STAYING ROM AN

What did it mean to be Roman once the Roman empire had collapsed
in the West? Staying Roman examines Roman identities in the region of
modern Tunisia and Algeria between the fifth-century Vandal conquest
and the seventh-century Islamic invasions. Using historical, archaeological, and epigraphic evidence, this study argues that the fracturing of the
empires political unity also led to a fracturing of Roman identity along
political, cultural, and religious lines, as individuals who continued to feel
Roman but who were no longer living under imperial rule sought to
redefine what it was that connected them to their fellow Romans elsewhere. The resulting definitions of Romanness could overlap, but were
not always mutually reinforcing. Significantly, in late antiquity, Romanness had a practical value, and could be used in remarkably flexible ways
to foster a sense of similarity or difference over space, time, and ethnicity,
in a wide variety of circumstances.
jonathan conant is Assistant Professor of History at Brown University,
where his teaching and research focus on the early medieval Mediterranean.

Cambridge Studies in Medieval Life and Thought

Fourth Series
General Editor:
rosamond mckitterick
Professor of Medieval History, University of Cambridge, and Fellow of Sidney Sussex College

Advisory Editors:
christine carpenter
Professor of Medieval English History, University of Cambridge

jonathan shepard

The series Cambridge Studies in Medieval Life and Thought was inaugurated
by G. G. Coulton in 1921; Professor Rosamond McKitterick now acts as
General Editor of the Fourth Series, with Professor Christine Carpenter and
Dr Jonathan Shepard as Advisory Editors. The series brings together outstanding
work by medieval scholars over a wide range of human endeavour extending
from political economy to the history of ideas.
A list of titles in the series can be found at:
www.cambridge.org/medievallifeandthought

STAYING ROM AN
Conquest and Identity in Africa and
the Mediterranean, 439700

JONATHAN CONANT

cambridge university press

Cambridge, New York, Melbourne, Madrid, Cape Town,
Singapore, Sao Paulo, Delhi, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge cb2 8ru, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521196970

c Jonathan Conant 2012

This publication is in copyright. Subject to statutory exception and

to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2012
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Conant, Jonathan, 1974
Staying Roman : conquest and identity in Africa and the Mediterranean, 439700 /
Jonathan Conant.
p. cm. (Cambridge studies in medieval life and though: fourth series ; 82)
Includes bibliographical references and index.
isbn 978-0-521-19697-0 (hardback)
1. Romans Africa, North. 2. Africa, North History To 647. 3. National characteristics,
Roman. 4. Africa, North Civilization Roman influences. 5. Africa, North Antiquities,
Roman. 6. Inscriptions, Latin Africa, North. I. Title.
dt170.c65 2012
939 .704 dc23
2011047925
isbn 978-0-521-19697-0 Hardback
Cambridge University Press has no responsibility for the persistence or
accuracy of URLs for external or third-party internet websites referred to
in this publication, and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.

To Vanessa

CONTENTS

List of figures
List of maps
List of tables
Acknowledgements
List of abbreviations

1
2
3
4
5
6
7

page viii
ix
x
xii
xv

introduction
the legitimation of vandal power
flight and communications
the old ruling class under the vandals
new rome, new romans
the moorish alternative
the dilemma of dissent
aftermath
conclusions

1
19
67
130
196
252
306
362
371

Bibliography
Index

379
420

vii

F I G U RE S

page 69
Africans abroad, 439533: churchmen and laity
Africans abroad, 439533: occupations
69
Africans abroad, 439533: distribution over time (by group)
76
The names Adeodatus and Benenatus in dated Italian
inscriptions, 350599 (CIL and ICVR n.s.)
123
2.5 The names Adeodatus and Benenatus, 410534 in PCBE 2
(Italy)
124
2.1
2.2
2.3
2.4

viii

MAPS

1.1
1.2
1.3
2.1
2.2

The Mediterranean world

Late antique North Africa, 439700
Vandals in North Africa? archaeological evidence
Africans abroad in the Mediterranean, 439533
Latin African names in Mediterranean inscriptions
(CIL only)

ix

page 20
22
50
77
120

TA BL E S

page 30
1.1 Roman embassies to the Vandals, 45584
2.1 Africans abroad, c.439c.533: Constantinople and
Chalcedon
78
2.2 Africans abroad, c.439c.533: The East other than
Constantinople and Chalcedon
82
2.3 Africans abroad, c.439c.533: Rome
83
2.4 Africans abroad, c.439c.533: The West other than Rome
84
2.5 Travel from the African interior to the nearest port,
c.439c.533
96
2.6 The travels of Fulgentius of Ruspe, c.484c.532
101
2.7 Comparison of nine names in PCBE 12 (Africa and Italy)
before the Vandal capture of Carthage, c.300439
116
2.8 Comparison of nine names in PLRE 12 before the
Vandal capture of Carthage, c.260c.439
117
2.9 Comparison of nine names in PLRE 23 in the Vandal
period, c.439533
117
2.10 Comparison of nine names in PCBE 2 (Italy) before and
after the Vandal capture of Carthage
118
2.11 Adeodatus/Adeodata: comparison between provinces
(CIL only)
121
2.12 Adeodatus/Adeodata: Rome (ICVR n.s. only)
121
2.13 Benenatus/Benenata: comparison between provinces
(CIL only)
121
2.14 Benenatus/Benanata: Rome (ICVR n.s. only)
122
2.15 Adeodatus/-a and Benenatus/-a beyond Africa: dated
inscriptions
123
3.1 Romano-African families in the late Roman and Vandal
administration
146
3.2 Consular and Imperial dating systems in late Roman
Africa, c.40754
151
3.3 Vandal regnal dates in African inscriptions
153
3.4 Anno and Anno Karthaginensis dates in African inscriptions
154
x

List of tables
4.1 Supreme commanders of the Byzantine forces in Africa:
regional origins
4.2 Praetorian prefects of Byzantine Africa: regional origins
4.3 Constantinople to Africa: late ancient itineraries
4.4 Supreme commanders of the Byzantine forces in Africa:
previous careers
4.5 Duces of Egypt, c.538641: previous and subsequent careers
4.6 Sixth- and seventh-century commanders in Africa:
previous careers
4.7 Early commanders in Africa: previous careers
4.8 Subordinate commanders in Italy: terms of appointment
5.1 Secular office-holders in Moorish Africa, fifth to seventh
centuries: the epigraphic evidence

xi

202
204
215
218
221
227
229
238
293

A C K N OW L E D G E M E N T S

I have incurred many debts of gratitude in writing this book. First thanks
go to Michael McCormick, who oversaw the Ph.D. dissertation on which
it is based. Every aspect of this work has been improved by his demanding,
insightful, probing, meticulous, and erudite comments, questions, and
suggestions. I would also like to thank my readers, Thomas N. Bisson
and Christopher P. Jones, who pushed me to rethink a number of my
conclusions, to dig deeper into the sources, and to formulate new answers
to the challenging questions that they posed. I owe my introduction to
field archaeology and indeed the original inspiration for the project
to Susan T. Stevens and a summer spent on the Bir Ftouha (Carthage)
excavation.
In the process of turning that inspiration into a book, I have benefited
greatly from the generous funding of a number of institutions. A Jacob
K. Javits Fellowship, a Packard Fellowship, and a Dumbarton Oaks Junior
Fellowship in Byzantine Studies made possible the research and writing of
the dissertation. The American Numismatics Society Summer Seminar
in Numismatics provided me with an essential grounding in the use of
coins as a source for the history of late antiquity and the early Middle
Ages. My participation in the Bir Ftouha excavation was enabled by
a Harvard University History Department Travelling Fellowship and a
Harvard University Graduate Society Summer Award for pre-dissertation
prospectus research. The revision and rewriting necessary to turn the
dissertation into a book were made possible by the American Academy
in Rome and the University of San Diego, which granted me a year
of leave and also generously funded my research with numerous Faculty
Research Grants.
Over the years, I have presented different aspects of my work at a
series of scholarly forums, including the American Academy in Rome,
the Catholic University of America, the Constantines Dream Project
(Hawarden, Wales), Oxford University, the Deutsches Archaologisches
Institut (Rome), Dumbarton Oaks, the Ianiculum Workshop (Rome),
Harvard University, the Institute of Historical Research (London), the
xii

Acknowledgements
International Congress on Medieval Studies (Kalamazoo, Michigan),
the International Medieval Congress (Leeds), Shifting Frontiers in Late
Antiquity, Syracuse University, the Universit`a degli Studi di Messina,
and Yale University. My work has benefited greatly from the many questions and comments that I received at all of these meetings, which I
acknowledge with gratitude.
I have also gained from the expertise and generosity of friends and colleagues. Adam J. Kosto and Warren C. Brown read early versions of this
work in its entirety and offered many useful suggestions for revision and
reorganization. Leslie Dossey and Mark Handley graciously shared with
me the fruits of their outstanding research in advance of its publication,
in each case reshaping my own work. I owe much, conceptually and substantively, to my conversations, communications, and collaborations with
Guido Berndt, Jonas Bjornebye, Ralf Bockmann, Andy Merrills, Philipp
von Rummel, Roland Steinacher, Alicia Walker, Ann Marie Yasin, and
Christine Zitrides Atiyeh. Jennifer Ball, Michael L. Bates, David Cook,
Nathaniel Cutajar, Guy Halsall, Kyle Harper, Emmanuel Papoutsakis,
Deborah G. Tor, and Robin Whelan all shared aspects of their diverse
expertise on the late ancient and early medieval world with me. I have
learned much from sustained or occasional discussions with Vincenzo
Aiello, Kim Bowes, Mike Clover, Kate Cooper, Florin Curta, Elizabeth
Fentress, Carmela Vircillo Franklin, Caroline Goodson, Walter Kaegi,
Michael Kulikowski, Conrad Leyser, Cecile Morrisson, Michele Salzman, Irfan Shahd, and Alice-Mary Talbot. My thoughts on every aspect
of this work were deepened by Diliana Angelova, Jennifer Davis, Brian
DeLay, Gregory A. Smith, Daniel Gutierrez, Edward Miller, Elizabeth
Mellyn, and Edward Watts. I am also especially grateful for the support
of Sahr Conway-Lanz, Daniel Fitzgerald, Susan Gundersen, and Jenifer
Van Vleck. As a visiting assistant professor at Columbia University in the
City of New York in 20045 I had the opportunity to teach a seminar
on early medieval North Africa, and I gained greatly from the exchanges
and insights of the students in that class.
I owe many thanks to the dedicated librarians of the American
Academy in Rome, the Bodleian, the Deutsches Archaologisches Insti
tut (Rome), Dumbarton Oaks, the Ecole
Francaise de Rome, UCLA,
UCSD, the University of San Diego, and Widener, within all of whose
collections I have worked on this project but above all to Deborah Brown Stewart, Paolo Brozzi, Rebecka Lindau, Linda Lott, and
Toni Stephens. Giulia Barra worked wonders obtaining access for me to
archaeological sites in Italy that are not generally open to the public; and I am particularly grateful to the Catacombe di Napoli, the
xiii

Acknowledgements
Commune di Roma, and the Pontificia Commissione di Archeologia
Sacra for granting that access.
Special thanks are due to Rosamond McKitterick, whose challenges,
comments, and observations on countless points of detail, context, and
interpretation have inestimably improved this book; to the anonymous
readers from Cambridge University Press; and to my editor, Liz FriendSmith, for her exceptional patience and generosity.
It goes without saying that this project has been nourished immeasurably by the love and kindness of my family. My mother, Barbara Conant,
and my father and stepmother, Roger and Shirley Conant, have provided me with warmth, encouragement, and unwavering support over
the years. Anyone who has met me knows how much I love my native
city of Chicago, and also how fortunate I am in my sister, Rebecca
Conant, my stepbrother Christian Fredrickson, my stepsister Abigayle
Shay, and their families, all of whom live there, and whom I do not visit
often enough. But, above all, I owe more than words can express on
the page or in person to my wife Vanessa and my daughter Evie.

xiv

A B B REV I A T I O N S

AASS
ACOec.
ACOec.2
AE
AL
BAR
BCTH
Cass. Var.
CCSG
CCSL
CIL
CJ
Coripp. Ioh.
CPL

Acta Sanctorum quotquot toto orbe coluntur: vel a catholicis

scriptoribus celebrantur (Paris, 1863 )
E. Schwartz and J. Straub (eds.), Acta Conciliorum
Oecuminicorum, 4 vols. (Berlin, 191483)
R. Riedinger (ed.), Acta Conciliorum Oecuminicorum
series secunda (Berlin, 1984 )
LAnnee epigraphique: revue des publications epigraphiques
relatives a lantiquite romaine
Anthologia Latina, ed. D. R. Shackleton Bailey, vol. 1,
Carmina in codicibus scripta, fasc. 1: Libri Salmasiani
aliorumque carmina (Stuttgart, 1982)
British Archaeological Reports
Bulletin archeologique du Comite des travaux historiques et
scientifiques
J. Fridh, in
Cassiodorus, Variarum libri xii, ed. A.
Magni Aurelii Cassiodori Senatoris Opera, pars I, CCSL
96 (Turnhout, 1973)
Corpus Christianorm, Series Graeca (Turnhout,
1977 )
Corpus Christianorum, Series Latina (Turnhout,
1953 )
Corpus Inscriptionum Latinarum (Berlin, 1862 )
Codex Justinianus, ed. P. Kruger, in Corpus Iuris
Civilis, ed. P. Krueger, T. Mommsen, R. Scholl, and
W. Kroll, vol. 2 (Berlin, 1892)
Flavius Cresconius Corippus, Iohannis seu de bellis
Libycis Libri VIII, ed. J. Diggle and F. R.D. Goodyear
(Cambridge, 1970)
E. Dekkers, Clavis Patrum Latinorum qua in Corpus
Christianorum edendum optimas quasque scriptorum
recensiones a Tertulliano ad Bedam, CCSL (3rd edn;
Steenbrugge, 1995)
xv

List of abbreviations
CSCO

Corpus scriptorum Christianorum Orientalium

(Louvain, 1903 )
Scr. Arab. Scriptores Arabici
Scr. Syr. Scriptores Syri

CSEL

Corpus Scriptorum Ecclesiasticorum Latinorum

(Vienna, 1866 )
CTh
Codex Theodosianus, T. Mommsen and P. Meyer,
2 vols. (Berlin, 1905)
Duval, Hadra 1 N. Duval, Recherches archeologiques a` Hadra, 2 vols.,

Collection de lEcole
francaise de Rome 18/12
(Rome, 197581), vol. 1, Les inscriptions chretiennes
Greg. Ep.
Gregory I, Registrum Epistularum, ed. D. Norberg,
2 vols., CCSL 140140a (Turnhout, 1982)
IAM 2
M. Euzennat and J. Marion, Inscriptions antiques du
Maroc, vol. 2, Inscriptions latines (Paris, 1982)
ICVR n.s.
Inscriptiones Christianae urbis Romae septimo saeculo
antiquiores, nova serie (Rome, 1922 )
ILAlg.
S. Gsell and H.-G. Pflaum (eds.), Inscriptions latines
de lAlgerie, 2 vols. (Paris, 192276)
ILCV
E. Diehl, Inscriptiones Latinae Christianae Veteres,
3 vols. (Berlin, 192531)
JRS
Journal of Roman Studies
JThS
Journal of Theological Studies
Just. Nov.
Justinian, Novellae, ed. R. Scholl and W. Kroll, in
Corpus Iuris Civilis, ed. P. Krueger, T. Mommsen,
R. Scholl, and W. Kroll, vol. 3 (Berlin, 1895; repr.
1954)
MGH
Monumenta Germaniae Historica (Hanover and
Berlin, 1819 )
AA
Auctores Antiquissimi
Chron. Min. Chronica Minora, 3 vols. = AA 9,
11, and 13
Epist.
Epistolae
Epist. Select. Epistolae Selectae
Poet.
Poetae Latini aevi Carolini
SRG
Scriptores rerum Germanicarum
SRL
Scriptores rerum Langobardicarum
SRM
Scriptores rerum Merovingicarum
Morizot 1989

P. Morizot, Pour une nouvelle lecture de lelogium de

Masties, Antiquites africaines 25 (1989), pp. 26384
xvi

List of abbreviations
Notitia

PBE
OCD
ODB
PCBE

Notitia provinciarum et civitatum Africae, ed.

M. Petschenig, in Victoris episcopi Vitensis historia
persecutionis Africanae provinciae, CSEL 7 (Vienna,
1881), pp. 11534
Prosopography of the Byzantine Empire,
ed. J. R. Martindale (Ashgate, 2001) (CD-Rom)
The Oxford Classical Dictionary, ed. S. Hornblower and
A. Spawforth (3rd edn; Oxford, 1996)
The Oxford Dictionary of Byzantium, ed. A. Kazhdan,
3 vols. (Oxford, 1991)
Prosopographie chretienne du Bas-empire
1 vol. 1, Prosopographie de lAfrique chretienne
(303533), ed. A. Mandouze (Paris, 1982)
2 vol. 2, Prosopographie de lItalie chretienne (313604),
ed. C. Pietri and L. Pietri (Rome, 19992000)

PG
PL
PLRE

Patrologiae Cursus Completus, Series Graeca,

ed. J.-P. Migne, 161 vols. (Paris, 185766)
Patrologiae Cursus Completus, Series Latina,
ed. J.-P. Migne, 221 vols. (Paris, 184464)
The Prosopography of the Later Roman Empire
1 vol. 1, ad 260395, ed. A. H. M. Jones, J. R.
Martindale, and J. Morris (Cambridge, 1971)
2 vol. 2, ad 395527, ed. J. R. Martindale
(Cambridge, 1980)
3 vol. 3, ad 527641, ed. J. R. Martindale
(Cambridge, 1992)

PLS
PO
Proc.

Patrologiae Cursus Completus, Series Latina,

Supplementum, ed. J.-P. Migne, 5 vols. (Paris,
195874)
Patrologia Orientalis (Paris, 1903 )
Procopius of Caesarea, Opera omnia, ed. J. Haury,
4 vols. (Leipzig, 190513; repr. 19624)
Aed.
Anecd.
BG
BP
BV

SC

De aedificiis
Anecdota sive Historia Arcana
De bello Gothico (De bellis libri, vviii)
De bello Persico (De bellis libri, iii)
De bello Vandalico (De bellis libri, iiiiv)

Sources chretiennes
xvii

List of abbreviations
Val. Nov.
V. Fulg.
Vict. Tonn.
Vict. Vit.

Valentinian III, Novellae, ed. T. Mommsen and

P. Meyer, CTh 2:69154.
Ferrandus, Vita S. Fulgentii episcopi Ruspensis,
ed. G.-G. Lapeyre, in Vie de Saint Fulgence de Ruspe
(Paris, 1929)
Victor Tonnennensis, Chronicon, ed. C. Cardelle de
Hartmann, CCSL 173A (Turnhout, 2001), pp. 155
Victor of Vita, Historia persecutionis Africanae provinciae,
ed. Michael Petschenig, CSEL 7 (Vienna, 1881)

xviii

I N T RO D U C T I O N

In 416, when preaching a sermon on the psalms in late Roman Carthage,

Augustine was able to ask his audience, Who now knows which nations
in the Roman empire were what, when all have become Romans, and
all are called Romans?1 Yet already by the time Augustine addressed his
Carthaginian audience the continued unity of the Roman Mediterranean
was being called into question. The defeat and death of the Roman
emperor Valens at Adrianople in 378 had set the stage for a new phase
of conflict between the empire and its non-Roman neighbours; and
over the course of the fifth century Roman power collapsed in the West,
where it was succeeded by a number of sub-Roman kingdoms. Questions
that had seemed trivial to Augustine were suddenly and painfully alive:
what did it mean to be Roman in the changed circumstances of the
fifth and later centuries? And (from a twenty-first-century perspective)
what became of the idea of Romanness in the West once Roman power
collapsed?
Empires can survive as identities long after they disappear as polities.
This book is an examination of that process in late antique North Africa.
The region lends itself to such a study above all because Romanness
was contested there over the long term and between multiple groups.
Roughly corresponding to the strip of modern Morocco, Algeria,
Tunisia, and western Libya between the Mediterranean and the Sahara,
Roman Africa was economically and politically one of the empires most
critical territories. Strategically located at the bottleneck between the
eastern and western Mediterranean, Africa was also the breadbasket of
Rome, providing through annual taxes in kind the grain, oil, and wine
that fed the Eternal City, the imperial court, and the administration.2 The
1

Augustine, Enarrationes in psalmos 58.1.21, ed. E. Dekkers and J. Fraipont, CCSL 3840 (Turnhout, 1956), 39:744: Quis iam cognoscit gentes in imperio Romano quae quid erant, quando
omnes Romani facti sunt, et omnes Romani dicuntur? Unless otherwise noted, all translations
are mine.
On the annona and its role in the transformation of the late Roman Mediterranean, see esp.
M. McCormick, Bateaux de vie, bateaux de mort: maladie, commerce, transports annonaires

Staying Roman
fate of Africa was intimately connected to that of the western Roman
empire writ large. Not surprisingly, then, in late antiquity Africa had a
troubled history of conquests and reconquests that forced North Africans
constantly to reconsider the terms in which their identities were defined.
In 406, a confederation of peoples known as the Vandals crossed the
empires Rhineland frontier into Gaul, passing next into Spain (where
they settled for a time) and then in 429 into Africa. There they established
an autonomous kingdom which, from 439, had as its capital the storied
metropolis of Carthage.3 Roughly one hundred years on, in 5334, the
East Roman or Byzantine empire managed to re-establish control of
Africa, only to see their domination of the region checked in the interior by indigenous kingdoms that from an imperial point of view were
thought of as Moorish.4 Finally, in the seventh century, the armies of
Islam began a fifty-year conquest of Africa, and by c.700, they had ended
for ever Byzantine control of the region.5
In this study, I argue that the fracturing of the political unity of the
Roman empire which followed from these developments (and similar ones across the Mediterranean) also led to a fracturing of Roman
identity above all along political, cultural, and religious lines as individuals who continued to feel Roman but who were no longer living
under imperial rule sought to define what it was that connected them to
their fellow Romans elsewhere. The multiple definitions of Romanness
this process produced could (and did) overlap and inform one another,
but they were not always mutually reinforcing. Significantly, though,
in the changed conditions of the fifth and later centuries, Romanness
was not just a question of sentiment or nostalgia; it had practical value,
which varied according to the context. Critically, late antique ideas about
Roman identity could be used in a remarkably flexible manner to foster
a sense of similarity (or difference) over space, time, ethnicity, and so
forth in a wide variety of situations and circumstances. For indeed, even
in the face of protracted political and social upheaval, both the African
elite and a succession of emperors struggled to ensure that Africa stay

et le passage e conomique du Bas-empire au moyen a ge, in Morfologie sociali e culturali in Europa fra
tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sullalto Medioevo
45 (Spoleto, 1998), pp. 35122.
The classic studies are L. Schmidt, Geschichte der Wandalen (2nd. edn; Munich, 1942) and
C. Courtois, Les Vandales et lAfrique (Paris, 1955); the most recent, A. Merrills and R. Miles, The
Vandals (Chichester, 2010).
The most recent synthetic treatment of Byzantine Africa remains C. Diehl, LAfrique byzantine:
histoire de la domination byzantine en Afrique (533709) (Paris, 1896). On the Moors, see Y. Moderan,

Les Maures et lAfrique romaine (IVeVIIe si`ecle), Biblioth`eque des Ecoles

francaises dAth`enes et de
Rome 314 (Rome, 2003).
See now W. Kaegi, Muslim Expansion and Byzantine Collapse in North Africa (Cambridge, 2010).

Introduction
Roman by actively seeking to ensure the regions continued integration
into the larger Mediterranean world.
The analysis that follows thus focuses heavily on the interconnectedness of Africa and the Mediterranean. Since Pirenne, questions of this
sort have been intimately bound up with the broader transition from
Roman antiquity to the early Middle Ages.6 Connectedness does not in
itself provide a definition of Romanness, a heavily freighted term whose
meaning was constantly being redefined over time and which was in a
continual process of mediation and renegotiation in different situations
and contexts. But the culture that had emerged by the fifth century of
the present era and which late antique North Africans (among others)
thought of as Roman was inherently international. One facet of its
preservation in the fifth to seventh centuries was the maintenance of
ties political, personal, religious, intellectual, and economic among
regions that had once been part of the empire, but now found themselves
following divergent political trajectories. It is this facet of the maintenance of Romanness that particularly interests me in this book.
1. conceptualizing romanness
The Romanness of Roman Africa has not always been taken for granted.
In his 1976 La Resistance africaine a` la romanisation, the Algerian scholar
Marcel Benabou explored the strength of pre-Roman African traditions
and the emergence of a distinctively African form of Roman civilization
by arguing that the empire had encountered not only military but also
cultural resistance in Africa.7 Over thirty years on, Benabous ideas remain
challenging.8 The notion that Africa had never really been Romanized
is also central to what are still two of the most influential books on
late antique North Africa, both written as French colonial rule in the
Maghrib lurched toward its eventual collapse: W. H. C. Frends The
Donatist Church and Christian Courtoiss Les Vandales et lAfrique.9 Both
6

7
8

Three notable recent works to take up the challenges of H. Pirenne, Mahomet et Charlemagne (7th
edn; Paris, 1937) are P. Horden and N. Purcell, The Corrupting Sea: A Study of Mediterranean History
(Oxford, 2000); M. McCormick, Origins of the European Economy: Communications and Commerce,
ad 300900 (Cambridge, 2001); and C. Wickham, Framing the Early Middle Ages: Europe and the
Mediterranean, 400800 (Oxford, 2005). See also, on a still wider canvas, B. Cunliffe, Europe between
the Oceans, 9000 bcad 1000 (New Haven, Conn., 2008).
M. Benabou, La Resistance africaine a` la romanisation (Paris, 1976).
See, e.g., G. Woolf, Becoming Roman: The Origins of Provincial Civilization in Gaul (Cambridge,
1998), pp. 1920 and G. Woolf, Beyond Romans and Natives, World Archaeology 28/3 (1997),
pp. 3401.
W. H. C. Frend, The Donatist Church: A Movement of Protest in Roman North Africa (Oxford, 1952);
for Courtois, see above, n. 3.

Staying Roman
authors believed that in Africa the empire had encountered a Berber
population that remained fundamentally unchanged by Greco-Roman
civilization. The idea of African resistance (or intransigence) poses an
obvious challenge to a study examining how Africa stayed Roman in late
antiquity: had Africa ever really become Roman in the first place?
J. Frank Gilliam once remarked that Being a Roman, like being an
American, was a matter of law, not of culture or the lack of it.10 Recent
analyses of Roman identity have nuanced this idea, focusing precisely
on the cultural and ethnic aspects of being Roman; but on at least
one level the statement is certainly true: cultural assimilation was not
a prerequisite of Roman citizenship.11 By the third century of this era
most free inhabitants of the empire were Roman citizens. Moreover,
as we will see, when fifth- and sixth-century Africans thought of things
Roman, they thought for the most part of the empire itself, its history and
army, its greatest poet (Virgil), and the Latin language: the empire and
its institutions defined Romanness. Accordingly, in the minds of some,
the Romanness of a particular provincial group could be lost or gained
according to the empires varying political and military fortunes as
some felt had happened in Africa in the Vandal period (see Chapter 4).
It also seems to have been the case that whatever notions the Senate
and people of Rome may have had about their civilizing mission in the
western Mediterranean, political control was the primary factor motivating the metropoliss relations with its conquered provinces. As often as
not, this was accomplished by working together with local elites. Again,
cultural change was not essential.12 In an important paper, P. D. A.
Garnsey has adduced evidence of both continuity and rupture in the
African ruling class after the Roman conquest of Africa. The region
unquestionably saw immigration from Italy and elsewhere in the Roman
world. Nonetheless, in accordance with their traditional policy of building up a network of families, groups and communities with vested interests in the prolongation of Roman rule, Romans also rewarded local,
10
11

12

J. F. Gilliam, Romanization of the Greek East: The Role of the Army, Bulletin of the American
Society of Papyrologists 2 (1965), p. 66.
Cultural aspects of Roman identity: see, e.g., A. Wallace-Hadrill, Romes Cultural Revolution
(Cambridge, 2008); E. Dench, Romulus Asylum: Roman Identities from the Age of Alexander to the
Age of Hadrian (Oxford, 2005); and Y. Syed, Vergils Aeneid and the Roman Self: Subject and Nation
in Literary Discourse (Ann Arbor, Mich., 2005). Citizenship: see, e.g., P. D. A. Garnsey, Romes
African Empire under the Principate, in P. D. A. Garnsey and C. R. Whittaker (eds.), Imperialism
in the Ancient World (Cambridge, 1978), p. 248; and, in general, Dench, Romulus Asylum,
pp. 93151 and A. N. Sherwin-White, The Roman Citizenship (2nd edn; Oxford, 1973).
See, inter alia, R. Laurence and J. Berry (eds.), Cultural Identity in the Roman Empire
(London, 1998), p. 3; D. J. Mattingly, Libyans and the Limes: Culture and Society in Roman
Tripolitania, Antiquites africaines 23 (1987), p. 80; Garnsey, Romes African Empire, pp. 2524.

Introduction
African notables for their support with land and other material benefits.
These benefits included access to positions in the central administration
and membership in the senatorial aristocracy.13 Indeed, D. J. Mattinglys
studies of Tripolitania (western Libya) seem to indicate that there, at least,
Romans preferred to leave existing power structures more or less intact as
long as local elites could be persuaded to reconcile themselves to Roman
authority.14 This was probably the case throughout the frontier zone in
Roman Africa, where representatives of the empire deployed much the
same techniques to ensure their hegemonic dominance.15
Though not necessarily aggressively promoted by the Roman state,
in the imperial period political control and cultural change nevertheless did go hand in hand. This process has traditionally been referred to
as Romanization, though the word is misleading if taken to imply a
unidirectional flow of culture.16 As Greg Woolf has recently observed,
there was no standard Roman civilization against which provincial cultures might be measured. The city of Rome was a cultural melting
pot and Italy experienced similar changes to the provinces.17 What
we seem to see instead is the acceleration of a process already under
way in the third century bc whereby the economies, societies, and
cultures of the disparate regions of the Mediterranean became ever
more tightly interwoven: an increased circulation of people, things,
and ideas, and the emergence of what can, even if only loosely, be
referred to as a pan-Mediterranean set of attitudes, outlooks, beliefs, and
values.
The result was a remarkably flexible cultural system that I refer to here
as Roman, though it was deeply indebted to the Hellenic tradition,
unthreatened by the survival of distinctively local customs and conventions, and easily capable of assimilating foreigners. Reinforced for
centuries by an intensely conservative educational system in the hands

13
14

15

16

17

Garnsey, Romes African Empire, passim; the quotation is ibid., p. 235.

Mattingly, Libyans and the Limes , pp. 803. As Ramsay MacMullen has recently shown
of Jubas Mauretanian kingdom, a high degree of acculturation could accompany such reconciliation: R. MacMullen, Romanization in the Time of Augustus (New Haven, Conn., 2000),
pp. 429.
Hegemonic dominance: D. J. Mattingly, War and Peace in Roman North Africa: Observations
and Models of State-Tribe Interaction, in R. B. Ferguson and N. Whitehead (eds.), War in the
Tribal Zone: Expanding States and Indigenous Warfare (Santa Fe, NM, 1992), pp. 3160. See also
D. Cherry, Frontier and Society in Roman North Africa (Oxford, 1998).
See, e.g., Wallace-Hadrill, Romes Cultural Revolution, esp. pp. 4478 and for Africa, D. J. Mattingly
and R. B. Hitchner, Roman Africa: An Archaeological Review, JRS 85 (1995), pp. 2045. See
also the similar debate surrounding the term Hellenization: e.g., G. W. Bowersock, Hellenism
in Late Antiquity (Ann Arbor, Mich., 1990), pp. 67.
Woolf, Becoming Roman, p. 7.

Staying Roman
of grammarians,18 Roman culture was nevertheless not static. Like all
cultural systems, it was the end-product of individual people living
together and in communication with one another. Cultures adapt to
the new circumstances in which they find themselves as a product of the
more personal adaptations of individuals. Given the vagaries of distinct
personalities and characters, let alone the absorption of new populations,
change is inevitable.
This was perhaps most famously the case with Roman religion.
Romans were, of course, generally willing to expand their pantheon to
include the gods of conquered peoples. By the fifth century of our era,
however, an even more profound transformation of Roman religion had
taken place as the Roman faith ( fides Romana) came to mean Nicene
Christianity (see Chapter 3). But the adaptability of the Roman cultural
system is visible in many different areas, from naming patterns to patterns
of thought. By the sixth century ad, for example, the old Roman tria
nomina or three names had for the most part given way to the use of
a single name. In the sixth century, the two most popular of these were
John and Theodore, neither of them Roman by, say, the standards of
the second century bc. Similarly, Peter Heather has recently shown how
even so profound a division in the Roman thought world as that between
Romans and barbarians could be adapted to the new realities of the
fifth century. As control over the western provinces of the Roman empire
was increasingly concentrated in the hands of non-Romans (barbari, or
barbarians), the very idea of Romanness came to signify a willingness to work alongside the empire.19 However it is defined, Roman
culture like all cultures changed over time.
Culture in general is, however, notoriously difficult to define.20 Like
ethnicity, culture seems to be something that is only ever visible in our
peripheral vision; on closer examination, it has a tendency to fall apart.
This results in an unavoidable degree of vagueness as to the defining
features of Roman culture and a corresponding lack of precision in our
18
19

20

For the role of the grammarians, see R. Kaster, Guardians of Language: The Grammarian and Society
in Late Antiquity, Transformation of the Classical Heritage 11 (Berkeley, Calif., 1988).
P. Heather, The Barbarian in Late Antiquity: Image, Reality, and Transformation, in Constructing
Identities in Late Antiquity, ed. R. Miles (London, 1999), pp. 23458; the quotation is from
p. 247.
M. Harris, Theories of Culture in Postmodern Times (Walnut Creek, Calif., 1999), p. 19 defines
culture as the socially learned ways of living found in human societies and sees culture as
embracing all aspects of social life, including both thought and behavior. C. Geertz, The
Interpretation of Cultures (New York, 1973), p. 145 defined it as the fabric of meaning in terms
of which human beings interpret their experience and guide their action. See also R. C. Ulin,
Understanding Cultures: Perspectives in Anthropology and Social Theory (2nd edn; Malden, Mass.,
2001) and J. Clifford, The Predicament of Culture: Twentieth-Century Ethnography, Literature, and Art
(Cambridge, Mass., 1988).

