of time? It now appears that the vessels from Sînnicolau Mare
(Nagyszentmiklós) were made in the seventh and eighth centuries.
As the results of antiquarian analysis of the treasure show that it is
most closely related to the Avar material and as Sînnicolau Mare
(Nagyszentmiklós) lies within the region of Avar settlement, it is pos-
sible that the objects in the gold hoard are in fact “left-overs” from
the Avar royal hoard, which remained in the Danube region when
most of the treasure was brought into the Frankish Empire after
Charlemagne’s Avar wars. This, however, does not neccessarily imply
that the Sînnicolau Mare (Nagyszentmiklós) Treasure was buried in
the course of the Avar wars—this could also have taken place later.
A number of different cultures meet in the treasure from Sînnicolau
Mare (Nagyszentmiklós); elements of Byzantine, south Russian/Central
Asian, post-Sassanian and Avar origin may be discerned. For now,
it may be assumed that the different parts of the Sînnicolau Mare
(Nagyszentmiklós) treasure originated in a heavily barbarised, provin-
cial Byzantine environment, and that the customer’s tastes and wishes
were taken into consideration when the vessels were made. The tech-
nical quality of some of the vessels is of the very highest standard.
The fact that there are few good parallels for the different types
of vessels included in the treasure, that they are basically individual
objects, and that the ornament of the vessels combines decorative
elements which occur—in the archaeological material pertaining to
the broad masses—in different chronological phases, makes research
on this treasure both difficult and interesting. An international research
project currently in progress in Vienna is aimed at providing a com-
prehensive analysis of the treasure and will, perhaps be able to answer
some important historical questions, and at least provide the basis
for further research.

Instead of a summary:
Constant factors, influences and breaks in Avar culture

Traditionalist historians frequently ask the archaeologist—finally—to

summarise his “historical” results. In fact, the results of archaeolog-
ical research are usually concerned with other aspects of human life
than the written sources. The latter take no notice of what is “obvi-
   () 517

ous”. Archaeological research is restricted to observing long-term

structures and processes, in as far as these leave traces in the ground,
for instance in types of settlements, burial customs, technology, cul-
tural exchange, crafts and trade. Only in very rare cases is it pos-
sible for archaeology to obtain data pertaining to political history.
Nevertheless some of our results are on the same level as our writ-
ten sources, thus shedding light on high politics, structures of sov-
ereignty, international relations, communication and value systems.
Without exaggerating, we can say that only the archaeological sources
and the results of their careful, methodological evaluation and inter-
pretation permit us to interpret a number of written records. One
example of this is the isolated story of an Avar delegation being sent
to Constantinople in 678/79 in order to congratulate the Emperor
on a victory—in the vicinity of the capital—of the Byzantine fleet
against the Arabs. The historian would have many ways to inter-
pret this event, however, only the archaeological sources demonstrate
that this diplomatic activity occured in a period during which Byzantine
coins and luxury goods entered Avar territory, and the Avar elite
almost exclusively followed Byzantine fashions (see for instance Ozora-
Tótipuszta), while at the same time, merchants and/or craftsmen
must have also been crossing the western border of the Avar Empire.
Only once this delegation is thus integrated into a clear archaeo-
logical picture does it make sense from a historical point of view:
The process during which the Avar Empire had gradually re-gained
its strength had now been more or less completed and, immediately
before the foundation of the Bulgarian Empire in 680, Byzantium
was looking for allies on its northern border.
However, let us first deal with the basic pattern and the main
lines of development of Avar culture:
What now presents itself to us as the Avar culture in the Carpathian
Basin was the result of mainly Byzantine but also western influences.
It was doubtlessly formed in situ, with various eastern traditions
grafted onto a local base. The traditions that once constituted the
“backbone” of Avar identity can only be glimpsed archaeologically
in minute fragments. The myths and songs remain unknown to us,
and to be honest we know nothing about the religious ideas; we can
only employ analogies as an aid to understanding. Only the way of
life and the appearance of the Avars may, if only to a certain degree,
be reconstructed. In the course of doing this, one is struck by the
impression that the most vital elements of Avar culture—as far as
518  

we can tell—remain the same. The basic concept of the Avar way
of life persisted until the end of the Avar Empire (around 800).
Although Roman field patterns would have lent themselves to inten-
sive cultivation, crop husbandry evidently played but a minor role.
The equestrian herdsman’s greatest source of pride is his livestock:
all other means of displaying his status are carried on his person.
Nevertheless, at best, in the earliest period (until the first quarter of
the seventh century) the Avars were nomads. Later, they inhabited
permanent settlements, the high population of which did not permit
a nomadic way of life. Although cattle still formed the basis of Avar
agriculture in the eighth century, the horse occupied a major role
in Avar life. Up to the Late Avar Period, equestrian graves appear
to be exceptionally well endowed, even though they become increas-
ingly uniform by comparison with the considerable variety present
in the first half of the seventh century. Even the Avar warrior of
the time of Charlemagne can be assumed to have differed only in
small details from his predecessor of the era of Justinian I or that
of Maurikios. The heavily-armoured cavalry as described in the
Strategikon no longer existed in the eighth century. It probably required
a kind of “war economy”, and would have been impossible to main-
tain in long periods of peace.
In what respects do the Avar means of representation differ from
those of the Byzantines and Franks? First of all, it is apparent that
not much attention is being paid to furnishing the home. Although
we may assume the existence of valuable tapestries and woodwork,
the houses—of the common people, at least—were small and not
very comfortable. Certainly, the wealth of an Avar family was deter-
mined by the number of cattle it owned, as it had been in the time
of their nomadic ancestors. This set of values, which must have com-
prised specific ideas about the right to exploit the land, had survived
the transition to sedentariness. A permanent settlement of the Late
Avar Period appears not to have differed in any fundamental way
from one of the Early or Middle Avar Period. Despite the Avars’
obvious enthusiasm for Byzantine culture, not a single stone build-
ing is known from the Carpathian Basin.
The material culture of the Early Avar Period appears to have
been culturally heterogeneous: drawing on local, late Roman, Germanic
as well as eastern elements, which had presumably been introduced
by the Avars and which included some late Hunnic reminiscences,
plus Byzantine dress-ornaments, shapes and motifs, as well as tech-
   () 519

nological know-how. At the same time, there seem to have been

decisive influences and imports from Lombard Italy, from southern
Germany and—of course to a lesser extent—from other regions of
Europe as well. The process of standardisation of Avar culture begins
only in the second half of the seventh century, some time after the
Avar wars. However, local groups with particular types of dress or
burial customs may be observed even in the eighth century. Avar
culture remains diverse and very colourful.
Both in the Early Avar Period and in the eighth century, the
region around Keszthely, at the western end of Lake Balaton, con-
stitutes an special element within the Avar Empire, although for
totally different reasons. In the late sixth and probably still in the
early seventh century, in Keszthely there is distinct evidence of con-
nections with the Adria region and Italy; here, on the fringes of the
area of settlement, there is a well-organised Christian community
linked with Catholic Italy. Even if we have no written sources con-
cerning this, the fact that pilgrimages, perhaps even to the Holy
Land, were common, is an argument for a working church-organi-
sation in the area around Keszthely, possibly even in Pécs (Fünfkirchen).
Typically Avar cemeteries are found in the region around Keszthely
only from the Middle Avar Period onwards (for instance Gyenesdias).
Therefore—prior to the struggle for power around the year 626—
Keszthely seems to have been the end-point of the route to Italy,
virtually the bridge-head of Mediterranean, Late Roman and Lombard
culture in present-day western Hungary. It is quite possible that this
is linked to the old pact between the Lombards and the Avars.
However this may be: after 626, the situation seems to have changed
fundamentally. In the Late Avar Period, the Keszthely-culture no
longer has anything whatsoever to do with Italy or Byzantium. The
almost “baroque” enlargement of once typically Byzantine dress acces-
sories (basket earrings, dress pins) demonstrates, if anything, the local
population’s adherence to old traditions. It is not evidence of exter-
nal contacts and we have no proof of a flourishing Christian com-
munity in Keszthely in the Late Avar Period. Neither can the
continuation of a burial custom which does not include placing food
in the grave, limited to a very small area, be viewed as proof of a
Christian community at the western end of Lake Balaton in the Late
Avar Period. At the moment, in fact, it looks rather as if the re-
organisation of the Avar Empire in the seventh century and the far-
reaching isolation of the Carpathian Basin must also have caused
520  

the activities of the Christian community to disappear almost com-

pletely. One find which may be perceived as evidence for Christian
burial customs in the eighth century is the existence of the iron
crosses on coffins from Székkutas. It is difficult to explain these as
a tradition without meaning.104 The other finds which might suggest
the existence of Late Avar Christians have no value as evidence,
because the transfer of motifs whose original meaning is, however,
lost in the process, is common on Late Avar belt-mounts.105 Apart
from that, it was some time ago that it was shown that the escutcheon-
shaped hinged fittings with eagle and “cross” from Boldog and
Komárno in fact imitate early Byzantine coins with eagle and the
numeric symbol X.106 At a synod in 796, it is reported that in the
eighth century, there were Christians living in the Avar Empire, but
they had lost contact with their brothers and sisters in faith and
were therefore in need of spiritual guidance.107 With this fact in mind,
the archaeological “evidence” for Late Avar Christianity should be
interpreted with particular care.
The archaeologist is frequently confronted with inquiries regard-
ing the world of beliefs of the population he is studying. Actually,
we know nothing about Avar religion, although the evidence from
analogous ethnographic studies seems to suggest that most groups of
people living in the Carpathian Basin in the Avar Period practised
a kind of shamanism. The evidence for the existence of a Christian
contingent in the early Avar Empire has already been mentioned.
However, the Byzantine cross from Ozora-Tótipuszta (Middle Avar
Period) is rather an expression of the wearer’s association with a
Byzantine form of representation than of her personal beliefs. The
co-existence and the mutual permutation of quite contradictory worlds

104
K.B. Nagy, “Székkutas-Kápolnadul  o avar temeto néhány 9. Szászadi síre-
gyüttese [Einige Grabkomplexe aus dem 9. Jahrhundert im awarischen Gräberfeld
bei Székkutas-Kápolnadul  o]”,
 Az Alföld a 9. szászadban, ed. G. L orinczy
 (Szeged
1993) pp. 151–69, esp. fig. 3,22–23; T. Vida, “Neue Beiträge zur Forschung der
frühchristlichen Funde der Awarenzeit”, Acta XIII congressus internationalis archaeologiae
christianae, Split—Pore‘ 1994 2, ed. N. Cambi and E. Marin, Studi di antichità cris-
tiana 54. Vjesnik za arheologiju i historiju Dalmatinsku 87/89 Suppl. (Split 1998)
pp. 529–40, esp. p. 534 with note 27 and fig. 7.
105
Vida, “Frühchristliche Funde”, pp. 534–6 and fig. 8.
106
Daim, Leobersdorf, p. 146 with notes 192–197.
107
W. Pohl, Die Awaren. Ein Steppenvolk in Mitteleuropa, 567–822 n. Chr. (München
1988) pp. 319–20.
   () 521

of beliefs, symbols and images, the syncretism of the Late Migration

Period, can frequently be deduced from the archaeological evidence.
What we now find so difficult to imagine, seems to have actually
existed in the Early Middle Ages: people adhered to several different
cults in the same way that nowadays it is (theoretically) quite possi-
ble to be a member of two or more different parties. Pictures, amulets
etc. could be interpreted according to Christian as well as to shaman-
istic beliefs. For the early medieval person this must have been, at
best, a theoretical problem. He was much more concerned with their
effectiveness in daily life.108 This is the reason why all motifs on
objects from the Late Avar Period, which seem to possess Christian
meaning, should be interpreted with care.109 Although we have no
positive evidence for a Christian church organisation within the Avar
Empire or for successful missionary activities during the seventh and
eighth centuries, it is quite possible that some Christian traditions
still existed at the time of the Avar wars. Maybe the types of dress
in Late Antique tradition, which we find in the regions around
Keszthely and Pécs (Fünfkirchen) in the eighth century are in fact
evidence for a small Christian community which survived the Avar
reign.
There are some exciting new ideas concerning the Avars’ cultural
orientation after their defeat at Constantinople in 626. If we are cor-
rect in assuming that the princely graves from Bócsa, Tepe and
Kunbábony belong to the second quarter of the seventh century,
they may in fact be interpreted as an attempt to create a represen-
tational culture which makes use of Byzantine technology in jew-
ellery-production and consciously looks for models in the East. The
act of focussing on a particular cultural root may have been deemed
necessary in order to build up the damaged ego and consolidate
power. A second way of interpreting the princely graves with golden
“pseudo-buckles” would be to perceive them as evidence of an alliance
(which was not at all adverse to Byzantium) with Kuvrat’s Bulgarian
Empire, which was then, after his death, followed by more inten-

108
On the amulet-capsules: T. Vida, “Frühmittelalterliche scheiben- und kugelför-
mige Amulettkapseln zwischen Kaukasus, Kastilien und Picardie”, Berichte der Römisch-
Germanischen Kommission des Deutschen Archäologischen Instituts 76 (1995) pp. 219–90, esp.
pp. 263 ff.
109
Vida, “Frühchristliche Funde”.
522  

sive relations with the Byzantine Empire. However, it is important

to stress that this shifting of the orientation of Avar cultural expres-
sion would be evident even if we had no knowledge of Avar polit-
ical history. It would be a grave methodological error to interprete
the archaeological sources from the point of view of the scanty writ-
ten evidence rather than empirically.
Thanks to the decoding of eighth-century Mediterranean orig-
inals in the Avar Empire and its periphery, our perception of the
relationship of Avar and Byzantine or Italo-Byzantine culture is now
becoming much clearer. We are also aware of the influence which
Byzantine technology had on Avar craftsmanship. On the other hand,
Avar culture was strong enough to consciously select from what Medi-
terranean art had to offer, to adapt it stylistically and even to create
totally new motifs on the basis of these stimuli. Due to the wealth of
archaeological remains, the Avar material and its Byzantine or Italo-
Byzantine models are well suited to studying the mechanisms of
cultural transfer and illuminating the system of values, their com-
munication through art and even some of their psychological aspects.
This set of questions was discussed with reference to the griffin-motif
and the boar, the Byzantine birds and the “Chinese” scroll with
blossoms.
The archaeological finds will make it necessary to re-evaluate the
role of the Bulgarians on the Balkans and along the middle and
lower Danube. On the one hand, we can observe strong cultural
relations between the Bulgarian centre of power from Pliska and
Preslav to Varna, and the Byzantine Adria region (Dyrrachium and
surroundings, present-day Albania); on the other hand, there seems
to be a strong link with the region of Avar culture. We cannot quite
predict at present, where the latest research will lead us; neverthe-
less, in the eighth century there is a network of connections between
the Carpathian Basin, Italy, the Balkans and the Black Sea Region
which seems to have been much more intensive than we would have
thought possible even a short while ago.
Christianity, which had frequently played an important role in the
stabilisation of power and social structure, had not (yet) gained access
to eighth-century Avar culture. As a result, when Charlemagne set
the army in motion from Bavaria, the Franks were able to use
“pagan-war” propaganda as a powerful lever. Apparently, Avar tra-
dition was no longer strong enough: certain members of the Avar
   () 523

elite changed sides rapidly without being able to build up positions

of power of their own.
The fundamental cultural changes which were brought about by
the Avars’ defeat by Charlemagne and by the re-organisation of the
eastern part of the Carolingian Empire are becoming increasingly
clear from the archaeological sources. However, some interpretations
are hindered by our problems with absolute fine chronology. There
is definite evidence for the establishment of political and church cen-
tres with a huge input of resources and of a surprisingly high stan-
dard (e.g. Zalavár-Mosapurc), and as a result for an attempt to create
an extensive church, political and economic infrastructure. There is
also evidence for the immigration of settlers from other Slav terri-
tories, which was presumably encouraged by the political leaders.
There is also evidence of immigration from Bavaria and of the short-
or middle-term survival of Avar population groups and settlements.
The extent of the latter, however, varies considerably from region
to region. At the moment it is still difficult to say, to what extent
the cultural integration of the population had been completed when
Magyar rule was established in the Carpathian Basin. Archaeological
research in areas of Slav settlement near Kremsmünster and in the
Enns valley, as well as in the precincts of the “Great Moravian
Empire”, in the princedom Thunau, have yielded good results. Al-
though in both cases—as the burial customs demonstrate—the Slav
ancestry of the population is evident (regardless of their self-percep-
tion!), we can also observe a stronger link with Bavarian and East
Carolingian culture. The written sources which are available both
for Kremsmünster and the Kamp valley, support the archaeological
evidence and illuminate the political background.
Avar traditions were strong enough to survive fundamental changes
in the political geography of Europe. However, the Avar Empire did
not come to grips with the transition to medieval state. This difficult
task was achieved, around 200 years later, by the Magyar king
Stephen. It is due to him that Hungary has survived until the pre-
sent day, while the Avars were assimilated by the Carolingian pop-
ulation and are only accessible to historians and archaeologists.
In one respect, the methods used by the historian and the archae-
ologist are very similar: both must attempt to abandon modern ter-
minological systems and to decipher the early medieval “code”. The
historian has to use the terminology of the written sources as a start-
524  

ing point and he must then examine what was meant by the names
given in the texts. The archaeologist discovers types of artefacts,
forms of dress, types of settlements, burial customs, evidence for other
customs and much more besides. He has to consider what status each
of these cultural elements once possessed within a given semiotic
structure and which of them were understood as a criterion for differ-
entiating between groups. We can assume with some certainty that
there were just as many possibilities of identification in early medieval
society as there are today; some of them manifest themselves in
the archaeological material. However it is highly problematic to
label one or more of these “cultural groups” as “ethnic”, without
extensive spatial comparison. If ethnicity is really a phenomenon of
“social psychology”—as defined by Leo S. Klejn—this would imply
that we are over-stressing our material by a long way.

[Translated by Birgit Bühler]

Illustration acknowledgements

Figures in the Text

Fig. 1. From W. Pohl, Die Awaren. Ein Steppenvolk in Mitteleuropa, 567–822 n. Chr.
(München 1988) map 2, revised. Graphic arts: Franz Siegmeth
Fig. 2. Graphic arts: Franz Siegmeth
Fig. 3. Reitervölker aus dem Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp.
344–5
Fig. 4. F. Daim et al., Leobersdorf p. 168
Fig. 5. Ibid., p. 148

Plates
Pl. 1. 1 Hampel, Alterthümer, pl. 208; 2–3 drawing by Franz Siegmeth I. Kovrig,
“Contribution au problème de l’occupation de la Hougrie par les Avars”,
Acta Archaeologica Hungarica Academiae Scientiarum Hungaricae 6 (1955) pl. VII
Pl. 2. 1 Daim, Leobersdorf, pl. 153; 2 Garam, Funde byzantinischer Herkunft in der
Awarenzeit vom Ende des 6. bis zum Ende des 7. Jahrhunderts (Budapest 2001) pl.
92; 3 H. Winter, Awarenzeitliche Grab- und Streufunde aus Ostösterreich, pl. 28; 4
Garam, Funde, pl. 94; 5 ibid., pl. 94; 6 Hampel, Alterthümer, pl. 180; 7
Garam, Funde, pl. 81; ibid., pl. 93
Pl. 3. Bóna, “A Szegvár-sápoldali lovassír”, fig. 1
Pl. 4. Ibid., fig. 2–5
Pl. 5. 1 Garam, Funde, pl. 3; 2 ibid., pl. 2; 3 ibid., pl. 1; 4 ibid., pl. 41; 5 ibid.,
pl. 31; 6 ibid., pl. 32; 7 ibid., pl. 31; 8 ibid., pl. 32
Pl. 6. Drawings by Péter Posztobányi
Pl. 7. Garam, Goldgegenstände, pl. 53–55; 58; 60–62
Pl. 8. Ibid., pl. 56–58
   () 525

Pl. 9. Das awarenzeitlich gepidische Gräberfeld von Kölked—Feketekapu A, pl. 34; 59; 76
Pl. 10. Ibid., pp. 42; 45 and pl. 36
Pl. 11. Kiss, “Előzetes jelentés” Part I, pp. 267; 279; 293–4; 304; 319
Pl. 12. Ibid., Part I, p. 31
Pl. 13. Reconstruction Tivadar Vida; drawing by Sándor Ősi
Pl. 14. Kiss, “Előzetes jelentés” Part I, p. 332
Pl. 15. 1 Garam, Goldgegenstände, pl. 43; 2 Bakay, “Az avarkor idő rendjéől,
pl. III
Pl. 16. 1 Bakay, “Az avarkor idő rendjéől”, pl. IV; 2 Das awarenzeitlich gepidische
Gräberfeld von Kölked—Feketekapú A, pl. 70; 3 Daim, Sechs Gräber mit
westlichen Gegenständen”, pl. 2; 4 Bende, “Pitvaros”, fig. 5; 5 F. Daim
and A. Lippert, Das awarische Gräberfeld von Sommerein am Leithagebirge, NÖ
(Wien 1984) pl. 15–16
Pl. 17. 1 Bóna, VII. sz-i avar települések és Arpád-kori magyar falu Dunaújvarosban
(Budapest 1973) p. 40; 2 Reitervölker aus dem Osten p. 378
Pl. 18. 1 Vida, Die awarenzeitliche Keramik I (6.–7. Jh.) (Budapest 1999) pl. 1; 2
ibid., pl. 92; 3 Vida, “Zu einigen handgeformten frühawarischen Keramik-
typen und ihren östlichen Beziehungen”, Awarenforschungen, ed. F. Daim
(Wien 1992) pl. 8; 4 ibid., pl. 13; 5 Daim and Lippert, Das awarische
Gräberfeld von Sommerein, pl. 1; 6 Vida, Keramik I, pl. 3
Pl. 19. E. Tóth and A. Horváth, Kunbábony. Das Grab eines Awarenkhagan (Kecskemét
1992) pl. I–IV
Pl. 20. Ibid., pl. II, XI–XII
Pl. 21. Ibid., pl. V, VII, XIII, XV
Pl. 22. Ibid., pl. IX–X, XXV–XXVI
Pl. 23. I. Popovic’, Zlatni avarski pojas iz okoline Sirmijuma (Beograd 1997) fig. 14;
16; 22; 24–25
Pl. 24. G. Fülöp, “Awarenzeitliche Fürstenfunde von Igar”, Acta Archaeologica
Hungarica Academiae Scientiarum Hungaricae 40 (1988) fig. 10; 13–14
Pl. 25. R. Müller, “Vorbericht über die Freilegung des Grabes eines hohen
Militärs aus der Mittelawarenzeit in Gyenesdiás”, Communicationes Archaeo-
logicae Hungariae (1989) fig. 2; 5–6
Pl. 26. Ibid., fig. 5; 10
Pl. 27. Ibid., fig. 8–9
Pl. 28. Garam, Goldgegenstände, pl. 85–87
Pl. 29–32. Mödling—Goldene Stiege, ed. F. Daim, K. Matzner and H. Schwammenhöfer
(forthcoming)
Pl. 33. Daim, “Zur nachgedrehten Keramik aus dem awarischen Gräberfeld
von Mödling—An der goldenen Stiege”, Slawische Keramik in Mitteleuropa
vom 8. bis 11. Jahrhundert, ed. C. Stana (Brno 1994) fig. 1–2
Pl. 34. Daim, “‘Byzantinische’ Gürtelgarnituren des 8. Jahrhunderts”, Die Awaren
am Rand der byzantinischen Welt. Studien zu Diplomatie, Handel und Technolo-
gietransfer im Frühmittelalter, ed. id. (Innsbruck 2000) pp. 113; 123; 137 ff.;
167
Pl. 35. 1 Hampel, Alterthümer, pl. 186,4; 2–5 R. Müller, “Keszthely-Kultur I”,
pp. 292–3
Pl. 36. Daim, Leobersdorf, pl. 69–71
Pl. 37. 1–4 A. Trugly, “Gräberfeld aus der Zeit des awarischen Reiches bei
der Schiffswerft in Komárno I”, Slovenská Archeológia 35 (1987) pp. 200;
256; 257; 5–9 Trugly, “Komárno I”, p. 302
Pl. 38. A. Trugly, “Gräberfeld aus der Zeit des awarischen Reiches bei der
Schiffswerft in Komárno II”, Slovenská Archeológia 41 (1993) pp. 244; 286;
290
Pl. 39. L. Bende, “A pitvarosi késo avar temeto 51—sírja (Adatok a késo avar
526  

kori lószerszámok díszítéséhez)”, A Móra Ferenc Múzeum Évkönyve—Studia

Archaeologica 4 (1998) p. 212
Pl. 40. 1–2 Ibid., p. 214; 3 p. 216
Pl. 41. 1 Ibid., p. 225; 2 p. 224
Pl. 42–44. Hampel, Alterthümer, Vol. III, pl. 288–319
   () 527

Plate 1:1: Early Avar bit with side-bars made of antler (Cikó, Hungary); 2 Thrust
lance (Zámoly, Hungary); 3 “Apple-shaped” stirrup (Bicske, Hungary). Scale: 1:2.
528  

Plate 2: Belt ornaments of Byzantine type from an Early Avar context. 1 Mask
fitting (Leobersdorf, Austria); 2 Mask fitting (Szekszárd, Hungary); 3 Cast silver
fitting (Bruckneudorf, Austria); 4 Silver fitting with “fringe decoration” cast in open-
work (Budakalász, Hungary); 5 Cast silver fittings and sheet silver strap ends (Kiskőrös,
Hungary); 6 Strap end and fitting (Keszthely – Fenékpuszta, Hungary); 7 Fittings
and strap ends with “triple hemispherical dent” ornament (Keszthely – Fenékpuszta,
Hungary); 8 Sheet silver strap end (Cikó, Hungary). Scale: 2:3.
   () 529

Plate 3: The Early Avar men’s grave from Szegvár – Sapoldal (Hungary), with
the complete body of a horse and a horse skin as grave accompaniments.
530  

Plate 4: Selected finds from the Early Avar men’s grave from Szegvár – Sapoldal,
Hungary. 1 Gold earring; 2 Sheet metal strap ends; 3 Rosette-shaped mount, pre-
sumably from the quiver strap; 4 Strap-loop made of bone; 5 Buckles; 6 Sword
and attachment loop; 7 Iron armour lamellae; 8–9 Harness mounts. Scale: 6 much
reduced, the others approximately 2:3.
   () 531

Plate 5: Finds from the early Keszthely-culture. 1 Byzantine gold earring (Keszthely
– Fenékpuszta, Hungary); 2 Byzantine gold earring (Bóly, Hungary); 3 Byzantine
basket earring (Keszthely – Fenékpuszta, Hungary); 4 Gold disc – fibula (with real
pearls and rock crystal inlay) and gold pin (Keszthely – Fenékpuszta, Hungary);
5 Box fibula with Bellerophon/equestrian saint (Keszthely – Fenékpuszta, Hungary);
6 Box fibula with Herkules and Omphale (Keszthely – Fenékpuszta, Hungary);
7 Box fibula with imperial apotheosis (?) (Keszthely – Dobogo, Hungary); 8 Box fibula
with archangel (Nagyharsány, Hungary). Scale 2:3.
532  

Plate 6: Selected finds from the Early Avar metalsmith’s grave from Kunszentmárton,
Hungary. 1 Three weights made of copper alloys and two made of glass; 2 Moulds
for pressing a belt set of Byzantine type; 3 Mould for pressing earring components,
fittings and strap ends. Scale: 2:3.
   () 533

Plate 7: Selected finds from the Early Avar equestrian grave from Kunágota.
1 Parts of the belt set. Bronze buckle, silvered; gold fittings and strap ends; 2 Gold
fingering; 3 Sheet silver harness mounts; 4 Sheet silver cup. Scale: 2:3.
534  

Plate 8: Sword from the Early Avar equestrian grave from Kunágota. Left:
Reconstruction; below left: buckle, copper alloy, silvered; functional components
and fittings made of gold foil. Reconstruction scaled down considerably, otherwise
scale: 2:3.
   () 535

Plate 9: Early Avar finds from cemetery A of Kölked – Feketekapú. 1 Sword;

2 “Four-part belt set” with “tooth-cut” (“Zahnschnitt”) decoration, copper-alloy;
3 Bow fibula. Reconstruction scaled down considerably, sword components scale 1:2,
otherwise M 2:3.
536  

Plate 10: Byzantine folding-chair from an Early Avar grave from Kölked, Hungary.
Iron, silver wire inlay. Oblique view scaled down considerably, scale of details
approximately 1:3.
   () 537

Plate 11: Selected finds of Germanic type from the Early Avar women’s grave 85
of Kölked – Feketekapú B. 1 Bronze gilded shield-tongue buckle decorated with
the image of the Germanic god Tyr (?); 2 Rectangular fitting made of gilded bronze;
3 Cast silver pin head with “tooth-cut” (“Zahnschnitt”) ornament; 4 Silver bow fibula
of Cividale type; 5 Silver strap end, part of a belt pendant. Scale: 2:3.
538  

Plate 12: The Early Avar women’s grave 85 from Kölked – Feketekapú B:
objects found in the pelvic- and upper thigh region.
   () 539

Plate 13: The woman from Kölked – Feketekapú B, grave 85. Reconstruction of
dress: Tivadar Vida.
540  

Plate 14: Kölked – Feketekapu B, grave 119. Wooden bowl with silver fittings.
1 Silver rim fitting with depictions of birds of prey and of animal heads (boar’s
heads below); 2 Central silver fitting (reconstructed); 3 Reconstruction of the bowl.
1, 2 scale 2:3.
   () 541

Plate 15: Early Avar “toothcut” (“Zahnschnitt”) decoration. 1 Parts of a gold sword-
pendant and a saddle- or shield-mount (so-called “Jankovich-gold”, provenance
unknown); 2 Parts of a belt-set from Zamárdi, Hungary. Scale 2:3.
542  

Plate 16: Seventh century belt ornaments of western type from Avar contexts.
1 “North Italian” belt set (first half of the seventh century, Zamárdi, Hungary);
2 Iron strap-end with wire inlay in the form of spiral ornaments (second quarter
of the seventh century, Kölked, Hungary); 3 Iron fitting with wire inlay in the form
of spiral ornaments (second quarter of the seventh century, Zillingtal, Austria);
4 Fitting and strap end with silver wire inlay (second quarter of the seventh century,
Pitvaros, Hungary); 5 Parts of an iron belt set with silver inlay in the form of stripes
(third quarter of the seventh century, Sommerein, Austria). Scale 2:3.
   () 543

Plate 17: Archaeological evidence of Avar settlements. 1 Early Avar sunken hut
with oven from Dunaújváros, Hungary; 2 Settlement with system of ditches and
well from Lébény, Hungary.
544  

Plate 18: Early Avar pottery. 1 With stamped decoration, of Germanic type
(Budakalász, Hungary); 2 “Water bottle”, “grey pottery” (Kisk őrös, Hungary);
3 Hand-made (Szeged, Hungary); 4 Hand-made (Budapest, Hungary); 5 Jug with
spout, “grey pottery” (Sommerein, Austria); 6 Pot, “grey pottery” (Budakalász,
Hungary). Scale: 1:3.
   () 545

Plate 19: Princely grave from Kunbábony, Hungary, selection of finds. 1 Gold
buckle with millefiori-inlay; 2 Fittings and strap-ends made of sheet gold with beaded
border. 1 Scale 3:2, otherwise 2:3.
546  

Plate 20: Princely grave from Kunbábony, Hungary, selection of finds. 1 Strap ends
and fittings made of sheet gold, partly decorated with granulation; 2 Gold earrings
3 Gold fingerings; 4 Gold bracelets; 5 Gold fitting with millefiori-inlay and niello-
decoration. 5 Scale 3:2, otherwise 2:3.
   () 547

Plate 21: Princely grave from Kunbábony, Hungary, selection of finds. 1 Sword;
2–3 Iron knives with sheaths made of sheet gold, partly decorated with granula-
tion. Sword reconstruction scaled down considerably, scale otherwise 2:3.
548  

Plate 22: Princely grave from Kunbábony, Hungary, selection of finds. 1 Jug made
of gold sheet; 2 Gold drinking-horn; 3 Finger tip ornament (?) and shroud trim-
ming (?), both made of gold foil. 3 Scale 1:3, otherwise 1:2.
   () 549

Plate 23: Gold belt set with “pseudo-buckles” from Sirmium, selection. Scale: 2:3.
550  

Plate 24: The Middle Avar men’s grave from Igar, Hungary, selection of finds.
1 Earring; 2 Plait clasp; 3 Sheet gold belt set with chain ornament, buckle with “tooth-
cut” (“Zahnschnitt”) ornament; 4 Sabre; 5 Decorative mount from the attachment
loops of the sabre sheath; 6 Fitting with hook; 7 Buckle of Byzantine type. Sabre
reconstruction scaled down considerably, parts of the sabre scale approximately 1:2,
other objects scale 2:3.
   () 551

Plate 25: The Middle Avar equestrian grave 64 from Gyenesdiás, Hungary, 1 Gold
earrings, 2 gold fmgerrings, 3 Buckle of Byzantine type; 4 Parts of the belt set.
Scale 2:3.
552  

Plate 26: The Middle Avar equestrian grave 64 from Gyenesdiás, Hungary. 1 Sabre;
2 Horse harness ornaments. Sabre reconstruction scaled down considerably, other-
wise scale 2:3.
   () 553

Plate 27: The Middle Avar equestrian grave 64 from Gyenesdiás, Hungary. Parts
of the composite reflex bow, made of deer antler. Scale: 2:3.
554  

Plate 28: The Middle Avar women’s grave from Ozora, selection of gold objects.
1 Earrings; 2 Torque with pendant; 3 Byzantine cross; 4 Leaf-shaped pendant;
5 Part of a coat fastener; 6–7 Fingerings. Scale 2:3.
   () 555

Plate 29: The cemetery of Mödling – An der goldenen Stiege, Austria. Selection
of finds. 1 Gold earrings and sheet silver belt ornaments with chain decoration.
Grave 35, Middle Avar Period I; 2 Belt ornament with interlace. Grave 93, Middle
Avar Period I; 3 Belt ornament from grave 79. Middle Avar Period EL Scale: 2:3.
556  

Plate 30: The cemetery of Mödling – An der goldenen Stiege, Austria. Selection
of finds. 1 and 2 Belt ornaments transition Middle-/Late Avar Period. 1 Grave 242;
2 Grave 135; 3 Belt ornaments from Late Avar Period III, Grave 140. Scale: 2:3.
   () 557

Plate 31: The cemetery of Mödling – An der goldenen Stiege, Austria. Women’s
grave 144, Late Avar Period III, selection of finds. 1 Gold and silver earrings;
2 Spiral fingerings; 3 Bead necklaces, 4 Bracelet, 5 Torque. Scale: 2:3.
558  

Plate 32: The cemetery of Mödling – An der goldenen Stiege, Austria. Women’s
grave 144, Late Avar Period III. Coat fastener, bronze, mercury gilded.
Scale: 2:1.
   () 559

Plate 33: The cemetery of Mödling – An der goldenen Stiege, Austria. Development
of pottery “finished on the slow wheel”. 1 Middle Avar Period I (Grave 93);
2–4 Late Avar Period I (Grave 135, 232, 242); 5 Late Avar Period II (Grave
121); 6–7 Late Avar Period III (Grave 311, 505 A). Scale: 1:3.
560  

Plate 34: Mediterranean belt ornaments from the eighth century. 1 Parts of the
belt set from Hohenberg, Austria. Brass, gilded; 2 Fragment of a main strap-end
from Mikulc§ice, Southern Moravia; 3 Chased sheet gold strap end from Aleppo,
Syria; 4 Silver fitting, gilded. Reputedly from Weiden am See, Austria. Scale 2:3.
   () 561

Plate 35: Finds from the late Keszthely-culture (eighth century). 1 Dress pin;
2–5 Selection of finds from Lesencetomaj, Hungary, women’s grave 6.2 Earrings;
3 Fingering; 4 Basket earrings with bead necklace; 5 Bracelet. Scale 2:3.
562  

Plate 36: Belt ornaments from the cemetery of Leobersdorf Austria, men’s grave
71 (Late Avar Period II), bronze, mercury gilded. 1 Buckle; 2 Strap loop;
3–4 Fittings decorated with images of griffins; 5 Propeller-shaped mount; 6 Hole-
reinforcements; 7 Subsidiary strap ed; 8 Main strap end with animal combat scene
(two griffins attacking a deer). Scale: 2:3.
   () 563

Plate 37: The Late Avar cemetery of Komarno, Slovakia. Selection of finds. 1–4
Equestrian grave 114:2 Round decorative horse-harness mount; 3 Buckle; 4 Decorative
capsule from horse-harness. 5–8 Grave 12:5 Main strap end with circus scene,
Bronze, gilded; 6 Subsidiary strap end; 7 Fitting with eagle and numerical symbol
X. 9 “Yellow pottery”, produced on the potter’s wheel, from grave 11.9 scale 1:3,
otherwise 2:3.
564  

Plate 38: The Late Avar cemetery of Komarno, Slovakia. Selection of finds.
1 Round decorative horse-harness mount with scroll- and flower-ornament of Chinese
type (grave 149); 2 Head ornamet from horse-harness (grave 142); 3 Round deco-
rative horse-harness mount (grave 103). Scale: 2:3.
   () 565

Plate 39: The Late Avar tunnel grave from Pitvaros, Hungary, grave 51. Grave
structure. Left: ground-plan: right: cross-section.
566  

Plate 40: The Late Avar tunnel grave from Pitvaros, Hungary. Horse-harness
decoration. 1 Head ornament; 2 Mounts; 3 Round decorative harness mount.
Scale: 2:3.
   () 567

Plate 41: The Late Avar tunnel grave from Pitvaros, Hungary. Above: Horse
skull with parts of the bridle; below: reconstruction.
568  

Plate 42: The Sînnicolau Mare Treasure, Romania (Hung. Nagyszentmiklós). Selection
of finds. 1 Cup Nr. 11; 2 Bowl with stem Nr. 22; 3 Jug Nr. 3; 4 Drinking horn
Nr. 17. (numbers according to Hampel). Scale: app. 1:2.
   () 569

Plate 43: The Sînnicolau Mare Treasure, Romania (Hung. Nagyszentmiklós). Selection
of finds. 1 Jug Nr. 7; 2 Jug Nr. 2 (numbers according to Hampel). Scale: app. 1:2.
570  

Plate 44: The Sînnicolau Mare Treasure, Romania (Hung. Nagyszentmiklós).

Selection of finds. 1 Bowl Nr. 9; 2 Bowl with bull’s head Nr. 13; 3 Bowl Nr. 19,
with imprint; 4 Bowl Nr. 21. Numbers according to Hampel. Scale: app. 1:3.
 A NON-ROMAN EMPIRE IN CENTRAL EUROPE:
THE AVARS

Walter Pohl

The role of steppe empires in early medieval Europe

Most of the kingdoms discussed in this volume had their centres in

former Roman provinces, and followed the Roman model to some
extent, although it may be disputed in every instance what was Roman
and what was “barbarian” about them; or, more fundamentally: Is
the opposition between “Roman” and “barbarian” (or “Germanic”)
an adequate way to describe these kingdoms? For much in these
societies had specifically been adapted or developed in the post-
Roman period.1 In every documented instance, supra-regional king-
ship had to be Christian to be successful, for that (and only that, it
seems) allowed the integration of a broad Christian population, and
the maintenance of some late-Roman infra-structure and regular com-
munication between different parts of the kingdom. Communities
such as the fourth-century Alamanni and Tervingi, the fifth- and sixth-
century kingdoms in the British Isles, the sixth- to eighth-century
Slavs or the continental Saxons in the eighth century, lacking some
of these Roman and Christian foundations, created only a territori-
ally and institutionally limited range of lordship. It is debatable
whether Scandinavian kings of the post-Roman period enjoyed a
more prominent position;2 on the continent, in any case, powerful king-
ship on a purely barbarian and pagan basis could not be maintained.3

1
See W. Pohl, Die Völkerwanderung. Eroberung und Integration (Stuttgart-Berlin-Köln
2002); with a critique of the binary opposites often used to describe post-Roman
societies: I.N. Wood, “Conclusion: Strategies of distinction”, Strategies of Distinction.
The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The
Transformation of the Roman World 2 (Leiden-New York-Köln 1998) pp. 297–303.
I would like to thank Bonnie Effros for corrections of the English text, and Falko
Daim for suggestions.
2
L. Hedeager, “Asgard reconstructed? Gudme—a ‘central place’ in the North”,
Places of Power, ed. M. de Jong, F. Theuws and C. van Rhijn, The Transformation
of the Roman World 6 (Leiden-Boston 2001) pp. 467–508.
3
This leaves little room for the arguments in favour of a Germanic character of
572  

There is an important exception to this rule, however: the steppe

empires based in the Carpathian Basin. These include: first, the king-
dom of the Huns in the first half of the fifth century; then, the Avar
khaganate; and finally, the Magyars/Hungarians in the tenth cen-
tury.4 They broadly (although with some specific traits) represent a
type of state current in the Eurasian steppes from the time of the
Scythians in the first millennium B.C. to the late medieval Mongols
and beyond. These are often classed as “nomadic” empires, but that
is a simplification because the degree and type of nomadism involved
varied considerably; especially in European steppe empires, their pop-
ulation (for example, the late Avars) might become rather sedentary
although pastoralism remained important. The defining trait was that
the ruling groups were steppe warriors on horseback. They were
able to build supra-regional networks of communication and depen-
dance that allowed them to control and govern huge areas.
Thus, the steppe empires provide an alternative to the post-Roman
model of government which offers valuable points of comparison.
Furthermore, they also directly or indirectly influenced the develop-
ment of the post-Roman societies. It has long been argued that the
arrival of the Huns in eastern Europe in c. 375 A.D. accelerated the
migration of barbarians into the Roman Empire and thus had grave
short- and long-term consequences.5 Hunnic dominion showed that
barbarians could wield power over a large area, and favoured the
establishment of new networks of power that cut across the fron-

early medieval kingship forcefully deployed by Otto Höfler in the 1930s—see

O. Höfler, “Das germanische Kontinuitätsproblem”, Historische Zeitschrift 157 (1938)
pp. 1–26—and shared by the Verfassungsgeschichte school of his time; see W. Pohl,
“Herrschaft”, Reallexikon der Germanischen Altertumskunde 14 (2nd edn., 1999) pp. 443–57.
4
E.A. Thompson, The Huns (2nd edn., Oxford 1996); W. Pohl, Die Awaren. Ein
Steppenvolk in Mitteleuropa, 567–822 (München 1988); id., “The role of steppe peo-
ples in eastern and central Europe in the first millennium A.D.”, Origins of Central
Europe, ed. P. Urbańczyk (Warsaw 1997) pp. 65–78.
5
A position argued forcefully by P.J. Heather, “The Huns and the end of the
Roman Empire in western Europe”, English Historical Review 110 (1995) pp. 4–41;
the sceptical arguments by W. Goffart, Rome’s Fall and After (London 1989) p. 130
(“the barbarian invasions after 370 were neither numerous nor long-lasting”), and
M. Kulikowski, “Nation vs. army: a necessary contrast?”, On Barbarian identity—
Critical Approaches to Ethnogenesis Theory, ed. A. Gillett (Turnhout forthcoming),
should be taken seriously but do not suffice to explain away that there was con-
siderable migration by barbarian warriors in the decades after 375. I would reckon
that about 100,000 barbarian warriors came into the Empire between 375 and
410.
   () 573

tiers. Around 450, the fathers of three successive rulers of Italy—the

last emperor, Romulus Augustulus; Odoacer; and Theoderic—were
in Attila’s service.
The rise of Avar power after 560 also had far-reaching conse-
quences and is much less known. It led directly and indirectly to
the collapse of the Justinianic system and the loss of most of the
Balkan provinces to the Byzantine Empire. Therefore, rather than
dealing with early medieval “non-Roman Europe” in general, this
paper, as the preceding one, make an exemplary case of the Avars.
One additional reason for this attention is that in early medieval
Europe, there is no comparable archaeological record; so far, about
60,000 Avar graves, many of them with grave-goods, have been exca-
vated. The Avars thus also constitute a test-case for the opportuni-
ties and the pitfalls of interdisciplinary research, recently an object
of heated methodological debate (see below and in Falko Daim’s
contribution). Written sources are, as may be expected, patchy, but
some aspects (for instance, the wars and negotiations between
Byzantium and the Avars until 626) are rather well documented in
the works of Menander, Theophylact Simocatta and others.6
The Avars appeared north of the Caucasus in the late 550s.7 It
is characteristic that as soon as they arrived in the neighbourhood
of the Empire on their flight from central Asia, they sent an embassy
to the aging Byzantine emperor Justinian in 558. They agreed to
fight for the Empire against unruly barbarians and in turn would
receive annual payments and other benefits. In 568, khagan (which
was the Avar rulers’ title) Baian, as the result of an alliance with
the Lombards, took possession of the Carpathian Basin and soon
established his supremacy over practically all groups of barbarian
warriors along the Danube frontier. Despite growing difficulties with
unruly groups of Slavs, his sons managed to preserve this position.
They used it to increase the pressure on the emperor and the pay-
ments in gold and kind reached the record sum of 200,000 gold
solidi (almost 3,000 pounds of gold) shortly before 626. These trea-
sures allowed the khagans to win the loyalty of their growing army;
nowhere could a barbarian warrior hope to gain more prestige than

6
M. Whitby, The Emperor Maurice and his Historian: Theophylact Simocatta on Persian
and Balcan Warfare (Oxford 1988).
7
For all aspects of Avar history, see Pohl, Awaren. An English translation is in
preparation.
574  

by following the khagan with his legendary victories and dazzling

wealth. Paradoxically, victories and successful expansion inescapably
led to crisis, for the khagan had to set himself more and more ambi-
tious goals to satisfy the expectations of his growing army. The great-
est and most daring attack on the Byzantine Empire was the siege
of Constantinople in 626, something that neither the Huns before,
nor the Hungarians after the Avars, ever undertook. The Avar army
had even carried hundreds of vessels with Slavic boatsmen from the
Danube to attack the city on water from the Golden Horn. But the
siege failed, and the Avar offensives came to an end. With the trea-
sures assembled in the decades before 626, however, the Avar kha-
gans managed to maintain their power over much of eastern central
Europe; they preserved their empire unchallenged for an additional
170 years.
Throughout the more than two centuries of their rule, the Avars
maintained rather peaceful relations with their western neighbours,
Franks, Bavarians and Lombards, with a few minor border conflicts
excepted. This may be a reason why the Huns and Hungarians,
who repeatedly ravaged western Europe, have been much more
prominent in medieval and modern western historiography. On the
contrary, it was an attack from the West that led to the downfall of
the Avars. In 795/96, Charlemagne’s armies destroyed the Avar kha-
ganate, almost without encountering resistance. The name “Avars”
fell into disuse within one generation. This did not mean that the
Avars had all physically disappeared; nonetheless, it simply proved
impossible to keep up an Avar identity after Avar institutions and
the high claims of their tradition had failed.

Avar origins

My father, the khagan, went off with seventeen men. Having heard
the news that [he] was marching off, those who were in the towns
went up mountains and those who were on mountains came down
[from there]; thus they gathered and numbered seventy men. Due to
the fact that Heaven granted strength, the soldiers of my father, the
khagan, were like wolves, and his enemies were like sheep. Having
gone on campaigns forward and backward, he gathered together and
collected men; and they all numbered seven hundred men. After they
had numbered seven hundred men, [my father, the khagan] organized
and ordered the people who had lost their state and their khagan, the
   () 575

people who had turned slaves and servants, the people who had lost
the Turkic institutions, in accordance with the rules of my ancestors.8
This is how an eighth-century Turkic inscription found in the Orkhon
area of central Asia describes the beginnings of a steppe empire.
The direct relationship between a new people and its institutions,
between gens, ruler and kingdom (or khaganate) could hardly be made
clearer. It is the military success of a leader that establishes a new
community, which he then “orders in accordance with the rules of
the ancestors”. Those who have “lost the Turkic institutions” and
become slaves of foreign rulers regain their freedom. There is, at
this stage, no trace of a myth of common descent. The reference to
Turkic identity is implicit, however, it was not particular to the new
group. Only later, when the new khagan had subdued all relevant
Turkic groups, would his realm become “the” empire of the Turks.
We can picture the origin of the Avars, and the relationship between
“gens and regnum” in the khaganate, in a similar manner. Until they
became “the” Avars, a rather small and aggressive group had to
establish a strong leadership and achieve extraordinary success.
In Latin and Greek historiography, the Avars were quite consis-
tently called Avari or ÖAbaroi. At the same time, they were identified
with the Huns and the Scythians. This was a broad ethnographic
category that underlined their way of life as steppe warriors and cat-
tle breeders (not necessarily, but predominantly nomads in the mod-
ern sense). However, the central Asian Sogdians and Turks knew
the European Avars by a different name: Varchonitai, as Menander
and Theophylact attest.9 In quite an extraordinary manner, these
two writers shed light on the central Asian origins of the Avars. The
second half of the sixth century, the period of formation of the first
Turk empire in central Asia, saw unusually intense exchanges of
embassies and information between Constantinople and the Eurasian
steppes, as far east as the Altai mountains; a similar intensity of com-
munication along the silk route was only reached again in the thir-
teenth century, in the heyday of the Mongol empire.

8
Pohl, Awaren, pp. 164–5; English translation of the text: P.J. Geary, The Myth
of Nations. The Medieval Origins of Europe (Princeton 2001) p. 94.
9
Menander Protector, Fragmenta 4,2, pp. 44–6; 10,1, pp. 114–6; 19,1, p. 174,
ed. R.C. Blockley, Arca 17 [The History of Menander the Guardsman] (Liverpool 1985)
[henceforth: Menander].
576  

Theophylact Simocatta (who wrote around 640) indeed gives some

precious, if confusing information about the origin of the Avars in
the context of a description of a diplomatic exchange with a kha-
gan of the Turks. The Turks, he writes, after their victory over the
Hephtalites
enslaved the Avar nation. But let no one think that we are distorting
the history of these times because he supposes that the Avars are those
barbarians neighbouring on Europe and Pannonia, and that their arrival
was prior to the times of the emperor Maurice. For it is by a mis-
nomer that the barbarians on the Ister have assumed the appellation
of Avars; the origin of their race will shortly be revealed. So, when
the Avars had been defeated, some of them made their escape [to
different places. Next, the Turkish khagan subdued the Ogurs,] one
of the strongest tribes on account of its large population and its armed
training for war [. . .] The earliest leaders of this nation were named
Var and Chunni; from them some parts of those nations were also
accorded their nomenclature, being called Var and Chunni. Then,
while the emperor Justinian was in possession of the royal power, a
small section of these Var and Chunni fled from that ancestral tribe
and glorified their leader with the appellation of khagan. Let us declare,
without departing in the least from the truth, how the means of chang-
ing their name came to them. When the Barselt, Onogurs, Sabir, and
other Hun nations in addition to these, saw that a section of those
who were still Var and Chunni had fled to their regions, they plunged
into extreme panic, since they suspected that the settlers were Avars.
For this reason they honoured the fugitives with splendid gifts and
supposed that they received from them security in exchange. Then,
after the Var and Chunni saw the well-omened beginning to their
flight, they appropriated the ambassadors’ error and named themselves
Avars; for among the Scythian nations that of the Avars is said to be
the most adept tribe. In point of fact even up to our present times
the Pseudo-Avars (for it is more correct to refer to them thus) are
divided in their ancestry, some bearing the time-honoured name of
Var while others are called Chunni.10
Scholars have come up with various interpretations of this story, try-
ing to detect the “true identity” of the Avars behind its narrative:
the Avars were in fact Juan-Juan, Hephthalites, Ogurs or possibly
other peoples. But Theophylact’s account is more complex than that.
Probably, the story is derived from the enemies of the Avars, the

10
Theophylactus Simocatta, Historiae 7,7–8, ed. C. de Boor (Leipzig 1887) [hence-
forth: Theophylactus Simocatta].
   () 577

Turks, who threatened to annihilate the Avars under their hooves

and on one occasion strongly reproached the Byzantines for having
concluded a treaty with the Avars.11 Byzantine writers had little rea-
son for sympathy with the Avars either. Thus, the account has a
distinct anti-Avar flavour; the conclusion that they were in fact only
Pseudo-Avars was surely intended as an insult, and Byzantine pro-
paganda might have used it for that purpose. In fact, it is strangely
reminiscent of the so-called Namensatz in the Germania of Tacitus
which says that the Germani were given that name because people
were afraid of a rather small group that had crossed the Rhine; this
story may probably be explained in the context of the propaganda
war around the revolt of Civilis.12 Still, the basic outline of Theo-
phylact’s report is trustworthy, for it fits well with what we know
about the construction of power and ethnic identity on the steppe.
A few basic principles emerge very clearly from the story:

1. Many steppe empires were founded by groups who had been

defeated in previous power struggles but had fled from the domin-
ion of the stronger group. This is especially true for European
steppe empires (the Hungarians also came to central Europe after
defeat against the Pechenegs); the Carpathian Basin was ideally
suited for protection against more powerful groups who domi-
nated the central Asian steppes.
2. These groups usually were of mixed origin, and each of its com-
ponents was part of a previous group. In the case of the Avars,
the heterogeneous character of the new unit was conceived of as
a dual structure, Var and Chunni, which was also reflected in the
name Varchonitai; these in turn were supposed to have formed
part of the Ogurs. But Theophylact also insists that only “a small
section” of these Varchonitai started the Avar movement west-
ward.
3. Crucial in the process was the elevation of a khagan, which signified
a claim to independent power and an expansionist strategy. This
group also needed a new name that would give all of its initial
followers a sense of identity.

11
Menander 19,1, p. 174.
12
Tacitus, Germania 2, ed. A. Lund (Heidelberg 1988); for an overview of the
debate, W. Pohl, Die Germanen, Enzyklopädie deutscher Geschichte 57 (München
2000) pp. 54–6.
578  

4. The name for a new group of steppe riders was often taken from
a repertoire of prestigious names which did not necessarily denote
any direct affiliation to or descent from groups of the same name;
in the early middle ages, Huns, Avars, Bulgars, and Ogurs, or
names connected with -(o)gur (Cutrigurs, Utigurs, Onogurs, etc.),
were most important. In the process of name-giving, both per-
ceptions by outsiders and self-designation played a role. These
names were also connected with prestigious traditions that directly
expressed political pretensions and programmes, and had to be
endorsed by success. In the world of the steppe, where agglom-
erations of groups were rather fluid, it was vital to know how to
deal with a newly-emergent power. The symbolical hierarchy of
prestige expressed through names provided some orientation for
friend and foe alike.13

Other gentes under Avar rule

A successful group could quickly grow as other groups joined it or

were subdued. In the case of the Avars, this process of expansion
may be reconstructed in some detail.14 In the course of their march
along the northern shore of the Black Sea in the years 559/60, the
Avars subdued the Cutrigurs, Utigurs, Zaloi, Sabirs, Antae and oth-
ers. Whereas the Sabirs and the Antae maintained some indepen-
dence until the beginning of the seventh century, the Cutrigurs and
Utigurs disappeared. 10,000 Cutrigurs are mentioned for the last
time as Avar subjects in 568. Before 582, Avar envoys repeatedly
demanded the yearly subsidies that Byzantium had paid to the Utigurs
for themselves. More central Asian warriors joined the khagan around
580; the sources name the three tribes of Tarniach, Zabender and
Kotzagir, alleged to comprise 10,000 men each.15
Not all of these steppe warriors of different origin simply became
Avars, although many of them did. Some ethnic divisions among
them were maintained, while others faded away. The names of

13
For the origin of the Avars and its context, see Pohl, Awaren, pp. 18–43. For
the practice of name-giving, H.-W. Haussig, Die Geschichte Zentralasiens und der Seidenstraße
in vorislamischer Zeit (Darmstadt 1983).
14
For this and the following, Pohl, Awaren pp. 43–8; 225–36.
15
Theophylactus Simocatta 7,8.
   () 579

specific groups already disappear from our quite extensive sources

in the course of the late sixth century. Soon, a two-layered struc-
ture emerges: whereas the warrior elite of the khaganate were called
Avars in the full sense, the less prestigious but still free warriors on
horseback of largely eastern- or central-European origin (among them,
Cutrigurs and Utigurs) came to be known as Bulgars. This name
had already been attested to in the 480s along the lower Danube,
when Theoderic the Great had defeated them.16 As Bulgars, they
formed recognizable units within Avar armies or acted independently
in numbers of a few hundred or thousand. This is repeatedly reported
by Theophylact Simocatta and in the sources on the Avar siege of
Constantinople in 626. After the siege failed, a Bulgar seems to have
laid claim to the throne of the Avar khagan, but was defeated in a
bitter civil war. 9,000 Bulgars with women and children fled to
Bavaria where the Frankish king Dagobert allowed them to spend
the winter. Soon, however, he ordered the Bavarians to massacre
them, and only a group of about 700 men, women and children
escaped under their duke Alciocus, a story that Fredegar tells at
some length.17 It seems that the same group later moved on, under
a leader of the same name, to the Lombard kingdom, where they
were accommodated in the duchy of Benevento.18 Other Bulgar
groups are mentioned in different peripheral regions of the kha-
ganate, especially on the Black Sea steppes, where Khan Kuvrat
formed a powerful but short-lived Bulgar empire in the middle of
the seventh century. A number of these Bulgars, under the leader-
ship of Asparuch, established an independent khaganate on the lower
Danube in the 680s.19
When the Avars invaded the Carpathian Basin in 567/68, several
groups already settled there were drawn into the Avar orbit. Most
notably, this included the Gepids whose kingdom the Lombards had
destroyed in 567. A minority of them joined the Lombards on their
way to Italy, some fled to the Byzantine Empire, but most of them

16
H. Wolfram, Die Goten. Von den Anfängen bis zur Mitte des 6. Jahrhunderts (3rd
edn., Wien-München 1990) pp. 276–7; P.J. Heather, The Goths (Oxford 1996) p. 162.
17
Fredegar, Chronicon 4,72, ed. B. Krusch, MGH SSrM 2 (Hannover 1888).
18
Paul the Deacon, Historia Langobardorum 5,29, ed. L. Bethmann and G. Waitz,
MGH SSrL (Hannover 1878).
19
J. Shepard, “Slavs and Bulgars”, The New Cambridge Medieval History, vol. 2:
700–900, ed. R. McKitterick (Cambridge 1995) pp. 228–48.
580  

remained settled in separate villages under Avar rule.20 They are

mentioned repeatedly as forces of considerable strength in Avar
armies. Not all Gepids, however, took part in war, even if it was at
their doorstep. In 599, when a Byzantine army victoriously marched
along the Tisza river, deep within the Avar heartland, they chanced
upon three Gepid villages. Although just a few days after a great
battle had been fought in the vicinity, the villagers were caught com-
pletely unawares. They had been feasting and were slaughtered by
the Byzantines in their sleep at dawn.21 The position of subdued
gentes in the Avar empire, therefore, was not necessarily the same for
all. Some Gepids evidently took the opportunity to join the khagan’s
army to improve their lot, while others lived peacefully as peasants.
Another group of (at least partly) pre-Avar inhabitants living under
Avar rule in Pannonia emerges from the archaeological material: the
Romano-barbarian population at Keszthely, who maintained their
own style of craft for centuries (see Falko Daim’s contribution).
Especially in the early period, the Avar empire provided room for
a variety of groupings and communities. Many of these component
groups seem to have been less coherent than under the rule of Attila,
who had encouraged royal leadership among his subject peoples,
such as the Goths, Gepids, Sciri and others, and thus reinforced
their identity. The archaeological evidence seems to reflect both the
common social and cultural space created by the khaganate and the
wide range of warrior groups it encompassed. Many of these groups
were rather small and mobile, as their scattered graves indicate. It
is important to note that cultural differences among the inhabitants
of the Carpathian Basin under the early Avars that emerge from the
archaeological record do not neatly coincide with ethnic divides
recorded in the written sources in some cases. Especially the spec-
tacular finds of Zamárdi or of the “Keszthely Culture” in Pannonia
do not express separate ethnic identities, although they differ from

20
W. Pohl, “Die Gepiden und die Gentes an der mittleren Donau nach dem
Zerfall des Attilareiches”, Die Völker an der mittleren und unteren Donau im fünften und
sechsten Jahrhundert, ed. H. Wolfram and F. Daim, Denkschriften der Österreichi-
schen Akademie der Wissenschaften, philosophisch-historische Klasse 145 (Vienna
1980) pp. 239–305. A monograph about Die Langobarden is in preparation. See also
id., “The Empire and the Lombards: treaties and negotiations in the sixth century”,
Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. id., The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997) pp. 75–134.
21
Theophylactus Simocatta 8,3,11–13.
   () 581

the standard early Avar culture that quickly spread across the Car-
pathian Basin and some adjacent areas.
Minorities and ethnic groups under Avar rule were neither fully
absorbed into an Avar culture and ethnic identity, nor did they sim-
ply remain as they were. The Gepids, for instance, are still attested
to at the end of Avar rule, after 250 years. Others became part of
new regional groups that were accepted to a certain degree, as a
remarkable story in the Miracula Demetrii, dated to the 680s, relates.22
The descendants of Roman prisoners that had been settled in the
old province of Pannonia achieved their freedom by participating in
Avar campaigns as warriors; then they mixed with Bulgars and Avars
so that in time they came to be recognized as a new people. Thus,
the khagan sent them a leader, the Bulgar Kuver. As the Romans
still nourished the hope of return to their ancient homelands in the
Balkans, the new group rebelled, and successfully emigrated to
Byzantine territory. There, they were perceived as Sermesianoi Bulgars,
the Bulgars from Sirmium. Kuver planned a surprise attack on the
city of Thessalonica which, however, failed, and the Sermesianoi dis-
persed; their core was accepted as a federate unit into the Byzantine
army. The story reveals an interesting paradox. The oral traditions
of the Romans living in captivity, “the knowledge of their origins
and the way of life according to the customs of the Romans”, served
as a motivating myth for the formation of a new group of steppe
warriors who broke away from the dominant steppe empire of the
region to form an independent power. They received a name from
the Byzantines who perceived them as Bulgars, as warriors who
left the Avar empire were often called. To affirm their position and
their identity, they needed a decisive victory. But even in defeat,
they represented a military force that was of value to the victorious
Byzantines who included them in their army. In many respects,
Byzantines and steppe empires played by the same rules. The ways
to control and direct military power and to integrate ethnic groups
were similar on both sides. The zone of contact and conflict was
not characterized by the neat juxtaposition of a universal Christian
empire and barbarian ethnic states. A group of Byzantines could

22
Miracula Sancti Demetrii 2,5, ed. P. Lemerle, 2 vols. [Les plus anciens recueils des
miracles de Saint Démétrius et la pénétration des Slavs dans les Balkans] (Paris 1979/81) vol.
1, pp. 222–34; Pohl, Awaren, pp. 278–81.
582  

become a gens among others and even be perceived as barbarians,

and Bulgars could be received in the Byzantine army although they
behaved like barbarians.