Introduction
ability to measure them. In a forceful critique of what he calls unworkable models of Romanization, David Cherry considers point by point
the provincial adoption of Roman or Roman-style architectural forms,
names, religious practices, styles of dress, and municipal government;
urbanization; the promotion of cities to the status of municipalia or coloniae; the use of coinage; the diffusion of Latin as a spoken and written
language, the epigraphic habit, Roman tastes in art, and Roman-style
graves; the distribution of goods of Roman manufacture or style; the
presence of the Roman army in the provinces, and the recruitment
of provincials into it. In themselves, Cherry argues, each of these is
an insufficient indicator of provincial acculturation.21 Cherrys critiques
are thoughtful and reasoned; his scepticism, sobering. Even if a precise
definition is impossible, however, it must be admitted that when taken
together the combination of factors that Cherry rejects one by one represent something approximating a working characterization of culture,
or at least of Roman culture.
Considering such a variety of factors also has the advantage of reflecting
late antique perceptions of what it was that distinguished peoples from
one another, and especially barbarians from Romans. Augustine wrote
of different rites and customs and a diversity of languages, weapons,
and varieties of dress.22 Other late antique writers added laws and forms
of government, religion, battle tactics, and marriage customs, as well
as diet, hairstyle, and other elements of physical appearance to the list
(see Chapter 5).
These marks of distinction are not always traceable 1,500 years or more
after the fact. By almost any indicator, however, Africa Proconsularis
(northern Tunisia), Byzacena (southern Tunisia), and Numidia (eastern
Algeria) participated fully in the broader culture of the Mediterranean
empire. They were the most heavily urbanized of the African provinces,
and Claude Lepelley has demonstrated that their cities and municipal
institutions continued to function right down to the period of the
Vandal invasion.23 D. J. Mattingly and R. B. Hitchner have observed
that the construction of fora, basilicas, Romanized temples, baths, theatres, amphitheatres, circuses, and aqueducts was a major concern of
21
22

23

Cherry, Frontier and Society, pp. 8299.

Augustine, De civitate Dei 14.1, ed. B. Dombart and A. Kalb, CCSL 478 (Turnhout, 1955),
48:414: cum tot tantaeque gentes per terrarum orbem diuersis ritibus moribusque uiuentes
multiplici linguarum armorum uestium sint uarietate distinctae, non tamen amplius quam duo
quaedam genera humanae societatis existerent.
C. Lepelley, Les Cites de lAfrique romaine au Bas-empire, 2 vols. (Paris, 197981); now see also
G. Sears, Late Roman African Urbanism: Continuity and Transformation in the City, BAR International
Series 1693 (Oxford, 2007) and A. Leone, Changing Townscapes in North Africa from Late Antiquity
to the Arab Conquest, Studi storici sulla Tarda Antichit`a 28 (Bari, 2007), pp. 45125.

Staying Roman
towns of all sorts, with most local schemes limited more by the scale of
resources than by resistance.24 Latin was so well established as an everyday language in parts of this region that it was said still to be spoken
in Tunisia as late as the twelfth century (see Chapter 7). To a Constantinopolitan observer of the sixth century, Africans spoke Latin more
pleasingly even than Italians (see Chapter 3). Africa Proconsularis and
Byzacena were the production-centres of African red slip ware or terra
sigillata, the quintessential late Roman fine ceramic tableware, enjoying as
it did a pan-imperial distribution in the fourth and fifth centuries. Mosaic
arts were highly developed in these provinces too, as demonstrated, for
example, by the magnificent collections of the Bardo Museum in Tunis.
The Roman educational system was firmly entrenched in Africa and may
have survived longer there than anywhere else in the West.25 The provincial archives of Africa were one of the major wellsprings of information
for the codification of Roman law.26 Nor was Roman legal and political
thought always restricted to a thin, highly Romanized elite. Leslie Dossey
has argued cogently that such ideas permeated rural aspirations in the late
empire.27 Language, lifestyle, arts, and institutions: by 439, the culture of
the central African provinces would have been comfortably familiar to
visitors from other parts of the empire.
Even before the influx of new blood in the fifth century, however,
local cultures had remained important throughout the Roman world. In
Africa, Benabou was quite right to observe the specifically African nature
of Roman civilization. Though Garnsey rejects the explanatory value of
the idea of resistance, he too concludes that a specific cultural complex
emerged in Roman Africa,28 while Mattingly and Hitchner write of
Roman Africa as a new world, different from what had gone before and
equally distinct from other parts of the Empire.29 Punic survived as a
spoken language alongside Latin.30 Pan-imperial artistic motifs such as
the four seasons could have a distinctive meaning in an African context.31
24
25

26
27

28
29
30
31

Mattingly and Hitchner, Roman Africa, p. 205.

H.-I. Marrou, Histoire de leducation dans lantiquite (7th edn; Paris, 1971), pp. 4923; P. Riche,
Education and Culture in the Barbarian West, from the Sixth through the Eighth Century, trans. J. J.
Contreni (Columbia, SC, 1978), pp. 379.
A. H. M. Jones, The Later Roman Empire, 284602: A Social, Economic and Administrative Survey,
2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986), 1:4745.
L. Dossey, Christians and Romans: Aspiration, Assimilation, and Conflict in the North African
Countryside, Ph.D. diss., Harvard University (1998) and now L. Dossey, Peasant and Empire in
Christian North Africa, Transformation of the Classical Heritage 47 (Berkeley, Calif., 2010).
Garnsey, Romes African Empire, pp. 2524.
Mattingly and Hitchner, Roman Africa, p. 205.
J. N. Adams, Bilingualism and the Latin Language (Cambridge, 2003), pp. 20045; on Libyan,
see ibid., pp. 2457. See also below, Chapter 3.4.
Mattingly and Hitchner, Roman Africa, p. 205.

Introduction
Even in the Christian period it is possible to speak of a characteristically
local flavour to the African name-stock, peppered with names such as
Victor, Adeodatus, Benenatus, Quodvultdeus, Saturninus, Cresconius,
and Felix (see Chapter 2). Under the Vandals an unmistakable pride in
Africa comes to the surface in the writings of local elites (see Chapter 1).
By the fifth century, then, the empires southern provinces had managed
to become Roman while remaining African.
2. africa and the mediterranean on the eve of the
vandal invasion
If Roman cultural identity was by definition trans-regional, then integration into the larger Mediterranean world was of the essence. And on
the eve of the Vandal invasion, Africa remained well integrated into the
empire. Proconsular Africa the chief province of Roman Africa had
long been governed by a proconsul of senatorial rank. Under Constantine (ad 31237) Byzacena and Numidia came to be administered by
senators as well. The governors of these two provinces were given the
title of consularis to distinguish them from the non-senatorial governors
or praesides of Tripolitania and the two Mauretanias (central and western
Algeria).32 Apart from the proconsul, all of these governors were under
the authority of the Vicar of Africa who, under Constantine, also came
to be drawn from ranks of the nobility.33
According to Mechtild Overbeck, whose study is the only full-length
investigation to date of the role the African elite played in the political and
social changes of the late antique world, the men who governed Africa
in the fourth century were for the most part Italian in origin. Officeholders from other regions, including Africa and the provinces of the
eastern Mediterranean, played a role as well. The regional origins of the
fourth-century governors of Byzacena and the Mauretanias are largely
unknown, but an outright majority of the known consulares of Numidia
were from Italo-Roman aristocratic families, including one of the most
important noble households of the late Roman world, the gens Ceionia.34
Similarly, a large number of the Vicars and Proconsuls of Africa stemmed
from the great families of the city of Rome, particularly the houses of the
32

33
34

M. T. W. Arnheim, The Senatorial Aristocracy in the Later Roman Empire (Oxford, 1972), pp. 52
and 567; A. Chastagnol, Les Consulaires de Numidie, in Melanges darcheologie, depigraphie et
dhistoire offerts a` Jerome Carcopino (Paris, 1966), pp. 21528; and A. Chastagnol, Les Gouverneurs
de Byzac`ene et de Tripolitaine, Antiquites africaines 1 (1967), pp. 11934.
Vicars: Arnheim, Senatorial Aristocracy, pp. 634.
M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Spatantike, Frankfurter althistorische Studien 7 (Kallmunz, 1973), pp. 2930; Chastagnol, Consulaires de Numidie, p. 219.

Staying Roman
Anicii and, again, the Ceionii.35 In the later Roman empire, government
was a family affair, and through the reign of Constantine the Italo-Roman
Proconsuls of Africa typically appointed their sons or younger brothers
to the post of legate.36 The deep, personal engagement in Africa of
these prominent metropolitan aristocrats created a human bridge linking the families of two of the wealthiest and most important provinces
of the western empire. The local contacts and clientele networks these
Italo-Roman proconsuls and legates established in Africa could later be
actualized by ambitious Africans who made their way to Rome, even
as the greatest families of the ancient capital lent a certain lustre to the
circles in which they moved during their African governorships.
Africans were, of course, also involved in the administration of their
own provinces. If we accept Overbecks judgement as to their origins,
perhaps 17 per cent of the known Proconsuls of Africa between the
years c.295 and 429 were themselves Africans.37 Overbeck also concludes
that two comites Africae military commanders of all the troops stationed in Africa and one Praetorian Prefect of Africa were of local
origin as well.38 Five of the late Roman senatorial governors of Numidia
were from African families, and after the reign of Constantine all of the
fourth- and early fifth-century proconsular legates appear to have been
Africans, too, even when the proconsuls were Roman nobles.39 Precision
is unattainable, but, as Garnsey once observed of Roman Africa in the
second century, this matters less than the fundamental fact that Africans
had access to the central administration and the highest status-group.
The empire was still Rome-based, but the ruling class that directed it
was cosmopolitan.40 Notwithstanding the displacement of Rome as the
ruling centre of empire in late antiquity, the comment applies with equal
validity to Africa in the fourth and early fifth centuries.
Beyond Africa, scholars have tended to comment on the relative
absence of Africans from positions of influence in the fourth and fifth
centuries.41 The data provided by Overbeck, however, further serve to
35
36

37

38
39
40
41

Overbeck, Senatsadel, pp. 23 and 33.

A. Chastagnol, Les Legats du proconsul dAfrique au Bas-empire, Libyca 6 (1958), p. 12, repr. in

A. Chastagnol, LItalie et lAfrique au Bas-empire: Etudes

administratives et prosopographiques, Scripta
varia (Lille, 1987), pp. 6782, here p. 72.
Overbeck, Senatsadel, pp. 238; see also PLRE 12, fasti. Overbeck, Senatsadel, p. 33 rejects the
argument of A. Chastagnol, La Prefecture urbaine a` Rome sous le Bas-empire, Publications de la
Faculte des lettres et sciences humaines dAlger 34 (Paris, 1960), p. 431 that Chilo, Proconsul of
Africa in 375, was himself African.
Overbeck, Senatsadel, pp. 323.
Chastagnol, Legats du proconsul, p. 12; Overbeck, Senatsadel, pp. 29 and 312.
Garnsey, Romes African Empire, p. 251.
Overbeck, Senatsadel, p. 40 and B. H. Warmington, The North African Provinces from Diocletian to
the Vandal Conquest (Cambridge, 1954), p. 107.

10

Introduction
highlight the fact that in the fourth century Italy and Africa so close
together geographically were also connected by social and political
ties among the ruling elite. Not only was Africa administered largely by
Italo-Roman aristocrats; insofar as Africans of senatorial rank served in
the imperial administration, they tended to do so in Italy. Overbeck notes
that for many Africans, service in the imperial administration began with
a post in an Italian province or in the capital itself.42 Gallic senators, by
comparison, only rarely administered Italian provinces: the main focus of
their political activity was Gaul.43 Similarly, from the time of Theodosius
(ad 37995), Spanish senators acted as governors of various provinces, but
rarely of Italy.44 By Theodosius day, however, African senators had also
by and large ceased to administer the Italian provinces. The particularly
close relationship to the imperial centres of power that a select handful
of the most elite Romano-Africans appear to have enjoyed in the fourth
century thus seems to have been a product of a special connection to the
house of Constantine.
The fact that circulation between Africa and Italy was not confined
to the ranks of the senatorial elite in the fourth century is witnessed by
the careers of men such Augustine, who in 382 set sail from Carthage to
Rome in hopes of furthering his career as a professor of rhetoric. He was
not alone in this movement. He had been preceded by the late fourthcentury bishop of Verona, Zeno, who was also an African. Augustine
himself was joined in Italy by his pupil and fellow-townsman Licentius,
as well as by his own brother Navigius and their redoubtable mother
Monica. Indeed, Peter Brown suggests that in Augustines biography we
can glimpse the reflected light of all the ambitious young men from
African provincial backwaters like Thagaste. The sons of a financially
precarious, small-time provincial gentry, they moved together in a restless
pursuit of advancement, travelling like Augustine to Carthage and then
to Italy. But outside the imperial administration, too, there seems to have
been a constriction in the number of Africans who could find positions of
influence outside of their native province, and all of Augustines fellowcountrymen returned home eventually, their ambitions frustrated, to
spend the rest of their lives in a thoroughly provincial setting, as the
42
43
44

Overbeck, Senatsadel, p. 33: Fur viele Afrikaner begann der Dienst in der Reichsverwaltung mit
einem Amt in einer italienischen Provinz oder in der Hauptstadt selbst.
Overbeck, Senatsadel, p. 34; on Gallic senators, see also K. F. Stroheker, Der senatorische Adel im
spatantiken Gallien (Tubingen, 1948), pp. 542.
Overbeck, Senatsadel, p. 34. On the Spanish senatorial elite, see also K. F. Stroheker, Spanische
Senatoren der spatromischen und westgotischen Zeit, in Germanentum und Spatantike (Zurich,
1965), pp. 5487; and A. Chastagnol, Les Espagnols dans laristocratie gouvernementale a`
lepoque de Theodose, in Les Empereurs romains dEspagne (Paris, 1965), pp. 26992.

11

Staying Roman
bishops of small African towns.45 For in the struggle to maintain the
integrity of the empires military frontiers, Brown argues, the late fourthcentury emperors no longer had a need for the highly cultivated and
literate services of these southerners.46 Africa, however, was the source
of Italys grain and oil, and as long as that was the case the western empire
would certainly have a need for Africa. Moreover, many Italo-Roman
aristocratic families especially, again, the Ceionii owned large estates
in the African provinces, ensuring that their interest in the south was
economic as well as administrative.47 Despite the slackening of its sons
success in the top ranks of the imperial administration, at the dawn of
the fifth century Africa was in no danger of falling out of the Roman
orbit.
3. sources, questions, and methods
This kind of integration political, social, economic was vital to the
formation of Roman culture as it had developed by the fifth century.
It is remarkable that, in an era of pre-modern communications, by the
late imperial period some sense of unity had come to characterize the
area stretching from Hadrians Wall to the Sahara and from the Strait of
Gibraltar to the Euphrates. This sense of unity defined what it meant
to be Roman, as opposed to Gaulish, Spanish, British, Italian, African,
Egyptian, or whatever. Irrespective of whether we choose to characterize
the developments of the later fourth, fifth, and sixth centuries as the fall
of Rome or the transformation of the Roman world, the empires
loss of direct political control of the West posed a serious challenge to
the maintenance of that sense of unity.48 Our understanding of how
western senatorial aristocrats ensured their own survival and salvaged
some kind of continuity in their way of life under barbarian rule has been
deepened immeasurably by studies that reveal nobles seeking to maintain
the tradition of imperial service by serving under Germanic regimes,
turning to ecclesiastical office (that exceptional repository of late Roman
ideals, customs, and ideas), and engaging in the literary pursuits that
45
46
47

48

P. Brown, Augustine of Hippo: A Biography (Berkeley, Calif., 1969), pp. 245; the quotation is
from p. 25. See also Warmington, North African Provinces, pp. 1068.
Brown, Augustine, p. 25.
Chastagnol, Consulaires de Numidie, p. 219; Overbeck, Senatsadel, pp. 345 and 412. On
the social and economic power of the fourth-century western senatorial elite in general, see
Arnheim, Senatorial Aristocracy, 14368. The varying economic fortunes of different regions of
the African countryside are briefly summarized by McCormick, Origins, p. 33.
On fall and transformation, see, e.g., the contrasting views of W. Goffart, Barbarian Tides: The
Migration Age and the Later Roman Empire (Philadelphia, Pa., 2006) and B. Ward-Perkins, The Fall
of Rome and the End of Civilization (Oxford, 2005).

12

Introduction
defined the Roman gentleman, including the maintenance of epistolary
ties.49 Such developments in late Roman aristocratic culture took place
alongside the emergence or redefinition of various non-Roman identities
that, with time, were to reshape the European ethnic landscape.50 On
both fronts, these same trends characterized African society, too, after the
Vandal conquest (see Chapters 3 and 1, respectively). But just as important
for staying Roman in late antique North Africa especially given the
provinces crucial role as a hub of trans-Mediterranean communications
was the maintenance of inter-regional ties, without which the centrifugal
tendencies of localism could not be restrained.
Such inter-regional integration operated on many different levels,
including the realm of political ideas and the legitimation of power,
participation in religious debates, the movement of saints cults, and the
physical circulation of individuals, books, letters, and other objects. In the
Byzantine period, political participation in the life of the empire would
once again become important; in the Vandal period, it was probably less
49

50

A few notable titles among many, in addition to the material cited above: D. Claude, Adel,
Kirche und Konigtum im Westgotenreich, Vortrage und Forschungen 8 (Sigmaringen, 1971); R. W.
Mathisen, The Ecclesiastical Aristocracy of Fifth-Century Gaul: A Regional Analysis of Family
Structure, Ph.D. diss., University of Wisconsin (1979); J. Matthews, Western Aristocracies and
Imperial Court, ad 364425 (Oxford, 1990); J. Drinkwater and H. Elton (eds.), Fifth-Century
Gaul: A Crisis of Identity? (Cambridge, 1992); R. W. Mathisen, Roman Aristocrats in Barbarian
Gaul: Strategies for Survival in an Age of Transition (Austin, Tex., 1993); H. Sivan, Ausonius of
Bordeaux: Genesis of a Gallic Aristocracy (New York, 1993); J. Harries, Sidonius Apollinaris and the
Fall of Rome, ad 407485 (Oxford, 1994); W. E. Klingshirn, Caesarius of Arles: The Making of a
Christian Community in Late Antique Gaul, Cambridge Studies in Medieval Life and Thought,
4th ser. 22 (Cambridge, 1994). Slightly different in their focus, but great influences on my
own thought: T. S. Brown, Gentlemen and Officers: Imperial Administration and Aristocratic Power in
Byzantine Italy ad 554800 (London, 1984) and M. Kulikowski, Late Roman Spain and its Cities
(Baltimore, Md., 2004).
The literature on barbarian identity in late antiquity is vast and ever-growing. Foundational studies include R. Wenskus, Stammesbildung und Verfassung: Das Werden der fruhmittelalterlichen Gentes
(Cologne, 1961) and H. Wolfram, History of the Goths, trans. T. J. Dunlap (Berkeley, Calif.,
1988); see also W. Pohl, Tradition, Ethnogenese und literarische Gestaltung: Eine Zwischenbi
lanz, in K. Brunner and B. Merta (eds.), Ethnogenese und Uberlieferung:
Angewandte Methoden der
Fruhmittelalterforschung (Vienna, 1994), pp. 926. A few other notable titles include: P. J. Geary,
Before France and Germany: The Creation and Transformation of the Merovingian World (New York,
1988); P. Heather, Goths and Romans, 332489 (Oxford, 1991); E. James, The Franks (Oxford,
1988); I. N. Wood, The Merovingian Kingdoms 450751 (London, 1994); P. Amory, People and
Identity in Ostrogothic Italy, 489554, Cambridge Studies in Medieval Life and Thought, 4th ser. 33
(Cambridge, 1997); P. Heather, The Goths (Oxford, 1996); I. N. Wood (ed.), Franks and Alamanni
in the Merovingian Period: An Ethnographic Perspective, Studies in Historical Archaeoethnology 3
(Woodbridge, 1998); P. Heather (ed.), The Visigoths from the Migration Period to the Seventh Century:
An Ethnographic Perspective, Studies in Historical Archaeoethnology 4 (Woodbridge, 1999);
F. Curta, The Making of the Slavs: History and Archaeology of the Lower Danube Region, ca. 500700,
Cambridge Studies in Medieval Life and Thought, 4th ser. 52 (Cambridge, 2001); F. Curta (ed.),
Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the Middle Ages (Turnhout, 2005);
E. James, Europes Barbarians, ad 200600 (New York, 2009).

13

Staying Roman
so. Over the course of my study, I have therefore repeatedly asked four
main questions of my sources, and subsequently applied and examined
my results in their light:
1. What made Africans (or at least elite Africans) Roman in late antiquity?
2. What were the links between politics, culture, and religion in Africa?
3. How did the powerful presence of foreign elements affect the
Romanness of the African elite?
4. What role did the circulation of people, things, and ideas play
in fostering, maintaining, and circumscribing a long-term, transMediterranean sense of political, cultural, and religious unity?
Though I have tried to cast as wide a social net as possible, the visibility
of the regional elite has ultimately rendered them my primary focus.
Sources for this group are comparatively abundant, and include letters,
ecclesiastical texts, chronicles and histories, poetry, sermons, legal sources,
and diplomatic correspondence, as well as archaeological, numismatic,
and epigraphic data. I have tried to draw on all of these sources in the
pages that follow, although as my study is ultimately an investigation of
the fate of an idea (or a complex of ideas) in general the textual sources
naturally figure more prominently than the non-textual ones.
In a search for answers to my questions, I have taken three main
approaches to these texts. First and foremost, I have tried to read as
widely as possible and with as open a mind as possible in the late
antique sources written in Africa, by Africans, or about either of the
two. Latin, Greek, and Arabic sources I have been able to consult in their
original languages; for Syriac, Armenian, and Coptic, I have relied on
the translations of others. Second, I have compiled a prosopographical
database of over 1,900 individuals with connections to North Africa in
the period from ad 439 to 700. Through this database, supplemented
by the three great prosopographies that cover this time and place (the
Prosopography of the Later Roman Empire, the Prosopography of the Byzantine Empire, and the Prosopographie chretienne du Bas-empire), I analyse the
names, ages, and occupations, ethnic identities and social origins, relatives, and associates of the shifting African elite, in addition to where
and when they were active, their movements from region to region, and
career patterns. This prosopographical approach allows me to address
the research questions outlined above on a case-by-case basis, while the
compilation of large sets of data ensures that the evidence is more than
simply anecdotal. Third, I have made extensive use of the electronic
resources that are increasingly available to historians of late antiquity and
14

Introduction
the Middle Ages, most of which are known through their acronyms: the
Patrologia Latina Database (PLD), Thesaurus Linguae Graecae (TLG),
Centre de traitement e lectronique des documents Library of Christian Latin Texts (CLCLT), digitized Monumenta Germaniae Historica
(dMGH), and Acta sanctorum databases. Together these provide a vast
corpus of Greek and Latin literature in electronically searchable format, allowing us to ask new questions and analyse long-familiar texts in
new ways.
The nature of my questions necessitates a broad view of the African
regional elite. I take the term elite to refer to a legally and economically
diverse group which, by virtue of their wealth and social position, wield
power and exercise influence within the society of which they are a part.
The definition holds equally for notables in a peasant village and for intimates of the emperor himself, though again my analysis is skewed towards
the higher echelons of power. In late antique North Africa the elite was
dynamic with respect to its character and composition, especially at its
highest levels. Africans played a role in their own governance throughout this period, but, as we have seen, at the dawn of the fifth century
African society was dominated by a civilian administration whose representatives were of largely Italian origins. The Vandal conquest brought
with it a new Germanic ruling class; after the Byzantine reconquest, the
sources reveal a class of military men, principally from the East, overseeing the defence of Africa. As in contemporary Italy, so in Byzantine
Africa reintegration into the political, military, and economic structures
of the empire came at a price. The senatorial aristocracy did not survive
the transition, at least not as a social group about which the sources have
anything to say. In examining the continued integration of Africa into
the larger Mediterranean, we would thus do well to consider not only
senatorial aristocrats who were themselves of African origins but also the
military commanders, exarchs, tribunes, and provincial administrators
sent to Africa from other parts of the empire; parvenus; and, indeed, even
those who existed on the fringes of the Roman world, and who came
to assimilate or to reject Roman culture.
A further word about some of the terminology I use in the course of
this study might be useful. I refer to Roman or Romanized natives of
Africa as Romano-Africans. The word does not necessarily refer only
to Romanized Africans of indigenous or Punic extraction; it is intended
to be a flexible term that can also refer to the descendants of Italians,
Greeks, and other inhabitants of the Roman empire who had settled in
the African provinces. The term Romano-African is used to distinguish
the inhabitants of the Roman provinces of Africa (who participated, even
15

Staying Roman
if only partially, in a broader Roman culture) from foreigners of all sorts,
including, for example, both Vandals on the one hand and Romans from
other provinces on the other; or in other words all those who were not
themselves Romano-Africans.
When greater precision is required, I refer to Romano-Africans specifically by their region (typically their province) of origin. The unmodified
term African (or North African) encompasses them all, but can also
be somewhat broader in its connotations: also African in my usage here,
as in late antiquity, are the inhabitants of this general region who lived
beyond the immediate pale of Roman control, referred to in the modern
literature as Moors or Berbers. However, as I use the adjective African
in a broadly late antique administrative sense, I do not include, for example, Egyptians under its rubric. Though of course African by modern
geographic standards, in this study (as in late antiquity) Egyptians, Cyrenaicans, Ethiopians and so forth are not considered Africans. In the
discussion that follows they are considered elements of the larger category of easterners. Similarly, the provinces of Egypt and Cyrenaica
and the kingdom of Ethiopia are not referred to by the terms Africa
and North Africa, which specifically designate the six provinces of
Roman Africa (Proconsularis, Byzacena, Numidia, Tripolitania, Mauretania Sitifensis, and Mauretania Caesariensis), occasionally extended to
include Mauretania Tingitana (which, under the Diocletianic reforms of
the third century, was administratively part of Spain).
For reasons that I hope will become clear in the course of the discussion in Chapter 5, late antique usage of the Latin term Maurus is
difficult to translate. Though it is problematic and in some instances
perhaps even misleading I have generally opted to use the noun Moor
(and its adjective, Moorish) as an English equivalent, and have tried to
avoid using the term Berber except where a clear linguistic or historical
connection to modern Berber societies or cultures exists. Though also
perhaps misleading, I use the term Byzantine more or less interchangeably with East Roman. It has become a deeply ingrained habit among
late antique North Africanists to refer to the period from ad 533 to 698
as the Byzantine period, and insofar as there were meaningful differences
between reconquest and pre-Vandal society it is probably useful to do so.
Byzantine, of course, is an entirely modern label. While I argue that the
legitimacy of their power in Africa was very much an open issue in the
critical years of the reconquest, no one in the late antique world questioned even for a moment the Romanness of those whom scholars call
Byzantines. With respect to place names, my usage is not as consistent as
perhaps it might have been. As a rule I have tended to use modern names
16

Introduction
and spellings for major sites like Rome, Carthage, and Jerusalem, though
I have sometimes preferred the historical equivalent, as for example with
Constantinople, Hadrumetum, and Sicca, where reference to Istanbul,
Sousse, and El Kef seems too evocative of a later period. With respect
to less-well-known sites, I have tried to use the name that seems likely
to be most familiar from the established literature. In general, this means
that I have tended to use the modern name for places that are primarily
known for their material remains and the ancient one for places that are
discussed more extensively in the written sources.
The organization of the study is largely chronological. The first three
chapters treat the Vandal period. Chapter 1 considers how the Vandal
kings sought to legitimate their rule in Africa and why that legitimation
took the particular form it did. Chapter 2 explores Africas sustained social
and cultural connections to the larger Mediterranean world in the fifth
and sixth centuries, looking in particular at the movements of individuals, letters, books, and cultural phenomena such as personal names and
saints cults. Chapter 3 examines Romano-African responses to the Vandal presence, how Africans sought to accommodate or resist the Vandals,
and how this reveals diverging attitudes towards Romanness. The next
three chapters consider the question of how well integrated Africa was
into the East Roman world in the wake of the sixth-century Byzantine
reconquest. In Chapter 4, I focus on the realities and nature of the Byzantine presence in sixth- and seventh-century Africa: the regions of origin
of its administrators (civil and military), the patterns and terms of their
appointment, their family life in Africa, and the languages they spoke,
as well as the extent to which an eastern presence extended beyond the
elite stratum of high officials. Chapter 5 examines how far the Moorish
kingdoms of the African interior presented a viable alternative to Byzantine rule in the minds of sixth-century Romano-Africans and how the
rhetoric of non-Romanness was deployed against Moors for both political and cultural ends. Chapter 6 examines Byzantine attempts to legitimate imperial power in Africa and reconsiders the notion of African
resistance to the sixth- and seventh-century empire. Finally, Aftermath
brings some additional evidence from the Islamic period to bear on the
larger questions that this study raises, while my conclusions recapitulate
the argument of the book as a whole.
The study of Romanness in late antique North Africa returns the
region to the position of central importance it enjoyed in the ancient
world, challenging still-lingering scholarly perceptions of African resistance to the late Roman empire and its involvement in African affairs.
Ultimately, the present study tests our understanding of the collapse
17

Staying Roman
of the Roman empire in the West and the transition between classical
antiquity and the early Middle Ages; it illuminates the long-term success of the Roman state in shaping attitudes, outlooks, and beliefs in
the Mediterranean world; and it highlights the problematic questions of
what it meant to become, to be, and to stay Roman.

18

Chapter 1

T H E L E G I T I M A T I O N O F VA N D A L P OWE R

Legitimacy, like beauty, is in the eye of the beholder. The Vandals

peregrinations through the western Roman empire were thought of
as invasions by the provincial intellectuals who wrote the sources that
survive to us. The intruders were seen as barbarians who had erupted
over the frozen Rhine into Gaul in 406, traversed the Pyrenees into Spain
in 409, passed into Mauretania in 429, and thence to Africa Proconsularis
a decade later. Their movements were assisted at every turn by the
treachery and betrayal of Romans: their passage into Mauretania was
the result of sedition; their capture of Carthage, the result of duplicity.
How the Vandals themselves perceived their progress through the western
empire, we have no way of knowing. We hear their voices only rarely
in the sources, and only after they came to Africa. They never speak of
what came before.
Once they had gained control of Carthage, however, the Vandals were
faced with a serious challenge: how to cope with the disruption of
Roman power in Africa. For the empire, the loss of one of its richest
agricultural provinces and a keystone of eastwest communications in
the Mediterranean was a crushing blow. Re-securing the region through
conquest or diplomacy was an unambiguous priority. But having conquered the region by force, the Vandals, too, had to secure it. They
had only limited numbers with which to accomplish this end. The only
surviving estimate, however unreliable it may be, places the number of
Vandals old, young, children, slaves and masters, even he whom the
belly of the womb poured forth into this [worlds] light that very day
at 80,000 at the time of their passage from Spain into Mauretania in
429.1 This would have represented only a fraction of the population of
1

Vict. Vit. 1.2, p. 3: usque ad illam diem quem huic luci uterus profuderat uentris . . . senes, iuuenes, paruuli, serui uel domini. On the question of the reliability of this estimate, see L. Schmidt,
Zur Frage nach der Volkszahl der Wandalen, Byzantinische Zeitschrift 15 (1906), pp. 6201 and
C. Courtois, Les Vandales et lAfrique (Paris, 1955), pp. 21517, with the sceptical assessment of
W. Goffart, Barbarians and Romans, ad 418584: The Techniques of Accommodation (Princeton, NJ,
1980), pp. 2314 and A. Merrills and R. Miles, The Vandals (Chichester, 2010), pp. 489.

19

Staying Roman

Map 1.1.

The Mediterranean world

Carthage, let alone the rest of Roman Africa. The security of Vandal
rule would therefore rely on the co-optation of Romans at home and
the good will of Romans abroad, and, when these failed, the continued
ability both to mobilize Vandals themselves and to persuade other barbarian rulers to join the African kingdom as allies. Thus the Vandal kings
had to legitimate their power to four audiences: the Roman emperors
who could (and did) try to eject the Vandals from their newly won
prize; the other barbarian kings throughout the late antique West whose
variable allegiances could quickly vacillate between Roman, Vandal, and
calculated non-intervention; the Vandals own newly conquered Roman
subjects; and the conglomeration of barbarians through the strength of
whose arms the Vandal kings had won their kingdom in the first place
(see Map 1.1).
1. the imperial woman: eudocia
There can be no question of the ideological significance to the Vandals
of the capture of Carthage from Roman control in ad 439. Geiseric, the
most accomplished of the Vandal kings and the man who had led his
20

Vandal power
people into Africa, dated his reign from the seizure of the metropolis;
not from his succession (together with his half-brother Guntharic) to the
Vandal throne some time after 411, nor yet from his assumption of sole
power upon Guntherics death in 428.2 Selecting the fall of Carthage as
his first regnal year sent a clear message: it was control of the city that
made Geiseric a king.3 Later Vandal rulers incorporated the metropolis
into their regnal formulas; at least, Gunthamund (ad 48496) dated his
reign to the year of Carthage.4 Coins struck in the Vandal kingdom also
celebrated the African metropolis and its foundation legend, as did court
poetry and less official media such as mosaics.5
The Vandal seizure of Carthage also forced the Roman emperor in the
West, Valentinian III (ad 42555), to renegotiate the terms under which
the Vandals held their portion of Africa. Each of the arrangements that
had settled the Vandals in Roman territory prior to 439 (in Spain in 411
and in Mauretania and Numidia in 435) had established the barbarians
as foederati, allies or federates of the empire, to which they owed military
service.6 The alliance, however such as it was proved to be an uneasy
one at the best of times, and more often than not fictive. Under the
settlement of 442, it seems to have been abandoned: the Vandals were no
longer bound to the empire as federates.7 Our best account of the treaty
2

4
5

Joint succession: Proc. BV 1.3.23, 1:322. Sole power: Hydatius, Chronicon 79 (ad 428), ed. and
trans. R. W. Burgess, in The Chronicle of Hydatius and the Consularia Constantinopolitana: Two
Contemporary Accounts of the Final Years of the Roman Empire (Oxford, 1993), p. 88; on Hydatius
and his chronicle, see Burgess, Chronicle, pp. 310 and S. Muhlberger, The Fifth-century Chroniclers:
Prosper, Hydatius, and the Gallic Chronicler of 452, ARCA Classical and Medieval Texts, Papers and
Monographs 27 (Leeds, 1990), pp. 193266. Geiserics regnal date: Laterculus regum Wandalorum
et Alanorum 3, ed. T. Mommsen, MGH AA 13/Chron. Min. 3 (Berlin, 1898), p. 458; and F. M.
Clover, Timekeeping and Dyarchy in Vandal Africa, Antiquite tardive 11 (2003), pp. 503. On
the Laterculus regum Wandalorum, see R. Steinacher, The So-Called Laterculus Regnum Vandalorum
et Alanorum: A Sixth-Century African Addition to Prosper Tiros Chronicle?, in A. H. Merrills
(ed.), Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot, 2004),
pp. 16380. On regnal dating, see also below, Chapter 3.2.3.
Clover, Timekeeping and Dyarchy, p. 46 argues that Geiserics claim to the very title rex dated
to c.439, and that to that point the Vandal kings could more accurately be described as phylarchs
or tribe leaders.
J.-P. Bonnal and P.-A. Fevrier, Ostraka de la region de Bir Trouch, Bulletin darcheologie algerienne
2 (19667), pp. 23942 and 2445, nos. 12 and 4; Clover, Timekeeping and Dyarchy, pp. 549.
F. M. Clover, Felix Karthago, Dumbarton Oaks Papers 40 (1986), pp. 116. For a good recent
synthesis on Vandal coinage, see C. Morrisson, LAtelier de Carthage et la diffusion de la monnaie
frappee dans lAfrique vandale et byzantine (439695), Antiquite tardive 11 (2003), pp. 6674.
This, at least, would seem to be the implication of Hydatius, Chronicon 41 (ad 411), p. 82 and
Prosper, Epitoma de chronicon s.a. 435, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin,
1892), p. 474. On Prosper and his chronicle, see Muhlberger, Fifth-Century Chroniclers, pp. 48135.
Prosper, Chronicon s.a. 442, p. 479; F. M. Clover, Flavius Merobaudes: A Translation and Historical
Commentary, Transactions of the American Philosophical Society, n.s. 61/1 (Philadelphia, Pa.,
1971), pp. 534 and F. M. Clover, Geiseric the Statesman: A Study of Vandal Foreign Policy,
Ph.D. diss., University of Chicago (1966), pp. 8990.

21

Staying Roman

Map 1.2.

Late antique North Africa, 439700

of 442 speaks unambiguously of the Vandals and the Romans dividing

Africa between them.8 Other sources indicate that the Vandals received
the more fertile provinces of Africa Proconsularis and Byzacena as well as
eastern Numidia and western Tripolitania; the empire regained control of
the less fertile Mauretania Caesariensis, Mauretania Sitifensis, and western
Numidia (see Map 1.2).9 The empire may not have recognized the full
juridical independence of the Vandal kingdom, but the barbarians had
nevertheless secured their practical autonomy.10
Perhaps more important, however, in terms of legitimating Vandal
power in the western Mediterranean was the engagement almost
certainly at the same time of Geiserics eldest son Huneric to
8
9

10

Prosper, Chronicon s.a. 442, p. 479.

Vict. Vit. 1.13, p. 7; for the retrocession of the western provinces to the empire, see also
Valentinians laws for Africa in the 440s and 450s: Val. Nov. 13 (ad 445) and 34 (ad 451),
pp. 957 and 1401.

Y. Moderan, LEtablissement
territorial des Vandales en Afrique, Antiquite tardive 10 (2002),
pp. 8897.

22

Vandal power
Valentinians daughter, the imperial princess Eudocia.11 The Vandal
capture of Carthage doubtless forced the commitment. The Roman
state placed overwhelming importance on the African grain supply, the
preservation of which was so imperative that the eastern emperor Theodosius II (ad 40850) had sent his general Aspar to North Africa in 431 in
an effort to help contain the advance of the Vandals.12 The treaty negotiated upon Aspars recall to Constantinople in 434 had manifestly failed
in this respect. Faced with the loss of Africa Proconsularis and Byzacena
and the alarming threat this posed to the continued flow of African grain
to the rest of the empire, the western emperor Valentinian III needed to
guarantee the solidity of RomanVandal relations. He therefore granted
Geiseric the promise of a marriage alliance. In exchange, Geiseric seems
to have agreed to pay a yearly tribute, presumably in the form of continued shipments of grain to Italy, and to send his son, the Vandal prince
Huneric, to Valentinians court as a promise of good behaviour.13 The
resulting engagement of Huneric and Eudocia in 442 was a significant
break with tradition, which to that point had not contemplated an officially sanctioned marriage between an imperial princess and a barbarian
prince.14 The connection that this engagement gave the Vandal royal
family to the imperial house of Theodosius was to colour the relations
between the empire and the African kingdom for the rest of the Vandal
century.
Certainly Geiseric seems to have viewed the marriage alliance as giving
him and his lineage a particular kind of legitimacy not conferred by
mere possession of Carthage alone. At least in subsequent years, after
11

12

13
14

Merobaudes, Panegyricus II, ed. F. Vollmer, MGH AA 14 (Berlin, 1905), pp. 1118, esp. 12,
ll. 249 and Merobaudes, Carmen 1.1718, ed. Vollmer in MGH AA 14:3, with the commentary
of Clover, Merobaudes, pp. 24 and 514, and Clover, Geiseric, pp. 967.
Proc. BV 1.3.356, 1:324; Evagrius Scholasticus, Historia Ecclesiastica 2.1, ed. J. Bidez and L. Parmentier, in The Ecclesiastical History of Evagrius with the Scholia (London, 1898), p. 37; Theophanes,
Chronographia AM 5931 and 5943, ed. C. de Boor, 2 vols. (Leipzig, 18835; repr. Hildesheim,
1963), 1:95 and 104; John Zonaras, Epitomae Historiarum 13.24, ed. L. Dindorf, 6 vols. (Leipzig,
186875), 3:2456; for the date, see PLRE 2:1649, s.n. Aspar, at p. 166. F. M. Clover, Carthage
in the Age of Augustine, in J. H. Humphrey (ed.), Excavations at Carthage 1976, Conducted by the
University of Michigan, vol. 4 (Ann Arbor, Mich., 1978), pp. 1314. On the importance of Africa
to the eastern as well as the western grain supply, see M. McCormick, Origins of the European
Economy: Communications and Commerce, ad 300900 (Cambridge, 2001), pp. 100 and 106, and
M. McCormick, Bateaux de vie, bateaux de mort: maladie, commerce, transports annonaires et
le passage e conomique du Bas-empire au moyen a ge, in Morfologie sociali e culturali in Europa fra
tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sullalto Medioevo
45 (Spoleto, 1998), pp. 778.
Proc. BV 1.4.13, 1:326 with Clover, Merobaudes, p. 53; see also Merobaudes, Carmen 1.78, p. 3
with the commentary by Clover, Merobaudes, pp. 201.
On this point, see P. Heather, The Fall of the Roman Empire: A New History of Rome and the
Barbarians (Oxford, 2006), p. 292.

23

Staying Roman
the marriage had taken place, the Vandal king felt a certain freedom
to intervene in imperial politics in an attempt to set his brother-inlaw Anicius Olybrius a Roman senator and the husband of Eudocias
sister Placidia on the throne of the western empire.15 Priscus, a wellinformed contemporary, stated explicitly that it was Geiserics connection
to Olybrius by marriage ( x pigamav suggneia) that motivated the
Vandal kings efforts on the senators behalf; and from the vantage point of
the sixth century John Malalas thought it similarly plausible that this same
connection would have given Olybrius a licence to speak with Geiseric
(pahsa) that was not available to other late Roman aristocrats.16
Geiseric also sought control of that portion of Valentinians inheritance
that was due to Eudocia, an issue that seems to have remained a matter
of contention between the Vandals and the East Roman court until
Geiserics death in 477.17 For, quite apart from the strategic importance of
the lands Geiseric claimed, gaining recognition from the eastern imperial
court of Eudocias marriage to Huneric seems to have been one of the
driving forces of the Vandal kings raids into imperial territory in the
middle of the fifth century.18
The terrific importance to Geiseric of an imperial connection is
underscored by the kings willingness to sacrifice his relations with
the Visigothic kingdom in order to secure the marriage of Eudocia
to his eldest son. Before becoming engaged to the Roman princess,
the young Huneric had been married to the daughter of the Visigothic king Theoderic I.19 Under Roman (and Christian) law, however,
Huneric could not have become engaged to Eudocia while still married
to another woman.20 The unfortunate Visigothic princess was accused
of preparing poison one of the few grounds for divorce in the later
15

16
17
18
19

20

Priscus, fragments 38.1 and 38.2, ed. and trans. R. C. Blockley, in The Fragmentary Classicising
Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and Malchus, 2 vols. (Liverpool,
1983), 2:3402; Proc. BV 1.6.6, 1:336. See also John Malalas, Chronographia, ed. L. Dindorf,
Corpus Scriptorum Historiae Byzantinae 15 (Bonn, 1831), pp. 3734, reporting Leo Is fears that
Geiseric and Olybrius might have ambitions for the senators rule in the East. On Malalas and
his chronicle, see E. Jeffreys, B. Croke, and R. Scott (eds.), Studies in John Malalas, Byzantina
Australiensia 6 (Sydney, 1990).
Priscus, frags. 38.1 and 38.2, pp. 3401; cf. Proc. BV 1.6.6, 1:336 (kdov). Malalas, Chronographia,
p. 373.
Priscus, frags. 38.2 and 39.1, pp. 3402; Malchus of Philadelphia, frag. 17, ed. Blockley, in
Fragmentary Classicising Historians, 2:424.
R. C. Blockley, East Roman Foreign Policy: Formation and Conduct from Diocletian to Anastasius,
ARCA Classical and Medieval Texts, Papers and Monographs 30 (Leeds, 1992), p. 72.
Jordanes, Getica 36.184, ed. F. Giunta and A. Grillone, in Iordanis De origine actibusque Getarum,
Fonti per la storia dItalia 117 (Rome, 1991), pp. 789. The alliance between the Vandals and
the Visigoths is discussed by Clover, Geiseric, pp. 1059.
J. Evans Grubbs, Law and Family in Late Antiquity (Oxford, 1995), p. 143.

24

Vandal power
Roman empire and was returned to the court of her father with her
nose and ears cut off.21
What motivated the brutality of the Visigothic princesss punishment
is difficult to tell. Perhaps Geiseric did actually believe the allegations to
be true.22 The Vandal king must certainly have wanted to strike hard,
not only at the hapless princess, but at her father; for in the sixthcentury account of the Gothic historian Jordanes there is every indication that Geiseric knew what he was about. A mutilated princess who
would always appear like a repulsive corpse (turpe funus miseranda semper
offerret)23 would do her father no good as a pawn in any further
diplomatic alliance. But whatever Geiserics motivations, the fury which
Theoderic felt at the treatment of his daughter doubtless accounts at
least in part for the chill which settled upon VandalVisigothic relations
for the following quarter-century. Not until Euric assumed the Visigothic throne in 467 do we explicitly hear of further attempts at contact
between the two kingdoms.24 Upon his accession, Euric immediately
sent envoys to announce the event, first to the emperor and the king of
the Sueves, and then to the courts of the Ostrogoths and the Vandals.
The mission to the Vandals, however, was cut short, for the ambassadors,
terrified by a rumour of hostilities between the Romans and the Vandal
kingdom, returned home in haste.25 Even so, the Visigothic and Vandal
courts seem to have re-established ties at some point in the following
decade, for Jordanes tells us that Geiseric once sent gifts to Euric by
which the Vandal king managed to persuade the Visigoth to harass the
Romans.26
The report of Hunerics previous marriage highlights another aspect
of the union between the Vandal prince and the imperial princess. It

21

22
24

25
26

Jordanes, Getica 36.184, p. 79; Clover, Geiseric, p. 109. Poison as grounds for divorce: CTh
3.16.1 (ad 331), pp. 1556 and CTh 3.16.2 (ad 421), pp. 1578, discussed by Evans Grubbs, Law
and Family, pp. 22931 and 2347.
23 Jordanes, Getica 36.184, p. 79.
Clover, Geiseric, p. 109.
Hydatius, Chronicon 186 (ad 458), p. 110 records the arrival in the Suevic kingdom of Galicia
and Lusitania of Vandal envoys together with envoys of the Visigoths: Legati Gothorum et
Vandalorum pariter ad Sueuos ueniunt et reuertuntur. Clover, Geiseric, pp. 1745 reads this
as evidence of a temporary VandalVisigothic alliance, which is certainly possible. It is perhaps
worth noting that this would still post-date the death of Theoderic I in 452. The arrival of
the African embassy is also discussed by E. A. Thompson, The End of Roman Spain, Part I,
Nottingham Medieval Studies 20 (1976), p. 11.
Hydatius, Chronicon 234 and 236, pp. 11820.
Jordanes, Getica 47.244, p. 100. On Eurics campaigns in Gaul, see Continuatio Hauniensis Prosperi
s.a. 476 (Auctarii Haun. ordo prior), ed. Mommsen in MGH AA 9:309 (see also ibid., s.a. 486,
p. 313) and Chronica Gallica a. dxi 657, ed. Mommsen in MGH AA 9:665.

25

Staying Roman
seems to have mattered to Geiseric that Eudocia marry Huneric specifically, the oldest of the Vandal kings sons.27 The union, as we have
seen, did not come without a cost. But at some point before his death
Geiseric established that succession to the Vandal throne should pass
in perpetuity to his eldest male descendant.28 Geiserics decision to
marry Eudocia to Huneric rather than to one of his younger brothers,
Theoderic or Genton may thus have been inspired by the Vandal kings
desire to ensure that his heir, too, was connected to the Theodosian
house.
Whether Geiseric thought out the implications of all of these arrangements beyond the first generation is impossible to know. In the thirteen
years or so that separated Huneric and Eudocias engagement and their
eventual marriage, both of Geiserics younger sons had had their own
children, at least four of them males, each with a claim on the Vandal
kingship that was prior to that of Huneric and Eudocias son Hilderic.29
Shortly after his own accession to sole power in 428, though, Geiseric
himself had been willing to execute his brothers widow and sons in
order to solidify his position; and so the Vandal king may well have contemplated with equanimity the kind of fraternal bloodletting in which
Victor of Vita tells us Huneric did indeed engage in the 480s in an effort
to secure his own sons succession to the throne.30
In the early days of his reign, however, Geiseric was clearly willing to
fight and fight fiercely to secure and defend his connection to the
imperial family. In 455, the Vandal king launched a military expedition
against the city of Rome in which his troops were said to have stripped
the Palatine palace of its treasures, removed half of the gilded bronze
roof tiles from the Capitoline temple of Jupiter Optimus Maximus, and
carried off thousands of the citys inhabitants into slavery. Critically, the
Vandals also took into their custody the princess Eudocia, along with
27

28

29

30

On the sons of Geiseric, see PLRE 2:5723, s.n. Hunericus; ibid., pp. 5023, s.n. Genton
1; ibid., p. 1073, s.n. Theodericus 4; though surely Proc. BV 1.8.6, 1:346, v atn gr
crnov fere t prwtea to Gizercou gnouv, refers to Gunthamund then Geiserics oldest
surviving male heir not Genton as suggested by PLRE 2:502, s.n. Genton.
Vict. Vit. 2.13, p. 28; Jordanes, Getica 33.169, p. 72; Proc. BV 1.7.29, 1:3445; Courtois, Vandales,
pp. 23842; H. Wolfram, The Roman Empire and its Germanic Peoples, trans. T. Dunlap (Berkeley,
Calif., 1997), p. 165. On the flexibility of Vandal succession in practice, see Merrills and Miles,
Vandals, pp. 747.
Sons of Theoderic: Vict. Vit. 2.1214, pp. 289 (one of whom was still an infantulus in the
480s and therefore younger than Hilderic). Sons of Genton: Vict. Vit. 2.14, p. 29; Proc. BV
1.8.6 and 1.8.8, 1:346; PLRE 2:515, s.n. Godagis; ibid., pp. 5256, s.n. Gunthamundus; ibid.,
pp. 111617, s.n. Thrasamundus 1.
Vict. Vit. 2.14, p. 29.

26

Vandal power
her mother and sister.31 The imperial women were brought to Africa,
where at long last Geiseric married his son to the daughter of Valentinian,
though perhaps not until the following year.32
The attack must be read in the light of the threat that the emperors
assassination earlier that same year and the accession of the usurper Petronius Maximus posed to the engagement of the Vandal prince and the
imperial princess. Eudocia had been only perhaps four years old at the
time of her engagement to Huneric in 442 too young to be wed
under Roman law and so the marriage itself was postponed for a time.
Even after his daughter reached a marriageable age, however, Valentinian
delayed following through on his promises to Geiseric, perhaps in part
at the behest of the leading general Flavius Aetius, at the time the most
influential man in the western empire, who himself harboured ambitions of establishing a marriage tie to the imperial house. The result was
that Huneric and Eudocia remained unmarried during the emperors
lifetime.33 Moreover, Valentinians killing of Aetius in 453 set in motion
a series of events that led to the emperors own assassination two years
later, and the usurpation of the senator Petronius Maximus. Upon seizing
the empire, Maximus may even have gone so far as to break the longstanding pledge of union between Huneric and Eudocia and to have
married his own son to the princess, for the Spanish chronicler Hydatius
tells us that Maximus married his son Palladius to the daughter of Valentinian (Valentiniani filia).34 While Hydatius does not specify which of the
emperors two daughters was united with the new Caesar, Valentinians
younger daughter, Placidia, had probably already wed Olybrius by the
time of her fathers death.35 At least, our earliest and best sources seem
to imply that the senator and the princess were already married when
31

32
33
34
35

Prosper, Chronicon s.a. 455, p. 484; Sidonius Apollinaris, Carmen 7.4419, ed. A. Loyen in Sidoine
Apollinaire, 3 vols. (Paris, 196070), 1:712; Hydatius, Chronicon 160 (ad 455), p. 104; Priscus,
frags. 30.13, pp. 3302; Jordanes, Romana 334, ed. T. Mommsen, MGH AA 5 (Berlin, 1882),
p. 43; Jordanes, Getica 45; Chronica Gallica a. dxi 623, p. 663; Proc. BV 1.5.15, 1:3312; Vict.
Tonn. s.a. 455.15, pp. 78; Malalas, Chronographia, pp. 3656; Theophanes, Chronographia AM
5947, 1:1089. See also Liber pontificalis 47.6, ed. T. Mommsen, MGH Gesta Pontificum Romanorum 1 (Berlin, 1898), p. 104, though it is not clear from this account whether churches were
subject to plunder. CIL 6.1663, 6.31890, and 6.31891 seem to refer to the restoration of public
buildings in the wake of the Vandal incursion. CIL 6.1750 is probably earlier; CIL 6.526, probably
later.
Courtois, Vandales, pp. 3967, no. 17, Eudocie, and Clover, Geiseric, p. 186.
Clover, Geiseric, pp. 12936. On Aetius, see PLRE 2:218, s.n. Aetius 7.
Hydatius, Chronicon 155 (ad 455), p. 104.
Thus PLRE 2:887, s.n. Placidia 1. However, F. M. Clover, The Family and Early Career of
Anicius Olybrius, Historia: Zeitschrift fur Alte Geschichte 27 (1978), pp. 17482 argues that the
two were only engaged in 455.

27

Staying Roman
(after seizing her with her mother and sister in 455) Geiseric eventually
returned Placidia to imperial custody in 462.36 It thus seems likely that
the daughter in question was Eudocia.37
As Frank Clover has rightly observed, in breaking the princesss
engagement to Huneric, Maximus risked Geiserics anger.38 Indeed, the
sixth-century African chronicler Victor of Tonnena even indicates that
Maximus expected Geiseric to attack Rome, but that the Vandals arrived
earlier than the usurper had anticipated.39 But if Olybrius was in fact
already married to Placidia, Maximus may have felt that he had very
little choice in the matter. The 450s were a critical period of dynastic
transition in the empire, for they saw the extinction of the male line of
the Theodosian house, which by then had ruled the Roman world for
three-quarters of a century. This happened first in the East with the death
of Theodosius II while hunting in 450; the assassination of his cousin
Valentinian III five years later had the same effect in the West. As the line
of Theodosian emperors expired, a solution to the worrying problem of
dynastic discontinuity (and the threat of civil war) was initially found in
political marriage. In the East, the last member of that branch of the
imperial family, Theodosius sister Aelia Pulcheria, agreed to marry
the tribune Marcian (ad 4507), whom she then raised to the purple.
The symbolic importance of this act is highlighted by the fact that Pulcheria was both a dedicated virgin and, at fifty-one, probably already past
her childbearing years. Even in its conception, the union was to remain
childless; but for the moment it provided an heir to the imperial throne
of unquestioned legitimacy. In the West, too, the major political actors
seem initially to have believed that after Valentinians death succession

36

37

38

Priscus, frag. 38.1, p. 340; Hydatius, Chronicon 211 (ad 462), p. 114; Proc. BV 1.5.6, 1:332; and,
much later, Theophanes, Chronographia AM 5947, 1:109. See also Malalas, Chronographia, p. 366,
who explicitly calls Placidia the wife of Olybrius ( gun to patrikou Olubrou); Malalas,
however, also claims that Theodosius II allowed Eudocia to languish in Africa as a punishment
for having sought the aid of the Vandals, when in fact Theodosius had died five years earlier.
Evagrius, Historia Ecclesiastica 2.7, pp. 545 seems to indicate that Olybrius and Placidia were
married only after the princesss return to Constantinople, which Evagrius places in the reign
of Marcian (ad 4507). In the twelfth century, Zonaras, Epitomae 13.25, 3:250 indicates that
Olybrius and Placidia were engaged but presumably not married at the time of Placidias
African sojourn.
PLRE 2:4078, s.n. Eudocia 1; contra S. I. Oost, Aetius and Majorian, Classical Philology 59
(1964), pp. 278 and Clover, Geiseric, pp. 1434. Malalas, Chronographia, p. 366 twice calls
Eudocia a parqnov (virgin or young woman) at the time of her marriage to Huneric; but on
the relationship between parqnov and neniv in general, see A. Kamesar, The Virgin of Isaiah
7:14: The Philological Argument from the Second to the Fifth Century, JThS n.s. 41 (1990),
pp. 5175 (thanks to Emmanuel Papoutsakis for this reference, and for a useful discussion of the
significance of the word parqnov in Malalas Syriac milieu).
39 Vict. Tonn. s.a. 455.14, p. 7.
Clover, Geiseric, p. 144 n. 2.

28

Vandal power
to the empire would follow from marriage into the Theodosian family.
The usurper Maximus not only wed his son to the former emperors
daughter; in a bid to legitimate his own seizure of power, the senator himself also forcibly married Valentinians widow Eudoxia (herself a
daughter of Theodosius II).40 Geiseric, too, seems to have continued to
harbour hopes of proximity to imperial power in the post-Theodosian
empire, though as we have seen his ambitions seem to have centred not
so much on himself or even his son as on his (prospective) brother-inlaw Olybrius. For a time he may perhaps also have enjoyed a working
relationship with the dowager empress Eudoxia herself. At least, in the
East it was rumoured that Geiseric had attacked Rome only after the
imperial widow appealed to him for aid.41 Thus, far from dying with
Valentinian, the significance of a marriage tie to the Theodosian house
may even have increased in the wake of the emperors assassination, at
least in the immediate term.
It is perhaps no surprise, then, that this moment accompanies a shift
in the quantity of the surviving information about the mechanics of
Vandal-imperial relations. Contact, of course, had existed between the
Vandal kingdom and the empire before 455. As we have seen, Valentinian
III negotiated two treaties with Geiseric, the first in 435, establishing the
Vandals in the Mauretanias and western Numidia as foederati, and the
second in 442, retroceding these territories to the empire and giving
the Vandals Africa Proconsularis, Byzacena, eastern Numidia and western Tripolitania.42 Huneric was sent as a hostage to the western imperial
40

41

42

Forced marriage of Eudoxia: Hydatius, Chronicon 155 (ad 455), p. 104; Vict. Tonn. s.a. 455.14,
p. 7; Proc. BV 1.4.36, 1:330. On this period in general, see P. Heather, The Western Empire,
42576, in A. Cameron, B. Ward-Perkins, and M. Whitby (eds.), Late Antiquity: Empire and
Successors, ad 425600, Cambridge Ancient History 14 (Cambridge, 2000), pp. 1827; A. D. Lee,
The Eastern Empire: Theodosius to Anastasius, ibid., pp. 4252; and K. Holum, Theodosian
Empresses: Women and Imperial Dominion in Late Antiquity, Transformation of the Classical Heritage
3 (Berkeley, Calif., 1982). See also PLRE 2:4078, s.n. Eudocia 1; ibid., pp. 41012, s.n.
Eudoxia 2; ibid., pp. 71415, s.n. Marcianus; ibid., p. 887, s.n. Placidia 1; ibid., pp. 92930,
s.n. Pulcheria; ibid., p. 1100, s.n. Theodosius 6; ibid., pp. 11389, s.n. Valentinianus 4; and
ibid., pp. 13089, stemmata 1 and 3.
Priscus, frag. 30.1, p. 330, picked up and repeated as fact by Marcellinus comes, Chronicon
s.a. 455.3, ed. T. Mommsen, MGH AA 11/Chron. Min. 2 (Berlin, 1894), p. 86; Proc. BV
1.4.369, 1:3301; Jordanes, Romana 334, p. 43; Malalas, Chronographia, p. 365; Evagrius, Historia
Ecclesiastica 2.7 and 4.17, pp. 54 and 167; Theophanes, Chronographia AM 5947, 1:108; Zonaras,
Epitomae, 13.25, 3:249; George Cedrenus, Compendium Historiarum, ed. I. Bekker, 2 vols., Corpus
Scriptorum Historiae Byzantinae 345 (Bonn, 18389), 1:606. In the western sources, the rumour
is repeated only in Hydatius, Chronicon 160 (ad 455), p. 104: Gaisericus sollicitatus a relicta
Valentiniani, ut malum fama dispergit . . . Romam ingreditur, on which, see Thompson, End
of Roman Spain, Part I, pp. 1112. The rumour is discussed and rejected by Clover, Geiseric,
pp. 1505. On the forced marriage of Eudoxia and Petronius Maximus, see PLRE 2:411, s.n.
Eudoxia 2.
Clover, Geiseric, pp. 54 and 87100.

29

Staying Roman
Table 1.1. Roman embassies to the Vandals, 45584
Name

Rank

From

Date

Bleda
Anonymous 1
Anonymous 2
Phylarchus

Arian bishop

Tatianus
Severus
Alexander
Reginus
Uranius

patricius
patricius
vir inlustris

Marcian (East)
Avitus (West)
Ricimer MVM (West)
Leo I (East)
Leo I (East)
Leo I (East)
Zeno (East)
Zeno (East)
Zeno (East)
Zeno (East)

455/6
455/6
461
462/3
467
c.464
474
c.480/1
483
484

vir devotus, agens in

rebus

court and returned to Africa, probably before 446.43 In 454, Valentinians

intervention secured the ordination of a new Nicene bishop of Carthage,
Deogratias, after the see had languished without a metropolitan for perhaps twelve or fourteen years.44 And again, Eudoxia may have appealed
to Geiseric for aid after her forced marriage to Petronius Maximus in
455. But with the removal of the empress and her daughters to Africa,
we begin to hear of specific embassies sent between the empire and the
Vandal kingdom.
Initially this exchange seems to have focused on the return of the
imperial women and the related issue of Vandal raiding. Thus, for example, in 455 or 456, the eastern emperor Marcian sent an Arian bishop
named Bleda to Carthage in an attempt to secure the release of Valentinians widow and daughters. However, neither Bledas arguments nor
his threats were able to persuade Geiseric, and the bishop returned to the
East empty-handed. In the same year, the short-lived western emperor
Avitus (ad 4556) also sent an embassy to Geiseric, threatening war if the
Vandal king broke the old treaty between the two powers.45 But break the
treaty Geiseric did. The following two decades saw numerous Vandal raids
in the central Mediterranean, primarily in the western islands, southern
Italy, and the Adriatic littoral.46 Geiseric openly demanded the properties of Valentinian that were due to Eudocia as her inheritance and
43

44
45

Huneric as hostage: Proc. BV 1.4.1314, 1:326. Clover, Geiseric, pp. 989, argues that Huneric
had probably returned to Africa by the time Merobaudes delivered his panegyric in honour of
Aetius third consulate on 1 Jan. 446, and certainly by the time of the empress Gallia Placidias
death on 27 Nov. 450.
Vict. Vit. 1.24, p. 11 with Prosper Tiro add. s.a. 454, ed. T. Mommsen in MGH AA 9:490; PCBE
1:2713, s.n. Deogratias 1, p. 271; and Clover, Geiseric, pp. 1346.
46 Merrills and Miles, Vandals, p. 118.
Priscus, frag. 31.1, p. 334.

30

Vandal power
the recognition of the legitimacy of the Vandal connection to the Theodosian house that this implied which the western court consistently
refused to supply.47 But as we have seen, Geiseric was also said to have
secretly hoped for Olybrius accession to the western throne, and thus
a continuation of his own personal connection to the imperial centre.48
Under Avitus successor Majorian (ad 45761) the situation degenerated into war. Outmanoeuvred by Geiseric, the western emperor was
compelled to make peace on shameful terms (sunqkai ascra);
when he returned to Italy he was executed by Ricimer, then the leading general in the West and the power behind the western throne.49 In
461, Ricimer sent an embassy to Geiseric protesting the Vandal kings
attacks on Italy and Sicily, and in 461 or 462 the eastern emperor Leo I
(ad 45774) sent his envoy Phylarchus on an embassy to Geiseric to the
same end. The latter mission resulted in a treaty between Geiseric and
Leo in which the Vandal king seems to have agreed to remand Eudoxia
and Placidia to Constantinople, in exchange for eastern recognition of
Hunerics marriage to Eudocia. The deal apparently also involved the
release to the princess of some of her fathers eastern properties and
probably also included a pledge on Leos part to support Olybrius candidacy for the western throne when it next became vacant. Geiseric,
however, refused to cease his attacks on Italy and Sicily until the West
Roman court surrendered Eudocias western inheritance.50
Perhaps two years later, Leo again sent an envoy Tatianus to
Carthage on behalf of the Italians. This time, however, Geiseric refused
even to grant the embassy an audience, and so Tatianus returned to Constantinople having accomplished nothing.51 East Roman relations with
the Vandal kingdom deteriorated still further in 467, when Leo nominated the eastern aristocrat Anthemius to fill the then-vacant western
throne, rather than Geiserics candidate Olybrius. Leo sent Phylarchus
back to Geiserics court, threatening to invade Africa if the Vandal king
did not surrender his claims to Italy and Sicily. Still Geiseric refused to
accept the emperors demands, complaining that the eastern emperor
had broken his word, and again war was the result.52 Once more the
Vandal king proved capable of outmanoeuvring his enemies: the massive
assault launched on Africa by land and by sea came to naught and in
47
48
49
50
51
52

Priscus, frags. 38.2 and 39.1, p. 342; Blockley, East Roman Foreign Policy, p. 72.
Priscus, frag. 38.2, pp. 3402.
Priscus, frag. 36.2, p. 338; Clover, Geiseric, p. 179. On the conflict, see Merrills and Miles,
Vandals, pp. 11920.
Priscus, frags. 38.1 and 39.1, pp. 340 and 342; Blockley, East Roman Foreign Policy, p. 72.
Priscus, frags. 41.12, p. 346.
Priscus, frag. 52, p. 360; Merrills and Miles, Vandals, p. 121.