Avars and Slavs

The most important ethnic group to live under Avar rule were the
Slavs. This is also the most problematic example; the relationship
between Avars and Slavs was among the most fiercely-discussed top-
ics in older scholarship. Recently, debate has been reopened on the
fundamental question of the ethnogenesis and expansion of the Slavs.
The traditional view was that the Slavs are descended from an Urvolk
that for many centuries lived somewhere north-east of the Carpathians
and for some reason began to spread rapidly after c. 500 A.D. because
it had become so populous. Usually, these earliest Slavs are identified
with the Venethi that Tacitus locates along the eastern frontiers of
the Germania.23 At first sight, this seems very likely because German
speakers called (and in some regions still call) Slavs Ven(e)di, Wenden,
or similar. But the analogous German name for their western and
southern, Romance-speaking neighbours, Walchen, Val(a)chi, Wallones,
Welsh or the like, is derived from the Volcae, a people that had noth-
ing to do with these peoples. The hypothesis of an ancient Slavic
people that spread through migration has always appealed to Slav
nationalists, and is still widely held, but there is little sound evidence
to recommend it. Most of all, it fails to make plausible why a regional
and hitherto virtually unknown group could take over almost the
whole of eastern and east central Europe in a relatively short period
from c. 500 to c. 650 A.D. The extraordinary phenomenon that has
to be accounted for is that much of the eastern half of the conti-
nent became thoroughly Slavicized whereas the more spectacular

23
Even recently, P. Dolukhanov, The Early Slavs. Eastern Europe from the Initial
Settlement to the Kievan Rus (London-New York 1996), begins Slavic history in the
neolithic, in complete contradiciton to the sources that only mention them from
the sixth century A.D. For a history of views on Slavic origins see F. Curta, The
Making of the Slavs. History and Archaeology of the Lower Danube Region c. 500 –700
(Cambridge 2001) pp. 6–36; id., “From Kossinna to Bromley: ethnogenesis in Slavic
archaeology”, On Barbarian Identity-Critical Approaches to Ethnogenesis Theory, ed. A. Gillett
(Turnhout forthcoming); P. Barford, The Early Slavs (Ithaca 2001) pp. 1–26.
   () 583

migrations in western Europe did not lead to a Germanization of

larger areas, with the exception of a strip of land west of the Rhine
and south of the upper Danube.
Obviously, the key to Slavicization is not Slavic power. Slavs did
not spread by expanding their dominion but under the direct or
indirect rule of Avars, Bulgars and other lords. I have therefore tried
to explain the expansion of the early Slavs by the attraction of a
social and cultural model that implied a lower level of social differ-
entiation and surplus production than both the Byzantine Empire
and the barbarian kingdoms. Sixth- and seventh-century Byzantine
sources indicate that Slavs, unlike all others, did not enslave their
prisoners of war but allowed them to settle freely among them. There
is some incidental evidence that the inhabitants of Slavic villages
north of the lower Danube were of quite mixed origin. This way of
life may have appealed to the rural population in a wide area.24
Their produce had previously gone to the Byzantine state, the Church
and/or barbarian warlords, whose abilities to offer protection in
return had seriously declined.25 Linguistic Slavicization may have
only come later.26
A different, but not necessarily incompatible hypothesis has recently
been proposed by Florin Curta. In the sixth century, he argues, Slav
was a name given on the basis of wholesale perceptions by the
Byzantines. Slavic identity then developed from confrontation with
the Empire along the lower Danube; a common Slavic culture may
have only developed later. The idea that the Slavs were “invented”
by the Byzantines is perhaps harder to argue than the comparable
case that, according to a now widely held opinion, Caesar “invented”
the Germani as an all-inclusive group. Self-perception as Slavs is
attested rather early, whereas during the middle ages there is hardly

24
Pohl, Awaren, pp. 94 ff., esp. pp. 125–7; see Maurikios, Strategikon 11,4, ed.
G.T. Dennis, transl. E. Gamillscheg, Corpus Fontium Historiae Byzantinae 17 (Wien
1981); Theophylactus Simocatta 6,8.
25
See also J. Haldon, Byzantium in the Seventh Century (Cambridge 1990).
26
It is puzzling that linguistic reconstructions of Old Common Slavonic have
problems to reach back beyond the 8th century; see H.G. Lunt, “Common Slavic,
Proto-Slavic, Pan-Slavic; what are we talking about?”, International Journal of Slavic
Linguistics and Poetics 41 (1997) pp. 7–67; H. Birnbaum, “Von ethnolinguistischer
Einheit zur Vielfalt: die Slawen im Zuge der Landnahme der Balkanhalbinsel”,
Südostforschungen 51 (1992) pp. 1–19; Z. Go∑ab, The Origins of the Slavs. A Linguist’s
View (Columbus Ohio 1992).
584  

any proof for a “Germanic” identity.27 But Byzantine perceptions in

the sixth century did not necessarily correspond to Slavic self-
identification; the sources variously identify or link the Slavs with
Huns, Sporoi, or Getae, and treat Slavs and Venethi in some cases
as two different peoples and in other instances as one and the same.28
In any case, it is evident that Avar rule and the spread of Slavic
population were closely interrelated. On the one hand, the early
Avars depended for some of their subsistence on Slavic agriculture,
and on the other hand, they protected Slavic settlers from foreign
interference. Slavs also played an important part in Avar armies,
especially as foot soldiers and boatsmen. For quite some time, how-
ever, the Avar khaganate blocked, as far as it could, attempts to
form independent Slavic powers. On the lower Danube, this was an
interest they shared with the Byzantines (an Avar attack in 578, for
instance, was supported by Byzantium). In the course of the seventh
century, the decline of Avar rule made the formation of regional
Slavic powers on the western periphery of the Avar empire possi-
ble. The first one was Samo’s kingdom, probably in Bohemia; another
one was the Carantanian duchy in modern Carinthia and Styria.
The relationship between Avars and Slavs was thus dynamic and
oscillated between Avar dominion, competition and interdependence.
But still, Slavic peasants provided an agricultural basis for the Avars
which replaced the more differentiated late Roman infrastructure
preserved in other parts of the post-Roman world.

Gens and regnum

So what was the relationship between gens and regnum, or in this case,
gens and the khaganate, in the Avar empire? The problem is made
more complex by the wide range of meanings that ethnic terms had
in the world of the steppe. One and the same name, such as Huns

27
Pohl, Germanen; self-designation as Slavs is found, for instance, in the Russian
Nestorchronik 1,27–36, ed. L. Müller (München 2001) p. 21.
28
Sporoi and Huns: Procopius, Wars 7,14, ed. J. Haury, transl. H.B. Dewing,
vol. 3–5, Loeb Classical Library (London-Cambridge Mass. 1953–54); Getae:
Theophylactus Simocatta 3,4; 6,6; 7,2. Sclaveni and Venethi: Jordanes, Getica 23,119,
ed. T. Mommsen, MGH AA 5,1 (Berlin 1882); Sclavi coinomenti Vinedi: Fredegar,
Chronicon 4,48.
   () 585

or Avars, could have several meanings. In more or less the modern

sense, it could denote a steppe empire that Byzantine (or other) his-
torians knew by that name, which probably coincided with its self-
definition. Second, the name could indicate the ruling elite of that
empire who distinguished themselves in this way from their subjects.
Third, in a slightly more general sense, the name referred to the
armed and free population of that empire, except, perhaps, those
who belonged to established sub-groups who had joined the army
as recognizable units. So far, these three meanings more or less cor-
respond to the significance of ethnic terms in other regna (for instance,
the Franks).29 The name mainly indicated those who acted together
as Avars. But in the steppe, the same or very similar names could
also cover other groups in different parts of the steppe, and some-
how suggest a symbolic union of all those who had appropriated
this tradition. In this sense, it could also be interpreted as a blood
relationship, however fictional (as in the case of the Bulgar myth of
five brothers found in ninth-century Byzantine chronicles).30 Finally,
certain groups or families within a people could claim to be specifically
descended from an eponymous ancestor, such as the Chunni among
the Varchonitae/Avars, which probably implied a claim to higher
prestige. Ethnic identities pervaded the whole society vertically, from
ruling elites to subject groups, and the same names could be used
for self-designation or description of groups of very different size.
Even in early-medieval western Europe, ethnic groups were not peo-
ples in the modern sense, well-delineated units that integrated soci-
eties from top to bottom; for steppe peoples, such a notion would
be totally inappropriate.
Still, ethnic identity formed a fundamental part of the political
language, and was also used to define power and those who held it
on every level. The few short Avar runic inscriptions found so far
have (in spite of several attempts) not yet been deciphered, which is

29
I.N. Wood, “Defining the Franks: Frankish Origins in Early Medieval
Historiography”, Concepts of National Identity in the Middle Ages, ed. S. Forde, L. Johnson
and A.V. Murray, Leeds Texts and Monographs. New Series 14 (Leeds 1995) pp.
47–57; W. Pohl, “Franken und Sachsen: die Bedeutung ethnischer Prozesse im 7.
und 8. Jahrhundert”, 799—Kunst und Kultur der Karolingerzeit. Karl der Große und Papst
Leo III. in Paderborn. Beiträge zum Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann
and M. Wemhoff (Mainz 1999) pp. 233–6.
30
Nikephoros Patriarcha, Breviarium Historicum 33, ed. C. Mango (Washington DC
1990); Theophanes, Chronographia AM 6171, ed. C. de Boor, 2 vols. (Leipzig 1883).
586  

partially due to the fact that we cannot be certain what language

the Avars spoke (or used for inscriptions).31 Seventh- and eighth-cen-
tury Turkic and Bulgarian inscriptions indicate that power was pri-
marily understood as power over peoples. There is no doubt that
Byzantine and western observers also perceived it that way, and this
shared understanding was perfectly suitable for complex negotiations
with the Avars. In the case of the Avars, furthermore (and quite
unusually), the Avar khaganate and Avar identity were so closely
related that one could not exist without the other. If we follow
Theophylact’s report discussed above, the adoption of an ethnic
name, Avars, coincided with the elevation of a khagan. Avar iden-
tity expressed a political programme and a claim for power; it worked
in the way of a self-fulfilling prophecy. It is remarkable that we know
only a single khagan’s name, that of the first ruler Baian who seems
to have died shortly after the occupation of Sirmium in 582. All
other khagans were only known by their title, in Byzantine and west-
ern sources alike. This is not due to a lack of information; Theophylact
could otherwise go into surprising detail, and in some occasions gave
the names of even minor Avar commanders. This omission can only
mean that the ideology of Avar rulership obliterated the individual-
ity of the khagan; it was inconceivable that there was another kha-
gan. In the late Avar period, a whole hierarchy of titles is attested
in Frankish sources; with all of them, the title is used without a
name. Only the Christian petty khagans after 800 were mentioned
with their names, but by this time, these were Christian names. Even
though the hierarchy changed in the course of Avar history, it formed
the backbone of Avar identity to the end.
How directly Avar identity was linked to the khaganate becomes
evident from a number of observations. In a way specific to the
Avars, the khaganate retained a monopoly over the name. We have
no evidence for Avars outside the khaganate, and that is extremely
unusual. All other important ethnic groups of the period are also
attested to in the Roman and Byzantine army, but Avars are not.
There is one Byzantine officer with the same name as a comman-
der of an Avar army in the late sixth century, Apsich, but in Byzantium

31
Pohl, Awaren pp. 223–4; attempt to decipher Avar inscription: A. Róna-Tas,
“Die Inschrift des Nadelbehälters von Szarvas (Ungarn)”, Ural-Altaische Jahrbücher
N.F. 9 (1990) pp. 1–30; language of the Huns unidentifiable: G. Doerfer, “Zur
Sprache der Hunnen”, Central Asiatic Journal 17 (1973) pp. 1–50.
   () 587

he was called a Hun.32 This does not necessarily mean that no Avar
individuals joined the Byzantine side. There was even an Avar magus,
or high priest, the Bookalabras, who flew to Constantinople after an
affair with one of the khagan’s wives; Theophylact presented him as
“a Scythian”, noting that “he suborned seven of the subject Gepids,
and made his way towards his ancestral tribe. These are the Huns,
who dwell in the east as neighbours of the Persians and whom it is
more familiar for the many to call Turks.”33 If the Bookalabras had
intimate access to one of the khagan’s wives, he must have belonged
to the Avar core group; however, Theophylact carefully avoided call-
ing him an Avar, but, within a short narrative, referred to him as
a Scythian, Hun, and Turk. The traditional labels Huns and Scythians
were, of course, repeatedly used for the Avars; yet, characteristically,
in the following passage of the Chronicle Theophylact presented the
Avar ambassador Targitios simply as an Avar.
A similar phenomenon may be observed with warriors and peas-
ants who emigrated from the Avar empire. Whoever migrated from
the Avar empire came to be called a Slav if he settled in a rural
community in the Slavic manner, for instance in the Peloponnes or
in Dalmatia. If he lived as a warrior, conversely, he was perceived
as a Bulgar. This is the case of the Alciocus/Alzeco group, or of
Kuver’s “Sermesianoi Bulgars” who included many Avars, as the
Miracula Demetrii remembered; no Avars appear in the narrative after
the group left the khaganate.34 When, however, within the Avar
empire, a Bulgar aspired to the office of the khagan, it was still the
Avar khaganate (like the Rugian Erarich who became king of the
Ostrogoths in the Gothic war).35 After the Avars were subdued by
the Franks, there were repeated efforts by Avar leaders to restore
the honor antiquus of the khagan with Frankish support; and indeed,
for a while, the Carolingians sustained a Christian khaganate in
Pannonia. But when these efforts failed, the name disappeared com-
pletely from the sources. After 822, apart from reminiscences, there
is no trace of the Avars.36 This does not mean that the whole Avar

32
Theophylactus Simocatta 1,14,5; 2,3 (Apsich the Hun); cf. 8,5 (Apsich the
Avar).
33
Ibid. 1,8; transl. Whitby, The Emperor Maurice and his Historian, p. 30.
34
Pohl, Awaren pp. 99–100; 243–5; 278–81.
35
Fredegar, Chronicon 4,72; Procopius, Wars 7,2.
36
Pohl, Awaren, pp. 320–8.
588  

population had been eradicated, but that Avar identity disappeared

as it had come: with the “ancient honour” of the khaganate. “Gone
like the Avars”, seems to have become proverbial among the Byzantines.
Thus, it is likely that Avar identity served as an important cohesive
factor within the Avar empire, but had no independent existence
without it in central and eastern Europe.

Changes and continuity in the khaganate

Archaeological evidence can be interpreted as the remains of a sym-

bolical language used to express identity (ethnic or other), although
with great methodological caution. The Avars, with their extraordi-
nary wealth of archaeological evidence, are in fact an interesting test
case for this approach.37 Current debate on the possibility to inter-
pret archaeological cultures as indicators for ethnic identities has
become rather controversial in recent years, especially in German-
speaking countries. The methodological principle put forward by
Gustaf Kossinna around 1900 that archaeological cultures necessar-
ily corresponded to ethnic units has long been abandoned for a more
pragmatic approach, although often with insufficient methodological
debate. Around 2000, we are faced with the opposite assertion by
some scholars: that archaeological cultures tell us nothing about eth-
nicity.38 To spur methodological debate and overcome schematic
arguments, this contention has to be taken seriously. But so far, it

37
F. Daim, “Archaeology, ethnicity and the structures of identification. The exam-
ple of the Avars, Carantanians and Moravians in the eighth century”, Strategies of
Distinction. The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Rei-
mitz, The Transformation of the Roman World 2 (Leiden-New York-Köln 1998)
pp. 71–94.
38
S. Brather, “Ethnische Identitäten als Konstrukte der frühgeschichtlichen
Archäologie”, Germania 78 (2000) pp. 139–77. Some previous positions: H.J. Eggers,
Einführung in die Vorgeschichte (München 1959); H. Härke, “All quiet on the western
front? Paradigms, methods and approaches in West German archeology”, Archaeological
Theory in Europe. The Last Three Decades, ed. I. Hodder (London-New York 1991) pp.
187–222; V. Bierbrauer, “Die Landnahme der Langobarden in Italien aus archäolo-
gischer Sicht”, Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters.
Methodische Grundlagendiskussion im Grenzbereich zwischen Archäologie und Geschichte 1, ed.
M. Müller-Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen
1993) pp. 103–72; U. Veit, “Ethnic concepts in German prehistory: a case study
on the relationship between cultural identity and archaeological objectivity”, Archaeological
Approaches to Cultural Identity, ed. S. Shennan (London-New York 1989) pp. 35–56;
   () 589

is mainly based on the cases of Franks and Alemanni where identi-

fication is indeed difficult, and a clear delineation should not be
expected given the rather blurred and dynamic ethnic identities on
both sides of the Rhine in the fourth and fifth centuries.39 The Avars
are a different case. We have quite precise information on the south-
ern and western limits of the Avars’ dominions, and about the core
area of settlement in the plains of the Carpathian Basin; the chrono-
logical frame of the khaganate between 568 and 796 (with about a
generation of scattered survivals) is also well established. The archae-
ological culture attested in this area at that given time is rather uni-
form in many respects, and can be connected with the Avar khaganate
beyond any reasonable doubt.
Having said that, a few caveats have to be kept in mind. First,
“Avar” archaeological culture cannot simply be identified with the
extension of Avar ethnic identity in any strict sense; neither does it
help to draw the boundaries of the Avar realm. Especially at the
margins, finds of “Avar” type materials cannot serve as proof that
individuals or local groups inhumed with them considered themselves
as Avars, or lived under the khagan’s rule. This is especially impor-
tant for the northern and eastern periphery of the khaganate, where
cultural influence and political power did not necessarily coincide.
Falko Daim has demonstrated that the eighth-century “Avar” objects
found in a Slavic context at Krungl and Hohenberg in modern
Austria cannot serve as indicators of Avar rule or identity; some of
them are in fact of Byzantine origin, but mainly preserved in Avar
graves.40 Second, within the cultural sphere of the khaganate, there
was a considerable degree of diversity within the overall frame, espe-
cially in the early period, which is indicative of, but by no means
co-extensive with, groups of non-Avar origin living under Avar rule.
Third, the distribution of artefacts over wide areas need not denote

F. Siegmund, Alemannen und Franken, F. Siegmund, Alemannen und Franken, Ergän-

zungsbände zum Reallexikon der Germanischen Altertumskunde 23 (Berlin-New
York 2000).
39
Pohl, Völkerwanderung, pp. 165–75.
40
F. Daim, “‘Byzantinische’ Gürtelgarnituren des 8. Jahrhunderts”, Die Awaren
am Rand der byzantinischen Welt. Studien zu Diplomatie, Handel und Technologietransfer im
Frühmittelalter, ed. id., Monographien zur Frühgeschichte und Mittelalterarchäologie
7 (Innsbruck 2000) pp. 77–204; id., “Byzantine belts and Avar birds. Diplomacy,
trade and cultural transfer in the eighth century”, The Transformation of Frontiers—
from Late Antiquity to the Carolingians, ed. W. Pohl, I.N. Wood and H. Reimitz, The
Transformation of the Roman World 10 (Leiden-Boston 2000) pp. 143–88.
590  

actual migration, but can result from a variety of forms of commu-

nication that may be maintained over several generations.41 And
finally, as Guy Halsall has stressed, ritual and representation used
in burial often do not set out to demonstrate large-scale ethnic iden-
tities but rather social self-identification on a local level, especially
among groups competing for prestige and power.42 This need not
mean that ethnicity had no part in their symbolic competition, but
it should warn us that the symbolic language of grave-goods and
burial practices may be rather complex.
All of this does not rule out conceiving of a cultural history of
the Avar empire on the basis of material remains; on the contrary,
it makes such an effort much more interesting and complex because
it allows us to reconstruct some of the ambiguities and contradic-
tions that resulted in a dynamic development of Avar society. Culture
should not be treated as a separate phenomenon but as an expres-
sion of society as a whole. This also means studying how Avar cul-
ture was linked to political power. If we follow Niklas Luhmann’s
striking definition, power is communication regulated by codes.43
These codes are a cultural achievement, and so is their communi-
cation. In the case of the Avars, this language of power has left rel-
atively few traces in written texts, for the speeches by the khagan
and his envoys in the chronicles of Menander and Theophylact are
literary creations, although surely on the basis of long experience
with barbarian neighbours. Some of it can be reconstructed from
the Avars’ political actions and claims. But artefacts and rituals also

41
H.W. Böhme, “Kontinuität und Traditionen bei Wanderungsbewegungen im
frühmittelalterlichen Europa vom 1.–6. Jahrhundert”, Archäologische Informationen 19
(1996) pp. 89–103; H. Steuer, “Archäologie und germanische Sozialgeschichte.
Forschungstendenzen in den 1990er Jahren”, Runische Schriftkultur in kontinental-skan-
dinavischer und angelsächsischer Wechselbeziehung, ed. K. Düwel (Berlin-New York 1994)
pp. 10–55.
42
G. Halsall, Early Medieval cemeteries: introduction to burial archaeology in the post-Roman
West (Glasgow 1995); F. Theuws and M. Alkemade, “A kind of mirror for men:
sword depositions in Late antique northern Gaul”, Rituals of Power. From Late Antiquity
to the Early Middle Ages, ed. F. Theuws and J.L. Nelson, The Transformation of the
Roman World 8 (Leiden-Boston-Köln 2000) pp. 401–76; see also B. Effros, “Dressing
conservatively: a critique of recent archaeological discussions of women’s brooches
as markers of ethnic identity”, Gender and the Transformation of the Roman World, ed.
L. Brubaker and J. Smith (Cambridge forthcoming), for a discussion of the gender
perspective.
43
N. Luhmann, Macht (Stuttgart 1975) p. 15: “Macht ‘ist’ codegesteuerte
Kommunikation”.
   () 591

formed part of the communication in which power and identity were

constantly re-negotiated within the Avar khaganate, and towards
outsiders.
On the basis of such an approach, Falko Daim’s synthesis in this
volume provides interesting insights into the changes and develop-
ments in the khaganate. In spite of the methodological problems of
harmonizing archaeological periodization with what is known about
the history from written sources, the years 567/68 in the Carpathian
Basin evidently constitute a clear terminus post quem for Avar burials.
The basic continuity of the cultural language that drew the Avar
“koiné” together corresponds to the continuity of the political sys-
tem. Just as would be expected from historical sources, some ele-
ments of this archaeological culture can be traced as far as central
Asia, many came from the steppes north of the Black Sea, but their
synthesis was only achieved in the Carpathian Basin. Before 568,
the Avars were a large but precarious accumulation of ambitious
warriors temporarily united by a common purpose and following a
leader whose victories had won him great prestige. Many of these
victories, however, had been more symbolic than decisive, such as
against the Franks near the Elbe in 566. In this case, the Franks
had to supply the Avars with food after their defeat, a condition
indicative of the difficulties the Avars found themselves in at the
time.44 Only settlement in the Carpathian Basin provided the basis
for the construction of a stable Avar power.
It took more than a decade until the khaganate was well-estab-
lished and had subdued the barbarians in and around the Carpathian
Basin. Archaeological evidence supplements this picture in interest-
ing ways. The Avars used Roman roads and remaining infrastruc-
ture (although they do not seem to have lived inside Roman towns
and buildings). According to what has been found so far, larger set-
tlements, if at all, were more common in Pannonia. The khagans
accepted ethnic and cultural diversity and communication with neigh-
bouring countries. This was not simply a case of tolerating a regional
subject population that continued its lifestyle. The Keszthely culture,
as the most obvious case, shows that a new Christian culture with
late Roman and western barbarian flavour was actively encouraged

44
Gregory of Tours, Historiae 4,23; 4,29, ed. B. Krusch and W. Levison, MGH
SSrM 1,1 (2nd edn., Hannover 1951).
592  

in the region south-west of Lake Balaton; it was probably even rein-

forced by the settling of Christians from different parts in- and out-
side the khaganate there. Western influences, especially from Lombard
Italy, continued. These cultural contacts correspond with the stable
alliance with the Lombards and with the relatively peaceful rela-
tionship maintained with the Franks. Avar culture in the narrower
sense was not homogeneous, either. Most notably, there is a great
variety in pottery styles; types of houses and settlement patterns also
varied until the end of the khaganate. Burial custom (for instance,
the layout of cemeteries) likewise showed considerable regional
differences. One example were the gallery graves (Stollengräber) in the
eastern half of the Carpathian Basin, a regional custom maintained
into the eighth century. These examples indicate that a homoge-
neous Avar culture was not created “bottom up”, but rather “top
down”. It did not involve every-day culture as much as prestige
objects, weapons and ornaments.
Crucial to an understanding of the Avar “culture of power” is the
relationship of the khaganate with Byzantium, well-attested in writ-
ten sources up to 626. Negotiations to settle the Avars in a Byzantine
border province before 567 remained inconclusive; after the Avars
had occupied Pannonia, integration in Byzantine territory was not
an issue any more. Khagan Baian’s territorial claims for the city of
Sirmium had strategic reasons; when Sirmium was taken after a long
siege in 582, no attempt appears to have been made to maintain
Roman-style civic life, and soon the city was more or less aban-
doned. Subsistence in the Carpathian Basin was based upon seden-
tary agriculture (not least, by Gepids and Slavs) and pastoralism.
The Avars thus remained firmly outside the Roman order.45 Still,
they could not have done so without the help of Byzantium. Yearly
subsidies, presents, ransom for prisoners and plunder brought about
an enormous influx of riches. Archaeological evidence attests to this
“Golden Age” in the Carpathian Basin. Finds of Byzantine coins are
relatively rare (mostly as single coins deposited as “oboli” in graves).46

45
See also A. Avenarius, “Struktur und Organisation der europäischen Steppen-
völker”, Popoli delle steppe: Unni, Avari, Magiari, Settimane di studio del centro ital-
iano di studi sull’alto medioevo 35 (Spoleto 1988) pp. 125–50.
46
É. Garam, “Die münzdatierten Gräber der Awarenzeit”, Awarenforschungen 1,
ed. F. Daim, Archaeologia Austriaca 1. Studien zur Archäologie der Awaren 4
(Wien 1992) pp. 135–250.
   () 593

Byzantine high-status artefacts, by contrast, are much more com-

mon, but a considerable part of the precious metals seems to have
been melted down and used as raw materials by Avar and other
goldsmiths.47 The distribution of these riches seems to have remained
chiefly under the control of the khagan. We do not have any writ-
ten evidence for the way in which precious objects circulated among
the Avars, but may safely hypothesize that gift exchange played an
important part. We do know, however, that the khagans remained
in control of a huge treasure until 795/96, when Charlemagne’s
armies seized most of it. Even a generation later, Einhard described
the Avar treasure, which was “collected over a long time”, and noted:
“Human memory cannot remember that in any war fought against
the Franks these were more enriched by booty”.48
Cultural developments in the seventh century, when written evi-
dence becomes patchy, may be linked with some confidence to his-
torical events, especially to the crisis that ensued after the siege of
Constantinople in 626. This defeat resulted in civil war (the “Bulgar”
revolt) and was in some way connected with the secession of Samo’s
Slavs, most likely, in Bohemia and perhaps also in Moravia, and of
those in Carantania.49 Many other steppe empires collapsed in such
a situation, whereas the Avar empire was quite successfully trans-
formed into a sedentary state for almost another two centuries. The
Avar defeat also meant the end of the substantial influx of Byzantine
subsidies (200,000 gold solidi in the years before 626); but the kha-
gans kept control of the treasure acquired up to this point. It seems
plausible to assume that without this treasure, there would have been
no khaganate. The Avars are thus a curious case of a peripheral
society that had lost most of its ties to the Mediterranean world
upon which it had once depended, but still relied on the gains and
exchanges from a distant past. For the cultural consequences of the
crisis, Falko Daim has a very attractive hypothesis to offer in his
contribution to this volume. In the second quarter of the seventh
century, high-status objects used in “princely” burials such as Bócsa

47
A. Kiss, “Die Goldfunde des 5.–10. Jahrhunderts im Karpatenbecken. Angaben
zu den Vergleichsmöglichkeiten der schriftlichen und archäologischen Quellen”, Acta
Archaeologica Academiae Scientiarum Hungaricae 18 (1986) pp. 105–45.
48
Einhard, Vita Karoli 13, ed. O. Holder-Egger, MGH SSrG 25 (6th edn.,
Hannover 1911).
49
Pohl, Awaren, pp. 256–68.
594  

and Kunbabony maintained their quality, but aimed for a distinc-

tively non-Byzantine language of representation.50 From the middle of
the century, Byzantine objects returned to high-status graves (such as
Ozora-Tótipuszta). The expedition of Constans II to Italy, but espe-
cially the foundation of the Bulgar empire on the lower Danube, may
have contributed to renewed Byzantine interest in the Avars, although
communication never reached the intensity it had before 626.
Written evidence for the whole period from 626 until the late
eighth century is scarce, and transformations in the Carpathian Basin
may only be reconstructed on the basis of archaeological evidence.
Although regional diversity faded, it did not disappear; permanent
settlements and cemeteries grew and indicate that the Avars adopted
a sedentary way of life. Communication with the outside world did
not stop altogether; some contacts (for instance, political exile of
Lombards at the Avar court) are attested by written evidence, oth-
ers (for instance to the Bulgar khaganate) by material traces. It would
be fascinating to know what caused the changes in the symbolic lan-
guage used in the eighth century; most notably, the appearance (and
disappearance) of the griffin motif and of the animal combat scenes.
These were worn on buckles throughout the khaganate and can lit-
erally be found in hundreds of graves. It is tempting to assume that
they constituted an emblem connected with rulership ideology and/or
ethnic identity; the griffin motif is attested in the culture of the steppe,
but also well-known in Mediterranean art.51
In open contrast to its non-expansionistic, sedentary agricultural
way of life, the political language underlined the steppe character of
the khaganate. When the Franks arrived at the end of the eighth
century, they encountered a number of titles and offices of central
Asian origin: among them tudun, tarkhan, kapkhan, iugurrus, and the
title of the queen, catun. This type of hierarchy was unknown to the
well-informed Byzantine sources of the earlier period, and we do not
know when the new formal hierarchy was introduced. In any case,
it means that right up to the end, the Avar state followed central

50
E.H. Tóth and A. Horváth, Kunbábony. Das Grab eines Awarenkhagans (Kecskemét
1992).
51
F. Daim, “Der awarische Greif und die byzantinische Antike”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. id. and H. Friesinger, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 273–303.
   () 595

Asian models and maintained a political structure that was com-

pletely different from that of its central European neighbours. Only
in the Bulgar empire, some similar titles (but otherwise, a different
hierarchy, with “khan” as the ruler’s title) are attested to in the
sources.
The military strength of the Avar empire, however, had not been
maintained in the long period of its relatively peaceful existence.
Defeat against the Franks, or rather, the incapability to put up a
serious defence, delegitimated the rule of the Avar gens so that the
khaganate disappeared altogether. Whether Charlemagne’s wars exactly
coincided with the end of Avar cemeteries, may perhaps be debated,
as the synchronisation of archaeological evidence with the history of
events in general. Especially in Pannonia, however, it does not seem
plausible to assume that Avar ways of life could have continued
undisturbed much after 800. The achievements of the khaganate are
evident from the power vacuum that ensued in the Carpathian Basin
throughout the ninth century.52 The “people that remained from Huns
and Slavs in those parts”, as the Conversio Bagoariorum et Carantanorum
puts it,53 could not be organized in a sufficiently stable manner in
spite of the efforts of the Franks. Bulgars and Moravians intervened
repeatedly but could not establish lasting control either. Carolingian
order in Pannonia finally collapsed when the Magyars arrived in the
late ninth century and established another steppe empire in the
Carpathian Basin. Once again, their state followed a non-Roman
model, but relied on the riches of post-Carolingian Europe. Only
gradually did Christian Europe come to control its ancient periph-
ery of warriors.

52
Overview in H. Wolfram, 378–907: Grenzen und Räume. Geschichte Österreichs vor
seiner Entstehung, Österreichische Geschichte 3 (Wien 1995) pp. 211–75.
53
Conversio Bagoariorum et Carantanorum 6, ed. F. Lo“ek, MGH Studien und Texte
15 (Hannover 1997): populum qui remansit de Hunis et Sclavis in illis partibus.
This page intentionally left blank
 REGNA AND GENTES: CONCLUSION1

Inquiring into the relationship between the early medieval gentes and
their (corresponding) regna, as has been attempted in this volume for
each people and kingdom separately, is no doubt a crucial, but
difficult enterprise. It is of importance because, at least in the eyes
of the Roman contemporary authors, there were “peoples”, and there
were “kingdoms” that were named after them; this close, but com-
plex relationship between these two historical phenomena has, how-
ever, never been analysed more closely before. It is nevertheless a
difficult and problematic question. Hence, it is no less complicated
(and should be regarded as a preliminary venture) to attempt a com-
parative conclusion. It may be prudent, therefore, to recall the inher-
ent problems first.

1. Some inherent problems

One of the problems concerns the terminology (modern and medieval)

of the two central themes of this book. What is meant by a “peo-
ple”? and what by a “kingdom”? The first question (“what is a peo-
ple?”) is a relatively obvious problem nowadays as we have become
aware of the historicity of peoples, and have hence become reluc-
tant about acknowledging constant ethnic elements.2 It seems to be
even more difficult in Germany where, on the one hand, Volk has be-
come burdened with ideologically incriminating connotations (though,
in the final analysis, there is no real alternative to continuing to use

1
A first draft and the final version of this conclusion was conceived by Hans-
Werner Goetz. It owes much to both the group’s discussion in Bellagio and to the
helpful remarks and additions made by Jörg Jarnut and the other contributors and
respondents, particularly Evangelos Chrysos, Bonnie Effros, Wolf Liebeschuetz, Walter
Pohl, Ian Wood and Barbara Yorke. I am grateful to Salina Braun, Barbara Cox
and Ian Wood for stylistic advice. Responsibility for contents and wording of the
final version and for any insufficiencies remain my own.
2
Cf. the introduction, but also the introductory remarks of Wolf Liebeschuetz’s
contribution.
598 

this term), and on the other hand, there are subtle nuances in the
terminological differentiations between Stamm (tribe), Volk (people)
and Völkerschaften (an untranslatable expression). At least it it no longer
common practice to speak of “tribes” with regard to the early medieval
peoples because this would imply an “archaic” state, but also that
“tribes” are still considered as part of an entire “people”. Another
aspect of this problem is the relationship between “peoples” and
“nation”. Whereas previous research has presupposed that peoples
were the original communities and, consequently, that a “nation”
was the (inevitable) product of the existence of a people, one of the
results of a (predominantly German) project on the so-called “birth”
of the European nations has been the recognition that it may well
have been the other way round, that is, (new) peoples originating
from the existence of (political) nations.
The other question (“what is a kingdom?”) is not as unreasonable
as it may at first appear, and becomes more apparent if we con-
sider the “date” of the origin of a kingdom (since there was no
official proclamation, as there was, for example, in the establishment
of a modern German Empire in 1871): from what time on, then,
are we allowed to speak of a (“Germanic”) kingdom, or: when did
a community begin to be a kingdom? With regard to the Franks,
for example, this raises the question whether it was not until Clovis
that the first kingdom emerged, or whether there were Frankish king-
doms before his reign (since we know that Roman authors refer to
Frankish kings previous to Clovis)? It will probably be impossible to
find common definitions for “people” and “kingdom” that all schol-
ars agree upon. For simplicity’s sake we may just call the (somehow)
“ethnic” and political bodies of the early medieval communities “peo-
ples” and “kingdoms”, whilst constantly bearing in mind the ambi-
guity of these terms. This coincides with the equally ambivalent and
wide-ranging meanings of Latin terms (such as gens or natio, or even
regnum), and, of course, there is a vast gap or shift of meaning between
early medieval and modern terminology. In considering “peoples”
and “kingdoms” in the transformation of the Roman world, we have
to consider both: early medieval and contemporary terms (and the
way they relate to each other).
A second, closely related problem is the use and meaning of the
terms “ethnicity” and “ethnogenesis” which are seen nowadays to denote
a process, developing on a political and ideological level and sub-
  599

ject to constant changes.3 Reinhard Wenskus’s theories have mean-

while been met with almost general consent, though it has to be
acknowledged that they are still embedded in a certain German tra-
dition which exaggerates the “Germanic” character of those peo-
ples.4 However, it has been accepted meanwhile that (ethnic) origin
is just one element of ethnicity among others, such as language, cul-
ture, political cohesion, name, and identification, and that political
factors should not be underestimated. The terms for most gentes
(Franks, Goths, Vandals, Alamans, or Saxons) were “collective terms”
for a “multiethnic” conglomeration. Another concept that had once
been widely accepted, but is now being disputed again, is that a gens
was built around some “kernel” (which Wenskus called a “nucleus
of tradition”, Traditionskern), whereas it is very difficult indeed to dis-
cern what this “nucleus of tradition” actually was. It should by no
means be understood as a stable, unchanging myth, but has to be
determined by looking into each the respective situation. Our prob-
lem is that we do not know anything substantial about possible “pre-
ethnographic, non-Roman traditions” (W. Pohl).5 Maybe, though, as
Wolf Liebeschuetz warns, we have moved too far from the aspect
of ethnicity as common descent (or at least the common belief in
it), a factor that no doubt had been overemphasized for a long time,
particularly, but by no means exclusively, in Germany, and perhaps
there is a tendency nowadays to minimize the relevance of ethnic
continuity and of distinguishing between the various gentes. At least
we should not forget that the early medieval authors believed in the
common descent of their peoples. The problem, however, remains
in finding methods or indicators on which a concept of ethnicity
could be based. It is difficult, for example, to investigate the “psy-
chological” depth of “ethnicity”, the consciousness of (a certain) sol-
idarity which no doubt played an important part. Meanwhile, we
are also reluctant to assume that language was a reliable indicator;

3
Cf. the introduction, p. 8 above.
4
Wenskus’ theory of ethnogenesis is, in fact, a general one and not one specific
to “Germanic” ethnogenesis, as is illustrated by the contributions to non-Germanic
peoples in this volume.
5
For the problem, see Walter Pohl, “Ethnicity, theory and tradition: a response”
(forthcoming), who meanwhile rejects the theory of a “nucleus of tradition” com-
pletely. This position solves the inherent ideological problem, but it does not solve
the problem of how to explain the coherence of these new expanding peoples.
600 

when (new) “peoples” were formed from parts of other peoples, there
must have been, at least for a while, a transitional period during
which several languages were spoken. Finally, it seems almost impos-
sible (and is sometimes ruled out altogether) to determine ethnicity
from archaeological findings6 which mainly comprise grave goods
and reflect only a small fragment of a past reality. Archaeology can
only define “cultural groups”, and it has become questionable today
(though it is still disputed) whether cultural groups may (if only very
cautiously) be identified with peoples (an identification that, in the
tradition of Gustav Kosinna, had been presupposed by former research).
The distinguishing of gentes by archaeological means will only become
possible if we define ethnicity according to archaeological criteria
(though this would not facilitate a comparison of archaeological results
with those of linguistic and historical studies). Moreover, “culture”,
as recognizable in archaeological findings, was not a constant value,
but a dynamic process (M. Schmauder). Changes in the material
culture of the Avars, for example, cannot simply be explained by
the migration of new groups (F. Daim). On the whole, the difficulty
seems to lie in the fact that we are presupposing the existence of
gentes when at the same time we are investigating their formation
and their ethnogenetic background.
A third problem is the use and meaning of “Germanic” in this period.7
The more we have reflected upon this problem, the more our uncer-

6
Cf. G. Halsall, “Social identities and social relationships in early Merovingian
Gaul”, Franks and Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N.
Wood, Studies in Historical Archaeoethnology 3 (San Marino 1998) pp. 141–65;
S. Brather, “Ethnische Identitäten als Konstrukte der frühgeschichtlichen Archäologie”,
Germania 78 (2000) pp. 139–77; S. Lucy, The Anglo-Saxon Way of Death (Stroud 2000).
Distribution maps of individual goods do not in themselves prove ethnic distinc-
tion, as is still often assumed and as M. Schmauder rightly observes.
7
There has been a long discussion in our group whether we should still con-
tinue to speak of “Germanic” peoples and kingdoms while being fully aware of the
problems this poses: first, the misuse of so-called Germanic values in former German
historiography, second, the result of modern research on ethnogenesis teaching us
that the so-called Germanic peoples (as others) were mixed populations, third, con-
sequently, the insight that nobody is able any more to define what “Germanic”
really means, and fourth, and most convincingly, that in this volume we are not
dealing exclusively with “Germanic” kingdoms. Naturally, these cannot be included
under this term. Personally, however, I (H.-W. Goetz) reject the alternative term,
“barbarian”, because it would just replace one ideology with another one. I also
am reluctant to follow the modern trend of abandoning words once they have
become problematic in one way or another. To replace them by other terms is no
solution. I would rather advocate that we become aware of the problems inherent
  601

tainty as to the actual meaning of the term “Germanic” has increased.

Walter Pohl, therefore, begins his latest survey on this subject with
the sentence: “A people that called itself Germanic possibly never
existed.”8 This term may be applied, of course, to the language, and
certainly the members of those peoples spoke their own vernacular
language which may have given them a certain sense of unity, but,
unfortunately, we have very little evidence of the language(s) used
in the kingdoms considered here. It is impossible to say, therefore,
which role the Germanic language played in the creation of the
regna. And whereas former research (again mainly in Germany) has
emphasized the “Germanic” character and mentality of the “bar-
barian” kingdoms and their constitution, recent studies have become
wary of such an assumption: most elements that have been classified
as being “Germanic” (for example the Eigenkirchenwesen, the Gefolg-
schaft and fidelity, or the “sib”, or “sept”) have turned out to be
widespread among the early medieval societies—regardless of diffe-
rent ethnic origin, and many developments that were typical of the
“Germanic” early Middle Ages had in fact already started in the
Late Antique Roman Empire. A comparison between “Germanic”
and other (“barbarian”) kingdoms, as, to a certain degree, has been
intended here (namely with regard to the Avars, Celts, and Islamic
Spain), seems to verify such an opinion. However, even in this respect
it is important to remain aware of certain differences: the Avar world,
as far as our questions are concerned, reveals significant similarities
to the “Germanic” kingdoms, though it had a completely different
political structure that followed central-Asian models (W. Pohl). In
the case of the sub-Roman Britons, however, there were even con-
siderable differences between the lowlands (with no kings at all, or
at least none which are recorded as surviving into the seventh cen-
tury) and the highlands, including Cornwall, Wales, the Pennines,
Rheged and Strathclyde with sizeable kingdoms that lasted into the
seventh century or even longer (A. Woolf ). Moreover, we acknowl-
edge anew the role and function of the Roman tradition—and of

in (all of ) these terms. Thus, in so far as these peoples might be considered as

being “Germanic” by their language and tradition, and also by a long historio-
graphical custom, I prefer to maintain the traditional terminology using quotation
marks to indicate the problems that are discussed in this passage.
8
W. Pohl, Die Germanen, Enzyklopädie deutscher Geschichte 57 (München 2000)
p. 1. For the ideological use and development of the term “Germanic”, cf. ibid.,
pp. 1–7 and 45–65.
602 

Christian ideology—in establishing and stabilizing the early medieval

kingdoms, an aspect to be reverted to in a later passage. Apart from
an older German “germanophile” tradition, there are also ideological
“barriers” elsewhere which have to be refuted as being unhistorical.
One example is the tendency of Spanish scholars to consider the Visi-
gothic kingdom as the beginning of a Spanish “nation” (I. Velázquez),
which has strong parallels to the older German conviction of having
developed from entirely “Germanic” origins.
A fourth (and probably not the last) problem lies in our sources.
We have to accept the fact that while comparing the different peo-
ples and kingdoms, first, we are relying on completely different evi-
dence for each, second, the sources do not lead us to a “past reality”,
but are reflections of the personal and subjective perception of their
authors, and third, until the end of the fifth century, our informa-
tion is based almost exclusively on Roman sources which provide us
with Roman perspectives, but lack any insight into how those peo-
ples may have perceived themselves. Sources deriving directly from
these peoples were written considerably later and do not provide
contemporary insights; they are retrospective accounts dating from
a period when the kingdoms were firmly established, such as Gregory
of Tours’ chronicles about the Franks, written about 590, or, as an
extreme case, Paul the Deacon’s reports about the Langobards, writ-
ten probably shortly before 796 (and even these early medieval writ-
ers continued to use the same terminology, although not necessarily
with exactly the same meaning). And, of course, in each case we
have to bear in mind the intent behind each script (causa scribendi )
and the “horizon of knowledge” of each chronicler or author. A
comparable problem is inherent in the archaeological evidence which
consists mainly of grave goods, whereas there is comparatively little
evidence which could throw a light on the settlements themselves.
Grave goods, however, are as much a reflection of the concept of
“the other world” as they are a mirror of earthly society. Moreover,
archaeological evidence, rather than giving an impression of stabil-
ity, is equally suggestive of dynamic changes (M. Schmauder).
Resulting from all these problems, the central question of this vol-
ume cannot be answered directly but has to be tackled from different
angles and by different approaches. From an academic perspective,
it can be seen from the historian’s, the archaeologist’s and the lin-
guist’s point of view. Most of the contributors to this volume are
historians, some are archaeologists (and some both), two are linguists,
  603

and it is a great help that at least some regna (Franks, Avars) have
been dealt with by both historians and archaeologists. The problem
of language and also place-names has been a central point in at least
two papers (I. Velázquez, A. Woolf ), but has been omitted or only
marginally touched on in the others. With regard to the subject-
matter, there are several “levels” which lend themselves to dealing
with the problem: first, the political level of the historic develop-
ment; second, the level of peoples and ethnogenesis; third, the level
of terminology and contemporary perceptions of the authors of our
sources; and fourth, the level of modern theory (of ethnogenesis).
Although these levels cannot be separated completely in the process
of research, it is advisable to remember which level we are talking
(or writing) about at any moment. Finally, a “conclusion” naturally
and necessarily tends to generalize. But we should not forget that
regnum had a history of its own, or, as Ian Wood writes, “each state
was created in different circumstances [. . .] and [. . .] ended in
different circumstances”.

2. The relationship between gentes and regna I:

Were there peoples before kingdoms?

Judging by the names of peoples used in the Roman sources, we

might get the impression that some of the “Germanic” peoples dealt
with here, such as the Vandals, the Langobards and the Goths, go
back a very long way (having already been mentioned by Tacitus
and his contemporaries). In these cases, however, we know that far-
ranging migrations took place between the first reference in the first
or second century and the establishment of kingdoms in the early
fifth (Goths and Vandals) or even sixth century (Langobards). It has
been believed—both in the Middle Ages and in modern times—that
(the) Goths had wandered possibly from Scandinavia to the Black
Sea and later on to southern Gaul and Spain (Visigoths) or to Italy
(Ostrogoths) respectively; (the) Vandals migrated from Silesia to Spain
and, after a brief period of 20 years, further to Africa, (the) Langobards
moved from the lower Elbe to the Danube and, soon afterwards, to
Italy, and there were no doubt migrations from the Continent to
England. In all these cases, we may assume that it was not whole
communities, but only parts of these that migrated, and that the
migrations were not one great systematic movement but consisted of
604 

many small wanderings, and, in some cases, there is even reason to

doubt the historicity of these migrations at all. We can, however, be
relatively certain that these later “peoples”, who had retained their
original name, meanwhile had changed their (ethnic) character and
structure completely during the Migration period. The Ostrogoths
who conquered Italy, for example, had been a comparatively recent
formation that was not united until the reign of Valamer (P. Heather).
The Burgundians of the fourth century lived in very different cir-
cumstances from the surviving Burgundians of the fifth century set-
tled around Sapaudia (I. Wood) whilst the Angles, Saxons and Jutes
in Britain who were still bearing these names turned out to be newly
defined peoples (B. Yorke). The same name does not necessarily mean
a continuity of the same people, but there must have been at least
some bond (or collective memory) that still allowed them to use the
same denomination. The prehistory of these peoples, therefore, may
be obscure, but it is not insignificant.9 The fact that a new name
does not necessarily imply the emergence of completely new peoples
has to be taken into consideration; it can equally lead to the con-
clusion that decisive changes took place. Most of these peoples seem
to have been a conglomeration of smaller peoples or tribes, such as
the Alamans, the Franks, or the Saxons, expressions that emerge in
the course of the third century. The changing of names should be
regarded as an important issue even though we do not know if these
terms were innovative expressions or if, previous to this period, they
had signified just one (name-giving) part of the later people.10 Such
a conclusion might be true for the Bavarians who were to form a
(new) people in parts of the formerly Roman provinces of Raetia
and Noricum at a time that cannot be fixed precisely, but was clearly
not before the second half of the sixth century (M. Hardt). (The
endless discussion about the origin of the Bavarians becomes less
significant once we acknowledge that they, too, consisted of various
communities.) It remains difficult to recognize what it was that held
these peoples together or to what extent they regarded themselves

9
This argument is emphasized by Wolf Liebeschuetz.
10
For a theory of these collective terms cf. Karl Strobel, Die Galater. Geschichte
und Eigenart der keltischen Staatsbildung auf dem Balkan des hellenistischen Kleinasien 1 (Berlin
1996) pp. 131 ff., who compares the ancient peoples with the Germanic peoples.
Strobel believes that these peoples’ names were their own before they were adopted
by others.
  605

as (ethnic) units and developed an ethnic identity. Although we may

assume that they shared some kind of solidarity and cohesion,11 it
remains uncertain on what elements these were based or if such ele-
ments were of greater import than the interests of the individual
(W. Liebeschuetz). In most cases, we lack clear indications either
of a common religious cult or of a common political leadership.
Nevertheless it has to be borne in mind that the Roman (and later
on also other) sources perceived all these communities as being peo-
ples ( gentes) that could be comprehended under one collective name,
without making great differences as to their age and origin. From
the Roman perspective, there were no great differences between
“Germanic” and other non-Roman peoples who were all considered
to be “barbarian”. If differences were made at all they were not
considered of great relevance (and the term Germani was in fact used
extremely rarely by the authors).12 Equally, in the (later) early medieval
tradition, there were also gentes before there were kings, as in the
case of the Franks or the Langobards who instituted a king ( J. Jarnut).
Similarly, at a later period, Spanish Christian chronicles, such as the
“Chronicle of 754”, perceived the Arabs who had conquered Spain
as a gens which was usually distinguished from the “Moors” in North
Africa. They not only described the new realm with the same words,
regnum and gens, and with expressions typical of Christian kingdoms
(as if regarding the influx merely as another migration), they even
failed to comment on any signs of distinctions between the inhabi-
tants and the newcomers; in fact, they did not even mention the
difference in religion in connection with the conquest of Spain,
although they knew, for example, that the Saracens who had defeated
the Emperor Heracleus had a new religion (A. Christys). Moreover,
it seems that the groups which were serving in the Roman army

11
This is emphasized by Wolf Liebeschuetz.
12
Wolf Liebeschuetz commented on this passage that there was little doubt that
all the “Germanic” tribes had common characteristics, besides their language, but
that there seemed to be little evidence that they themselves were conscious of this,
although it must have made communication between different “Germanic” peoples
much easier. This view, however, is not shared by everybody, and certainly many
features that were characterized as specifically “Germanic” by former research have
to be regarded either as “barbarian” or even as “late antique”, although, as
Liebeschuetz argues, Romans that had lived among or campaigned against these
peoples noticed that they were different, for example, from the Gauls, but that they
shared certain cultural characteristics among themselves.
606 

experienced a stronger solidarity with the Romans than with their

own peoples, even to the extent of facing them in battle (which fre-
quently occurred).
Evidently then “peoples” existed on Roman soil before the estab-
lishment of large kingdoms that were named after them, however
tenuous and prone to change these communities may have been,13
and it also seems obvious, at least in many cases, that they were not
united by a common leadership before that time. On the contrary,
the reports on these peoples suggest that they did not act as a sin-
gle unit, but that the political and military actions that have been
alluded to in various sources were carried out by smaller groups (for
example, by Franks but not by “the” Franks). Nevertheless there
were kings even in the earlier history of these peoples, and in most
cases we have a far clearer concept of these kings than we have of
the peoples themselves: The Langobards, contrary to the statements
of a later tradition, were probably ruled by kings from the begin-
ning of their historical existence and possibly united even in these
early times ( J. Jarnut). This may have been an exception to the rule,
but going by the Roman sources at least there were other peoples
who had their own “kings”: before Clovis, the Franks were ruled by
several kings who governed smaller territories (H.-W. Goetz), other
marauding groups, such as the Visigoths and also the Vandals, were
more or less united under military kings (Heerkönige), and also the
Avars were associated from their historical beginnings with a kha-
ganate (W. Pohl). There were even indications, though not in all
peoples, of a dynastic hereditary kingship (such as, above all, the
Merovingians of Tournai, the Vandals under Geiseric’s clan, but also
the so-called Ostrogothic “Amals”, the Visigothic “Balths”, or the
Burgundian “Gibichungs”). Such dynasties could, however, take on
different forms: e.g. as the solidarity of sons (or relatives) but often
enough the succession was contested by the sons or brothers of the
last king.14 Some of these kings had a double function, serving as
magistri militum in the Roman army besides being leaders of their

13
This is not a contradiction to the fact that barbarian troops could be a great
threat to the Roman Empire, but Alamans who threatened Rome, or Goths who
marauded through the Empire, simply were not “the” Alamans or Goths but sin-
gle groups under individual leaders.
14
Wolf Liebeschuetz argues that the fact that a people lacked a dynasty was a
stronger indication of tribal cohesion than their having one because in this case
they depended on their own sense of oneness to unite under a single new ruler,
  607

peoples (or part of their peoples). In some cases, as with Theoderic,

the Roman military function seems to have been closely linked with
the establishment of a “Germanic” kingdom (even if, as Wolf Liebe-
schuetz argues with regard to Theoderic, he was given a function
by the Emperor only in order to be rid of him). Moreover, there was
a certain degree of Romanization in these leaders as well as their
soldiers and perhaps also in the population of these “barbarian” peo-
ples (for example, the Franks in Gaul, the Ostrogoths in Italy, the
Vandals, Alans and Sueves in Spain, the Visigoths in Spain, or the
Burgundians); a Romanization that derived from military service in
the Roman army, proximity to the Roman Empire, settlement on
Roman soil, and cultural influence. Thus, many of the immigrants
were not strangers, but already Romanized when migrating into the
Empire, a process that was continued in the “barbarian” kingdoms.
(It is striking that, later on, we have a similar situation when it was
no longer the Romans, but the Franks who were the “model” for
new peoples: In Bavaria, for example, the role played by the duke
with regard to the Frankish king is comparable with the role of the
magistri militum towards the Roman Emperor.) The new barbarian
peoples that had now become “civilized” and orthodox might even
have considered themselves as the representatives of an ancient Roman
Romanitas, like the Visigoths in Spain according to Isidore’s Historia
Gothorum (I. Velázquez).
Thus we know that there were peoples before the establishment
of greater realms and that in many cases they were already led by
kings; nevertheless, there are difficulties in “defining” them distinctly.
Little can be discerned about the early peoples by way of archaeo-
logical evidence (from grave goods, costumes or weapons), and the
same applies to their customs and habits which do not reveal clear
ethnic distinctions, resulting in a degree of uncertainty regarding their
ethnogenesis. The Franks, for example, are not discernible as an
archaeological (or cultural) unit, that is, they are not distinguishable
as Franks, but at least there seems to be sufficient evidence of an
(increasing) non-Roman settlement in northern Gaul (which probably
was mainly Frankish). And yet even in this respect it is not easy to
discern “non-Roman” settlement archaeologically since most objects

whereas the dynasties of the Goths were in fact short-lived, but given invented ret-
rospective pedigrees.
608 

buried as grave goods were of Roman craftsmanship.15 Archaeologi-

cally, the Avars were (and always remained) a heterogeneous con-
glomeration of various groups with local or regional traditions. From
the beginning there were various underlying influences, from Byzan-
tium as well as from the West, and yet their way of life seems to have
remained unaltered (F. Daim). Apparently these peoples embodied
“mixed cultures” from the beginning (a statement that corresponds
neatly with the modern understanding of ethnogenesis), and it seems
that acculturation played an important part as can be clearly observed
in the case of the Franks in Gaul (M. Schmauder). Nevertheless, we
may distinguish between different phases when considering these
peoples (or their ethnogenesis) although it remains difficult to say if
similar phases can be applied to different peoples. We may clearly
discriminate between the early Vandals in Silesia and those in Spain
in the early fifth century (even though we are unable to say what
differences occurred during the migration of Silingi and Hasdingi
from Silesia to Spain). The same can be said about the situation in
Africa after Geiseric’s conquests, but it is hard to deduce the degree
of impact these phases had on the ethnogenetic development of this
people. For lack of evidence, it is also difficult, or nearly impossi-
ble, to make statements about a “gentile” identity of these peoples
(a phrase borrowed from the German term “gentile Identität” in the
sense of ethnic identity, whilst at the same time acknowledging and
emphasizing the fact that the peoples this term refers to were not
really “ethnic” units).16 It is equally difficult to make any conclusive
statement about their common origins, since the written Origines gen-
tium (such as the Frankish Trojan myth) deviated to a high degree
from historical facts, and in addition were handed down at a later
period. Thus they are good indicators of later traditions, but ones

15
Thus Guy Halsall, in a forthcoming article of the Transformation of the Roman
World-series. M. Schmauder still retains the traditional view that the abandonment
of burials with grave goods follows the “Roman model”, which is probably true on
a large scale (and can actually be regarded more as a “Christian” than a “Roman”
model), although both evidently became increasingly intertwined.
16
There has been a discussion in our group whether it is appropriate to use
terms such as “gentile” which might cause misunderstandings, but knowing that,
first, alternative terms, and primarily “ethnic” would lead to more misunderstan-
dings, and, second, one inevitably has to “define” these words in some way, I
decided to keep “gentile” as a term directly referring to gentes (which were not, or
at least not necessarily, “ethnic” units).
  609

without great value with regard to the early peoples (see § 7 below).
In conclusion, we are forced to accept the existence of early peo-
ples, without being able to arrive at a comprehensive understanding
of their nature and without being able to define them as political
communities until such time that these peoples turned into king-
doms, which was obviously a decisive factor (see § 4 below). This is
confirmed by Arab Spain where, at variance with the Arab point of
view, the new (Islamic) “kingdom” was obviously of greater impor-
tance to the Christian perception than the new, foreign people
(A. Christys). We have to ask, however, if there really was such a
development from “people” to “kingdom”, or if the establishment of
kingdoms was the consequence of some political formation under the
reign of a king, or even if the kingdom (or people) were those pre-
pared to follow or accept the authority of a king.