31

Staying Roman
468 Leo and Geiseric negotiated another treaty.53 Finally, in 474, the
new emperor in the East, Zeno (ad 47491), sent a Constantinopolitan senator by the name of Severus to Carthage to seek an end to the
raids and the establishment of a lasting peace with the Vandals. By this
point, the aging Geiserics long-cherished hopes for an intimate connection to the western imperial centre will have been dashed. Eudocia had
borne Huneric a male heir descended from the house of Theodosius,
and Olybrius had ascended the western imperial throne in 472, just as
Geiseric had hoped; but the new emperor had died suddenly in that
same year.54 The Theodosian connection remained prestigious after his
passing, but it was no longer politically useful. At much the same time,
Geiseric seems to have lost interest in Eudocia. According to the early
ninth-century Byzantine chronicler Theophanes, after sixteen years in
Africa the princess was allowed to travel to Jerusalem, where she died
after only a few days.55 In any case, Severus mission was an unparalleled success. Despite some occasional later friction, the peace which he
negotiated lasted until 5334, when the armies of the eastern emperor
Justinian reconquered Africa for Rome.56 But Severus also secured the
release of many recently seized Roman captives, and persuaded Geiseric
to recall the Nicene clergy of Carthage from exile and to allow Nicene
Christians to worship with some measure of tolerance.57
Zenos first embassy to the court of Huneric (ad 47784) was equally
successful. In what was surely a calculated concession to Vandal dynastic claims, the eastern emperor Zeno sent Alexander, the guardian of
Hunerics sister-in-law Placidia, as an ambassador to the new king in 478,
the year following his succession to the throne. The move seems to have
worked. Alexander returned to Constantinople with envoys from Africa
who brought word that Huneric was pleased that Zeno had honoured
Placidia, that the new Vandal king loved all things Roman (strgoi t
Rwmawn), and that he was abandoning his fathers claims on Eudocias
inheritance.58 As a concession to the emperor and to his own sister-inlaw, Huneric also allowed the Nicene Carthaginian church to ordain a
new metropolitan bishop after the see had lain vacant for twenty-four
years.59
53
54
55
56

57
59

Blockley, East Roman Foreign Policy, p. 76; Merrills and Miles, Vandals, pp. 1212.
PLRE 2:7968, s.n. Olybrius 6, pp. 7978.
Theophanes, Chronographia AM 5964, 1:118; see also Zonaras, Epitomae 13.25, 3:250 and
Nicephorus Callistus, Historia Ecclesiastica 15.12, PG 147:40B.
Malchus, frag. 5, p. 410; Proc. BV 1.7.268, 1:344, but see also ibid., 1.16.1314, 1:384, where
Justinian is at pains to claim that he was not breaking the treaty. Blockley, East Roman Foreign
Policy, pp. 7980.
58 Malchus, frag. 17, p. 424.
Malchus, frag. 5, p. 410, and Vict. Vit. 1.51, p. 22.
Vict. Vit. 2.34, p. 25.

32

Vandal power
Presumably as a result of these actions, Huneric established himself as
a friend of the emperor (flov te t basile).60 The relationship was
official and diplomatic: the phraseology here is evocative of the title rex
sociusque et amicus (king, ally, and friend) traditionally granted by the
Roman state to its client-kings.61 As such, Zeno and Huneric probably
renewed a formal association that had existed between the Vandal kingdom and the empire (or at least its western half) from 442 down to the
450s,62 but that had fallen into abeyance thereafter as the empires new,
post-Theodosian rulers once again sought a military solution to the threat
posed by Vandal raids, by the Vandal kings insistence on his connection
to the old imperial house, and by the claims to strategic Roman territories and to the right to intervene in questions concerning the imperial
succession that (in his mind) this entailed. At least, the sixth-century
Byzantine historian Procopius seems to have understood Geiserics status
as friend and ally (flov te ka xmmacov) as playing an important role in
structuring the Vandals relations with the West Roman imperial family
down to 455.63 Yet John Malalas, Procopius younger contemporary, was
aware that this status no longer characterized RomanVandal relations
in the 470s, and thought it plausible only that at that time the eastern
emperor Leo I might desire to have Geiseric as his friend (flov mou).64
Once re-established, the diplomatic friendship between the empire
and the Vandal kingdom seems to have proven remarkably resilient. To
be sure, it seems to have become strained towards the end of Hunerics
reign. Zenos envoy Reginus was present in Carthage in May of 483
when the Vandal king had an edict publicly read out insisting that the
regions Nicene bishops come to Carthage in the following year to debate
the principles of Christian faith.65 A year later another of Zenos legates,
Uranius, was sent to Hunerics court in an attempt to persuade the king
to stop the ensuing persecution of Africas Nicene Christians.66 And we
hear nothing of VandalRoman relations under Hunerics nephew and
successor, Gunthamund (ad 48496); though our evidence for contact
60
61

62
63
65

Malchus frag. 17, p. 424.

territorial, pp. 924; D. Braund, Rome and the Friendly King: The CharModeran, LEtablissement
acter of the Client Kingship (London, 1984), pp. 2337. In classical antiquity the title was granted
most famously to Herod (see, e.g., Braund, Rome, pp. 245); in Africa it was also extended
to the Mauretanian king Ptolemy: Tacitus, Annales 4.26, ed. H. Heubner (Stuttgart, 1983),
p. 147. On the meaning of this phrase, see also A. Rodolfi, Procopius and the Vandals: How
the Byzantine Propaganda Constructs and Changes African Identity, in G. Berndt and R.

Steinacher (eds.), Das Reich der Vandalen und seine (Vor-) Geschichten, Osterreichische
Akademie der
Wissenschaften, Philosophisch-historische Klasse, Denkschriften 366/Forschungen zur
Geschichte des Mittelalters 13 (Vienna, 2008), pp. 23342.

territorial, pp. 923; Heather, Fall of the Roman Empire, p. 292.

Moderan, LEtablissement
64 Malalas, Chronographia, pp. 3734.
Proc. BV 1.4.39, 1:331.
66 Vict. Vit. 3.32, p. 88.
Vict. Vit. 2.389, pp. 389.

33

Staying Roman
between the two powers in general diminishes after Hunerics death
in 484. Procopius our major textual source for late fifth- and early
sixth-century Africa was not particularly interested in the niceties of
diplomatic exchange between the Vandals and Constantinople. However,
Procopius does tell us that the Vandal king Thrasamund (ad 496523)
was exceptionally well disposed towards the emperor Anastasius or his
particular friend (flov v t mlista) and that the same was true
of Hilderic (ad 52330) and Justinian.67 Envoys also passed between
Justinians uncle and predecessor, Justin, and Thrasamund in 519.68 Thus
the empire and the Vandal kingdom appear to have continued to enjoy
the peace of a negotiated treaty well into the sixth century.
After Hilderics accession, in fact, we are told that imperialVandal
relations were strong enough to make the Ostrogothic king Theodoric
think twice about launching an expedition against Africa.69 Here the
official friendship that the empire and the Vandal kingdom enjoyed was
only part of the story, for Hilderic was also said to have been Justinians
guest-friend (xnov).70 This in turn would seem to imply that the Vandal prince had spent time in Constantinople, most probably during the
reign of his cousin Thrasamund, at some point between the accession of
Justinians uncle Justin to the imperial throne in 518 and Hilderics own
succession to the Vandal kingdom five years later in 523. Hilderic will
thus have been a mature man at the time of his sojourn in the imperial
capital. Considered an old man when he was deposed from his throne in
530, Hilderic had been born between c.456 and c.471 (the marriage of his
parents and the death of his mother); if his stay in Constantinople overlapped with Justinians period of power, then the prince must have been
somewhere between the ages of 46 and 67 at the time.71 In the imperial
capital, the Vandals descent from the Theodosian house would unquestionably have been recognized and honoured. Ten or fifteen years after
his visit, in the wake of the fall of the Vandal kingdom, Justinian granted
Hilderics children copious amounts of money (crmata kan) as the
great-grandchildren of Valentinian III.72 Indeed, once acknowledged,
the Vandal royal familys connection to the house of Theodosius was
67
68

69
70
72

Proc. BV 1.8.14, 1:347 (Thrasamund and Anastasius) and ibid., 1.9.5, 1:351 (Hilderic and
Justinian).
Collectio Avellana 212, ed. O. Guenther, in Epistulae imperatorum, pontificum, aliorum inde ab a.
ccclxvii usque ad a. dliii datae Avellana quae dicitur collectio, 2 vols., CSEL 35 (Vienna, 18958),
2:6701.
Proc. BV 1.9.5, 1:3512. Hilderic is also known to have sent at least one embassy to the court of
Justin: Proc. BV 1.9.8, 1:352.
71 An old man: Proc. BV 1.9.10, 1:353.
Proc. BV 1.9.5, 1:351.
Proc. BV 2.9.13, 1:458.

34

Vandal power
probably seen by all parties as something that could draw the empire and
the African kingdom closer together.
Those embassies that we can see in greater depth show that the Roman
emperors took some care in their selection of envoys to send to the
Vandals.73 Marcians choice of an Arian bishop to lead the mission to
Geiseric in 455 or 456 was doubtless calculated to appeal to the Vandal
kings religious sensibilities. Alexanders selection as an envoy to Huneric
was, as we have seen, certainly a concession to Hunerics connections
to the Theodosian house. Tatianus had already enjoyed a distinguished
career in the Roman civil service as urban prefect of Constantinople, and
perhaps as governor of Caria, when Leo selected him to lead an embassy
to the Vandals.74 All three of the Roman envoys sent to the African
kingdom between c.464 and c.481 were of the illustris grade the highestranking grade of senators at the time of their appointment, and two of
them were patricians. Nor was this entirely a coincidence. The historian
Malchus tells us that Zeno elevated Severus to patrician status specifically
to lend his embassy greater weight.75 Perhaps significantly, Uranius
who, though probably a career diplomat, was not of senatorial status
not only saw his mission to Hunerics court fail, he was forced to witness
atrocities committed against the very Nicene Christians whose interests
he had been sent to Africa to defend.76 The move was surely a deliberate
humiliation of the ambassador on Hunerics part. Of course, the failure
of Uranius embassy cannot be blamed on the envoys comparatively
humble status; patrician status did not guarantee a missions success either.
Geiseric, after all, had refused even to meet with the first Roman patrician
known to have been sent on an embassy to the Vandal kingdom. Even
so, the dispatch of high-ranking envoys flattered the Vandal kings, for
ambassadors of the illustris grade were only regularly sent to the Persians.77
As for Vandal embassies to the empire, we are almost completely ignorant. As we have seen, Huneric sent envoys back to Constantinople with
Alexander; but we hear nothing specific about them. In any case, it was
the reception of ambassadors that provided the Vandal kings with their
greatest opportunity to impress the splendour, majesty, and strength of
Vandal power upon the emperors representatives.78 This point was made
73

74
75
76
77
78

On the selection of envoys in general, see A. Gillett, Envoys and Political Communication in the
Late Antique West, 411533, Cambridge Studies in Medieval Life and Thought, 4th ser., 55
(Cambridge, 2003), pp. 2318.
For Tatianus career, see PLRE 2:10534, s.n. Tatianus 1.
Malchus, frag. 5, p. 410.
Vict. Vit. 3.32, p. 88. For his earlier career, see PLRE 2:11867, s.n. Vranius 4.
Blockley, East Roman Foreign Policy, p. 153.
On this point in general, see Gillett, Envoys, pp. 24459.

35

Staying Roman
quite forcefully by Geiseric just before the arrival of Severus embassy
in Africa. Learning of the departure of the diplomatic mission, Geiseric
sent a raid against the Roman city of Nicopolis, capturing and enslaving its inhabitants. Upon his arrival in Carthage, Severus complained to
Geiseric about the raid. Geiseric is said to have replied that he attacked
Nicopolis as an enemy of the Roman empire, but that now the embassy
had arrived he was willing to discuss a treaty. Impressed with Severus
character, Geiseric gave the ambassador those captives that had fallen to
the lot of the king and his family.79 The message could hardly be clearer:
the Vandals made dangerous enemies, but peace would bring its own
rewards. Even so, we do not hear of Roman ambassadors being overwhelmed by the splendour of Vandal kingship. Quite the reverse: the
representatives of the Roman state maintained a calculated indifference
to the Vandals display of imperial majesty. In the account that survives
to us, it was the personal integrity of the Byzantine ambassador that was
said to have struck Geiseric, not the other way round.
2. barbaria
To the Moorish kings and chieftains living along and across the frontiers of Roman Africa, on the other hand, the Vandals seemed to be
the legitimate successors of the empire. Procopius tells us that Moorish
leaders sought and received the symbols of rule from the new Germanic
monarchs just as they had done earlier from the Roman proconsuls.80
For a time, these Moorish rulers appear to have been clients of the
Vandals. Victor of Vita tells us that both Geiseric and Huneric exiled
Nicene Christians among the Moors, and the Moorish king Capsur sent
Geiseric a report (relatio) on the activities of the exiles, whose executions
the Vandal king subsequently ordered.81 If we are to believe that the
fifth-century Notitia provinciarum et civitatum Africae does in fact represent
a list of the Nicene bishops who came to Carthage in 484 by reason
of the royal command (ex praecepto regali) to debate the faith with their
Arian counterparts, at that date the Vandal kings writ still ran beyond
even well beyond the immediate borders of his kingdom, for the
bishops came from such far-flung provinces as Mauretania Caesariensis,
Mauretania Sitifensis, and Numidia, which at the time would have been
79
80

81

Malchus, frag. 5, p. 410.

Proc. BV 1.25.56, 1:413. On MoorishVandal relations, see Y. Moderan, Les Maures et lAfrique
romaine (IVe VIIe si`ecle), Biblioth`eque des e coles francaises dAth`enes et de Rome 314 (Rome,
2003), esp. pp. 54161.
Geiseric: Vict. Vit. 1.358, pp. 1617. Huneric: Vict. Vit. 2.2637, pp. 338; see also ibid., 2.4,
p. 25. Report and execution: Vict. Vit. 1.37, pp. 1617.

36

Vandal power
under Moorish control (see below, Chapter 5).82 And despite the deterioration in relations between the Vandals and some of their Moorish allies
over the course of the later fifth and early sixth century, even after the
fall of his kingdom the last Vandal king, Gelimer (ad 5303), was able
to seek refuge in the ancient city of Medeus, among the Moors of Mt.
Papua.83
However, we have little evidence to suggest how the Vandals appeared
to the other barbarian kings throughout the late Roman West. We know
that they maintained diplomatic relations of some kind, and in fact these
were probably much more common than the sparse record of them in
our sources might seem to suggest.84 In 458, Vandal ambassadors arrived
in Suevic Galicia, and within a decade legates were once more travelling
between the two kingdoms.85 As mentioned above, Geiseric is said once
to have sent gifts to the Visigothic king Euric, which also implies the
dispatch of African legates.86 Jordanes claims that Geiseric sent gifts to the
Hunnish king Attila as well, though this is not clear from the fragment
of Priscus Jordanes source preserved in the tenth-century Excerpta
de Legationibus.87 In both of these latter two exchanges, Geiseric is said
to have sought the aid of foreign barbarians against his enemies: that of
the Huns against the Visigoths, and later, in the changeable world of late
antique Mediterranean politics, that of the Visigoths against the Romans.
In 533, Gelimer sent his envoys Phuscias and Gotthaeus to Spain to seek
a further alliance with the Visigoths. The embassy, however, arrived only
after Carthage had fallen to the Byzantine army. News of the Vandal
defeat had already reached the Spanish court, and so the ambassadors
were rebuffed by the Visigothic king.88 It was the last act in the troubled
history of VandalVisigothic relations, and in some ways it parallels the
rejection met by the deposed Visigothic king Gisaleic when he fled to
Africa a quarter of a century before. There he sought help from the Vandal
king Thrasamund to regain his realm from the Ostrogothic king who had
82

83
84
85

86
87
88

Notitia provinciarum et civitatum Africae, ed. M. Petschenig, in Victoris episcopi Vitensis historia
persecutionis Africanae provinciae, CSEL 7 (Vienna, 1881), pp. 11534 (quotation, ibid., p. 117);
Warmington, North African Provinces, p. 72; Y. Moderan, La Notitia provinciarum et civitatum
Africae et lhistoire du royaume vandale, Antiquite tardive 14 (2006), pp. 16585.
Proc. BV 2.4.268, 1:436.
On the regularity of diplomatic contact, see Gillett, Envoys, pp. 745.
Hydatius, Chronicon 186 (ad 458), p. 110 and ibid., p. 234 (ad 4667), p. 118; Gillett, Envoys,
pp. 689. Hydatius also mentions an embassy from the Gallo-Roman general Aegidius to Africa
that presumably sought an alliance against the Visigoths or Ricimer (then the power behind
the West Roman throne) or both: Hydatius, Chronicon 220 (ad 4645), p. 116; Gillett, Envoys,
p. 69.
Jordanes, Getica 47.244, p. 100.
Jordanes, Getica 36.184, pp. 789 (= Priscus, frag. 20.2, p. 306) and Priscus, frag. 20.1, pp. 3046.
Proc. BV 1.24.718, 1:41112.

37

Staying Roman
ousted him. Thrasamund appears to have provided Gisaleic with money
but no other support.89 Nevertheless, to the last of the Vandal kings,
Spain must have appeared his best hope for safety. Procopius informs us
that Gelimer had arranged for the Vandal treasury to be shipped to Spain,
where he himself hoped to flee, if the Vandals should be conquered by
the Byzantines. Unfavourable winds prevented the treasure ship from
leaving the port of Hippo Regius (mod. Annaba), and so the wealth of
the Vandals like the Vandal king himself fell into the hands of the
Byzantine army.90
In the later fifth and sixth centuries, however, it was with Italy that
Africa enjoyed its best-documented, if not always its most amicable,
diplomatic exchanges. In 476, for example, Geiseric ceded Sicily to
Odoacer in exchange for the payment of an annual tribute, which itself
probably implied the new master of Italys recognition of the legitimacy
of Vandal claims to Sicily, and thus of Huneric and Eudocias marriage.91
Despite Hunerics own renunciation of his claims to Eudocias inheritance
two years later (above, section 1), his successor Gunthamund may have
sought to revive Vandal overlordship of Sicily. At least in 491, amidst
the disorder of Odoacer and Theodorics struggle for control of Italy, a
Vandal attack on the island was defeated and a peace treaty negotiated.92
Even so, the Sicilian question may well have continued to simmer in the
background of OstrogothicVandal relations through what was left of the
fifth century. It seems still to have been a live issue around the year 500,
when the Vandal king Thrasamund married Theodorics newly widowed
sister, Amalafrida.93 The marriage was probably intended to strengthen
the political ties between the Vandal and Ostrogothic kingdoms, for
89

90
91

92

93

Isidore of Seville, De origine Gothorum 3738, ed. and trans. C. Rodrguez Alonso, in Las Historias
de los godos, vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica y traduccion, Fuentes y estudios
de historia leonesa 13 (Leon, 1975), pp. 2324; Isidore, Historia Gothorum s. aera 545, ed. and
J. Fridh, in
trans. C. Rodrguez Alonso, ibid., pp. 2324; Cassiodorus, Variarum libri xii, ed. A.
Magni Aurelii Cassiodori senatoris opera, Pars 1, CCSL 96 (Turnhout, 1973), 5.434, pp. 2202; see
also Proc. BG 1.12.46, 2:69; J. Moorhead, Theoderic in Italy (Oxford, 1992), p. 190.
Proc. BV 2.4.3441, 1:4378.
Vict. Vit. 1.14, p. 7. F. M. Clover, A Game of Bluff: The Fate of Sicily after ad 476, Historia:
Zeitschrift fur alte Geschichte 48 (1999), pp. 23544 is sceptical that the Vandals ever exercised
effective control over Sicily. On Vandal relations with Odoacer, see also G. Berndt, Konflikt und
Anpassung. Studien zu Migration und Ethnogenese der Vandalen, Historische Studien 489 (Husum,
2007), p. 201.
Cassiodorus, Chronicon s.a. 491, ed. T. Mommsen, MGH AA 11:159; Cass. Var. 1.4.14, p. 16;
and see also Ennodius, Panegyricus dictus clementissimo regi Theoderico 70, ed. F. Vogel, MGH AA
7 (Berlin, 1885), p. 211. Courtois, Vandales, p. 193; Wolfram, Roman Empire, p. 175; Heather,
Goths, p. 231.
Excerpta Valesiana 12.68, ed. J. Moreau and V. Velkov (2nd edn; Leipzig, 1968), p. 20; Cass. Var.
5.43, p. 220; Proc. BV 1.8.11, 1:347; Jordanes, Getica 58.299, p. 123. On the alliance, see Moorhead, Theoderic, pp. 635; and on the question of who initiated negotiations, H.-J. Diesner, Die

38

Vandal power
the bride was accompanied to Africa by 1,000 noble Goths and 5,000
warriors. But her dowry also smacks of a political deal, for Theodoric
gave his sister Lilybaeum (the westernmost of Sicilys three promontories)
as a wedding gift.94
The marriage of Amalafrida to Thrasamund was the most important in
a series of marriage ties that Theodoric established early in his reign.95 Yet
it does not seem to have brought about much in the way of an alignment
of Vandal and Ostrogothic interests. To be sure, the kingdoms seem
to have remained in fairly close contact. The twenty-five diplomatic
missions undertaken by Theodorics envoy Senarius brought him not
just to Constantinople and to Spain but also to Carthage.96 Agnellus,
one of the higher officials of Theodorics court, was in Africa in the
winter of 5056, and then again probably in 5078. These embassies took
place within the context of worsening relations between Ravenna and
Constantinople: by 5078, an East Roman fleet was harrying the Italian
coast and had landed troops in Apulia. It was presumably to seek Vandal
naval assistance that Agnellus travelled to Carthage; Theodorics official
correspondence optimistically claims that by seeking the kingdom of
another, he will be of service to our advantages.97 As we have seen,
though, Thrasamund was also closely allied to the empire, and in this
affair he maintained a steady neutrality. Several years later, in 511, the
same king must have been fully cognizant that receiving and funding the
royal refugee Gisaleic, whom Theodoric had deposed from the Visigothic
throne, was a move calculated to enrage his brother-in-law. Indeed, the
Ostrogothic king wrote to Thrasamund infuriated that the latter had
violated the marriage alliance between them in this way. Theodoric
found it impossible to believe that the Vandal king had consulted with
Amalafrida on this decision, and urged him to look on his wife as a

94

95
96

97

Auswirkungen der Religionspolitik Thrasamunds und Hilderichs auf Ostgoten und Byzantiner, Sitzungsberichte der Sachsischen Akademie der Wissenschaften zu Leipzig, Philologisch-historische
Klasse 113/3 (Berlin, 1967), p. 4.
Proc. BV 1.8.1213, 1:347. Procopius account seems to receive archaeological support from
a boundary marker at Lilybaeum (modern Marsala) which reads The border between the
Vand[als] and the [Go]ths . . . 4 (Fines / inter / Vand[a/los] et / [Go]th[os] / . . . IIII): CIL 10.7232
= Courtois, Vandales, p. 383, no. 138. S. J. B. Barnish, Pigs, Plebeians and Potentes: Romes
Economic Hinterland c.350600 ad, Papers of the British School at Rome 55 (1987), p. 181 and
Heather, Goths, pp. 2312 have read Thrasamunds Africa as a client-kingdom of Theodorics
Italy.
Moorhead, Theoderic, p. 63 and H. Wolfram, History of the Goths, trans. T. J. Dunlap (Berkeley,
Calif., 1988), p. 513 n. 347.
Senarii v.i. comitis patrimonii et patricii epitaphium, quoted by T. Mommsen in his index personarum
to Cassiodorus, Variae, MGH AA 12 (Berlin, 1984), p. 499; and on Senarius career in general,
PLRE 2:9889, s.n. Senarius, and Gillett, Envoys, pp. 190219.
Cass. Var. 1.15.2, p. 25: qui regnum petens alterius nostris est utilitatibus seruiturus; Moorhead,
Theoderic, pp. 1823. For the dates and Agnellus possible offices, see PLRE 2:3536, s.n.

39

Staying Roman
source of counsel.98 Thrasamund evidently wrote back, offering what
were interpreted in Ravenna as his sincere and humble excuses, and
sending along a diplomatic gift. Theodoric accepted the apology but
rejected the gift.99 From that point onwards we hear of no further specific
instances of diplomatic exchange between the two kings courts.
OstrogothicVandal relations reached their nadir in the reign of
Thrasamunds cousin and successor Hilderic. On Thrasamunds death
in 523, Amalafrida fled the Vandal court and sought refuge among the
Moors living on the edge of the desert in Byzacena. There, at Gafsa
(class. Capsa), she was captured by Hilderics agents. The new king had
Amalafrida put into prison, where, some time before 526, she died.100
In Italy it was believed that she had been murdered.101 According to
Procopius, the Gothic entourage which had accompanied her to Africa
was entirely destroyed (dif{eiran) as well.102 The incident provoked a
diplomatic crisis between the Vandal and Ostrogothic kingdoms. Procopius indicates that Theodoric hesitated to invade Africa only because
he did not believe that he could gather a fleet strong enough to carry
out the attack, and because he did not want to provoke a further deterioration of relations with Constantinople, where Hilderics guest-friend,
Justinian while not yet emperor was already the power behind the
throne.103 Yet in 526, Theodoric did order the immediate construction
of a massive fleet, which was to be 1,000 dromons strong. Greeks and
Africans were very much on his mind at the time, and it is hard to
believe that the timing and the urgency were completely coincidental.104
The Ostrogothic kings death that same year forestalled whatever attack
he may have been planning, but the issue nevertheless prompted some
diplomatic exchange between the court of Theodorics successor and the
Vandal king Hilderic.105
We owe our knowledge of OstrogothicVandal relations to the survival
of Cassiodorus Variae, official correspondence which the Italo-Roman
bureaucrat wrote on behalf of the Ostrogothic kings and later revised for
publication. It is the only such documentation that survives as evidence
for the nature of Vandal foreign relations, though of course it primarily records the Ostrogothic not the Vandal side of that exchange.
98
100
101
103
104

105

99 Cass. Var. 5.44, pp. 2212.

Cass. Var. 5.43, p. 220.
Vict. Tonn. s.a. 523.106, p. 34; Proc. BV 1.9.4, 1:351. For the date, see below, next note.
102 Proc. BV 1.9.4, 1:351.
Cass. Var. 9.1, p. 345.
Proc. BV 1.9.5, 1:3512; see also Moorhead, Theoderic, pp. 21617 and 2468.
Cass. Var. 5.1620, pp.1959; Moorhead, Theoderic, pp. 2468. See, however, the sceptical assessment of J. Pryor and E. M. Jeffreys, The Age of the Dromwn: The Byzantine Navy
ca 5001204, The Medieval Mediterranean 62 (Leiden, 2006), pp. 1314.
Cass. Var. 9.1, pp. 3456.

40

Vandal power
From this perspective, OstrogothicVandal relations appear as the natural
continuation of later Roman diplomacy. The interactions between
Ravenna and Carthage took place in the measured cadences of Cassiodorus meticulous late antique Latin, between kings with a studied
affiliation to the classical tradition. Of the two ambassadors whom we
see travelling from the Ostrogothic to the Vandal court, Senarius brought
with him the polish of a career diplomat with considerable experience at
the most important courts of the late antique Mediterranean.106 Perhaps
tellingly, however, his name is ambiguous: it may be either Roman or
Germanic, or may perhaps function in both registers.107 By the end of his
career, Senarius had achieved the exalted status of vir inlustris and patrician. Through Cassiodorus, Theodoric praised his envoys eloquence.108
As the voice of kings (vox regum), beauty and fluency of expression seem
to have been highly prized qualities in an ambassador.109 Agnellus, too,
was repeatedly praised for his eloquent Latin, and by no less a stylist than
Ennodius.110 At the time of his first mission to the Vandals, Agnellus
enjoyed the rank of magnificus vir.111 By 508, he had been elevated to
patrician status, and in the years after his second mission to Africa, the
sometime envoy held one of the highest posts at Theodorics court.112
Neither Agnellus nor Senarius can have appealed much to barbarian
nationalist sentiment, if such there was in the Vandal kingdom. Their
offices, dignities, command of the Latin language virtually everything
we can see about them was Roman. Their selection as ambassadors to
the Vandals, and specifically to the court of the highly cultured Vandal
king Thrasamund, was surely a conscious and deliberate move intended
to stress the integration of the Ostrogothic kingdom into the Roman
world, and perhaps to appeal to the Vandal kings sense of his own
refined sophistication.
As with Vandal ambassadors to the Romans, we are completely ignorant about the envoys sent from the African kingdom to the Ostrogoths.
Phuscias and Gotthaeus Gelimers envoys to the Visigoths are little
more than names to us; the fifth-century envoys to the Sueves and Huns,
not even that. Phuscias bore a name of the sort that only occasionally
percolated its way into the onomasticon of the late Roman elite. When it
106
107

108
109
110
111

Senarii epitaphium, p. 499. On Senarius career, see PLRE 2:9889, s.n. Senarius.

M. Schonfeld, Worterbuch der altgermanischen Personen- und Volkernamen nach der Uberlieferung
des klassischen Altertums bearbeitet (Heidelberg, 1911; repr. 1965), p. 202 and Gillett, Envoys,
pp. 1989.
Cass. Var. 4.3, pp. 1445 and Senarii epitaphium, p. 499.
The quotation is from Senarii epitaphium, p. 499.
Ennodius, Epistulae 7.11, 7.16, and 7.26, ed. Vogel in MGH AA 7:2356, 240, and 258.
112 PLRE 2:356, s.n. Agnellus.
Ennodius, Ep. 4.18, p. 143.

41

Staying Roman
did, it seems to have had particularly Italo-Roman and Romano-African
connotations.113 Gotthaeus is presumably an East Germanic name.114 But
as with the contemporary Ostrogothic kingdom, it is difficult to know
how far to push the name evidence from sixth-century Vandal Africa.
That Gelimer sent two envoys and not just one to the Visigothic court
is, however, probably indicative of the importance of the mission, for the
norm in the late Roman world was to send a single principal together
with a number of subordinates.115 Beyond this, the prosopography of
Vandal diplomacy is of little help.
On the other hand, we may have one of the diplomatic gifts sent
from Gelimers Carthage to Ostrogothic Ravenna in what was a regular feature of late antique foreign relations. Around the central rosette
of a silver plate discovered in Italy, an apparently hasty hand added the
inscription +GEILAMIR REX VANDALORVM ET ALANORVM
(+Gelimer, king of the Vandals and Alans) at some point after the platter was originally produced, presumably converting it into an instrument
of royal Vandal self-presentation. Of course, we have no way of knowing whether the platter made its way to Italy as a royal gift, as plunder
from the defeated Vandal kingdom, or in some other manner.116 But
the idea that the plate may have served as an element of Gelimers official representations to the Ostrogothic kingdom remains an intriguing
possibility.
We are left, then, with little certain evidence with which to evaluate
Vandal foreign relations from an African perspective. The Vandals maintained wide-ranging contacts with the empire, the Ostrogoths, Visigoths, Sueves, Moors, and possibly the Huns and their exchanges seem
113

114
115

116

PLRE 1:376, s.nn. Flavius Fuscenillus, Caeionia Fusciana, Fabia Fuscinilla, and L. Allius
F(uscus?), along with the presumably apocryphal, ibid., 3767, s.nn. Fuscus 12; PLRE 2:489,
s.nn. Fuscianus and Fuscina; and PLRE 3:497, s.n. Fuscus. The female names are attested
primarily in Italy; the male names are particularly concentrated in the late antique west, from
Spain to Italy, though PLRE 2:489, s.n. Fuscianus was apparently Isaurian: see PLRE 2:3067,
s.n. Conon 4. In addition to the women, only PLRE 3:497, s.n. Fuscus (= PCBE 2/1:879,
s.n. Fuscus 1) and PCBE 2/1:87980, s.n. Fuscus 2, were certainly Italian. Africans: PLRE
1:376, Flavius Fuscenillus, vir clarissimus, is attested in a, perhaps, fourth-century inscription
from Byzacena. Also PCBE 1:514, Fuscius, a fourth-century duumvir from a town on the
frontier between Byzacena and Proconsularis, and PCBE 1:515, s.n. Fusculus (= Vict. Vit.
2.45, p. 42), a late fifth-century African bishop.
Schonfeld, Altgermanischen Person- und Volkernamen, pp. 11213.
Greater numbers of principals indicated greater weight: Blockley, East Roman Foreign Policy,
p. 152 with p. 250 n. 6; see also R. Helm, Untersuchungen u ber den auswartigen diplomatischen
Verkehr des romischen Reiches im Zeitalter der Spatantike, Archiv fur Urkundenforschung 12
(19312), p. 402 n. 4.
C. Morrisson, C. Brenot, and J.-N. Barrandon, LArgent chez les Vandales: plats et monnaies,
in F. Baratte (ed.), Argenterie romaine et byzantine: actes de la table ronde, Paris 1113 octobre 1983
(Paris, 1988), pp. 12333.

42

Vandal power
to have taken place within the usual framework of late Roman diplomacy. Emperors and kings announced their accessions to one another
through official envoys, they negotiated treaties, concluded marriage and
military alliances, sought help, sent gifts, and complained of one anothers
conduct. In general, the Vandal kings seem naturally to have been flattered to receive embassies of high status and envoys of great eloquence
or great personal integrity; and though we know almost nothing about
the conduct of Vandal diplomacy, it seems reasonable to suppose that the
Vandals similarly sought to project an image of power and sophistication
that also played on their connections to the Roman state. In Africa, the
Moors certainly seem to have viewed the Vandal kings as the legitimate
heirs of Rome. Even if the same was not true elsewhere, in sixth-century
Mediterranean politics, the close relations of the Vandal kings and the
Roman emperors were a factor to be reckoned with, and the joint
descent of Hilderic and his children from the Hasding royal family and
the house of Theodosius was both recognized and acknowledged in
Carthage, Constantinople, and presumably throughout the sub-Roman
West.
3. the cultural implications of conquest
The ideological significance to the Vandal kings themselves and to
their Romano-African subjects of the marriage of Huneric and Eudocia is most spectacularly illustrated in a poem penned by the sixthcentury Latin poet Luxorius that addresses the couples son, King Hilderic
(ad 52330), in splendidly imperial terms:
Mighty Vandalric, heir of a twin crown, you have adorned your own name
through momentous deeds. Theodosius the avenger conquered warlike battlelines, making the foreign peoples captives with easy effort. Honorius subdued
[his] adversaries with peaceful arms [i.e., diplomacy], whose most powerful
success conquers all the better. With [our] enemies enslaved, the great manliness of Valentinian, well-known to the world, is exhibited in the wile of [his]
grandson.117

With the very opening word of the poem, Luxorius hails Hilderic as king
of the Vandals (Vandalirice). From that point onwards, however, the poet
117

AL 206, p. 154: Vandalirice potens, gemini diadematis heres,/ ornasti proprium per facta
ingentia nomen. / belligeras acies domuit Theodosius ultor, / captivas facili reddens certamine
gentes. / adversos placidis subiecit Honorius armis, / cuius prosperitas melior fortissima vicit. /
ampla Valenti<ni>ani virtus cognita mundo / hostibus addictis ostenditur arte nepotis. This
poem was discussed by F. M. Clover, Time-Keeping, Monarchy and the Heartbeat of Vandal
and Proto-Byzantine Africa, paper delivered to the 26th Byzantine Studies Conference, Harvard
University (4 Nov. 2000).

43

Staying Roman
praises Hilderic, not as the son or grandson of the Vandal kings Huneric
and Geiseric, but as the scion of the imperial house of Theodosius, the
heir of Honorius and Valentinian.118 By the second quarter of the sixth
century, Vandal kingship could be legitimated through an appeal to a
thoroughly Roman lineage.
It should therefore come as no surprise that Vandal rulers cast their
kingship in a very Roman light.119 From the start, the Vandal kings had
established themselves in the palace of the proconsul in Carthage, long
the seat of Roman power in North Africa.120 There the Vandal kings
maintained themselves in imperial splendour. From 455, the palace was
adorned with booty captured from the palace and the temple of Jupiter
Capitolinus in Rome itself: carriages for the royal women, jewellery,
golden drinking cups, silver, and the treasures captured by Titus from
Jerusalem at the end of the Jewish Revolt of ad 6670.121 By the sixth
century, at least, the Vandal king seems to have clothed himself in the
imperial purple and to have presided over his kingdom from a throne.122
Probably from the reign of Huneric, Vandal kings styled themselves
dominus noster rex, our lord king, in what was surely a conscious imitation
of the imperial title.123 From the reign of his successor Gunthamund (ad
48496) the epithet appeared on the silver and copper coinage, and the
only known royal inscription from Africa which reads simply Domn(us)
Geilimer, lord Gelimer publicly proclaims the title as well.124
118

119

120
121
122
123

124

J. H. W. G. Liebeschuetz, Gens into Regnum: The Vandals, in H.-W. Goetz, J. Jarnut, and
W. Pohl (eds.), Regna et Gentes: The Relationship between Late Antique and Early Medieval Peoples
and Kingdoms in the Transformation of the Roman World, Transformation of the Roman World 13
(Leiden, 2003), p. 74.
The Roman cast of Vandal kingship has been extensively discussed in the literature. See esp.
Merrills and Miles, Vandals, pp. 703; Berndt, Konflikt und Anpassung, pp. 2056; M. E. Gil
Egea, Africa en tiempos de los vandalos: continuidad y mutaciones de las estructuras sociopolticas romanas,
Memorias del Seminario de Historia Antigua 7 (Alcala de Henares, 1998), pp. 31418; M.
McCormick, Eternal Victory: Triumphal Rulership in Late Antiquity, Byzantium and the Early
Medieval West (Cambridge, 1986), pp. 2616; F. M. Clover, Carthage and the Vandals, in
J. H. Humphrey (ed.), Excavations at Carthage 1978, Conducted by the University of Michigan,
vol. 7 (Ann Arbor, Mich., 1982), pp. 122; Schmidt, Wandalen, pp. 1567; and Courtois,
Vandales, pp. 2428.
Clover, Carthage and the Vandals, p. 9.
Proc. BV 1.5.35 and 2.9.45, 1:3312 and 1:4567. On the Vandal sack of Rome, see Berndt,
Konflikt und Anpassung, pp. 1945.
AL 371, ll. 1214, p. 287; Proc. BV 2.9.12, 1:457 (purple garments); ibid., 1.20.21, 1:399 (throne).
Title: Courtois, Vandales, p. 243 n. 5; Clover, Timekeeping and Dyarchy, pp. 509; see also
A. Gillett, Was Ethnicity Politicized in the Earliest Medieval Kingdoms?, in A. Gillett (ed.),
On Barbarian Identity: Critical Approaches to Ethnicity in the Early Middle Ages, Studies in the Early
Middle Ages 4 (Turnhout, 2002), pp. 85121, esp. 10910.
Coinage: W. Hahn, Moneta Imperii Byzantini: Rekonstruktion des Prageaufbaues auf
synoptisch-tabellarischer Grundlage, vol. 1, Von Anastasius I. bis Justinianus I. (491565) einschliesslich der ostgotischen und vandalischen Pragungen, Veroffentlichungen der Numismatischen

44

Vandal power
Michael McCormick has demonstrated that the Roman imperial ideology of triumphal rulership continued to flourish under the Vandal
kings.125 In his legislation Huneric pointedly referred to his own triumphal majesty. Vandal coinage continued to employ Victory types, and
formal victory celebrations were apparently also staged in the African
kingdom. The Nicene theologian Fulgentius of Ruspe praised the
Vandal king Thrasamund (ad 496523) with epithets of imperial victory
ideology; Romano-African court poets celebrated the victories of their
barbarian kings in Latin verse; and a dowry from the African hinterland
was dated to The ninth year of the lord most unconquered king (anno
nono domini inuictissimi regis).126
Only three pieces of legislation survive in whole or in part from the
Vandal kingdom, and these demonstrate a clear connection to Roman
law, in both form and content.127 All three are royally issued documents
dating to Hunerics reign; one, indeed, is dated according to the Vandal
kings regnal year.128 Otherwise the protocol, textual structure, language,
and rhythmic prose of the texts show no deviation from contemporary
Roman documents.129 Moreover, Huneric not only issued edicts and
praecepta, legislation which could be enacted by provincial officials; the
Vandal king arrogated to himself the right to proclaim laws (leges), which
in the Roman world remained the strict prerogative of the emperor.130
Like the emperor before him, then, in Vandal Africa the king was to be
seen as the font of all justice.131
Co-opting another imperial privilege, Huneric renamed the coastal
city of Hadrumetum after himself, calling the port Unuricopolis.132
Other Vandal rulers concerned themselves with public works: Hilderic
had an audience hall built in the Carthaginian suburb of Anclae;
Thrasamund ordered the construction of baths, public fortifications

125
126
127
128
129

130
131
132

Akademie der Wissenschaften, Philosophisch-Historische

Kommission 1/Osterreichische
Klasse, Denkschriften 109 (Vienna, 1973), pp. 945 and pl. 42.28, 10, 12 and 1516. Inscription:
CIL 8.10862.
McCormick, Eternal Victory, pp. 2636.
Tablettes Albertini: actes prives de lepoque vandale (fin du Ve si`ecle), ed. C. Courtois, L. Leschi,
C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952), act 1, 1:215.
Full edicts: Vict. Vit. 2.39 and 3.314, pp. 39 and 728; fragment: ibid., 2.34, p. 25.
Vict. Vit. 2.39, p. 39.
P. Classen, Kaiserreskript und Konigsurkunde: Diplomatische Studien zum Problem der Kontinuitat
zwischen Altertum und Mittelalter, Vyzantina Keimena kai Meletai 15 (Thessaloniki, 1977),
p. 109.
Ibid., p. 110, with special reference to Vict. Vit. 3.2, p. 72 and the following law (ibid., 3.314,
pp. 728).
Gil Egea, Africa, pp. 31820.
Notitia, Byz. 107, p. 127; Courtois, Vandales, p. 243 n. 6.

45

Staying Roman
(publica moenia), and a church.133 The Vandal kings probably also redirected towards themselves the (now secularized) cultic veneration that
had traditionally been dedicated to the Roman emperor.134 They certainly preserved other Roman institutions, fashioning them to their own
ends. Carthage continued to be the seat of a proconsul, and the old dignities of illustris, spectabilis, and clarissimus continued to distinguish the
Romano-African elite. The survival of a handful of flamines perpetui and
a sacerdotalis provinciae Africae may suggest that the provincial council
continued to meet in Carthage as late as the sixth century, only now
presumably communicating matters of concern with the Vandal king
rather than the emperor.135 The Vandal kings also maintained the cursus
publicus, by which a number of couriers bore official information from
one end of the kingdom to the other on government horses, supervised
by a public overseer.136
Such actions manoeuvred carefully between emphasizing the continuities that linked late Roman and Vandal power on the one hand, and, on
the other, highlighting the clear discontinuity represented by the exercise
of this power in Africa by a king based at Carthage rather than by an
emperor based at Ravenna or even Constantinople. In other words, even
as they tried to reassure their Romano-African subjects that Romanness
would be safe under the new order, the Vandal kings sought to underscore their own equality with and autonomy from the emperors. One
innovation in particular betrays what we might call the imperial pretensions of the Vandal kings: they seem to have had patriarchal ambitions for
their metropolis. At the council Huneric convened to debate the proper
definition of orthodoxy itself an act of Roman rulership the Nicene

133

134

135
136

AL 194, p. 145 (audience hall); AL 2015, pp. 1503 (baths); AL 204, l. 1, p. 152 (fortifications);
AL 204, ll. 56, p. 152 (church); for this building as a church, see R. Miles, The Anthologia
Latina and the Creation of a Secular Space in Vandal Carthage, Antiquite tardive 13 (2005),
p. 310. The baths were built at Alianae; the location of the fortifications is unclear. Proc. BV
1.21.11, 1:402 claims that by the time of Gelimer the circuit-wall of Carthage had been so
neglected that in many places it had become accessible to whomever wished, and easy to assail
(n gr Karchdnov perbolov otw d phmelhmnov ste sbatv n croiv pollov
t boulomn ka ejodov gegnei).
A. Chastagnol and N. Duval, Les Survivances du culte imperial dans lAfrique du Nord a`
lepoque vandale, in Melanges dhistoire ancienne offerts a` William Seston (Paris, 1974), pp. 87118;
but see F. M. Clover, Emperor Worship in Vandal Africa, in G. Wirth (ed.), RomanitasChristianitas: Untersuchungen zur Geschichte und Literatur der romischen Kaiserzeit (Berlin, 1982),
pp. 66174 and F. M. Clover, Le Culte des empereurs dans lAfrique vandale, BCTH ser. 2,
1516 (1984), pp. 1218, who suggests that the later Vandal kings allowed their Roman subjects
to continue to venerate the emperor.
Clover, Carthage and the Vandals, pp. 1213; however, see Wickham, Framing, p. 637.
Proc. BV 1.16.12, 1:384.

46

Vandal power
bishops of Africa derided the Arian Cyrila for styling himself patriarch.137
The innovation of a patriarchate centred on the African metropolis highlights an important aspect of the Roman cast of Vandal kingship. The
move cannot have been intended to appeal to Romano-Africans of the
Nicene confession, who were, all in all, intensely conservative in their
religious sensibilities and, as we have seen, reacted negatively to what they
regarded as the usurpation of the patriarchal title. Indeed, the only audience who could have supported and endorsed the creation of an Arian
patriarchate at Carthage were Arians themselves.138 And though there
was unquestionably a significant amount of Romano-African conversion
to Arianism in the course of the fifth and sixth centuries (see below, and
Chapter 3), throughout the Vandal period there was still an unavoidably
strong association between Arians and Vandals. Cyrilas assumption of
the patriarchal title was presumably aimed at a Vandal audience.
By the time they captured Carthage in 439 an entire generation of
Vandals had been born and grown to adulthood completely within the
borders of the Roman empire. But even when they had lived outside the empire Vandals had presumably been exposed to and absorbed
Roman cultural influences. The appeal on the part of the Vandal kings
to Roman institutions and the Roman vocabulary of power was not
addressed to Romano-Africans alone. Already by the time the Vandals
seized Carthage, Roman trappings would have provided their kings with
the most eloquent language through which to buttress their social status
and pretensions to authority, even among Vandals themselves.
These trends would only have intensified as Vandal notables became
ever more tightly interwoven into the fabric of African society, and perhaps above all as they came to own property there. After the capture of
Carthage, Geiseric is said to have set aside the best and most numerous
estates of Africa Proconsularis many of which would have been imperial
properties as hereditary tax-free allotments for his army, while less productive land was left in the hands of its original Romano-African owners.
In Byzacena and Numidia, the Vandal king is similarly said to have taken
over the estates of local Romano-African landholders who abounded in
137
138

Vict. Vit. 2.54, p. 45.

On this point, see also R. W. Mathisen, Barbarian Bishops and the Churches in Barbaricis Gentibus during Late Antiquity, Speculum 72 (1997), pp. 6868, with the corrective
offered by M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain,
ad 300750, BAR International Series 1135 (Oxford, 2003), p. 2 in the light of N. Duval,

Recherches archeologiques a` Hadra, 2 vols., Collection de lEcole

francaise de Rome 18/12
(Rome, 197581), 1:878, no. 58 = AE (1968), 204, no. 638 (though note that Vandalorum
seems to be a later addition to this inscription: see Duval, Hadra, 1:88 and 2:119 (La basilique

I dite de Melleus ou de Saint-Cyprien, Collection de lEcole

francaise de Rome 18/2 (Rome,
1981)).

47

Staying Roman
wealth (plot kmzwn), and to have established these properties as
part of the royal domain.139 What such arrangements meant in practical
terms is still debated, and it does not help that Vandal settlement is difficult to see archaeologically.140 However, we are on rather firmer ground
with two observations. First, that, while some wealthy private houses
were certainly abandoned in the Vandal period, it is also clear that others
were maintained and even refurbished.141 And second, that, whatever the
legal framework within which Vandal settlement was accommodated, the
contemporary textual sources would lead us to believe that at least some
of these sumptuously decorated villas and townhouses quickly came to
be occupied by the Vandal kings most prominent followers. Thus, for
example, the metropolitan residence of a certain Gordian, a Carthaginian
senator, was given to some Arian priests after the aristocrat fled with
his family to Italy. The senators estate in Byzacena, by contrast, seems
to have been taken over by the Vandal king, for it proved recoverable
when Gordians son returned to Africa to regain what he could of the
family property.142 Victor of Vita tells of a Vandal millenarius (leader of
1,000 troops) who seems to have lived, together with his wife, children,
household, and animals, on an estate near Tabarka (class. Thabraca), on
the Mediterranean coast of Proconsularis; and a sixth-century nobleman
named Fridamal similarly owned a seaside villa that was surrounded by
gardens where sea birds nested. These gardens were watered by fountains
and adorned with a statue of Diana, and also had a tower where Fridamal
had himself depicted killing a boar.143
139

140

141
143

Proc. BV 1.5.1115, 1:3334; Vict. Vit. 1.13, p. 7; see also Val. Nov. 34.23 (ad 451),
p. 141, which specifically mentions despoliations in Africa Proconsularis and Byzacena. On
the geography, see J. Desanges, Un Temoignage peu connu de Procope sur la Numidie vandale
et byzantine Byzantion 33 (1963), pp. 4956. On imperial estates in Africa Proconsularis, see
M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Spatantike, Frankfurter althistorische Studien 7 (Kallmunz, 1973), pp. 401 and D. P. Kehoe, The Economics of Agriculture
on Roman Imperial Estates in North Africa, Hyptomnemata: Untersuchungen zur Antike und zu
ihrem Nachleben 89 (Gottingen, 1988).

On Vandal settlement, see esp. Moderan, LEtablissement

territorial, passim (who argues in
favour of the idea of a territorial settlement); and J. Durliat, Le Salaire de la paix sociale dans
les royaumes barbares (Ve VIe si`ecles), in H. Wolfram and A. Schwarcz (eds.), Anerkennung und
Integration. Zu den wirtschaftlichen Grundlagen der Volkerwanderungszeit 400600. Berichte des Symposions der Kommission fur Fruhmittelalterforschung, 7. bis 9. Mai 1986, Stift Zwettl, Niederosterreich,
Veroffentlichungen der Kommission fur Fruhmittelalterforschung 11 (Vienna, 1988), pp. 2172;
J. Durliat, Les Transferts fonciers apr`es la reconquete byzantine en Afrique et en Italie, in E.
Magnou-Nortier (ed.), Aux sources de la gestion publique, 2 vols. (Lille, 19935), 2:89121 ; and
A. Schwarcz, The Settlement of the Vandals in North Africa, in Merrills, Vandals, Romans, and
Berbers, pp. 4957 (who reject it). Vandal settlement is not discussed by Goffart, Barbarians and
Romans.
142 V. Fulg. 1, p. 11.
Leone, Changing Townscapes, pp. 1458.
Millenarius: Vict. Vit. 1.305, pp. 1315. Fridamal: AL 299300, pp. 2467.

48

Vandal power
Though nobles like Fridamal and Victors millenarius owned rural or
suburban estates, Vandals seem for the most part to have taken up residence in Africas cities, and above all in Carthage.144 The metropolis and
its suburbs were the centre of Vandal court life, as Victor of Vita and the
poems of the Latin Anthology clearly attest. The fifth-century moralist
Salvian of Marseilles, too, thought of the Vandals activities as primarily urban, and the non-narrative evidence, such as it is, may well bear
him out.145 There are, however, problems with the data. The majority of the inscriptions from North Africa commemorating individuals
with Germanic names are not dated. While most probably belong to
the Vandal period, there is still reason for caution. The Vandals were not
the only peoples with Germanic names in Africa even in the fifth and
sixth centuries: at the time of the Vandal invasion, for example, Goths
were among the troops stationed in Hippo.146 If, however, we accept
that in aggregate these inscriptions probably do broadly reflect Vandal patterns of settlement, the largest single concentration of Vandals
nearly half of the attested cases would seem to have been in and around
the royal capital. The remainder were scattered around the cities of the
African hinterland, particularly in northern Byzacena. Recent scholarship has tended to be sceptical of the extent to which taste in and
use of so-called barbarian-style jewellery as grave goods in late antique
North Africa is likely to be indicative of Vandal identity. While clearly
expressing wealth and local social status and thus perhaps in that sense
Vandal in a fifth- and sixth-century African context in the absence
of supporting evidence we cannot safely use such artefacts as a guide to
their owners ethnic self-identification. Nevertheless, finds of these goods
show a similar distribution to the inscriptions, though in this case

144

145

146

territorial,
For the localization of Vandal settlement in general, see Moderan, LEtablissement
p. 89, but also Schwarcz, Settlement of the Vandals, p. 57, who rightly rejects evidence of
Arian persecution as evidence of Vandal settlement.
Salvian of Marseilles, De gubernatione Dei 7.21.8991, ed. G. Lagarrigue, in Du Gouvernement de
Dieu, vol. 2 of uvres, SC 220 (Paris, 1975), pp. 4946; Schwarcz, Settlement of the Vandals,
p. 57.
Goths in Africa: Possidius of Calama, Vita Augustini 28.12, ed. M. Pellegrino, in Vita di S.
Agostino, Verba seniorum 4 (Alba, 1955), p. 154; see also Augustine, Epistulae, ed. A. Goldbacher,
CSEL 34, 44, 578 (Vienna, 18951923), here Ep. 185.1.1, CSEL 57:2; Olympiodorus, frag. 40,
ed. Blockley, in Fragmentary Classicising Historians, 2:2024; and pseudo-Augustine, Ep. 4, PL
33:1095. Clover, Geiseric, pp. 2830 argues that the Goths in Bonifaces army were foederati of
the treacherous general Sanoex (?), himself possibly a Goth, sent to Africa to oppose Boniface,
but who betrayed the imperial expedition to him. Possidius, Vita Augustini 28.4, pp. 14850
also indicates that the invading Vandal confederation included Goths. On the difficulty of using
names as an indicator of ethnicity in a multilingual environment, see, e.g., R. Bagnall, Egypt in
Late Antiquity (Princeton, NJ, 1993), pp. 2323.

49

Staying Roman

Map 1.3.

Vandals in North Africa? archaeological evidence

particularly concentrated at Hippo Regius, even more than at Carthage

(see Map 1.3).147
The process of settling down and acquiring property gave the new
Vandal elite a stake in local society. To be sure, raiding and warfare in
the Mediterranean continued to supply Vandal warriors with a source
147

The grave goods are discussed by G. Koenig, Wandalische Grabfunde des 5. und 6. Jhs., Madrider
Mitteilungen 22 (1981), pp. 299360; J. Kleemann, Quelques reflexions sur linterpretation
ethnique des sepultures habillees considerees comme vandales, Antiquite tardive 10 (2002),
pp. 1239; P. von Rummel, Habitus Vandalorum? Zur Frage nach einer gruppen-spezifischen
Kleidung der Vandalen in Nordafrika, Antiquite tardive 10 (2002), pp. 1347; Christopher Eger,
Vandalisches Trachtzubehor? Zu Herkunft, Verbreitung und Kontext ausgewahlter Fibeltypen
in Nordafrika, in G. Berndt and R. Steinacher (eds.), Das Reich der Vandalen und seine (Vor-)

Geschichten, Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse,
Denkschriften 366/Forschungen zur Geschichte des Mittelalters 13 (Vienna, 2008), pp. 18396;
and S. Brather, Kleidung, Grab und Identitat in Spatantike und Fruhmittelalter, in Berndt
and Steinacher, Das Reich der Vandalen, pp. 28394. See also, in general, P. Amory, People and
Identity in Ostrogothic Italy, 489554, Cambridge Studies in Medieval Life and Thought, 4th ser.
33 (Cambridge, 1997), pp. 334 and 3347.

50

Vandal power
of income in the form of slaves, at least some of whom were sold on
local markets to the profit of their captors.148 But from the mid fifth
century onwards the prosperity of the Vandal ruling class was probably
increasingly bound up with that of the African countryside. Our picture
of the African economy in late antiquity is still fragmentary and it is
difficult to generalize, but early returns suggest that at least parts of
Africa Proconsularis (where Vandal settlement seems for the most part
to have concentrated) may have enjoyed a fair degree of affluence under
the Vandals.149 In the fifth century, the prosperity of rural sites seems
to have increased dramatically in the immediate hinterland of Carthage,
and this trend continued into the sixth century.150 The Segermes Valley,
slightly to the south, saw a similar expansion in the number and density
of rural sites in the first half of the sixth century (including the early
Byzantine period), after having held more or less steady for the previous
hundred fifty years.151 At the same time, pottery workshops in the Vandal
kingdom saw a renewal in the large-scale production of African red slip
ware for export, and numerous urban kilns produced fine wares for
local consumption.152 Amphora production also continued, and from
the Vandal period onwards a considerable amount of olive oil appears
to have been bottled on the farms where it was produced before being
148

149

150

151

152

Vict. Vit. 1.25, p. 12 and Malchus, frag. 5, p. 410; see also Theodoret of Cyrrhus, Ep. 70, ed.
Y. Azema, in Correspondance, 4 vols., SC 40, 98, 111, and 429 (Paris, 196498), 2:1524 and
Vict. Vit. 1.12 and 1.14, pp. 78.
On the African rural economy in this period in general, see A. Leone and D. J. Mattingly,
Vandal, Byzantine and Arab Rural Landscapes in North Africa, in N. Christie (ed.), Landscapes of Change: Rural Evolutions in Late Antiquity and the Early Middle Ages (Aldershot, 2004),
pp. 13562 and Leone, Changing Townscapes, pp. 12734. Note the apparent contraction of rural
settlement in the inland regions of Dougga and Cillium-Thelepte: M. de Vos, Rus Africum:
Terra, acqua, olio nellAfrica settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell
tunisino) (Trent, 2000), pp. 201, 725, and 812 and R. B. Hitchner, The Kasserine Archeological Survey 19821986 (University of Virginia, USA Institut national darcheologie et dart
de Tunisie), Antiquites africaines 24 (1988), pp. 741, respectively.
J. A. Greene, Une Reconnaissance archeologique dans larri`ere-pays de la Carthage antique,
in A. Ennabli (ed.), Pour sauver Carthage: exploration et conservation de la cite punique, romaine et
byzantine (Tunis, 1992), pp. 1957.
S. Dietz, A Summary of the Field Project, in S. Dietz, L. Ladjimi Seba, H. Ben Hassen,
P. rsted, and J. Carlsen (eds.), Africa Proconsularis: Regional Studies of the Segermes Valley of
Northern Tunisia, 3 vols. (Aarhus, 19952000), 2:7812.

Leone, Changing Townscapes, pp. 1289; M. Bonifay, Etudes

sur la ceramique romaine tardive
dAfrique, BAR International Series 1301 (Oxford, 2004), pp. 49 and 537; D. Barraud,
M. Bonifay, F. Dridi, and J. F. Pichonneau, LIndustrie ceramique de lantiquite tardive, in
H. Ben Hassen and L. Maurin (eds.), Oudhna (Uthina): La Redecouverte dune ville antique de
Tunisie (Bordeaux, 1998), pp. 13967; and M. Mackensen, Die spatantiken Sigillata- und Lampentopfereien von El Mahrine (Nordtunesien): Studien zur Nordafrikanischen Feinkeramik des 4. bis 7.
Jahrhunderts, Munchner Beitrage zur Vor- und Fruhgeschichte 50, 2 vols. (Munich, 1993).

51

Staying Roman
exported abroad.153 As was the case elsewhere in the Mediterranean,
access to a share of this wealth seems to have brought with it a certain
alignment in the interests of the new Vandal elite and what remained of
the old Romano-African aristocracy.154
In Africa, Vandal notables also quickly assimilated to the lifestyle of
the late Roman gentry.155 Indeed, in general, in the generations after
the conquest, it becomes more and more difficult for us to distinguish
between Vandals and Romano-Africans at the level of the secular aristocracy. A wealthy Carthaginian woman from the district of Koudiat-Zateur
was buried, decked in so-called barbarian-style gold, in a marble sarcophagus decorated with Season reliefs.156 A Roman matron with a taste for
barbarian jewellery? Or a barbarian noblewoman with a flair for Roman
funerary culture? More probably the barbaric qualities of the jewellery
are a red herring.157 The distinction is not much easier to make from the
written record. Here we are inevitably at the mercy of our sources; and
by and large the means they used to discriminate between social groups
have proved too ephemeral to be preserved in the historical record.
This is not to deny the existence of individuals or groups who were
clearly Roman or clearly Vandal. The aristocrats Dracontius, Fulgentius
of Ruspe and his younger brother Claudius would fall into the former
category; Victors Vandal millenarius, the praepositus regni Heldica, his wife,
and his brother Gamuth would fall into the latter.158 But we speak with
much less certainty when we say that Fridamal, for example, was a Vandal
rather than a Romano-African whose parents had given him (or who
had himself adopted) a Germanic name. This would seem to be the case
with, for example, the sixth-century aristocrat Becca, whom Luxorius
153

154
155

156
157

158

D. P. S. Peacock, F. Bejaoui, and N. Ben Lazreg, Roman Amphora Production in the Sahel
Region of Tunisia, in Amphores romaines et histoire economique: dix ans de recherche, Collection

de lEcole
francaise de Rome 114 (Rome, 1989), p. 200; C. Panella, Merci e scambi nel
Mediterraneo tardoantico, in A. Momigliano and A. Schiavone (eds.), Storia di Roma, 4 vols.
(Turin, 1988), 3/2, pp. 6413; Leone, Changing Townscapes, p. 130.
See Heather, Goths, p. 305 for similar processes in Visigothic Spain and Ostrogothic Italy.
Romanization among the Vandals in Africa has been discussed, esp. by Clover, Carthage
and the Vandals, passim and P. Riche, Education and Culture in the Barbarian West, from the
Sixth through the Eighth Century, trans. J. J. Contreni (Columbia, SC, 1978), pp. 379 and 64.
K. Vossing, Schule und Bildung in Nordafrika der romischen Kaiserzeit (Brussels, 1997), pp. 62431
takes a darker view of the relations between Vandals and Romano-Africans.
Koenig, Wandalische Grabfunde, pp. 3089; Kleemann, Quelques reflexions, p. 126; Eger,
Vandalisches Trachtzubehor?, p. 189.
Thus Rummel, Habitus Vandalorum, pp. 1367; see also Rummel, Habitus barbarus: Kleidung
und Reprasentation spatantiker Eliten im 4. und 5. Jahrhundert, Erganzungsbande zum Reallexikon
der Germanischen Altertumskunde 55 (Berlin, 2007), pp. 270323 and Kleemann, Quelques
reflexions, pp. 1245. See, further, Amory, People and Identity, pp. 3434.
Claudius: V. Fulg. 5, pp. 2931. Heldica and his family: Vict. Vit. 2.15, p. 29. Also two Vandal
brothers: Vict. Vit. 3.38, p. 91.

52

Vandal power
accused of sodomy. The name Becca appears to be Germanic, yet the
poet refers to the wealth and great banquets and many gifts that Beccas
grandfather and great-grandfathers and great-great-great-grandfathers
had left him.159 This seems to evoke Roman ancestry, or perhaps a mixed
marriage, rather than strictly barbarian descent. Similarly, a boy who died
at the tragically young age of two and a half and who was buried in Aquae
Caesaris (mod. Youks, Algeria) at some point in the fifth or sixth century
bore the mixed Latin-Germanic name Flavius Vitalis Vitarit.160 At some
point after the Byzantine reconquest, the thirty-nine-year-old Gregoria daughter of Theoderic was buried at Thysdrus (mod. el-Djem,
Tunisia).161 We do not know whether either Vitarit or Gregoria was
born and named under the Vandal regime, but that is somewhat beside
the point; for in both of these two families Roman and Germanic names
could exist comfortably side by side.
Insofar as we can tell from our sources, Romano-African aristocrats
and Vandal nobles shared much the same tastes, the same interests, the
same attitudes. Thus, for example, the Vandal and Roman elite seem
to have shared a devotion to hunting. Procopius tells us that it was a
favourite Vandal pastime, and mosaics depicting hunting scenes (a popular motif in the artwork of Roman-era North Africa) apparently continued to be commissioned by the regional elite through the fifth and sixth
centuries.162 Luxorius dedicated a number of poems to hunting themes,
including the painting of Fridamal killing a boar, and the archaeological,
epigraphic, and textual evidence further suggests that beast hunts continued to be staged in the amphitheatres both of the metropolis and of the
African hinterland into the sixth century.163 Baths too were a common
interest. Like Thrasamund, the sixth-century Vandal prince Gebamund
oversaw the construction of a suburban bath complex near Tunis, and
commemorated the accomplishment in a now fragmentary dedicatory
159

160
161
162
163

AL 316, p. 256: Divitias grandesque epulas et munera multa, / quod proavi atque atavi quodque
reliquit avus. On the name Becca: E. Forstemann, Altdeutsches Namenbuch, vol. 1, Personennamen (Munich, 1966), pp. 3001, s.n. Bic.
AE (1974), 198 no. 705. On the name Vitarit, see N. Francovich Onesti, I Vandali: Lingua e
storia, Lingue e letterature Carocci 14 (Rome, 2002), pp. 1789.
ILCV 1349 A.
Proc. BV 2.6.7, 1:444; K. Dunbabin, The Mosaics of Roman North Africa: Studies in Iconography
and Patronage (Oxford, 1978), pp. 4664, esp. 59 and pls. XVIII.40XIX.43.
Hunting themes: AL 299, 302, 32930, and 355, pp. 2467, 2489, 264, and 278; see also AL
287 and 295, pp. 239 and 243. Beast hunts and amphitheatres: AL 341, 3489, and 368, pp. 270,
2735, and 285. Archaeology: Leone, Changing Townscapes, p. 140. Epigraphy: N. Duval and J.
Mallon, Les Inscriptions de la chapelle vandale a` Hadra dapr`es labbe Delapard, Bulletin de
la Societe nationale des Antiquaires de France (1969), pp. 11824 = AE (1973), 198, no. 622: Fecit
Va/ricos ludos; the name seems to be the Punic Baric, but see Francovich Onesti, I Vandali,
pp. 1778, s.n. Varica.