3. The relationship between gentes and regna II:

The development from “peoples” to “kingdoms”

The “development” from being a “people” to becoming a kingdom

was a consequence of the collapse of the West Roman Empire
(P. Heather) and could occur in various ways. One possibility was
a political federation led by kings or warlords that conquered new
lands, preferably on Roman soil. Examples include the Vandals’ con-
quest of Africa under Geiseric (whereas the Alans and Vandals in
Spain did not conquer the whole country [ J. Arce]), or Theoderic’s
Ostrogoths, and then later on Alboin’s Langobards who conquered
Italy. In other cases, for example, the Visigoths in southern France,
probably also the Burgundians around the Lake Leman (the Sapaudia
region, which was to become the central region of the later realm)
and to a certain degree one can also include Theoderic’s early reign
in Italy, these “peoples” (or marauding barbarian armies) were officially
settled on Roman soil and given duties as federates. Some hordes
may also have conquered a country without becoming politically
united, as in the case of the Anglo-Saxons who, according to Gildas,
subdued the Britons or expelled them from eastern England—
archaeologically, we can detect a “Germanization” of the country—
in order to form their own, small kingdoms. (Moreover, most his-
torians do not believe that all the Britons were expelled, but imagine
610 

a gradual process in which the replacement of Romano-British cul-

ture by a “Germanic” one seems to signal a shift in the balance of
power from one side to the other.) We cannot, however, give any
definite dates to this process although it seems to have been definitely
completed in the late sixth century (and therefore we cannot ascer-
tain any direct development from migrating to establishing kingdoms
which, contrary to the Franks, for example, may have been a delayed
secondary development rather than an integral part, or result, of
that process [B. Yorke]). Bede’s “Anglo-Saxons” (or rather Angli ) were
an expression for the whole Anglo-Saxon population; they were, how-
ever, divided into several kingdoms. It may have been Jutes who set-
tled in Kent as Bede wants us to believe; the sources, on the other
hand, reveal a close-knit relationship between the people and the
kingdom of Kent. The king, therefore, seems to have been the deci-
sive factor in this process. In fact, as Barbara Yorke emphasizes in
her contribution, regnum and gens were closely entwined, and all the
Anglo-Saxon gentes mentioned in Bede had (or were) kingdoms at
the same time. Furthermore it may well be that the (archaeological)
differences between these kingdoms were not distinctions of the im-
migrating peoples but were adopted later after their migration to
Britain. There may also have been some Frankish influence, or rather
model, for the establishment of kingdoms in Britain (B. Yorke). If the
Bavarians migrated into their new, formerly Roman settlement area,
they certainly did not come as a “closed” people (M. Hardt) and
did not really conquer the country, but may only then have grown
together to become Bavarians in “Bavaria” itself. Such an immigra-
tion was not necessarily a single event but rather a long process, as
in the case of the Franks in northern Gaul. It may well be (and
would be in accordance with our modern theories of ethnogenesis)
that the name-giving peoples originated from a synthesis or even
“syncretism” (in the sense of growing together) of various popula-
tion groups (H.-W. Goetz). At first sight, the Franks seem to be a
special case (or one in which several elements were combined) because
they did not conquer northern Gaul, but appear to have settled there
gradually, in steadily increasing numbers (notwithstanding that sin-
gle groups, such as the so-called Salian Franks under Julian, were
settled, or rather allowed to retain their settlements, by the Romans).
Nevertheless, this process may also have served as a model for other
peoples, and there may have been a (more or less peaceful) period
of settlement before the formation of “barbarian” kingdoms else-
  611

where; this was probably true for the Visigoths in Spain. Settlement,
as Ian Wood emphasizes, was not a single event, but a long process.
Later on, again, the Franks also conquered important parts of the
remaining Roman dominion (the realm of Aegidius and Syagrius)
which was to become the basis for their huge kingdom. Unfortunately,
we cannot reconstruct this process of settlement by studying place-
names because these normally date from a much later period. In
France, for example, they may originate from the time before or
after Clovis, but in most cases there are no records that can be
dated before the eighth or ninth centuries.

4. The relationship between gentes and regna III:

The establishment of (great) kingdoms

Consequently, the establishment of a new kingdom (in most cases

on Roman soil) could derive from a gradual infiltration into Roman
areas, either in the form of expansion (as in the case of the Franks
and the Visigoths in Spain), or due to the displacement of the set-
tlement area (as in the case of the Tolosan Visigoths, the Ostrogoths
in Italy, the Vandals in Africa, the Burgundians, Anglo-Saxons and
Langobards). Nowhere do we discern the formation of a kingdom
as a single meditated act. In some cases, for example the Anglo-
Saxons, it is nearly impossible to put the establishment of the new
kingdoms into any kind of chronological order; in other cases we
can only hazard a guess at the formation of new kingdoms over a
more or less broad period of time; the Burgundian king Gundobad,
for example, seems to have founded a kingdom some time between
474 and 494 (I. Wood). It is significant, though, that Geiseric counted
his regnal years only from the year of the capture of Carthage
(439), ten years after the conquest of most parts of northern Africa
(W. Liebeschuetz). The clearest cases, however, that allow us to date
such “foundations” within a comparatively short period, were those
where the new kingdom may be seen as the achievement of indi-
vidual kings: Clovis (the Franks in Gaul), Geiseric (the Vandals in
Africa), or Alboin (the Langobards in Italy). At the same time we
have to bear in mind that it was naturally also the work of their
followers who were normally characterized as their “people” in the
Roman sources (a “people” that had probably grown together from
various gentes and armies). These individuals may have been “men
612 

of exceptional ability” (W. Liebeschuetz), and they hardly ever had

successors with comparable qualities, but nevertheless they were most
important for the “foundation” and continuity of their kingdoms.
Comparing Clovis and Alboin, we discover significant similarities as
well as differences between these “founder generations”: Both men
had been kings of their respective peoples before, but created new
kingdoms by conquering large areas. Whilst Clovis enlarged his ter-
ritory by successive conquests, Alboin found new domiciles in a “for-
eign” country (Italy) after a migration. Though archaeological findings
provide only minor information of political developments, the famous
grave of Childeric displays his elated rank, even before Clovis’s time.
Luxurious graves in other kingdoms (such as the Avars) may also be
taken as proof of the existence of princes.
Most of the “barbarian” kingdoms were established on Roman
soil (and it would be an interesting question to see how further devel-
opments varied in different cases). As Peter Heather claims, they
could not have succeeded without at least some kind of coexistence
with the Roman world, or population. More than that, it seems obvi-
ous that many, if not most elements of the newly built kingdoms
were based on the Roman tradition (and, as has been mentioned
before, we may assume a certain Romanization of the “barbarian”
peoples even prior to that time). As far as we can see, Roman admin-
istration was adopted, or rather continued, in all kingdoms on Roman
soil, probably with the exception of England, where the severence
from the Empire in the beginning of the fifth century marked a
stronger separation although it did not necessarily mean a complete
collapse of the Roman system (Barbara Yorke). (It could, in fact,
prove to be a very worthwhile future project to investigate the differ-
ences between the single kingdoms in adopting the Roman system.)
The same is true of the imperial demesnes, which now became the
basic possessions of the new kings, for the royal fisc, the monetary
system, the issue of charters, and the codification of laws. These
Roman traditions were maintained, but possibly to a varying degree
and in different ways, and sometimes, as in fifth-century Spain, the
country remained “Roman” throughout ( J. Arce). The Vandals were
the first to even restore a stable currency in Roman tradition (W. Lie-
beschuetz). The “barbarian” kings also adopted imperial titles and
epithets, they used corresponding insignia, and chose Roman cities
as their capitals, or sedes regiae. It is significant that all Merovingian
sedes regiae were situated in the former Roman province of Aegidius
  613

and Syagrius,17 whereas the Visigoths chose one of the less impor-
tant Roman cities as their capital. Only Toledo in Spain and Pavia
in Langobard Italy, and probably also Ravenna in Ostrogoth Italy,
may deserve being classified as “capitals”.18 Another, probably even
more decisive factor of Roman continuity was the Church. Moreover,
we should not underestimate the whole complex of Roman culture
surviving, at least to a certain degree, in the successor states. The
“Germanic” elite in the Burgundian kingdom, for example, adopted
a Roman life-style (I. Wood). Accordingly, it is probable that the
Latin language prevailed, as Alex Woolf argues, even in the Celtic
world of Britain, particularly in the lowlands. We do not know
whether Latin was the official language (or if an “official” language
existed at all), but it is at least significant that all the administrative
correspondence, and also tombstones, were written or engraved in
Latin. (I do not consider this to be just a consequence of the illitera-
cy of the “Germanic” peoples, because they probably could have
developed a “Germanic” literature if there had been a need for it
before the later eighth century. And it is not by chance that a con-
siderable vernacular literature was developed in Anglo-Saxon England,
and Celtic Britain, where Roman traditions were less dominant than
elsewhere.) Finally, the infrastructure (particularly the Roman roads)
continued to be used for centuries. But, again, we should not neglect
the fact that we still lack a comparative study of Roman traditions
in the single kingdoms.
These kingdoms, therefore, may have carried on the Roman tra-
dition in wide areas of political, cultural and everyday life, and, at
least in the beginning, the kings themselves may have felt and acted
much more as Roman officials than barbarian kings (as Ian Wood
claims throughout for the Burgundian kings who should “be con-
sidered not just alongside the kings of the Vandals, Visigoths and
Ostrogoths, but also alongside Odoacer, Ricimer, Aëtius and Stilicho”).19

17
Though one should note the scepticism of Edward James over the size of their
kingdom: E. James, The Franks (London 1988) pp. 67–71.
18
Cf. now Sedes regiae (ann. 400–800), ed. G. Ripoll and J.M. Gurt with A. Cha-
varría (Barcelona 2000).
19
Ian Wood’s suggestion that the Burgundians did not conceive themselves as
part of a “Burgundian” kingdom has been questioned during the discussions by at
least some members of the group and may be the result of the Roman perspective
of our sources. As Wolf Liebeschuetz argues, Constitutiones Extravagantes 21, ed. L.R.
von Salis, MGH LL nationum Germanicarum 2,1 (Hannover 1892) (Capitulus quem
614 

In fact, they were often both Roman “soldiers” and barbarian war-
lords. Politically, however, they were (or became increasingly) inde-
pendent, not Roman federates any more, but rulers of independent
successor states which had liberated themselves from the imperial
claims of a remote Byzantium. It remains to be seen what effects
that may have had on their political identity.
These new kingdoms, in many cases at least, may be regarded as
the achievement of single outstanding kings, and were based on
Roman traditions. Nevertheless it is significant that they were named
after peoples. These peoples (or rather their leading classes) may but
need not be identified with Wenskus’s “nucleus of tradition” (Traditions-
kern), but formed a demographic minority in most parts of the coun-
try. The “people”, therefore, continued to play some part in the king’s
realm. This seems to apply to the Franks, although expressions like
rex Francorum or regnum Francorum are not handed down until the end
of the sixth century. It may also be true for most of the other king-
doms, and is particularly apparent in Langobard Italy.
In reality, however, these kingdoms were “multinational” territo-
ries throughout, including not only the name-giving people, but also
the resident Roman population and other “barbarian” peoples (such
as Berbers in Vandalic Africa, Slavs, but also Bulgars or Gepids, in
the Avar khaganate, or Britons in the Anglo-Saxon kingdoms, prob-
ably also remaining Ostrogoths in Langobard Italy). In most king-
doms, the Roman population formed a vast majority and, though
“Roman” by law, was itself of varied origins. The clearest case of
a “multinational” (or multiethnic), and also “multi-kingdom” realm
is the Frankish kingdom, consisting of the former Visigothic and
Burgundian kingdoms and including trans-Rhenian peoples, such as
Alamans, Thuringians, and Bavarians. From an “ethnic” point of
view, therefore, we find various cultures as well as (or rather: and
increasingly) a “mixed culture” in these realms, a phenomenon which
is confirmed by archaeological evidence. Nevertheless, for people of
the time, these groups still seem to have been distinguishable,20 at
least for a while. According to Roman sources, the Vandals remained

domnus noster in conventu Burgundionum instituit) is strong evidence for the existence of
a Burgundian assembly, and therefore of a Burgundian people (habito nunc cum
comitibus nostris tractatu, praesenti constitutione decrevimus, quod in populo nostro debeat cus-
todire). See also Wood’s discussion of this assembly, this volume, p. 256.
20
Again, Wolf Liebeschuetz puts emphasis on this argument.
  615

“Vandals” till the collapse of their kingdom, and, in spite of their

assimilation and a process of acculturation, they never became “Ro-
mans”. In cases where a kingdom survived for a long period, how-
ever, such as the Frankish realm, we can observe an increasing
“syncretism” (in the sense of growing together). Nevertheless it is also
evident that a process of acculturation began and was achieved much
earlier and more successfully in “Roman” Gaul than in the Austrasian
parts of the kingdom (where, for example, the custom of grave goods
was maintained much longer). The fact that Frankish women in
northern Gaul adopted Roman clothing (without fibula) earlier than
their men (M. Schmauder), might have been the result of inter-
marriage (which at the same time would explain the increase of
Germanic personal names in all population groups).

5. The relationship between gentes and regna IV: Development and changes

The establishment of kingdoms was no doubt a deciding factor which

bore an essentially political character. As kingdoms were named after
a people that had existed before the kingdom itself, we cannot ignore
the relationship between gens and regnum, but it would be daring to
conclude a straightforward development from one to the other. It
is, however, only after the formation of kingdoms, that these peo-
ples assumed a more institutionalized character (at least on a polit-
ical level). The kingdoms, therefore, seem to have been a crucial
element and at the same time a caesura in the ethnogenesis of these
peoples which consequently underwent changes during the subse-
quent periods. Francus/Franci were ambiguous expressions with con-
notations that could be ethnic, legal (as in the Lex Salica), or political
(and even the ethnic term was political since it normally did not
mean “the” Franks, but the politically active and responsible Frankish
elite). Significantly enough authors used the same term to charac-
terize (or “label”) the people and the kingdom. Moreover, in the
process of Frankish history, the political (and territorial) connotation
prevailed increasingly. First, the people adopted a political (or “reg-
nal”) meaning by being the leading people, or the body responsible,
in this realm, the Reichsvolk. Second, however, due to increasing “syn-
cretism”, the Reichsvolk turned into a Reichsbevölkerung, the population
of the whole kingdom. Consequently, the ethnic terms (again) changed
their meaning (H.-W. Goetz). The kingdom became more and more
616 

important to the degree that the significance of the people dimin-

ished (unless the people’s name had come to denote the kingdom).
A similar process can be seen in Visigothic Spain in official docu-
ments (Church councils), though here this process seems to have
begun and lasted much later: In Toledo III (589), which established
a Catholic kingdom, the gens Gothorum had indeed become Catholic,
although here again the gens still did not include the whole kingdom,
but was apparently restricted to the Goths as the ruling people
(that is, excluding the Hispani), a concept that can also be observed
in John of Biclar and Isidore of Seville. Nevertheless, we can note
a certain consciousness of unity of the whole of Spain in Toledo IV
(633); finally, in Toledo V (636) and VI (638), even this original
separation of Goths and Hispani has been abandoned: gens had
ceased to be an “ethnic” and had become a political term (I. Veláz-
quez). As in the Frankish kingdom, this “growing together” was deeply
rooted in the common Catholic religion.
Most kingdoms were characterized by two (seemingly) contradic-
tory features (which can be observed most clearly in the Frankish
kingdom): unity and integration on the one hand, disunity and dis-
integration on the other (H.-W. Goetz). The most important ele-
ments of unity were, first, the king and the kingdom itself as well
as a tendency towards establishing royal dynasties, stirpes regiae, then,
the Church (that is, if the Church was Catholic), and the fact that
the king was head of this Church, then the (predominantly Roman)
administration, and also the social structure. Elements of disunity
were the divisions of the kingdoms among the heirs (the king’s sons
or brothers), as in the kingdoms of the Franks, the Church (in such
instances where the “Germanic” peoples were Arians), and, of course,
the “multinational” society which, in those cases where the peoples
lived according to their own laws, also meant legal disunity, although
in some kingdoms, such as the Frankish or Langobard, and also the
Burgundian kingdoms, the laws seem to refer to the whole popula-
tion i.e. to its “Germanic” and Roman parts (P. Wormald). Actually,
after the second generation, the Merovingian kings frequently opposed
and attacked one another, though there was still a concept, or an
illusion, of a united Frankish kingdom. In reality, the parcelled king-
doms (Teilreiche) became increasingly independent.
These kingdoms were stabilized by several elements: Catholicism,
an imitatio imperii, the integration of the conquered population and a
further Romanization of the king’s people (which actually had begun
  617

long before), and, not to be underestimated, the role of the army

(which was also rooted to some degree in Late Roman traditions).
The army even gained in significance as the barbarian soldiers to
an increasing extent were the army, thus playing an important role
in the state (W. Liebeschuetz). Moreover, the laws which were codified
in most “Germanic” kingdoms (except, however, in the Vandalic
realm in Africa) were a significant element of state-building which
probably had a stabilizing effect, too. They undoubtedly derived from
the Roman model (whereas their contents were mainly “Germanic”,
however “adapted and modified” these customs may have been by
the “law-makers”), but they equally originated in the king’s duty to
preserve justice (P. Wormald). In southern Europe they seem to
have complemented the Roman law (which was imitated everywhere),
with a “Germanic” version beside a “Roman” one, and in some
regions, such as Burgundy, as Stefan Esders has shown in his dis-
sertation, the Roman legal tradition survived into late Merovingian
times (the seventh century). These laws, however, which were undoubt-
edly linked with royal power and legislation, were not necessarily
laws that applied to the entire kingdom. In some instances this was
the case, but in other instances the laws were applicable only to cer-
tain groups that may have been identical with the known “ethnic”
groups (though it is also possible that they applied to a certain region).
There were several laws in the Frankish kingdom, and the Alamans
and Bavarians had laws of their own, though, remarkably enough,
these leges too, were initiated by the Frankish kings. As such they
can be considered to bear a political as well as an ethnic significance
(P. Wormald).
When speaking about stabilizing factors, we should not neglect
the importance of religion.21 Supra-regional kingship had to be
Christian to be successful with regard to broad integration, although
this does not include the Avars (W. Pohl) where indications of Chris-
tianity were an exception (F. Daim). We cannot be sure if Arianism
was understood as a relevant sign of distinction between “barbar-
ians” and Romans, or, as Wolf Liebeschuetz argues, was some-
times even used as an instrument to keep them separate; actually
both the “Catholic” as well as the “Arian” Church considered them-
selves to be universal. In some instances, however, Arianism did

21
This has again been suggested by Wolf Liebeschuetz.
618 

play an important role, particularly throughout the Vandalic king-

dom which, under Huneric, even organized persecutions of Cath-
olics, and in the case of the Ostrogoths who formed a “Homoian”
Church in Italy (P. Heather).22 It seems that religious conflicts and
tensions between a “Germanic” identity (either heathen or Arian) and
a Romano-Christian tradition (which to a certain degree was existent
everywhere) were more acute in these two cases than in all other
kingdoms, notwithstanding the fact that there were Roman Arians
as well as barbarian Catholics. On the other hand, Arianism was
often no more than a phase in the history of most kingdoms (such
as the Visigoths, and perhaps also the Burgundians and the Lango-
bards; in the case of the Burgundians, however, religion does not
seem to have been a clear sign of distinction). In Bavaria, we find
different denominations; here, a Christian mission of any import did
not start until the first half of the eighth century; significantly, in
these years, under Theodo and Odilo, it was organized by the duke
as well as by the Carolingian maior domus (M. Hardt). In the Frankish
kingdom, as a consequence of the baptism of Clovis, Catholicism
spread more or less “naturally” as the common belief, although, of
course, supported and propagated by the kings and the Church; the
fact that Catholicism was the common belief of Romans and Franks
in the Frankish kingdom(s) definitely contributed towards the “syn-
cretism” of these populations, and the same is true for Visigothic
Spain after 585 (I. Velázquez). In later Langobard Italy, it was even
decreed as a “state religion” (Staatsreligion), and Anglo-Saxon law-
codes contain legislation making certain Christian acts (for example,
baptism) compulsory so that Christianity might be said to be a “state
religion” in the Anglo-Saxon kingdoms as well (B. Yorke). The king’s
position as a ruler by God’s grace (Gottesgnadentum), a concept which
became widespread in all Christian realms, contributed considerably
towards the development of stabilizing the kingdom. Catholicism and
the king’s sacrality, as Jörg Jarnut emphasizes, were crucial elements
in this process. In fact, Catholicism seems to have been an impor-
tant precondition for successful “syncretism”. And yet, as Wolf Liebe-
schuetz reminds us, the importance of religion seems to have been
different in each kingdom.

22
On the other hand, Ian Wood argues that “there is nothing to suggest that
religion acted as a distinguishing mark between Burgundian and Roman”.
  619

The further development of these kingdoms, however, varied greatly

from one kingdom to the next. There were those, such as Langobard
Italy, where we can observe the growth of political order and sta-
bilization, preferably supported by three elements: increasing cen-
tralization, hierarchization, and institutionalization, including the
establishment of new administrative offices ( J. Jarnut). There were
others, such as the Frankish kingdom(s), where similar developments
were halted by internal political weakness in a later phase of its his-
tory, or the Vandals, where the royal stabilization did not really lead
to a comprehensive integration of all its peoples. We should not for-
get that only some of these kingdoms were of longer duration allow-
ing us to pursue their development over a longer period of time:
those of the Franks, the Visigoths in Spain, the Langobards; with
regard to the Avars such developments can only be reconstructed
by archaeological material. In fact, most of these kingdoms perished
after a short while, for the most part because they were conquered
by foreign kings: the Visigoths were beaten and conquered by Clovis
(507), the Ostrogoths (who themselves had challenged and killed
Odoakar) and the Vandals were defeated by the Emperor’s gener-
als (Narses, Belisarius), and the Spanish Visigoths were subdued by
the Arabs (711) or took refuge in the regions of the Asturias, the
Picos and the Pyrenees. The Avars and Langobards as well as (again)
the Bavarian dukes were subjected by Charlemagne (though the Avar
khaganate had a remarkably long existence of more than two cen-
turies, but in the end was incapable of putting up sufficient defence).
Sometimes we hear of heroic resistance (as in the case of the Ostro-
goths and Vandals), but in other cases one single battle (Vouillé 507)
or a few campaigns decided the destiny of a kingdom (Visigothic
Spain, Langobard Italy). This should be taken as an indication of
feeble political and social structures as well as a lack of support by
the population. As Wolf Liebeschuetz claims for the Vandals, one
might argue that the disappearance of these peoples was due to the
fact that their gens was deliberately broken up; that is, if we define
gens in other than political terms. One could therefore conclude that
kingdoms such as these often rose and fell with their kings.
620 

6. The relationship between gentes and regna V:

The people(s) inside the kingdom

As has already been stated above, all early medieval kingdoms included
a “Germanic” and a Roman population, that is, a “Germanic” minor-
ity and a Roman majority, and some were even “multinational” cul-
tures. The Anglo-Saxons constitute an apparent exception, although
this may be because we categorise the Romano-British as British and
not as Romans. This is also underlined by the archaeological evi-
dence, which shows that the Merovingian kingdom, for example, was
far from giving the impression of cultural uniformity. Different pop-
ulations were held together by the king’s rule. We may assume that
the Roman majority accepted “barbarian” rule, and we know that
leading Romans were ready to work with it (W. Liebeschuetz). We
may also assume that the leading classes at least were bilingual in
so far as they knew Latin besides their native tongue (and perhaps
vice versa), even in regions with considerable “Germanic” settlement
(as in northern France). In the long run (again, with the exception
of the Anglo-Saxons), the Latin language (as the language of the
majority) was able to assert itself indicating that the “Germanic” (or
rather the non-Roman) population must have adopted it gradually.
In Visigothic Spain, we have no clear indication that the Gothic
language was prevalent, though certain Gothic words were adopted
(I. Velázquez). On the other hand, however, it is also significant that
at the same time Roman personal names were decreasing because
Roman people obviously preferred to give their children “Germanic”
names, a development that can be easily observed in the Frankish
and Langobard kingdoms, but also elsewhere. All in all, it is likely
that there was mutual influence. On the whole, “Romanization” was
dominant in the linguistic sphere whereas the political denomina-
tions of the kingdoms remained those of the “barbarian” victors. In
England, there was a similar but reverse process: here it was the
Celtic (and Roman) language that was superseded by Anglo-Saxon
dialects in most areas—there is little evidence of any Celtic influence
on the Old English language, except in some place-names and espe-
cially in river names—, whereas Latin did not prevail, probably
because the Romanization of the population and country had been
less intense than on the continent. Whereas Ine’s law still distin-
guished between “British” (wealh) and “Saxon”, 200 years later, in
  621

the reign of Alfred, there were only “Saxons” in Wessex, presum-

ably a sign of the acculturation which had occurred (B. Yorke).
Apparently there were marriage prohibitions in some kingdoms
(or at least in the earliest phases of those kingdoms), and there were
(several) forms of coexistence, as between “Germanic” peoples and
Romans in early fifth-century Spain ( J. Arce). In the Italy of that
time, the Ostrogoths preserved a sense of distinction between their
own people and the Italo-Romans, in spite of the beginnings of
assimilation, as Peter Heather argues convincingly contrary to Patrick
Amory’s theories, and the same may be said for the Vandals in
Africa. It is difficult to say to what extent the Slavs were integrated
into the Avar khaganate, but there seems to have been at least a
kind of “symbiosis” and cooperation, and the seventh century even
saw a rise of regional Slavic powers (W. Pohl). Other (or perhaps
most) kingdoms, however, such as the Franks, the early Vandals,
and, above all, the Burgundians, who even encouraged immigration,
were “open societies”.23 The Frankish kingdom is a clear case of the
increasing “syncretism” of different populations, to such a degree
that, from the seventh century on, it often seems impossible to ascer-
tain the “ethnic” origin of the individual. Again, however, it was the
political unit, the kingdom, that decided on the “affiliation” for, as
has already been stated above, the whole population came to be
known as “Franks”, a term which adopted a new, though ambiva-
lent, meaning in so far as it increasingly denoted the members of
the (whole) Frankish kingdom: to belong to this kingdom evidently
became more important than remembering one’s “ethnic” descent,
and an equivalent development can be seen in Visigothic Spain in
the second quarter of the seventh century (I. Velázquez). There are
no direct parallels of this kind of integrative “syncretism” mentioned
in the other contributions to this volume. Nevertheless it cannot be
definitely said that this was a specifically Frankish (and Visigothic)
feature: As Matthias Hardt suggests with regard to the Bavarians,
the integration of the whole population, particularly the Romans,
was an important presupposition for successful ethnogenesis.

23
This has been emphasized by Wolf Liebeschuetz.
622 

7. The role of a “gentile” (or ethnic) consciousness

It is difficult to say whether and how the continuity of “ethnic”

names on the one hand and the re-formation of these groups in a
political kingdom on the other hand affected the self-conception of
these peoples, and whether they retained an “ethnic” identity, or re-
gained a new, political one.24 It is true that in Gaul, according to
our sources, it was “the Franks” who were the (politically) leading
classes, who elected the king, took part in the general assemblies,
who went to war or made peace with other peoples and, significantly
enough, it was “the Franks” in each of the “parcelled kingdoms”
(Teilreiche). In many cases, however, these “Franks” were no longer
the original “Germanic” people, but, as has already been stated,
were obviously—and increasingly—conceived as the (whole) popula-
tion of the Frankish kingdom, or “parcelled kingdom”. Similarly, in
Visigothic Spain, the gens Gothorum was the dominating group in the
kingdom, and there was a consciousness of group identity at least
until the first decades of the seventh century (I. Velázquez). One
may assume that it was similar in the other kingdoms although
this has not been explicitly stated in the other contributions to this
volume.
A “gentile” (or ethnic) consciousness is most clearly recognizable
in the so-called Origines gentium, which, however, are not handed down
in all kingdoms. They are, for example, lacking in the case of the
Vandals, the Burgundians (as it seems, the Burgundians of the
Sapaudia did not even trace back their origin to the Rhineland
Burgundians) and the Bavarians. On the other hand they are well-
known and frequently discussed in the case of the Goths, the Franks,
or the Langobards. Even so some problems still remain: All these
Origines gentium were written down comparatively late, and, in the
final analysis, we do not know whether they really are testimonies
of widespread popular convictions or just scholarly constructions.
(This is particularly true of the completely unhistorical Frankish origo
claiming Trojan origins for the Franks, a theory that obviously had
its roots in the writings of Virgil). The Burgundians, according to

24
In my opinion, identity did not derive from a common language, as Matthias
Hardt suggests for the Bavarians, because language was not a deciding factor for
ethnic distinction.
  623

Ammianus, knew of their Roman descent (I. Wood), whereas Anglo-

Saxon elites seemed to have traced their descent from Scandinavian
ancestors (B. Yorke). Moreover, judging from the different versions
of Origines gentium in the case of the Franks and Langobards, there
was not just one common belief in the people’s origin, but varying
myths. Nevertheless, these traditions prove the importance of the
belief in a—remote—common origin and also in a long migration,
even in (or perhaps particularly in) later periods, that is, in periods
where the “people” had already changed considerably in their new
environment of a (multinational) kingdom. In contrast to ideological
research in the nineteenth and the first half of the twentieth cen-
turies, however, one cannot emphasize clearly enough that all these
Origines gentium look back to the early history of their respective peo-
ples (as distinct from those of other groups) and that there was
absolutely no conception of a common “Germanic” origin. Moreover,
it is also significant that all Origines associate the origin of their peo-
ple with the reign of kings, thereby confirming the importance of a
political element for “gentile” (or ethnic) consciousness. Burgundian
and Langobard law-codes contained kinglists, which indicate the
importance these peoples attached to the dynastic line of their kings.
Avar identity, too, was so closely connected with the khaganate that
one could not have existed without the other, and it is not by chance
that the Avars disappear almost completely from our sources once
the khaganate had been destroyed by Charlemagne (W. Pohl). Bavarian
identity, in its turn, was linked with their dukes, indicating, in both
cases, a political rather than an ethnic identity (M. Hardt). On the
whole, therefore, our sources point towards the fact that collective
identity was linked to a far greater extent to the kingdom than to
its name-giving gens. Sometimes, however, former gentes maintained
their identity which can be observed, for example, when the Vandal
kings in Spain (and also later in Africa) were called “kings of the
Alans and Vandals”.

8. The role of the kings

In view of this, the eminent role of the kings should be evident. It

has already been emphasized that the establishment of a kingdom
was often the achievement of a (single) king, that even before this
occurred, the “people” was constituted by a political leadership
624 

(normally also under kings) and that, afterwards, the kingdom changed
the structure and “definition” of the people: Kingship, or a compa-
rable political institution, such as the Avar khaganate or the Bavarian
ducatus, seem to have been the decisive element of ethnicity, too, and
it is again worth noting, as has already been said, that even the
Origines gentium were associated with the reign of kings. It is significant
that the function (king) sometimes became more important than the
individual king’s name, as in the Avar khaganate where all rulers
were simply called “khagan” (W. Pohl). The kings, or monarchy,
were therefore of paramount importance for this process. The ten-
year interruption of royal rulership in the early history of Langobard
Italy (574–584) was an absolute exception, and it is significant that
it was followed by a prompt and, as far as we know, unproblem-
atic reconstitution of the monarchy. In the course of Langobard his-
tory we actually perceive an increasing importance and intensification
of kingship which has to be considered as an effective political and
military form of organization. According to Jörg Jarnut, the king was
not only the central focus of reference of his people, but even the
“incorporation” of a “gentile” (or ethnic) consciousness. In Bavaria,
the dukes, who by law held a hereditary office, seem to have gained
a comparable function as an integrative force, particularly in the
beginning under the auspices of the Frankish king; in some periods
at least, the power of the Bavarian duke did not differ from that of
a royal government (M. Hardt). In regions like Bavaria, therefore,
we have to investigate a double role: that of the Frankish king and
the Bavarian duke. In all these kingdoms, therefore, kingship was
not only the deciding factor for political unity, but, since the peo-
ple, too, was a political unit, it was also decisive in the process of
ethnogenesis: this becomes particularly evident after the conversion
of individual rulers to Catholicism (in the case of the Franks, the
Visigoths and the Langobards). Perhaps one can go as far as to assert
that it was the kings who formed the famous “nucleus of tradition”
that was the centre of “peoples’ building” in this age, although we
should be careful not to become caught in the vicious circle of com-
prehending a people simply from a study of its kingship. In Langobard
Italy, it was the king (and his court) who even defined (by legisla-
tion) who was a “Langobard”, by freeing and integrating slaves or
members of other peoples into the Langobard army, as Jörg Jarnut
argues. If this is true (here and in other kingdoms) it would under-
line the pivotal role played by royalty in the process of ethnogene-
  625

sis. And yet even this assumption leaves unanswered the decisive
question as to the criteria that defined or allowed a person to become
a “Langobard” (or member of any other “king’s people”). Any answer
to this question would probably reveal an extremely flexible system.
The Langobards, again, were exceptional in another way: they
chose (or elected) their kings from various families, namely from the
circle of the dukes of the Langobard territories. By contrast, a dynas-
tic system prevailed in most kingdoms, either as a strict dynastic suc-
cession (such as, primarily, in the case of the Merovingians who
ruled uncontestedly until Pippin’s “coup d’état” of 751, but also the
Burgundian Gibichungs, or the Vandalic Hasdings), or, in other
cases, where there was at least a tendency towards a dynastic monar-
chy (as in all Gothic realms: in the Visigothic kingdoms both of
Toulouse and Toledo, but also in the earlier Ostrogothic kingdom
until the extinction of the Amals). And yet in spite of this Peter
Heather warns us not to overestimate the role the dynasty played
as a shaping factor.

9. The role of the Emperor/Empire

The role of the Emperor, however, seems to have been rather differ-
ent. 476, of course, was not the end of the Empire, but only of the
existence of an Emperor in the western part of the Imperium Romanum.
Actually connections between the “barbarian” kings and the Emperor
in Byzantium continued throughout and there are links between the
transformation of gentes into regna and the Empire (which was also
subject to changes [E. Chrysos]). Considering the role of the Empire
in this process of ethnogenesis and the formation of independent
kingdoms, it seems wise to distinguish different phases, for example,
phases of confrontation and integration (as P. Delogu commented at
the meeting in Bellagio). The first is represented by various sieges
of Constantinople, for example by the Goths in 400 or the Avars in
626 which at the same time constituted a significant caesura in the
history of the Avars (F. Daim). The second may be divided into
phases of imitatio, aemulatio and translatio imperii (E. Chrysos), though
the later kingdoms do not fit entirely into such a scheme. In the
early phases, the Emperor undoubtedly had a tremendous impact
on the formation of kingdoms, though it is difficult to be concrete
about these affiliations. Frequently, the relationship between a certain
626 

kingdom and the Empire changed several times. Sometimes, as

in fifth-century Spain, the “barbarian” influence was weak, some-
times the “barbarian” kings governed by permission of the Emperor,
or, as in the case of the Vandals and Sueves in Spain, they usurped
the imperial throne ( J. Arce), or they were even encouraged by the
Emperor to conquer new territories (as in Theoderic’s case). In the
sixth, seventh and eighth centuries, however, when the kings were
increasingly governing without the Emperor’s permission, imperial
influence declined and imperial policy resigned itself to accepting the
development of the kingdoms. Actually, these kingdoms themselves
played a crucial part in the disintegration of the Roman Empire. In
this age, therefore, we have to distinguish between the role of the
Emperor (which was comparatively weak) and the role of the “Empire”,
or, rather, the role of Roman structures, or Romanitas, which were
predominant everywhere. “Barbarian” kingship was often explicitly
an imitatio imperii (for example, in the name or title Flavius, the sacrum
palatium, the dating of royal charters, or economic structures), and it
remained so for a long time, even under the comparatively late ruler-
ship of the Langobards ( J. Jarnut). The Roman laws which served
as a model for the codification of the leges barbarorum, and Roman
administration, as well as minting and coinage were continued through-
out. The cultural impact of Romanitas was equally great in the field
of literacy and writing as well, and is apparent from the archaeo-
logical evidence of all regions that were formerly Roman (and also
in the Avar khaganate, though with characteristic adaptations). One
is tempted to conclude that imitatio imperii, along with a certain “sacral-
ity” of a king who reigned “by God’s grace” was one of the essen-
tial sources of royal power ( J. Jarnut).

10. General conclusion

Finally, it may be prudent to remember the distinction between the

early medieval and the modern concepts of gens and of regnum as
well as their relationship. With modern research beginning to with-
draw from or even abandon the traditional concept of gens as an
ethnic community and now emphasizing its political and cultural
essence or nature, it seems at the same time to have alienated itself
from the early medieval concept which no doubt did include ethnic
thinking to some degree. Nevertheless, this is not the whole truth,
  627

and it should be borne in mind that from the beginning the con-
temporary perception of peoples in those times comprised two ele-
ments: an ethnic (or “popular”) and a political (or “territorial”) one.
It seems that the names of peoples were linked to an increasing
extent to the (whole) population of the kingdom, and to the king-
dom or the royal house itself (as in the Venerable Bede and in
Frankish historiography), and that territorial names derived from
these peoples were increasingly used alongside “gentile” (or ethnic)
expressions; that is, that political (and territorial) concepts gradually
prevailed. From the modern point of view, this development can be
interpreted as a sign of successful integration, or as a new phase of
ethnogenesis. However, perhaps it is even more important to lay
aside a strict (modern) distinction between these two elements (“gen-
tile” and “political”) and acknowledge that they were (always?) linked
to some degree in the early medieval mind, notwithstanding that
there were differences between gentes and regna: In Bede, for example,
the gens Anglorum could even comprise several kingdoms. Accord-
ingly, it was not attached to one political unit, but preceded the
political unification of the whole of England which was not accom-
plished until the tenth century. It seems, therefore, as Barbara Yorke
suggested, that Bede was aware of certain elements that were shared
by all the Anglo-Saxon kingdoms and which allowed him to “clas-
sify” them under one “ethnic” term. It may be that in Bede’s Historia
ecclesiastica Anglorum the gens Anglorum had some religious connotation.
Nevertheless, it took a long time for Angli to become the predo-
minant term for the population of the “Anglo-Saxon” kingdoms
(W. Liebeschuetz).
The “standard” case, however, was different. There were (name-
giving) peoples before the establishment of kingdoms that were
named after them, but the kings that founded these realms were as
important as the Roman tradition upon which they were based. The
formation of peoples and kingdoms, therefore, went hand in hand,
as Barbara Yorke suggested at the symposium in Bellagio with re-
gard to the Anglo-Saxons. This statement is probably true for all
or most cases dealt with in this volume, to a degree that, although
the kingdom was a “secondary development”, gens and regnum were
equated (Barbara Yorke). This necessarily meant a decisive trans-
formation of these so-called “Germanic” (or other “barbarian”) peo-
ples in the era subsequent to that of the creation of the kingdoms,
a change that included a “politicization” (or a further, important
628 

step of politicization) of these gentes whose names were increasingly

linked with the(ir) kingdom, whereas the kingdom, or its population,
was widely characterized by an acculturation of its different groups
(which also holds true for the Avars). It is true that this accultura-
tion always went in two directions, though the Romanization of the
“barbarians” seems to have been the more important one. It was
embedded, however, in a gradual but fundamental change which we
may continue to call the “transformation of the Roman world”. We
have to admit that even after thoroughly dealing with this subject
in a long project that led to this volume, there are still numerous
unanswered questions, problems and enigmatic areas concerning the
relationship of peoples and kingdoms. And yet having delved into
these questions, however, we see more clearly the distinctions as well
as the similarities between the peoples and kingdoms involved. The
“transformation” of gentes and regna, of gentes into regna and of regna
into new or altered gentes was an essential part of the process of the
transformation of the Roman World.
 BIBLIOGRAPHY

1. Primary sources

Agathias Myrinaei, Historiarum libri quinque, ed. R. Keydell, Corpus Fontium Historiae
Byzantinae 2 (Berlin 1967)
Alcuin, Epistolae, ed. E. Dümmler, MGH EE 4 (Berlin 1895)
Al-Maqqari, Analectes sur l’histoire et la littérature des Arabes d’Espagne par al-Makkari, ed.
R. Dozy, G. Dugat, L. Krehl and W. Wright, 2 vols. (Leiden 1855/61; repr.
Amsterdam 1967)
Ammianus Marcellinus, Res Gestae, ed. C.U. Clark, 2 vols. (Berlin 1910/15)
——, ed. J.C. Rolfe, 3 vols. (Cambridge Mass. 1935–39)
——, ed. W. Seyfarth, 4 vols. (Berlin 1970)
——, ed. W. Seyfarth, 2 vols., Bibliotheca Teubneriana (Leipzig 1978)
Anglo-Saxon Chronicle: Two of the Saxon Chronicles Parallel, ed. C. Plummer, 2 vols.
(Oxford 1899)
Annales Cambriae, ed. J. Morris [Nennius, British History and the Welsh Annals] (Chichester
1980)
Annales Fuldenses, ed. F. Kurze, MGH SSrG 7 (Hannover 1891)
Annales Laureshamenses, ed. G.H. Pertz, MGH SS 1 (Hannover 1826) pp. 22–39
Annales Nazariani, ed. G.H. Pertz, MGH SS 1 (Berlin 1826) pp. 23–31; 40–44
Annales regni Francorum, ed. F. Kurze, MGH SSrG 6 (Hannover 1895)
Annals of Ulster (to A.D. 1131), ed. S. Mac Airt and G. Mac Niocaill (Dublin 1983)
Das Annolied, ed. M. Roediger, MGH Deutsche Chroniken 1,2 (Berlin 1895) pp.
63–132
——, ed. E. Nellmann (4th edn., Stuttgart 1996)
Anonymous of Ravenna, Cosmographia, ed. M. Pinder and G. Parthey [Ravennatis
Anonymi Cosmographia et Guidonis Geographica] (Berlin 1860; repr. Aalen 1962)
——, transl. J. Schnetz, Nomina Germanica 10 [Cosmographia. Eine Erdbeschreibung
um das Jahr 700] (Uppsala 1951)
Anonymus Valesianus, Fragmenta, ed. and transl. J.C. Rolfe, Loeb Classical Library
[Ammianus Marcellinus, Res Gestae] (Cambridge Mass. 1935–39)
Anthologia latina, ed. F. Bücheler and A. Riese (Leipzig 1894)
Arbeo, Vita Corbiniani, ed. H. Glaser, F. Brunhölzl and S. Benker (München-Zürich
1983)
Arbeo, Vita Haimhramni, ed. B. Krusch, MGH SSrG 13 (Hannover 1920) pp. 1–99
Augustinus, Opera, ed. J. Divjack, CSEL 98 (Vienna 1981)
Augustinus, De civitate Dei, ed. J. Divjack, CSEL 98 [Augustinus, Opera] (Vienna
1981)
Avitus of Vienne, Opera, ed. R. Peiper, MGH AA 6,2 (München 1883)
Avitus of Vienne, Letters and Selected Prose, ed. D. Shanzer and I.N. Wood (Liverpool
2002)
Bede, Historia Ecclesiastica gentis Anglorum, ed. B. Colgrave and R.A.B. Mynors [Bede’s
Ecclesiastical History of the English People] (Oxford 1969)
Boethius, Consolatio philosophiae, ed. and transl. S.J. Tester, Loeb Classical Library
[Boethius, The theological tractates: The Consolation of Philosophy] (2nd edn., London
1973)
Breves Notitiae, ed. W. Hauthaler, Salzburger Urkundenbuch, vol. 1: Traditionskodizes
(Salzburg 1910) pp. 17–52
630 

Brut y Tywysogion or the Chronicle of the Princes. Peniarth MS 20 Version, ed. T. Jones
(Cardiff 1952)
Brut y Tywysogion or the Chronicle of the Princes. Red Book of Hergest Version, ed. T. Jones
(Cardiff 1955)
Les canons des conciles mérovingiens (VI e –VII e siècles), vol. 1, ed. J. Gaudemet and
B. Basdevant, Sources Chrétiennes 353 (Paris 1989)
Carmen de synodo Ticinensi, ed. L. Bethmann, MGH SSrL (Hannover 1878)
Cassiodorus, Variae, transl. T. Hodgkin [The Letters of Cassiodorus] (Oxford 1886)
——, ed. T. Mommsen, MGH AA 12 (Berlin 1894) pp. 1–385
——, ed. and transl. S.J.B. Barnish, Translated Texts for Historians 12 (Liverpool
1992)
Cassius Dio, Roman History, vol. 9, ed. and transl. E. Cary (London-Cambridge
Mass. 1969)
Chartae Latinae Antiquiores, ed. A. Bruckner and R. Marichal (Zurich 1982)
Chronica Byzantia-Arabica, ed. J. Gil, Corpus Scriptorum Muzarabicorum 1 (Madrid
1973) pp. 7–14
Chronica Muzarabica, ed. J. Gil, Corpus Scriptorum Muzarabicorum 1 (Madrid 1973)
pp. 15–54 [transl. K.B. Wolf, Conquerors and Chroniclers of Early Medieval Spain
(Liverpool 1990) pp. 111–58]
Chronicorum a. CCCCLII, ed. T. Mommsen, MGH AA 9 (Berlin 1892) pp. 615–66
Chronicorum a. DXI, ed. T. Mommsen, MGH AA 9 (Berlin 1892) pp. 615–66
Chroniques Asturiennes ( fin IXe siècle), ed. and transl. Y. Bonnaz (Paris 1987)
Cicero, Academicorum reliquiae cum Lucullo, ed. O. Plasberg (Leipzig 1969, repr. 1996)
Claudian, De bello Pollentino sive Gothico, ed. T. Birt, MGH AA 10 (Berlin 1892; repr.
1962) pp. 259–83
Codex Euricianus, ed. K. Zeumer, MGH LL nationum Germanicarum 1 (Hannover
1902) pp. 1–32
Codex Iustinianus, ed. P. Krüger (Berlin 1877)
Codex Theodosianus, ed. T. Mommsen and P.M. Meyer, 3 vols. (3rd edn., Berlin 1962)
Codice Diplomatico Longobardo, ed. L. Schiaparelli, 2 vols., Fonti per la storia d’Italia
(Rome 1929/33)
Codices Latini Antiquiores, ed. E.A. Lowe, 11 vols. and supplement (Oxford 1934–71)
La Colección Canónica Hispana, ed. G. Martínez Díez and F. Rodríguez, 5 vols. (Madrid
1966–92)
Concilios visigóticos e hispano-romanos, ed. J. Vives, España cristiana 1 (Madrid-Barcelona
1963)
Concilium Toletanum, ed. J. Vives, Concilios Visigóticos e Hispano-romanos, España cris-
tiana 1 (Madrid-Barcelona 1963)
Concilium universale Constantinopolitanum sub Justiniano habitum, ed. E. Schwartz, Acta
conciliorum oecumenicorum 4,2 (Straßburg 1914)
Conquerors and Chroniclers of Early Medieval Spain, transl. K.B. Wolf (Liverpool 1990)
Constantius of Lyon, Vie de St Germain d’Auxerre, ed. R. Borius, Sources Chrétiennes
112 (Paris 1965)
Continuatio Havniensis Prosperi, ed. T. Mommsen, MGH AA 9 (Berlin 1892; repr.
Munich 1981) pp. 298–339
Conversio Bagoariorum et Carantanorum, ed. F. Lo“ek, MGH Studien und Texte 15
(Hannover 1997)
Corippus, Eloge de l’empereur Justin II, ed. S. Antès (Paris 1981)
Crónica del Moro Rasis, ed. D. Catalán, D. de Andrés and M.S. de Andrés (Madrid
1974)
Deuxippus, Fragmenta, ed. C. Müller, Fragmenta Historicorum Graecorum 3 (Paris
1849; repr. 1928)
Diplomata, Chartae, Epistolae aliaque instrumenta ad res Gallo-Francicas spectantia, ed. J.-M.
Pardessus, 2 vols. (Paris 1843/49)
  631

Diplomatica Hispano-visigoda, ed. A. Canellas Lopez (Zaragoza 1979)

Dracontius, Romulea, ed. F. Vollmer, MGH AA 14 (Berlin 1905) pp. 132–96
Early Welsh Genealogical Tracts, ed. P.C. Bartrum (Cardiff 1966)
Edictus Rothari, ed. F. Bluhme, MGH LL 4 (Hannover 1868) pp. 1–90
Edictus Theodorici, ed. J. Baviera, Fontes Iuris Romani Anteiustiniani 2 (2nd edn., Florence
1968–69) pp. 683–710
Egilbert, Vita Ermenfredi, ed. J. Carnandet, Acta Sanctorum Ordinis sancti Benedicti
(Paris-Rom 1867) pp. 106–13
Einhard, Vita Karoli, ed. O. Holder-Egger, MGH SSrG 25 (6th edn., Hannover 1911)
Ekkehard, Waltharius of Gaeraldus, ed. A.K. Bate (Reading 1978)
Ennodius, Opera, ed. W. Hartel, CSEL 6 (Vienna 1882)
——, ed. F. Vogel, MGH AA 7 (Berlin 1885)
Ennodius, Panegyricus Theoderico, ed. W. Hartel, CSEL 6 [Ennodius, Opera] (Vienna 1882)
——, ed. F. Vogel, MGH AA 7 (Berlin 1885) pp. 203–14
Ennodius, Vita Epifani, ed. W. Hartel, CSEL 6 [Ennodius, Opera] (Vienna 1882)
——, ed. F. Vogel, MGH AA 7 (Berlin 1885) pp. 84–109
Epistolae Arelatenses Genuinae, ed. W. Gundlach, MGH EE 3 (Berlin 1892) pp. 1–83
Epistolae Austrasicae, ed. W. Gundlach, MGH EE 3 (Berlin 1892) pp. 110–53
Epistulae Wisigothicae, ed. J. Gil, Miscellanea Wisigothica, Publicaciones de la Universidad
de Sevilla (Sevilla 1972) pp. 1–49
Epitome de Caesaribus, ed. F. Pichelmayer (Leipzig 1961)
Eugippius, Vita Sancti Severini, ed. R. Noll and E. Vetter, Schriften und Quellen der
Alten Welt 11 (Berlin 1963)
Eulogius, Apologeticus martyrum, ed. J. Gil, Corpus Scriptorum Muzarabicorum 1
(Madrid 1973) pp. 475–95
Eutropius, Breviarium ab urbe condita, ed. F.L. Müller (Stuttgart 1995)
Fasti Vindobonenses Priores, ed. T. Mommsen, MGH AA 9 (Berlin 1892) pp. 274–336
Felix, Life of St Guthlac, ed. B. Colgrave (Cambridge 1956)
Ferrandus, Vita Fulgentii, PL 65, cols. 117–50
——, transl. R.B. Eno, Fathers of the Church 95 [Fulgentius Selected Works] (Washington
1996) pp. 4–56
Fontes Iuris Romani Anteiustiniani, ed. S. Riccobono et al., 3 vols. (2nd edn., Florence
1968–69)
Fredegar, Chronicon, ed. B. Krusch, MGH SSrM 2 (Hannover 1888) pp. 1–193
Die Gesetze der Angelsachsen, vol. 1, ed. F. Liebermann (Halle 1903)
Gildas, The Ruin of Britain and other works, ed. and transl. M. Winterbottom (London-
Chichester 1978)
The Gododdin of Aneirin: Text and Context from Dark Age North Britain, ed. J.T. Koch
(Cardiff 1997)
Gregorii I Papae Registrum Epistolarum, ed. P. Ewald, MGH EE 1 (Berlin 1887)
Gregory of Tours, Historiae, ed. B. Krusch and W. Levison, MGH SSrM 1,1 (2nd
edn., Hannover 1951)
——, ed. R. Buchner [Zehn Bücher Geschichten] (Berlin 1955)
Gregory of Tours, Liber de Virtutibus sancti Juliani, ed. B. Krusch, MGH SSrM 1,2
(Hannover 1885) pp. 112–34
Griechische und lateinische Quellen zur Frühgeschichte Mitteleuropas bis zur Mitte des 1.
Jahrtausends u. Z., vol. 3: Von Tacitus bis Ausonius, ed. J. Herrmann, Schriften und
Quellen der Alten Welt 37,3 (Berlin 1991)
Hildegar of Meaux, Vita Faronis, ed. J. Mabillon, Acta Sanctorum Ordinis sancti
Benedicti (Mâcon 1936) pp. 606–25
Historia Brittonum, ed. J. Morris [Nennius, British History and the Welsh Annals] (Chichester
1980)
Hydatius, Chronicon, ed. T. Mommsen, MGH AA 11 [Continuatio Chronicorum
Hieronymianorum] (Berlin 1894) pp. 1–36
632 

——, ed. A. Tranoy, 2 vols., Sources Chrétiennes 218–219 (Paris 1974)

——, ed. R.W. Burgess [The Chronicle of Hydatius and the Consularia Constantinopolitana]
(Oxford 1993)
Ibn al-Fara∂ì, Ta"rikh 'ulama" al-Andalus, ed. F. Codera and J. Ribera y Tarragó,
2 vols., Bibliotheca Arabico-Hispana (Madrid 1891/92)
Ibn al-Qùtßìya, Historia de la conquista de España de Abenalcotía de Córdoba, ed.
P. Gayangos, E. Saavedra and F. Codera (Madrid 1868) [repr. with Spanish
transl. by J. Ribera y Tarragó (Madrid 1926)]
Ibn Óabìb, Kitàb al-Ta"rìkh, ed. J. Aguadé (Madrid 1991)
Ibn Óazm, Jamharat ansab al-'arab, ed. 'A.M. Harun (Cairo 1971)
Ibn Idhàrì, Al-Bayàn al-Mughrib, ed. R. Dozy, E. Lévi-Provençal and G. Colin,
2 vols. (Leiden 1948/51)
Ibn Khaldûn, The Muqaddimah, an Introduction to History, transl. F. Rosenthal, 3 vols.
(London 1958)
Ildefonsus de Toledo, De viris illustribus, ed. C. Codoñer [El “de viris illustribus” de
Ildefonso de Toledo. Estudio y edición crítica] (Salamanca 1972)
Interpolationes chronicis Prosperi insertae saec. XV, ed. T. Mommsen, MGH AA 9,1 (Berlin
1882)
Isidore of Seville, Chronica maiora, ed. T. Mommsen, MGH AA 11 (Berlin 1894) pp.
391–506
Isidore of Seville, Etymologiarum sive originum libri XX, ed. W.M. Lindsay, 2 vols.
(Oxford 1911)
——, ed. J. Oroz Reta and M.A. Marcos Casquero, 2 vols. (Madrid 1982)
Isidore of Seville, Historia Gothorum, Wandalorum, Sueborum, ed. T. Mommsen, MGH
AA 11 (Berlin 1894) pp. 241–303
——, transl. G. Donini and G.B. Ford, History of the Goths, Vandals and Suevi (Leiden
1970)
——, ed. C. Rodríguez Alonso [Las historias de los godos, vándalos y suevos de Isidoro de
Sevilla. Estudio, edición crítica y traducción] (León 1975)
——, transl. K.B. Wolf [Conquerors and Chroniclers of Early Medieval Spain: History of
the Kings of the Goths] (Liverpool 1990) pp. 81–110
Isidore of Seville, De viris illustribus, ed. C. Codoñer Merino (Salamanca 1964)
Jerome, Lettres, ed. J. Labourt, vol. 7 (Paris 1961)
John of Antioch, Fragmenta, ed. C. Müller, Fragmenta Historicorum Graecorum, vols.
4 and 5 (Paris 1868/70) [transl. C.D. Gordon, The Age of Attila (Ann Arbor 1960)]
John of Biclar, Chronica, ed. T. Mommsen, MGH AA 11 (Berlin 1894) pp. 207–20
Jonas of Bobbio, Vita Columbani, ed. B. Krusch, MGH SSrG 37 (Hannover 1905)
Jordanes, Getica, ed. T. Mommsen, MGH AA 5,1 (Berlin 1882) pp. 53–138
Jordanes, De origine actibusque Getarum, ed. F. Giunta and A. Grillone, Fonti per la
storia d’Italia 117 (Rome 1991)
Justinian, Novellae, ed. R. Scholl (Berlin 1895)
Die Kaiserchronik eines Regensburger Geistlichen, ed. E. Schröder, MGH Deutsche Chroniken
1,1 (Hannover 1892)
Laws of Æthelberht, ed. F. Liebermann, Die Gesetze der Angelsachsen 1 (Halle 1903)
Laws of Ine, ed. F. Liebermann, Die Gesetze der Angelsachsen 1 (Halle 1903)
Laws of Wihtred, ed. F. Liebermann, Die Gesetze der Angelsachsen 1 (Halle 1903)
Leges Ahistulfi, ed. F. Bluhme, MGH LL 4 (Hannover 1868) pp. 194–205
Leges Alamannorum, ed. K. Lehmann and K.A. Eckhardt, MGH LL nationum
Germanicarum 5,1 (rev. 2nd edn., Hannover 1966)
Leges Liutprandi, ed. F. Bluhme, MGH LL 4 (Hannover 1868) pp. 96–182
Leges Visigothorum antiquiores, ed. K. Zeumer, MGH Fontes iuris Germanici antiqui
5 (Hannover 1894)
Lex Baiwariorum, ed. E. von Schwind, MGH LL nationum Germanicarum 5,2
(Hannover 1926)
  633