53

Staying Roman
inscription.164 The enterprise may also have been celebrated by Luxorius in one of his epigrams; at least it is tempting to see the first word
of Gebamunds inscription, cerne (behold!), as a play on the place name
Cirne, where Luxorius tells us another bath complex was erected at
much the same time.165 Some baths that were already falling into disrepair at the beginning of the fifth century seem to have been abandoned
in the Vandal period, but many others both public and (increasingly)
private were restored and refurbished.166 Gardens, too, are a recurring theme in descriptions of the lifestyle of the North African elite in
the Vandal period. Procopius provides a glowing account of the springs
and fruit-trees of the gardens on an estate at Grasse, some distance from
Carthage.167 In two separate poems, Luxorius too praised the gardens of
one Eugetius and of the Vandal prince Hoageis, the latter of which was
devoted to medicinal herbs.168 Little fish were kept in the pools of the
palace gardens.169 Luxorius wrote an epigram celebrating the Egyptian
lily which one aristocratic family cultivated inside their house (where it
was said to grow better than in the garden) and another in praise of a
hundred-leaved rose.170 The rose was a recurrent theme in North African
mosaic art, and a cycle of three poems by an anonymous author contained in the Latin Anthology also celebrate the flower, as, indeed, does
the pseudo-Dracontian On the Origins of Roses.171
In addition to the poems written at the Vandal court from the reign of
Huneric onwards (see below, Chapter 3), a Carthaginian noble with the
East Germanic name Fridus commissioned Luxorius to write a cento a
patchwork poem stitched together from the verses of other poets work,
in this case Virgils Aeneid and Georgics to celebrate his marriage.172 The
164
165
166

167
168

169
170
171

172

CIL 8.25362; on Thrasamunds baths, see above n. 133. On Gebamund, see Proc. BV 1.18.1
and 1.18.1219, 1:388 and 1:3901 and Vict. Tonn. s.a. 534.38, p. 38.
AL 345 (De aquis calidis Cirnensibus), pp. 2712.
Y. Thebert, Thermes romains dAfrique du Nord et leur contexte mediterraneen: etudes dhistoire
et darcheologie, Biblioth`eque des e coles francaises dAth`enes et de Rome 315 (Rome, 2003),
pp. 41821 and 4823; Leone, Changing Townscapes, pp. 1401 and 1579; and Miles, Anthologia
Latina, pp. 31012.
Proc. BV 1.17.10, 1:3867.
AL 327 and 364, pp. 2623 and 283, respectively. The name Eugetius is Greek egh{v, joyous
or cheerful, not a corruption of Hoageis; on whom, see Courtois, Vandales, p. 399, no. 24,
Hoageis, and ibid., p. 390, Tableau genealogique des Hasdings.
AL 286, p. 2389.
AL 367 and 361, pp. 284 and 281, respectively.
AL 724, pp. 768; pseudo-Dracontius, De origine rosarum, ed. E. Baehrens and F. Vollmer, in
Poetae latini minores, 6 vols. (Leipzig, 18811911), 5:237. See also the poem by Florus, AL 75,
p. 78.
Anthologia Latina 18, ed. A. Riese, F. Buecheler, and E. Lommatzsch, 2 vols. (Leipzig, 1894
1926), 1/1:7982. This poem is not included in D. R. Shackleton Baileys edition of the
Anthologia Latina, vol. 1/1 (Stuttgart, 1982).

54

Vandal power
poem is doubly interesting, both for its form and sources and because
the work is evocative of the Roman custom of reciting a marriage poem
at the wedding.173 Luxorius satirized an aristocrat named Blumarit in
another epigram, which Pierre Riche takes as an indication of Vandal
sensitivity to Latin verse.174 In the fifth century, the son of the Vandal prince Theoderic received an education in the liberal arts, and the
Carthaginian grammaticus Felicianus encouraged Vandals and Romans
to mingle in his auditorium.175 One of Luxorius poems celebrates a
mime, a dwarf named Macedonia who, he says, always portrayed the
roles of Andromache and Helen in dance.176 During his sufferings on
Mt. Papua, one of the Vandal king Gelimers three requests was for a
cithara, to the accompaniment of which he could sing his miseries.177
The Vandal prince Hoamer was called the Achilles of the Vandals, and
in a curious exchange of pleasantries a certain Parthemius presbyter saw
fit to praise one Sigisteus comes in similar terms with the verse mighty
Larissa did not beget such an Achilles.178 In the Vandal kingdom, high
culture was Roman culture.
But Roman culture meant different things in different places, and in
Africa the Vandals were exposed to Romanness of a distinctively African
stamp. Parthemius praise of Sigisteus displays a fierce pride in Africa,
specifically with respect to the Greek East. Greece had not produced
such learning, nor such an Achilles, but valiant and fertile Africa bore
such a fruit to us a man who, shining, of course, with bright light,
was then raised from our gentle breast to the stars.179 This local pride
was deeply rooted and would survive the Byzantine conquest, when
Fulgentius the Mythographer himself probably a product of the late
Vandal literary milieu would recount how Calliope, the epic muse, had
wandered from Athens to Rome to Alexandria and thence to Fulgentius

173
174
175
176

177
178
179

On which see OCD, p. 928, s.v. marriage ceremonies, Roman, and ibid., p. 548, s.v.
epithalamium.
AL 321, p. 259; Riche, Education and Culture, p. 64.
Vict. Vit. 2.13, p. 28 (magnis litteris institutus); Blossius Aemilius Dracontius, Romulea 1.14,
ed. J. Bouquet and E. Wolff, uvres, 4 vols. (Paris, 198596), 3:134.
AL 305, pp. 2501. Luxorius also wrote about a female cithara-player named Gattula, whose
body disgusted the poet: AL 3567, pp. 27980; Victor of Vita refers to Geiserics chief pantomime (archimimum) Masculas: Vict. Vit. 1.47, p. 20. Clover, Carthage and the Vandals,
p. 10.
Proc. BV 2.6.30 and 2.6.33, 1:447.
Parthemii presbyteri rescriptum ad Sigisteum, PLS 3:448: Nequae larissa potens similem procreavit
achillem. Hoamer: Proc. BV 1.9.2, 1:351.
Parthemii rescriptum, PLS 3:448: nostris qualem armipotens tam fertilis africa frugum vexit ad
astra virum quem claro lumine fulgens scilicet tunc placido nostro de pectore tolli.

55

Staying Roman
own rural estate somewhere in Africa.180 As far as the Mythographer was
concerned, Africa had become the reigning heir of the classical literary
tradition. But the greatest paean to the region to emerge from the Vandal
kingdom itself was Florentius In laudem regis, addressed to Thrasamund,
who alone gathered to himself all the best things that the world produced.
More than that, his capital, Carthage, was the pre-eminent city in Libya,
populous, strongly defended, distinguished in learning and adorned with
buildings and fortifications, and sweet in its charms.181 Significantly,
Florentius also praises Carthage as the mother of the Hasdings (Carthago
Asdingis genetrix).182 The Vandals had come to identify themselves with
Africa; so much so that in the sixth century, when the conquering
Byzantine army tried to deport defeated Vandal warriors to the Persian
frontier, perhaps four hundred of the Vandals seized control of the boat
transporting them to the East, forced the sailors to turn it around, and
sailed back to Africa. There they made their home among the Moors of
the Aur`es Mountains and in Mauretania183 rather than among, say, the
Visigoths in Spain. It is surely this association on the part of the Vandals
with the Romano-Africans pride in their province and its metropolis
that explains the revival of images of Felix Karthago and figures from
the citys foundation legend on the coinage from Vandal Africa as well as
other Vandal-era celebrations of the metropolis.184
It is even conceivable that there may be specifically African overtones
to Procopius description of the Vandals as the most luxurious of the
barbarian peoples:
Indeed, from the time when they captured Africa, all of them made use of the
baths every day, and [they enjoyed] a table abounding in all things, the sweetest
and the best that earth and sea yield up. And they wore gold as much as possible,
and, wrapped in Persian clothes, which they now call serike [i.e., silk], they spent
their time in theatres and in hippodromes and in the enjoyment of other good
things, and most of all in hunting. And they had dancers and actors and many
things to hear and watch which happen to be musical and especially worth
seeing among people. And most of them lived in gardens, which abounded in
180

181
183

Fabius Planciades Fulgentius Mythographer, Mitologiarum libri tres 1.praef., ed. R. Helm, in
Opera (Leipzig, 1898; repr. 1970), pp. 89. Fulgentius wrote after c.550, but probably not long
after: G. Hays, The Date and Identity of the Mythographer Fulgentius, Journal of Medieval Latin
13 (2003), pp. 2414. His writings share a common literary culture with those of the late Vandal
period, and there is no reason to suppose that his career could not have spanned the Byzantine
reconquest: G. Hays, Romuleis Libicisque Litteris: Fulgentius and the Vandal Renaissance ,
in Merrills, Vandals, Romans, and Berbers, pp. 10132.
182 AL 371, l. 30, p. 288.
AL 371, pp. 2868, esp. ll. 918 and 2836.
184 Discussed by Clover, Felix Karthago, pp. 110.
Proc. BV 2.14.1819, 1:4845.

56

Vandal power
water and trees; and they held as many drinking-parties as possible, and all kinds
of sexual activities were widely practised by them.185

The parallels between this brief caricature and the image of the late
Vandal-era aristocratic lifestyle that emerges from the contemporary
African sources are striking, and serve to reinforce the emerging consensus as to the reliability of much of the factual information that Procopius
reports, if not his interpretations.186 Though they differ completely over
the nature of the barbarians themselves, the Byzantine historians description of Vandal decadence is also remarkably similar to the picture painted
by Salvian of Marseilles of the moral dissipation of the Roman inhabitants
of Carthage on the eve of the Vandal conquest. The barbarians had, wrote
Salvian, descended upon a province over-ripe with all kinds of evils and
corruption, most especially with sexual immorality, and they had closed
the brothels, forced the prostitutes to marry, and put an end to the adulterous and homoerotic pleasures in which the Carthaginians in particular
were said to have indulged. Even more remarkable, the Vandals themselves had remained untainted by the morally putrefying atmosphere of
this most decadent of Roman provinces.187 Alas, they were not to remain
so for long. The East Roman historian Malchus of Philadelphia tells us
that after the death of Geiseric and the accession of Huneric in 477 the
Vandals lapsed into every kind of weakness (v psan malakan), in
which state they continued to languish until Procopius day.188
For Procopius, however, the decadence of the Vandals may simply have
been a product of their urban lifestyle. The historian describes Antiochenes in similar terms as frivolous and ridiculous, caring for nothing other
than festivals, luxuriousness, and the theatre.189 Salvian was certainly
hostile to the pleasures of the late antique urban lifestyle, which he saw
as depraved, dissolute, and morally corrupting. Amphitheatres, odeons,
games, processions, athletes, tumblers, and pantomimes all received
Salvians censure, but the bishop reserved his strongest condemnation
185
186

187

188

Proc. BV 2.6.59, 1:4434.

On Procopius as an ethnographer, see the fuller discussion below, Chapter 5.1. On the ongoing
reassessment of Procopius Quellenwert, see the collected articles in Antiquite tardive 8 (2000), a

special issue dedicated to the De aedificiis, esp. D. Feissel, Les Edifices

de Justinien au temoignage
de Procope et de lepigraphie, Antiquite tardive 8 (2000), pp. 81104; and A. Kaldellis, Procopius
of Caesarea: Tyranny, History, and Philosophy at the End of Antiquity (Philadelphia, Pa., 2004).
Salvian, De gubernatione Dei 6.12.6671, 7.13.547.22.100, and 8.2.98.5.25, pp. 4048, 468
502, and 51727, esp. ibid., 7.13.56 and 7.22.94100, pp. 470 and 498502. On these points, see
D. Lambert, The Barbarians in Salvians De Gubernatione Dei, in S. Mitchell and G. Greatrex
(eds.), Ethnicity and Culture in Late Antiquity (London, 2000), pp. 10315.
189 Proc. BP 1.17.37 and 2.8.6, 1:88 and 1:1845.
Malchus, frag. 17, p. 424.

57

Staying Roman
for the circuses and theatres.190 The vice and impurity which these
entertainments visited upon urban populations left an indelible stain on
their moral character. Such an infection could be annihilated only when
the cities themselves had similarly perished or, for whatever other reason, lacked the means to continue producing fodder for their debauched
diversions.191 Immorality was a crisis of the cities.
But for both Procopius and Salvian, decadence was also an explanatory
principle in human history. In their thought world, strong and warlike
peoples easily conquered the effete, the luxury-loving, the degenerate.
Salvian unusually for a late antique intellectual directed his diatribe
against his own society. The Roman world, to his mind, was hopelessly
corrupt: Vice and impurity are, as it were, one of the ties that unite
the Roman people in common descent and, as it were, [their] mind and
nature, since wherever there are Romans, there too especially is vice.192
It mattered to Salvians apologetic endeavour that the barbarians remain
outside that society, morally pure, untainted by the depravity of Roman
civilization. Procopius and Malchus, on the other hand, drew on one
of the most venerable traditions of ancient ethnography in their own
depictions of the Vandals: that of the luxurious and debased barbarian.193
To Procopius, at least, barbarians were inherently warlike. They could
threaten the empire or defend it, they could fight among themselves,
or they could simply strike mindlessly at anyone or anything weaker
than themselves; but they were, at heart, warriors. A barbarian in pursuit of high culture was a barbarian gone soft. Both for Procopius and
for Malchus, barbarian decadence provided the only intellectual framework through which to understand the fact that the Vandals as a people
had so quickly accommodated themselves to the leisured lifestyle of the
Romano-African aristocracy.
4. the limits of romanness
Even so, to late antique observers, Vandals remained recognizably Vandal.
In the wake of the Byzantine reconquest of North Africa, two waves of
190
191
192

193

Salvian, De gubernatione Dei 6.3.15, p. 370. Salvians denunciation of the circuses and theatres
provides much of the matter for book 6 of his De gubernatione Dei.
Salvian, De gubernatione Dei 6.8.3945, pp. 38892.
Ibid., 6.8.40, p. 388: uitiositas et impuritas quasi germanitas quaedam est hominum Romanorum
et quasi mens atque natura, quia ibi praecipue uitia ubicumque Romani. On these points, see
D. Lambert, The Uses of Decay: History in Salvians De Gubernatione Dei, Augustinian Studies
30 (1999), pp. 11530.
Y. A. Dauge, Le Barbare: recherches sur la conception romaine de la barbarie et de la civilisation,
Collection Latomus 176 (Brussels, 1981), pp. 4334.

58

Vandal power
deportations were said to have cleansed the region of its Vandal population. Precisely whose displacement this entailed may be obscure to us,
but it seems to have been reasonably clear to contemporary witnesses like
Procopius.194 Dracontius, too, seems to have known a barbarian when
he saw one.195 And, according to Victor of Vita, the great persecution
of the Nicene Christians under Huneric began when the Vandal king
stationed his torturers at the doors of Nicene churches with orders to
scalp anyone trying to enter who looked like a Vandal (in specie suae
gentis). Thus mutilated, the women curiously, in Victors account, only
the women, although he is quite clear that men, too, were subjected
to scalping were paraded behind heralds through the wide streets of
Carthage in what must have been a horrific public spectacle.196
The gendered aspects of this particular act of brutality and humiliation are so interesting in part because we glimpse Vandal women only
fleetingly in our sources, typically as wives, mothers, daughters, and
sisters.197 When late antique authors wrote about Vandals, they usually
either spoke of them as a people (a gens or gentis) or alluded specifically
to their army or warrior class.198 Indeed, though we have very little
194

195
196
197

198

First wave: Proc. BV 2.5.1, 1:439 (readied for the trip to Constantinople), ibid., 2.9.1, 1:455
(arrive in Constantinople), ibid., 2.14.1718, 1:4845 (organized into cavalry units). Second
wave: ibid., 2.19.3, 1:508. On this point, see Liebeschuetz, Gens into Regnum, pp. 712 and,
more broadly, R. Steinacher, Gruppen und Identitaten: Gedanken zur Bezeichnung vandalisch, in Berndt and Steinacher, Das Reich der Vandalen, pp. 24360. On ethnic markers
in general, see W. Pohl, Telling the Difference: Signs of Ethnic Identity, in W. Pohl and
H. Reimitz (eds.), Strategies of Distinction: The Construction of Ethnic Communities, 300800, Transformation of the Roman World 2 (Leiden, 1998), pp. 1769.
This, at least, would seem to be the implication of Dracontius, Romulea 1.14, 3:134.
Vict. Vit. 2.9, p. 27. For parallels to this parade of infamy elsewhere in the late antique
Mediterranean, see McCormick, Eternal Victory, pp. 1345, 1423, 1812, 186, 249, and 258.
Rarely can we be confident that they were themselves considered Vandals: Sidonius Apollinaris,
Carmen 5.4367, 1:44 (Geiserics sister); Vict. Vit. 2.14, p. 29 (Theoderics daughters); AL
340, pp. 26970 (Damira). More usually they are the wives or mothers of Vandals: Vict. Vit.
1.35, p. 15 (millenariuss widow); ibid., 2.12 (Theoderics wife), p. 28; ibid., 2.14 (Godagis wife;
Gunderics widow), p. 29; ibid., 2.15 (wife of Heldica; Teucharia?), p. 29; ibid., 3.33, p. 88
(Dagilas wife); and ibid., 3.38, p. 91 (mother of Vandal confessors); Proc. BV 1.8.11, 1:347 (first
wife of Thrasamund); Malalas, Chronographia, pp. 459 (wife of Hilderic) and ibid., pp. 4789
(wives of Gelimer). See also Proc. BV 2.2.8, 2.14.8 and 2.19.3, 1:424, 1:483, and 1:508.
Vandals as a gens/gentis or ethnos: see, e.g., Jerome, Adversus Jovinianum 2.7, PL 23, col. 295;
Olympiodorus, frag. 13, p. 172; Orosius, Historiae adversum paganos 7.15.8, 7.38.3, and 7.40.3,
ed. C. Zangemeister, CSEL 5 (Vienna, 1882), pp. 471, 543, and 549; Prosper, Chronicon s.a.
427, p. 472; Vict. Vit. 1.1, p. 3; Proc. BV 1.2.2, 1:311; see also ibid., 1.5.21, 1:334. As in
Hydatius, Chronicon 79 (ad 428), p. 88 this is presumably also the sense in which the royal title
rex Vandalorum was intended, though see Gillett, Was Ethnicity Politicized?, 10910. Vandals
in military context: see, e.g., Possidius, Vita Augustini 28.45, pp. 14850; Prosper, Chronicon
s.aa. 422, 430, 435, and 441, pp. 469, 473, 474, and 478; Hydatius, Chronicon 63 (ad 419), 66
(ad 420), 77 (ad 425), 123 (ad 445), 169 (ad 4567), 195 (ad 460), 223 (ad 464), 232
(ad 466), and 241 (ad 468), pp. 86, 88, 96, 108, 112, 116, 118, and 120; Priscus, frags. 10,
38.1, 39.1, and 62, pp. 242, 340, 342, and 370; Malchus, frag. 2, p. 408; Candidus frag. 2, ed.

59

Staying Roman
evidence from which to reconstruct how Vandals themselves imagined
their own identities, it seems likely that military service was a constituent
element. Martial prowess certainly seems to have been expected of Vandal
princes, whom we regularly encounter leading troops into battle.199 The
husband of Geiserics sister similarly headed a naval expedition against
Campania in 458.200 Beyond the royal family we are not able to trace the
careers of individual warriors, though in 533 we do encounter Vandal
millenarii commanding the right and left wings of the troops assembled at
Tricamarum in a vain last stand against the invading Byzantine forces.201
Throughout the fifth and sixth centuries, the Vandal kings were also regularly able to field sizeable armies, largely (it would seem) of mounted
warriors.202 Significantly, too, the Vandals deported from Africa after the
Byzantine reconquest were sent east as cavalry regiments.203 By contrast,
if there were Vandal blacksmiths, say, or Vandal merchants, we do not
hear about them. From the outside looking in and probably from the
inside looking out at least part of what it meant to be a Vandal in late
antique North Africa was membership in the military elite.
Indeed, what made some individuals look Vandal as they entered
Carthages Nicene churches in the spring of 484 probably had as much
to do with markers of social status as with ethnicity and its signifiers. To
Victor of Vita, clothing was the critical factor that distinguished Vandals
from Romans on first sight. Victor refers to men and women in barbarian
dress (in habitu barbaro) and speaks of the confusion that arose from the fact

199

200
201
202

203

Blockley, in Fragmentary Classicising Historians, 2:470; Sidonius Apollinaris, Carmen 2.34850,

3625, 36870 and 37980, 1:1718 and ibid., 5.3902 and 41920, 1:434; Vict. Vit. 1.30, p. 13;
Paulinus, Epigramma, ed. C. Schenkl, CSEL 16/1 (Vienna, 1888), p. 504; Ennodius, Panegyricus
Theoderico 70, p. 211; Vict. Tonn. s.a. 534.118, p. 38; and Proc. BV passim. See also Hydatius,
Chronicon 80 (ad 429), p. 90 and Vict. Vit. 1.2, p. 3, which imply that families, children, the
elderly, and slaves were not normally reckoned among the Vandals, on which see also above,
previous n.
Genton: Proc. BV 1.6.24, 1:339; Hoamer: ibid., 1.9.2, 1:351; Hoageis: AL 340, l. 15, p. 269;
Tzazon: Proc. BV 1.11.23, 1.24.14, and 2.3.814, 1:3634, 1:410, and 1:42930. Gunthimer:
Vict. Tonn. s.a. 534.118, p. 38. Ammatas: Proc. BV 1.17.11 and 1.18.17, 1:3879. Gebamund:
ibid., 1.18.1 and 1.18.1219, 1:388 and 1:3901 and Vict. Tonn. s.a. 534.118, p. 38.
Sidonius Apollinaris, Carmen 5.4357, 1:44.
Proc. BV 2.3.8, 1:429; see also ibid., 1.5.18, 1:334. On the social ties among the highest Vandal
elite, see also Vict. Vit. 1.35, pp. 1516.
Cavalry: Sidonius Apollinaris, Carmen 5.397424, 1:434; Coripp. Ioh. 3.23655, pp. 567; Proc.
BV 1.8.20 and 1.19.15, 1:349 and 1:393, and esp. ibid., 1.8.27, 1:350; and see also CJ 1.27.1.3
(ad 534), p. 77. On the social status of cavalry in the late Roman world, see P. Rance, Battle,
in P. Sabin, H. Van Wees, and M. Whitby (eds.), The Cambridge History of Greek and Roman
Warfare, 2 vols. (Cambridge, 2007), 2:349; for the early middle ages, see also G. Halsall, Warfare
and Society in the Barbarian West, 450900 (New York, 2003), pp. 1808. Procopius estimated the
Vandal force sent to recapture Sardinia in 533 at 5,000 (BV 1.11.23, 1:3634) and the detachment
under Gebamunds command at Ad Decimum in the same year at 2,000 (BV 1.18.1, 1:388).
Proc. BV 2.14.17, 1:484; Schwarcz, Settlement of the Vandals, p. 57.

60

Vandal power
that a large number of Romano-Africans had adopted this costume.204
Unfortunately we do not know how barbarian garments differed from
Roman clothing. Scholars have long associated Vandal dress with the
loose-fitting tunics, trousers, and boots depicted in (among other places) a
late antique hunting-scene mosaic from the Bordj Djedid neighbourhood
of Carthage. Yet such attire was widespread geographically, common at
all levels of society, and, by the late fifth century, hardly new to the world
of late antique Mediterranean fashion.205 It seems unlikely in itself to
have been sufficient to distinguish Vandals from Romans. Moreover,
the contexts in which we hear about this habitus barbarus (court, church)
suggest that it was put on principally, perhaps even exclusively, for formal
public occasions. This in turn may imply that Victors barbarian dress
consisted of symbols of rank or office which were barbaric only insofar
as they symbolized social prominence, access to power, and cooperation
with the new Vandal regime.206 But, in any case, the overall impression
that Victor of Vita wished to leave was that to the Vandals the clothes
to all external appearances made the man.
Victor, of course, also wanted to challenge that assumption. For
Romano-Africans, the decision to put on barbarian clothes was a political one. The crowd of our Catholics in their dress was huge, Victor
explains, because they served at the royal household.207 The ruling
class of the Vandal kingdom was never a closed group defined by shared
descent real or imagined from a common ancestor or ancestors. As
we have seen, royal titulature drew a distinction between Vandals and
Alans from the fifth century down to the Byzantine reconquest; thereafter Justinian similarly took the honorific conqueror of the Alans and
Vandals (Alanicus Vvandalicus).208 In the 530s, one of Gelimers slaves,
a Goth by the name of Godas, was entrusted with the governorship of
Sardinia. In the event, the appointment was a mistake, but what had
recommended him at the time was his apparent loyalty to the king.209 As
long as they evinced a similar loyalty, access to power was likewise available to Romans (see below, Chapter 3). That such individuals should take

204
205
206
207
208
209

Vict. Vit. 2.89, p. 27: mares uel feminas in habitu barbaro. Rummel, Habitus Vandalorum,
pp. 1334 is sceptical of Victor on this point.
Rummel, Habitus Vandalorum, pp. 13840; Rummel, Habitus barbarus, pp. 23145; and Amory,
People and Identity, pp. 3413.
On the changing significance of symbols in a changing cultural context, see Heather, Goths,
p. 309.
Vict. Vit. 2.8: ingens fuerat multitudo nostrorum catholicorum in habitu illorum . . . ob hoc
quod domui regiae seruiebant.
CJ 1.27.1 (ad 534), p. 77. See also Gillett, Was Ethnicity Politicized?, pp. 10910.
Proc. BV 1.10.2533, 1:35960; cf. Heather, Goths, pp. 3023.

61

Staying Roman
to wearing Victors habitus barbarus was probably interpreted by the Vandal royal family as a sign of their own ideological penetration of African
society: Romano-Africans, too, had come to identify with the Vandal
ruling class, adopting their dress and (at least in theory) their religion,
and thus establishing themselves as faithful subjects of the regions new
kings.210 To target such individuals who attended the Nicene liturgy was
to target renegades and apostates a point that underscores the importance of Arianism to the royal definition of Vandalness, to which we
will return (see below, Chapter 3). To Victor, and to those unfortunate
Romano-Africans who found themselves on the wrong end of the torturers toothed stakes, external appearance and internal identity were less
closely linked. One could dress like a barbarian, and remain a Catholic.
Romano-Africans may have had good material reasons for integrating into the Vandal ruling class, for the latter enjoyed a privileged legal
status under the new regime, as holders of tax-free property allotments
if nothing else.211 But changing ones identity was not invariably as
easy as changing ones clothes. Though not mentioned by our African
sources, and an uncomfortable topic in modern intellectual circles, perceived physical differences could conceivably have played a role in signalling Vandalness: at least, in the sixth century, Procopius tells us that,
from his eastern Mediterranean perspective, Vandals like all Gothic
peoples were white, fair-haired, and tall.212 The British UNESCO
excavations in Carthage have uncovered evidence of the butchering of
horses, whose meat, Jerome tells us, delighted Vandals. The idea offended
Jeromes Roman sensibilities, and so diet too may be another area where
Vandals clung to their own traditions, even as they adapted to their new
environment.213 Then, as now, other more intangible factors such as
210
211

212
213

Liebeschuetz, Gens into Regnum, pp. 789; Berndt, Konfikt und Anpassung, pp. 2234.
Proc. BV 1.5.14, 1:333. On the disputed nature of these allotments, see above, n. 140. The
long-term importance of this issue in the early medieval West in general is explored by M.
Innes, Land, Freedom and the Making of the Medieval West, Transactions of the Royal Historical
Society 16 (2006), pp. 3974.
Proc. BV 1.2.4, 1:311: leuko te gr pantev t smat esi ka tv kmav xanqo, emkeiv
te ka gaqo tv yeiv.
McCormick, Origins, p. 36, on the evidence of J. Schwartz, The (Primarily) Mammalian Fauna,
in H. R. Hurst and S. P. Roskams, Excavations at Carthage: The British Mission, 2 vols. (Sheffield
1984), 1/1:2306, with Jerome, Adversus Jovinianum 2.7, col. 295. Schwartz, Mammalian fauna,
p. 230 points out that the osteological remains alone are insufficient to distinguish between the
species Equus caballus (horse), E. asinus (donkey), and E. hemionis (mule). The discovery of
butchered ostrich (a delicacy) at the British site probably militates against the idea that the
butchering and eating of horses was a result of the famine that struck Africa in 484, but see
Vict. Vit. 3.5560, pp. 99102 with Clover, Carthage and the Vandals, pp. 1517 for the
archaeological evidence of a contemporary mass grave.

62

Vandal power
grooming and gait may have helped cue individuals into one anothers
cultural identities and social standing.214 Finally, the language of the
Vandals probably also served to set them apart from their RomanoAfrican subjects, at least to the end of the fifth century.215 Victor of Vita
relates that, at the Council of Carthage called by Huneric in February
of 484,
our people [the Nicene bishops] said to [the Arian patriarch of Carthage] Cyrila,
Tell [us] what you are arranging. Cyrila said, I dont know Latin. Our bishops
said, We know very well that you always spoke in Latin; you should not excuse
[yourself] now, especially since you kindled the fire of this matter.216

Cyrila could not seriously claim not to speak Latin if Vandalic had not
continued to enjoy some currency as a spoken language, at least within
the restricted circles of the Vandal ruling class. But Latin was the common
language between Africans in the Vandal kingdom. It was certainly the
language spoken by the Romano-African literate elite; in the case of
Cyrilas interlocutors, probably the only language unless perhaps they
happened to speak some Greek, or the lingua punica of the North African
hinterland. In any case, a Vandal could clearly be expected to speak Latin;
a Romano-African could not be expected to speak Vandalic.
Still, even the shadowy existence this language enjoys in our sources
seems to indicate that not all Romano-Africans were completely oblivious to the idiom of the new ruling class. Cultural adaptation went both
ways. A fragment from the Vandal liturgy survives in the richly problematic context of a pseudo-Augustinian polemical treatise;217 Luxorius
hails Hilderic as Vandalric, a Germanic compound meaning king of the
Vandals.218 A poem from the Latin Anthology lamenting the decline of
Latin poetry amongst the more boisterous Gothic entertainments contains another fragment of East Germanic speech:
214
215

216

217

218

In general: Pohl, Telling the Difference, pp. 5161.

die Sprache der Wandalen: Ein Beitrag zur germanischen Namen- und Dialektforschung,
F. Wrede, Uber
Quellen und Forschungen zur Sprach- und Kulturgeschichte der germanischen Volker 59
(Strasburg, 1886); Francovich Onesti, Vandali, pp. 133202; but see also T. L. Markey, Germanic
in the Mediterranean: Lombards, Vandals, and Visigoths, in F. M. Clover and R. S. Humphreys
(eds.), Tradition and Innovation in Late Antiquity (Madison, Wis., 1989), pp. 5171 and cf. Amory,
People and Identity, pp. 1028.
Vict. Vit. 2.55, pp. 456: Conuersique nostri Cyrilae dixerunt: propone quod disponis. Cyrila
dixit: nescio Latine. Nostri episcopi dixerunt: semper te Latine esse locutum manifesto
nouimus; modo excusare non debes, praesertim quia tu huius rei incendium suscitasti .
Collatio beati Augustini cum Pascentio ariano, PL 33, col. 1162. For a discussion of this fragment, see
H. Tiefenbach, Das wandalische Domine miserere, Historische Sprachforschung (Historical Linguistics)
104 (1991), pp. 25168.
AL 206, p. 154.

63

Staying Roman
Among the Eils of the Goths, Scapia matzia ia drincan,
No-one dares to proclaim proper verses.219

The Gothic is still intelligible to modern readers: To your health! Bring

food and drink!220 Of course, this poem could very well have been
written in Italy, Gaul, or Spain, possibly even by a cultured barbarian.
Even so, in the second quarter of the sixth century, someone in Africa
understood enough of both languages to get the joke.
In order to secure and control their new realm, Africas Vandal rulers
made a political and cultural choice to emphasize the continuities
between their regime and that of the late Roman state. This decision
appears to have been intended at least in part to reassure multiple audiences throughout the Mediterranean that the Romanness of Africa would
be safe under barbarian rule; but it should not blind us to the fact that
the very existence of an autonomous Vandal kingdom in what had for
centuries been Roman Africa represented a profound historical discontinuity not just for this region, but for the empire as a whole. Quite
apart from the violence and disruption that it caused, the Vandal conquest reconfigured Africas political ties to the rest of the late Roman
world: what had once been a matter of internal affairs was now one of
international diplomacy. In addition, Vandal dominion brought with it
a new ruling elite, as well as new offices such as praepositus regni and
millenarius. Language and dress may have served to differentiate Vandals
and Romans in Africa to the end of the fifth century and beyond; and
even if the newcomers did fully assimilate to the leisured lifestyle of the
Romano-African elite their ability to do so was to no small degree predicated on the flight and dispossession of Roman landowners. Though
the Vandal kings issued laws in good Roman style, Matthew Innes has
recently demonstrated the dangers of confusing continuity of legal form
with continuity in social practice.221 Moreover, the obvious point that
whatever his descent the king of the Vandals and Alans was not the
Roman emperor bears some emphasis. Indeed, according to Procopius,
the sixth-century emperor Justinian tried to justify his armys invasion
of Africa to the regions Vandal population through an appeal to the
219

220

221

AL 279, p. 201: Inter eils Goticum scapia matzia ia drincan / non audet quisquam dignos
edicere versus. For the Vandals as Goths, see, e.g., Liber genealogus 618, ed. Mommsen in MGH
AA 9:195; Procopius also includes the Vandals among the Gothic peoples (Got{ik {nh) at
BV 1.2.2, 1:311.
In the app. crit. to AL 279, p. 201, Shackleton Bailey cites Rieses note to this effect: eils
salutem, skapja procuratorem peni vel skap procura, praebe, jah matjan jah dringkan, et cibum et
potum interpretatur.
Innes, Land, p. 54.

64

Vandal power
law of succession that Geiseric had established for his kingdom and that
Gelimer had violated by imprisoning Hilderic and seizing the throne
for himself.222 As far as Justinian knew, among Vandals themselves the
legitimacy of a ruler derived not so much from his relationship to the
Roman state or his maintenance of Roman forms of power as from his
place in a line of descent that extended only three generations into the
past.
I suspect, however but cannot prove that, even among Vandals, both
of these other factors mattered. Ultimately the Vandal kings addressed all
four of their audiences at much the same time and in much the same
language. The Vandal capture of Carthage proved Geiserics strength not
just to the Carthaginians, but to Vandals and other barbarians, as well as
to Romans across Africa and throughout the rest of the Mediterranean.
These Romans, both at home and abroad, realized that like it or not
they would have to deal with Geiseric. But the Vandal king also proved
his power and importance by securing and maintaining the diplomatic
friendship of the Roman emperor and, indeed, of the other powers
throughout the Mediterranean world: the kings of the Ostrogoths, the
Visigoths, even the Sueves and the Huns. This was true not only for
Geiseric (who only partially succeeded in this respect) but for all of
his successors as well. The union of Huneric and Eudocia like that
of Thrasamund and Amalafrida was the supreme expression of that
friendship.
The marriage of the Vandal prince and the imperial princess cast a
long shadow over the exercise of Vandal kingship. We have little evidence
through which to understand the nature of Vandal relations with the other
kingdoms of the barbarian West, but what we do have suggests that those
exchanges that took place did so within the general framework already
established by the conventions of late Roman diplomacy. In this respect,
a marriage connection with the Theodosian house can only have been an
additional asset. This was even more the case when those exchanges were
between the Vandal kingdom and the Roman state. Geiseric, Huneric,
and Hilderic had a personal connection to the Theodosian imperial house
that continued to be recognized even after that family had ceased to rule
in any other part of the empire, East or West. Geiserics intervention in
imperial politics may have irritated eastern emperors, but this connection
to the house of Theodosius gave the Vandals rule in Africa a kind of
legitimacy it would otherwise have lacked. And, in Hilderics Africa, the
poetry of Luxorius vividly illustrates that it mattered deeply, at least to
222

Proc. BV 1.16.1314, 1:384.

65

Staying Roman
some Romano-Africans, that an heir of Valentinian sat on the Vandal
throne.
In Africa itself, the Vandal kings adopted and adapted the structures of
Roman provincial rule and the Roman vocabulary of power to suit their
own needs. They cast their rule in a triumphal light, arrogated to themselves the legislative and other prerogatives of emperors, and attempted to
establish Carthage as an Arian patriarchate. These moves were addressed
to Romano-Africans, to be sure, but also to Vandals. For already in the
fifth century Vandals and Romano-Africans must have shared many of
the same attitudes, perspectives, and tastes. From the vantage point of the
twenty-first century their ways of life are virtually indistinguishable. Even
as they remained recognizably Vandal, the barbarians were also becoming
Romanized. What we perceive as the Roman aspects of Vandal kingship the building programmes, court poetry, and gardens as much as the
regalia, the laws, and offices would doubtless have appealed as much to
Vandals as to Romano-Africans. So too the fierce pride in Africa; for in
the fifth and sixth centuries the Vandals seem to have found Romanness
in Africanness.

66

Chapter 2

F L I G H T A N D C O M MU N I C A T I O N S

In the Vandal century, Africa maintained its connections across the

Mediterranean through the circulation of people, goods, and ideas.
Africans themselves appear unexpectedly in remote corners of the late
Roman world as well as in its busy centres, individuals in search of
advancement, opportunity, theological solidarity, and security. For their
movements were not always voluntary: in consolidating his control of
Africa, the Vandal king Geiseric seems to have been intent on ridding
the province of the most threatening elements of the existing power structure. He banished the bishop of Carthage, Quodvultdeus, and eventually
drove the remainder of the Carthaginian clergy into penal exile.1 Victor
of Vita and Fulgentius of Ruspes biographer both also agree that Geiseric exiled the senators and honorati of Carthage from the metropolis,
ultimately banishing them from Africa altogether.2 If they chose not to
leave, Victor informs us, they were reduced to slavery.3 The flow of such
refugees and exiles, alongside that of other travellers from Vandal Africa,
clearly reveals to us some of the regions points of contact with the rest of
the Mediterranean world. But we can also trace this network of communications through the distribution of traded commodities, the exchange
of letters and books, and even the diffusion of distinctively African personal names and saints cults, which were borne outwards from Africa by
these same travellers. In their collective movement, we catch a reflected
glimpse of the connections that bound the late Roman world together,
even as the political structures that for centuries had unified the empire
began to collapse.

2
3

Vict. Vit. 1.1516, p. 8: poenali exilio, doubtless with a pun on poena, punishment, and Poenus,
Carthaginian or Punic. This was presumably after the death of Quodvultdeus successor, Bishop
Deogratias, in 457: Vict. Vit. 1.51, p. 22.
Vict. Vit. 1.15, p. 8, and V. Fulg. 1, p. 11; see also Val. Nov. 34.23 (ad 451), p. 141, which refers
to the Vandal expulsion of distinguished African property owners.
Vict. Vit. 1.14, pp. 78; see also ibid., 1.12, p. 7, and Proc. BV 1.5.11, 1:333.

67

Staying Roman
1. individuals
Even before the fall of the North African metropolis, Africa had already
seen the displacement of at least some of its notables as local leaders
proved unable or unwilling to meet the challenges posed by the Vandal
presence. In 429, bishops from the region surrounding Hippo Regius
fled to the seaport for protection, where, Possidius of Calama tells us,
they found themselves together with the comes Africae Boniface and
his army of Goths blockaded and besieged by the Vandals for fourteen
months.4 Bishop Optatus of Vescera in Mauretania Sitifensis may very
well have fled to Rome at the same time.5 But in the wake of the Vandal
capture of Carthage, aristocratic refugees from the recently conquered
African provinces began to flood the lands of the Roman Mediterranean.
Over the course of the following century they were to be joined by
other Africans, who travelled abroad for reasons of their own some of
them fleeing persecution, others probably not. Ascertaining who they
were, when they left, and where they decided to go are the first steps
in unravelling their relationship to the Vandal venture in Africa, and
the significance of their movement to the larger question of African
integration into the late Roman world.
1.1. Social profile
Not counting the political and military contacts already discussed (see
above, Chapter 1), we can trace the movements of at least fifty-four
African travellers perhaps more, probably no fewer beyond the territory of the Vandal kingdom with a greater or lesser degree of accuracy.
For the most part they were of remarkably high social status. Clergymen and monks (29 out of 54) solidly outnumber the laity (22 out of
54), with only three e migres of undetermined status (see Figure 2.1).
Despite canonical prohibitions against bishops travelling overseas without first obtaining a forma from their primate, no fewer than twenty-two
of the fifty-four were bishops at the time they left Africa.6 Another was
4

Possidius, Vita Augustini 28.1213, p. 154. On Geiserics later expulsion of some African bishops
from their sees, see Prosper, Chronicon s.a. 437, p. 475, with PCBE 1:7834, s.n. Novatus; ibid.,
pp. 8906, s.n. Possidius 1, at p. 895; and ibid., p. 1069, s.n. Severianus 3.
ICVR n.s. 9370, 9516, and 9517, with Gesta conlationis Carthaginiensis, anno 411 1.120, ed. S. Lancel,
CCSL 149A (Turnhout, 1974), p. 99; see also G. B. de Rossi, La Roma sotterranea cristiana, 3 vols.
(Rome, 186477), 2:2215; PCBE 2/2:1561, s.n. Optatus 2; PCBE 1:8012, s.n. Optatus 4;
and J. Conant, Europe and the African Cult of Saints, circa 350900: An Essay in Mediterranean
Communications, Speculum 85 (2010), p. 43.
Breviarium Hipponense, Brevis statutorum 27, ed. C. Munier, in Concilia Africae a.345a.525, CCSL
149 (Turnhout, 1974), p. 41; Canones in causa Apiarii 23, ed. Munier, in CCSL 149:108; Registri
Ecclesiae Carthaginensis Excerpta 56 and 94d, ed. Munier, in CCSL 149:193 and 214.

68

Flight and communications

undetermined
5%

laity
41%

Fig. 2.1.

churchmen
54%

Africans abroad, 439533: churchmen and laity

miscellaneous
7%
merchants
4%
slaves
7%
teachers
6%

bishops
43%

monks
and lesser
clergy
11%

aristocrats
22%

Fig. 2.2.

Africans abroad, 439533: occupations

to become Pope. Twelve more were aristocrats, wealthy landowners, or

officeholders. The balance of the Africans whom we see abroad in the
Mediterranean over the course of the Vandal century all in all, another
nineteen individuals are a motley crew of lesser clergy and monks,
merchants, grammatici, a Latin tutor, a water-diviner, and a handful of
slaves (see Figure 2.2).
The top-heavy distribution serves as a caution, highlighting the role of
social status in the creation of the surviving documentary record. Indeed,
69

Staying Roman
many of our travellers belonged not just to the elite, but to the super-elite
of late Roman society. Two of our twenty-two bishops were metropolitans of Carthage, while one of the monks, Fulgentius of Ruspe, was from
a leading aristocratic family and was himself later to become the most
prominent African churchman of the early sixth century. Fulgentius
grandfather (also one of our travellers) and another refugee named Caelestiacus were both said to have been members of the Carthaginian curia.7
The terms that the eastern bishop Theodoret of Cyrrhus (ad 42349 and
45166) used to refer to this Caelestiacus most excellent and most
magnificent (4 : / )) leave little
doubt that the man also belonged to the ranks of the high senatorial
nobility.8 Theodoret refers to other African travellers as most well-born
() and most magnificent ( )).9 One of
these was well placed enough to secure an office for himself after his
flight from Africa.10 Under the western emperor Majorian (ad 45761),
yet another African (and skilled poet) named Domnulus similarly rose to
the post of Quaestor of the Sacred Palace.11 In short, then, most of the
Africans whom we see abroad in the later fifth and early sixth centuries
were members of the powerful, privileged, and literate social order which
wrote primarily for and about itself, thus producing the documents on
which our analysis is based.
In this sense our sample is unlikely to be representative of the Africans
who circulated abroad in general in the fifth and sixth centuries. In his
examination of the epigraphic evidence for travel and travellers in the late
antique western Mediterranean as a whole, Mark Handley has found that
secular individuals of very high status represent only about 5 per cent of
the entire corpus, and that bishops account for a comparable proportion

7
8

9
10
11

V. Fulg. 1, p. 11; Theodoret, Ep. 33, 2:94. On the Carthaginian curia, see CTh 12.1.27 (ad 339),
p. 669.
Theodoret, Epp. 2932 and Epp. 346, 2:8692 and 968; also most magnificent and esteemed
( )1 / 1 ) at Theodoret, Ep. 33, 2:94. For these terms of address,
see esp. Theodoret, Epp. 33 and 37, 2:94 and 1002 with PLRE 2:972, s.n. Sallustius 5 (perhaps
comes Orientis) and ibid., p. 1028, s.n. Stasimus (also a comes). For Your greatest excellence
( & 4 ) alone, see Theodoret, Ep. 124, 3:92 (Marana scholasticus) and Ep. 126, 3:102
(magistrates of Zeugma). For Your magnificence ( & )) alone, see the Index
des mots grecs to Theodoret, Correspondance 3:261 s.v. ); those addressed with
this title include comites, magistri militum, sophists, quaestors, and curators.
Theodoret, Ep. xxiii, 1:94 (Maximian) and Ep. 70, 2:1524 (Maria and Eudaemon).
Theodoret, Ep. 70, 2:154, probably in the West (  ?0), though C. Courtois, Les Vandales
et lAfrique (Paris, 1955), p. 280 n. 8 suggests Dusae, i.e., Bithynia.
Sidonius Apollinaris, Epistulae 9.13.4 and 9.15.1 v. 38, ed. and trans. A. Loyen, in Sidoine
Apollinaire, 3 vols. (Paris, 196070), 3:164 and 176. PLRE 2:374, s.n. Domnulus 1.

70

Flight and communications

of the total.12 This is in stark contrast to these two groups combined
dominance in our sample. Similarly, 81 per cent of the epigraphically
attested late antique travellers were not religious office-holders: a much
more secular cross-section of late Roman society than the Africans whose
movements are attested in the textual sources of the Vandal period.13
These trends are borne out in the five inscriptions collected by Handley that may perhaps be attributed to the Vandal fifth or sixth century
(unfortunately none of which is precisely datable and which are therefore
used here primarily for comparative purposes). None of the individuals
they commemorate is identified with respect to social rank, and if they
were office-holders, bishops, members of the clergy, or monks, we do
not hear about it.14 In fact, only one indicates his profession at all: a
certain Ithallas, a ship captain (0)) from Leptis Magna who died
in Syracuse, perhaps in the fifth century.15 Whenever he moved from
Tripolitania to Sicily, Ithallas serves to remind us of an important group
that is probably severely under-represented in the textual sources, for the
movements of all of our travellers no less than the literally millions of
sherds of African ceramics distributed on archaeological sites throughout
the Mediterranean attest to the circulation of sailors and merchants in
the Vandal era, many of them doubtless African crews who have left no
trace in the written record.16
Our travellers were also predominantly men. This too is probably a
product of the biases of the documentary evidence: in general, when
we know their gender, women seem to have comprised about 16 per
12

13
14

15

16

M. Handley, Dying on Foreign Shores: Travel and Mobility in the Late Antique West, Journal of Roman
Archaeology Supplementary Series 86 (Portsmouth, RI, 2011), pp. 413 and 49. Many thanks
to Mark Handley for providing me with a copy of the manuscript in advance of its publication.
Ibid., p. 49.
G. Alfoldy, Die romischen Inschriften von Tarraco, Madrider Forschungen 10, 2 vols. (Berlin, 1975),
1:4389, no. 996 = Handley, Dying on Foreign Shores, no. 68 (Tarragona); A. Silvagni (ed.),
Monumenta epigraphica christiana saeculo XIII antiquiora quae in Italiae finibus adhuc existant iussu Pii
XII pontificis maximi, 4 vols. (Vatican City, 1943), 2/1:tab. 9.5 = M. Handley, Dying on Foreign
Shores, no. 102 (Milan); G. Brusin, Inscriptiones Aquileiae, Pubblicazioni della Deputazione di
storia patria per il Friuli 20, 3 vols. (Udine, 19913), 3:1111, no. 3180 = Handley, Dying on
sel and J. Sa
sel, Inscriptiones latinae quae in Iugoslavia inter
Foreign Shores, no. 135 (Aquileia); A. Sa
annos MCMII et MCMXL repertae et editae sunt, Situla: Dissertationes Musei Nationalis Labacensis
25 (Ljubljana, 1986), p. 352, no. 2671 = Handley, Dying on Foreign Shores, no. 450 (Salona); and
below, next n. See also, perhaps, Handley, Dying on Foreign Shores, no. 87.
AE (1985), 124, no. 484 = C. Wessel, A. Ferrua, and C. Carletti (eds.), Inscriptiones Graecae
Christianae veteres Occidentis, Inscriptiones Christianae Italiae Subsidia 1 (Bari, 1989), p. 25,
no. 93 = Handley, Dying on Foreign Shores, no. 322 (Syracuse).
African merchants in the East in the 460s: Malchus, frag. 17, p. 424. See also Fulgentius of Ruspe,
Abecedarium ll. 2478, ed. C. Lambot, in S. Fulgentii episcopi Ruspensis opera, ed. J. Fraipont,
2 vols., CCSL 9191A (Turnhout, 1968), 2:884 for Vandal-era African ship masters. On African
economic connections abroad, see also below, section 2.

71

Staying Roman
cent of the late antique travellers attested in the inscriptions.17 Yet (apart
from the princess Eudocia) the only women we see travelling abroad
without a male relative in the Vandal period in either the written or the
epigraphic sources were both slaves. In one of the many letters that he
wrote on behalf of displaced Africans, Theodoret of Cyrrhus informs us
that a certain Maria (the daughter of an aristocrat named Eudaemon)
and her anonymous slave girl were taken captive in the disaster which
befell Libya, and sold into slavery together in northern Syria.18 When
their daily work was done Marias former handmaid, now fellow slave,
continued to tend to her one-time mistresss needs, washing her feet,
looking after her bedding, and so on. The slave girls ministrations to
Maria caught the attention of their common masters and then, Theodoret
tells us, of the entire city. Their story became well known, and the soldiers
of the citys garrison pooled their resources to pay Marias purchase-price,
redeeming her and returning her to freedom. Maria learned that her
father had escaped Africa and was now an office-holder in the West.
She was determined to find him, and so Theodoret wrote to Bishop
Eustathius of Aegae asking him to help the young aristocrat on her way.19
Concerning the fate of Marias slave girl (whose devotion to her former
mistress won Maria her freedom) we are completely ignorant, for this was
apparently of no interest either to Theodoret or to his correspondent.
It is not certain that Maria and her handmaiden were inhabitants of
Carthage or, for that matter, even victims of the Vandals.20 Indeed, if
other Africans were enslaved and sold overseas when their province fell
to the Germanic invaders in 439, what became of them is unknown to
us.21 But Marias story also serves to highlight once again the difficulties
of seeing non-elite Africans abroad in the late antique Mediterranean. We
hear about these slaves only because Maria was a Roman citizen of noble
birth. Her story of captivity and redemption were, in Theodorets opinion, worthy of a tragedian.22 But the disturbing fact that Romans could
buy other Romans as slaves on eastern markets as a result of barbarian
17
18
19

20
21

22

Handley, Dying on Foreign Shores, pp. 378.

Theodoret, Ep. 70, 2:152: '@ A  *0  ; & B*0.
Ibid., 2:1524. The incident provides an interesting illustration of the psychological effects of
slavery on the enslaved themselves; see K. Bales, The Social Psychology of Modern Slavery,
Scientific American (Apr. 2002), pp. 808.
Courtois, Vandales, p. 280 n. 8 is sceptical about Marias Carthaginian origins.
Courtois, Vandales, p. 280, notwithstanding the statements of Vict. Vit. 1.14, pp. 78 and Proc.
BV 1.5.11, 1:333 to the effect that African senatorial aristocrats were enslaved by the Vandals;
see also Vict. Vit. 1.12, p. 7: nam et senatorum urbis non paruam multitudinem captiuauit.
Theodoret, Ep. 70, 2:152: C
 D. Theodoret seems to have made the same assessment of the fall of Carthage in general: Epp. 29 and 33, 2:86 and 94; cf. Ep. 86, 2:226 and
Ep. 120, 3:82 (of Theodorets own affairs).

72

Flight and communications

activity in the West and the ambiguities of Romanness this underscores for a modern observer seem not to have troubled Theodoret.
Nor does the situation seem to have troubled Marias owners, who from
Theodorets account appear to have been quite happy to exercise their
legal right to keep their freeborn slave in bondage for five years or until
she could refund her purchase price.23 To Theodorets credit, he did
feel some sense of obligation to the young woman. He seems slightly
embarrassed that it was the local garrison that redeemed Maria, quickly
explaining to Bishop Eustathius that he himself had been away from
Cyrrhus at the time. Once freed, Theodoret placed Maria in the charge
of one of his deacons, provided her with provisions for ten months, and
assisted her in the search for her father. But we are left with the sense
that in Theodorets mind this was something due to Maria primarily
because of her noble birth, and only secondarily (if at all) because she
was a Roman.
Marias story unexpectedly illustrates another aspect of movement in
the late antique Mediterranean: the importance of companions. Most of
our fifty-four African e migres appear in the sources singly, lone individuals whose African origins were noted by someone else. Some of
them may indeed have travelled alone: a fifth-century inscription from
Aquileia commemorates a certain Restutus, seemingly an African who
had come to the city and died there having established local bonds, but
nevertheless far from his family.24 Yet whether by choice or by necessity,
many travellers clearly moved in groups. Caelestiacus fled east with his
wife, children, and slaves.25 A Carthaginian citizen named Cris . . . who
died in late antique Milan was commemorated by his wife and brother,
probably implying family movement of some sort.26 The citizens of
Tipasa seem to have emigrated together in two waves, the first to Spain,
the second to Constantinople.27 Bishop Quodvultdeus of Carthage was
banished from Africa together with the biggest crowd of clergymen
(maxima turba clericorum).28 Bishop Possessor of Zabi (modern Bechilga,
Algeria) appears in Constantinople accompanied by his deacon Justin.29
23
24

25
26
27
29

CTh 5.7.2 (ad 409), pp. 2234.

Brusin, Inscriptiones Aquileiae 3:1111, no. 3180 = Handley, Dying on Foreign Shores, no. 135;
G. Rinaldi, Osservazioni sullepitaffio di Restuto, in Aquileia e lAfrica, Antichit`a altoadriatiche
5 (Udine, 1974), pp. 1819.
Theodoret, Epp. 29, 31, and 356, 2:88, 90, and 98.
Silvagni, Monumenta epigraphica christiana 2/1, tab. 9.5 = Handley, Dying on Foreign Shores,
no. 102. Handley favours a sixth-century date for this inscription: ibid., p. 60.
28 Vict. Vit. 1.15, p. 8.
Vict. Vit. 3.2930, pp. 867.
Hormisdas, Epistula 115.2, ed. A. Thiel, in Epistulae Romanorum pontificum genuinae, vol. 1
(Braunsberg, 1867; repr. Hildesheim, 1974), p. 917. On his career, see also Notitia, M. Sitif. 40,
p. 133 and, in general, PCBE 1:889, s.n. Possessor.

73

Staying Roman
Fulgentius of Ruspe chose a fellow monk named Redemptus to go with
him on his planned journey to Alexandria.30 All together, including
single travellers, our fifty-four individuals can be divided into perhaps
forty-two separate groups, each moving separately across the Mediterranean.
Most of these individuals seem to have been fairly mature by the standards of late antiquity when they embarked on their travels. Firm statistics
are impossible to come by, but, as we have seen, both Caelestiacus and
Eudaemon were old enough to have married and had children, and the
implication seems to be that this was true of Fulgentius grandfather
Gordian as well.31 Fulgentius himself would have been about thirty-two
when he reached Rome, having abandoned his plans to head east.32 Similarly, the Cris . . . buried in Milan and another African named Titzanus
commemorated in an inscription in Tarragona had travelled abroad and
died by the ages of roughly forty and thirty-five, respectively.33 By contrast, the fact that Maria appears to have been unmarried may imply that
she was still quite young when she was captured and sold into slavery.34
What we can trace of the earlier careers of our ecclesiastical refugees
suggests that they were considerably older at the time of their flight or
exile, at least in social terms. Thus, for example, while he was still a deacon Bishop Cyprian (who fled to Syria after the Vandal invasion) may
well have served as the intermediary between Augustine and Jerome in
a complex exchange of letters in 4025.35 The African cannons differed
slightly as to whether men could be ordained as clerics at age twenty or
twenty-five, but even if Cyprian had been a twenty-year-old deacon in
30
31
32

33

34

35

V. Fulg. 8, p. 47.
Caelestiacus: Theodoret, Epp. 29, 31, and 356, 2:88, 90, and 98. Eudaemon: Theodoret,
Ep. 70, 2:152. Gordian: V. Fulg. 1, p. 11.
He was in Rome for Theoderics visit there in 500 (V. Fulg. 9, pp. 557). Fulgentius is said to
have died in 533 or 534 at the age of sixty-five, which would place his birth c.468: see, in general,
PCBE 1:50713, s.n. Fulgentius 1.
Silvagni, Monumenta epigraphica christiana 2/1, tab. 9.5 = Handley, Dying on Foreign Shores, no. 102
and Alfoldy, Die romischen Inschriften von Tarraco i. 4389, no. 996 = Handley, Dying on Foreign
Shores, no. 68.
Theodoret, Ep. 70, 2:1524 with M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs
from Gaul and Spain, ad 300750, BAR International Series 1135 (Oxford, 2003), pp. 98100, who
cites eight inscriptions from late antique North Africa with evidence for girls age at marriage: the
young women in this small sample were on average married and dead by twenty-four. Elsewhere
in the contemporary Mediterranean girls and young women married at ages from as young as
twelve to as old as thirty-six.
Cyprian may possibly also be the bishop of Thuburbo Maius (mod. Henchir Kasbat), who signed
the mandatum of the Council of Carthage in 411 (as no. 229) and was perhaps one of the two
bishops of the same name who signed the synodal letter of the anti-Pelagian synod of Carthage
in 416: see PCBE 1:2578, s.n. Cyprianus 34 and the sources cited there; see also Azema,
Correspondence, 2:130 n. 1.

74

Flight and communications

402, by the time of the Vandal capture of Carthage he would have been
fifty-seven.36 Similarly, Quodvultdeus of Carthage was probably already
a deacon around 408, and is therefore likely to have been in his early
fifties at least when he was banished from Africa.37 Quintianus of Urusi
had been a bishop for at least twenty-seven years when he escaped the
Vandal persecution for the East, and was therefore almost certainly in his
late forties and probably considerably older when he fled.38
In sum, then, the Africans revealed to us in the documentary sources
as having been abroad in the wider Mediterranean over the course of
the Vandal century are for the most part exactly the kind of people who
produced those sources in the first place. They are by and large mature
males of an elevated social class, both secular and ecclesiastical. Of course
this does not mean that women, children, the elderly, and people of a
lower social status did not travel beyond the frontiers of Africa in the fifth
and sixth centuries; it simply means they are more difficult to see.
1.2. Chronology
If we consider the range of dates at which these groups of Africans first
become visible beyond Africa, we notice two particular spikes (see Figure
2.3). The first is in the period 43950; the second is a smaller rise beginning in the 480s and continuing into the second decade of the 500s. In the
case of the first spike, the cause is clear: the fall of Carthage to the Vandals
and the disruption of African elite society through the flight or exile of
many of the provinces leading secular and ecclesiastical aristocrats. The
case of the later rise is more ambiguous. There is a correlation with
the persecution of Huneric (ad 484), which certainly precipitated the
flight of a handful of the travellers we see in the 480s, including Quinitanus of Urusi and the citizens of Tipasa. But the grammatici Priscian and
Pomerius could have lived abroad for quite some time before they first
emerged in the sources around the turn of the sixth century, and Gelasius
certainly cannot have been new to Rome when he was elected bishop in
492 (see below, section 1.3). A third, and much smaller, rise in the early
530s is probably equally deceptive, for the travellers we see abroad in that
decade were not necessarily recent migrants; and, indeed, in some cases,
the reverse is probably true.
36

37
38

Twenty-five: Breviarium Hipponense, Breuis statutorum 1b, p. 33; Concilium Carthaginense a.525, ed.
C. Munier in CCSL 149:264, ll. 3912; and Ferrandus, Breviatio canonum 121, ed. C. Munier in
CCSL 149:297. Twenty: Canones in causa Apiarii 16, p. 105.
See PCBE 1:9479, s.n. Quodvultdeus 5, at p. 947.
Vict. Vit. 1.29, p. 13. Quintianus seems still to have been in Africa down to 484: Vict. Vit. 2.22,
p. 32 with ibid., 2.17, p. 30; Notitia, Proc. 20, p. 118.

75

Staying Roman
14

Number of groups

12
10
8
6
4
2
53
1
33

20

1
30
52

0
51

00

1
1
50

90

49
1

1
48

80
1
47

70

60

1
46

1
45

43

50

First visible

Fig. 2.3. Africans abroad, 439533: distribution over time (by group)

1.3. Directions of movement

Our forty-two groups of travellers are evenly divided in their destinations between East and West. The sources written in Africa itself that
discuss the movement of Africans abroad in the Vandal period mention
destinations such as Syracuse, Naples, Rome, Edessa in Macedonia, Constantinople, and Alexandria, or speak in terms of broad areas like Spain,
Sicily, and Italy.39 The general impression left by this handful of examples
is that Africans immediate connections to the rest of the Mediterranean
did not often penetrate far inland from major port cities and centres of
power, and in any case were closest with Italy. This picture is broadened,
however, by the local, non-African sources, which show us Africans in
southern Gaul, Ravenna, and most especially the Syrian hinterland (see
Map 2.1).
East
In the eastern Mediterranean, there can be little question that Constantinople was the primary destination of a majority of our African
travellers. Of the twenty-one traceable groups who headed east, fourteen ended or sought to end their journey in the imperial capital or
its immediate vicinity (see Table 2.1). To contemporary observers, perhaps the most famous Africans in fifth- and sixth-century Constantinople were the confessors of Tipasa. In 484, the Vandal king Huneric
ordered that the Catholics (perhaps only the Catholic clergy) of Tipasa in
39

The two sources are Vict. Vit. and V. Fulg. For specific references, see below.

76

Flight and communications

Map 2.1.

Africans abroad in the Mediterranean, 439533

Mauretania Caesariensis have their tongues and right hands cut off for
publicly celebrating the liturgy. Some of these mutilated confessors were
said to have made their way to the eastern capital, where as they were
miraculously still able to speak perfectly they were warmly received
by the emperor Zeno (ad 47491) and the empress Ariadne. There they
appear to have remained quite the cause cel`ebre well into the sixth century. Victor of Vita mentions only the subdeacon Reparatus as being
present at Constantinople, but Procopius refers to many such victims of
Hunerics persecution who had fled to the imperial capital, including
two who frequented the fleshpots of the metropolis.40 Marcellinus comes,
writing in the early sixth century, claimed that he himself had seen the
African refugees in Constantinople speaking without tongues; so did
Marcellinus older contemporary Aeneas of Gaza, who further claimed
to have been sceptical of the miracle before witnessing it himself.41
40

41

Vict. Vit. 3.30, p. 87; Proc. BV 1.8.4, 1:3456; see also CJ 1.27.1.4 (ad 534), p. 77. W. Lackner,
Westliche Heilige des 5. und 6. Jahrhunderts im Synaxarium Ecclesiae Constantinopolitanae,
Jahrbuch der o sterreichischen Byzantinistik 19 (1970), pp. 1929.
Marcellinus comes, Chronicon s.a. 484.2, p. 93; Aeneas of Gaza, Theophrastus, PG 85, cols. 10001.

77

Staying Roman
Table 2.1. Africans abroad, c.439c.533: Constantinople and Chalcedon
Name

No. Itinerary

Anonymous1

2+

Africa Constantinople

Aurelius 1
Aurelius 2

1
1

Pupput Constantinople
448/51
Hadrumetum Constantinople 448/51
Chalcedon

bishop
bishop

Florentius

Africa Cyrrhus Ankyra

Constantinople

440/9

bishop (refugee)

Januarius
Possessor
Justin

1
2

Byzacena Constantinople
Zabi Constantinople

448
517

bishop
bishop and
deacon

Priscian

491/518

grammaticus

Reparatus 2+
fellow
citizen(s)
Restitutianus 1

Caesarea (in Mauretania?)

Constantinople
Tipasa Constantinople

484

subdeacon
(refugee)

Africa Chalcedon

451

bishop

Sacconius
Speciosus

1
1

Uzalis Constantinople
Africa (?) Constantinople

484/93
532

bishop (exile)
teacher of Latin

Valerian 1

448?/451

bishop

Valerian 2

Africa Constantinople?
Chalcedon
Bassianensis Constantinople?
Chalcedon

448?/451

bishop

Date

Status

before 533 dispossessed

landowners

Unlikely to have travelled together; therefore counted as two groups.

Evagrius Scholasticus perhaps confuses Aeneas account and that of Procopius when he rather enthusiastically records that the historian had
met and spoken with the refugees a claim Procopius himself did not
make.42 The miraculous confessors were, however, dead by the time the
African bishop Victor of Tonnena wrote his chronicle later in the sixth
century, and Gregory the Great could only claim that when he was in
Constantinople in the 580s he had met a certain elderly bishop who had
himself seen the famous refugees.43
42
43

Evagrius, Historia Ecclesiastica 4.14, pp. 1634.

Vict. Tonn. s.aa. 479.50 and 566/7.173, pp. 16 and 54; Gregory I, Dialogi 3.32, ed. A. de Vogue , in
Diologues, 3 vols., SC 251, 260, 265 (Paris, 197880), 2:3902. Perhaps the second-hand nature of
his information helps to explain why Gregory misdates the persecution to the time of Justinian.