Lex Frisionum, ed. K.F. von Richthofen, MGH LL 3 (Hannover 1863) pp. 656–697
Lex Ribuaria, ed. F. Beyerle and R. Buchner, MGH LL nationum Germanicarum
3,2 (Hannover 1954)
Lex Salica, ed. K.A. Eckhardt, MGH LL nationum Germanicarum 4,2 (Hannover
1969)
Lex Saxonum, ed. K.F. von Richthofen, MGH LL 5 (Hannover 1875–89) pp. 47–84
Lex Thuringorum, ed. K.F. von Richthofen, MGH LL 5 (Hannover 1875–89) pp.
103–44
Lex Visigothorum, ed. K. Zeumer, MGH LL nationum Germanicarum 1 (Hannover
1902) pp. 33–456
Liber Constitutionum, ed. L.R. von Salis, MGH LL nationum Germanicarum 2,1
(Hannover 1892) pp. 1–122
Liber historiae Francorum, ed. B. Krusch, MGH SSrM 2 (Hannover 1888) pp. 215–328
Liber pontificalis, ed. L. Duchesne, 3 vols., Bibliothèque des Ecoles Françaises d‘Athènes
et de Rome, 2e sér. (Paris 1886; repr. 1955)
Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historicorum Byzantinae
(Bonn 1831)
Malchus of Philadelphia, Fragmenta, ed. and transl. R.C. Blockley, The Fragmentary
Classicising Historians of the Later Roman Empire 2 (Liverpool 1882; repr. 1983)
pp. 378–455
Marius of Avenches, Chronique, ed. J. Favrod (Lausanne 1991)
Maurikios, Strategikon, ed. G.T. Dennis, transl. E. Gamillscheg, Corpus Fontium
Historiae Byzantinae 17 (Wien 1981)
Menander Protector, Fragmenta, ed. R.C. Blockley, Arca 17 [The History of Menander
the Guardsman] (Liverpool 1985)
MGH Capitularia regum Francorum, vol. 1, ed. A. Boretius (Hannover 1883)
MGH Concilia aevi Merovingici, ed. F. Maassen (Hannover 1893)
MGH DD regum Francorum e stirpe Merovingica, ed. T. Kölzer, 2 vols. (Hannover
2001)
MGH Formulae Merowingici et Karolini Aevi, ed. K. Zeumer (Hannover 1886)
Miracula Sancti Demetrii, ed. P. Lemerle, 2 vols. [Les plus anciens recueils des miracles de
Saint Démétrius et la pénétration des Slavs dans les Balkans] (Paris 1979/81)
Nestorchronik, ed. L. Müller (München 2001)
Die nichtliterarischen lateinischen Papyri Italiens aus der Zeit 445–700, ed. J.-O. Tjäder,
3 vols. (Lund-Stockholm 1954–82)
Nikephoros Patriarcha, Breviarium Historicum, ed. C. Mango (Washington DC 1990)
Nithard, Historiarum libri IIII, ed. E. Müller, MGH SSrG 44 (Hannover 1907)
Notitia Dignitatum accedunt Notitia Urbis Constantinopolitanae et Latercula Prouinciarum, ed.
O. Seeck (Berlin 1876; repr. Frankfurt a. M. 1962)
Olympiodorus of Thebes, Fragmenta, ed. and transl. R.C. Blockley, The Fragmentary
Classicising Historians of the Later Roman Empire 2 (Liverpool 1982; repr. 1983)
pp. 151–209
Origo gentis Langobardorum, ed. G. Waitz, MGH SSrL (Hannover 1878)
Origo gentis Langobardorum codicis Gothani, ed. G. Waitz, MGH SSrL (Hannover 1878)
Orosius, Historiarum adversum paganos libri VII, ed. K. Zangemeister, CSEL 5 (Vienna
1882; repr. Hildesheim 1967)
Otfrid of Weißenburg, Evangelienbuch, ed. O. Erdmann (Halle 1882; repr. 6th edn.
Tübingen 1973)
Pactus legis Salicae, ed. K.A. Eckhardt, Germanenrechte 1 (Göttingen 1954)
——, ed. K.A. Eckhardt, MGH LL nationum Germanicarum 4,1 (Hannover 1962)
Panegyrici latini, ed. R.A.B. Mynors (Oxford 1964)
Passio Sigismundi, ed. B. Krusch, MGH SSrM 2 (Hannover 1888) pp. 329–40
Paul the Deacon, Historia Langobardorum, ed. L. Bethmann and G. Waitz, MGH
SSrL (Hannover 1878) pp. 12–187
634 

Paul the Deacon, Historia Romana, ed. H. Droysen, MGH SSrG 49 (Hannover 1879)
——, ed. A. Crivellucci, Fonti per la storia d’Italia 51 (Rome 1914)
Paulinus, Vita Sancti Ambrosii, ed. M. Pellegrino (Rom 1961)
Pliny, Historia Naturalis, ed. and transl. H.R. Rackham, 10 vols. (London-Cambridge
1949–68)
——, ed. G. Winkler and R. König (Munich 1988)
The Poems of Taliesin, ed. I. Williams (Dublin 1987)
Possidius, Vita Augustini, PL 32, cols. 33–66
——, transl. M.M. Muller and R.J. Deferrari, Early Christian Biographies (Washington
1952) pp. 73–124
Priscus of Panium, Fragmenta, ed. and transl. R.C. Blockley, The Fragmentary
Classicising Historians of the Later Roman Empire 2 (Liverpool 1982) pp. 221–377
——, transl. E. Doblhofer, Byzantinische Geschichtsschreiber 4 [Byzantinische Diplo-
maten und östliche Barbaren. Aus den Excerpta de legationibus des Konstantin Porphyrogenetos
ausgewählte Abschnitte des Priskos und Menander Protector] (Graz-Wien-Köln 1955)
Procopius, Gotenkriege, ed. O. Veh [Prokop, Werke] (München 1966)
Procopius, Wars, ed. J. Haury, transl. H.B. Dewing, 7 vols., Loeb Classical Library
(London-Cambridge Mass. 1914–28)
Prosper of Aquitaine, Epitoma Chronicon, ed. T. Mommsen, MGH AA 9 (Berlin 1892;
repr. Munich 1981) pp. 341–499
Ptolemy, Geographia, ed. C. Müller (Paris 1883)
——, ed. C.F. Nobbe and A. Diller (Hildesheim 1966)
Quellen zur Geschichte der Alamannen, vol. 2, transl. C. Dirlmeier and G. Gottlieb,
Schriften der Heidelberger Akademie der Wissenschaften, Kommission für Alaman-
nische Altertumskunde 3 (Heidelberg-Sigmaringen 1978)
Regino of Prüm, Chronicon, ed. F. Kurze, MGH SSrG 50 (Hannover 1890)
Salvian, De Gubernatione Dei Libri VIII, ed. F. Pauly, CSEL 8 (Vienna 1883)
Sidonius Apollinaris, ed. A. Loyen, 3 vols. (Paris 1960–70)
Sidonius Apollinaris, Poems and Letters, ed. and transl. W.B. Anderson, 2 vols., Loeb
Classical Library (Cambridge Mass. 1936/65)
Sozomen, Kirchengeschichte, ed. J. Bidez (Berlin 1960)
Tablettes Albertini, actes privés de l’époque vandale, ed. C. Courtois, L. Leschi, C. Perrat
and C. Saumagne (Paris 1952)
Tacitus, Germania, ed. R. Much and W. Lange with H. Jankuhn (3rd edn., Heidelberg
1967)
——, ed. A. Lund (Heidelberg 1988)
——, ed. G. Perl, Griechische und lateinische Quellen zur Frühgeschichte Mitteleuropas bis zur
Mitte des 1. Jahrtausends unserer Zeit, vol. 2, ed. J. Herrmann, Schriften und Quellen
der Alten Welt 37,2 (Berlin 1990)
Themistius, Orationes, ed. G. Downey (Leipzig 1965)
Theophanes, Chronographia, ed. C. de Boor, 2 vols. (Leipzig 1883)
Theophylactus Simocatta, Historiae, ed. C. de Boor (Leipzig 1887)
Trioedd Ynys Prydain, ed. R. Bromwich (Cardiff 1978)
Velleius Paterculus, Historiarum libri duo, ed. W.S. Watt, Bibliotheca Teubneriana
(Leipzig 1988)
Venantius Fortunatus, Vita Sancti Martini, ed. F. Leo, MGH AA 4,1 (Berlin 1881)
pp. 293–370
Victor of Vita, Historia persecutionis Africanae provinciae, ed. C. Halm, MGH AA 3,1
(Berlin 1879) pp. 1–58
——, ed. M. Petschenig, CSEL 7 (Bonn 1881)
——, transl. J. Moorhead, Liverpool Translated Texts for Historians 10 [Victor of
Vita: History of the Vandal Persecution] (Liverpool 1992)]
Vita Abbatum Acaunensium absque epitaphiis, ed. B. Krusch, MGH SSrM 7 (Hannover
1920) pp. 322–36
  635

Vita Altmanni, ed. W. Wattenbach, MGH SS 12 (Berlin 1856) pp. 226–43

Vita Patrum Iurensium, ed. F. Martine, Sources Chrétiennes 142 [Vie des Pères du Jura]
(Paris 1968)
Vita Sadalbergae, ed. B. Krusch, MGH SSrM 5 (Hannover-Leipzig 1910) pp. 40–66
Vitae Sanctorum Britanniae et Genealogiae, ed. A.W. Wade-Evans (Cardiff 1944)
Vitas Sanctorum Patrum Emeretensium, ed. A. Maya Sánchez, Corpus Christianorum
Series Latina 116 (Turnhout 1992)
Widsith, ed. K. Malone (Copenhagen 1962)
The works of the Emperor Julian, ed. W.C. Wright (New York 1930)
Zosimus, Historia Nova, ed. J.J. Buchanan and H.T. Davis (San Antonio 1967)
——, ed. and transl. F. Paschoud, 4 vols. (Paris 1971–89)
——, ed. O. Veh and S. Rebenich, Bibliothek der griechischen Literatur 31 (Stuttgart
1990)

2. Secondary literature

799—Kunst und Kultur der Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn.
Beiträge zum Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann and
M. Wemhoff (Mainz 1999)
H. Abdul Wahab and F. Dachraoui, “Le régime foncier en Sicile aux IXème–
Xème siècles: edition et traduction d’un chapitre du ‘Kitab al-amwal’ de al-
Dawudi”, Etudes d’orientalisme dédiées à la mémoire de Lévi-Provençal 2 (Paris 1962)
pp. 401–55
L. Abrams, “Edward the Elder’s Danelaw”, Edward the Elder 899–924, ed. N. Higham
and D. Hill (Manchester 2001) pp. 128–43
Acta XIII congressus internationalis archaeologiae christianae, Split—Poreè 1994, 3 vols., ed.
N. Cambi and E. Marin, Studi di antichità cristiana 54. Vjesnik za arheologiju
i historiju Dalmatinsku 87/89 Suppl. (Split 1998)
Actas del I Congreso Andaluz de Estudios Clásicos ( Jaén 1982)
Actas del II Congreso Andaluz de Estudios Clásicos, 3 vols. (Málaga 1987)
Actas del XIV Centenario del Concilio III de Toledo, 589–1989 (Toledo 1991)
Actes du 8ème Congrès international de numismatique 1973 (Paris-Basel 1976)
After Empire. Towards an Ethnology of Europe’s Barbarians, ed. G. Ausenda, Studies in
Historical Archaeoethnology (Woodbridge 1995)
After Rome’s Fall. Narrators and Sources of Early Medieval History. Essays presented to Walter
Goffart, ed. A.C. Murray (Toronto-Buffalo-London 1998)
The Age of Sutton Hoo, ed. M. Carver (Woodbridge 1992)
B. Ager, “The alternative quoit brooch: an update”, Anglo-Saxon Cemeteries: A Reappraisal,
ed. E. Southworth (Stroud 1990) pp. 153–61
Die Alamannen. Ausstellungskatalog Stuttgart, Zürich, Augsburg (Stuttgart 1997)
L. Alcock, “Gwyr y Gogledd: an archaeological appraisal”, Archaeologia Cambrensis 132
(1983) pp. 1–18
——, Economy, Society and Warfare Among the Britons and Saxons (Cardiff 1987)
——, Cadbury Castle, Somerset: the Early Medieval Archaeology (Cardiff 1995)
Althochdeutsch, vol. 2: Wörter und Namen. Forschungsgeschichte, ed. R. Bergmann,
H. Tiefenbach and L. Voetz, Germanische Bibliothek NF, Reihe 3, Untersuchungen
(Heidelberg 1987)
J. Alvarado Planas, El problema del germanismo en el derecho español (Madrid 1997)
H. Ament, Fränkische Adelsgräber von Flonheim in Rheinhessen, Germanische Denkmäler
der Völkerwanderungszeit B5 (Berlin 1970) pp. 130–63
——, “Das Gräberfeld von Dieue-sur-Meuse, ein Bestattungsplatz von Franken und
Romanen”, Acta Praehistorica et Archaeologica 7/8 (1976/77) pp. 301–9
——, “Franken und Romanen im Merowingerreich als archäologisches Forschungs-
problem”, Bonner Jahrbücher 178 (1978) pp. 377–94
636 

——, Eschborn, Main-Taunus-Kreis, Grabfunde des 5. Jahrhunderts: ein alamannisches Gräberfeld

an der Wende vom Altertum zum Mittelalter, Archäologische Denkmäler in Hessen 41
(Wiesbaden 1984)
——, “Die Ethnogenese der Germanen aus der Sicht der Vor- und Frühgeschichte”,
Ethnogenese europäischer Völker. Aus der Sicht der Anthropologie und Vor- und Frühgeschichte,
ed. W. Bernhard and A. Kandler-Pálsson (Stuttgart-New York 1986) pp. 247–56
——, “Romanen an Rhein und Mosel im frühen Mittelalter”, Bonner Jahrbücher 192
(1992) pp. 261–71
——, Das alamannische Gräberfeld von Eschborn (Main-Taunus-Kreis), Materialien zur Vor-
und Frühgeschichte in Hessen 14 (Wiesbaden 1992)
——, “Die Franken in den Römerstädten der Rheinzone”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 129–37
H. Ament, H.H. Anton, H. Beck, A. Quak, F. Rexroth, K. Schäferdiek, H. Steuer,
D. Strauch, N. Voorwinden, “Franken”, Reallexikon der Germanischen Altertumskunde
9 (2nd edn., 1995) pp. 373–461
P. Amory, “The meaning and purpose of ethnic terminology in the Burgundian
laws”, Early Medieval Europe 2 (1993) pp. 1–28
——, “Names, ethnic identity, and community in fifth- and sixth-century Burgundy”,
Viator 25 (1994) pp. 1–30
——, People and Identity in Ostrogothic Italy, 489–534, Cambridge Studies in Medie-
val Life and Thought (Cambridge 1997)
H. Amrein, A. Rast-Eicher and R. Windler with a contribution by E. Langenegger,
“Neue Untersuchungen zum Frauengrab des 7. Jahrhunderts in der reformierten
Kirche von Bülach (Kanton Zürich)”, Zeitschrift für Schweizerische Archäologie und
Kunstgeschichte 56 (1999) pp. 73–114
T. Anderson Jr., “Roman military colonies in Gaul, Salian ethnogenesis and the
forgotten meaning of Pactus Legis Salicae 59.5”, Early Medieval Europe 4 (1995) pp.
129–44
Die Anfänge des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberösterreichischen
Landesarchivs, Ergänzungsband 2 (Linz 1978)
A. Angenendt, “Le parrainage dans le haut Moyen Âge. Du rituel liturgique au
cérémonial politique”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événe-
ment, ed. M. Rouche (Paris 1997) pp. 243–54
Anglo-Saxon Cemeteries 1979: the fourth Anglo-Saxon Symposium at Oxford, ed. P.A. Rahtz,
T. Dickinson and L. Watt (Oxford 1980)
Anglo-Saxon Cemeteries: A Reappraisal, ed. E. Southworth (Stroud 1990)
The Anglo-Saxon Cemetery on Mill Hill, Deal, Kent, ed. K. Parfitt and B. Brugmann,
Society for Medieval Archaeology, Monograph series 14 (London 1997)
The Anglo-Saxons from the Migration Period to the Eighth Century: An Ethnographic Perspective,
ed. J. Hines (Woodbridge 1997)
H.H. Anton, “Antike Großländer, politisch-kirchliche Traditionen und mittelalter-
liche Reichsbildung”, Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Kanonistische
Abteilung 86 (2000) pp. 33–85
——, “Troja-Herkunft, origo gentis und frühe Verfaßtheit der Franken in der gal-
lisch-fränkischen Tradition des 5. bis 8. Jahrhunderts”, Mitteilungen des Instituts für
Österreichische Geschichtsforschung 108 (2000) pp. 1–30
H. Aouni, “Das spätantik-frühmittelalterliche Gräberfeld von Jülich—die ‘einfachen’
Gürtelgarnituren”, Archäologie des Frankenreiches—Neueste Forschungen. Internationales
Kolloquium anläßlich der Ausstellung “Die Franken—Les Francs” Berlin, 10. bis 12. Oktober
1997, ed. W. Menghin, Acta Praehistorica et Archaeologica 30 (1998) pp. 19–37
J. Arce, “La Notitia Dignitatum et l’armée romaine dans la Diocesis Hispaniarum”,
Chiron 10 (1980) pp. 539 ff.
  637

——, “La Penisola Iberica”, Storia di Roma, vol. 3,2: L’età tardoantica. I luoghi e le cul-
ture, ed. A. Momigliano (Turin 1993) pp. 379–404
——, “El catastrofismo de Hydacio y los camellos de la Gallaecia”, Los últimos
romanos en Lusitania, ed. A. Velázquez et al. (Mérida 1995) pp. 219–29
——, El último siglo de la España romana (284–409 d.C.) (3rd repr., Madrid 1997)
pp. 31–63
——, “Los gobernadores de la Diocesis Hispaniarum (ss. IV–V) y la continuidad
de las estructuras administrativas romanas en la Peninsula Ibérica”, Antiquité Tardive
7 (1999) pp. 73–83
——, “La epistula de Honorio a las tropas de Pompaelo: comunicaciones, ejército
y moneda en Hispania (s.IV–V d.C.)”, Rutas, ciudades y moneda en Hispania, ed.
P. Centeno, M.P. García y Bellido and G. Mora, Anejos de Archivo Español de
Arqueología 20 (Madrid 1999) pp. 461–88
——, Esperando a los bárbaros en Hispania (409–507 A.D.) (forthcoming)
Archaeological Approaches to Cultural Identity, ed. S. Shennan (London-New York 1989)
Archaeological Theory in Europe. The Last Three Decades, ed. I. Hodder (London-New
York 1991)
Archaeometrical Research in Hungary, ed. M. Járó and L. Költö (Budapest 1988)
Archäologie des Frankenreiches—Neueste Forschungen. Internationales Kolloquium anläßlich der
Ausstellung “Die Franken—Les Francs” Berlin, 10. bis 12. Oktober 1997, ed. W. Menghin,
Acta Praehistorica et Archaeologica 30 (1998)
M. Archibald, M. Brown and L. Webster, “Heirs of Rome: the shaping of Britain
A.D. 400–900”, The Transformation of the Roman World, ed. L. Webster and
M. Brown (London 1997) pp. 208–48
Armut, Liebe, Ehre. Studien zur historischen Kulturforschung, ed. R. van Dülmen (Frankfurt
a. M. 1988)
C. Arnold, Anglo-Saxon Cemeteries of the Isle of Wight (London 1982)
——, The Archaeology of Anglo-Saxon Kingdoms (2nd edn., London 1997)
W. Ashburner, “The Farmer’s Law”, Journal of Hellenic Studies 32 (1912) pp. 68–95
Atti del Convegno Internazionale sul Tema: La civiltà dei Longobardi in Europa. Roma 1971
(Rome 1974)
Atti del primo Congresso internazionale di Studi Longobardi (Spoleto 1952)
Atti del VI° Congresso internazionale di studi sull’alto medioevo (Spoleto 1980)
Aufstieg und Niedergang des römischen Welt II 29,2, ed. H. Temporini and W. Haase
(Berlin 1983)
Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner, Schriftenreihe zur
bayerischen Landesgeschichte 62 (München 1962)
Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters. Methodische
Grundlagendiskussion im Grenzbereich zwischen Archäologie und Geschichte, vol. 1, ed.
M. Müller-Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen 1993)
Die Ausgrabungen in St. Ulrich und Afra in Augsburg 1961–1968, ed. J. Werner, Münchener
Beiträge zur Vor- und Frühgeschichte 23 (München 1977)
A. Avenarius, “Struktur und Organisation der europäischen Steppenvölker”, Popoli
delle steppe: Unni, Avari, Magiari, Settimane di studio del centro italiano di studi
sull’alto medioevo 35 (Spoleto 1988) pp. 125–50
Die Awaren am Rand der byzantinischen Welt. Studien zu Diplomatie, Handel und Technologietransfer
im Frühmittelalter, ed. F. Daim, Monographien zur Frühgeschichte und Mittelal-
terarchäologie 7 (Innsbruck 2000)
Awarenforschungen, 2 vols., ed. F. Daim, Archaeologia Austriaca 1. Studien zur
Archäologie der Awaren 4 (Wien 1992)
Das awarenzeitlich gepidische Gräberfeld von Kölked—Feketekapu A, ed. A. Kiss and F. Daim,
Monographien zur Frühgeschichte und Mittelalterarchäologie 2. Studien zur
Archäologie der Awaren 5 (Innsbruck 1996)
638 

Az Alföld a 9. szászadban, ed. G. Lõrinczy (Szeged 1993)

V. Babucke, “Nach Osten bis an den Lech. Zur alamannischen Besiedlung der
westlichen Raetia secunda”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg
(Stuttgart 1997) pp. 249–60
B.S. Bachrach, Merovingian Military Organisation 481–751 (Minneapolis 1972)
Baiernzeit in Oberösterreich. Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum
Ende des 8. Jahrhunderts, Kataloge des Oberösterreichischen Landesmuseum 97
(Linz 1977)
Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988)
K. Bakay, “Az avarkor idõrendjéõl. Újabb avar temetök a Balaton Környékén [Zur
Chronologie der Awarenzeit. Neue awarenzeitliche Gräberfelder in der Umgegend
des Plattensees]”, Somogyi Múzeumok Közleményei 1 (1973) pp. 1 ff.
A. Balil, “Aspectos sociales del Bajo Imperio en Hispania, IV–VI”, Latomus 24 (1965)
pp. 886–904
C. Bálint, Die Archäologie der Steppe (Wien-Köln 1989)
——, Die spätawarenzeitliche Siedlung von Eperjes, Varia Archaeologica Hungarica 4
(Budapest 1991)
——, “Kontakte zwischen Iran, Byzanz und der Steppe. Das Grab von Üè Tepe
(Sowj. Azerbajd_an) und der beschlagverzierte Gürtel im 6. und 7. Jahrhundert”,
Awarenforschungen 1, ed. F. Daim, Archaeologia Austriaca 1. Studien zur Archäologie
der Awaren 4 (Wien 1992) pp. 309–496
——, “Probleme der archäologischen Forschung zur awarischen Landnahme”,
Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters. Methodische
Grundlagendiskussion im Grenzbereich zwischen Archäologie und Geschichte 1, ed. M. Müller-
Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen 1993) pp.
195–273
——, “Byzantinisches zur Herkunftsfrage des vielteiligen Gürtels”, Kontakte zwischen
Iran, Byzanz und der Steppe im 6.–7. Jahrhundert, ed. id., Varia Archaeologica Hungarica
10 (Budapest 2000) pp. 99–162
——, “Der Goldschatz von Nagyszentmiklós”, Varia Archaeologica Hungarica
(Budapest forthcoming c. 2003)
C.J. Balkwill, “Old English wic and the origins of the hundred”, Landscape History
15 (1993) pp. 5–11
K. Banghard, “Eine frühmittelalterliche Gürtelgarnitur und ihre Motivgeschichte”,
Archäologische Nachrichten aus Baden 59 (1998) pp. 24–35
M. Banniard, Viva voce. Communication écrite et communication orale du IV e au IX e siècle en
Occident latin (Paris 1992)
A. Barbero, “El pensamiento político visigodo y las primeras unciones regias en la
Europa medieval” La sociedad visigoda y su entorno histórico, ed. J. Faci (Barcelona
1992) pp. 1–77
A. Barbero and M. Vigil, La formación del feudalismo en la Península Ibérica (Barcelona
1978)
M. Barceló, “El Califa patente: el ceremonial omeya de Cordoba o la escenificación
del poder”, Estructuras y formas del poder en la historia, ed. R. Pastor, I. Kieniewicz,
E. García de Enterría et al. (Salamanca 1991) pp. 51–71
R. Bärenfänger, “Fundmeldungen: 6. Leer FstNr. 2710/5:10”, Emdener Jahrbuch 77
(1997) p. 269
P. Barford, The Early Slavs (Ithaca 2001)
L. Barkóczi, “Das Gräberfeld von Keszthely-Fenekpuszta aus dem 6. Jahrhundert
und die frühmittelalterlichen Bevölkerungsverhältnisse am Plattensee”, Jahrbuch des
Römisch-Germanischen Zentralmuseums Mainz 18 (1971) pp. 179–91
J. Barlow, “Gregory of Tours and the Myth of the Trojan Origins of the Franks”,
Frühmittelalterliche Studien 29 (1995) pp. 86–95
  639

S.J.B. Barnish, “Taxation, Land and barbarian settlement in the Western Empire”,
Papers of the British School at Rome 54 (1986) pp. 170–95
——, “Maximian, Cassiodorus, Boethius, Theodahad: literature, philosophy and
politics in Ostrogothic Italy”, Nottingham Medieval Studies 34 (1990) pp. 16–32
P.S. Barnwell, Emperors, Prefects and Kings. The Roman West 395–565 (London 1992)
——, “Hlafæta, ceorl, hid and scir: Celtic, Roman or Germanic?”, Studies in Anglo-
Saxon Archaeology and History 9 (1996) pp. 58–61
——, Kings, Courtiers and Imperium. The Barbarian West, 565–725 (London 1997)
P. Bartholomew, “Fifth-century facts”, Britannia 13 (1982) pp. 261–70
R. Bartlett, The Making of Europe. Conquest, Colonization and Cultural Change, 950–1350
(London 1994)
B. Basdevant-Gaudemet, “L’évêque, d’après la législation de quelques conciles méro-
vingiens”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed.
M. Rouche (Paris 1997) pp. 471–94
S. Bassett, “Lincoln and the Anglo-Saxon see of Lindsey”, Anglo-Saxon England 18
(1989) pp. 1–32
——, “In search of the origins of Anglo-Saxon kingdoms”, The Origins of Anglo-Saxon
Kingdoms, ed. id. (Leicester 1989) pp. 3–27
T. Bauer, “Graf/Grafio (Historisches)”, Reallexikon der Germanischen Altertumskunde 12
(2nd edn., 1999) pp. 532–55
Die Bayern und ihre Nachbarn, vol. 1, ed. H. Wolfram and A. Schwarcz, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische
Klasse 179. Veröffentlichungen der Kommission für Frühmittelalterforschung 8
(Wien 1989)
Bayerisch-schwäbische Landesgeschichte an der Universität Augsburg 1975–1977, ed. P. Fried,
Veröffentlichungen der Schwäbischen Forschungsgemeinschaft bei der Kommission
für Bayerische Landesgeschichte, Reihe 7, Augsburger Beiträge zur Landesgeschichte
Bayerisch-Schwabens 1 (Sigmaringen 1979)
M. Becher, Eid und Herrschaft. Untersuchungen zum Herrscherethos Karls des Großen, Vorträge
und Forschungen, Sonderband 39 (Sigmaringen 1993)
——, Rex, Dux und Gens. Untersuchungen zur Entstehung des sächsischen Herzogtums im 9.
und 10. Jahrhundert, Historische Studien 444 (Husum 1996)
——, “Die Sachsen im 7. und 8. Jahrhundert”, 799—Kunst und Kultur der Karolingerzeit.
Karl der Große und Papst Leo III. in Paderborn. Beiträge zum Katalog der Ausstellung
Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz 1999) pp. 183–94
——, “Non enim habent regem idem Antiqui Saxones . . . Verfassung und Ethnogenese in
Sachsen während des 8. Jahrhunderts”, Studien zur Sachsenforschung 12, ed. H.-J.
Häßler (Oldenburg 1999) pp. 1–31
H. Beck, S. Hamann and H. Roth, “Bajuwaren”, Reallexikon der Germanischen
Altertumskunde 1 (2nd edn., 1973) pp. 601–27
C. Behr, “The origins of kingship in early medieval Kent”, Early Medieval Europe 9
(2000) pp. 25–52
F. Beisel, Studien zu den fränkisch-römischen Beziehungen. Von ihren Anfängen bis zum
Ausgang des 6. Jahrhunderts, Wissenschaftliche Schriften im Wissenschaftlichen Verlag
Dr. Schulz-Kirchner, Reihe 9: Geschichtswissenschaftliche Beiträge 105 (Idstein
1987)
Beiträge zur mittelalterlichen Reichs- und Nationsbildung in Deutschland und Frankreich, ed.
C. Brühl and B. Schneidmüller, Historische Zeitschrift Beiheft Neue Folge 24
(München 1997)
A. Ben Abed and N. Duval, “Carthage et les villes de la Tunisie”, Sedes regiae (ann.
400–800), ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona 2000) pp.
163–218
K. Benda, “Souèasny stav studia zlatych nádob pokladu ze Sânnicol>ul Mare
(Nagyszentmiklós) [Gegenwärtiger Studienstand über die Goldgefässe aus Sânnicol>ul
640 

Mare (Nagyszentmiklós)]”, Slovenská Archeológia 13,2 (1965) pp. 399–414 [German

summary: pp. 412–4]
L. Bende, “A pitvarosi késõ avar temetõ 51—sírja (Adatok a késõ avar kori lós-
zerszámok díszítéséhez) [Das Grab 51 im spätawarenzeitlichen Gräberfeld von
Pitvaros (Beiträge zur Verzierung der spätawarenzeitlichen Pferdegeschirre)]”, A
Móra Ferenc Múzeum Évkönyve—Studia Archaeologica 4 (1998) pp. 195–230 [German
summary: pp. 210–2]
——, “Tausírozott díszû övgarnitúra a pitvarosi avar temetõbõl [ Tauschierte
Gürtelgarnitur im awarischen Gräberfeld von Pitvaros]”, A Móra Ferenc Múzeum
Évkönyve—Studia Archaeologica 6 (2000) pp. 199–217 [German summary: pp. 209–11]
H. Bengtson, “Neues zur Geschichte der Naristen”, Historia 8 (1959) pp. 213–21
H. Berg, “Bischöfe und Bischofssitze im Ostalpen- und Donauraum vom 4. bis zum
8. Jahrhundert”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 179. Veröffentlichungen der Kommission für Frühmittelalter-
forschung 8 (Wien 1989) pp. 61–108
——, “Christentum im bayerischen Raum um 700”, Der heilige Willibald—Klosterbischof
oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter, Eichstätter
Studien NF 30 (Regensburg 1990) pp. 69–113
——, “Zur Organisation der bayerischen Kirche und zu den bayerischen Synoden
des 8. Jahrhunderts”, Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern
1, ed. W. Pohl and H. Wolfram, Denkschriften der Österreichischen Akademie
der Wissenschaften, philosophisch-historische Klasse 201. Veröffentlichungen der
Kommission für Frühmittelalterforschung 12 (Wien 1990) pp. 181–97
F. Berger, Untersuchungen zu römerzeitlichen Münzfunden in Nordwestdeutschland, Studien
zu Fundmünzen der Antike 9 (Berlin 1992)
W. Bergmann, “Untersuchungen zu den Gerichtsurkunden der Merowingerzeit”,
Archiv für Diplomatik 22 (1976) pp. 1–186
E. Bernareggi, Moneta Langobardorum (Milano 1983) [English transl. Moneta Langobardorum
(Lugano 1989)]
H. Bernhard et al., Der Runde Berg bei Urach, Führer zu archäologischen Denkmälern
in Baden-Württemberg 14 (Stuttgart 1991)
C. Bernis Madrazo, Indumentaria medieval española (Madrid 1959)
C. Bertelli, “The Production and Distribution of Books in Late Antiquity”, The Sixth
Century. Production, Distribution and Demand, ed. R. Hodges and W. Bowden, The
Transformation of the Roman World 3 (Leiden 1998) pp. 41–60
E. Besta, “Le Fonti dell’ Editto di Rotari”, Atti del primo Congresso internazionale di
Studi Longobardi (Spoleto 1952) pp. 51–69
H. Beumann, “Unitas Ecclesiae—Unitas Imperii—Unitas Regni. Von der imperi-
alen Reichseinheitsidee zur Einheit der Regna”, Nascita dell’Europa ed Europa
Carolingia: Un’equazione da verificare, Settimane di studio del centro italiano di studi
sull’alto medioevo 27 (Spoleto 1981) pp. 531–71
F. Beyerle, “Über Normtypen und Erweiterungen der Lex Salica”, Zeitschrift der
Savigny-Stiftung für Rechtsgeschichte, Germanistische Abteilung 44 (1924) pp. 394–419
——, “Süddeutschland in der politischen Konzeption Theoderichs des Großen”,
Grundfragen der alemannischen Geschichte, Vorträge und Forschungen 1 (Sigmaringen
1955) pp. 65–81
J. Biel, “Die Ausstattung des Toten. Reichtum im Grabe—Spiegel der Macht”, Der
Keltenfürst von Hochdorf. Methoden und Ergebnisse der Landesarchäologie (Stuttgart 1985)
pp. 79–105
V. Bierbrauer, Die Ostgotischen Grab- und Schatzfunde in Italien, Biblioteca degli studi
medievali 7 (Spoleto 1975)
——, “Frühgeschichtliche Akkulturationsprozesse in den germanischen Staaten am
Mittelmeer (Westgoten, Ostgoten, Langobarden) aus der Sicht des Archäologen”,
Atti del VI° congresso internazionale di studi sull’alto medioevo (Spoleto 1980) pp. 89–105
  641

——, “Alamannische Besiedlung Augsburgs und seines näheren Umlandes”, Geschichte

der Stadt Augsburg von der Römerzeit bis zur Gegenwart, ed. G. Gottlieb, W. Baer,
J. Becker, J. Bellot, K. Filser, P. Fried, W. Reinhard and B. Schimmelpfennig
(Stuttgart 1984) pp. 87–100
——, “Das Reihengräberfeld von Altenerding in Oberbayern und die bajuwarische
Ethnogenese—Eine Problemskizze”, Zeitschrift für Archäologie des Mittelalters 13 (1985)
pp. 7–25
——, “Die germanische Aufsiedlung des östlichen und mittleren Alpengebietes im
6. und 7. Jahrhundert aus archäologischer Sicht”, Frühmittelalterliche Ethnogenese im
Alpenraum, ed. H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985) pp.
9–47
——, “Die Landnahme der Langobarden in Italien aus archäologischer Sicht”,
Ausgewählte Probleme europäischer Landnahmen des Früh- und Hochmittelalters 1, ed.
M. Müller-Wille and R. Schneider, Vorträge und Forschungen 41 (Sigmaringen
1993) pp. 103–72
——, “Romanen im fränkischen Siedelgebiet”, Die Franken—Wegbereiter Europas. Vor
1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim 1 (Mainz 1996) pp. 110–20
——, “Les Wisigoths dans le royaume franc”, Antiquités Nationales 29 (1997) pp. 167–200
K. Binding, Das burgundisch-romanische Königreich (Leipzig 1868)
H. Birnbaum, “Von ethnolinguistischer Einheit zur Vielfalt: die Slawen im Zuge
der Landnahme der Balkanhalbinsel”, Südostforschungen 51 (1992) pp. 1–19
J. Blair, Anglo-Saxon Oxfordshire (Oxford-Stroud 1994)
W. Bleiber, Das Frankenreich der Merowinger (Wien-Köln-Graz 1988)
A. Boddington, “Models of burial, settlement and worship: the final phase reviewed”,
Anglo-Saxon Cemeteries: A Reappraisal, ed. E. Southworth (Stroud 1990) pp. 177–99
G.P. Bognetti, L’età longobarda, 4 vols. (Milano 1966–68)
——, “S. Maria foris portas di Castelseprio e la storia religiosa dei Longobardi”,
id., L’età longobarda 2 (repr. Milano 1966) pp. 12–683
——, “L’influsso delle istituzioni militari romane sulle istituzioni longobarde del sec-
colo VI e la natura della fara”, id., L’età longobarda 3 (Milano 1967) pp. 1–46
H.W. Böhme, Germanische Grabfunde des 4. bis 5. Jahrhunderts. Zwischen unterer Elbe und
Loire. Studien zur Chronologie und Bevölkerungsgeschichte, Münchener Beiträge zur Vor-
und Frühgeschichte 19 (München 1974)
——, “Studien zu Gallien in der Spätantike”, Jahrbuch des Römisch-Germanischen Zentral-
museums Mainz 33 (1986) pp. 844–8
——, “Das Ende der Römerherrschaft in Britannien und die angelsächsische
Besiedlung Englands im 5. Jahrhundert”, Jahrbuch des Römisch-Germanischen Zentral-
museums Mainz 33 (1986) pp. 469–574
——, “Zur Bedeutung des spätrömischen Militärdienstes für die Stammesbildung
der Bajuwaren”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer
and H. Dopsch (Rosenheim-Salzburg 1988) pp. 23–37
——, “Les Thuringiens dans le Nord du royaume franc”, Revue Archéologique de Picardie
3–4 (1988) pp. 57–69
——, “Eine elbgermanische Bügelfibel des 5. Jahrhunderts aus Limetz-Villez (Yvelines,
Frankreich)”, Archäologisches Korrespondenzblatt 19 (1989) pp. 397–406
——, “Adelsgräber im Frankenreich. Archäologische Zeugnisse zur Herausbildung
einer Herrenschicht unter den merowingischen Königen”, Jahrbuch des Römisch-
Germanischen Zentralmuseums Mainz 40 (1993) pp. 397–534
——, “Der Frankenkönig Childerich zwischen Attila und Aëtius. Zu den Goldgriffs-
pathen der Merowingerzeit”, Festschrift für Otto-Herman Frey, ed. C. Dobiat, Marburger
Studien zur Vor- und Frühgeschichte 16 (Hitzeroth 1994) pp. 69–110
——, “Söldner und Siedler im spätantiken Nordgallien”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 91–101
642 

——, “Kontinuität und Traditionen bei Wanderungsbewegungen im früh-mittelal-

terlichen Europa vom 1.–6. Jahrhundert”, Archäologische Informationen 19 (1996) pp.
89–103
——, “Franken und Romanen im Spiegel spätrömischer Grabfunde im nördlichen
Gallien”, Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed.
D. Geuenich, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
19 (Berlin-New York 1998) pp. 31–58
——, “Ethnos und Religion der Bewohner Westfalens. Methodische und historische
Problematik”, 799—Kunst und Kultur der Karolingerzeit. Karl der Große und Papst Leo
III. in Paderborn. Beiträge zum Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann
and M. Wemhoff (Mainz 1999) pp. 237–45
——, “Franken oder Sachsen? Beiträge zur Siedlungs- und Bevölkerungsgeschichte
in Westfalen vom 4.-7. Jahrhundert”, Studien zur Sachsenforschung 12, ed. H.-J. Häß-
ler (Oldenburg 1999) pp. 43–73
——, “Sächsische Söldner im römischen Heer. Das Land zwischen Ems und
Niederelbe während des 4. und 5. Jahrhunderts”, Über allen Fronten. Nordwestdeutschland
zwischen Augustus und Karl dem Großen, ed. M. Fansa (Oldenburg 1999) pp. 49–73
K. Böhner, “Zur historischen Interpretation der sogenannten Laetengräber”, Jahrbuch
des Römisch-Germanischen Zentralmuseums Mainz 10 (1963) pp. 139–167
——, “Romanen und Franken im Trierer Land nach dem Zeugnis der archäolo-
gischen Quellen”, Siedlung, Sprache und Bevölkerungsstruktur im Frankenreich, ed. F. Petri,
Wege der Forschungen 49 (Darmstadt 1973) pp. 346–82
I. Bóna, “Beiträge zu den ethnischen Verhältnissen des 6.–7. Jahrhunderts in
Westungarn I—Norditalischer Bronze-Gürtelschmuck in frühawarischen Gräberfel-
dern”, Alba Regia 2/3 (1963) pp. 49–63
——, VII. sz-i avar települések és Arpád-kori magyar falu Dunaújvarosban [Awarische Siedlungen
aus dem VII. Jh. und arpadenzeitliches Dorf in Dunaújvaros], Fontes Archaeologici
Hungariae (Budapest 1973)
——, “A Szegvár-sápoldali lovassír. Adatok a korai avar temetkezési szokhásokhoz
[Das Reitergrab von Szegvár-Sápoldal. Beiträge zu den frühawarischen Bestattungs-
sitten]”, Archeologiai Értesítõ 106,1 (1979) pp. 3–32
——, “Studien zum frühawarischen Reitergrab von Szegvár”, Acta Archaeologica
Hungarica Academiae Scientiarum Hungaricae 32 (1980) pp. 31–95
E. Boshof, “Agilolfingisches Herzogtum und angelsächsische Mission: Bonifatius und
die bayerische Bistumsorganisation von 739”, Ostbairische Grenzmarken 31 (1989)
pp. 11–26
K. Bosl, “Das ‘jüngere’ bayerische Stammesherzogtum der Luitpoldinger”, Zeitschrift
für bayerische Landesgeschichte 18 (1955) pp. 144–72 [repr. Zur Geschichte der Bayern,
ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 329–63]
——, “Das bayerische Stammesherzogtum”, Zeitschrift für bayerische Landesgeschichte 25
(1962) pp. 275–82 [repr. Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung
60 (Darmstadt 1965) pp. 1–11]
——, “Der ‘Adelsheilige’, Idealtypus und Wirklichkeit, Gesellschaft und Kultur im
merowingerzeitlichen Bayern des 7. und 8. Jahrhunderts”, Speculum Historiale.
Geschichte im Spiegel von Geschichtsschreibung und Geschichtsdeutung, ed. C. Bauer, L.
Boehm and M. Müller (Freiburg-München 1965) pp. 167–87
——, “Regensburgs politische Stellung im frühen Mittelalter”, Zeitschrift für bayerische
Landesgeschichte 34 (1971) pp. 3–14
——, Bayerische Geschichte (7th edn., München 1990)
J.-L. Boudartchouk, “La nécropole francque de Ictium à L’Isle-Jourdain (Gers, Midi-
Pyrénées, France)”, Archäologie des Frankenreiches—Neueste Forschungen. Internationales
Kolloquium anläßlich der Ausstellung “Die Franken-Les Francs” Berlin, 10. bis 12. Oktober
1997, ed. W. Menghin, Acta Praehistorica et Archaeologica 30 (1998) pp. 126–36
D. Boyson, “Romano-Burgundian society in the Age of Gundobad: some legal,
  643

archaeological and historical evidence”, Nottingham Medieval Studies 32 (1988) pp.

91–118
S. Brather “Ethnische Identitäten als Konstrukte der frühgeschichtlichen Archäologie”,
Germania 78 (2000) pp. 139–77
M. Brett, “The Islamization of Morocco from the Arabs to the Almoravids”, Journal
of Moroccan Studies 2 (1992) pp. 57–71 [repr. id., Ibn Khaldun and the Medieval Maghreb
(Aldershot 1999)]
M. Brett and E. Fentress, The Berbers (Oxford 1996; repr. 1998)
V. Brieske, Schmuck und Trachtbestandteile des Gräberfeldes von Liebenau, Kr. Nienburg/Weser,
Studien zur Sachsenforschung 5,6 (Oldenburg 2001)
H. Brink-Kloke, A. von Bohlen, M. Doll, E. Lietz and C. Poniecki, “Ein (kleines)
germanisches Dorf—Die jüngerkaiserzeitliche Siedlung von Dortmund-Oespel”,
Millionen Jahre Geschichte. Fundort Nordrhein-Westfalen, ed. H.G. Horn, H. Hellenkemper,
G. Isenberg and H. Koschik (Mainz 2000) pp. 343–8
Britain 400 –600: Language and History, ed. A. Bammesberger and A. Wollmann
(Heidelberg 1990)
G.P. Brogiolo, “Capitali e residenze regie nell’Italia longobarda”, Sedes regiae (ann.
400–800), ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona 2000) pp.
135–62
E.R. Brooks, “The Emperor Zenon and the Isaurians”, English Historical Review 8
(1893) pp. 209–38
N. Brooks, The Early History of the Church of Canterbury (Leicester 1984)
——, “The creation and structure of the kingdom of Kent”, The Origins of Anglo-
Saxon Kingdoms, ed. S. Bassett (Leicester 1989) pp. 55–74
——, Bede and the English ( Jarrow 1999)
P. Brown, The World of Late Antiquity (London 1971)
R. Browning, Byzantium and Bulgaria: a comparative study across the early medieval frontier
(London 1975)
R. Bruce-Mitford, The Sutton Hoo Ship-Burial II: Arms, Armour and Regalia (London
1978)
B. Brugmann, “Britons, Angles, Saxons, Jutes and Franks”, The Anglo-Saxon Cemetery
on Mill Hill, Deal, Kent, ed. K. Parfitt and B. Brugmann, Society for Medieval
Archaeology, Monograph series 14 (London 1997) pp. 110–24
C. Brühl, Deutschland—Frankreich. Die Geburt zweier Völker (Köln-Wien 1990; repr. 1995)
R. Brulet, “Tournai und der Bestattungsplatz um Saint-Brice”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 163–70
——, “La tombe de Childéric et la topographie funéraire de Tournai à la fin du
Ve siècle”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed.
M. Rouche (Paris 1997) pp. 59–78
H. Brunner, “Über ein verschollenes merowingisches Königsgesetz des 7. Jahr-
hunderts”, Sitzungsberichte der königlichen preußischen Akademie der Wissenschaften 309
(1901) pp. 932–55
——, Deutsche Rechtsgeschichte, 2 vols. (Leipzig 1906, 2nd edn. 1928)
L. Buchet, “Die Landnahme der Franken in Gallien aus der Sicht der Anthropologen”,
Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben.
Katalog zur Ausstellung im Reiss-Museum Mannheim 2 (Mainz 1996) pp. 662–7
R. Buchner, Die Rechtsquellen, Wattenbach-Levison, Deutschlands Geschichtsquellen
im Mittelalter, Vorzeit und Karolinger, Beiheft (Weimar 1953)
R.W. Burgess, “The Dark Ages Return to Fifth-Century Britain: the ‘restored’ Gallic
Chronicle exploded”, Britannia 21 (1990) pp. 185–95
T.S. Burns, A History of the Ostrogoths (Bloomington 1984)
——, Barbarians within the Gates of Rome (Bloomington 1994)
A. Burzler, Archäologische Beiträge zum Nobilifizierungsprozeß in der jüngeren Merowingerzeit
(unpublished doct. diss. München 1991)
644 

The Byzantine and Early Islamic Near East, vol. 3: States, Resources and Armies, ed.
A. Cameron (Princeton NJ 1995)
A. Cameron, “Byzantine Africa—the literary evidence”, Excavations at Carthage 1978
conducted by the University of Michigan 7, ed. J.H. Humphrey (Ann Arbor 1982) pp.
29–62
A. Cameron, “How did the Merovingian kings wear their heir?”, Revue belge de philo-
logie et d’histoire 48 (1965) pp. 1203–14
E. Campbell, “Trade in the Dark Age West: a Peripheral Activity”, Scotland in Dark
Age Britain, ed. B.E. Crawford (Aberdeen 1996) pp. 79–93
J. Campbell, Bede’s Reges et Principes ( Jarrow 1979)
J. Campos, Juan de Biclaro. Obispo de Gerona, su vida y su obra (Madrid 1960)
C. Cardelle de Hartmann, “The textual transmission of the Mozarabic chronicle
of 754”, Early Medieval Europe 8 (1999) pp. 13–29
C. von Carnap-Bornheim and J. Ilkjær, Illerup Ådal 5: Die Prachtausrüstungen, Jutland
Archaeological Society Publications 25,5 (Aarhus 1996)
——, “Kaiserzeitliche germanische Traditionen im Fundgut des Grabes des “chef
militaire” in Vermand und im Childerich-Grab in Tournai”, Germanen beiderseits
des spätantiken Limes, ed. Th. Fischer, G. Precht and J. Tejral, Spisy Archeologického
Ústavu Av ÈR Brno 14 (Köln-Brno 1999) pp. 47–61
M. Carver, Sutton Hoo: Burial Ground of Kings (London 1998)
H. Castritius, “Die Grenzverteidigung in Rätien und Noricum im 5. Jh. n. Chr.
Ein Beitrag zum Ende der Antike”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram
and A. Schwarcz, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 179. Veröffentlichungen der Kommission für
Frühmittelalterforschung 8 (Wien 1989) pp. 17–28
The Celtic World, ed. M. Green (London 1995)
M. Cesa, Impero tardoantico e barbari (Como 1994)
H.M. Chadwick, The Origin of the English Nation (Cambridge 1907)
R.A. Chambers, “The Late- and Sub-Roman Cemetery at Queensford Farm,
Dorchester-on-Thames, Oxon.”, Oxoniensia 52 (1987) pp. 35–69
T.M. Charles-Edwards, Early Irish and Welsh Kinship (Oxford 1993)
——, “Language and society among the insular Celts, A.D. 400–1000”, The Celtic
World, ed. M. Green (London 1995) pp. 703–36
A. Chastagnol, “Les gouverneurs de la Byzance et de Tripolitaine”, Antiquités
africaines 1 (1967) pp. 119–34
——, “Les espagnols dans l’aristocratie gouvernamentale à l’époque de Théodose”,
id., Aspects de l’Antiquitè tardive (Rome 1994) pp. 11–42
——, Aspects de l’Antiquitè tardive (Rome 1994)
A. Chastagnol and N. Duval, “Les survivances du culte impériale dans l’Afrique
du Nord à l’époque Vandale”, Mélanges d’histoire anciennes offerts à William Seston
(Paris 1974) pp. 87–118
A. Chauvot, “Images positives, images négatives des Barbares dans les sources latines
à la fin du Ve siècle et au début du VIe siècle après J.-C.”, Clovis. Histoire et
mémoire, vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp.
3–14
Das Christentum im bairischen Raum von den Anfängen bis ins 11. Jahrhundert, ed. E. Boshof
and H. Wolff, Passauer historische Forschungen 8 (Köln-Wien 1994)
Christianity in Roman Britain, ed. M.W. Barley and R.P.C. Hanson (Leicester 1968)
Christians and Moors in Spain 3, ed. C.P. Melville and A. Ubaydli (Warminster 1992)
N. Christie, The Lombards, The Peoples of Europe (Oxford UK-Cambridge Mass.
1995)
R. Christlein, “Besitzabstufungen zur Merowingerzeit im Spiegel reicher Grabfunde
aus West- und Süddeutschland”, Jahrbuch des Römisch-Germanischen Zentralmuseums
Mainz 20 (1973) pp. 147–80
  645

A. Christys, “The History of 'Abd al-Malik ibn Óabìb and ethnogenesis in al-
Andalus”, Power and the Construction of Communities, ed. W. Pohl and H. Reimitz
(forthcoming)
——, Christians in al-Andalus 711—c. 1000 (Richmond, Surrey 2002)
E. Chrysos, “Legal Concepts and patterns for the Barbarians’ Settlement on Roman
Soil”, Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989)
pp. 13–23
——, “Perceptions of the International Community of States during the Middle
Ages”, Ethnogenese und Überlieferung. Angewandte Methoden der Frühmittelalterforschung, ed.
K. Brunner and B. Merta, Veröffentlichungen des Instituts für Österreichische
Geschichtsforschung 31 (Wien-München 1994) pp. 293–307
The Churches, Ireland and the Irish, ed. W.J. Sheils and D. Wood, Studies in Church
History 25 (Oxford 1989)
Le cimitière et la fortification du Bas-Empire de Vireux-Molhain, Dép. Ardennes, ed. J.-P.
Lemant (Mainz 1985)
A. Clark, “Varieties of uniformity: the first century of the Church of Ireland”, The
Churches, Ireland and the Irish, ed. W.J. Sheils and D. Wood, Studies in Church
History 25 (Oxford 1989) pp. 105–22
P. Classen, “Bayern und die politischen Mächte im Zeitalter Karls des Großen”,
Die Anfänge des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberöster-
reichischen Landesarchivs, Ergänzungsband 2 (Linz 1978) pp. 169–87
D. Claude, “Die Bestellung der Bischöfe im merowingischen Reiche”, Zeitschrift der
Savigny-Stiftung für Rechtsgeschichte, Kanonistische Abteilung 49 (1963) pp. 1–75
——, Geschichte der Westgoten (Stuttgart-Berlin-Köln-Mainz 1970)
——, Adel, Kirche und Königtum im Westgotereich (Sigmaringen 1971)
——, “Probleme der vandalischen Herrschaftsnachfolge”, Deutsches Archiv 30 (1974)
pp. 329–55
——, “Zur Königserhebung Theoderichs des Großen”, Geschichtsschreibung und geistiges
Leben im Mittelalter. Festschrift für Heinz Löwe zum 65. Geburtstag, ed. K. Hauck and
H. Mordek (Cologne 1978) pp. 1–13
——, “Remarks about relations between Visigoths and Hispano-Romans in the sev-
enth century”, Strategies of Distinction. The Construction of Ethnic Communities, 300–800,
ed. W. Pohl with H. Reimitz, The Transformation of the Roman World 2 (Leiden-
New York-Köln 1998) pp. 117–30
F.M. Clover, “Carthage and the Vandals”, Excavations at Carthage 1978 conducted by
the University of Michigan 7, ed. J.H. Humphrey (Ann Arbor 1982) pp. 1–22
——, “Emperor worship in Vandal Africa”, Romanitas-Christianitas. Untersuchungen zur
Geschichte und Literatur der römischen Kaiserzeit, ed. G. Wirth (Berlin 1982) pp. 663–74
——, “Felix Carthago”, Dumbarton Oaks Papers 40 (1986) pp. 1–6
——, “Le culte des empereurs dans l’Afrique vandale”, id., The Late Roman West
and the Vandals (Aldershot 1993) ch. 8, pp. 121–8
——, The Late Roman West and the Vandals (Aldershot 1993)
——, “Time-keeping, monarchy and the heartbeat of Vandal and proto-Byzantine
Africa” (forthcoming)
Clovis. Histoire etvmémoire, vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997)
R. Coates, “On some controversy surrounding Gewissae/Gewissei, Cerdic and Ceawlin”,
Nomina 13 (1989/90) pp. 1–11
C. Codoñer, “Literatura hispano-latina tardía”, Unidad y pluralidad en el mundo antiguo.
Actas VI Congreso Español de Estudios Clásicos 1 (Madrid 1983) pp. 435–65
——, “La literatura”, Historia de España de Ramón Menéndez Pidal, Part 3,2: España
visigoda (Madrid 1991) pp. 209–67
R. Collins, Early Medieval Spain. Unity in Diversity, 400–1000, New Studies in Medieval
History (London 1983)
——, “Conclusion”, The Settlement of Disputes in early medieval Europe, ed. W. Davies
and P. Fouracre (Cambridge 1986) p. 210
646 

——, The Arab Conquest of Spain, 710–797 (Oxford 1989)

——, “Isidore, Maximus and the Historia Gothorum”, Historiographie im frühen Mittelalter,
ed. A. Scharer and G. Scheibelreiter (Vienna-Munich 1994) pp. 345–58
The Community, the Family and the Saint. Patterns of Power in Early Medieval Europe. Selected
Proceedings of the International Medieval Congress. University of Leeds, 4 –7 July 1994,
10–13 July 1995, ed. J. Hill and M. Swan, International Medieval Research 4
(Turnhout 1998)
Concepts of National Identity in the Middle Ages, ed. S. Forde, L. Johnson and A.V.
Murray, Leeds Texts and Monographs. New Series 14 (Leeds 1995)
The Construction of Communities in the Early Middle Ages, ed. R. Corradini, M. Diesenberger
and H. Reimitz (forthcoming)
La conversione al cristianesimo nell’Europa dell’alto medioevo, Settimane di studio del cen-
tro italiano di studi sull’alto medioevo 14 (1967)
C. Courtois, Victor de Vita et son oeuvre (Algiers 1954)
——, Les Vandales et l’Afrique (Paris 1955; repr. Aalen 1964)
L. Cracco Ruggini, “Les généraux francs aux IVe–Ve siècles et leurs groupes aris-
tocratiques”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed.
M. Rouche (Paris 1997) pp. 673–88
D. Csallány, A kunszentmártoni avarkori ötvössír [Goldschmiedegrab aus der Awarenzeit von
Kunszentmárton] (Szentes 1933)
——, “A hajdúdorogi avar mellpáncél [Der avarische Brustpanzer von Hajdúdorog]”,
A Debreceni Déri Múzeum Évkönyve 45 (1958/59) pp. 17–23 [German summary:
p. 23]
C.E. Cubitt, Anglo-Saxon Church Councils c. 650–c. 850, Studies in the Early History
of Britain (Leicester 1995) pp. 153–90
F. Curta, The Making of the Slavs. History and Archaeology of the Lower Danube Region
c. 500–700 (Cambridge 2001)
——, “From Kossinna to Bromley: ethnogenesis in Slavic archaeology”, On Barbarian
Identity—Critical Approaches to Ethnogenesis Theory, ed. A. Gillett (Turnhout forth-
coming)
H. Dachs, “Germanischer Uradel im frühbairischen Donaugau”, Verhandlungen des
Historischen Vereins für Oberpfalz und Regensburg 86 (1936) pp. 179–92 [repr. Zur
Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp.
85–106]
——, “Römerkastelle und frühmittelalterliches Herzogs- und Königsgut an der
Donau”, Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner, Schriftenreihe
zur bayerischen Landesgeschichte 62 (München 1962) pp. 293–320 [repr. Zur
Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp.
44–84]
F. Daim, “Der awarische Greif und die byzantinische Antike”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. id. and H. Friesinger, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 273–303
——, “Zur nachgedrehten Keramik aus dem awarischen Gräberfeld von Mödling—
An der goldenen Stiege”, Slawische Keramik in Mitteleuropa vom 8. bis 11. Jahrhundert,
ed. È. Staòa, Internationale Tagungen in Mikulcice 1 (Brno 1994) pp. 29–52
——, “Das awarische Gräberfeld von Zillingtal: Sechs Gräber mit westlichen Gegen-
ständen”, Wissenschaftliche Arbeiten aus dem Burgenland 100 (1998) pp. 97–135
——, “Archaeology, ethnicity and the structures of identification. The example of
the Avars, Carantanians and Moravians in the eighth century”, Strategies of Distinction.
The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The
Transformation of the Roman World 2 (Leiden-New York-Köln 1998) pp. 71–94
——, “Byzantine belts and Avar birds. Diplomacy, trade and cultural transfer in
  647

the eighth century”, The Transformation of Frontiers—from Late Antiquity to the Carolingians,
ed. W. Pohl, I.N. Wood and H. Reimitz, The Transformation of the Roman
World 10 (Leiden-Boston 2000) pp. 143–88
——, “‘Byzantinische’ Gürtelgarnituren des 8. Jahrhunderts”, Die Awaren am Rand
der byzantinischen Welt. Studien zu Diplomatie, Handel und Technologietransfer im Frühmittelalter,
ed. id., Monographien zur Frühgeschichte und Mittelalterarchäologie 7 (Innsbruck
2000) pp. 77–204
——, “Komárno/Komárom”, Reallexikon der Germanischen Altertumskunde 17 (2nd edn.,
2001) pp. 177–9
——, “Kunbábony”, Reallexikon der Germanischen Altertumskunde 17 (2nd edn., 2001)
pp. 490–5
F. Daim et al., Das awarische Gräberfeld von Leobersdorf, Niederösterreich, 2 vols., Studien
zur Archäologie der Awaren 3. Denkschriften der Österreichischen Akademie der
Wissenschaften, philosophisch-historische Klasse 194. Veröffentlichungen der
Kommission für Frühmittelalterforschung 10 (Wien 1987)
F. Daim and A. Distelberger, “Die awarische Siedlung von Zillingtal—Die Grabungen
1994–95”, Reitervölker aus dem Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996)
pp. 372–7
F. Daim and A. Lippert, Das awarische Gräberfeld von Sommerein am Leithagebirge, NÖ,
Studien zur Archäologie der Awaren 1. Studien zur Ur- und Frühgeschichte des
Donau- und Ostalpenraumes 2. Denkschriften der Österreichischen Akademie der
Wissenschaften, philosophisch-historische Klasse 170 (Wien 1984)
F. Daim and P. Stadler, “Der Goldschatz von Sînnicolaul Mare (Nagyszentmiklós)”,
Reitervölker aus dem Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp.
439–45
Dal Concilio di Calcedonia (451) a Beda. I Padri Latini, ed. A. Di Berardino, Patrologia
4 (Genova 1996)
K.R. Dark, “A sub-Roman defence of Hadrian’s Wall?”, Britannia 22 (1992) pp.
111–20
——, Civitas to Kingdom (Leicester 1994)
——, Britain and the End of the Roman Empire (Stroud 2000)
A. Dasnoy, “La nécropole de Furfooz”, Annales de la Société Archéologique de Namur 55
(1969) pp. 121–94
J.R. Davies, “Liber Landavensis: Its Date and the Identity of its Editor”, Cambrian
Medieval Celtic Studies 35 (1998) pp. 1–12
W. Davies, “Land and Power in Early Medieval Wales”, Past and Present 81 (1978)
pp. 3–23
——, The Llandaff Charters (Aberystwyth 1979)
——, Wales in the Early Middle Ages (Leicester 1982)
——, Patterns of Power in Early Wales (Oxford 1990)
J. Dekan, “Herkunft und Ethnizität der gegossenen Bronzeindustrie des VIII.
Jahrhunderts”, Slovenská Archeológia 20,2 (1972) pp. 317–452
P. Delogu, “Il regno longobardo”, Storia d’Italia, vol. 1: Longobardi e Bizantini, ed. id.,
A. Guillou and G. Ortalli (Torino 1980) pp. 17–24
——, “Longobardi e Romani: Altre congetture”, Langobardia, ed. S. Gasparri and
P. Cammarosano (Udine 1990) pp. 111–68
A. Demandt, Der Fall Roms. Die Auflösung des Römischen Reiches im Urteil der Nachwelt
(München 1984)
——, “The Osmosis of Late Roman and Germanic Aristocracies”, Das Reich und
die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989) pp. 75–86
——, Die Spätantike. Die Römische Geschichte von Diocletian bis Justinian, 284–565 n. Chr.,
Handbuch der Altertumswissenschaft III,6 (München 1989)
E. Demougeot, La formation de l’Europe et les invasions barbares, vol. 1: Des origines
germaniques à l’avènement de Dioclétien (Paris 1969); vol. 2: De l’avènement de Dioclétien
648 

(284) à l’occupation germanique de l’Empire romain d’Occident (début du VI e siècle), Collection

historique (Paris 1979)
Deutsche Königspfalzen. Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 4:
Pfalzen—Reichsgut—Königshöfe, ed. L. Fenske, Veröffentlichungen des Max-Planck-
Instituts für Geschichte 11,4 (Göttingen 1996)
M.C. Díaz y Díaz, “La transmisión textual del Biclarense”, Analecta Sacra Tarraconensia
35 (1962) pp. 57–78
——, De Isidore al siglo XI. Ocho estudios sobre la vida literaria peninsular (Barcelona 1976)
——, “Titulaciones regias en la monarquía visigoda”, Revista Portuguesa de História
16 (1978) pp. 133–41
——, “Más sobre epítetos regios en la Hispania visigótica”, Studi medievali 19 (1978)
pp. 318–33
——, “Los discursos del rey Recaredo: El Tomus”, Actas del XIV Centenario del Concilio
III de Toledo, 589–1989 (Toledo 1991) pp. 223–36
——, “Scrittori della Penisola Iberica”, Dal Concilio di Calcedonia (451) a Beda. I Padri
Latini, ed. A. Di Berardino, Patrologia 4 (Genova 1996) pp. 61–118
P. Díaz, “Visigothic Political Institutions”, The Visigoths. From the Migration Period to
the Seventh Century. An Ethnographic Perspective, ed. P.J. Heather (San Marino 1999)
pp. 321–55
H. Dickerhof and S. Weinfurter, “Summa historica”, Der heilige Willibald—Klosterbischof
oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter, Eichstätter
Studien NF 30 (Regensburg 1990) pp. 245–61
T. Dickinson, “Material culture as social expression: the case of Saxon saucer
brooches with running spiral decoration”, Studien zur Sachsenforschung 7, ed. H.-J.
Häßler (Hildesheim 1991) pp. 39–70
——, “Early Saxon saucer brooches: a preliminary overview”, Anglo-Saxon Studies in
Archaeology and History 6 (1993) pp. 11–44
G. Diepolder, “Altbayerische Laurentiuspatrozinien”, Aus Bayerns Frühzeit. F. Wagner
zum 75. Geburtstag, ed. J. Werner, Schriftenreihe zur bayerischen Landesgeschichte
62 (München 1962) pp. 371–96
A. Dierkens, “Die Taufe Chlodwigs”, Die Franken—Wegbereiter Europas. Vor 1500 Jahren:
König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim 1
(Mainz 1996) pp. 183–91
A. Dierkens and P. Périn, “Death and burial in Gaul and Germania, 4th–8th cen-
tury”, The Transformation of the Roman World AD 400–900, ed. L. Webster and
M. Brown (London 1997) pp. 79–95
A. Dierkens and P. Périn, “Les sedes regiae mérovingiennes entre Seine et Rhin”,
Sedes regiae (ann. 400 –800), ed. G. Ripoll and J.M. Gurt with A. Chavarría
(Barcelona 2000) pp. 267–304
H.-J. Diesner, Das Vandalenreich (Stuttgart 1966)
——, “Zum vandalischen Post-und Verkehrswesen”, Philologus 112 (1968) pp. 283–7
K. Dietz, U. Osterhaus, S. Rieckhoff-Pauli and K. Spindler, Regensburg zur Römerzeit
(Regensburg 1979)
A. Distelberger, Das awarische Gräberfeld von Mistelbach, Niederösterreich, Studien zur
Archäologie der Awaren 6 (Innsbruck 1996)
G. Doerfer, “Zur Sprache der Hunnen”, Central Asiatic Journal 17 (1973) pp. 1–50
F. Dölger, “Die ‘Familie der Könige’”, id., Byzanz und die europäische Staatenwelt
(Darmstadt 1964) pp. 34–69
——, Byzanz und die europäische Staatenwelt (Darmstadt 1964)
P. Dolukhanov, The Early Slavs. Eastern Europe from the Initial Settlement to the Kievan
Rus (London-New York 1996)
U. Domínguez del Val, Historia de la antigua literatura latina hispano-cristiana, 4 vols.
(Madrid 1998)
F.M. Donner, “Centralised authority and military autonomy in the early Islamic
  649

conquests”, The Byzantine and Early Islamic Near East, vol. 3: States, Resources and
Armies, ed. A. Cameron (Princeton NJ 1995) pp. 337–60
O. Doppelfeld, “Das fränkische Knabengrab unter dem Chor des Kölner Domes”,
Germania 42 (1964) pp. 156–88
H. Dopsch, “Zum Anteil der Romanen und ihrer Kultur an der Stammesbildung
der Bajuwaren”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer
and H. Dopsch (Rosenheim-Salzburg 1988) pp. 47–54
H. Drescher and K. Hauck, “Götterthrone des heidnischen Nordens”, Frühmittelalterliche
Studien 16 (1982) pp. 237–301
K.F. Drew, The Burgundian Code (Philadelphia 1949)
J.F. Drinkwater, “The Bacaudae of Fifth-Century Gaul”, Fifth-century Gaul: a crisis of
identity, ed. id. and H. Elton (Cambridge 1992) pp. 208–17
——, “The usurpers Constantine III (407–11) and Jovinus (411–13)”, Britannia 29
(1998) pp. 269–98
D.N. Dumville, “Kingship, genealogies and regnal lists”, Early Medieval Kingship, ed.
P.H. Sawyer and I.N. Wood (Leeds 1977) pp. 72–104
——, “Sub-Roman Britain: History and Legend”, History 62 (1977) pp. 173–92
——, “Gildas and Maelgwn: problems of dating”, Gildas: New Approaches, ed. id. and
M. Lapidge (Woodbridge 1984) pp. 51–60
——, “The West Saxon Genealogical Regnal List and the chronology of early
Wessex”, Peritia 4 (1985) pp. 21–66
——, “Early Welsh Poetry: Problems of Historicity”, Early Welsh Poetry: Studies in the
Book of Aneirin, ed. B.F. Roberts (Aberystwyth 1988) pp. 1–16
——, “The idea of government in sub-Roman Britain”, After Empire. Towards an
Ethnology of Europe’s Barbarians, ed. G. Ausenda, Studies in Historical Archaeoethno-
logy (Woodbridge 1995) pp. 177–216
J. Durliat, “Cité, impôt et intégration des barbares”, Kingdoms of the Empire. The
Integration of Barbarians in Late Antiquity, ed. W. Pohl, The Transformation of the
Roman World 1 (Leiden-New York-Köln 1997) pp. 153–79
B. Eagles, “The archaeological evidence for settlement in the fifth to seventh cen-
turies A.D.”, The Medieval Landscape of Wessex, ed. M. Aston and C. Lewis (Oxford
1994) pp. 13–32
T. Earle, “Chiefdoms in archaeological and historical perspective”, Annual Review of
Anthropology 16 (1987) pp. 279–308
Early Medieval Kingship, ed. P.H. Sawyer and I.N. Wood (Leeds 1977)
Early Welsh Poetry: Studies in the Book of Aneirin, ed. B.F. Roberts (Aberystwyth 1988)
W. Ebel-Zepezauer, “Ostgermanische Blechfibeln des 5. und 6. Jahrhunderts zwi-
schen Rhein und Garonne”, Archäologisches Korrespondenzblatt 29 (1998) pp. 297–
303
H. Ebling, Prosopographie der Amtsträger des Merowingerreiches. Von Chlothar II. (613) bis
Karl Martell (714), Beihefte der Francia 2 (München 1974)
H. Ebling, J. Jarnut and G. Kampers, “Nomen et gens. Untersuchungen zu den
Führungsschichten des Franken-, Langobarden- und Westgotenreiches im 6. und
7. Jahrhundert”, Francia 8 (1980) pp. 687–745
R. Eckes and H. Zeiß, “Bairische Reihengräber des 6. Jahrhunderts bei Irlmauth,
BA. Regensburg”, Bayerische Vorgeschichtsblätter 15 (1938) pp. 44–56
K.A. Eckhardt, Merowingerblut II. Agilolfinger und Etichonen, Deutschrechtliches Archiv
11 (Witzenhausen 1965)
L. Eckhart, “Das Nach- und Weiterleben der Römerzeit in Oberösterreich”, Baiernzeit
in Oberösterreich. Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum Ende
des 8. Jahrhunderts, Kataloge des Oberösterreichischen Landesmuseums 97 (Linz
1977) pp. 27–38
Edward the Elder 899–924, ed. N. Higham and D. Hill (Manchester 2001)
B. Effros, “Dressing conservatively: a critique of recent archaeological discussions
650 

of women’s brooches as markers of ethnic identity”, Gender and the Transformation

of the Roman World, ed. L. Brubaker and J. Smith (Cambridge forthcoming)
C. Eger, “Vandalische Grabfunde aus Karthago”, Germania 79,2 (2001) pp. 347–90
H.J. Eggers, Einführung in die Vorgeschichte (München 1959)
S.R. Elze, “Per la storia della corona del tesoro di Monza”, Atti del VI° congresso
internazionale di studi sull’alto medioevo (Spoleto 1980) pp. 393–400
The End of Roman Britain, ed. P.J. Casey, British Archaeological Reports, British series
71 (Oxford 1979)
L. Ennabli, Carthage, une métropole chrétienne du IV e à la fin du VIIe siècle (Paris 1997)
W. Ensslin, Theoderich der Große (Munich 1947)
M. Erdrich, “Terra Nigra-Fußschalen wie Chenet 342 oder Gellep 273: eine salisch-
fränkische Keramikgattung”, Germania 76 (1998) pp. 875–83
——, “Wirtschaftsbeziehungen zwischen der Germania inferior und dem german-
ischen Vorland—ein Wunschbild”, Germania inferior. Besiedlung, Gesellschaft und
Wirtschaft an der Grenze der römisch-germanischen Welt, ed. T. Grünewald with H.-J.
Schalles, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
28 (Berlin-New York 2001) pp. 306–35
F.R. Erkens, “Die Ursprünge der Lorcher Tradition im Lichte archäologischer, his-
toriographischer und urkundlicher Zeugnisse”, Das Christentum im bairischen Raum
von den Anfängen bis ins 11. Jahrhundert, ed. E. Boshof and H. Wolff, Passauer his-
torische Forschungen 8 (Köln-Wien 1994) pp. 423–59
S. Esders, Römische Rechtstradition und merowingisches Königtum. Zum Rechtscharakter poli-
tischer Herrschaft in Burgund im 6. und 7. Jahrhundert, Veröffentlichungen des Max-
Planck-Instituts für Geschichte 134 (Göttingen 1997)
A.S. Esmonde Cleary, The Ending of Roman Britain (London 1989)
Estructuras y formas del poder en la historia, ed. R. Pastor, I. Kieniewicz, E. García de
Enterría et al. (Salamanca 1991)
Et multum et multa. Beiträge zur Literatur, Geschichte und Kultur der Jagd. Festgabe für Kurt
Lindner zum 27. November 1971, ed. S. Schwenk, G. Tilander and C.A. Willemsen
(Berlin-New York 1971)
P. Ethelberg, “The Chieftain’s Farms of the Over Jerstal Group”, Journal of Danish
Archaeology 11 (1992/93) pp. 111–35
Ethnic Groups and Boundaries: The Social Organization of Cultural Difference, ed. F. Barth
(Boston 1969)
Ethnische und kulturelle Verhältnisse an der mittleren Donau vom 6. bis zum 11. Jahrhundert.
Symposium Nitra 6. bis 10. November 1994, ed. D. Bialeková and J. Zábojník (Bratislava
1996)
Ethnogenese europäischer Völker. Aus der Sicht der Anthropologie und Vor- und Frühgeschichte,
ed. W. Bernhard and A. Kandler-Pálsson (Stuttgart-New York 1986)
Ethnogenese und Überlieferung. Angewandte Methoden der Frühmittelalterforschung, ed. K. Brunner
and B. Merta, Veröffentlichungen des Instituts für Österreichische Geschichts-
forschung 31 (Wien-München 1994)
Etudes d’orientalisme dédiées à la mémoire de Lévi-Provençal, 2 vols. (Paris 1962)
Europe Between Late Antiquity and the Middle Ages, ed. J. Bintliff and H. Hamerow,
British Archaeological Reports, International series 617 (Oxford 1995)
D.E. Evans, “Language Contact in pre-Roman and Roman Britain”, Aufstieg und
Niedergang des Römischen Welt II, 29,2, ed. H. Temporini and W. Haase (Berlin
1983) pp. 948–87
——, “Insular Celtic and the Emergence of Welsh”, Britain 400–600: Language and
History, ed. A. Bammesberger and A. Wollmann (Heidelberg 1990) pp. 149–77
N. Everett, “Literacy and the law in Lombard government”, Early Medieval Europe
9 (2000) pp. 93–127
E. Ewig, “Beobachtungen zur politisch-geographischen Terminologie des Fränki-
schen Großreiches und der Teilreiche des 9. Jahrhunderts”, id., Spätantikes und
  651

fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-München

1976) pp. 323–61
——, “Descriptio Franciae”, id., Spätantikes und fränkisches Gallien 1, ed. H. Atsma,
Beihefte der Francia 3/1 (Zürich-München 1976) pp. 274–322
——, “Das Fortleben römischer Institutionen in Gallien und Germanien”, id.,
Spätantikes und fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-
München 1976) pp. 409–34
——, “Die fränkischen Teilungen und Teilreiche (511–613)”, id., Spätantikes und
fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-München
1976) pp. 114–71
——, “Die fränkischen Teilreiche im 7. Jahrhundert (613–714)”, id., Spätantikes und
fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-München
1976) pp. 172–230
——, “Résidence et capitale pendant le haut Moyen Age”, id., Spätantikes und
fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-München
1976) pp. 362–408
——, “Volkstum und Volksbewußtsein im Frankenreich des 7. Jahrhunderts”, id.,
Spätantikes und fränkisches Gallien 1, ed. H. Atsma, Beihefte der Francia 3/1 (Zürich-
München 1976) pp. 231–73
——, Spätantikes und fränkisches Gallien. Gesammelte Schriften (1952–1973), 2 vols., ed.
H. Atsma, Beihefte der Francia 3 (Zürich-München 1976/79)
——, Die Merowinger und das Frankenreich, Urban 392 (Stuttgart 1988)
——, “Troiamythos und fränkische Frühgeschichte”, Die Franken und die Alemannen
bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände
zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-New York 1998)
pp. 1–30
Excavations at Carthage 1978 conducted by the University of Michigan, vol. 7, ed. J.H.
Humphrey (Ann Arbor 1982)
R. Faith, The English Peasantry and the Growth of Lordship (Leicester 1997)
Fälschungen im Mittelalter. Internationaler Kongreß der Monumenta Germaniae Historica, vol.
1: Kongreßdaten und Festvorträge. Literatur und Fälschung, Schriften der MGH 33,1
(Stuttgart 1988)
S. Fanning, “Emperors and Empires in fifth-century Gaul”, Fifth-century Gaul: a cri-
sis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge 1992) pp. 288–97
B. Fauvarque, “Le baptême de Clovis, ouverture du millénaire des saints”, Clovis.
Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris
1997) pp. 271–86
J. Favrod, Histoire politique du royaume burgonde (443–534) (Lausanne 1997)
S. Feist, Vergleichendes Wörterbuch der Gotischen Sprache (Leiden 1939)
H. de Felipe, “Leyendas árabes sobre el orígen de los beréberes”, Al-Qantara 11/2
(1990) pp. 379–98
A. Ferreiro, “Linguarum diversitate: Babel and Pentecost in Leander’s Homily at the
third Council of Toledo”, Actas del XIV Centenario del Concilio III de Toledo, 589–1989
(Toledo 1991) pp. 237–48
Festschrift für Helmut Beumann zum 65. Geburtstag, ed. K.-U. Jäschke and R. Wenskus
(Sigmaringen 1979)
Festschrift für Otto-Herman Frey, ed. C. Dobiat, Marburger Studien zur Vor- und
Frühgeschichte 16 (Hitzeroth 1994)
Festschrift Adolf Hofmeister zum 70. Geburtstag, ed. U. Scheil (Halle 1955)
N. Fettich, Die Metallkunst der landnehmenden Ungarn, Archaeologia Hungarica 21
(Budapest 1937)
U. Fiedler, “Zu Hermann Aments These eines romanischen Bestattungsareals auf
dem merowingerzeitlichen Gräberfeld von Dieue-sur-Meuse”, Archäologie des Franken-
reiches—Neueste Forschungen. Internationales Kolloquium anläßlich der Ausstellung “Die
652 

Franken—Les Francs” Berlin, 10. bis 12. Oktober 1997, ed. W. Menghin, Acta Praehistorica
et Archaeologica 30 (1998) pp. 244–52
Fifth-century Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge
1992)
W. Filmer-Sankey, “The ‘Roman Emperor’ in the Sutton Hoo Ship Burial”, Journal
of the British Archaeological Association 149 (1996) pp. 1–9
G. Fingerlin, “Franken am Kaiserstuhl”, Archäologische Nachrichten aus Baden 44 (1990)
pp. 7–15
——, “Ein früher Stützpunkt fränkischer Herrschaft am Oberrhein. Neue mero-
wingerzeitliche Grabfunde aus Bad Krotzingen, Kreis Breisgau-Hochschwarzwald”,
Archäologische Ausgrabungen in Baden-Württemberg (1998) pp. 200–3
Th. Fischer, “Der Übergang von der Spätantike zum frühen Mittelalter in Ostbayern”,
Regensburg—Kelheim—Straubing 1, ed. S. Rieckhoff-Pauli and W. Torbrügge, Führer
zu archäologischen Denkmälern 5 (Stuttgart 1984) pp. 236–43
——, “Römer und Germanen an der Donau”, Die Bajuwaren. Von Severin bis Tassilo
488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp.
39–45
——, “Zur Archäologie des fünften Jahrhunderts in Ostbayern”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 204. Veröffentlichungen der Kommission für Frühmittelalter-
forschung 13 (Wien 1990) pp. 101–22
——, Das Umland des römischen Regensburg, Münchener Beiträge zur Vor- und Früh-
geschichte 42 (München 1990)
——, Das bajuwarische Reihengräberfeld von Staubing. Studien zur Frühgeschichte im bayerischen
Donauraum, Kataloge der Prähistorischen Staatssammlung 26 (Kallmünz 1993)
——, Die Römer in Deutschland (Stuttgart 1999)
Th. Fischer and H. Geisler, “Herkunft und Stammesbildung der Baiern aus archäol-
ogischer Sicht”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer
and H. Dopsch (Rosenheim-Salzburg 1988) pp. 61–8
G. Fisher, “Kingdom and community in early Anglo-Saxon eastern England”, Regional
Approaches to Mortuary Analysis, ed. L. Anderson Beck (New York 1995) pp. 147–66
J. Fontaine, Isidore de Seville et la culture classique dans l’Espagne wisigothique (Paris 1959)
——, “Théorie et pratique du style chez Isidore de Seville”, Vigiliae Christianae 14,2
(1960) pp. 65–101
——, “Conversion et culture chez les Wisigoths d’Espagne”, La conversione al cris-
tianesimo nell’Europa dell’alto medioevo, Settimane di studio del centro italiano di studi
sull’alto medioevo 14 (1967) pp. 87–147
——, “El de viris illustribus de Ildefonso de Toledo: tradición y originalidad”, Anales
Toledanos 3 (1971) pp. 59–96
——, Culture et spiritualité en Espagne du VI e au VII e siècle (London 1986)
——, “La homilía de San Leandro ante el III Concilio de Toledo: Temática y
forma”, Actas del XIV Centenario del Concilio III de Toledo, 589–1989 (Toledo 1991)
pp. 249–70
——, Isidore de Seville. Genèse et originalité de la culture Hispanique au temps des Wisigoths
(Turnhout 2000)
The Formation of al-Andalus, vol. 2: Language, Religion, Culture and Sciences, ed. M. Fierro
and J. Samsó (Aldershot 1998)
P.J. Fouracre, “The Nature of Frankish Political Institutions in the Seventh Century”,
Franks and Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N.
Wood, Studies in Historical Archaeoethnology 3 (San Marino 1998) pp. 285–301;
301–16
J. Frána and A. Ma“talka, “Röntgenfluoreszenzanalyse von frühmittelalterlichen
Bronzen aus Böhmen und Mähren”, Awarenforschungen, ed. F. Daim, Archaeologia
Austriaca 1. Studien zur Archäologie der Awaren 4 (Wien 1992) pp. 779–801
  653

Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich,
Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-
New York 1998)
Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog
zur Ausstellung im Reiss-Museum Mannheim, 2 vols. (Mainz 1996)
Franks and Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N. Wood,
Studies in Historical Archaeoethnology 3 (San Marino 1998)
E.A. Freeman, Western Empire in the Fifth Century (London 1904)
P. Fried, “Bischof Simpert und das Bistum Neuburg-Staffelsee”, Jahrbuch des Vereins
für Augsburger Bistumsgeschichte 12 (1978) pp. 181–5
A. Friese, Studien zur Herrschaftsgeschichte des fränkischen Adels. Der mainländisch-thüringi-
sche Raum vom 7. bis 11. Jahrhundert, Geschichte und Gesellschaft. Bochumer Histo-
rische Studien 18 (Stuttgart 1979)
H. Fröhlich, Studien zur langobardischen Thronfolge. Von den Anfängen bis zur Eroberung
des italienischen Reiches durch Karl den Großen (774), 2 vols. (doct. diss. Tübingen
1980)
From the Baltic to the Black Sea. Studies in medieval archaeology, ed. D. Austin and
L. Alcock, One world archaeology 18 (London 1990) pp. 138–46
Früh- und hochmittelalterlicher Adel in Schwaben und Bayern, ed. I. Eberl, W. Hartung and
J. Jahn, Regio 1 (Sigmaringendorf 1988)
Frühe Baiern im Straubinger Land (Straubing c. 1995)
Frühes Mönchtum in Salzburg, ed. E. Zwink, Salzburg Diskussionen 4 (Salzburg 1983)
Frühmittelalterliche Ethnogenese im Alpenraum, ed. H. Beumann and W. Schröder, Nationes
5 (Sigmaringen 1985)
D. Frye, “Gundobad, the Leges Burgundionum, and the struggle for sovereignty in
Burgundy”, Classica et Medievalia 41 (1990) pp. 203–4
G. Fülöp, “Awarenzeitliche Fürstenfunde von Igar”, Acta Archaeologica Hungarica
Academiae Scientiarum Hungaricae 40 (1988) pp. 151–90
——, “New research on finds of Avar chieftain-burials at Igar, Hungary”, From the
Baltic to the Black Sea. Studies in medieval archaeology, ed. D. Austin and L. Alcock,
One world archaeology 18 (London 1990) pp. 138–46
Gallien in der Spätantike. Von Kaiser Constantin zu Frankenkönig Childerich (Mainz 1980)
É. Garam, “A bõcsi kesõavarkori lelet és kore [Der spätawarenzeitliche Fund von
Bõcs und sein Kreis]”, Archeologiai Értesítõ 108 (1981) pp. 34–50
——, “Bemerkungen zum ältesten Fundmaterial der Awarenzeit”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denks-
chriften der Österreichischen Akademie der Wissenschaften, philosophisch-historische
Klasse 204. Veröffentlichungen der Kommission für Frühmittelalter-forschung 13
(Wien 1990) pp. 253–72
——, “Über Halsketten, Halsschmucke mit Anhängern und Juwelenkragen byzan-
tinischen Ursprungs aus der Awarenzeit”, Acta Archaeologica Hungarica Academiae
Scientiarum Hungaricae 43 (1991) pp. 151–79
——, “Die münzdatierten Gräber der Awarenzeit”, Awarenforschungen 1, ed. F. Daim,
Archaeologia Austriaca 1. Studien zur Archäologie der Awaren 4 (Wien 1992)
pp. 135–250
——, “Die awarenzeitlichen Scheibenfibeln”, Communicationes Archaeologicae Hungariae
(1993) pp. 99–143
——, Katalog der awarenzeitlichen Goldgegenstände und der Fundstücke aus den Fürstengräbern
im Ungarischen Nationalmuseum, Catalogi Musei Nationalis Hungarici, Seria Archeologica
1 (Budapest 1993)
——, “Gürtelverzierungen byzantinischen Typs im Karpatenbecken des 6.–7. Jahrhun-
derts”, Acta Archaeologica Hungarica Academiae Scientiarum Hungaricae 50 (1999/2000)
pp. 379–91
——, Funde byzantinischer Herkunft in der Awarenzeit vom Ende des 6. bis zum Ende des 7.
Jahrhunderts, Monumenta Avarorum Archaeologica 5 (Budapest 2001)
654 

Y. García López, “La cronología de la Historia Wambae”, Anuario de Estudios medievales

23 (1993) pp. 121–39
L.A. García Moreno, El fin del reino visigodo de Toledo (Madrid 1975)
——, Historia de España visigoda (Madrid 1989)
——, “Gothic Survivals in the Visigothic Kingdom”, Francia 21/1 (1994) pp. 1–15
R. Gardner, “Gildas’ New Testament Models”, Cambrian Medieval Celtic Studies 30
(1995) pp. 1–12
S. Gasparri, I duchi Longobardi, Studi storici 109 (Roma 1978)
——, La cultura tradizionale dei Longobardi. Struttura tribale e resistenze pagane (Spoleto
1983)
——, “Il regno Longobardo in Italia”, Langobardia, ed. id. and P. Cammarosano
(Udine 1990) pp. 237–305
——, Prima delle nazioni. Popoli, etnie e regni fra Antichità e Medioevo (Rom 1997)
——, “Kingship rituals and ideology in Lombard Italy”, Rituals of Power. From Late
Antiquity to the Early Middle Ages, ed. F. Theuws and J.L. Nelson, The Transformation
of the Roman World 8 (Leiden-Boston-Köln 2000) pp. 95–114
H.L.G. Gastroph, Herrschaft und Gesellschaft in der Lex Baiuvariorum. Ein Beitrag zur
Strukturanalyse des Agilolfingischen Stammesherzogtums vom 6. zum 8. Jahrhundert, Miscellanea
Bavarica monacensia 53. Neue Schriftenreihe des Stadtarchivs München 71
(München 1974)
N. Gauthier, “La topographie chrétienne entre idéologie et pragmatisme”, The Idea
and Ideal of the Town Between Late Antiquity and the Early Middle Ages, ed. G.P. Brogiolo
and B. Ward-Perkins, The Transformation of the Roman World 4 (Leiden-Boston-
Köln 1999) pp. 195–209
H. Geake, The Use of Grave-Goods in Conversion-Period England c. 600–c. 850, British
Archaeological Reports, British series 261 (Oxford 1997)
P.J. Geary, “Ethnic identity as a situational construct in the early middle ages”,
Mitteilungen der Anthropologischen Gesellschaft in Wien 113 (1983) pp. 15–36
——, Before France and Germany. The Creation and Transformation of the Merovingian World
(New York-Oxford 1988) [German transl. Die Merowinger: Europa vor Karl dem Großen
(München 1996)]
——, The Myth of Nations. The Medieval Origins of Europe (Princeton 2001)
M. Gebühr, “Angulus desertus?”, Studien zur Sachsenforschung 11, ed. H.-J. Häßler
(Oldenburg 1998) pp. 43–8
H. Geisler, “Barbing-Kreuzhof ”, Regensburg—Kelheim—Straubing 1, ed. S. Rieckhoff-
Pauli and W. Torbrügge, Führer zu archäologischen Denkmälern 5 (Stuttgart 1984)
pp. 164–73
——, “Neue archäologische Quellen zur frühesten Geschichte der Baiern: spätan-
tike und frühmittelalterliche Gräberfelder aus Straubing”, Typen der Ethnogenese unter
besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
204. Veröffentlichungen der Kommission für Frühmittelalterforschung 13 (Wien
1990) pp. 89–100
M. Gelling, “English place-names from the compound wicham”, Medieval Archaeology
12 (1967) pp. 87–104
——, “Why aren’t we speaking Welsh?”, Anglo-Saxon Studies in Archaeology and History
6 (1993) pp. 51–6
Gender and the Transformation of the Roman World, ed. L. Brubaker and J. Smith
(Cambridge forthcoming)
R.A. Gerberding, The Rise of the Carolingians and the Liber Historiae Francorum, Oxford
Historical Monographs (Oxford 1987)
Germanen, Awaren, Slawen in Niederösterreich: das 1. Jahrtausend nach Christus, ed. H. Windl
(Wien 1977)
Germanen beiderseits des spätantiken Limes, ed. Th. Fischer, G. Precht and J. Tejral, Spisy
Archeologického Ústavu Av ÈR Brno 14 (Köln-Brno 1999)
  655

Germanen, Hunnen und Awaren. Schätze der Völkerwanderungszeit, ed. W. Menghin (Nürnberg
1988)
Germania inferior. Besiedlung, Gesellschaft und Wirtschaft an der Grenze der römisch-germani-
schen Welt, ed. T. Grünewald with H.-J. Schalles, Ergänzungsbände zum Reallexikon
der Germanischen Altertumskunde 28 (Berlin-New York 2001)
Geschichte der Stadt Augsburg von der Römerzeit bis zur Gegenwart, ed. G. Gottlieb, W. Baer,
J. Becker, J. Bellot, K. Filser, P. Fried, W. Reinhard and B. Schimmelpfennig
(Stuttgart 1984)
Geschichte Salzburgs, vol. 1,1: Vorgeschichte, Altertum, Mittelalter, ed. H. Dopsch (Salzburg
1981)
Geschichtsschreibung und geistiges Leben im Mittelalter. Festschrift für Heinz Löwe zum 65.
Geburtstag, ed. K. Hauck and H. Mordek (Cologne 1978)
Geschichtswissenschaft und Archäologie. Untersuchungen zur Siedlungs-, Wirtschafts- und Kirchen-
geschichte, ed. H. Jankuhn and R. Wenskus (Sigmaringen 1979)
D. Geuenich, “Personennamen als Personengruppennamen”, Proceedings of the Thirteenth
International Congress of Onomastic Sciences 1, ed. K. Rymut (Wroclaw 1981) pp.
437–45
——, “Die Alamannen unter fränkischer Herrschaft”, Die Alamannen. Ausstellungskatalaog
Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp. 204–8
——, Geschichte der Alemannen, Urban 575 (Stuttgart-Berlin-Köln 1997)
——, “Chlodwigs Alemannenschlacht(en) und Taufe”, Die Franken und die Alemannen
bis zur “Schlacht bei Zülpich” (496/97), ed. id., Ergänzungsbände zum Reallexikon
der Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 423–37
D. Geuenich and H. Keller, “Alamannen, Alamannien, Alamannisch im frühen
Mittelalter. Möglichkeiten und Schwierigkeiten des Historikers beim Versuch der
Eingrenzung”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 179. Veröffentlichungen der Kommission für Frühmittelalterfor-
schung 8 (Wien 1989) pp. 135–57
D. Geuenich and W. Kettemann, “Das Pilotprojekt zur gens Alamannorum. Erste
Erfahrungen mit einem Teilprojekt von ‘Nomen et gens’”, Nomen et gens. Zur hi-
storischen Aussagekraft frühmittelalterlicher Personennamen, ed. D. Geuenich, W. Haubrichs
and J. Jarnut, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
16 (Berlin-New York 1997) pp. 279–303
A. Giardina and F. Grelle, “La tavola di Trinitapoli: una nueva costituzione di
Valentiniano I”, Mélanges de l’École Française de Rome 95 (1983) pp. 249–303
Gildas: New Approaches, ed. M. Lapidge and D.N. Dumville (Woodbridge 1984)
La giustizia nell’alto medioevo (secoli V–VIII), Settimane di studio del centro italiano di
studi sull’alto medioevo 42 (Spoleto 1995)
M. Gluckman, “The Peace in the Feud”, id., Custom and Conflict in Africa (Oxford
1956) pp. 1–26
——, Custom and Conflict in Africa (Oxford 1956)
R. Göbl and A. Róna-Tas, Die Inschriften des Schatzes von Nagy-Szentmiklós. Eine paläo-
graphische Dokumentation, Denkschriften der Österreichischen Akademie der Wissen-
schaften, philosophisch-historische Klasse 240 (Wien 1995)
H.-W. Goetz, “‘Nobilis’. Der Adel im Selbstverständnis der Karolingerzeit”, Vierteljahres-
schrift für Sozial- und Wirtschaftsgeschichte 70 (1983) pp. 153–91
——, Moderne Mediävistik. Stand und Perspektiven der Mittelalterforschung (Darmstadt
1999)
——, “Gentes et linguae. Völker und Sprachen im Ostfränkisch-deutschen Reich in der
Wahrnehmung der Zeitgenossen”, Theodisca. Beiträge zur althochdeutschen und alt-
niederdeutschen Sprache und Literatur in der Kultur des frühen Mittelalters, ed. W. Hau-
brichs, E. Hellgardt, R. Hildebrandt, S. Müller and K. Ridder, Ergänzungsbände
zum Reallexikon der Germanischen Altertumskunde 22 (Berlin-New York 2000)
pp. 290–312
656 

——, “Gentes. Zur zeitgenössischen Terminologie und Wahrnehmung ostfränkischer

Ethnogenese im 9. Jahrhundert”, Mitteilungen des Instituts für Österreichische Geschichts-
forschung 108 (2000) pp. 85–116
——, “Die germanisch-romanische (Kultur-)Synthese in der Wahrnehmung der
merowingischen Geschichtsschreibung”, Akkulturation. Probleme einer germanisch-roman-
ischen Kultursynthese in Spätantike und frühem Mittelalter, ed. D. Hägermann, W. Haubrichs
and J. Jarnut (forthcoming)
——, “Gens. Terminology and Perception of the Germanic Peoples from Late Anti-
quity to the Early Middle Ages” (in preparation, to be published in a volume of
the series “The Transformation of the Roman World”, ed. W. Pohl)
——, “Zur Wandlung des Frankennamens im Frühmittelalter”, Integration und Herrschaft.
Ethnische Identitäten und soziale Organisation im Frühmittelalter, ed. M. Diesenberger and
W. Pohl (Vienna 2002)
W. Goez, “Über die Anfänge der Agilulfinger”, Jahrbuch für fränkische Landesforschung
34/35 (1975) pp. 145–61
W. Goffart, Barbarians and Romans A.D. 418–584. The Techniques of Accommodation
(Princeton NJ 1980)
——, “The supposedly ‘Frankish’ table of nations: an edition and study”, Frühmittelal-
terliche Studien 17 (1983) pp. 98–130
——, Rome’s Fall and After (London 1989)
——, “Conspicuously absent: martial heroism in the Histories of Gregory of Tours
and its likes”, The World of Gregory of Tours, ed. K. Mitchell and I.N. Wood (Leiden
2002) pp. 365–93
Z. G∑ab, The Origins of the Slavs. A Linguist’s View (Columbus Ohio 1992)
H. Grahn-Hoek, Die fränkische Oberschicht im 6. Jahrhundert. Studien zu ihrer rechtlichen
und politischen Stellung, Vorträge und Forschungen Sonderband 21 (Sigmaringen
1976)
A.S. Gratwick, “Latinitas Britannica: Was British Latin Archaic?”, Latin and the Vernacular
Languages in Early Medieval Britain, ed. N. Brooks (Leicester 1982) pp. 1–79
D. Greene, “The Making of Insular Celtic”, Proceedings of the Second International Congress
of Celtic Studies (Cardiff 1966) pp. 121–36
——, “Some Linguistic Evidence Relating to the British Church”, Christianity in
Roman Britain, ed. M.W. Barley and R.P.C. Hanson (Leicester 1968) pp. 75–86
Grenze und Differenz im frühen Mittelalter, ed. W. Pohl and H. Reimitz, Forschungen
zur Geschichte des Mittelalters 1 (Wien 2000)
Grenzen und Grenzregionen. Frontières et régions frontalières. Borders and Border Regions, ed.
W. Haubrichs and R. Schneider, Veröffentlichungen der Kommission für
Saarländische Landesgeschichte und Volksforschung 22 (Saarbrücken 1993)
P. Grierson, M. Blackburnand L. Travaini, Medieval European Coinage, 2 vols. (Cambridge
1986/98)
B. Grosskopf-Theune, “Die Kontrolle der Verkehrswege. Ein Schlüssel zur fränkischen
Herrschaftssicherung”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg
(Stuttgart 1997) pp. 237–42
Grundfragen der alemannischen Geschichte, Vorträge und Forschungen 1 (Sigmaringen 1955)
Ch. Grünewald, “Neues zu Sachsen und Franken in Westfalen”, Studien zur Sachsenfor-
schung 12, ed. H.-J. Häßler (Oldenburg 1999) pp. 83–108
M. Grünewald, “Worms zwischen Burgund und Saliern”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 160–2
J. Guillaume, “Les nécropoles mérovingiennes de Dieue/Meuse (France)”, Acta
Praehistorica et Archaeologica 5/6 (1976/77) pp. 301–9
R. Günther, “Laeti, foederati und Gentilen in Nord- und Nordwestgallien im
Zusammenhang mit der sogenannten Laetenzivilisation”, Zeitschrift für Archäologie
5 (1971) pp. 39–59
  657

——, “Einige neue Untersuchungen zu den Laeten und Gentilen in Gallien im 4.

Jahrhundert und zu ihrer historischen Bedeutung”, Klio 69 (1977) pp. 311–21
R. Günther and A.R. Korsunskij, Germanen erobern Rom. Der Untergang des Weströmischen
Reiches und die Entstehung germanischer Königreiche bis zur Mitte des 6. Jahrhunderts,
Veröffentlichungen des Zentralinstituts für Alte Geschichte und Archäologie der
Akademie der Wissenschaften der DDR 15 (Berlin/Ost 1986; 2nd edn. 1988)
S.L. Guterman, From Personal to Territorial Law. Aspects of the History and Structure of the
Western Legal-Constitutional Tradition (New York 1972)
B. Gutmann, Studien zur römischen Außenpolitik in der Spätantike (364–395 n. Chr.) (Bonn
1991)
W. Haarnagel, Die Grabung Feddersen Wierde. Methode, Hausbau, Siedlungs- und Wirtschafts-
formen sowie Sozialstrukturen (Wiesbaden 1979) pp. 230–8
O. Hageneder, “Die kirchliche Organisation im Zentralalpenraum vom 6. bis 10.
Jahrhundert”, Frühmittelalterliche Ethnogenese im Alpenraum, ed. H. Beumann and
W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 201–35
S. Haider, “Zur Baugeschichte des Salzburger Virgil-Domes”, Mitteilungen des Instituts
für Österreichische Geschichtsforschung 80 (1972) pp. 35–47
——, “Oberösterreich im bairischen Stammesherzogtum”, Baiernzeit in Oberösterreich.
Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts,
Kataloge des Oberösterreichischen Landesmuseums 97 (Linz 1977) pp. 13–26
J. Haldon, Byzantium in the Seventh Century (Cambridge 1990)
G. Halsall, “The origins of the Reihengräberzivilisation: forty years on”, Fifth-century
Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge 1992) pp.
196–207
——, Early Medieval Cemeteries: An archaeological introduction to burial in the Post-Roman
West (Glasgow 1995)
——, Settlement and Social Organization. The Merovingian Region of Metz (Cambridge 1995)
——, “Social identities and social relationships in early Merovingian Gaul”, Franks
and Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N. Wood,
Studies in Historical Archaeoethnology 3 (San Marino 1998) pp. 141–65; 165–75
——, “Archaeology and the late Roman frontier in northern Gaul: The so-called
‘Föderatengräber’ reconsidered”, Grenze und Differenz im frühen Mittelalter, ed.
W. Pohl and H. Reimitz, Forschungen zur Geschichte des Mittelalters 1 (Wien
2000) pp. 167–80
H. Hamerow, “The archaeology of rural settlements in early medieval Europe”,
Early Medieval Europe 3 (1994) pp. 167–79
——, “Migration theory and the Anglo-Saxon ‘identity crisis’”, Migrations and Invasion
in Archaeological Explanation, ed. J. Chapman and H. Hamerow, British Archaeological
Reports, International series 664 (1997) pp. 33–44
——, “The earliest Anglo-Saxon kingdoms”, New Cambridge Medieval History, vol. 1:
c. 500–700 (Cambridge forthcoming)
J. Hampel, Alterthümer des frühen Mittelalters in Ungarn I–III (Braunschweig 1905)
Handbuch der bayerischen Geschichte, vol. 1, ed. M. Spindler (2nd edn., München 1981)
M.A. Handley, The Inscribed Charters of Early Medieval Wales (unpublished M.Phil. dis-
sertation, University of Cambridge 1995)
J. Hannig, Consensus fidelium. Frühfeudale Interpretationen des Verhältnisses von Königtum und
Adel am Beispiel des Frankenreiches, Monographien zur Geschichte des Mittelalters
27 (Stuttgart 1982)
——, “Ars donandi. Zur Ökonomie des Schenkens im früheren Mittelalter”, Armut,
Liebe, Ehre, ed. R. van Dülmen, Studien zur historischen Kulturforschung 1
(Frankfurt a. M. 1986) pp. 11–37; pp. 275–8
H. Härke, “‘Warrior graves?’ The background of the Anglo-Saxon burial rite”, Past
and Present 126 (1990) pp. 22–43
——, “All quiet on the western front? Paradigms, methods and approaches in West
658 

German archeology”, Archaeological Theory in Europe. The Last Three Decades, ed.
I. Hodder (London-New York 1991) pp. 187–222
——, “Early Anglo-Saxon social structure”, The Anglo-Saxons from the Migration Period
to the Eighth Century: An Ethnographic Perspective, ed. J. Hines (Woodbridge 1997) pp.
125–60
——, “Sächsische Ethnizität und archäologische Deutung im frühmittelalterlichen
England”, Studien zur Sachsenforschung 12, ed. H.-J. Häßler (Oldenburg 1999) pp.
109–22
K.W. Harl, Coinage in the Roman Economy, 300 B.C. to A.D. 700 (Baltimore-London
1996)
J. Harries, Sidonius Apollinaris and the Fall of Rome (Oxford 1994)
D. Harrison, The Early State and the Towns. Forms of Integration in Lombard Italy, A.D.
568–774, Lund Studies in International History 29 (Lund 1993)
——, “Political rhetoric and political ideology in Lombard Italy”, Strategies of Distinc-
tion. The Construction of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz,
The Transformation of the Roman World 2 (Leiden-New York-Köln 1998)
pp. 241–54
G. Harrison, “Quoit brooches and the Roman-Medieval transition”, Theoretical Roman
Archaeology Conference, ed. P. Baker (Oxford 1999) pp. 108–20
K. Harrison, “Early Wessex annals in the Anglo-Saxon Chronicle”, English Historical
Review 86 (1971) pp. 527–33
C. Hart, “The Aldewerk and minster at Shelford, Cambridgeshire”, Anglo-Saxon Studies
in Archaeology and History 8 (1995) pp. 43–68
L.M. Hartmann, Geschichte Italiens im Mittelalter, vol. 2,1: Römer und Langobarden bis zur
Teilung Italiens (Leipzig 1900)
W. Hartmann, “Das Recht”, Die Bajuwaren. Von Severin bis Tassilo 488–788, ed.
H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 266–72
W. Hartmann and H. Dopsch, “Bistümer, Synoden und Metropolitanverfassung”,
Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988) pp. 318–26
W. Hartung, Süddeutschland in der frühen Merowingerzeit. Studien zu Gesellschaft, Herrschaft,
Stammesbildung bei Alamannen und Bajuwaren, Vierteljahreshefte zur Sozial- und Wirt-
schaftsgeschichte, Beiheft 73 (Wiesbaden 1983)
A. Harvey, “The Significance of Cothraige”, Ériu 36 (1985) pp. 1–9
I. Haselbach, Aufstieg und Herrschaft der Karlinger in der Darstellung der sogenannten Annales
Mettenses priores. Ein Beitrag zur Geschichte der politischen Ideen im Reiche Karls des Großen,
Historische Studien 412 (Lübeck-Hamburg 1970)
W. Haubrichs, “Germania submersa. Zu Fragen der Quantität und Dauer ger-
manischer Siedlungsinseln im romanischen Lothringen und Südbelgien”, Verborum
amor. Studien zur Geschichte und Kunst der deutschen Sprache. Festschrift Stefan Sonderegger,
ed. H. Burger, A.M. Haas and P. von Matt (Berlin-New York 1992) pp. 633–66
——, “Über die allmähliche Verfertigung von Sprachgrenzen. Das Beispiel der
Kontaktzonen von Germania und Romania”, Grenzen und Grenzregionen. Frontières
et régions frontalières. Borders and Border Regions, ed. W. Haubrichs and R. Schneider,
Veröffentlichungen der Kommission für Saarländische Landesgeschichte und
Volksforschung 22 (Saarbrücken 1993) pp. 99–129
——, “Fränkische Lehnwörter, Ortsnamen und Personennamen im Nordosten der
Gallia. Die ‘Germania submersa’ als Quelle der Sprach- und Siedlungsgeschichte”,
Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich,
Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-
New York 1998) pp. 102–29
G. Hauptfeld, “Die Gentes im Vorfeld von Ostgoten und Franken im sechsten
Jahrhundert”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
  659

historische Klasse 179. Veröffentlichungen der Kommission für Frühmittelalter-

forschung 8 (Wien 1989) pp. 121–34
H.-W. Haussig, Die Geschichte Zentralasiens und der Seidenstraße in vorislamischer Zeit
(Darmstadt 1983)
S.C. Hawkes, H.R.E. Davidson and C. Hawkes, “The Finglesham man”, Antiquity
39 (1965) pp. 17–32
M. Haycock, “Metrical models for the poems in the Book of Taliesin”, Early Welsh
Poetry: Studies in the Book of Aneirin, ed. B.F. Roberts (Aberystwyth 1988) pp. 155–78
P.J. Heather, “Cassiodorus and the Rise of the Amals: Genealogy and the Goths
under Hun Domination”, Journal of Roman Studies 79 (1989) pp. 103–28
——, Goths and Romans 332–489 (Oxford 1991)
——, “The Historical Culture of Ostrogothic Italy”, Teoderico il grande e i Goti d’Italia.
Atti del XIII Congresso internazionale di studi sull’Alto Medioevo (Spoleto 1993) pp. 317–53
——, “The Huns and the end of the Roman Empire in western Europe”, English
Historical Review 110 (1995) pp. 4–41
——, “Theoderic, king of the Goths”, Early Medieval Europe 4 (1995) pp. 145–73
——, The Goths (Oxford 1996)
——, “Disappearing and reappearing tribes”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transforma-
tion of the Roman World 2 (Leiden-New York-Köln 1998) pp. 95–111
——, “The Creation of the Visigoths”, The Visigoths. From the Migration Period to the
Seventh Century. An Ethnographic Perspective, ed. id. (San Marino 1999) pp. 42–73
——, “State, Lordship and Community in the West (c. A.D. 400–600)”, Cambridge
Ancient History 14 (2nd edn., 2000) pp. 437–68
P.J. Heather and J.F. Matthews, The Goths in the fourth Century (Liverpool 1991)
L. Hedeager, Iron Age Societies: From Tribe to State in Northern Europe, 500 B.C. to A.D.
700 (Oxford 1992)
——, “Myth and art: a passport to political authority in Scandinavia during the
migration period”, The Making of Kingdoms, ed. T. Dickinson and D. Griffiths,
Anglo-Saxon Studies in Archaeology and History 10 (Oxford 1999) pp. 151–6
——, “Asgard reconstructed? Gudme—a ‘central place’ in the North”, Places of
Power, ed. M. de Jong, F. Theuws and C. van Rhijn, The Transformation of the
Roman World 6 (Leiden-Boston 2001) pp. 467–508
Der heilige Willibald—Klosterbischof oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and
S. Weinfurter, Eichstätter Studien NF 30 (Regensburg 1990)
M. Heinzelmann, Bischofsherrschaft in Gallien. Zur Kontinuität römischer Führungsschichten
vom 4. bis zum 7. Jahrhundert. Soziale, prosopographische und bildungsgeschichtliche Aspekte,
Beihefte der Francia 5 (Zürich-München 1976)
H. Hencken, “Lagore: a royal residence of the seventh to tenth centuries”, Proceedings
of the Royal Irish Academy (C) 42 (1950) pp. 1–76
M.F. Hendy, “Mint and Fiscal Administration under Diocletian, his Colleagues and
his Successors, A.D. 305–324”, Journal of Roman Studies 62 (1972) pp. 75–82
——, Studies in the Byzantine Money Economy (Cambridge 1985)
J. Henning, “Germanische Siedlungen im spätrömischen Nordgallien. Neue Ergebnisse
archäologischer Forschungen”, Zeitschrift für Archäologie 23 (1989) pp. 211–7
J. Heuclin, “Le concile d’Orléans de 511, un premier concordat?” Clovis. Histoire et
mémoire, vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp.
435–50
J. Heywood, Dark Age Naval Power (London 1991)
N. Higham, The English Conquest. Gildas and Britain in the Fifth Century (Manchester
1994)
J.N. Hillgarth, “Historiography in Visigothic Spain”, La storiografia altomedievale,
Settimane di studio del centro italiano di studi sull’alto medioevo 17 (Spoleto
1970) pp. 261–311
660 

C. Hills, “Did the people of Spong Hill come from Schleswig-Holstein?”, Studien
zur Sachsenforschung 11, ed. H.-J. Häßler (Oldenburg 1998) pp. 145–54
——, “Spong Hill and the Adventus Saxonum”, Spaces of the Living and Dead: an
Archaeological Dialogue, ed. C.E. Karkov, K.M. Wickham-Crowley and B.K. Young,
American Early Medieval Studies 3 (Oxford 1999) pp. 15–26
J. Hines, The Scandinavian Character of Anglian England in the Pre-Viking Period, British
Archaeological Reports, British series 124 (Oxford 1984)
——, “Philology, archaeology and the adventus Saxonum vel Anglorum”, Britain 400–600:
Language and History, ed. A. Bammesberger and A. Wollmann (Heidelberg 1990)
pp. 17–36
——, “The Scandinavian character of Anglian England: an update”, The Age of
Sutton Hoo, ed. M. Carver (Woodbridge 1992) pp. 315–29
——, “The becoming of the English: identity, material culture and language in
early Anglo-Saxon England”, Anglo-Saxon Studies in Archaeology and History 7 (1994)
pp. 49–59
D. Hinton, “The Inscribed Stones in Lady St Mary Church, Wareham”, Proceedings
of the Dorset Natural History and Archaeological Society 114 (1992) pp. 260
Historia de España de Ramón Menéndez Pidal, Part 3,2: España visigoda (Madrid 1991)
pp. 209–67
Historiographie im frühen Mittelalter, ed. A. Scharer and G. Scheibelreiter (Vienna-
Munich 1994)
E. Hlawitschka, “Studien zur Genealogie und Geschichte der Merowinger und
frühen Karolinger”, Rheinische Vierteljahrsblätter 43 (1979) pp. 1–99
O. Höfler, “Das germanische Kontinuitätsproblem”, Historische Zeitschrift 157 (1938)
pp. 1–26
H. Hofmann, “Roma caput mundi? Rom und Imperium Romanum in der liter-
arischen Diskussion zwischen Spätantike und dem 9. Jahrhundert”, Roma fra
oriente e occidente, Settimane di studio del centro italiano di studi sull’alto medio-
evo 49 (Spoleto 2002) pp. 493–556
Y. Hollevoet and J.-P. van Roeyen, “Germanic settlers at Sint-Gillis-Waas? (Prov.
East-Flanders)”, Archaeologie in Vlandern 2 (1992) pp. 209–21
K. Holter, “Kunstschätze der Gründungszeit”, Die Anfänge des Klosters Kremsmünster,
ed. S. Haider, Mitteilungen des Oberösterreichischen Landesarchivs, Ergänzungsband
2 (Linz 1978) pp. 111–43
A.M. Honoré, Emperors and Lawyers (London 1981)
——, “The Making of the Theodosian Code”, Zeitschrift der Savigny-Stiftung für Rechts-
geschichte, Romanistische Abteilung 103 (1986) pp. 133–222
——, Law in the Crisis of Empire, 379–455 A.D. The Theodosian Dynasty and its Quaestors
(Oxford 1998)
J. Hopkins, Medieval Government in Barbary (London 1958)
N. Howe, Migration and Mythmaking in Anglo-Saxon England (Yale 1989)
E.H. Huber, “Neu entdeckte Awarengräber in Wien, Simmering”, Fundort Wien.
Berichte zur Archäologie 1 (1998) pp. 117–43
J. Huggett, “Imported grave goods and the early Anglo-Saxon economy”, Medieval
Archaeology 32 (1988) pp. 63–96
H.J. Hummer, “Franks and Alamanni: a discontinuous ethnogenesis”, Franks and
Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N. Wood, Studies
in Historical Archaeoethnology 3 (San Marino 1998) pp. 9–21; 21–32
H.-J. Hundt, “Textilreste aus awarischen Gräbern von Leobersdorf und ein Exkurs
über gegossene Textilstrukturen an der Rückseite bronzener Riemenzungen”,
F. Daim et al., Das awarische Gräberfeld von Leobersdorf, Niederösterreich 2, Studien zur
Archäologie der Awaren 3. Denkschriften der Österreichischen Akademie der
Wissenschaften, philosophisch-historische Klasse 194. Veröffentlichungen der
Kommission für Frühmittelalterforschung 10 (Wien 1987) pp. 9–18
  661

Iconologia sacra. Festschrift für K. Hauck zum 75. Geburtstag, ed. H. Keller and N. Staubach,
Arbeiten zur Frühmittelalterforschung 23 (Berlin-New York 1994)
The Idea and Ideal of the Town Between Late Antiquity and the Early Middle Ages, ed. G.P.
Brogiolo and B. Ward-Perkins, The Transformation of the Roman World 4
(Leiden-Boston-Köln 1999)
Integration und Herrschaft. Ethnische Identitäten und soziale Organisation im Frühmittelalter, ed.
M. Diesenberger and W. Pohl (Vienna 2002)
F. Irsigler, Untersuchungen zur Geschichte des frühfränkischen Adels, Rheinisches Archiv 70
(Bonn 1969)
K.H. Jackson, Language and History in Early Britain (Edinburgh 1953)
——, “The Britons of Southern Scotland”, Antiquity 29 (1955) pp. 77–88
——, “Varia II: Gildas and the names of the British Princes”, Cambridge Medieval
Celtic Studies 3 (1982) pp. 30–40
J. Jahn, “Urkunde und Chronik. Ein Beitrag zur historischen Glaubwürdigkeit der
Benediktbeurer Überlieferung und zur Geschichte des agilolfingischen Bayern”,
Mitteilungen des Instituts für Österreichische Geschichtsforschung 95 (1987) pp. 1–51
——, “Bayerische ‘Pfalzgrafen’ im 8. Jahrhundert? Studien zu den Anfängen Herzog
Tassilos (III.) und zur Praxis der fränkischen Regentschaft im agilolfingischen
Bayern”, Früh- und hochmittelalterlicher Adel in Schwaben und Bayern, ed. I. Eberl,
W. Hartung and J. Jahn, Regio 1 (Sigmaringendorf 1988) pp. 80–114
——, Ducatus Baiuvariorum. Das bairische Herzogtum der Agilolfinger, Monographien zur
Geschichte des Mittelalters 35 (Stuttgart 1991)
——, “Hausmeier und Herzöge. Bemerkungen zur agilolfingisch-karolingischen
Rivalität bis zum Tode Karl Martells”, Karl Martell in seiner Zeit, ed. J. Jarnut,
U. Nonn and M. Richter, Beihefte der Francia 37 (Sigmaringen 1994) pp. 317–44
E. James, The Merovingian archaeology of south-west Gaul, 2 vols., British Archaeological
Reports, Supplementary Series 25 (Oxford 1977)
——, The Franks (London 1988)
——, “The origins of barbarian kingdoms: the continental evidence”, The Origins of
Anglo-Saxon Kingdoms, ed. S. Bassett (Leicester 1989) pp. 40–52
——, “Gregory of Tours and the Franks”, After Rome’s Fall. Narrators and Sources of
Early Medieval History. Essays presented to Walter Goffart, ed. A.C. Murray (Toronto-
Buffalo-London 1998) pp. 51–66
J. Jarnut, “Beiträge zu den fränkisch-bayerisch-langobardischen Beziehungen im 7.
und 8. Jahrhundert”, Zeitschrift für bayerische Landesgeschichte 39 (1976) pp. 331–52
——, “Studien über Herzog Odilo (736–748)”, Mitteilungen des Instituts für Österrei-
chische Geschichtsforschung 85 (1977) pp. 273–84
——, Geschichte der Langobarden, Urban 339 (Stuttgart 1982)
——, “Zur Frühgeschichte der Langobarden”, Studi Medievali 3a ser. 24/1 (1983)
pp. 1–16
——, Agilolfingerstudien. Untersuchungen zur Geschichte einer adligen Familie im 6. und 7.
Jahrhundert, Monographien zur Geschichte des Mittelalters 32 (Stuttgart 1986)
——, “I Longobardi nell’epoca precedente all’occupazione dell’Italia”, Langobardia,
ed. S. Gasparri and P. Cammarosano (Udine 1990) pp. 3–34
——, “Genealogie und politische Bedeutung der agilolfingischen Herzöge”, Mitteilungen
des Instituts für Österreichische Geschichtsforschung 90 (1991) pp. 1–22
——, “Die Landnahme der Langobarden aus historischer Sicht”, Ausgewählte Probleme
europäischer Landnahmen des Früh- und Hochmittelalters. Methodische Grundlagendiskussion
im Grenzbereich zwischen Archäologie und Geschichte 1, ed. M. Müller-Wille and
R. Schneider, Vorträge und Forschungen 41 (Sigmaringen 1993) pp. 173–94
——, “Gregor von Tours, Frankengeschichte II,12: Franci Egidium sibi regem ads-
ciscunt. Faktum oder Sage?”, Ethnogenese und Überlieferung. Angewandte Methoden der
Frühmittelalterforschung, ed. K. Brunner and B. Merta, Veröffentlichungen des Instituts
für Österreichische Geschichtsforschung 31 (Wien-München 1994) pp. 129–34
662 

——, “Münzbilder als Zeugnisse langobardischer Herrschaftsvorstellungen”, Iconologia

sacra. Festschrift für K. Hauck zum 75. Geburtstag, ed. H. Keller and N. Staubach,
Arbeiten zur Frühmittelalterforschung 23 (Berlin-New York 1994) pp. 283–90
——, “Die langobardische Herrschaft über Rugiland und ihre politischen Hinter-
gründe”, Westillyricum und Nordostitalien in der spätrömischen Zeit, ed. R. Brato≥
(Ljubljana 1996) pp. 207–13
——, “Teotischis homines (a. 816). Studien und Reflexionen über den ältesten
(urkundlichen) Beleg des Begriffs ‘theodiscus’”. Mitteilungen des Instituts für Österre-
ichische Geschichtsforschung 104 (1996) pp. 26–40
——, “Nomen et gens: Political and linguistic aspects of personal names between
the third and the eighth century—presenting an interdisciplinary project from a
historical perspective”, Strategies of Distinction. The Construction of Ethnic Communities,
300–800, ed. W. Pohl with H. Reimitz, The Transformation of the Roman World
2 (Leiden-Boston-Köln 1998) pp. 113–6
——, “Die Langobarden zwischen Pannonien und Italien”, Slovenija in sosednje de≥ele
med antiko in karolin“ko dobo. Zaèetki slovenske etnogeneze—Slowenien und die Nachbarländer
zwischen Antike und karolingischer Epoche. Anfänge der slowenischen Ethnogenese 1, ed.
R. Brato≥ (Ljubljana 2000) pp. 73–9
E. John, “The point of Woden”, Anglo-Saxon Studies in Archaeology and History 5 (1992)
pp. 127–34
A.H.M. Jones, “The constitutional position of Odoacer and Theoderic”, Journal of
Roman Studies 52 (1962) pp. 126–30
——, The Later Roman Empire 284 –602. A social, economic and administrative survey,
3 vols. (Oxford 1964)
——, The Decline of the Ancient World (London 1966)
M.E. Jones and P.J. Casey, “The Gallic Chronicle Restored: a Chronology for the
Anglo-Saxon Invasions and the End of Roman Britain”, Britannia 19 (1988) pp.
367–98
P.F. Jones, A Concordance to the Historia Ecclesiastica of Bede (Cambridge Mass. 1929)
S. Jones, The Archaeology of Ethnicity (London 1997)
M. de Jong, “Adding insult to injury: Julian of Toledo and his Historiae Wambae”,
The Visigoths. From the Migration Period to the Seventh Century. An Ethnographic Perspective,
ed. P.J. Heather (San Marino 1999) pp. 373–89
H.D. Kahl, “Die Baiern und ihre Nachbarn bis zum Tode des Herzogs Theodo
(717/18)”, Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
historische Klasse 179. Veröffentlichungen der Kommission für Frühmittelalter-
forschung 8 (Wien 1989) pp. 159–225
R. Kaiser, “Bistumsgründung und Kirchenorganisation im 8. Jahrhundert”, Der heilige
Willibald—Klosterbischof oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Wein-
furter, Eichstätter Studien NF 30 (Regensburg 1990) pp. 29–67
——, Das römische Erbe und das Merowingerreich, Enzyklopädie deutscher Geschichte
26 (2nd edn., München 1997)
——, Die Franken: Roms Erben und Wegbereiter Europas?, Historisches Seminar N.F. 10
(Idstein 1997)
G. Kampers, Personengeschichtliche Studien zum Westgotenreich in Spanien (Münster 1979)
Karl der Große. Lebenswerk und Nachleben, vol. 1: Persönlichkeit und Geschichte, ed.
H. Beumann (Düsseldorf 1965)
Karl Martell in seiner Zeit, ed. J. Jarnut, U. Nonn and M. Richter, Beihefte der Francia
37 (Sigmaringen 1994)
M. Kazanski, “À propos de quelques types de fibules ansées de l’époque des Grandes
Invasions trouvées en Gaule”, Archéologie Médiévale 14 (1984) pp. 7–27
——, “La diffusion de la mode danubienne en Gaule (fin du IVe siècle—début du
VIe siècle): essai d’interpretation historique”, Antiquités Nationales 21 (1989) pp.
59–73
  663

——, “Les barbares ‘orientaux’ dans l’armée romaine en Gaule”, Antiquités Nationales
29 (1997) pp. 201–17
H. Keller, “Strukturveränderungen in der westgermanischen Welt am Vorabend
der fränkischen Großreichbildung. Fragen, Suchbilder, Hypothesen”, Die Franken
und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich. Ergän-
zungsbände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-New
York 1998) pp. 580–607
Kelten, Germanen, Römer im Mittelgebirgsraum zwischen Luxemburg und Thüringen, ed.
A. Haffner and S. von Schnurbein, Kolloquien zur Vor- und Frühgeschichte 5
(Bonn 2000)
Der Keltenfürst von Hochdorf. Methoden und Ergebnisse der Landesarchäologie (Stuttgart 1985)
H. Kennedy, Muslim Spain and Portugal (London 1996)
J. Kent, “The end of Roman Britain: the literary and numismatic evidence reviewed”,
The End of Roman Britain, ed. P.J. Casey, British Archaeological Reports, British
series 71 (Oxford 1979) pp. 15–27
P.D. King, Law and Society in the Visigothic Kingdom, Cambridge Studies in Medieval
Life and Thought (Cambridge 1972)
——, “King Chindasvind and the First Territorial Law-code of the Visigothic
Kingdom”, Visigothic Spain: New Approaches, ed. E. James (Oxford 1980) pp. 131–57
Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. W. Pohl, The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997)
A. Kiss, “A Keszthely-kultura helye a pannoniai római kontinuitás kérdésében”,
Archeologiai Értesítõ 95 (1968) pp. 93–101
——, “Die Goldfunde des 5.–10. Jahrhunderts im Karpatenbecken. Angaben zu
den Vergleichsmöglichkeiten der schriftlichen und archäologischen Quellen”, Acta
Archaeologica Academiae Scientiarum Hungaricae 18 (1986) pp. 105–45
——, “Elõzetes jelentés (II.) A Kölked—Feketekapui avarkori települéles és temetõk
ásatasáról [Vorbericht (II.) über die Ausgrabung der awarenzeitlichen Siedlung
und Gräberfelder von Kölked-Feketekapu]”, Folia Archaeologia 39 (1988) pp. 173–94
G. Kiss, “Funde der Awarenzeit in Wiener Museen—1. Funde aus der Umgebung
von Keszthely”, Archaeologia Austriaca 68 (1984) pp. 61–201
——, “A késõi avar aranyozott övdíszek [Die spätawarischen vergoldeten Gürtel-
beschläge]”, Somogyi Múzeumok Közleményei 11 (1995) pp. 99–126 [German summary:
p. 106]
E. Klebel, “Langobarden, Bajuwaren, Slawen”, Mitteilungen der Anthropologischen Gesellschaft
in Wien 69 (1939) pp. 41–116 [repr. id., Probleme der bayerischen Verfassungsgeschichte,
Schriftenreihe zur bayerischen Landesgeschichte 57 (München 1957) pp. 1–89]
——, “Baierische Siedlungsgeschichte”, Zeitschrift für bayerische Landesgeschichte 16 (1951)
pp. 75–82
——, “Bayern und der fränkische Adel im 8. und 9. Jahrhundert”, Grundfragen der
alemannischen Geschichte, Vorträge und Forschungen 1 (Sigmaringen 1955) pp.
193–208
——, “Zur Geschichte des Herzogs Theodo”, Verhandlungen des Historischen Vereins für
die Oberpfalz und Regensburg 99 (1958) pp. 165–205 [repr. Zur Geschichte der Bayern,
ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 172–224]
J. Kleemann, “Zum Aufkommen der Körperbestattung in Niedersachsen”, Studien
zur Sachsenforschung 13, ed. H.-J. Häßler (Oldenburg 1999) pp. 253–62
W. Kleiber, “Mosella Romana. Hydronomie, Toponymie und Reliktwortdistribu-
tion”, Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed.
D. Geuenich, Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde
19 (Berlin-New York 1998) pp. 130–55
W. Kleiber and M. Pfister, Aspekte und Probleme der römisch-germanischen Kontinuität.
Sprachkontinuität an Mosel, Mittel- und Oberrhein sowie im Schwarzwald (Stuttgart 1992)
H. Kleinschmidt, “The Geuissae and Bede: On the Innovations of Bede’s Concept
of the Gens”, The Community, the Family and the Saint. Patterns of Power in
664 

Early Medieval Europe. Selected Proceedings of the International Medieval

Congress. University of Leeds, 4–7 July 1994, 10–13 July 1995, ed. J. Hill and
M. Swan, International Medieval Research 4 (Turnhout 1998) pp. 77–102
A. Koch, “Fremde Fibeln im Frankenreich. Ein Beitrag zur Frage nichtfränkischer
germanischer Ethnien in Nordgallien”, Archäologie des Frankenreiches—Neueste Forschungen.
Internationales Kolloquium anläßlich der Ausstellung “Die Franken—Les Francs” Berlin, 10.
bis 12. Oktober 1997, ed. W. Menghin, Acta Praehistorica et Archaeologica 30 (1998)
pp. 69–89
——, “Zur Problematik der sog. fränkischen Landnahme”, id., Bügelfibeln der Mero-
wingerzeit im westlichen Frankenreich, Römisch-Germanisches Zentralmuseum Mainz,
Monographien 41,2 (Mainz 1998) pp. 565–80
——, Bügelfibeln der Merowingerzeit im westlichen Frankenreich, Römisch-Germanisches
Zentralmuseum Mainz, Monographien 41,2 (Mainz 1998)
J.T. Koch, “When Was Welsh Literature First Written Down?”, Studia Celtica 20/21
(1986) pp. 43–66
——, “The Cynfeirdd poetry and the language of the sixth century”, Early Welsh
Poetry: Studies in the Book of Aneirin, ed. B.F. Roberts (Aberystwyth 1988) pp. 17–42
U. Koch, Die Grabfunde der Merowingerzeit aus dem Donautal um Regensburg, Germanische
Denkmäler der Völkerwanderungszeit A 10 (Berlin 1968)
——, “Das fränkische Gräberfeld von Herbolzheim, Kreis Heilbronn”, Fundberichte
aus Baden-Württemberg 7 (1982) pp. 387–474
——, “Stätten der Totenruhe—Grabformen und Bestattungssitten der Franken”,
Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben.
Katalog zur Ausstellung im Reiss-Museum Mannheim 2 (Mainz 1996) pp. 723–37
——, “Ethnische Vielfalt im Südwesten. Beobachtungen in merowingerzeitlichen
Gräberfeldern an Neckar und Donau”, Die Alamannen. Ausstellungskatalaog Stuttgart,
Zürich, Augsburg (Stuttgart 1997) pp. 219–32
——, “Besiegt, beraubt, vertrieben. Die Folgen der Niederlagen von 496/497 und
506”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart 1997) pp.
191–201
——, “Fernbeziehungen im Spiegel merowingerzeitlicher Grabfunde—Wer waren
die Kontaktpersonen?” Archäologisches Nachrichtenblatt 3 (1998) pp. 107–17
——, “Der Beginn fränkischer Besiedlung im Rhein-Neckar-Raum. Gräber von
Mannheim-Vogelstang und Mannheim-Straßenheim”, Mannheimer Geschichtsblätter
NF 7 (2000) pp. 57–106
H. Koller, “Die bairische Kirchenorganisation des 8. Jahrhunderts: Ansätze, Konzepte,
Verwirklichung”, Das Christentum im bairischen Raum von den Anfängen bis ins 11.
Jahrhundert, ed. E. Boshof and H. Wolff, Passauer historische Forschungen 8 (Köln-
Wien 1994) pp. 273–89
L. Kolmer, “Zur Kommendation und Absetzung Tassilos III.”, Zeitschrift für bay-
erische Landesgeschichte 43 (1980) pp. 291–327
——, “Die Inschriften aus dem Grab des Bischofs Gregorius und die Herkunft der
Baiern aus Armenien”, Ostbairische Grenzmarken 28 (1986) pp. 11–21
L. Költö, “Avar kori bronztárgyak röntgenemissziós analízise [X-ray emission analy-
sis of bronze objects from the Avar age]”, Somogyi Múzeumok Közleményei 5 (1982)
pp. 5–68 [English summary: p. 66]
G.G. König, “Wandalische Grabfunde des 5. und 6. Jahrhunderts”, Madrider Mitteilungen
22 (1981) pp. 299–360
Kontakte zwischen Iran, Byzanz und der Steppe im 6.–7. Jahrhundert, ed. C. Bálint, Varia
Archaeologic Hungarica 10 (Budapest 2000)
I. Kovrig, “Contribution au problème de l’occupation de la Hongrie par les Avars”,
Acta Archaeologica Hungaria Academiae Scientiarum Hungaricae 6 (1995) pp. 163–92
——, “Megjegyzések a Keszthely-kultura kérdéséhez”, Archeologiai Értesítõ 85 (1958)
pp. 66–72
  665

——, Das awarenzeitliche Gräberfeld von Alattyán, Archaeologia Hungarica 40 (Budapest

1963)
H. Krahwinkler, “Beiträge zu Namen und Geschichte der bayerischen Genealogiae”,
Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern 1, ed. W. Pohl and
H. Wolfram, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 201. Veröffentlichungen der Kommission für
Frühmittelalterforschung 12 (Wien 1990) pp. 217–34
L. Kraskovská, Slovansko-avarské pohrebisko pri Záhorskej Bystrici, Fontes archeologického
ústavu Slovenského národného múzea v Bratislava 1 (Bratislava 1972)
A. Kraus, “Zweiteilung des Herzogtums der Agilolfinger? Die Probe aufs Exempel”,
Blätter für deutsche Landesgeschichte 112 (1976) pp. 16–29
——, “Die Herkunft der Bayern. Zu Neuerscheinungen des letzten Jahrzehnts”,
Bayerisch-schwäbische Landesgeschichte an der Universität Augsburg 1975–1977, ed.
P. Fried, Veröffentlichungen der Schwäbischen Forschungsgemeinschaft bei der
Kommission für Bayerische Landesgeschichte, Reihe 7, Augsburger Beiträge zur
Landesgeschichte Bayerisch-Schwabens 1 (Sigmaringen 1979) pp. 27–46
——, “Der heilige Willibald von Eichstätt: Person, Zeit, Werk”, Der heilige Willibald—
Klosterbischof oder Bistumsgründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter,
Eichstätter Studien NF 30 (Regensburg 1990) pp. 9–28
S. Krautschick, “Die Familie der Könige in Spätantike und Frühmittelalter”, Das
Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989) pp. 109–42
K.-H. Krüger, Königsgrabkirchen der Franken, Angelsachsen und Langobarden bis zur Mitte
des 8. Jahrhunderts, Münstersche Mittelalter-Schriften 4 (München 1971)
H.-P. Kuhnen, “Zwischen Reichs- und Stadtgeschichte—Trier in Spätantike und
Frühmittelalter”, Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig
und seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim 1 (Mainz 1996)
pp. 138–44
M. Kulikowski, “Barbarians in Gaul, usurpers in Britain”, Britannia 31 (2000) pp.
325 ff.
——, “Nation vs. army: a necessary contrast?”, On Barbarian Identity—Critical Approaches
to Ethnogenesis Theory, ed. A. Gillett (Turnhout forthcoming)
M. Kunter and U. Wittwer-Backofen, “Die Franken—Anthropologische Bevölkerungs-
rekonstruktionen im östlichen Siedlungsgebiet”, Die Franken—Wegbereiter Europas.
Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim 2 (Mainz 1996) pp. 653–61
G.B. Ladner, “On Roman Attitudes toward Barbarians in Late Antiquity”, Viator 7
(1976) pp. 1–26
S. Ladstätter, Die materielle Kultur der Spätantike in den Ostalpen. Eine Fallstudie am Beispiel
der westlichen Doppelkirchenanlage auf dem Hemmaberg, Mitteilungen der Prähistorischen
Kommission der Österreichischen Akademie der Wissenschaften 35,1 (Wien 2000)
The largest cemetery from the Avar period in the Carpathian basin. Selection from the restored
material of the Avar cemetery at Zamárdi (Kaposvár 1998)
J. Laing, Warriors of the Dark Ages (Stroud 2000)
Der Landkreis Augsburg, vol. 2: Vor- und Frühgeschichte, Archäologie einer Landschaft, ed.
W. Pötzl and O. Schneider (Augsburg 1996)
A. Lane, “Trade, Gifts and Cultural Exchange in Dark Age Western Scotland”,
Scotland in Dark Age Europe, ed. B.E. Crawford (St Andrews 1994) pp. 103–15
Langobardia, ed. S. Gasparri and P. Cammarosano (Udine 1990)
R. Lapesa, Historia de la Lengua española (8th edn., Madrid 1985)
M. Lapidge, “Gildas’ education and the Latin culture of sub-Roman Britain”, Gildas:
New Approaches, ed. id. and D.N. Dumville (Woodbridge 1984) pp. 27–50
W. Laske, “Die Mönchung Herzog Tassilos III. und das Schicksal seiner Angehörigen”,
Die Anfänge des Klosters Kremsmünster, ed. S. Haider, Mitteilungen des Oberöster-
reichischen Landesarchivs, Ergänzungsband 2 (Linz 1978) pp. 189–97
666 

G. László, Études Archeologiques sur l´histoire de la Société des Avars, Archaeologia Hungarica
34 (Budapest 1955)
G. László and I. Rácz, Der Goldschatz von Nagyszentmiklós (Budapest 1983)
Latin and the Vernacular Languages in Early Medieval Britain, ed. N. Brooks (Leicester
1982)
R. Le Jan, Famille et pouvoir dans le monde franc (VIIe–IXe siècle) (Paris 1995) pp. 388–95
——, “Convents, violence, and competition for power in seventh-century Francia”,
Topographies of Power in the Early Middle Ages, ed. M. de Jong, F. Theuws and
C. van Rhijn, The Transformation of the Roman World 6 (Leiden 2001) pp.
243–69
Ch. Lécrivain, “Etudes sur le Bas-Empire”, Mélanges d’Archéologie et d’Histoire 10 (1890)
pp. 253–83
K.H. Lenz, Siedlungen der Römischen Kaiserzeit auf der Aldenhovener Platte (Köln-Bonn 1999)
W. Liebeschuetz, Barbarians and Bishops (Oxford 1990)
——, “Cities, taxation and the accommodation of the barbarians: the theories of
Durliat and Goffart”, Kingdoms of the Empire. The Integration of Barbarians in Late
Antiquity, ed. W. Pohl, The Transformation of the Roman World 1 (Leiden-New
York-Köln 1997) pp. 135–51
——, “Citizen status and law”, Strategies of Distinction. The Construction of Ethnic Communi-
ties, 300–800, ed. W. Pohl with H. Reimitz, The Transformation of the Roman
World 2 (Leiden-New York-Köln 1998) pp. 131–52
The Long Eighth Century, ed. I.L. Hansen and C. Wickham, The Transformation of
the Roman World 11 (Leiden-Boston-Köln 2000)
J.E. López Pereira, Estudio Crítico sobre la Crónica Mozárabe de 754 (Zaragoza 1980)
S.T. Loseby, “Gregory’s cities: Urban functions in sixth-century Gaul”, Franks and
Alamanni in the Merovingian Period. An Ethnographic Perspective, ed. I.N. Wood, Studies
in Historical Archaeoethnology 3 (San Marino 1998) pp. 239–70; 270–84
F. Lot, “La formation de la nation française”, Revue des deux mondes 15 mai et 1er
juin (1950) pp. 256–78; 418–35
F. Lotter, “Zur Rolle der Donausueben in der Völkerwanderungszeit”, Mitteilungen
des Instituts für Österreichische Geschichtsforschung 76 (1968) pp. 275–98
——, “Die germanischen Stammesverbände im Umkreis des Ostalpen-Mitteldonau-
Raumes nach der literarischen Überlieferung zum Zeitalter Severins”, Die Bayern
und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz, Denkschriften der Öster-
reichischen Akademie der Wissenschaften, philosophisch-historische Klasse 179.
Veröffentlichungen der Kommission für Frühmittelalterforschung 8 (Wien 1989)
pp. 29–59
G. Loud, “The Gens Normannorum—myth or reality?”, Proceedings of the Battle Conference
of Anglo-Norman Studies IV, ed. R. Allen Brown (Woodbridge 1982) pp. 104–16
H. Löwe, “Die Herkunft der Bajuwaren”, Zeitschrift für bayerische Landesgeschichte 15
(1949) pp. 5–67
——, “Von Theoderich dem Großen zu Karl dem Großen. Das Werden des
Abendlands im Geschichtsbild des frühen Mittelalters”, Deutsches Archiv 9 (1952)
pp. 353–401
——, “Bonifatius und die bayerisch-fränkische Spannung”, Jahrbuch für fränkische
Landesforschung 15 (1955) pp. 85–127 [repr. Zur Geschichte der Bayern, ed. K. Bosl,
Wege der Forschung 60 (Darmstadt 1965) pp. 264–328]
S. Lucy, The Early Anglo-Saxon Cemeteries of Eastern Yorkshire. An Analysis and Re-inter-
pretation, British Archaeological Reports, British series 272 (Oxford 1998)
——, The Anglo-Saxon Way of Death (Stroud 2000)
N. Luhmann, Macht (Stuttgart 1975)
A. Lund, Die ersten Germanen: Ethnizität und Ethnogenese (Heidelberg 1998)
H.G. Lunt, “Common Slavic, Proto-Slavic, Pan-Slavic; what are we talking about?”,
International Journal of Slavic Linguistics and Poetics 41 (1997) pp. 7–67
  667

M. Lupoi, Alle radici del mondo giuridico europeo (Rome 1994) [English transl. The Origins
of the European Legal Order (Cambridge 2000)]
W. Lütkenhaus, Constantius III., Studien zu seiner Tätigkeit und Stellung im Westreich 411–21
(Bonn 1998)
C.H. Lynch, Saint Braulio, Bishop of Saragossa (631–51). His Life and Writings (Washington
DC 1938) pp. 137–40
——, San Braulio (Madrid 1950)
M. Maas, “Ethnicity, Orthodoxy and Community in Salvian of Marseilles”, Fifth-
century Gaul: a crisis of identity?, ed. J.F. Drinkwater and H. Elton (Cambridge 1992)
pp. 275–84
M. Mackensen, Die spätantiken Sigillata- und Lampentöpfereien von El Mahrine (Nordtunesien):
Studien zur nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts (Munich 1993)
É. Magnou-Nortier, “Remarques sur la genèse du Pactus legis Salicae et sur le priv-
ilège d’immunité (IVe–VIIe siècles)”, Clovis. Histoire et mémoire, vol. 1: Le baptême de
Clovis, l’événement, ed. M. Rouche (Paris 1997) pp. 495–538
J.L. Maier, L’épiscopat de l’Afrique romaine, vandale et Byzantine (Rome 1973)
The Making of Kingdoms, ed. T. Dickinson and D. Griffiths, Anglo-Saxon Studies in
Archaeology and History 10 (Oxford 1999)
M.A. Makki, “Egipto y los origines de la historiografía árabe-española”, Revista del
Instituto Egipcio de Estudios Islámicos 5 (1957) pp. 157–248 [English transl. The
Formation of al-Andalus, vol. 2: Language, Religion, Culture and Sciences, ed. M. Fierro
and J. Samsó (Aldershot 1998) pp. 173–234]
A. Mandouze, Prosopographie de l’Afrique chrétienne (Paris 1982)
Maritime Celts, Frisians and Saxons, ed. S. McGrail, Council for British Archaeology
Research Report 71 (London 1990)
H.I. Marrou, “Le valeur historique de Victor de Vita”, Les Cahiers de Tunisie 15
(1967) pp. 205–8
B.I. Marschak, Silberschätze des Orients. Metallkunst des 3.-13. Jahrhunderts und ihre Kontinuität
(Leipzig 1986)
A. Marshall and G. Marshall, “Differentiation, change and continuity in Anglo-
Saxom buildings”, The Archaeological Journal 150 (1994) 366–402
R. Marti, “Das Grab eines wohlhabenden Alamannen in Altdorf UR, Pfarrkirche
St. Martin”, Jahrbuch der Schweizerischen Gesellschaft für Ur- und Frühgeschichte 78 (1995)
pp. 83–130
M. Martin, Das fränkische Gräberfeld von Basel-Bernering (Basel 1976)
——, Das spätrömisch-frühmittelalterliche Gräberfeld von Kaiseraugst, Kt. Aargau, 2 vols.,
Basler Beiträge zur Ur- und Frühgeschichte 5 (Derendingen 1976/91)
——, “Awarische und germanische Funde in Männergräbern von Linz-Zizlau und
Környe. Ein Beitrag zur Chronologie der Awarenzeit”, A Wosinsky Mór Múzeum
Évkönyve 15 (1990) pp. 65–90
——, “Tradition und Wandel der fibelgeschmückten frühmittelalterlichen Frauenk-
leidung”, Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 38,2 (1991)
pp. 629–80
——, “Fibel und Fibeltracht. § 41. W-germ. F. und F.-Tracht. A. Diskontinuität
der F.-Tracht zur RGZ und MZ”, Reallexikon der Germanischen Altertumskunde 8 (2nd
edn., 1994) pp. 549–56
——, “Die Gräberfelder von Straubing-Bajuwarenstraße und Straßkirchen—zwei
erstrangige Quellen zur Geschichte der frühen Baiern im Straubinger Land”,
Frühe Baiern im Straubinger Land (Straubing c. 1995) pp. 17–39
——, “Zu den tauschierten Gürtelgarnituren und Gürtelteilen der Männergräber
von Kölked-Feketekapu A”, Das awarenzeitlich gepidische Gräberfeld von Kölked—Fekete-
kapu A, ed. A. Kiss and F. Daim, Monographien zur Frühgeschichte und
Mittelalterarchäologie 2. Studien zur Archäologie der Awaren 5 (Innsbruck 1996)
pp. 345–61
668 

J. Martindale, The Prosopography of the Later Roman Empire 2 (Cambridge 1980)

G. Martínez-Gros, L’idéologie omeyyade: La construction de la légitimité du califat de Cordue
(Xe–XIe siècles) (Madrid 1992)
R. Mathisen, “Clovis, Anastase et Grégoire de Tours: consul, patrice et roi”, Clovis.
Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris
1997) pp. 395–407
J.F. Matthews, “Olympiodorus of Thebes and the history of the west (A.D. 407–425)”,
Journal of Roman Studies 60 (1970) pp. 79–97
——, Western Aristocracies and Imperial Court, A.D. 364–425 (Oxford 1975)
K. Matzner, “Völkerwanderung und Awarenzeit”, Mödling. Landschaft, Kultur, Wirtschaft
(Mödling 1975) pp. 67–88
N. Mavrodinov, Le trésor protobulgare de Nagyszentmiklós, Archaeologia Hungarica 29
(Budapest 1943)
E. Mayer, “Übersehene Quellen zur bayerischen Geschichte des 6.–8. Jahrhunderts”,
Zeitschrift für bayerische Landesgeschichte 4 (1931) pp. 1–36
G. Mayr, “Zur Todeszeit des Hl. Emmeram und zur frühen Geschichte des Klosters
Herrenchiemsee. Bemerkungen zur Schenkung des Ortlaip in Helfendorf ”, Zeitschrift
für bayerische Landesgeschichte 34 (1971) pp. 358–73
——, “Frühes Christentum in Baiern”, Die Bajuwaren. Von Severin bis Tassilo 488–788,
ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 281–6
——, “Neuerliche Anmerkungen zur Todeszeit des Heiligen Emmeram und zur
Kirchenpolitik Herzog Theodos”, Typen der Ethnogenese unter besonderer Berücksichtigung
der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 201. Veröffentlichun-
gen der Kommission für Frühmittelalterforschung 12 (Wien 1990) pp. 199–215
M. McCormick, Eternal Victory. Triumphal rulership in late antiquity, Byzantium, and the
early medieval West (Cambridge 1986)
——, “Clovis at Tours, Byzantine Public Ritual and the Origins of Medieval Ruler
Symbolism”, Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna
1989) pp. 155–80
D. McManus, “A Chronology of Latin loan-words in early Irish”, Ériu 34 (1983)
pp. 21–71
——, “Linguarum Diversitas: Latin and the Vernaculars in Early Medieval Britain”,
Peritia 3 (1984) pp. 151–88
A. Meaney, A Gazetteer of Early Anglo-Saxon Burial Sites (London 1964)
Medieval Christian Perceptions of Islam: A Book of Essays, ed. J.V. Tolan (New York-
London 1996)
Medieval Europeans. Studies in ethnic identity and national perspectives in medieval Europe, ed.
A.P. Smyth (Basingstoke 1998)
The Medieval Landscape of Wessex, ed. M. Aston and C. Lewis (Oxford 1994)
Mélanges d’histoire anciennes offerts à William Seston (Paris 1974)
J. Mellado, “Veladas discrepancias y pugna por el poder en el III Concilio de
Toledo”, Boletín de la Real Academia de Córdoba 141 (2001) (forthcoming)
R. Menéndez Pidal, Los godos y la epopeya española. “Chansons de geste” y baladas nórdi-
cas (2nd edn., Madrid 1969)
W. Menghin, “Schwerter des Goldgriffspathahorizonts im Museum für Vor- und
Frühgeschichte, Berlin”, Acta Praehistorica et Archaeologica 26/27 (1994/95) pp. 140–91
M. Menke, “Archeologia longobarda tra la bassa Elba e l’Ungheria”, Langobardia,
ed. S. Gasparri and P. Cammarosano (Udine 1990) pp. 35–106
——, “Zu den Fibeln der Awarenzeit aus Keszthely”, A Wosinsky Mór Múzeum
Évkönyve 15 (1990) pp. 187–214
——, “150 Jahre Forschungsgeschichte zu den Anfängen des Baiernstammes”, Typen
der Ethnogenese unter besonderer Berücksichtigung der Bayern 2, ed. H. Friesinger and F. Daim,
Denkschriften der Österreichischen Akademie der Wissenschaften, philosophisch-
  669

historische Klasse 204. Veröffentlichungen der Kommission für Frühmittelal-

terforschung 13 (Wien 1990) pp. 123–220
B. Merta, “Politische Theorie in den Königsurkunden Pippins”, Mitteilungen des Instituts
für Österreichische Geschichtsforschung 100 (1992) pp. 117–31
H. Messmer, Hispania-Idee und Gotenmythos (Zürich 1960)
D. Messner, “Salzburgs Romanen”, Virgil von Salzburg. Missionar und Gelehrter, ed.
H. Dopsch and R. Juffinger (Salzburg 1985) pp. 103–11
Method and Theory in Historical Archaeology—Papers of the Medieval Brugge 1997 Conference,
ed. G. de Boe and F. Verhaeghe (Zellik 1997)
Migrations and Invasion in Archaeological Explanation, ed. J. Chapman and H. Hamerow,
British Archaeological Reports, International series 664 (1997)
R. Miket, “A re-statement of evidence for Bernician Anglo-Saxon burials”, Anglo-
Saxon Cemeteries 1979: the fourth Anglo-Saxon Symposium at Oxford, ed. P.A. Rahtz,
T. Dickinson and L. Watt (Oxford 1980) pp. 289–305
G. Mildenberger, “Terra Nigra aus Nordhessen”, Fundberichte aus Hessen 12 (1972)
pp. 104–26
F. Millar, The Emperor in the Roman World, 31 B.C.–A.D. 337 (London 1977)
M. Miller, “Bede’s use of Gildas”, English Historical Review 90 (1975) pp. 241–61
——, “The last British entry in the ‘Gallic Chronicles’”, Britannia 9 (1978) pp. 315–8
W.I. Miller, Bloodtaking and Peacemaking: Feud, Law and Society in Saga Iceland (Chicago
1990)
M. Millett, The Romanization of Britain (Cambridge 1990)
Millionen Jahre Geschichte. Fundort Nordrhein-Westfalen, ed. H.G. Horn, H. Hellenkemper,
G. Isenberg and H. Koschik (Mainz 2000)
V. Miloj‘iÆ, “Zur Frage der Zeitstellung des Oratoriums von Mühlthal an der Isar”,
Bayerische Vorgeschichtsblätter 28 (1963) pp. 117–38
A. Miron, “Die babylonische Verwirrung. Überlegungen zur Terminologie der
Spätlatène-Chronologie”, Studien zur Archäologie der Kelten, Römer und Germanen in
Mittel- und Westeuropa. Alfred Haffner zum 60. Geburtstag gewidmet, ed. A. Müller-Karpe
et al., Internationale Archäologie, Studia honoraria 4 (Rahden Westf. 1998) pp.
429–38
——, “Zentrale Fragestellung—Das Grab als kulturhistorische Quelle”, Kelten, Germanen,
Römer im Mittelgebirgsraum zwischen Luxemburg und Thüringen, ed. A. Haffner and
S. von Schnurbein, Kolloquien zur Vor- und Frühgeschichte 5 (Bonn 2000)
pp. 257–61
H. Mitscha-Märheim, “Die Herkunft der Baiern”, Mitteilungen der Anthropologischen
Gesellschaft in Wien 80 (1950) pp. 213–44
Y. Modéran, “La chronologie de la vie de Saint Fulgence de Ruspe”, Mélanges de
l’École Française de Rome 105,1 (1993) pp. 135–88
Mödling—Goldene Stiege, ed. F. Daim, K. Matzner and H. Schwammenhöfer (forth-
coming)
Mödling. Landschaft, Kultur, Wirtschaft (Mödling 1975)
H. Moisl, “Anglo-Saxon genealogies and Germanic oral tradition”, Journal of Medieval
History 7 (1981) pp. 215–48
A. Momigliano, “Cassiodorus and the Italian Culture of His Time”, Proceedings of
the British Academy 4 (1955) pp. 207–45
F. Monfrin, “La conversion du roi et des siens”, Clovis. Histoire et mémoire, vol. 1: Le
baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp. 289–320
Monks and Nuns, Saints and Outcasts, ed. S. Farmer and B.H. Rosenwein (Ithaca
2000)
J. Moorhead, “Italian loyalties during Justinian’s Gothic War”, Byzantion 53 (1983)
pp. 575–96
——, “Libertas and Nomen Romanum in Ostrogothic Italy”, Latomus 46 (1988) pp.
161–8
670 

——, Theoderic in Italy (Oxford 1993)

——, The Roman Empire Divided, 400–700 (Harlow-London 2001)
C. Morrisson, “Les origines du monnayage vandale”, Actes du 8 ème Congrès international
de numismatique 1973 (Paris-Basel 1976) pp. 461–72
J. Moysidou, To Byzantio kai oi boreioi geitones tou ton 10o aiona (Athens 1995)
S. Muhlberger, “The Gallic Chronicle for 452 and its Authority for British Events”,
Britannia 14 (1983) pp. 23–33
——, “Heroic kings and unruly generals: the ‘Copenhagen’ continuation of Prosper
reconsidered”, Florilegium 6 (1984) pp. 50–70
——, “The Copenhagen continuation of Prosper: a translation”, Florilegium 6 (1984)
pp. 71–95
——, The Fifth-century Chroniclers: Prosper, Hydatius and the Gallic Chronicle of 452 (Leeds
1990)
M. Müller, “Die bayerische ‘Stammessage’ in der Geschichtsschreibung des Mittelalters.
Eine Untersuchung zur mittelalterlichen Frühgeschichtsforschung in Bayern”,
Zeitschrift für bayerische Landesgeschichte 40 (1977) pp. 341–71
R. Müller, “Die spätrömische Festung Valcum am Plattensee”, Germanen, Hunnen und
Awaren. Schätze der Völkerwanderungszeit, ed. W. Menghin (Nürnberg 1988) pp. 270–3
——, “Vorbericht über die Freilegung des Grabes eines hohen Militärs aus der
Mittelawarenzeit in Gyenesdiás”, Communicationes Archaeologicae Hungariae (1989) pp.
141–64
——, Der Häuptling von Gyenes. Awarenzeit in der Umgebung von Keszthely, Kataloge des
Niederösterreichischen Landesmuseums N.F. 256 (Wien 1990)
——, “Das Gräberfeld von Gyenesdiás”, Reitervölker aus dem Osten. Hunnen + Awaren,
ed. F. Daim (Halbturn 1996) pp. 411–6
——, “Die Festung ‘Castellum’, Pannonia Inferior”, Reitervölker aus dem Osten. Hunnen
+ Awaren, ed. F. Daim (Halbturn 1996) pp. 91–5
——, “Die Keszthely-Kultur”, Reitervölker aus dem Osten. Hunnen + Awaren, ed.
F. Daim (Halbturn 1996) pp. 265–74
M. Müller-Wille, “Königtum und Adel im Spiegel der Grabfunde”, Die Franken—
Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur
Ausstellung im Reiss-Museum Mannheim 1 (Mainz 1996) pp. 206–21
M. Müller-Wille, Zwei religiöse Welten: Bestattungen der fränkischen Könige Childerich und
Chlodwig, Akademie der Wissenschaften und der Literatur, Abhandlungen der
Geistes- und Sozialwissenschaftlichen Klasse, Jahrgang 1998, 1 (Stuttgart 1998)
A.C. Murray, Germanic Kinship Structure. Studies in Law and Society in Antiquity and the
early Middle Ages, Pontifical Institute of Mediaeval Studies, Studies and Texts 65
(Toronto 1983)
——, “The Position of the Grafio in the Constitutional History of Merovingian
Gaul”, Speculum 61 (1986) pp. 787–805
——, “From Roman to Frankish Gaul: ‘Centenarii’ and ‘Centenae’ in the Administra-
tion of the Merovingian Kingdom”, Traditio 44 (1988) pp. 59–100
——, “Fredegar, Merovech, and ‘Sacral Kingship’”, After Rome’s Fall. Narrators and
Sources of Early Medieval History. Essays presented to Walter Goffart, ed. id. (Toronto-
Buffalo-London 1998) pp. 121–51
J.N.L. Myres, “The Angles, the Saxons and the Jutes”, Proceedings of the British Academy
56 (1970) pp. 1–32
K.B. Nagy, “Székkutas-Kápolnadûlo avar temeto néhány 9. Szászadi síregyüttese
[Einige Grabkomplexe aus dem 9. Jahrhundert im awarischen Gräberfeld bei
Székkutas-Kápolnadûlõ]”, Az Alföld a 9. szászadban, ed. G. Lõrinczy (Szeged 1993)
pp. 151–69
M. Nagy, “Awarenzeitliche Gräberfelder im Stadtgebiet von Budapest”, Monumenta
Avarorum Archaeologica 2 (Budapest 1998)
Nascita dell’Europa ed Europa Carolingia: Un’equazione da verificare, Settimane di studio
del centro italiano di studi sull’alto medioevo 27 (Spoleto 1981)
  671

V.E. Nash-Williams, Early Christian Monuments of Wales (Cardiff 1950)

U. Näsman, “The ethnogenesis of the Danes and the making of a Danish king-
dom”, The Making of Kingdoms, ed. T. Dickinson and D. Griffiths, Anglo-Saxon
Studies in Archaeology and History 10 (Oxford 1999) pp. 1–10
Neuburg an der Donau. Archäologie rund um den Stadtberg, ed. K.H. Rieder and A. Tillmann
(Neuburg 1993)
G. Neumann, “Baias”, Reallexikon der Germanischen Altertumskunde 1 (2nd edn., 1973)
p. 600
La Neustrie. Les pays au nord de la Loire de 650 à 850, 2 vols., ed. H. Atsma, Beihefte
der Francia 16 (Sigmaringen 1989)
La Neustrie. Les pays au nord de la Loire de Dagobert à Charles le Chauve (VIIe–IXe siècle),
ed. P. Périn and L.-C. Feffer (Rouen 1985)
E. Neusz, “Sprachraumbildung am Niederrhein und die Franken. Anmerkungen zu
Verfahren der Sprachgeschichtsschreibung”, Die Franken und die Alemannen bis zur
“Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände zum Reallexikon
der Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 156–92
S. Newton, The Origins of Beowulf and Pre-Viking East Anglia (Woodbridge 1993)
Nibelungenlied und Klage. Sage und Geschichte, Struktur und Gattung. Passauer Nibelungen-
Gespräche 1985, ed. F.P. Knapp (Heidelberg 1987)
C. Nolte, Conversio und Christianitas. Frauen in der Christianisierung vom 5. bis 8. Jahrhundert,
Monographien zur Geschichte des Mittelalters 41 (Stuttgart 1995)
Nomen et gens. Zur historischen Aussagekraft frühmittelalterlicher Personennamen, ed. D. Geuenich,
W. Haubrichs and J. Jarnut, Ergänzungsbände zum Reallexikon der Germanischen
Altertumskunde 16 (Berlin-New York 1997)
U. Nonn, “Merowingische Testamente: Studien zum Fortleben einer römischen
Urkundenform im Frankenreich”, Archiv für Diplomatik 18 (1972) pp. 1–129
——, “Zur Verwaltungsorganisation in der nördlichen Galloromania”, Die Franken
und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungs-
bände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-New York
1998) pp. 82–94
E. Norden, Die germanische Urgeschichte in Tacitus’ Germania (Leipzig-Berlin 1929)
G. von Olberg, Die Bezeichungen für soziale Stände, Schichten und Gruppen in den Leges
Barbarorum, Arbeiten zur Frühmittelalterforschung 11 (Münster 1991)
L. Oliver, The Language of the Early English Laws, Harvard Ph.D. 1995
On Barbarian Identity—Critical Approaches to Ethnogenesis Theory, ed. A. Gillett (Turnhout
forthcoming)
The Origins of Anglo-Saxon Kingdoms, ed. S. Bassett (Leicester 1989)
Origins of Central Europe, ed. P. Urba…czyk (Warsaw 1997)
J. Orlandis and D. Ramos-Lissón, Historia de los Concilios de la España romana y visigoda
(Pamplona 1986)
P. van Ossel, Établissements ruraux de l’Antiquité tardive dans le nord de la Gaule, Gallia,
Suppl. 51 (Paris 1992)
G. Owen-Crocker, Dress in Anglo-Saxon England (Manchester 1986)
O.J. Padel, “A New Study of the Gododdin”, Cambrian Medieval Celtic Studies 35
(1998) pp. 45–56
P. de Paepe and L. van Impe, “Historical Context and Provenancing of Late Roman
Hand-Made Pottery from Belgium, the Netherlands and Germany. First Report”,
Archaeologie in Vlanderen 1 (1991) pp. 145–80
B. Päffgen and S. Ristow, “Die Römerstadt Köln zur Merowingerzeit”, Die Franken—
Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur
Ausstellung im Reiss-Museum Mannheim 1 (Mainz 1996) pp. 145–59
D. Parsons, “British *Caratîcos, Old English Cerdic”, Cambrian Medieval Celtic Studies
33 (1997) pp. 1–8
A. Pásztór and T. Vida, “Eine frühbyzantinische Bronzekanne aus dem awaren-
zeitlichen Gräberfeld von Budakalász”, Kontakte zwischen Iran, Byzanz und der Steppe
672 

im 6.–7. Jahrhundert, ed. C. Bálint, Varia Archaeologica Hungarica 10 (Budapest

2000) pp. 303–11
Les Pays du Nord et Byzance, ed. R. Zeitler, Acta Universitatis Upsaliensis N.S. 19
(Uppsala 1981)
L.V. Pekarskaja and D. Kidd, Der Silberschatz von Martynovka (Ukraine) aus dem 6. und
7. Jahrhundert, Monographien zur Frühgeschichte und Mittelalterarchäologie 1
(Innsbruck 1994)
People and Places in Northern Europe 500–1600. Essays in Honour of Peter Hayes Sawyer,
ed. I.N. Wood and N. Lund (Woodbridge 1991)
Á.S. Peremi, “Lesencetowmaj-piroskereszt keszthely-kultúrás temető fülbevalói”, A
Vesztprém Megyei Múzeumok Közleményei 21 (2000) pp. 41–75
P. Périn, “A propos de publications étrangères récentes concernant le peuplement
en Gaule à l’époque mérovingienne: La ‘question franque’”, Francia 8 (1980) pp.
537–52
——, “Paris, merowingische Metropole”, Die Franken—Wegbereiter Europas. Vor 1500
Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim
1 (Mainz 1996) pp. 121–8
——, “La Progression des Francs en Gaule du Nord au Ve siècle. Histoire et
archéologie”, Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97),
ed. D. Geuenich, Ergänzungsbände zum Reallexikon der Germanischen Alter-
tumskunde 19 (Berlin-New York 1998) pp. 59–81
P. Périn and L.-C. Feffer, Les Francs, 2 vols. (Paris 1987)
P. Périn and M. Kazanski, “Das Grab Childerichs I.”, Die Franken—Wegbereiter Europas.
Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim 1 (Mainz 1996) pp. 173–82
C. Perrat and A. Audin, “Alcimi Ecdicii viennensis episcopi homilia dicta in ded-
icatione superioris basilicae”, Studi in onore di Aristide Calderini e Roberto Paribeni 2
(Milan 1957) pp. 433–51
F. Petri, Die fränkische Landnahme und die Entstehung der germanisch-romanischen Sprachgrenze
in der interdisziplinären Diskussion, Erträge der Forschung 70 (Darmstadt 1977)
Peuples du Moyen Âge. Problèmes d’identification. Séminaires Sociétés, Idéologies et Croyances au
Moyen Âge, ed. C. Carozzi and H. Taviani-Carozzi, Publications de l’Université
de Provence (Aix-en-Provence 1996)
M. Pfister, “Entstehung, Verbreitung und Charakteristik des Zentral- und Ostalpen-
Romanischen vor dem 12. Jahrhundert”, Frühmittelalterliche Ethnogenese im Alpenraum,
ed. H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 49–95
La Picardie, berceau de la France. Clovis et les derniers Romains. 1500ème anniversaire de la
bataille de Soissons 486–1986 (Amiens 1986)
P. Piel and P. Kremer, Hispanogotisches Namenbuch (Heidelberg 1976)
A. Piganiol, L’Empire chrétien (Paris 1947)
——, “Les causes de la ruine de l’empire romain”, id., L’Empire chrétien (Paris 1947)
pp. 411–22 [repr. “Die Ursachen des Untergangs des Römischen Reiches”, Der
Untergang des Römischen Reiches, ed. K. Christ, Wege der Forschung 269 (Darmstadt
1970) pp. 270–85]
Places of Power, ed. M. de Jong, F. Theuws and C. van Rhijn, The Transformation
of the Roman World 6 (Leiden-Boston 2001)
E. Pohl, “Der Neuburger Stadtberg und sein Umfeld am Übergang von der Antike
zum Mittelalter (4. bis 7. Jhdt.) aus archäologischer Sicht”, Neuburg an der Donau.
Archäologie rund um den Stadtberg, ed. K.H. Rieder and A. Tillmann (Neuburg 1993)
pp. 109–32
W. Pohl, “Die Gepiden und die Gentes an der mittleren Donau nach dem Zerfall
des Attilareiches”, Die Völker an der mittleren und unteren Donau im fünften und sechsten
Jahrhundert, ed. H. Wolfram and F. Daim, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 145 (Vienna 1980)
pp. 239–305
  673

——, Die Awaren. Ein Steppenvolk in Mitteleuropa, 567–822 n. Chr. (München 1988)
——, “Tradition, Ethnogenese und literarische Gestaltung: eine Zwischenbilanz”,
Ethnogenese und Überlieferung. Angewandte Methoden der Frühmittelalterforschung, ed.
K. Brunner and B. Merta, Veröffentlichungen des Instituts für Österreichische
Geschichtsforschung 31 (Wien-München 1994) pp. 9–26
——, “The Barbarian Successor States”, The Transformation of the Roman World A.D.
400–900, ed. L. Webster and M. Brown (London 1997) pp. 33–47
——, “The Empire and the Lombards: treaties and negotiations in the sixth cen-
tury”, Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, ed. id.,
The Transformation of the Roman World 1 (Leiden-New York-Köln 1997) pp.
75–134
——, “The role of steppe peoples in eastern and central Europe in the first millen-
nium A.D.”, Origins of Central Europe, ed. P. Urba…czyk (Warsaw 1997) pp. 65–78
——, “Ethnic names and identities in the British Isles: a comparative perspective”,
The Anglo-Saxons from the Migration Period to the Eighth Century: An Ethnographic Perspective,
ed. J. Hines (Woodbridge 1997) pp. 7–40
——, “Introduction: Strategies of distinction”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. id. with H. Reimitz, The Transformation of
the Roman World 2 (Leiden-Boston-Köln 1998) pp. 1–15
——, “Telling the difference: Signs of ethnic identity”, Strategies of Distinction. The
Construction of Ethnic Communities, 300–800, ed. id. with H. Reimitz, The Trans-
formation of the Roman World 2 (Leiden-Boston-Köln 1998) pp. 17–69
——, “Gentilismus”, Reallexikon der Germanischen Altertumskunde 11 (2nd edn., 1998)
pp. 91–101
——, “Herrschaft”, Reallexikon der Germanischen Altertumskunde 14 (2nd edn., 1999) pp.
443–57
——, “Franken und Sachsen: die Bedeutung ethnischer Prozesse im 7. und 8.
Jahrhundert”, 799—Kunst und Kultur der Karolingerzeit. Karl der Große und Papst Leo
III. in Paderborn. Beiträge zum Katalog der Ausstellung Paderborn 1999, ed. C. Stiegemann
and M. Wemhoff (Mainz 1999) pp. 233–6
——, “Zur Bedeutung ethnischer Unterscheidungen in der frühen Karolingerzeit”,
Studien zur Sachsenforschung 12, ed. H.-J. Häßler (Oldenburg 1999) pp. 193–298
——, Die Germanen, Enzyklopädie deutscher Geschichte 57 (München 2000)
——, Le origini etniche dell’Europa. Barbari e Romani tra antichità e medioevo (Roma 2000)
——, Die Völkerwanderung. Eroberung und Integration (Stuttgart-Berlin-Köln 2002)
Popoli delle steppe: Unni, Avari, Magiari, Settimane di studio del centro italiano di studi
sull’alto medioevo 35 (Spoleto 1988)
I. Popovic’, Zlatni avarski pojas iz okoline Sirmijuma [Golden avarian belt from the vicinity
of Sirmium], National Museum, Belgrade, Monographies 10. Archaeological Institute,
Belgrade, Monographies 32 (Beograd 1997)
Power and the Construction of Communities, ed. W. Pohl and H. Reimitz (forthcoming)
D. Powlesland, “Early Anglo-Saxon settlements, structures, form and layout”, The
Anglo-Saxons from the Migration Period to the Eighth Century: An Ethnographic Perspective,
ed. J. Hines (Woodbridge 1997) pp. 101–24
F. Prinz, “Herzog und Adel im agilulfingischen Bayern. Herzogsgut und Konsens-
s-chenkungen vor 788”, Zeitschrift für bayerische Landesgeschichte 25 (1962) pp. 283–311
[repr. Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt
1965) pp. 225–63]
——, “Zur Herrschaftsstruktur Bayerns und Alemanniens im 8. Jahrhundert”, Blätter
für deutsche Landesgeschichte 102 (1966) pp. 11–27
——, “Salzburg zwischen Antike und Mittelalter”, Frühmittelalterliche Studien 5 (1971)
pp. 10–36
——, “Fragen der Kontinuität zwischen Antike und Mittelalter am Beispiel Bayerns”,
Zeitschrift für bayerische Landesgeschichte 37 (1974) pp. 699–727
674 

——, “Augsburg im Frankenreich”, Die Ausgrabungen in St. Ulrich und Afra in Augsburg
1961–1968, ed. J. Werner, Münchener Beiträge zur Vor- und Frühgeschichte 23
(München 1977) pp. 375–98
——, “Nochmals zur ‘Zweiteilung des Herzogtums der Agilolfinger’”, Blätter für
deutsche Landesgeschichte 113 (1977) pp. 19–32
——, Frühes Mönchtum im Frankenreich. Kultur und Gesellschaft in Gallien, den Rheinlanden
und Bayern am Beispiel der monastischen Entwicklung (4. bis 8. Jahrhundert) (2nd edn.,
München 1988)
Proceedings of the Battle Conference of Anglo-Norman Studies IV, ed. R. Allen Brown
(Woodbridge 1982)
Proceedings of the Second International Congress of Celtic Studies (Cardiff 1966)
Proceedings of the Thirteenth International Congress of Onomastic Sciences, 2 vols., ed.
K. Rymut (Wroclaw 1981)
D. Quast, Die merowingerzeitlichen Grabfunde aus Gültlingen (Stadt Wildberg, Kreis Calw),
Forschungen und Berichte zur Vor- und Frühgeschichte in Baden-Württemberg
52 (Stuttgart 1993)
——, “Vom Einzelgrab zum Friedhof. Beginn der Reihengräbersitte im 5. Jahr-
hundert”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich, Augsburg (Stuttgart 1997)
pp. 171–90
Z. Rácz, Awarische Goldschmiedegräber, Monographien zur Frühgeschichte und Mitte-
lalterarchäologie (forthcoming)
P.A. Rahtz, “Celtic Society in Somerset, A.D. 400–700”, Bulletin of the Board of Celtic
Studies 30 (1982) pp. 176–200
——, Cadbury-Congresbury 1968–73: a Late/Post-Roman Hilltop Settlement in Somerset
(Oxford 1992)
P.A. Rahtz and L. Watt, “The end of Roman temples in the west of Britain”, The
End of Roman Britain, ed. P.J. Casey, British Archaeological Reports, British series
71 (Oxford 1979) pp. 183–210
J.D. Randers-Pehrson, Barbarians and Romans. The Birth Struggle of Europe, A.D. 400–700
(London-Canberra 1983)
Ranking, Resource and Exchange. Aspects of the Archaeology of Early European Society, ed.
C. Renfrew and S. Shennan (Cambridge 1982)
Regensburg—Kelheim—Straubing, vol. 1, ed. S. Rieckhoff-Pauli and W. Torbrügge,
Führer zu archäologischen Denkmälern 5 (Stuttgart 1984)
Regional Approaches to Mortuary Analysis, ed. L. Anderson Beck (New York 1995)
Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz (Vienna 1989)
Ch. Reichmann, “Frühe Franken in Germanien”, Die Franken—Wegbereiter Europas.
Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim 1 (Mainz 1996) pp. 55–65
——, “Krefeld [2352/003]”, Bonner Jahrbücher 197 (1997) pp. 314–6
I. Reiffenstein, “Stammesbildung und Sprachgeschichte. Das Beispiel der bairischen
Ethnogenese”, Althochdeutsch, vol. 2: Wörter und Namen. Forschungsgeschichte, ed.
R. Bergmann, H. Tiefenbach and L. Voetz, Germanische Bibliothek NF, Reihe
3, Untersuchungen (Heidelberg 1987) pp. 1333–41
K. Reindel, “Die Bistumsorganisation im Alpen-Donau-Raum in der Spätantike und
im Frühmittelalter”, Mitteilungen des Instituts für Österreichische Geschichtsforschung 72
(1964) pp. 277–310
——, “Das Zeitalter der Agilolfinger (bis 788)”, Handbuch der bayerischen Geschichte 1,
ed. M. Spindler (2nd edn., München 1981) pp. 101–245
——, “Die Bajuwaren. Quellen, Hypothesen, Tatsachen”, Deutsches Archiv 37 (1981)
pp. 451–73
——, “Regensburg als Sitz der Herrscher bis zum 10. Jahrhundert”, Regensburg—
Kelheim—Straubing 1, ed. S. Rieckhoff-Pauli and W. Torbrügge, Führer zu archäol-
ogischen Denkmälern 5 (Stuttgart 1984) pp. 243–54
  675

——, “Salzburg und die Agilolfinger”, Virgil von Salzburg. Missionar und Gelehrter, ed.
H. Dopsch and R. Juffinger (Salzburg 1985) pp. 66–74
——, “Herkunft und Stammesbildung der Bajuwaren nach den schriftlichen Quellen”,
Die Bajuwaren. Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch
(Rosenheim-Salzburg 1988) pp. 56–60
Reitervölker aus dem Osten. Hunnen + Awaren, ed. F. Daim (Halbturn 1996)
Religion and National Identity, ed. S. Mews, Studies in Church History 18 (Oxford
1982)
C. Renfrew, “Post collapse resurgence: culture process in the Dark Ages”, Ranking,
Resource and Exchange. Aspects of the Archaeology of Early European Society, ed. id. and
S. Shennan (Cambridge 1982) pp. 113–5
A. Rettner, “Thüringisches und Fränkisches in Zeuzleben”, Archäologie des Frankenreiches—
Neueste Forschungen. Internationales Kolloquium anläßlich der Ausstellung “Die Franken—Les
Francs” Berlin, 10. bis 12. Oktober 1997, ed. W. Menghin, Acta Praehistorica et
Archaeologica 30 (1998) pp. 113–25
T. Reuter, Germany in the Early Middle Ages 800–1050 (Harlow 1991)
M. Reydellet, “Les intentions idéologiques et politiques dans la Chronique d’Isidore
de Seville”, Mélanges d’Archéologie et d’Histoire 82 (1970) pp. 363–400
——, La royauté dans la littérature latine de Sidoine Apollinaire à Isidore de Séville (Roma
1981)
S. Reynolds, “Medieval origines gentium and the community of the realm”, History
(1983) pp. 375–90
——, Kingdoms and Communities in Western Europe, 900–1300 (Cambridge 1984)
G. Ripoll, La necrópolis visigoda de El Carpio del Tajo (Toledo) (Madrid 1985)
——, La occupación visigoda a través de sus necrópolis (Hispania), Thesis Microfiche 912
(Barcelona 1991)
——, “Acerca de la supuesta frontera entre el regnum Visigothorum y la Hispania bizan-
tina”, Pyrenae 27 (1996) pp. 251–67
——, “The arrival of the Visigoths in Hispania: Population problems and the process
of acculturation, Strategies of Distinction. The Construction of Ethnic Communities, 300–800,
ed. W. Pohl with H. Reimitz, The Transformation of the Roman World 2 (Leiden-
New York-Köln 1998) pp.153–87
——, Toréutica de la Bética (siglos VI–VII d.C.) (Barcelona 1998)
——, “Symbolic Life and Signs of Identity in Visigothic Times”, The Visigoths. From
the Migration Period to the Seventh Century. An Ethnographic Perspective, ed. P.J. Heather
(San Marino 1999) pp. 403–31
G. Ripoll and J. Arce, “The Transformation and End of Roman villae in the West
(IV-VIIth centuries): Problems and Perspectives, Towns and their Territories between
Late Antiquity and the Early Middle Ages, ed. G.P. Brogliolo, N. Gauthier and
N. Christie, The Transformation of the Roman World 9 (Leiden 2000) pp. 63–114
G. Ripoll and I. Velázquez, La Hispania visigoda. Del rey Ataúlfo a Don Rodrigo (Madrid
1995)
Rituals of Power. From Late Antiquity to the Early Middle Ages, ed. F. Theuws and J.L.
Nelson, The Transformation of the Roman World 8 (Leiden-Boston-Köln 2000)
Roma fra oriente e occidente, Settimane di studio del centro italiano di studi sull’alto
medioevo 49 (Spoleto 2002)
Romanitas-Christianitas. Untersuchungen zur Geschichte und Literatur der römischen Kaiserzeit,
ed. G. Wirth (Berlin 1982)
A. Róna-Tas, “Die Inschrift des Nadelbehälters von Szarvas (Ungarn)”, Ural-Altaische
Jahrbücher N.F. 9 (1990) pp. 1–30
H. Rosenfeld, “Die Völkernamen Baiern und Böhmen, die althochdeutsche Lautver-
schiebung und W. Mayerthalers These ‘Baiern = Salzburger Rätoromanen’.
Völkernamen, Völkerwanderung, Stammesgenese und die Namen Baiern, Bayern,
Bajuwaren”, Althochdeutsch, vol. 2: Wörter und Namen. Forschungsgeschichte, ed.
676 

R. Bergmann, H. Tiefenbach and L. Voetz, Germanische Bibliothek NF, Reihe

3, Untersuchungen (Heidelberg 1987) pp. 1305–32
B.H. Rosenwein, “Perennial prayer at Agaune”, Monks and Nuns, Saints and Outcasts,
ed. S. Farmer and B.H. Rosenwein (Ithaca 2000) pp. 37–56
——, “One site, many meanings: Saint-Maurice d’Agaune as a place of power in
the early Middle Ages”, Topographies of Power in the Early Middle Ages, ed. M. de
Jong, F. Theuws and C. van Rhijn, The Transformation of the Roman World
6 (Leiden 2001) pp. 271–90
G. Rosner, “Keramikherstellung und Handel im Karpaten-Becken in der frühen
Awarenzeit”, A Wosinsky Mór Múzeum Évkönyve 15 (1989) pp. 125–33
M. Rouche, “Die Bedeutung der Taufe Chlodwigs”, Die Franken—Wegbereiter Europas.
Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum
Mannheim 1 (Mainz 1996) pp. 192–9
A. Rovelli, “Some considerations on the coinage of Lombard and Carolingian Italy”,
The Long Eighth Century, ed. I.L. Hansen and C. Wickham, The Transformation
of the Roman World 11 (Leiden-Boston-Köln 2000) pp. 195–223
J. Rowland, Early Welsh Saga Poetry: a Study and Edition of the Englynion (Woodbridge
1990)
I. Runde, “Die Franken und Alemannen vor 500. Ein chronologischer Überblick”,
Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich,
Ergänzungsbände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-
New York 1998) pp. 656–82
Runische Schriftkultur in kontinental-skandinavischer und angelsächsischer Wechselbeziehung, ed.
K. Düwel (Berlin-New York 1994)
P. Russell, “Recent Work on British Latin”, Cambridge Medieval Celtic Studies 9 (1985)
pp. 19–29
Rutas, ciudades y moneda en Hispania, ed. P. Centeno, M.P.García y Bellido and
G. Mora, Anejos de Archivo Español de Arqueología 20 (Madrid 1999)
E. Rynne, “An important seventh-century sword from Co. Roscommon”, Journal of
the Old Athlone Society 2 (1985) 148–51
W. Sage, “Frühes Christentum und Kirchen aus der Zeit des Überganges”, Geschichte
der Stadt Augsburg von der Römerzeit bis zur Gegenwart, ed. G. Gottlieb, W. Baer,
J. Becker, J. Bellot, K. Filser, P. Fried, W. Reinhard and B. Schimmelpfennig
(Stuttgart 1984) pp. 100–12
P. Salway, Roman Britain (Oxford 1981)
A. Sandberger, “Römisches Straßensystem und bairische Siedlung im Osten von
München”, Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner,
Schriftenreihe zur bayerischen Landesgeschichte 62 (München 1962) pp. 287–92
E. Savage, A Gateway to Hell, A Gateway to Paradise: the North African Response to the
Arab Conquest (Princeton NJ 1997)
P.H. Sawyer, From Roman Britain to Norman England (London 1978)
V. Saxer, “Les rites du baptême de Clovis dans le cadre de la pratique paléochré-
tienne”, Clovis. Histoire et mémoire, vol. 1: Le baptême de Clovis, l’événement, ed.
M. Rouche (Paris 1997) pp. 229–41
Ch. Schäfer, Der weströmische Senat als Träger antiker Kontinuität unter den Ostgotenkönigen
(490–540 n.Chr.) (St. Katharinen 1991)
K. Schäferdiek, “Gab es eine gotisch-arianische Mission im süddeutschen Raum?”,
Zeitschrift für bayerische Landesgeschichte 45 (1982) pp. 239–57
P. Schaffner, “Britain’s Iudices”, Gildas: New Approaches, ed. M. Lapidge and D.N.
Dumville (Woodbridge 1984) pp. 151–5
G. Scheibelreiter, Der Bischof in merowingischer Zeit, Veröffentlichungen des Instituts
für Österreichische Geschichtsforschung 27 (Wien-Köln-Graz 1983)
——, “Vester est populus meus. Byzantinische Reichsideologie und germanisches
Selbstverständnis”, Das Reich und die Barbaren, ed. E. Chrysos and A. Schwarcz
(Vienna 1989) pp. 203–20
  677

W. Schlesinger, “Über germanisches Heerkönigtum”, Das Königtum. Seine geistigen und

rechtlichen Grundlagen, ed. T. Mayer, Vorträge und Forschungen 3 (Lindau-Konstanz
1956, 4th edn. 1973) pp. 105–41
M. Schmauder, “Die Oberschichtgräber und Verwahrfunde Südosteuropas und das
Childerichgrab von Tournai. Anmerkungen zu den spätantiken Randkulturen”,
Archäologie des Frankenreiches—Neueste Forschungen. Internationales Kolloquium anläßlich der
Ausstellung “Die Franken—Les Francs” Berlin, 10. bis 12. Oktober 1997, ed. W. Menghin,
Acta Praehistorica et Archaeologica 30 (1998) pp. 55–68
——, “Der Verwahrfund von Lengerich, Ldkr. Emsland: Spiegel innerrömischer
Kämpfe?”, Die Kunde 50 (1999) pp. 91–118
——, “Überlegungen zur östlichen Grenze des karolingischen Reiches unter Karl
dem Großen”, Grenze und Differenz im frühen Mittelalter, ed. W. Pohl and H. Reimitz,
Forschungen zur Geschichte des Mittelalters 1 (Wien 2000) pp. 79–81
——, “Verzögerte Landnahme? Die Dacia Traiana und die sogenannten decumates
agri”, Integration und Herrschaft. Ethnische Identitäten und soziale Organisation im Frühmittelalter,
ed. M. Diesenberger and W. Pohl (forthcoming)
A. Schmid, “Bayern und Italien vom 7. bis zum 10. Jahrhundert”, Die transalpinen
Verbindungen der Bayern, Alemannen und Franken bis zum 10. Jahrhundert, ed. H. Beumann
and W. Schröder, Nationes 6 (Sigmaringen 1987) pp. 51–91
——, “Regensburg zur Agilolfingerzeit”, Die Bajuwaren. Von Severin bis Tassilo 488–788,
ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 136–40
P. Schmid, “Feddersen Wierde”, Reallexikon der Germanischen Altertumskunde 8 (2nd edn.,
1994) pp. 249–66
——, “König—Herzog—Bischof. Regensburg und seine Pfalzen”, Deutsche Königspfalzen.
Beiträge zu ihrer historischen und archäologischen Erforschung, vol. 4: Pfalzen—Reichsgut—
Königshöfe, ed. L. Fenske, Veröffentlichungen des Max-Planck-Instituts für Geschichte
11,4 (Göttingen 1996) pp. 53–83
W.P. Schmid, “Vidivarii”, Sprach- und Kulturkontakte im Polnischen, Specimina Philologicae
Slawicae, Supplementband 23 (München 1987) pp. 349–58
H. Schmidinger, “Das Papsttum und die bayerische Kirche—Bonifatius als Gegenspieler
Virgils”, Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and R. Juffinger
(Salzburg 1985) pp. 92–101
L. Schmidt, Geschichte der deutschen Stämme bis zum Ausgang der Völkerwanderung: Die
Ostgermanen (2nd edn., Munich 1933)
——, “Zum Ursprung der Baiern”, Zeitschrift für bayerische Landesgeschichte 10 (1937)
pp. 12–8
——, Die Westgermanen (München 1940; repr. 1970)
——, Geschichte der Vandalen (2nd edn., Munich 1942)
R. Schmidt-Wiegand, “‘Gens Francorum inclita’. Zu Gestalt und Inhalt des län-
geren Prologes der Lex Salica”, Festschrift Adolf Hofmeister zum 70. Geburtstag, ed.
U. Scheil (Halle 1955) pp. 233–50
——, Stammesrecht und Volkssprache. Ausgewählte Aufsätze zu den Leges barbarorum (Weinheim
1991)
——, “Rechtsvorstellungen bei den Franken und Alemannen vor 500”, Die Franken
und die Alemannen bis zur “Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergän-
zungsbände zum Reallexikon der Germanischen Altertumskunde 19 (Berlin-New
York 1998) pp. 545–57
R. Schneider, Königswahl und Königserhebung im Frühmittelalter. Untersuchungen zur
Herrschaftsnachfolge bei den Langobarden und Merowingern, Monographien zur Geschichte
des Mittelalters 3 (Stuttgart 1972)
——, “Fränkische Alpenpolitik”, Die transalpinen Verbindungen der Bayern, Alemannen und
Franken bis zum 10. Jahrhundert, ed. H. Beumann and W. Schröder, Nationes 6
(Sigmaringen 1987) pp. 23–49
——, Das Frankenreich, Oldenbourg Grundriß der Geschichte 5 (3rd edn., München
1995)
678 

——, “König und Königsherrschaft bei den Franken”, Von Sacerdotium und Regnum.
Geistliche und weltliche Gewalt im frühen und hohen Mittelalter. Festschrift für Egon Boshof
zum 65. Geburtstag, ed. F.-R. Erkens and H. Wolff (Köln-Weimar-Wien 2002) pp.
11–26
J. Schnetz, “‘Baias’ und der Baiernname”, Zeitschrift für bayerische Landesgeschichte 16
(1951) pp. 1–19
M.D. Schön, Der Thron aus der Marsch. Ausgrabungen an der Fallward bei Wremen im
Landkreis Cuxhaven, Begleitheft zu Ausstellungen/Museum Burg Bederkesa, Landkreis
Cuxhaven 1 (Bremerhaven 1995)
——, “Gräber und Siedlungen bei Ottendorf-Westerwörden, Landkreis Cuxhaven”,
Probleme der Küstenforschung im südlichen Nordseegebiet 26 (1999) pp. 123–208
K. Schreiner, “Adel oder Oberschicht? Bemerkungen zur soziologischen Schichtung
der fränkischen Gesellschaft im 6. Jahrhundert”, Vierteljahresschrift für Sozial- und
Wirtschaftsgeschichte 68 (1981) pp. 225–31
P. Schrijver, “The Celtic Contribution to the Development of the North Sea Germa-
nic Vowel System”, North-Western European language evolution 35 (1999) pp. 3–47
M. Schulze, “Einflüsse byzantinischer Prunkgewänder auf die fränkische Frauentracht”,
Archäologisches Korrespondenzblatt 6 (1976) pp. 149–61
H. Schutz, The Germanic Realms in Pre-Carolingian Central Europe, 400–750 (New York
2000)
H. Schwammenhöfer, “Ein awarenzeitlicher Bestattungsplatz in Mödling bei Wien”,
Antike Welt 7,2 (1976) pp. 11–8
E. Schwarz, “Baiern und Naristen in Burgund”, Südostdeutsche Forschungen 2 (1937)
pp. 379–82
——, “Die bairische Landnahme um Regensburg im Spiegel der Völker- und
Ortsnamen”, Beiträge zur Namenforschung 1 (1949/50) pp. 51–71
——, “Herkunft und Einwanderungszeit der Baiern”, Südost-Forschungen 12 (1953)
pp. 21–47
——, “Neues und Altes zur Geschichte der Naristen”, Jahrbuch für fränkische Landes-
forschung 22 (1962) pp. 281–9
——, “Das Ende der Völkerwanderungszeit in Böhmen und die Herkunftsfrage der
Baiern”, Bohemia 8 (1967) pp. 23–58
——, “Die Naristenfrage in namenkundlicher Sicht”, Zeitschrift für bayerische Landesgeschichte
32 (1969) pp. 397–476
——, “Baiern und Walchen”, Zeitschrift für bayerische Landesgeschichte 33 (1970) pp.
857–958
——, “Ermunduren—Thüringer, Böhmen—Baiern”, Et multum et multa. Beiträge zur
Literatur, Geschichte und Kultur der Jagd. Festgabe für Kurt Lindner zum 27. November
1971, ed. S. Schwenk, G. Tilander and C.A. Willemsen (Berlin-New York 1971)
pp. 341–9
K. Schwarz, “Regensburg während des ersten Jahrtausends im Spiegel der Ausgra-
bungen in Niedermünster”, Jahresbericht der Bayerischen Bodendenkmalpflege 13/14
(1972/73) pp. 20–98
Scotland in Dark Age Britain, ed. B.E. Crawford (Aberdeen 1996)
Scotland in Dark Age Europe, ed. B.E. Crawford (St Andrews 1994)
C. Scull, “Archaeology, early Anglo-Saxon society and the origins of kingdoms”,
Anglo-Saxon Studies in Archaeology and History 6 (1993) pp. 65–82
——, “Approaches to material culture and social dynamics of the migration period
of eastern England”, Europe Between Late Antiquity and the Middle Ages, ed. J. Bintliff
and H. Hamerow, British Archaeological Reports, International series 617 (Oxford
1995) pp. 71–83
——, “Social archaeology and Anglo-Saxon kingdom origins”, The Making of Kingdoms,
ed. T. Dickinson and D. Griffiths, Anglo-Saxon Studies in Archaeology and
History 10 (Oxford 1999) pp. 7–24
  679

Sedes regiae (ann. 400–800), ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona
2000)
H. Sedlmayr, “Die politische Bedeutung des Virgildomes”, Mitteilungen der Gesellschaft
für Salzburger Landeskunde 115 (1975) pp. 145–60
E. Seebold, “Was ist jütisch? Was ist kentish?” Britain 400–600: Language and History,
ed. A. Bammesberger and A. Wollmann (Heidelberg 1990) pp. 335–52
C. Seillier, “Les tombes de transition du cimetière germanique de Vron (Somme)”,
Jahrbuch des Römisch-Germanischen Zentralmuseums Mainz 36 (1989) pp. 599–634
——, “Céramique de type anglo-saxon du cimetière de Waben (Pas-de-Calais)”,
Antiquités Nationales 21 (1989) pp. 83–9
H.R. Sennhauser, “Die Salzburger Dombauten im Rahmen der frühmittelalterlichen
Baukunst Europas”, Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and
R. Juffinger (Salzburg 1985) p. 326
C. Servatius, “‘Per ordinationem principis ordinetur’. Zum Modus der Bischofs-
ernennung im Edikt Chlothars II. vom Jahre 614”, Zeitschrift für Kirchengeschichte
84 (1973) pp. 1–29
The Settlement of Disputes in early medieval Europe, ed. W. Davies and P. Fouracre
(Cambridge 1986)
B. Shaw, “Bandits in the Roman Empire”, Past and Present 105 (1984) pp. 3–52
J. Shepard, “Slavs and Bulgars”, New Cambridge Medieval History, vol. 2: 700–900,
ed. R. McKitterick (Cambridge 1995) pp. 228–48
J. Shephard, “The social identity of the individual in isolated barrows and barrow
cemeteries in Anglo-Saxon England”, Space, Hierarchy and Society, ed. B. Burnham
and J. Kingsbury, British Archaeological Reports, International series 59 (Oxford
1979) pp. 47–79
Siedlung, Sprache und Bevölkerungsstruktur im Frankenreich, ed. F. Petri, Wege der Forschungen
49 (Darmstadt 1973)
F. Siegmund, “Kleidung und Bewaffnung der Männer im östlichen Frankenreich”,
Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben.
Katalog zur Ausstellung im Reiss-Museum Mannheim 1 (Mainz 1996) pp. 691–706
——, Alemannen und Franken, Ergänzungsbände zum Reallexikon der Germanischen
Altertumskunde 23 (Berlin-New York 2000)
H. Siems, “Lex Baiuvariorum”, Handwörterbuch zur deutschen Rechtsgeschichte 2 (1978)
pp. 1887–901
P. Sims-Williams, “Gildas and the Anglo-Saxons”, Cambridge Medieval Celtic Studies 6
(1983) pp. 1–30
——, “The settlement of England in Bede and the Chronicle”, Anglo-Saxon England
12 (1983) pp. 1–41
——, “Gildas and vernacular poetry”, Gildas: New Approaches, ed. M. Lapidge and
D.N. Dumville (Woodbridge 1984) pp. 169–90
——, “Dating the Transition to Neo-Brittonic”, Britain 400–600: Language and History,
ed. A. Bammesberger and A. Wollmann (Heidelberg 1990) pp. 217–61
——, “The Emergence of Old Welsh, Cornish and Breton orthography, 600–800:
the Evidence of Archaic Old Welsh”, Bulletin of the Board of Celtic Studies 38 (1991)
pp. 20–86
——, “The Death of Urien”, Cambrian Medieval Celtic Studies 32 (1996) pp. 25–56
K. Sisam, “Anglo-Saxon royal genealogies”, Proceedings of the British Academy 39 (1953)
pp. 287–348
H. Sivan, “The appropriation of Roman law in barbarian hands: ‘Roman-barbar-
ian’ marriage in Visigothic Gaul and Spain”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transforma-
tion of the Roman World 2 (Leiden-New York-Köln 1998) pp. 189–203
The Sixth Century. Production, Distribution and Demand, ed. R. Hodges and W. Bowden,
The Transformation of the Roman World 3 (Leiden 1998)
680 

W.F. Skene, The Four Ancient Books of Wales containing the Cymric poems attributed to the
bards of the sixth century, vol. 1 (Edinburgh 1868)
Slovenija in sosednje de≥ele med antiko in karolin“ko dobo. Zaèetki slovenske etnogeneze—Slowenien
und die Nachbarländer zwischen Antike und karolingischer Epoche. Anfänge der slowenischen
Ethnogenese, vol. 1, ed. R. Brato≥ (Ljubljana 2000)
A.D. Smith, The Ethnic Origins of Nations (Oxford 1986)
C. Smith, “Vulgar Latin in Roman Britain: epigraphic and other evidence”, Aufstieg
und Niedergang der römischen Welt II 29,2, ed. H. Temporini and W. Haase (Berlin
1983) pp. 893–948
C.A. Snyder, An Age of Tyrants: Britain and the Britons, A.D. 400–600 (Stroud 1998)
A. Søby Christensen, “Cassiodorus and the making of a history of the Goths”,
Analecta Romana 26 (1999) pp. 173–7
Social Identity in Early Medieval Britain, ed. B. Frazer and A.J. Tyrell (London 2000)
La sociedad visigoda y su entorno histórico, ed. J. Faci (Barcelona 1992)
J. Sofer, Lateinisches und Romanisches aus den Etymologiae des Isidorus von Sevilla (Göttingen
1930)
P. Somogyi, Byzantinische Fundmünzen der Awarenzeit, Monographien zur Frühgeschichte
und Mittelalterarchäologie 5. Studien zur Archäologie der Awaren 8 (forthcom-
ing)
P. Sørensen, “Jutes in Kent? Considerations of the problem of ethnicity in south-
ern Scandinavia and Kent in the migration period”, Method and Theory in Historical
Archaeology—Papers of the Medieval Brugge 1997 Conference, ed. G. de Boe and
F. Verhaeghe (Zellik 1997) pp. 165–73
Á. Sós and Á. Salamon, Cemeteries of the Early Middle Ages (6th-9th c.) at Pókaszepetk
(Budapest 1995)
Space, Hierarchy and Society, ed. B. Burnham and J. Kingsbury, British Archaeological
Reports, International series 59 (Oxford 1979)
Spaces of the Living and Dead: an Archaeological Dialogue, ed. C.E. Karkov, K.M. Wickham-
Crowley and B.K. Young, American Early Medieval Studies 3 (Oxford 1999)
Speculum Historiale. Geschichte im Spiegel von Geschichtsschreibung und Geschichtsdeutung, ed.
C. Bauer, L. Boehm and M. Müller (Freiburg-München 1965)
G.R. Spohn, “Armenien und Herzog Naimes. Zur bayerischen Stammessage im
Mittelalter und bei Peter Harer”, Zeitschrift für bayerische Landesgeschichte 34 (1971)
pp. 185–210
Sprach- und Kulturkontakte im Polnischen, Specimina Philologicae Slawicae, Supplementband
23 (München 1987)
M. Springer, “Salier und Salisches Recht—Beobachtungen zu den Wörtern Salii
und Salicus”, Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und
seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim 1 (Mainz 1996) pp.
485–7
——, “Gab es ein Volk der Salier?”, Nomen et gens. Zur historischen Aussagekraft frühmittel-
alterlicher Personennamen, ed. D. Geuenich, W. Haubrichs and J. Jarnut, Ergänzungs-
bände zum Reallexikon der Germanischen Altertumskunde 16 (Berlin-New York
1997) pp. 58–83
——, “Riparii—Ribuarier—Rheinfranken nebst einigen Bemerkungen zum Geographen
von Ravenna”, Die Franken und die Alemannen bis zur “Schlacht bei Zülpich” (496/97),
ed. D. Geuenich, Ergänzungsbände zum Reallexikon der Germanischen Altertums-
kunde 19 (Berlin-New York 1998) pp. 201–69
——, “Geschichtsbilder, Urteile und Vorurteile. Franken und Sachsen in den
Vorstellungen unserer Zeit und in der Vergangenheit”, 799—Kunst und Kultur der
Karolingerzeit. Karl der Große und Papst Leo III. in Paderborn. Beiträge zum Katalog der
Ausstellung Paderborn 1999, ed. C. Stiegemann and M. Wemhoff (Mainz 1999) pp.
224–32
  681

T. Springer, “Germanenfunde der Völkerwanderungszeit in Nordbayern. Bemerkungen

zur Keramik vom Typ Friedenhain-PÏre“t’ovice”, Archäologisches Korrespondenzblatt 15
(1985) pp. 235–43
St Augustine and the Conversion of England, ed. R. Gameson (Stroud 1999)
F. Staab, “Les royaumes francs au Ve siècle”, Clovis. Histoire et mémoire, vol. 1: Le
baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp. 539–66
Die Stadt in Oberitalien und in den nordwestlichen Provinzen des Römischen Reiches. Deutsch-
Italienisches Kolloquium im italienischen Kulturinstitut Köln, ed. W. Eck and H. Galsterer,
Kölner Forschungen 4 (Mainz 1991)
H. Steuer, Frühgeschichtliche Sozialstrukturen in Mitteleuropa, Abhandlungen der Akademie
der Wissenschaften Göttingen, philologisch-historische Klasse 3,128 (Göttingen
1982)
——, “Höhensiedlungen des 4. und 5. Jahrhunderts in Südwestdeutschland”, Freiburger
Forschungen zum ersten Jahrtausend in Südwestdeutschland 1 (Sigmaringen 1990) pp.
139–205
——, “Archäologie und germanische Sozialgeschichte. Forschungstendenzen in den
1990er Jahren”, Runische Schriftkultur in kontinental-skandinavischer und angelsächsischer
Wechselbeziehung, ed. K. Düwel (Berlin-New York 1994) pp. 10–55
——, “Herrschaft von der Höhe. Vom mobilen Söldnertrupp zur Residenz auf
repräsentativen Bergkuppen”, Die Alamannen. Ausstellungskatalaog Stuttgart, Zürich,
Augsburg (Stuttgart 1997) pp. 149–62
C.E. Stevens, “Marcus, Gratian and Constantine”, Athenaeum 35 (1957) pp. 316–47
J. Stevenson, “The beginning of literacy in Ireland”, Proceedings of the Royal Irish
Academy (C) 89 (1989) pp. 127–65
N. Stoodley, The Spindle and the Spear: A Critical Enquiry into the Construction and Meaning
of Gender in the Early Anglo-Saxon Inhumation Burial Rite, British Archaeological Reports,
British series 288 (Oxford 1999)
Storia di Roma, vol. 3,2: L’età tardoantica. I luoghi e le culture, ed. A. Momigliano (Turin
1993)
Storia d’Italia, vol. 1: Longobardi e Bizantini, ed. P. Delogu, A. Guillou and G. Ortalli
(Torino 1980)
La storiografia altomedievale, Settimane di studio del centro italiano di studi sull’alto
medioevo 17 (Spoleto 1970)
W. Störmer, “Fernstraße und Kloster. Zur Verkehrs- und Herrschaftsstruktur des
westlichen Altbayern im frühen Mittelalter”, Zeitschrift für bayerische Landesgeschichte
29 (1966) pp. 299–343
——, Früher Adel. Studien zur politischen Führungsschicht im fränkisch-deutschen Reich vom 8.
bis 11. Jahrhundert, 2 vols., Monographien zur Geschichte des Mittelalters 6 (Stuttgart
1973)
——, “Die Agilolfinger im politischen Kräftefeld vom 6. bis zum 8. Jahrhundert”,
Baiernzeit in Oberösterreich. Das Land zwischen Inn und Enns vom Ausgang der Antike bis
zum Ende des 8. Jahrhunderts, Kataloge des Oberösterreichischen Landesmuseums
97 (Linz 1977) pp. 1–12
——, “Nibelungentradition als Hausüberlieferung in frühmittelalterlichen Adelsfamilien?
Beobachtungen zu Nibelungennamen im 8./9. Jahrhundert vornehmlich in Bayern”,
Nibelungenlied und Klage. Sage und Geschichte, Struktur und Gattung. Passauer Nibelungen-
Gespräche 1985, ed. F.P. Knapp (Heidelberg 1987) pp. 1–20
——, “Beobachtungen zu Aussagen und Intentionen der bayerischen Stammes-
‘Sage’ des 11./12. Jahrhunderts. Fiktionen—Sage—‘Geschichtsklitterung’”, Fälschungen
im Mittelalter. Internationaler Kongreß der Monumenta Germaniae Historica, vol. 1: Kongreßdaten
und Festvorträge. Literatur und Fälschung, Schriften der MGH 33,1 (Stuttgart 1988)
pp. 451–70
——, “Das Herzogsgeschlecht der Agilolfinger”, Die Bajuwaren. Von Severin bis Tassilo
682 

488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp.

141–52
——, “Bayerisch-ostfränkische Beziehungen vom 7. bis zum frühen 9. Jahrhundert”,
Die Bayern und ihre Nachbarn 1, ed. H. Wolfram and A. Schwarcz, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
179. Veröffentlichungen der Kommission für Frühmittelalterforschung 8 (Wien
1989) pp. 227–52
——, “Die bayerische Herzogskirche”, Der heilige Willibald—Klosterbischof oder Bistums-
gründer, ed. H. Dickerhof, E. Reiter and S. Weinfurter, Eichstätter Studien NF
30 (Regensburg 1990) pp. 115–42
——, “Zum Prozeß sozialer Differenzierung bei den Bayern von der Lex Baiuvariorum
bis zur Synode von Dingolfing”, Typen der Ethnogenese unter besonderer Berücksichtigung
der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften der Österreichischen
Akademie der Wissenschaften, philosophisch-historische Klasse 201. Veröffentlichun-
gen der Kommission für Frühmittelalterforschung 12 (Wien 1990) pp. 155–70
W. Störmer and G. Mayr, “Herzog und Adel”, Die Bajuwaren. Von Severin bis Tassilo
488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp.
153–9
Strategies of Distinction. The Construction of Ethnic Communities, 300–800, ed. W. Pohl
with H. Reimitz, The Transformation of the Roman World 2 (Leiden-New York-
Köln 1998)
K. Strobel, Die Galater 1 (Berlin 1996)
K.F. Stroheker, Eurich, König der Westgoten (Stuttgart 1937)
Studi in onore di Aristide Calderini e Roberto Paribeni, 3 vols. (Milan 1957)
Studien zur Archäologie der Kelten, Römer und Germanen in Mittel- und Westeuropa. Alfred
Haffner zum 60. Geburtstag gewidmet, ed. A. Müller-Karpe et al., Internationale
Archäologie, Studia honoraria 4 (Rahden Westf. 1998)
Studien zur Archäologie des Ostseeraumes. Festschrift für Michael Müller-Wille (Neumünster
1999)
Studien zur Sachsenforschung 2, ed. H.-J. Häßler (Hildesheim 1980)
Studien zur Sachsenforschung 7, ed. H.-J. Häßler (Hildesheim 1991)
Studien zur Sachsenforschung 11, ed. H.-J. Häßler (Oldenburg 1998)
Studien zur Sachsenforschung 12, ed. H.-J. Häßler (Oldenburg 1999)
Studien zur Sachsenforschung 13, ed. H.-J. Häßler (Oldenburg 1999)
Studies presented to R.H.C. Davis, ed. H. Mayr-Harting and R.I. Moore (London 1985)
J. Sturm, “Romanische Personennamen in den Freisinger Traditionen”, Zeitschrift
für bayerische Landesgeschichte 18 (1955) pp. 61–80
F. Stutz, Les objets mérovingiens de type septentrional dans le Sud de la Gaule, Mémoire de
Maîtrise à l’Université Paris 1 (Paris 1993)
——, Les objets mérovingiens de type septentrional dans la moitié sud de la Gaule. Du corpus
à l’histoire: position des problèmes, Mémoire de DEA à l`Université (Paris 1994)
W. Suerbaum, Vom antiken zum frühmittelalterlichen Staatsbegriff (3rd edn., Münster
1977)
G.V. Sumner, “The chronology of the governors of al-Andalus to the accession of
'Abd al-Rahßmàn I”, Medieval Studies 48 (1986) pp. 422–69
B. Svoboda, “Zum Verhältnis frühgeschichtlicher Funde des 4. und 5. Jahrhunderts
aus Bayern und Böhmen”, Bayerische Vorgeschichtsblätter 28 (1963) pp. 97–116
S. Szadeczky-Kardoss, “Bagaudae”, Realencyclopädie der classischen Altertumswissenschaft,
Supplement 11 (1968) pp. 346–54
M. Takács, “Die awarenzeitlichen Siedlungen von Lébény”, Reitervölker aus dem Osten.
Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp. 378–82
S. Teillet, Des Goths à la nation Gothique. Les origines de l’idée de nation en Occident du V e
au VIIe siècle (Paris 1984)
Teoderico il grande e i Goti d’Italia. Atti del XIII Congresso internazionale di studi sull’Alto
Medioevo (Spoleto 1993)
  683

Theodisca. Beiträge zur althochdeutschen und altniederdeutschen Sprache und Literatur in der Kultur
des frühen Mittelalters, ed. W. Haubrichs, E. Hellgardt, R. Hildebrandt, S. Müller
and K. Ridder, Ergänzungsbände zum Reallexikon der Germanischen Altertums-
kunde 22 (Berlin-New York 2000)
Theoretical Roman Archaeology Conference, ed. P. Baker (Oxford 1999)
F. Theuws and M. Alkemade, “A kind of mirror for men: sword depositions in
Late antique northern Gaul”, Rituals of Power. From Late Antiquity to the Early Middle
Ages, ed. F. Theuws and J.L. Nelson, The Transformation of the Roman World
8 (Leiden-Boston-Köln 2000) pp. 401–76
F. Theuws and H.A. Hiddink, “Der Kontakt zu Rom”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 66–80
C. Thomas, Christianity in Roman Britain to A.D. 500 (London 1981)
——, “The Early Christian Inscriptions of Southern Scotland”, Glasgow Archaeological
Journal 17 (1992) pp. 1–10
——, Whithorn’s Christian beginnings (Whithorn 1992)
——, And Shall these Mute Stones Speak (Cardiff 1994)
E.A. Thompson, “The settlement of the barbarians in southern Gaul”, Journal of
Roman Studies 46 (1956) pp. 65–75 [repr. id., Romans and Barbarians: the Decline of
the Western Empire (Madison 1982) pp. 23–37]
——, “The Visigoths from Fritigern to Euric”, Historia 12 (1963) pp. 105–26 [repr.
id., Romans and Barbarians: the Decline of the Western Empire (Madison 1982) pp.
38–57]
——, The Early Germans (Oxford 1965)
——, The Visigoths in the Time of Ulfila (Oxford 1966)
——, The Goths in Spain (Oxford 1969) [Spanish transl. Los godos en España (3rd edn.,
Madrid 1990)]
——, “Britain A.D. 406–10”, Britannia 8 (1977) pp. 303–18
——, “Zosimus 6.10.2 and the letters of Honorius”, Classical Quarterly 32 (1982)
445–62
——, Romans and Barbarians: the Decline of the Western Empire (Madison 1982)
——, “Fifth-century facts”, Britannia 14 (1983) pp. 272–4
——, Saint Germanus of Auxerre and the End of Roman Britain (Woodbridge 1984)
——, The Huns (2nd edn., Oxford 1996)
R. Thurneysen, “Wann sind die Germanen nach England gekommen?”, Englische
Studien 22 (1896) pp. 163–79
P. Tomka, “Frühawarenzeitliche Hirten in der kleinen Tiefebene”, Ethnische und kul-
turelle Verhältnisse an der mittleren Donau vom 6. bis zum 11. Jahrhundert. Symposium Nitra
6. bis 10. November 1994, ed. D. Bialeková and J. Zábojník (Bratislava 1996) pp.
141–9
Topographies of Power in the Early Middle Ages, ed. M. de Jong, F. Theuws and C. van
Rhijn, The Transformation of the Roman World 6 (Leiden 2001)
E. Tóth, “Mosaburg und Moosburg”, Acta Archaeologica Hungarica Academiae Scientiarum
Hungaricae 51 (1999/2000) pp. 439–56
E. Tóth and A. Horváth, Kunbábony. Das Grab eines Awarenkhagan (Kecskemét 1992)
Towns and their Territories between Late Antiquity and the Early Middle Ages, ed. G.P.
Brogliolo, N. Gauthier and N. Christie, The Transformation of the Roman World
9 (Leiden 2000)
Die Transalpinen Verbindungen der Bayern, Alemannen und Franken bis zum 10. Jahrhundert,
ed. H. Beumann, W. Schröder, Nationes 6 (Sigmaringen 1987)
The Transformation of Frontiers—from Late Antiquity to the Carolingians, ed. W. Pohl, I.N.
Wood and H. Reimitz, The Transformation of the Roman World 10 (Leiden-
Boston 2000)
The Transformation of the Roman World A.D. 400–900, ed. L. Webster and M. Brown
(London 1997)
684 

M. Trier, “Die frühmittelalterliche Besiedlung des Lechtals im Spiegel der archäol-

ogischen Überlieferung”, Der Landkreis Augsburg, vol. 2: Vor- und Frühgeschichte,
Archäologie einer Landschaft, ed. W. Pötzl and O. Schneider (Augsburg 1996) pp.
267–322
A. Trugly, “Gräberfeld aus der Zeit des awarischen Reiches bei der Schiffswerft in
Komárno I”, Slovenská Archeológia 35 (1987) pp. 251–344
A. Trugly, “Gräberfeld aus der Zeit des awarischen Reiches bei der Schiffswerft in
Komárno II”, Slovenská Archeológia 41 (1993) pp. 191–307
Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern, vol. 1, ed. W. Pohl and
H. Wolfram, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 201. Veröffentlichungen der Kommission für
Frühmittelalterforschung 12 (Wien 1990)
Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern, vol. 2, ed. H. Friesinger
and F. Daim, Denkschriften der Österreichischen Akademie der Wissenschaften,
philosophisch-historische Klasse 204. Veröffentlichungen der Kommission für
Frühmittelalterforschung 13 (Wien 1990)
Über allen Fronten. Nordwestdeutschland zwischen Augustus und Karl dem Großen, ed.
M. Fansa (Oldenburg 1999)
S. Uenze, Die spätantiken Befestigungen von Sadovec, Münchener Beiträge zur Vor- und
Frühgeschichte 43 (München 1992)
B. Ulm, “Patrozinien in Spätantike und Agilolfingerzeit”, Baiernzeit in Oberösterreich.
Das Land zwischen Inn und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts,
Kataloge des Oberösterreichischen Landesmuseums 97 (Linz 1977) pp. 189–212
K. Ulmschneider, “Archaeology, History and the Isle of Wight in the Middle Saxon
period”, Medieval Archaeology 43 (1999) pp. 19–44
Los últimos romanos en Lusitania, ed. A. Velázquez et al. (Mérida 1995)
Unidad y pluralidad en el mundo antiguo. Actas VI Congreso Español de Estudios Clásicos, 2
vols. (Madrid 1983)
Der Untergang des Römischen Reiches, ed. K. Christ, Wege der Forschung 269 (Darmstadt
1970)
R. von Uslar, Westgermanische Bodenfunde des ersten bis dritten Jahrhunderts nach Christus
aus Mittel- und Westdeutschland, Germanische Denkmäler der Frühzeit 3 (Berlin
1938)
V. Väänänen, Introduction au Latin vulgaire (Paris 1963)
M. Vallejo Girvés, Bizancio y la España tardoantigua (ss. V–VIII): Un capítulo de historia
mediterránea (Alcalá de Henares 1993)
J. Vallvé Bermejo, “España en el siglo VIII: ejército y sociedad”, Al-Andalus 43
(1978) pp. 51–112
R. Van Dam, Leadership and Community in Late Antique Gaul (Berkeley 1985)
U. Veit, “Ethnic concepts in German prehistory: a case study on the relationship
between cultural identity and archaeological objectivity”, Archaeological Approaches
to Cultural Identity, ed. S. Shennan (London-New York 1989) pp. 35–56
I. Velázquez, Innovaciones léxico-semánticas en las Etimologías de Isidore de Sevilla (Madrid
1981) (unpublished study)
——, “Vigencia y alcance de los términos innovados en las Etimologías de Isidoro
de Sevilla” Actas del I Congreso Andaluz de Estudios Clásicos ( Jaén 1982) pp. 461–5
——, “Formación de neologismos en las Etimologías de Isidoro de Sevilla”, Actas
del II Congreso Andaluz de Estudios Clásicos (Málaga 1987) pp. 167–72
——, “Wamba y Paulo: dos personalidades enfrentadas y una rebelión”, Espacio,
tiempo y forma, Serie II, Historia Antigua 2 (1989) pp. 213–22
——, “Léxico isidoriano en las Etimologías. Problemas para su estudio”, Euphrosyne
22 (1994) pp. 235–43
——, “Ámbitos y ambientes culturales en la Hispania visigoda. De Martín de Braga
a Isidoro de Sevilla”, Studia Ephemeridis Augustinianum 46 (1994) pp. 328–51
  685

——, “Jural relations as an indicator of Syncretism from the Law of Inheritance

to the Dum inlicita of Chindaswinth”, The Visigoths. From the Migration Period to the
Seventh Century. An Ethnographic Perspective, ed. P.J. Heather (San Marino 1999) pp.
225–58
——, “Impronta religiosa en el desarrollo jurídico de la Hispania visigoda”, ‘Ilu.
Revista de ciencias de las religiones 2 (1999) pp. 97–121
I. Velázquez and G. Ripoll, “Toletum, la construcción de una urbs regia”, Sedes regiae
(ann. 400–800), ed. G. Ripoll and J.M. Gurt with A. Chavarría (Barcelona 2000)
pp. 521–78
Verborum amor. Studien zur Geschichte und Kunst der deutschen Sprache. Festschrift Stefan
Sonderegger, ed. H. Burger, A.M. Haas and P. von Matt (Berlin-New York 1992)
A.D. Verlinde and M. Erdrich, “Eine germanische Siedlung der späten Kaiserzeit
mit umwehrter Anlage und umfangreicher Eisenindustrie in Heeten Overijssel,
Niederlande”, Germania 76 (1998) pp. 693–719
L. Verslype, “L’occupation mérovingienne aux confins de l’Austrasie et de la Neustrie
septentrionales et l’image archéologique des aristocraties”, Clovis. Histoire et mémoire,
vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp. 567–605
H. Vetters, “Die mittelalterlichen Dome zu Salzburg”, Frühmittelalterliche Studien 5
(1971) pp. 413–35
——, “Die mittelalterlichen Dome Salzburgs”, Virgil von Salzburg. Missionar und Gelehrter,
ed. H. Dopsch and R. Juffinger (Salzburg 1985) pp. 286–316
T. Vida, “Zu einigen handgeformten frühawarischen Keramiktypen und ihren
östlichen Beziehungen”, Awarenforschungen, ed. F. Daim, Archaeologia Austriaca 1.
Studien zur Archäologie der Awaren 4 (Wien 1992) pp. 517–77
——, “Frühmittelalterliche scheiben- und kugelförmige Amulettkapseln zwischen
Kaukasus, Kastilien und Picardie”, Berichte der Römisch-Germanischen Kommission des
Deutschen Archäologischen Instituts 76 (1995) pp. 219–90
——, “Das Töpferhandwerk in der Awarenzeit”, Reitervölker aus dem Osten. Hunnen +
Awaren, ed. F. Daim (Halbturn 1996) pp. 362–4
——, “Neue Beiträge zur Forschung der frühchristlichen Funde der Awarenzeit”,
Acta XIII congressus internationalis archaeologiae christianae, Split—Poreè 1994 2, ed.
N. Cambi and E. Marin, Studi di antichità cristiana 54. Vjesnik za arheologiju
i historiju Dalmatinsku 87/89 Suppl. (Split 1998) pp. 529–40
——, Die awarenzeitliche Keramik I (6.–7. Jh.), Varia Archaeologica Hungarica 8
(Budapest 1999)
——, “Die Ziergehänge der awarenzeitlichen Frauen im Karpatenbecken”, Acta
Archaeologica Hungarica Academiae Scientiarum Hungaricae 51 (1999/2000) pp. 367–77
——, “Merowingische Spathagurte der Awarenzeit”, Communicationes Archaeologicae
Hungariae (2000) pp. 161–75
——, Die Keramik der Spätawarenzeit (working title, forthcoming)
T. Vida and A. Pásztor, “Der beschlagverzierte Gürtel der Awaren am Beispiel des
Inventars von Budakalász-Dunapart, Ungarn, Grab 696”, Reitervölker aus dem Osten.
Hunnen + Awaren, ed. F. Daim (Halbturn 1996) pp. 341–7
H. Vierck, “Ein westfälisches “Adelsgrab” des 8. Jahrhunderts n. Chr. Zum archäol-
ogischen Nachweis der frühkarolingischen und altsächsischen Oberschichten”,
Studien zur Sachsenforschung 2, ed. H.-J. Häßler (Hildesheim 1980) pp. 457–88
——, “La ‘chemise de Sainte-Bathilde’ à Chelles et l’influence byzantine sur l’art
de la cour mérovingien au VIIe siècle”, Actes du Colloque International d’Archéologie,
Rouen 1975 (Rouen 1978) pp. 521–70
——, “Imitatio imperii und interpretatio Germanica vor der Wikingerzeit”, Les Pays
du Nord et Byzance, ed. R. Zeitler, Acta Universitatis Upsaliensis N.S. 19 (Uppsala
1981) pp. 64–113
Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and R. Juffinger (Salzburg
1985)
686 

Visigothic Spain: New Approaches, ed. E. James (Oxford 1980)

The Visigoths. From the Migration Period to the Seventh Century. An Ethnographic Perspective,
ed. P.J. Heather (San Marino 1999)
Die Völker an der mittleren und unteren Donau im fünften und sechsten Jahrhundert, ed.
H. Wolfram and F. Daim, Denkschriften der Österreichischen Akademie der
Wissenschaften, philosophisch-historische Klasse 145 (Vienna 1980)
W. Volkert, “Die Ortsnamen des Hachinger Tales”, Herrschaft, Kirche, Kultur. Beiträge
zur Geschichte des Mittelalters. Festschrift für F. Prinz zum 65. Geburtstag, ed. G. Jenal,
Monographien zur Geschichte des Mittelalters 37 (Stuttgart 1993) pp. 43–60
Von Sacerdotium und Regnum. Geistliche und weltliche Gewalt im frühen und hohen Mittelalter.
Festschrift für Egon Boshof zum 65. Geburttag, ed. F.-R. Erkens and H. Wolff (Köln-
Weimar-Wien 2002)
N. Wagner, “Zur Herkunft der Agilolfinger”, Zeitschrift für bayerische Landesgeschichte
41 (1978) pp. 19–48
J.M. Wallace-Hadrill, “The Bloodfeud of the Franks”, Bulletin of the John Rylands
Library 41 (1958/59) pp. 459–87 [repr. id., The Long-Haired Kings (London 1962)
pp. 121–47]
——, The Long-Haired Kings (London 1962)
——, Early Germanic Kingship in England and on the Continent (Oxford 1971)
H. Wanderwitz, “Der Libellus Virgilii und das Verhältnis von Herzögen und Bischöfen
in Bayern”, Virgil von Salzburg. Missionar und Gelehrter, ed. H. Dopsch and R. Juffinger
(Salzburg 1985) pp. 357–61
J. Ward, “Vortigern and the End of Roman Britain”, Britannia 3 (1972) pp. 277–89
B. Ward-Perkins, “Why did the English not become British?”, English Historical Review
115 (2000) pp. 513–33
M. Weber, Grundriss der Sozialökonomik, Part 3: Wirtschaft und Gesellschaft (2nd edn.,
Tübingen 1925)
M. Weidemann, Kulturgeschichte der Merowingerzeit nach den Werken Gregors von Tours, 2
vols. (Mainz 1982)
——, “Adel im Merowingerreich. Untersuchung zu seiner Rechtsstellung”, Jahrbuch
des Römisch-Germanischen Zentralmuseums Mainz 40 (1993) pp. 535–55
S. Weinfurter, “Das Bistum Willibalds im Dienste des Königs. Eichstätt im frühen
Mittelalter”, Zeitschrift für bayerische Landesgeschichte 50 (1987) pp. 3–40
R. Wenskus, Stammesbildung und Verfassung. Das Werden der frühmittelalterlichen gentes (Köln-
Graz 1961, 2nd edn. Köln-Wien 1977)
——, “Wie die Nibelungen-Überlieferung nach Bayern kam”, Zeitschrift für bayerische
Landesgeschichte 36 (1973) pp. 393–449
——, Sächsischer Stammesadel und fränkischer Reichsadel, Abhandlungen der Akademie
der Wissenschaften in Göttingen, philosophisch-historische Klasse 3,93 (Göttingen
1976)
——, “Randbemerkungen zum Verhältnis von Historie und Archäologie, insbesondere
mittelalterlicher Geschichte und Mittelalterarchäologie”, Geschichtswissenschaft und
Archäologie. Untersuchungen zur Siedlungs-, Wirtschafts- und Kirchengeschichte, ed. H. Jankuhn
and R. Wenskus (Sigmaringen 1979) pp. 637–57
J. Werner, “Die Herkunft der Bajuwaren und der ‘östlich-merowingische Reihengrä-
berkreis’”, Aus Bayerns Frühzeit. F. Wagner zum 75. Geburtstag, ed. J. Werner, Schrift-
enreihe zur bayerischen Landesgeschichte 62 (München 1962) pp. 229–50 [repr.
Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965)
pp. 12–43]
——, “Nomadische Gürtel bei Persern, Byzantinern und Langobarden”, Atti del
Convegno Internazionale sul Tema: La civiltà dei Longobardi in Europa. Roma 1971 (Rome
1974) pp. 109–56
K.F. Werner, “Bedeutende Adelsfamilien im Reich Karls des Großen”, Karl der
Große. Lebenswerk und Nachleben, vol. 1: Persönlichkeit und Geschichte, ed. H. Beumann
(Düsseldorf 1965) pp. 83–142
  687

——, Les origines (avant l’an mil), Histoire de France 1 (Paris 1984) [German transl.
Die Ursprünge Frankreichs bis zum Jahr 1000 (Stuttgart 1989)]
——, “Volk, Nation, Nationalismus, Masse, III–V”, Geschichtliche Grundbegriffe. Historisches
Lexikon zur politisch-sozialen Sprache in Deutschland 7 (1992) pp. 171–281
——, “Völker und Regna”, Beiträge zur mittelalterlichen Reichs- und Nationsbildung in
Deutschland und Frankreich, ed. C. Brühl and B. Schneidmüller, Historische Zeitschrift
Beiheft Neue Folge 24 (München 1997) pp. 15–44
——, “Die ‘Franken’. Staat oder Volk?”, Die Franken und die Alemannen bis zur “Schlacht
bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände zum Reallexikon der
Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 95–101
——, Naissance de la Noblesse (Paris 1998)
Westillyricum und Nordostitalien in der spätrömischen Zeit, ed. R. Brato≥ (Ljubljana 1996)
M. Whitby, The Emperor Maurice and his Historian: Theophylact Simocatta on Persian and
Balcan Warfare (Oxford 1988)
R. White and P. Barker, Wroxeter: the Life and Death of a Roman City (Stroud 1998)
C. Wickham, Early Medieval Italy. Central Power and Local Society 400 –1000, New
Studies in Medieval History (London 1981)
——, “Lawyers’ Time: History and Memory in Tenth and Eleventh-Century Italy”,
Studies presented to R.H.C. Davis, ed. H. Mayr-Harting and R.I. Moore (London
1985) pp. 53–71
——, “Problems of comparing rural societies in early medieval western Europe”,
Transactions of the Royal Historical Society 6th series 2 (1992) pp. 221–46
F. Wieacker, Allgemeine Zustände und Rechtszustände gegen Ende des weströmischen Reichs,
Ius Romanum Medii Aevi 1,2a (Milano 1963)
A. Wieczorek, “Die Ausbreitung der fränkischen Herrschaft in den Rheinlanden
vor und seit Chlodwig I.”, Die Franken—Wegbereiter Europas. Vor 1500 Jahren: König
Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim 1 (Mainz
1996) pp. 240–60
P. Wiesinger, “Gotische Lehnwörter im Bairischen. Ein Beitrag zur sprachlichen
Frühgeschichte des Bairischen”, Frühmittelalterliche Ethnogenese im Alpenraum, ed.
H. Beumann and W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 153–200
——, “Antik-romanische Kontinuitäten im Donauraum von Ober- und Niederös-
terreich am Beispiel der Gewässer-, Berg- und Siedlungsnamen”, Typen der Ethnogenese
unter besonderer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
201. Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien
1990) pp. 261–328
D.G. Wigg, Münzumlauf in Nordgallien um die Mitte des 4. Jahrhunderts n. Chr., Studien
zu Fundmünzen der Antike 8 (Berlin 1991)
T. Williamson, The Origins of Norfolk (Manchester 1993)
——, The Origins of Hertfordshire (Manchester 2000)
K.-H. Willroth, “Siedlungen und Gräber als Spiegel der Stammesbildung. Gedanken
zur Abgrenzung germanischer Stämme in der ausgehenden vorrömischen Eisenzeit
in Norddeutschland und Südskandinavien”, Studien zur Archäologie des Ostseeraumes.
Festschrift für Michael Müller-Wille (Neumünster 1999) pp. 359–71
H. Winter, Awarenzeitliche Grab- und Streufunde aus Ostösterreich. Ein Beitrag zur Siedlungs-
geschichte, Monographien zur Frühgeschichte und Mittelalterarchäologie 4 (Innsbruck
1997)
P. Wobrauschek, W. Haider and C. Streli, “Röntgenfluoreszenzanalyse von Bronze-
funden aus der Awarenzeit”, F. Daim et al., Das awarische Gräberfeld von Leobersdorf,
Niederösterreich 2, Studien zur Archäologie der Awaren 3. Denkschriften der Öster-
reichischen Akademie der Wissenschaften, philosophisch-historische Klasse 194.
Veröffentlichungen der Kommission für Frühmittelalterforschung 10 (Wien 1987)
pp. 45–55
688 

K.B. Wolf, “Christian views of Islam in early medieval Spain”, Medieval Christian
Perceptions of Islam: A Book of Essays, ed. J.V. Tolan (New York-London 1996) pp.
85–108
H. Wolff, “Die Anfänge des Christentums in Ostraetien, Ufernoricum und Nordwest-
pannonien: Bemerkungen zum Regenwunder und zum hl. Florian”, Ostbairische
Grenzmarken 31 (1989) pp. 27–45
——, “Die Kontinuität städtischen Lebens in den nördlichen Grenzprovinzen des
römischen Reiches und das Ende der Antike”, Die Stadt in Oberitalien und in den
nordwestlichen Provinzen des Römischen Reiches. Deutsch-Italienisches Kolloquium im ital-
ienischen Kulturinstitut Köln, ed. W. Eck and H. Galsterer, Kölner Forschungen 4
(Mainz 1991) pp. 287–318
——, “Die Kontinuität der Kirchenorganisation in Raetien und Noricum bis an
die Schwelle des 7. Jahrhunderts”, Das Christentum im bairischen Raum von den Anfängen
bis ins 11. Jahrhundert, ed. id. and E. Boshof, Passauer historische Forschungen 8
(Köln-Wien 1994) pp. 1–17
H. Wolfram, Lateinische Königs- und Fürstentitel bis zum Ende des 8. Jahrhunderts, Intitulatio
1 (Graz 1967)
——, “Das Fürstentum Tassilos III., Herzogs der Bayern”, Mitteilungen der Gesellschaft
für Salzburger Landeskunde 108 (1968) pp. 157–79
——, “Der heilige Rupert und die antikarolingische Adelsopposition”, Mitteilungen
des Instituts für Österreichische Geschichtsforschung 80 (1972) pp. 4–34
——, “Die Christianisierung der Baiern”, Baiernzeit in Oberösterreich. Das Land zwi-
schen Inn und Enns vom Ausgang der Antike bis zum Ende des 8. Jahrhunderts, Kataloge
des Oberösterreichischen Landesmuseums 97 (Linz 1977) pp. 177–88
——, “Gotisches Königtum und römisches Kaisertum von Theodosius dem Großen
bis Justinian I.”, Frühmittelalterliche Studien 13 (1979) pp. 1–28
——, “Theogonie, Ethnogenese und ein kompromittierter Großvater im Stammbaum
Theoderichs des Großen”, Festschrift für Helmut Beumann zum 65. Geburtstag, ed.
K.-U. Jäschke and R. Wenskus (Sigmaringen 1979) pp. 80–97
——, Geschichte der Goten. Von den Anfängen bis zur Mitte des sechsten Jahrhunderts. Entwurf
einer historischen Ethnographie (München 1979) [3rd edn. Die Goten (1990); English
transl. History of the Goths (Berkeley 1988)]
——, “Die Zeit der Agilolfinger. Rupert und Virgil”, Geschichte Salzburgs, vol. 1,1:
Vorgeschichte, Altertum, Mittelalter, ed. H. Dopsch (Salzburg 1981) pp. 121–56
——, “Der Heilige Rupert in Salzburg”, Frühes Mönchtum in Salzburg, ed. E. Zwink,
Salzburg Diskussionen 4 (Salzburg 1983) pp. 81–92
——, “Ethnogenesen im frühmittelalterlichen Donau- und Ostalpenraum (6. bis 10.
Jahrhundert)”, Frühmittelalterliche Ethnogenese im Alpenraum, ed. H. Beumann and
W. Schröder, Nationes 5 (Sigmaringen 1985) pp. 97–151
——, “Baiern und das Frankenreich”, Die Bajuwaren. Von Severin bis Tassilo 488–788,
ed. H. Dannheimer and H. Dopsch (Rosenheim-Salzburg 1988) pp. 130–5
——, “Tassilo III. und Karl der Große—Das Ende der Agilolfinger”, Die Bajuwaren.
Von Severin bis Tassilo 488–788, ed. H. Dannheimer and H. Dopsch (Rosenheim-
Salzburg 1988) pp. 160–6
——, Das Reich und die Germanen. Zwischen Antike und Mittelalter, Das Reich und die
Deutschen (Berlin 1990) [English transl. The Roman Empire and its Germanic Peoples
(Berkeley 1997)]
——, “Origo et religio. Ethnic traditions and literature in early medieval texts”, Early
Medieval Europe 3 (1994) pp. 19–38
——, 378–907: Grenzen und Räume. Geschichte Österreichs vor seiner Entstehung, Österre-
ichische Geschichte 3 (Wien 1995)
——, Salzburg, Bayern, Österreich. Die Conversio Bagoariorum et Carantanorum und die Quellen
ihrer Zeit, Mitteilungen des Instituts für Österreichische Geschichtsforschung,
Ergänzungsband 31 (Wien-München 1995)
  689

——, “Neglected evidence on the accommodation of the barbarians in Gaul”, King-

doms of the Empire. The Integration of Barbarians in Late Antiquity, ed. W. Pohl, The
Transformation of the Roman World 1 (Leiden-New York-Köln 1997) pp. 181–3
——, “Typen der Ethnogenese. Ein Versuch”, Die Franken und die Alemannen bis zur
“Schlacht bei Zülpich” (496/97), ed. D. Geuenich, Ergänzungsbände zum Reallexikon
der Germanischen Altertumskunde 19 (Berlin-New York 1998) pp. 608–27
I.N. Wood, “Kings, Kingdoms and Consent”, Early Medieval Kingship, ed. P.H. Sawyer
and I.N. Wood (Leeds 1977) pp. 6–29
——, The Merovingian North Sea (Alingsås 1983)
——, “Gregory of Tours and Clovis”, Revue Belge de Philologie et d’Histoire 63 (1985)
pp. 249–72
——, “Disputes in late fifth- and early sixth-century Gaul: some problems”, The
Settlement of Disputes in Early Medieval Europe, ed. W. Davies and P. Fouracre
(Cambridge 1986) pp. 7–22
——, “The fall of the western empire and the end of Roman Britain”, Britannia
18 (1987) pp. 251–62
——, “Clermont and Burgundy: 511–534”, Nottingham Medieval Studies 32 (1988) pp.
119–25
——, “Ethnicity and the ethnogenesis of the Burgundians”, Typen der Ethnogenese unter
besonderer Berücksichtigung der Bayern 1, ed. W. Pohl and H. Wolfram, Denkschriften
der Österreichischen Akademie der Wissenschaften, philosophisch-historische Klasse
201. Veröffentlichungen der Kommission für Frühmittelalterforschung 12 (Wien
1990) pp. 53–69
——, “The Channel from the 4th to the 7th centuries A.D.”, Maritime Celts, Frisians
and Saxons, ed. S. McGrail, Council for British Archaeology Research Report 71
(London 1990) pp. 93–7
——, “The Franks and Sutton Hoo”, People and Places in Northern Europe 500–1600.
Essays in Honour of Peter Hayes Sawyer, ed. id. and N. Lund (Woodbridge 1991)
pp. 1–14
——, “Continuity of calamity?: the constraints of literary models”, Fifth-century Gaul:
a crisis of identity, ed. J.F. Drinkwater and H. Elton (Cambridge 1992) pp. 9–18
——, “The mission of Augustine of Canterbury to the English”, Speculum 69 (1994)
pp. 1–17
——, The Merovingian kingdoms, 450–751 (London-New York 1994)
——, “Defining the Franks: Frankish Origins in Early Medieval Historiography”,
Concepts of National Identity in the Middle Ages, ed. S. Forde, L. Johnson and A.V.
Murray, Leeds Texts and Monographs. New Series 14 (Leeds 1995) pp. 47–57
——, “Die Franken und ihr Erbe—‘Translatio Imperii’”, Die Franken—Wegbereiter
Europas. Vor 1500 Jahren: König Chlodwig und seine Erben. Katalog zur Ausstellung im
Reiss-Museum Mannheim 1 (Mainz 1996) pp. 358–64
——, “Before and after migration to Britain”, The Anglo-Saxons from the Migration
Period to the Eighth Century: An Ethnographic Perspective, ed. J. Hines (Woodbridge
1997) pp. 41–64
——, “Conclusion: Strategies of distinction”, Strategies of Distinction. The Construction
of Ethnic Communities, 300–800, ed. W. Pohl with H. Reimitz, The Transforma-
tion of the Roman World 2 (Leiden-New York-Köln 1998) pp. 297–303
——, “The barbarian invasions and first settlements”, Cambridge Ancient History 13
(2nd edn., 1998) pp. 516–37
The World of Gregory of Tours, ed. K. Mitchell and I.N. Wood (Leiden 2002)
P. Wormald, “Lex Scripta and Verbum Regis: Legislation and Germanic Kingship, from
Euric to Cnut”, Early Medieval Kingship, ed. P.H. Sawyer and I.N. Wood (Leeds
1977) pp. 105–38 [repr. and corrected id., Legal Culture in the Early Medieval West.
Law as Text, Image and Experience (London 1999) pp. 1–43]
——, “In Search of King Offa’s ‘Law-code’”, People and Places in Northern Europe
690 

500–1600. Essays in Honour of Peter Hayes Sawyer, ed. I.N. Wood and N. Lund
(Woodbridge 1991) pp. 25–45 [repr. id., Legal Culture in the Early Medieval West.
Law as Text, Image and Experience (London 1999) pp. 201–23]
——, “‘Inter Cetera Bona . . . Genti Suae’: Law-making and Peace-Keeping in the ear-
liest English Kingdoms”, La giustizia nell’alto medioevo (secoli V–VIII), Settimane di
studio del centro italiano di studi sull’alto medioevo 42 (Spoleto 1995) pp. 963–96
——, Legal Culture in the Early Medieval West. Law as Text, Image and Experience (London
1999)
——, The Making of English Law. King Alfred to the Twelfth Century, vol. 1: Legislation
and its Limits (Oxford 1999)
N. Wright, “Gildas’ prose style and its origins”, Gildas: New Approaches, ed. M. Lapidge
and D.N. Dumville (Woodbridge 1984) pp. 107–28
Ph. Wynn, “Frigeridus, the British Tyrants and the Early Fifth Century Barbarian
Invasions of Gaul and Spain”, Athenaeum 85 (1997) pp. 69–117
B.A.E. Yorke, “The Jutes of Hampshire and Wight and the origins of Wessex”,
The Origins of Anglo-Saxon Kingdoms, ed. S. Bassett (Leicester 1989) pp. 84–96
——, Wessex in the Early Middle Ages (London 1995)
——, “The reception of Christianity at the Anglo-Saxon royal courts”, St Augustine
and the Conversion of England, ed. R. Gameson (Stroud 1999) pp. 152–73
——, “Political and ethnic identity: a case study of Anglo-Saxon practice”, Social
Identity in Early Medieval Britain, ed. B. Frazer and A.J. Tyrell (London 2000) pp.
69–89
——, “Gregory of Tours and sixth-century Anglo-Saxon England”, The World of
Gregory of Tours, ed. K. Mitchell and I.N. Wood (Leiden 2002) pp. 113–30
B.K. Young, “Pratiques funéraires et mentalités païennes”, Clovis. Histoire et mémoire,
vol. 1: Le baptême de Clovis, l’événement, ed. M. Rouche (Paris 1997) pp. 15–42
H. Zeiß, “Bemerkungen zur frühmittelalterlichen Geschichte Baierns I”, Zeitschrift
für bayerische Landesgeschichte 2 (1929) pp. 343–60
——, “Bemerkungen zur frühmittelalterlichen Geschichte Baierns II”, Zeitschrift für
bayerische Landesgeschichte 4 (1931) pp. 351–66
——, “Von den Anfängen des Baiernstammes”, Bayerische Vorgeschichtsblätter 13 (1936)
pp. 24–40
G. Zeller, “Tracht der Frauen”, Die Franken—Wegbereiter Europas. Vor 1500 Jahren:
König Chlodwig und seine Erben. Katalog zur Ausstellung im Reiss-Museum Mannheim 2
(Mainz 1996) pp. 672–83
E. Zöllner, “Die Herkunft der Agilulfinger”, Mitteilungen des Instituts für Österreichische
Geschichtsforschung 59 (1951) pp. 245–64 [repr. Zur Geschichte der Bayern, ed. K. Bosl,
Wege der Forschung 60 (Darmstadt 1965) pp. 107–34]
——, “Der bairische Adel und die Gründung von Innichen”, Mitteilungen des Instituts
für Österreichische Geschichtsforschung 68 (1960) pp. 362–87 [repr. Zur Geschichte der
Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965) pp. 135–71]
——, Geschichte der Franken bis zur Mitte des sechsten Jahrhunderts (München 1970)
——, “Das Geschlecht der Agilolfinger”, Die Anfänge des Klosters Kremsmünster, ed.
S. Haider, Mitteilungen des Oberösterreichischen Landesarchivs, Ergänzungsband
2 (Linz 1978) pp. 83–110
T. Zotz, “Adel, Oberschicht, Freie. Zur Terminologie der frühmittelalterlichen
Sozialgeschichte”, Zeitschrift für die Geschichte des Oberrheins 125 (1977) pp. 2–20
Zur Geschichte der Bayern, ed. K. Bosl, Wege der Forschung 60 (Darmstadt 1965)
 INDEX OF PEOPLES

Alamans 40, 41, 61, 62, 244, 260, 441, 444, 605, 608, 610, 612, 614,
275, 294, 305, 307, 309, 310, 311, 619, 622, 623
316, 318, 320, 321, 325, 332, 335, Byzantines: see Romans/East Romans
336, 337, 571, 590, 600, 601, 605,
607, 615, 618 Celts 11, 322, 339, 370, 386, 404,
Alans 63, 65, 66, 67, 68, 106, 108, 506, 602
137, 138, 139, 140, 141, 146, 147, Chamavi 276, 309
148, 149, 151, 155, 156, 252, 608, Chasuarii 309
610, 624 Chatti 309
Ampsivarii 276, 309 Chattuarii 276, 309
Angles 361, 385, 386, 387, 388, 389, Chauci 309
392, 395, 396, 400, 401, 404, 605, Chazars 483
611, 628 Cutrigurs 579, 580
Anglo-Saxons 2, 11, 98, 249, 352,
376, 381–407, 506, 610, 611, 612, Danes 387, 402
621, 628. See also Angles, Saxons,
Jutes Egyptians 244
Anglo-Saxons/Angles 385, 387, 388,
392 Falchovarii 276
Anglo-Saxons/Jutes 385, 387, 388, Franks 2, 7, 11, 18, 28, 30, 32–4, 36,
391, 394 40, 41, 61, 62, 64, 66, 83, 90, 98,
Angrivarii 276 100, 111, 117, 119, 174, 178, 186,
Antae 579 205, 213, 229, 233, 243, 244, 247,
Arabs 56, 219–20, 222–4, 226–7, 256, 258, 260–2, 265, 266, 269,
230, 232–4, 237–40, 517, 606, 620. 271–344, 352, 380, 387, 388, 394,
See also Berbers, Moors, Saracens 401, 403, 406, 415, 416, 438–41,
Armoricans 347 444, 457, 461, 518, 522, 574, 586,
Arverni 345 588, 590, 592, 593, 594, 595, 596,
Avars 11, 21, 442, 457, 463–595, 599–601, 603–9, 611, 612, 614–7,
601, 602, 604, 607, 609, 613, 618, 619, 620, 622–5
620, 624, 626, 629 Franks/Ripuarians 40–1, 313–4
Franks/Salians 276, 310, 313
Bavarians 11, 40, 41, 325, 429–61, Frisians 44, 387, 404
574, 580, 605, 611, 615, 618, 622,
623 Gauls 59, 60, 61, 136, 311, 606
Berbers 56, 219, 220, 224, 226, 233, Gepids 103, 118, 250, 415, 416,
234, 237, 238, 239, 615 422, 473, 580, 581, 582, 588, 593, 615
Boii 60, 430 Germans 2, 5, 7, 59, 60, 61, 96, 278,
Britons 60, 345, 347, 351, 353, 354, 279, 281–4, 303, 316, 578, 584, 606
357, 360, 361, 369, 376, 379, 380, Getae 585
388, 392, 398, 602, 610, 615 Goths 1, 24, 26–8, 85–132, 141, 144,
Bructeri 276, 309, 387 146, 154–5, 161–217, 222–3, 232–4,
Bulgars 21, 422, 522, 579, 580, 582, 240, 248, 249, 251, 261–2, 265,
583, 584, 588, 596, 615 267, 335–6, 345, 352, 380, 416,
Burgundians 11, 27, 59, 113, 243–69, 436, 580–1, 600, 604, 607, 617,
320, 321, 325, 335, 336, 366, 392, 623, 626
692   

Goths/Ostrogoths 1, 11, 58, 68, 77, Ogurs 576, 577, 578, 579
78, 85–132, 243, 248, 260, 265, Onogurs 576, 579
267, 269, 303, 319, 320, 338, 425, Ostrogoths: see Goths/Ostrogoths
438, 588, 604, 605, 608, 610, 612,
614, 615, 619, 620, 622 Pannonians 63, 64, 87, 91, 101, 104,
Goths/Visigoths 2, 11, 15, 58, 61, 65, 105, 106, 422
66, 70, 74, 77, 78, 92, 96, 107, 108, Pechenegs 578
119, 120, 148, 155, 161–217, 219, Persians 224, 425, 588
233, 234, 250, 252, 253, 265, 269,
303, 320, 321, 325, 441, 604, 607, Romans 1, 7, 14, 15, 16, 18, 22,
608, 610, 612, 614, 619, 620, 625 24–7, 31, 32, 33, 34, 35, 36, 38,
Greeks: see Romans/East Romans 55, 57, 59, 60, 61, 63, 64, 65, 66,
Greuthungs 61, 88, 100, 108. See also 68, 70, 71, 73, 74, 75, 78, 81, 83,
Goths/Ostrogoths 86–9, 91, 94, 100, 101, 102, 105,
107–10, 117, 118, 120–3, 125, 126,
Hastings: see Vandals/Hastings 129, 130, 135, 140, 141, 144, 145,
Hephtalites 576 148, 150, 152, 157, 161, 173,
Heruls 153, 414, 416, 422 178–80, 182, 186, 191, 192, 201,
Hispani 173–5, 178, 186, 191, 195, 233, 244, 245, 252, 256–60, 262,
198, 201, 205, 213, 617 263, 265, 279, 287, 300, 305, 309,
Hispano-Romans 15, 119, 143, 145, 312, 315, 316, 322, 323, 327, 329,
157, 168, 172, 173, 180, 181, 182, 330–2, 334, 335, 340, 342, 347,
195, 204, 233, 240 350, 426, 427, 470, 506, 582,
Honoriaci 65 606–8, 611, 616, 618, 619, 621,
Hungarians 572, 574, 578 622
Huns 11, 89, 91, 104, 246, 252, 387, Romans/East Romans 18, 19, 62, 75,
414, 469, 474–6, 481, 483, 485, 91, 101, 131, 132, 174, 179, 180,
486, 492, 515, 572, 574–6, 578, 189, 205, 222, 223, 234, 415, 484,
579, 585–8, 596 509, 518, 578, 581, 582, 584, 585,
589
Iceni 396 Rugians 86, 96, 103, 123, 124, 387

Jews 37, 38, 185, 190, 210, 216 Sabirs 579

Juan-Juan 576 Saracens 224, 226–8, 232, 233,
Jutes 385, 387–91, 393, 394, 401, 606
605, 611 Sarmatians 103, 422
Saxons 41, 44, 62, 243, 277–8,
Lombards 2, 11, 22, 34, 35, 38, 58, 309–11, 318, 335–6, 349–54, 361,
178, 243, 249, 267, 303, 380, 381, 385, 387–9, 391, 392, 393,
409–27, 454, 457, 460, 471, 474, 395, 396, 399–401, 404, 406, 422,
519, 573, 574, 580, 581, 593, 595, 571, 600, 605, 622
603, 604, 606, 607, 610, 612, 619, Sciri 89, 435, 437, 581
620, 623–7 Scythians 572, 575, 588
Silings: see Vandals/Silings
Magyars 572, 596 Slavs 442, 443, 571, 573, 580, 582,
Marcomanni 108, 431, 432, 434 583, 584, 585, 593, 594, 596, 615,
Mongols 572 622
Moors 68, 81, 83, 223, 224, 226–8, Sogdians 575
233, 234, 606 Sporoi 585
Sueves 59, 62, 63, 65, 66, 67, 89,
Naristi: see Warasks 108, 137–42, 146, 147, 148, 149,
Nervii 60 151–6, 169, 172, 173, 176, 178,
Noricans 422 179, 222, 223, 250, 349, 352, 422,
Norwegians 387 435, 441, 442, 608, 627
    693

Tencteri 309 Vandals/Hastings 61–6, 72, 140, 141,

Tervings 61, 88, 100, 108, 571. 146, 148, 149, 151
See also Goths/Visigoths Vandals/Silings 61–3, 65, 66, 107,
Thuringians 44, 325, 335, 336, 615 140, 141, 146–9, 609
Treveri 60 Varchonitae 586. See also Avars
Trinovantes 396 Varisti: see Warasks
Tubantes 276, 309 Vascones 213
Tungri 59 Venethi 583, 585
Turks 575, 576, 578, 588 Visigoths: see Goths/Visigoths
Volcae 583
Usipites 309
Utigurs 579, 580 Wallones 583
Warasks 434, 449, 450
Vandals 11, 16, 55–83, 97, 106–8, 131, Wenden 583
137–41, 146–53, 155, 156, 222, 223,
243, 265, 267, 269, 600, 604, 607–10, Zaloi 579
612–6, 620, 622–4, 627
 INDEX OF PERSONS

Adaloald (son of Agilulf ) 420 Anagastes (Gothic count) 87

Adax (Alanic king) 66 Anastasius (Roman emperor) 16, 117,
Aegidius (king of Soissons) 253, 266, 255, 257
269, 312, 319, 320, 321, 338, 366, Aneirin (poet) 369, 373, 374, 375,
399, 611, 612 376
Ælfwold (Northumbrian king) 45 Ansoald (notary) 35
Ælle (Anglo-Saxon ruler) 404 Anthemius (Roman emperor) 252
Æthelbald (Mercian king) 391 Antonius (bishop) 71
Æthelberht (Kentish king) 31, 32, 33, Apsich (Avar commander) 586–7
34, 40, 41, 394, 404 Apsich (Byzantine officer) 586–7
Aëtius (magister militum) 119, 246, Arbogast (magister militum) 311
250, 252, 269, 283, 352, 353, 354, Aripert I (Lombard king) 420
372, 613 Arogast (Frankish judge) 28
Agilmund (Burgundian king) 410, Aspar (Byzantine politician) 86, 87,
411, 412, 413, 414, 417 88, 102, 103
Agilolfings (Alamannic-Bavarian Asparuch (Bulgar khan) 484, 579
dynasty) 429, 439–42, 444–6, Athalaric (Ostrogothic king) 24, 27,
452–8, 460 94, 116, 124
Agilulf (Lombard king) 418, 420, Athanagild (Visigothic king) 176–7
423, 439 Athanaric (Visigothic leader) 190–1,
Agilus (abbot) 449 268
Agio (Burgundian king) 411, 412, Athaulf (Visigothic king) 6, 17, 162
413, 414 Attila (Hunnic king) 8, 40, 91, 100,
Agrestius (missionary) 450 124, 252, 283, 473, 474, 485, 486,
A˙mad al-Ràzì (historian) 236–7 573, 580
Aistulf (Lombard king) 419, 421 Audoin (Lombard king) 412, 420
Alaric I (Visigothic king) 63, 72, 77, Augustine (Saint) 138, 144, 164, 183,
88, 346, 347 189, 390, 394
Alaric II (Visigothic king) 27, 28, 46, Aurelius Caninus (British king) 357
107, 155 Ausonius (poet) 23, 144, 276
Alboin (Lombard king) 412, 414, Authari (Lombard king) 415, 420,
422, 609, 611, 612 426, 427
Alciocus (Bulgar duke) 579, 587 Avitus of Vienne (bishop) 26, 247,
Alcuin (scholar) 45 254, 255, 256, 257, 261, 262, 264,
Alfred (king of Wessex) 21, 42, 89, 265, 266, 268, 269, 329, 331
272, 621
al-Óakam II (Moorish leader) 236 Baian (Avar khagan) 573, 586, 592
al-Sam˙ (Moorish leader) 226, 229, 230 Balamber (Hunnic king) 89, 90, 92
al-Walìd (Moorish leader) 226, 228 Balths (Visigothic dynasty) 92, 606
Amalasuintha (Ostrogothic queen) 16, Barbas (bishop) 71
93, 94 Basiliscus (brother of emperor Leo)
Amals (Ostrogothic dynasty) 90, 92, 72
93, 606, 625 Bauto (magister militum) 311
Ambrosius Aurelianus (British leader) Bede the Venerable (scholar) 5, 33,
345, 354 34, 336, 348, 350, 352, 353, 354,
Ammianus Marcellinus (historian) 25, 381, 382, 385, 386, 387, 389, 390,
90, 92, 96, 106, 136, 243–4, 264, 391, 392, 393, 394, 397, 400, 403,
280, 310, 319, 339 404, 405, 610, 627
    695

Belisarius (Roman general) 16, 58, Chintila (Visigothic king) 205, 206,
67, 69, 76, 81, 82, 83, 111, 112, 210
129, 130, 131, 619 Chlodomer (Frankish king) 262, 268
Beremund (grandson of Hunimund) 90 Chlothar I (Frankish king) 32, 39, 40,
Bluchbard (poet) 373–4 268, 323, 324, 326, 329, 332, 336,
Boethius (Roman senator) 24, 116, 440
118, 130 Chlothar II (Frankish king) 32, 39,
Boniface (Saint) 127, 452 40, 323, 324, 329, 332
Boruth (Carinthian leader) 443 Chrotechildis (wife of Clovis) 268, 269
Braulius of Saragossa (bishop) 38, Cian (poet) 373, 374
165, 166, 189, 208, 209 Clef (Lombard king) 415, 426
Burgundofarones (Franco-Burgundian Clovis (Frankish king) 7, 16, 28–9,
family) 247 33, 39, 45, 72, 90, 100, 253, 255,
258, 262, 268–9, 282, 285–6, 307–9,
'Abd al-Malik (Moorish leader) 230, 311, 313, 316–7, 319–33, 335, 338,
234, 236 341–3, 352, 436, 598, 606, 611–2,
'Abd al-Ra˙màn I (Moorish leader) 618–9
221, 230, 238 Consentius (monk) 138, 142–5
'Abd al-Ra˙màn III (Moorish leader) Constans II (Roman emperor) 493,
236 594
'Abd al-Ra˙màn al-Ghafiqì (Moorish Constantine (usurper) 137, 156, 346,
leader) 234 349
Carloman (Frankish maiordomus) 453 Constantine III (Roman emperor) 64,
Carolingians (Frankish dynasty) 38, 65, 136, 137, 141, 347, 354, 367
44, 46, 324, 340, 343, 443, 453, Constantine the Great (Roman
587, 589 emperor) 23, 62, 312, 354
Cassiodorus (Roman senator) 16, Constantinus (British king) 357, 369
24–7, 61, 90, 92, 93, 94, 110, Constantius (Roman emperor) 24, 64,
112–4, 116–8, 120–2, 124, 125, 139, 141, 308, 312, 348, 399
128, 165, 248, 254, 263, 265, 266, Cuneglasus (British king) 357, 359
294, 435, 436 Cunigast 24
Ceawlin (Anglo-Saxon ruler) 369, Cunincpert (Lombard king) 420
399, 404 Cynric 369, 391
Cerdic 369, 391, 399 Cyprianus (Roman politician) 118,
Charibert (Frankish king) 394 127, 130
Charlemagne (Frankish king and Cyrila (bishop) 71, 78
emperor) 19, 44, 45, 455, 460,
463, 516, 518, 522, 574, 593, 595, Dagobert I (Frankish king) 32, 40,
619, 623 186, 324, 336, 445, 579
Charles Martel (Frankish maiordomus) Darwin 55
220, 452, 453, 454 Desiderius (Lombard king) 421, 447,
Childebert I (Frankish king) 39, 268, 455, 460
336 Desiderius of Cahors (bishop) 32
Childebert II (Frankish king) 39, 445 Didymus 137, 345
Childeric I (Frankish king) 285, 313, Drusus (Roman general) 245
316, 319, 400, 612
Childeric II (Frankish king) 324 Ecdicius 345
Chilperic I (Burgundian king) 248, Egica (Visigothic king) 210, 215, 225
250, 251, 253, 261, 268 Einhard (scholar) 44, 593
Chilperic II (Burgundian king) 251, Eligius of Limoges (Saint) 32
253, 268 Emmeram of Poitiers (Saint) 451
Chindaswinth (Visigothic king) 37, Ennodius of Pavia (bishop) 16, 110,
166, 182, 202, 206, 207, 208, 210, 113, 128, 253, 262, 266, 294
215, 217 Eormenric (Kentish king) 394, 395
696   

Eraric (Ostrogothic king) 94, 123, Gregory of Tours (historian) 26,

587 63–4, 66, 118–9, 128, 165, 186,
Ermenaric (Gothic king) 90, 92, 93 247, 253, 262, 264, 268–9, 311–3,
Erwig (Visigothic king) 37, 38, 197, 319, 322, 324–6, 328–9, 332,
210, 213, 214, 215, 233 335–6, 338, 340–1, 345, 394, 401,
Eugenius I of Toledo (bishop) 166 403–4, 440, 591, 602
Eugenius II of Toledo (bishop) 208 Gregory the Great (pope) 128, 164,
Eulogius (historian) 225 172, 184
Euric (Visigothic king) 17, 21, 26, 27, Grimoald (Bavarian duke) 447
28, 35, 36, 96, 122, 154, 252, 253, Grimoald (Lombard king) 419, 420,
265 422
Eusebius of Caesarea (bishop and Gundemar (Visigothic king) 166
historian) 222, 223, 348 Gunderic (Vandal king) 66–7
Eustasius of Luxeuil (missionary) 449 Gundichar (Burgundian king) 246,
Eutharic (husband of Amalasuintha) 247, 248, 249
16, 93, 94 Gundilas 129–30
Gundioc (Burgundian king) 248, 250,
Florentinus (poet) 74 251, 252, 261
Folchaid (wife of Theodo) 451 Gundobad (Burgundian king) 26, 27,
Fredbal (Vandal king) 141 245, 247, 248, 251–7, 259–69, 335,
Fredegar (historian) 186, 250, 254, 611
259, 309, 323, 329, 335, 336, Gunthamund (Vandal king) 76
338–40, 579, 584, 587 Guteric (Vandal king) 141
Fritgil (Marcomannic queen) 432 Guthlac (Saint) 391–2
Froia (usurper) 209
Hagen, Haganus 248
Gambara (mother of Ibor and Agio) Hengist 382, 392
412–3 Heracleus (Roman emperor) 222,
Garibald I (Bavarian duke) 442, 454 224, 605
Garibald II (Bavarian duke) 442 Hermenegild (Visigothic prince) 166
Geiseric (Vandal king) 63, 67, 68, 72, Hermeric (Suevic king) 66, 142
73, 75, 77, 79, 80, 83, 606, 608, Hilderic (Vandal king) 74, 80
609, 611 Hiltrud (wife of Odilo) 453
Gelimer (Vandalic king) 16, 81, 82 Honorius (Roman emperor) 136–7,
Gensemund (son of Hunimund) 90, 139–41, 152, 156, 346, 347, 349
125 Horsa 382, 392
George of Ostia (bishop) 45 Huneric (Vandal king) 75, 78, 80,
Gerontius (usurper) 65, 137, 139, 81, 618
141, 156 Hunimund (Gothic leader) 90, 433
Gibich (Burgundian king) 248 Hydatius (historian) 64–6, 79, 108,
Gibichungs (Burgundian dynasty) 247, 138–41, 144, 146–9, 151–4, 156–7,
249, 258–63, 266, 267, 269, 606, 246, 348
625
Gildas (historian) 225, 345, 347–8, ibn al-Qù†ìya (historian) 227, 229
350–60, 366–7, 372, 375–6, 381–2, ibn Óabìb (historian) 220, 221, 227,
385, 392, 395–9, 609 230, 235–9
Gislahar (Burgundian king) 248 ibn Óazm (historian) 238
Glycerius (Roman emperor) 252 ibn Khaldûn (historian) 56
Godegisel (Burgundian king) 64, 253, Ibor (Burgundian king) 412, 413
254, 268 Ida (Bernician king) 373, 374
Godeoc (Lombard king) 414 Ildebad (Gothic leader) 94, 123
Godomar (Burgundian king) 248, Illus (Roman general) 102
253, 256, 266 Ine (king of Wessex) 32, 33, 41, 42,
Gratianus (usurper) 354 376, 620
    697

Innocent III (pope) 18 Magnentius (Roman emperor) 311

Isidore of Seville (historian) 17, 26–7, Magnus Maximus (usurper) 64, 65,
35, 63, 155, 165–7, 169–70, 174, 137, 139, 140, 141, 156, 193, 223,
182–3, 188–9, 196, 198–9, 201, 351, 354
204, 232–3, 265, 345, 391, 616 Majorian (Roman emperor) 72, 152,
Iucundus (bishop) 71 250, 251, 253
Iudila (Visigothic king) 200–1 Malalas (historian) 86, 104
Malchus of Philadelphia (historian)
Jerome (Saint) 63, 64, 164, 166, 246, 87–9, 96
348 Marbod (Marcomannic king) 8
John of Antioch (historian) 86, 87, Marcianus (Roman emperor) 177
94, 104, 252 Marculf (monk) 43, 44
John of Biclar (historian) 165, 167, Marcus Aurelius (Roman emperor) 61
169, 173–8, 180–2, 186, 188, Marius of Avenches (bishop) 250, 251,
191–2, 222–3, 233–4, 616 253, 255, 262
Jonas of Bobbio (monk) 247, 430, Martin of Braga (bishop) 169, 188
449, 450 Masona (bishop) 172, 176, 182, 186
Jordanes (historian) 25–6, 61–2, 64, 72, Matasuentha (wife of Witigis) 94, 127
88–92, 99, 104–5, 125, 165, 169, 173, Maurikios (Roman emperor) 466, 518,
246, 250, 262, 312, 436, 584 583
Jovian (Roman emperor) 246 Menander (historian) 304, 573, 575,
Julian (Roman emperor) 152, 280, 312 577, 590
Julian of Toledo (bishop) 165, 166, Mendel 55
198, 205, 212, 215 Merovingians (Frankish dynasty) 305,
Julius Nepos (Roman emperor) 252 319, 323–6, 336–7, 341, 343, 403,
Justinian (Roman emperor) 1, 13, 16, 440, 442, 454, 606, 625
18, 23, 37, 78, 94, 109, 209, 417, Mùsà ibn Nusayr (Moorish leader)
420, 440, 457, 518, 573, 576 220–1, 225–6, 228–9, 235–6, 238,
240
Kuver (Bulgar leader) 497, 581, 587 Mu˙ammad (prophet) 224–5, 238
Kuvrat (Bulgar khan) 521, 579
Narses (Roman general) 619
Laiamicho (Lombard king) 414
Lantfrid (Alamannic duke) 40 Odilo (Bavarian duke) 40, 446–8,
Leander of Seville (bishop) 166, 169, 452–5, 458, 618
170, 188, 216 Odoacer (Germanic leader) 85–6, 99,
Leo (Roman emperor) 4, 26, 72, 86, 102–3, 108–9, 114, 117, 119, 255,
276, 277, 334, 400, 449, 524, 585 269, 437, 457, 573, 613, 619
Leth (Lombard king) 414 Offa (Mercian king) 45
Leudastus (Frankish count) 324 Olympiodorus of Thebes (historian)
Libius Severus (Roman emperor) 253 107, 137, 346
Liudolfings (Bavarian dynasty) 456 Opilio (Roman politician) 118
Liutpirc (wife of Tassilo III) 455 Ostrys (Gothic commander) 88
Liutpoldings (Bavarian dynasty) 456 Otfrid of Weißenburg (poet) 314
Liutprand (Lombard king) 38, 39,
418, 419, 421, 423 Paul the Deacon (historian) 17–8,
Liuva I (Visigothic king) 174, 176 131, 165, 184, 409–11, 413–5, 417,
Liuva II (Visigothic king) 186, 193 419, 420, 422, 426, 440–1, 447,
Liuvigild (Visigothic king) 27, 35, 36, 579, 602
162, 166, 169, 174–9, 192, 203, 264 Pinta (bishop) 71
Pippin II (Frankish maiordomus) 451,
Macrianus (Alamannic king) 244, 248 453, 455
Maglocunus (British king) 357, 359, Pippin III (Frankish king) 29, 453,
375, 376 455, 625
698   

Pitzas (Gothic leader) 99, 123 Silvanus (magister militum) 310, 311
Pliny (scholar) 59, 243 Simplicius 251
Pompeius (Roman general) 191, 340 Sisebut (Visigothic king) 37, 166, 186,
Postumus (Roman emperor) 319 193, 194
Priscus (historian) 96, 153, 304, 485 Sisenand (Visigothic king) 166, 186,
Probus (Roman emperor) 243 196–201, 204, 223
Procopius (historian) 58, 63, 67–73, Sisigis (Gothic commander) 123
75–6, 79, 81, 86–7, 94, 97–9, Sozomen (historian) 136, 346
109–13, 118–20, 123, 129, 131, Stephen (Magyar king) 523
136, 148, 150, 181, 420, 584, 587 Stilicho (magister militum) 63, 98,
Prosper (historian) 63, 67, 73, 80, 269, 613
151, 246, 250, 348, 411 Sunna (bishop) 176, 182, 186
Ptolemy (scholar) 61, 243 Swinthila (Visigothic king) 162, 166,
174, 186, 193, 195–6, 198, 200–1,
Quidila (Gothic leader) 124 204, 223, 233
Syagrius (king of Soissons) 26, 269,
Radagaisus (Gothic leader) 64, 65, 320, 321, 345, 366, 399, 611, 613
98, 100, 108
Raedwald (East Anglian king) 404 ˇàriq ibn Ziyàd (Moorish leader) 220,
Reccared I (Visigothic king) 37, 166, 221, 226, 228, 234, 235, 236, 237,
169–74, 176–78, 180, 186, 188, 238, 240
191, 193, 202, 223 Tacitus (historian) 59, 60, 61, 62, 95,
Reccared II (Visigothic king) 186 169, 276, 450, 577, 582, 603
Reccaswinth (Visigothic king) 37, 38, Talhaearn (poet) 373–4
130, 166, 187, 208–12, 217 Taliesin (poet) 373–5
Rechiarius (Suevic king) 153 Tanca 24
Rechila (Suevic king) 153 Targitios (Avar ambassador) 587
Recitach (Gothic leader) 87, 90, 93, Tassilo I (Bavarian duke) 442, 443,
94, 99, 102 447, 453, 454, 455, 460
Remigius of Rheims (bishop) 331 Tassilo III (Bavarian duke) 443, 447,
Ricimer (magister militum) 252, 253, 453, 454, 455, 460
254, 269, 613 Tato (Lombard king) 414, 422
Riothamus (British king) 399 Teias (Ostrogothic king) 94, 97, 131
Rodrigo (Visigothic king) 162, 180, Teodomir (bishop) 169
220, 223, 225, 226, 227, 228, 235 Tetricus (Roman emperor) 319
Romulus Augustulus (Roman emperor) Theodagunda 122
108, 573 Theodahad (nephew of Theoderic the
Rothari (Lombard king) 22, 34–5, Great) 16, 94, 110, 112, 116, 122,
38–40, 249, 410, 413–4, 416–7, 130
419–22, 426, 439 Theodelinde (wife of Agiluf ) 420
Rupert of Worms (bishop) 451, 459 Theoderic Strabo (Gothic leader) 87,
88, 93, 99
Sagittius of Ilerda (bishop) 138, 143–4 Theoderic the Great (Ostrogothic king)
Salegast (Frankish judge) 28 7–8, 16, 18, 25, 27, 35, 72, 76–7,
Samo (Franco-Slavic leader) 584, 593 85–95, 97–100, 102–3, 107–17,
Servatus (Roman commander) 118 119–30, 153, 190, 248, 255, 265–7,
Severus of Ilerda (bishop) 138–9, 294, 319, 432, 437, 573, 579, 607,
142–3 609, 626
Sidonius Apollinaris (bishop) 26, 136, Theodo (Bavarian duke) 437, 446,
145, 250, 251, 253, 345 447, 450, 451, 452, 458, 618
Sigibert (Frankish king) 336 Theodoric I (Visigothic king) 26, 250
Sigibert (king of Cologne) 320 Theodosius (Roman emperor) 13, 74,
Sigismund (Burgundian king) 26, 27, 88, 128, 136
28, 245, 247, 254–9, 261–9 Theophanes (historian) 86, 88, 585
    699

Theophylact Simocatta (historian) Venerius 24

573, 575–7, 579, 586–7, 590 Verinianus 137, 345
Theudebald (Frankish king) 440, 441, Victor of Vita (bishop) 58, 67–71,
445, 457 73–5, 77–80, 148, 264
Theudebert I (Frankish king) 294, Victorianus (proconsul) 75
323, 336, 438, 439, 440, 457 Vidimer (Gothic king) 89
Theudebert II (Frankish king) 323 Vinitharius (Gothic king) 90
Theudis (Visigothic king) 125, 181, Virgil (poet) 340, 622
192, 193, 203, 213, 215 Visimar (Vandal ruler) 61
Thiudimer (Gothic king) 85, 86, 89 Vitarit (notary) 74
Thorismund (Visigothic king) 193 Vithimiris (Gothic king) 90
Thrasamund (Vandal king) 74, 97 Vortigern (British king) 354, 369, 381
Tiberius (Roman emperor) 245 Vortiporix (British king) 357, 358,
Totila (Ostrogothic king) 94, 97, 99, 359, 371
131
Trapsta 266 Wacho (Lombard king) 414, 432
Tulga (Visigothic king) 206 Walderada (Lombard princess) 440
Tuluin 94, 99, 112 Wamba (Visigothic king) 166, 174,
198, 199, 210, 212, 213, 214, 215,
Ullibos (Gothic commander) 88 217
Umayyads (Moorish dynasty) 230, Widogast (Frankish judge) 28
236 Wihtred (Kentish king) 34, 42
Urais (nephew of Witigis) 112 Wisogast (Frankish judge) 28
Urban of Llandaff (bishop) 364 Witeric (Visigothic king) 186, 189,
Ursus (Roman commander) 118, 127 193
Witigis (Ostrogothic king) 94, 112,
Vadamerca (wife of Valamer) 90 123, 131, 457
Valamer (Gothic king) 89, 90, 125, Witiza (Visigothic king) 37, 162, 229
604 Wodan (Germanic God) 385, 392,
Valentinian I (Roman emperor) 136, 393, 414, 420
244
Valentinian III (Roman emperor) 74, Zeno (Roman emperor) 100–3
81, 119 Zosimus (historian) 14, 63, 65, 107,
Velleius Paterculus (historian) 409, 136, 137, 243, 277, 280, 346, 347,
424 349, 395, 397
 INDEX OF PLACES

Abaritana 69 Belgica 283, 316, 318

Aeron 359 Belgium 280
Africa 16, 30, 43, 57–8, 61, 63, Benevento 418, 579
65–73, 75–6, 78–9, 81–2, 107, 136, Bernicia 361
138, 147–50, 152–3, 156, 219–20, Bischofshofen 448
224, 226, 230, 234, 238–9, 243, Black Sea 91, 104, 105, 470, 472,
603, 605, 608–9, 611, 614, 617, 479, 484, 490, 510, 522, 578, 579,
621, 623 591, 603
Agaune 256, 264, 266, 268 Bócsa 482, 483, 521, 593
Aguntum 438 Bohemia 411, 431, 432, 433, 434,
Alamannia 288, 296, 325, 327 439, 441, 457, 502, 584, 593
Albania 477, 509, 522 Börcs-Nagydomb 469, 488
Aleppo 510 Bordeaux 23, 285
Alps 97, 110–2, 123, 432, 434, 435, Bracara, Bracara Augusta 152, 153, 154
438, 439, 442, 457, 458, 476, 504 Bratislava 431, 469, 495
Alsace 454 Brioude 262
Alsópáhok 475 Britain 21, 23, 34, 36, 45, 60, 137,
Altai mountains 575 225, 246, 316, 345–407, 604, 610,
Aquileia 475 613. See also England
Aquincum 476 Britannia Prima 398
Aquitaine, Aquitania 3, 63, 67, 73, 80, British Isles 147, 387, 571
137, 141, 220, 246, 320, 325–6, Brittany, Armorica 147, 347
327, 333, 337, 352 Bruttium, Brettia 346
Arabia 232 Brycheiniog 360, 379
Armorica: see Brittany Budakalász 470, 473
Aschheim 455 Bulgaria 21, 472, 507
Asia 481, 573, 575, 591 Burgundy, Burgundia 3, 243, 247, 251,
Asturias 162, 221, 619 254, 255, 260, 262, 263, 267, 268,
Asturica 152 321, 325–7, 332, 337, 343, 344,
Atlantic 140, 147, 398 450, 617
Augsburg 293, 294, 295, 297, 305, Byzantium 11, 21, 38, 190, 224, 226,
430, 438, 449, 450 255, 426, 457, 470, 472, 475, 481,
Austrasia 325, 326, 343 484, 489, 504, 517, 519, 521, 573,
Austria 424, 450, 457, 510, 589 578, 583, 584, 586, 592, 608, 614,
Auvergne 262 625

Baetica 66, 140, 146, 147, 152, 200 Cadbury 362

Baghdad 238 Cadiz 152
Baia 430, 431, 433, 435, 441 Caesaraugusta 144, 152, 155
Balearic Islands, Insulae Balearum 67, Calabria 476
138, 147 Cannington 361
Balkans 79, 86, 87, 89, 91, 93, 95, Canterbury 385, 394, 403
96, 97, 99, 101–7, 110, 122, 123, Carinthia 584
126, 472, 510, 522, 581 Carnuntum 475
Barcino 152, 153 Carouge 254
Bavaria 288, 325, 327, 433, 445, Carpathian Basin 463, 465, 469–72,
447, 460, 477, 522, 523, 579, 607, 474, 475, 479, 481, 486, 490, 495,
610, 618, 624 497, 503, 507, 509, 515, 517–9,
    701

522, 572, 573, 577, 579–81, 589, Emerita 147, 152, 153
591, 592, 594, 595 England 21, 42, 277, 303, 316, 353,
Carpathians 91, 582 356, 360, 361, 373, 379, 381–407,
Carthage 58, 72, 73, 75, 76, 78, 79, 603, 609, 612, 613, 620, 627.
81, 82, 106, 611 See also Britain
Carthaginiensis 66, 146, 188 English Channel 403–4
Carthago Nova 147, 152 Eperjes 484, 512, 513
Castille 163 Escaut: see Scheldt
Caucasus 469, 470, 479, 574 Essex 367
Celtica 347, 371 Etruria 130
Chalcedon 177 Eurasian steppe 486, 572, 575
Chartres 286 Europe 1, 2, 4, 5, 22, 29, 35, 42, 45,
Cheshire 356 46, 95, 98, 104, 108, 120, 161, 194,
Chester 375 222, 243, 248, 257, 308, 336, 339,
Cirencester 398 366, 386, 387, 388, 389, 398, 402,
Cologne 85, 286, 314 403, 407, 413, 417, 480, 481, 484,
Coninbriga 152 487, 497, 504, 518, 523, 571–7,
Constantinople 18, 19, 76, 85, 86, 582, 583, 585, 588, 595, 617
93, 99, 102, 103, 105–7, 113, 116,
125, 127, 131, 153, 178, 188, 215, Feddersen Wierde 279, 290, 291
257, 426, 482, 484, 490, 504, 517, Flavia Caesariensis 398
521, 574, 575, 579, 587, 593, 625 France 96, 105, 282, 284, 285, 300,
Cordoba 152, 179, 180, 226, 228, 305, 308, 315, 316, 317, 328, 330,
229, 236, 238 334, 609, 611, 620
Cornwall 357, 379, 601 Francia 3, 42, 44–5, 96, 198, 212–3,
Crimea 510 229, 233, 247, 277, 286, 308, 317,
Croatia 508, 510 329, 333, 336–8, 394, 403, 441
Freising 446, 451
Dacia 61, 277 Furfooz 282
Dalmatia 97, 99, 587
Damascus 220, 221, 224, 228, 229, Gaetulia 69
230, 231 Gallaecia 65, 66, 140, 146, 147, 151,
Danube 279, 339, 430–5, 438, 441, 169, 172, 174, 178, 180, 188, 196,
444, 450, 452, 457, 470, 474, 484, 197, 250
516, 522, 573, 574, 579, 582, 583, Gallia Narbonensis: see Narbonne
584, 594, 603 Gallia Ulterior 252
Dartmoor 356 Gaul, Gallia 3, 15, 25, 26, 28, 29, 32,
Dee 358 33, 36, 39, 42, 43, 46, 59, 63, 64,
Denmark 291, 385, 402 65, 70, 75, 105, 111, 119, 137, 147,
Derbyshire 356 152, 153, 170, 172–5, 178, 180,
Devon 356 184, 187, 192, 196, 197, 198, 212,
Dieue-sur-Meuse 300–2 216, 243, 246, 248, 250–3, 255,
Dingolfing 445, 455 258, 262, 267, 274, 278, 279,
Dorset 356 281–5, 291, 303–5, 307, 308, 311–3,
Dumnonia 361, 376 315–21, 325, 327, 328, 330, 332,
Dyfed 358–9, 379 333, 340, 347, 349, 350, 352, 360,
Dyrrachium 522 366–8, 378, 380, 399, 401, 460,
590, 600, 603, 607, 608, 610, 611,
East Anglia 367, 392, 403 615, 622
Ebro 138 Germany, Germania 3, 59, 61, 69, 95,
Egypt 235, 236, 244, 367 96, 169, 245, 272, 274, 278, 280,
Eichstätt 453 281, 283, 293–5, 305, 306, 308,
Elbe 61, 278, 316, 409, 411, 412, 315, 316, 330, 333, 334, 340, 352,
413, 416, 417, 420, 431, 591, 603 385, 386, 387, 389, 400, 450, 474,
702   

477, 489, 494, 518, 577, 582, 588, Komárno 506, 511, 513, 520
597, 599, 600, 601 Krungl 589
Glamorgan 356 Kunbábony 482, 483, 513, 521, 594
Gloucestershire 356 Kunszentmárton 472, 479, 480, 481,
Golaida 413, 417 488, 489
Greece 62, 77
Gwent 356 Lake Balaton 89, 472, 474, 510, 519,
Gwynedd 358, 359, 375 592
Gyenesdiás 472, 489, 490, 491, 492, Lake Geneva 246, 609
495 Lake Leman: see Lake Geneva
Lancashire 360
Hadrianople 100, 101 Lech 286, 305
Haemus Mountains 101 Legio 152
Hampshire 356, 388, 391, 393, 394, Leobersdorf 473, 486, 487, 488, 499,
403 502, 504, 514, 520
Herefordshire 356 León 155, 162, 165
Hispalis: see Seville Liguria 111, 112, 346
Hispania: see Spain Lincoln 396
Hohenberg 508, 509, 589 Lindsey 396
Huesca, Osca 138, 143, 144 Llandaff 364, 366, 379
Hungary 61, 113, 469–72, 474, 483, Lleida, Ilerda 138, 142, 143, 144
486, 488, 490, 502, 506, 512, 513, Loire 278, 281, 283, 316, 320
514, 519, 523 Lotharingia 300
Lower Saxony 276, 277, 278, 279,
Iberian Peninsula 135, 136, 139–41, 288
145, 146, 148–52, 154, 157, 198, Lucus 152, 153
216 Lusitania 66, 140, 146, 151, 152, 200
Igar 488, 489, 490, 493, 496 Lyons 253, 257
Ile-de-France 337
Ilerda: see Lleida Maghreb 235, 237, 239
Illerup 291 Main 338
Innichen 447 Marne 286
Insulae Balearum: see Balearic Islands Mauriac Plains 246, 248, 249, 250
Ireland 77, 363, 371 Mauritania Tingitana 140, 147, 157
Isle of Wight 393, 394, 403 Maxima Caesariensis 398
Italy, Italia 1, 3, 16, 18, 24, 26, 28, Maxima Sequanorum 246, 249
30, 34–6, 38–9, 43, 46, 63, 64, 75, Medina 235
76, 85–7, 91–5, 97–9, 102, 103, Mediterranean 15, 34, 46, 72, 147,
106–11, 113–8, 120–3, 125–9, 152, 153, 255, 307, 362, 378, 425,
131–2, 175, 243, 246, 252, 254–5, 471, 497, 503, 505, 507–11, 519,
262, 265, 267, 312, 320, 325, 338, 521, 522, 593
344, 346, 352, 380, 413–5, 417–26, Melun 247, 452
457, 471–2, 474–7, 489, 494, 497, Mercia 45, 358, 391
504, 510, 518–9, 522, 573, 579, Mérida 151, 180, 182, 211
592, 594, 603–4, 607, 609, 611–4, Meuse 286, 300, 316
618–9, 621, 624 Mödling 490, 494–7, 499
Mons Badonicus 350, 351
Kent 33, 37, 381, 387, 388, 392–6, Montenegro 476
403, 404, 610 Moravia 431, 433, 439, 502, 510, 593
Keszthely 463, 474–7, 488, 492, 495, Moselle 315
510, 519, 521, 580, 591 Mount Sondis 96, 97
Kirkmadrine 360
Kölked-Feketekapú 473, 474, 477, Naab 450
478, 486, 488, 506 Nagyszentmiklós 515–6
    703

Naples 111, 112, 131 Regensburg 431, 433–7, 439, 440,

Narbonne, Gallia Narbonensis 26, 170, 442, 445, 446, 448, 450–3, 456,
174, 186, 187, 220 458, 459, 460
Navarre 222 Rheged 601
Netherlands 276–8, 280 Rhine 29, 32, 60, 63–6, 245, 246,
Neuburg 295, 452 248, 249, 250, 277, 281, 283, 286,
Neuching 455 299, 300, 308, 309, 312, 315, 316,
Neustria 247, 325, 343 320, 339, 340, 347, 454, 577, 583,
Nicaea 125, 177 589
Nordgau 439, 453 Rhône 267, 320
Noricum 118, 124, 149, 429, 432, 433, Rieti 112, 124
435, 436, 438, 456, 457, 460, 604 Romania 19, 61, 315
North Sea 289, 309, 377, 385, 387, Rome 2, 13, 15, 18, 21, 33, 40, 41,
389, 394, 399, 402–5 43, 58, 72, 107, 111–3, 116, 131,
Northumbria 45 136, 139, 144, 145, 153–5, 185,
Norway 388 190, 191, 193, 194, 199, 216, 223,
Numidia 69, 75 224, 246, 250, 251, 253, 264, 323,
Nursia 124 336, 343, 346, 348, 350, 381, 390,
392, 395, 405, 447, 459, 471, 477,
Osca: see Huesca 508, 572, 606
Rouen 286, 303, 332
Palestine 239 Rugiland 414, 417, 424
Pannonia 64, 86, 89, 94, 340, 414,
417, 424, 425, 431, 432, 437, 439, Salzburg 430, 432, 435–7, 442,
469, 470, 471, 475, 477, 576, 580, 446–8, 450–3, 456, 458, 459
581, 587, 591, 592, 595 Samnium 111, 112, 120, 123
Passau 433, 446, 451 Sapaudia 245–50, 252, 258, 259, 261,
Pavia 131, 253, 410, 418, 419, 422, 262, 604, 609, 622
426, 613 Saragossa 42, 186, 220, 222
Pécs 519, 521 Scandinavia 21, 386, 387, 393, 412,
Phrygia 339 603
Picenum 111, 112, 120 Scheldt (Escaut) 316
Pitvaros 478, 513 Schleswig-Holstein 386–7
Pliska 522 Scotland 356, 360, 369, 378, 379
Po 131, 449, 493, 506 Seine 286, 316, 320, 413
Poitiers 220, 451 Septimania 180, 186, 207
Poland 77 Severn 363
Pompaelo 152 Seville, Hispalis 152, 153
Portugal 147, 219, 220 Shropshire 356
Portus Cale 153 Silesia 61, 603, 608
Powys 356, 358, 375 Sirmium 118, 476, 483, 484, 581,
Preslav 522 586, 592
Provence 4, 220, 321, 327 Slovakia 431, 506, 510, 513
Pyrenees 64, 65, 220, 619 Somerset 356, 361, 362, 376
Somme 282, 283, 316
Queensford Farm 361, 362, 384 Sommerein 478, 488, 495
Spain, Hispania 3, 11, 15, 30, 37–9,
Raetia 63, 118, 127, 301, 305, 429, 43, 46, 64–5, 70, 74–5, 83, 106,
431, 433, 435, 436, 438, 440, 441, 108, 119, 125, 132, 135–240, 336,
456, 460, 604 345, 347, 349, 352, 380, 442, 601,
Ravenna 25, 42, 110–2, 116, 120, 603, 605, 607–9, 611–3, 616,
127, 129, 313, 346, 431, 457, 613 618–23, 626
Reccopolis 179 Spoleto 18, 33, 39, 111, 116, 165,
Regen 450 223, 303, 414, 418, 423, 592
704   

Stadevanga 450 Vermand 279, 282, 291

Staffordshire 356 Verona 97, 131
Straits of Gibraltar 147 Verulamium 363, 384, 399
Strathclyde 601 Vézeronce 262, 268
Straubing 433, 434, 436, 458, 459 Via Augusta 152
Styria 508, 584 Via de la Plata 152
Sudetenland 61 Via Herculea 152
Sutton Hoo 388, 390, 392, 403 Vienne 253–6, 264–6, 268–9, 329,
Sweden 392 331
Syria 222, 232, 510 Vindolanda 360
Vireux-Molhain 282
Tarraco 138, 143, 144, 152, 153 Virunum 438
Tarraconense 138, 140, 141, 142, Vouillé, Vogladum 118, 119, 161, 320,
144, 147, 154, 155, 157, 202, 207, 619
208, 345 Vron 282
Tepe 471, 482, 483, 521
Teurnia 438 Waal 316
Thrace, Thracia 87, 88, 91, 104 Wales 353, 355, 358–61, 364, 365,
Thuringia 325, 327, 454 369, 370, 376, 379, 380, 397, 601
Tisza 470, 488, 581 Weser 279
Toledo 38, 130, 161, 162, 164–6, Wessex 33, 356, 376, 390, 391, 396,
168–77, 179, 180, 182, 183, 185, 397, 404, 621
187–90, 193, 195–8, 202, 204, 205, Westphalia 276–8, 288
207, 208, 210, 212, 214–7, 220, Whithorn 360
221, 225, 227, 228, 231, 235, 613, Wight 390, 391, 393, 394, 403
616, 625 Worcestershire 356
Toulouse 28, 141, 155, 229, 230, Worms 247, 248, 271, 451
252, 285, 625 Wremen 272, 289, 290
Tournai 279, 286, 305, 313, 319, Wroxeter 363
320, 606
Trier 271, 295, 301, 319 Xanten 277, 286
Troy 224, 339, 340
Yemen 232, 239
Ulixipo 152 Yorkshire 356, 389

Valence 252 Zamárdi 470, 473, 478, 494, 513,

Vandal mountains 61 580
Varna 522 Zeugitana 69
Venetia 111, 123 Zeuzleben 299, 300
Venice 435, 438, 439 Zillingtal 478, 486, 487, 488, 489
 INDEX OF SUBJECTS

[The index covers only the main categories within the scope of this volume.]

Acculturation 13–9, 23–8, 69–72, 437–41, 443–4, 574–82, 603–9,

128–32, 144–50, 172–5, 226–9, 622–3
265–6, 285–7, 300–4, 321–2,
332–5, 369–70, 396–9, 435–7, Justice/Law 21–53, 121–2, 206–9,
448–53, 456–9, 473–4, 517, 591–3, 257–8, 313–4, 327, 338–9, 445–8
612–21
Administration 72–6, 114–24, 320–1, Kingdom: see Realm
327–30, 395–401, 417–9, 439–41, Kingship 72–6, 189–91, 201–2,
445–53, 594–5, 619 229–31, 254–7, 261–53, 311–4,
Arianism 69–72, 77–81, 173–7, 322–7, 356–60, 366–7, 410–6,
330–2, 419–21, 448–50 422–4, 442–3, 453–6, 577, 586–8,
Aristocracy: see Nobility 616–7, 623–5
Army 67–8, 85–106, 334, 421–2, 465–8
Language 181–6, 314–6, 360–1,
Catholicism 77–81, 168–70, 191–3, 370–3, 377–9, 443–4
330–2, 419–21, 448–53, 519–21 Law: see Justice
Christianity/Christian Church 77–81,
125–8, 164–7, 192–3, 198–202, Migration 13–4, 61–8, 86–106,
206–8, 212–5, 330–2, 419–21, 139–42, 245–50, 317, 339–41,
443–4, 448–53, 616–8 431–7, 497, 572–4
Cities 142–4, 153–4, 395–9, 446–8, Military: see Army
591 Monarchy: see Kingship
Conversion 79–81, 125–8, 330–2,
406, 419–21 Nobility 73–4, 97–9, 181, 190, 207,
Culture/Customs 42–4, 142–4, 287–92, 328–9, 585
272–4, 317–8, 355–7, 373–6,
386–9, 469–73, 487–8, 496–7, Realm/Kingdom 69–76, 81–3,
522–3, 588–94 175–81, 196–8, 226, 305–6, 319–22,
390–5, 401–7, 412–6, 584–8,
Empire 13–9, 23–8, 60–8, 81–3, 609–17
86–106, 135–7, 151–4, 173–6, Romanization: see Acculturation
251–3, 312–3, 343, 354–5, 402–5,
424–6, 456–9, 481–2, 521–2, 625–6 Settlement 108–14, 139–48, 234–8,
Ethnogenesis/Ethnicity 28–37, 55–62, 246–50, 258–60, 281–5, 311–3,
81–3, 86–106, 128–32, 167–9, 316–8, 362–5, 458, 484–5,
171–6, 187–95, 214–5, 224–6, 609–11
232–4, 265–7, 277–81, 305–6, Society 97–9, 108–14, 237–9, 248–9,
308–11, 319, 336–44, 373–9, 287–8, 327–30, 365–71, 382–5,
385–95, 409–12, 416–7, 429–31, 512–5, 517–8, 578–82