78

Flight and communications

By the late fifth century, though, the path from Africa to the imperial
capital had become a well-travelled one, and the African ecclesiastical
presence in the Queen of Cities seems to have been particularly palpable
at this time. Thus, for example, on 22 November 448 at least two bishops from Byzacena Aurelius of Pupput and Januarius of Macriana
participated in the Constantinopolitan synod that condemned Eutyches,
the archimandrite of a local monastery, as a heretic.44 They would appear
to have been joined by a compatriot, Bishop Aurelius of Hadrumetum, who was to remain in the imperial capital through the following
spring.45 Also present in Constantinople in November 448 was a bishop
named Valerian, another African, who intervened in an earlier session
of the same proceedings with a statement (in Latin) of his own Nicene
orthodoxy.46 Like Aurelius of Hadrumetum, this Valerian appears to
have taken up residence in the metropolis, and together with two of
their fellow-countrymen in October 451 the bishops both seem to have
taken part in the council of Chalcedon, just across the Bosporus from the
imperial capital.47 Somewhat earlier, the African bishop Florentius, too,
had travelled to Constantinople, this time by way of northern Syria;
and at the end of the fifth century, Bishop Sacconius of Uzalis similarly sought refuge there from the Vandals and their Arianizing policies.
Yet the imperial city was itself not free of theological entanglements. In
493, Pope Gelasius I wrote to Bishop Sacconius, praising him for the
constancy of his faith, but rebuking him for being in communion with
an eastern ecclesiastical establishment that, it was felt in the West, had
44

45

46
47

Aurelius: Collectio novariensis de re Eutychis 2.161.27, in ACOec. 2/2/1:20; Gesta actionis primae
552.28, in ACOec. 2/1/1:146 (Greek) and ACOec. 2/3/1:130 (Latin); PCBE 1:1289, s.n. Aurelius 4. Januarius: Collectio novariensis de re Eutychis 2.161.29, in ACOec. 2/2/1:20; PCBE 1:593,
s.n. Ianuarius 28. Perhaps also African: PCBE 1:10501, s.n. Secundinus 5.
Gesta actionis primae 555.35, in ACOec. 2/1/1:149 (Greek) and ACOec. 2/3/1:133 (Latin);
ibid., 753, in ACOec. 2/1/1:170 (Greek) and ACOec. 2/3/1:157 (Latin); PCBE 1:129130, s.n.
Aurelius 5.
Gesta actionis primae 330, in ACOec. 2/1/1:11920 (Greek) and ACOec. 2/3/1:100 (Latin); PCBE
1:11368, s.nn. Valerianus 12.
Aurelius and Restitutianus: Gesta actionis primae 3.341 and 3.343, in ACOec. 2/1/1:64 (Greek)
and ACOec. 2/3/1:39 (Latin); Concilii Chalcedonensis actio iii 1.303 and 1.305, in ACOec. 2/1/2:77;
Concilii Chalcedonensis actio iiii 1.303 and 1.305, in ACOec. 2/1/2:92; Concilii Chalcedonensis actio
vi 1.322 and 324, in ACOec. 2/1/2:138 (Greek) and ACOec. 2/3/2:148 (Latin); ibid., 9.331
and 9.338, in ACOec. 2/1/2:151 (Greek); ibid., 9.330 and 9.337, in ACOec. 2/3/2:170 (Latin);
Concilii Chalcedonensis actio xvi 9.75 and 9.168, in ACOec. 2/3/3:105 and 107 (Latin) = Concilii
Chalcedonensis actio xvii 9.75 and 9.168, in ACOec. 2/1/3:91 and 94 (Greek), where the Greek
version gives Restitutianus name as -. 
; see also Concilii Chalcedonensis actionis iii appendix,
in ACOec. 2/3/2:100 and in general PCBE 1:12930, s.n. Aurelius 5 and ibid., p. 967, s.n.
Restitutianus 3. Valerians: Canones Chalcedonenses secundum versiones Dionysii exigui, in ACOec.
2/2/2:77; PCBE 1:11368, s.nn. Valerianus 12. A Valerian was also in Constantinople in the
spring of 449: Gesta actionis primae 750, in ACOec. 2/1/1:169 (Greek) and ACOec. 2/3/1:157
(Latin).

79

Staying Roman
turned its back on the definition of orthodoxy articulated forty years
earlier at Chalcedon. No less than Africa, Gelasius reminded Sacconius,
the East too was a battleground between Christ and Antichrist.48 A
quarter-century later, Pope Hormisdas exchanged a series of letters with
Possessor of Zabi, then resident in Constantinople, in which the African
bishop correspondingly sought the guidance of his patriarch as to where
the bounds of orthodox belief lay.49
Thus it may be significant that from the early sixth century onwards
we see mainly secular Africans abroad in the eastern metropolis. The
grammaticus Priscian, for example, was from Caesarea in Mauretania but
taught Latin in Constantinople during the reign of Anastasius I (ad
491518).50 Prominent African landowners, presumably living in Constantinople, were said to have had Justinians ear in the 530s, and to have
urged the emperor to reconquer Africa with glowing descriptions of its
wealth.51 At much the same time, in 532, the Praetorian Prefect Phocas
asked John the Lydian to teach him Latin. The prefect conceded, however, that he was looking for an African teacher, for he said that they
learned to speak more elegantly than the Italians.52 A common acquaintance who happened to be present and to overhear Phocas statement
recommended one Speciosus to the prefect, who engaged the man for
a time. The implication of the story is that Speciosus was himself an
African, though John never explicitly clarifies this point.53
48
49
50

51
52

53

Florentius: Theodoret, Ep. xxii, 1:924. Sacconius: Gelasius, Ep. 9, ed. Thiel in Epistolae Romanorum pontificum, pp. 33941; see also Notitia, Proc. 7, p. 117.
Hormisdas, Epp. 31, 115, and 124, pp. 8056, 91617, and 92631.
He is called Priscianus Caesariensis grammaticus in the lemma to his Institutiones: Priscian, Institutionum grammaticarum libri xviii, ed. M. Hertz, in Grammatici latini, ed. H. Keil, 8 vols. (Leipzig,
18559), 2:1. On the association of Caesarea with Caesarea in Mauretania, see Hertz, ibid., 2:p.
vii; M. Schanz and C. Hosius, Geschichte der romischen Literatur bis zum Gesetzgebungswerk des
Kaisers Justinian, 4 vols. (Munich, 190720), 4/2:2212; and PLRE 2:905, s.n. Priscianus 2,
which is more cautious, but concedes the probability of the identification.
Zachariah of Mitylene, Chronicle 9.17, The Syriac Chronicle Known as that of Zachariah of Mitylene,
trans. F. J. Hamilton and E. W. Brooks (London, 1899), p. 262.
John Lydus, De Magistratibus Populi Romani 3.73, ed. R. Wunsch (Leipzig, 1903), p. 166: [B
]* 
%EF G[]    !  )!!  G H I)G )+.
On distinctively African Latin, see J. N. Adams, The Regional Diversification of Latin 200 bcad
600 (Cambridge, 2007), pp. 25970 and 51676.
John Lydus, De Magistratibus 3.73, pp. 1657. On the date, see PLRE 2:8812, s.n. Phocas 5, at
p. 882. I. Kajanto, Onomastic Studies in the Early Christian Inscriptions of Rome and Carthage, Acta
Instituti Romani Finlandiae 2/1 (Helsinki, 1963), pp. 657 indicates that though names ending
in osus predominated in Africa (where they had probably first come into use), by the Christian
period they had spread to other parts of the empire, including Rome. No individual named
Speciosus appears either in PCBE 1 (Africa) or PLRE 1. PLRE 2:10245, s.nn. Speciosa, and
Speciosus 12, were all active in Italy, as was PLRE 3:1181, s.n. Speciosus 2. In addition to
these individuals, sixth-century Italy boasted PCBE 2/2:21002, s.nn. Speciosus 26. With

80

Flight and communications

The pull of Constantinople and the nature of our surviving sources is
such that we would see very little of the circulation of Africans in the East
beyond the imperial capital and its suburbs if it were not for the letter
collection of Bishop Theodoret of Cyrrhus. Indeed, the African sources
tell us of only one eastbound traveller not headed for Constantinople:
Bishop Quintianus of Urusi, who fled from the Vandal persecution of
the 480s to settle along the Via Egnatia at Edessa in Macedonia.54 As we
have seen, the monks Fulgentius and Redemptus intended to travel from
Africa to Alexandria, though when they reached Sicily both the bishop
of Syracuse and an African bishop living in exile nearby admonished
them not to continue on their journey given the schism that existed at
the time between the Roman and the eastern patriarchates.55 EastWest
ecclesiastical relations still seem to have been tense when a certain Peter
(probably but not certainly an African) planned to go to Jerusalem and
asked Fulgentius, now bishop of Ruspe, for a rule of the true faith so as
to avoid being lured into heresy.56 The letters of Theodoret of Cyrrhus
thus provide us with an invaluable glimpse at the circulation of travellers
beyond the major ports, communications hubs, and power centres of
the Mediterranean. Including Maria and Caelestiacus, Theodoret wrote
thirteen surviving letters on behalf of five different refugees and their
dependants, all of whom, for one reason or another, passed through his
city in the mid fifth century (see Tables 2.12). These refugees included
bishops and lay people, four men and one woman, all of whom seem to
have followed different paths to northern Syria.
Theodorets letters allow us to trace the movements of only two African
refugees before their arrival in Cyrrhus. Both were forwarded on to
Theodoret by metropolitan bishops: Eusebius of Ankyra (the metropolis
of the central Anatolian province of Galatia) and Patriarch Juvenal of
Jerusalem (ad 41958).57 Theodoret, in turn, wrote letters introducing
these refugees both to bishops and to secular authorities in the eastern
provinces. Based on the surviving documents, however, he wrote to

54
56
57

such strong peninsular connotations to the name, the joke may in fact have been that Speciosus
was Italian.
55 V. Fulg. 89, pp. 4755.
Vict. Vit. 1.29, p. 13.
Fulgentius of Ruspe, De fide ad Petrum seu de regula fidei, ed. J. Fraipont, in CCSL 91A:71160,
esp. ibid., 1, p. 711.
Theodoret, Ep. xxiii, 1:94 (Maximian), and Epp. 523, 2:12830 (Cyprian). Azema is of the opinion that the last of these two letters were written together, and that, like the letters for Florentius
(Theodoret, Ep. xxii, 1:924) and Maximian, they were written before the composition of those
for Caelestiacus (Theodoret, Epp. 2936, 2:86100). As Azema dates the latter to ad 4434,
Theodoret, Epp. xxiii and 523 may be dated slightly earlier (Azema, Correspondence, 2:1289,
n. 2).

81

Staying Roman
Table 2.2. Africans abroad, c.439c.533: The East other than
Constantinople and Chalcedon
Name

No.

Itinerary

Anonymous1

2+

Africa East

468

merchants

Caelestiacus
wife children
slaves

Carthage Cyrrhus Syria

c.440

aristocrat
(refugee)

Cyprian

Africa (Thuburbo Maius?)

Ankyra Cyrrhus Edessa
Constantina in Osrhoene

435/57

bishop
(refugee)

Maria
handmaiden

Africa Cyrrhus
Aegae West

443/8

slaves

Maximian

439/43

Quintianus

Carthage Jerusalem
Cyrrhus ?
Urusi Edessa in Macedonia

aristocrat
(refugee)
bishop
(refugee)

Date

484/489?

Status

Unlikely to have travelled together; therefore counted as two groups.

bishops twice as often as to lay people.58 Theodorets network of personal connections extended to Antioch, Edessa, Syrian Beroea (Aleppo),
Emesa, and Tyre in Syria; to Constantina in the frontier province of
Osrhoene; probably to Ankyra in Galatia; and of course to Jerusalem.59
It is a broad geographic distribution, but a broad personal distribution
as well. In the thirteen that survive to us, Theodoret only once wrote
the same person more than a single letter seeking his help for an African
refugee. This was the sophist Aerius, apparently a native of Cyrrhus, but
the location of whose school is now unknown.60
Theodoret increased this network of personal connections by rendering assistance to African refugees. The Bishop Florentius, whom
Theodoret helped on his way to Constantinople, is presumably the same
Bishop Florentius to whom Theodoret himself appealed for help after he
was deposed from his see in 449 at the Robber Council of Ephesus.61
Though we have no real way of knowing whether later support was a
condition Theodoret placed on his earlier assistance, he writes to Florentius as to one who will be sympathetic to his plight. And whether or
58
59
60

Bishops (8 people): Theodoret, Ep. xxii, 1:92 and Epp. 312, 356, 523 and 70, 2:902, 968,
12830. Laity (4 people): Theodoret, Ep. xxiii, 1:94, and Epp. 2930 and 334, 2:868 and 946.
Theodoret, Epp. 312, 356 and 523, 2:904, 96100, and 12830. Ankyra: Theodoret,
Ep. xxii, 1:92 with n. 7.
61 Theodoret, Ep. 117, 3:724.
Theodoret, Ep. xxiii, 1:94 and Ep. 30, 2:8890.

82

Flight and communications

Table 2.3. Africans abroad, c.439c.533: Rome
Name

No.

Itinerary

Date

Status

Anonymous
Donatus
Fulgentius
Redemptus
Gelasius I
Octavius
Pardulius
Restitutus
Rusticus

1
1
2

507/11
487
before 500

1
1
1
1
1

before 492
465
487
465
487

diviner
bishop
monks
(pilgrims)
Pope
bishop
bishop
bishop
bishop

Victor

Africa Rome
Africa Rome
Medidi Carthage
Syracuse Rome
Africa Rome
Africa Rome
Macomades Rome
Africa Rome
Tipasa in Numidia?
Tetcita? Rome
Africa Rome

487

bishop

not their support had any connection to Theodorets efforts on behalf

of their refugee compatriots, the bishop of Cyrrhus seems to have won
the enduring sympathies of the episcopate in Africa; for a century later
many of them staunchly refused to condemn his writings in the affair of
the Three Chapters.62
West
Rome enjoyed a status in the West similar to that of Constantinople in
the East: at least nine of our twenty-one groups of western travellers
went there (see Table 2.3).63 Here too an African ecclesiastical presence
was strongly felt. Indeed, among the nine groups of Africans who passed
through Rome, only a water-diviner (aquilex) who visited the city in
the early sixth century seems not to have been a monk or cleric.64
The other travellers all had a connection to the Church, most notably
Pope Gelasius I (ad 4926), who was, according to his biography in the
Liber Pontificalis, African by birth (natione Afer).65 In November 465, two
62

63

64
65

Interestingly Theodoret also forwarded an African refugee on to Bishop Ibbas of Edessa, whose
writings African bishops also refused to condemn in the Three Chapters Controversy: Theodoret,
Ep. 52, 2:12830.
On movement to Rome in general in late antiquity, see D. Noy, Immigrants in Late Imperial
Rome, in S. Mitchell and G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London,
2000), pp. 1530.
Cass. Var. 3.53, pp. 1378.
Liber Pontificalis 51.1, ed. Mommsen, p. 116. When addressing the emperor Anastasius I, Gelasius
himself indicates that he was born a Roman (Romanus natus: Ep. 12.1, p. 350), though this
probably only signifies that like Anastasius himself he was born in imperial territory before
it came under barbarian control. Anastasius was born at Dyrrachium (see PLRE 2:7880, s.n.
Anastastius 4, at p. 78 and sources cited there) which was captured by the Goths in 483, but
which seems to have returned to imperial control when the Goths left the Balkans for Italy:

83

Staying Roman
Table 2.4. Africans abroad, c.439c.533: The West other than Rome
Name

No. Itinerary

Date

Anonymous 1

1+

Anonymous 2

1+

Anonymous 3
Domnulus

2+
1

Africa Proconsularis Italy?

Numidia/Mauretania
Byzacena Italy? Numidia/
Mauretania
Tipasa Spain
Africa Ravenna

Eudaemon

Africa West

Eugenius
Gaudiosus
Gordian

1
1
1

Carthage Albi
Abitana Naples
Carthage Italy

Pomerius
Quintianus
Quodvultdeus
clergy
Rufinianus

1
1
3+

Africa Arles
Africa Rodez
Carthage Naples

Byzacena island off Sicily

Status

dispossessed
landowner
451
dispossessed
landowner
484
refugees
457/61
Quaestor of the
Sacred Palace
443/8
aristocrat
(refugee)
before 505 bishop (exile)
?
bishop
after 439
senator
(refugee)
c.500
grammaticus
before 506 bishop
439
bishop and
clergy (exiles)
before 499 bishop (refugee)
451

African bishops named Restitutus and Octavius were present at a Roman

synod convoked by Pope Hilarius.66 Several years later, in March 487, the
African bishops Victor, Donatus, Rusticus, and Pardulius took part in
another Roman synod, this one held at the Vatican and dealing, in part,
with the problems caused by the rebaptism of African clerics as Arians in
the face of the Vandal persecution.67 Later still, after abandoning his idea
of sailing to Alexandria, the monk Fulgentius went instead to Rome,
and was there for the Ostrogothic king Theoderics visit to the city
in 500.68
Even outside Rome we are able to see more Africans abroad in Italy
than anywhere else in the western Mediterranean over the course of
the Vandal century (see Table 2.4). Including Fulgentius and Redemptus, at least eight groups of travellers passed through Sicily or the Italian
peninsula in the period between the Vandal and Byzantine conquests.
Occasionally we cannot tell where they went. We know, for example,

66
67
68

Malchus, frags. 1 and 20, pp. 402 and 43850 (Epidamnus = Dyrrachium: see, e.g., Proc. BV
1.1.16 and 1.11.8, 1:310 and 1:362); Moorhead, Theoderic, p. 16. The city was certainly under
Byzantine control in 536, when Constantianus raised an army there: Proc. BG 1.7.278, 2:37.
Hilarius, Ep. 15.1, ed. Thiel in Epistolae Romanorum pontificum, p. 160.
Felix II (III), Ep. 13.13, ed. Thiel in Epistolae Romanorum pontificum, pp. 25961.
V. Fulg. 9, pp. 557.

84

Flight and communications

that Fulgentius grandfather, the Carthaginian senator Gordian, removed
himself to Italy in the face of the Vandal conquest of Africa. But Fulgentius biographer does not tell us where, specifically, Gordian and his
family sat out the 440s and 450s, before two of his sons returned to Africa
upon their fathers death.69 Some went to Ravenna. As we have seen,
the Quaestor of the Sacred Palace Domnulus was an African (see above,
section 1.1). Dignitaries and landholders who lost their estates in Africa
Proconsularis and Byzacena were resettled in Roman Numidia and the
Mauretanias, where in July 451 Valentinian III granted or leased them
lands from the imperial estate.70 This in turn would seem to imply that
dispossessed African landowners had pressed their case at the western
imperial court in the decade following the Vandal conquest of Africa.
We certainly hear in Cassiodorus Variae of an African seeking to inherit
the estate of another African in Italy who had died without heirs.71
A number of Africans who travelled to Italy (but not to Rome) also
ended up in the south. After the fall of Carthage to the Vandals in 439,
Bishop Quodvultdeus and his clergy found themselves in Naples.72 They
were perhaps joined there by Bishop Gaudiosus, whose see, according
to a tenth-century hagiographical text, was the African town of Abitina.
According to the Roman martyrology, Gaudiosus came from Africa to
Naples, where he lived as a monk.73 Similarly, the African bishop who
warned Fulgentius off travelling to the East was a refugee from the Vandal persecution in Byzacena named Rufinianus who had settled in a
monastery on a small island off Sicily.74

69
70
71

72
73

74

V. Fulg. 1, p. 11. It is not clear from the Latin that these sons (or indeed any family) had
accompanied Gordian into exile.
Val. Nov. 34 (ad 451), pp. 1401.
Cass. Var. 12.9, pp. 4734. The date of the letter (ad 533/7) probably places the exchange
in Africas Byzantine period, and so the individuals involved are otherwise excluded from my
analysis of the situation in the Vandal century.
Vict. Vit. 1.15, p. 8 and PCBE 1:9479, s.n. Quodvultdeus 5, at p. 949.
Martyrologium romanum s.d. v Kal. Nov. (Oct. 28), ed. H. Delehaye, et al., AASS Propyl. Dec.
(Brussels, 1940), p. 481; CIL 10.1538; U. M. Fasola, Le catacombe di S. Gennaro a Capodimonte
(Rome, 1975), p. 157, fig. 100; F. Lanzoni, Le diocesi dItalia dalle origini al principio del secolo
VII (an. 604), Studio critico, 2 vols. (2nd edn; Faenza, 1927), 2:1094. Localization of Abitina:
A. Beschaouch, Communication sur la localisation dAbitina, la cite des cel`ebres martyrs
africains, Comptes rendus des seances de lAcademie des inscriptions et belles-lettres (1976), 25566.
See also D. Mallardo, Il calendario marmoreo di Napoli s.d. 27 Oct., Biblioteca Ephemerides
Liturgicae 18 (Rome, 1947), p. 24. This calendar, which dates to the ninth century, also commemorates the African saints Augustine (s.d. 28 Aug., p. 23), Cyprian (s.d. 14 Sept., p. 24),
Saturninus, the companion of Perpetua and Felicitas (s.d. 7 Feb., p. 21), and the Scillitan martyr
Speratus (s.d. 17 July, p. 23). See also the Gesta episcoporum Neapolitanorum 2.42, ed. G. Waitz,
MGH SRL (Hanover, 1878), p. 426.
V. Fulg. 9, p. 55.

85

Staying Roman
Gaul and Spain seem to have attracted considerably less migration
from Vandal Africa. In the late fifth century, only those citizens of Tipasa
who managed to flee their city before the Vandal persecution began
there in earnest were said to have sought refuge in Spain.75 Around the
turn of the sixth century, an African grammaticus named Pomerius then
resident in Gaul taught Caesarius of Arles philology and corresponded
with Ennodius, the future bishop of Pavia.76 In the first quarter of the
sixth century, we also hear that the city of Rodez, in the Massif Central, had an African bishop.77 Earlier, in the late fifth century, the Gallic
rhetorician Lampridius was said by his friend Sidonius Apollinaris to have
consulted astrologers, the citizens of African cities (urbium ciues Africanarum), to learn when he would die though whether through letters
or in person, whether in Africa, Gaul, or elsewhere we do not hear.78
Connections such as these render more plausible the garbled account
that Gregory of Tours gives of the banishment of Bishop Eugenius of
Carthage to Albi, presumably some time early in the reign of the Vandal
king Thrasamund.79
1.4. Travellers to Africa
Over the course of the Vandal century, individuals from across the
Mediterranean also travelled to Africa, in capacities other than as ambassadors, soldiers, and captives. Some time between 523 and 526, the Ostrogothic king Theoderic the Great appointed the count Liwirit and the vir
inlustris Ampelius to investigate charges that a supply of wheat from Spain,
intended to relieve a famine in Rome, had been diverted by unscrupulous
shippers to Africa, where they were said to have illegally sold the grain for
their own profit. Given Africas sustained economic connections to the
75
76

77
78
79

Vict. Vit. 3.29, pp. 867.

Bishops Cyprian, Firminus, and Viventius, Messianus presbyter and Stephen the Deacon, Vita S.
Caesarii episcopi Arelatensis 1.9, ed. G. Morin, in Sancti Caesarii episcopi Arelatensis Opera omnia, 2
vols. (Maredsous, 193742), 2:299300; Gennadius, De viris inlustribus 99, ed. E. C. Richardson, in
Hieronymus liber De viris inlustribus; Gennadius liber De viris inlustribus, Texte und Untersuchungen
zur Geschichte der altchristlichen Literatur 14/1 (Leipzig, 1896), p. 96; Ennodius, Ep. 2.6,
pp. 378.
Gregory of Tours, Liber vitae patrum 4.1, ed. B. Krusch, MGH SRM 1/2 (Hanover, 1885),
p. 674.
Sidonius Apollinaris, Ep. 8.11.9, 3:115.
Gregory of Tours, Historiae 2.3, in Libri Historiarum X, ed. B. Krusch and W. Levison, MGH
SRM 1/1 (2nd edn; Hanover, 1951), pp. 445 (dated to the reign of Huneric) and PCBE 1:362
5, s.n. Eugenius 2, at p. 365. Eugenius was still alive (and seemingly not in Albigensian exile)
in c.496: Gennadius, De viris inlustribus 98, pp. 956. He is said to have died in 505: Vict. Tonn.
s.a. 505.86, p. 27. See also Gregory of Tours, In gloria martyrum 57, ed. Krusch, in MGH SRM
1/2:5278.

86

Flight and communications

rest of the Mediterranean world (see below, section 2), the appearance
of these shippers in the region is not particularly surprising. However,
in this case their grain ships had also carried paying passengers, whose
fares accounted for nearly three-quarters of the shipmasters profits from
the voyage a testament to the continued movement of individuals as
well as goods from Spain to Vandal Africa.80 The inscriptions from late
antique Carthage similarly reveal a handful of non-locals who died and
were buried in the metropolis, including one who was interred in the
suburban church of Bir el-Knissia in the fifth or sixth century and was
honoured with a Greek inscription dated according to the Egyptian calendar: presumably in this case the commemorand was an Egyptian.81
One or two of the chariot-drivers whom Luxorius praises and criticizes
in his epigrams were also from Egypt, as was the beast hunter Olympius,
who performed in Carthage in the sixth century.82 These entertainers
had probably moved to Africa in search of economic opportunities, but
at least one traveller sought out the Vandal kingdom as a place of refuge.
In one of the more curious careers of the fifth century, a certain Sebastian
comes criss-crossed the Mediterranean as he successively wore out his welcome first in Ravenna, then in Constantinople, and finally in Visigothic
Barcelona before travelling to Africa where, for a time, he became an
adviser to Geiseric.83 The Vandal kingdom even saw a certain amount of
religious traffic. At some point in the mid fifth century a foreign monk
(monachus transmarinus) named John came to Hadrumetum, where he
was received by the local bishop, Felix.84 There may have been a political
dimension to Johns visit, for Felix found himself exiled as a result of
this act of hospitality; but by the early sixth century the monastery of
Hadrumetum had a reputation for receiving individuals from overseas
(de transmarinis partibus) who came to pursue a monastic vocation.
According to a certain Abbot Peter and his monks, the same was widely
80
81

82
83

84

Cass. Var. 5.35, pp. 20910; Handley, Dying on Foreign Shores, p. 11. On Liwirit and Ampelius,
see also Cass. Var. 5.39, pp. 21215.
L. Ennabli, Inscriptions de Bir el Knissia, in S. Stevens (ed.), Bir el Knissia at Carthage: A ReDiscovered Cemetery Church, Report No. 1, Journal of Roman Archaeology Supplementary Series
7 (Ann Arbor, Mich., 1993), p. 258, no. 4 = Handley, Dying on Foreign Shores, no. 364. See
also L. Ennabli, Les Inscriptions funeraires chretiennes de la basilique dite de Sainte-Monique a` Carthage,

Collection de lEcole
francaise de Rome 25 (Rome, 1975), p. 289, no. 200 and p. 345, no. 378
and L. Ennabli, Les Inscriptions funeraires chretiennes de Carthage II: la basilique de Mcidfa, Collection

de lEcole
francaise de Rome 62 (Rome, 1982), p. 277, no. 510 = Handley, Dying on Foreign
Shores, nos. 3613, all of whom are commemorated as peregrini.
AL 288, 319, and 3489, pp. 240, 258, and 2735.
Hydatius, Chronicon 89 (ad 432), 95 (ad 434), 121 (ad 444), 124 (ad 445) and 136 (ad 450),
pp. 92, 96, and 98; Vict. Vit. 1.1921, pp. 910; Marcellinus comes, Chronicon s.a. 435.2, p. 79.
See also PLRE 2:9834, s.n. Sebastianus 3.
Vict. Vit. 1.23, p. 11; Schmidt, Geschichte der Wandalen, pp. 934.

87

Staying Roman
known to be true of their own monastic community, somewhere else in
Byzacena.85 Similarly, though potentially either Byzantine or Vandal in
date, a Greek inscription from Tebessa (class. Theveste) strikingly commemorates an eastern monk (#) J#) named Eulogis who died in
that city in the sixth century.86
1.5. Patterns of movement
The overall impression left by such movements, both to and from Africa,
is of sustained contact between the major port cities and communication hubs of the Mediterranean. Carthage, Hadrumetum, Pupput,
Caesarea in Mauretania, and Tipasa all of them ports figure prominently among the places from which our travellers began their journeys.
Foreigners abroad in Africa in the Vandal period are only clearly visible in the first of these two cities, though we do not know where in
Byzacena the monastery of Abbot Peter was located. Similarly, many
of the cities where we see Africans themselves overseas in the fifth and
early sixth centuries were ports: Arles, Naples, Syracuse, Constantinople,
Chalcedon, Tyre. Rome, Ravenna and Antioch are within thirty kilometres of the Mediterranean, Jerusalem about twice that distance, and
all four were served by neighbouring seaports. Macedonian Edessa and
Ankyra were on major land routes across the eastern empire. Other destinations were more remote. Cyrrhus was separated from the Mediterranean by the Amanus range of the Taurus Mountains. Constantina,
Beroea, Emessa, and a second Edessa are situated deep in the Syrian hinterland; Rodez in the highlands of southern Gaul. If an individual were
fleeing the Vandals, he might not feel safe in a coastal harbour town, but
the Massif Central and the Persian frontier were well beyond the reach
of these particular barbarians.
Such flight clearly played an important role in shaping Africas connections to the Mediterranean world in the fifth and sixth centuries.
However, all together the movement of only somewhat fewer than half
of our groups can be more or less clearly attributed to the Vandal invasion
or to the subsequent Arian persecution. These include Eugenius and his
predecessor Quodvultdeus, as well as their fellow bishops Cyprian, Florentius, Quintianus, Rufinianus, and Sacconius; the confessors of Tipasa;
and the Carthaginian aristocrats Caelestiacus, Maximian, Gordian, and
their fellow displaced landowners. It seems likely that Maria, Eudaemon,
85
86

Concilium Carthaginense a. 525, p. 279.

Supplementum Epigraphicum Graecum, vol. 18, ed. A. G. Woodhead (Amsterdam, 1962), p. 244,
no. 777 = Handley, Dying on Foreign Shores, no. 420.

88

Flight and communications

Gaudiosus, and Possessor were refugees from the Vandals as well. Yet
a small minority of our travellers, at least, clearly moved of their own
volition, and for professional reasons. These include the water-diviner
who came to Rome in the early sixth century as well as some African
merchants in the East whose goods were confiscated as the empire and
the Vandal kingdom lurched towards war in the 460s.87 The remaining cases are all ambiguous. The decision of Fulgentius and the monk
Redemptus to travel to Egypt was putatively motivated by a desire to
emulate the asceticism of the Desert Fathers, though this may well be
a pious fiction.88 As for the grammatici Priscian and Pomerius and the
Latin tutor Speciosus, the quaestor and poet Domnulus, Pope Gelasius,
and numerous African bishops in most cases we do not even know
when or why they or their families moved away from Africa, let alone
the circumstances surrounding that movement. As Christian Courtois
long ago observed, the fact that Caelestiacus was able to leave Africa
with his family and slaves probably indicates that their departure was not
under the conditions of ferocious cruelty that Victor of Vita is at pains to
stress.89 The separation of Maria and her father, however, and Marias sale
into slavery seem to stand as a poignant counter-example to Courtois
observation.
The direction of these individuals movements is probably also significant. Maria, captured and sold into slavery, had no choice but to travel to
Syria. The same was true of Caelestiacus family slaves. All of the others,
however, exercised at least a degree of volition in the direction of their
flight, and their movements suggest four general trends. First, security
rather than proximity to Africa seems to have been the primary concern
of those refugees who did flee the Vandals. Our travellers could cover long
distances to reach the relative safety of imperial territory, for example,
and yet Spain saw surprisingly little immigration. Second, and perhaps
most striking, is the importance of Constantinople as a destination for
travellers from Africa over the whole course of the Vandal century. This
is in stark contrast to late Roman Ravenna, to which relatively few of
our individuals seem to have made their way, and indeed movement
to Constantinople seems even to have outstripped that to Rome. The
eastern capital, of course, was a centre of both imperial and ecclesiastical
power; and, perhaps importantly, in the fifth and sixth centuries it would
also still have been a multilingual city. Indeed, the succession of a series
of Illyrian emperors Marcian, Leo I, Anastasius, Justin, Justinian
and the inevitable rise to prominence of their family members and
87
88

See above, nn. 16 (merchants) and 64 (water-diviner).

89 Courtois, Vandales, p. 282.
V. Fulg. 8, p. 47.

89

Staying Roman
compatriots doubtless ensured that Latin, in particular, continued to
flourish in the Queen of Cities.90 This in turn would probably have
increased the comfort level of the Latin-speaking African refugees there,
and thus have increased its appeal as a destination. Their presence in Constantinople was to be of critical importance to the history of Africa and
the empire in the sixth century, for it meant that on the eve of Justinians
reconquest of the Vandal kingdom strong ties already linked the region
to the imperial capital. Third, even in the face of these developing ties
with Constantinople, Africans continued to maintain significant social
and ecclesiastical ties to Italy, and above all to the city of Rome, which
were grounded in a deep history of contact between the regions. Finally,
while circulation beyond the two poles of Italy and Constantinople was
far more limited, the movements of at least a handful of travellers suggest
some interesting social ties with southern Gaul and Egypt, which are
probably indicative of the regions economic ties in the Vandal period.
2. goods
The history of Africas economic connections to the rest of the Mediterranean in the Vandal period has been well studied archaeologically, and
except in its details our understanding of the subject has not substantially
changed in the last twenty years.91 The general overview nonetheless
deserves to be rehearsed here, in part because trade constituted one of
the most important ties that bound Africa to that broader world, and in
part because the observable patterns in the regions economic contacts
both deepen and complicate our understanding of the movements of
people and ideas in the Vandal period.
We should begin by conceding that there are important gaps in our
understanding of Africas links to the larger late antique Mediterranean
90

91

On Latin in the East in late antiquity, see C. Rapp, Hagiography and Monastic Literature between
Greek East and Latin West in Late Antiquity, in Cristianit`a doccidente e cristianit`a doriente (secoli
VIIX), 2 vols., Settimane di studio della Fondazione Centro italiano di studi sullalto medioevo
51 (Spoleto, 2004), 2:122838.
The major syntheses are C. Panella, Merci e scambi nel Mediterraneo tardoantico, in
A. Momigliano and A. Schiavone (eds.), Storia di Roma, 4 vols. (Turin, 1988), 3/2:61397;
P. Reynolds, Trade in the Western Mediterranean, ad 400700: The Ceramic Evidence, BAR International Series 604 (Oxford, 1995); J.-P. Sodini, Productions et e changes dans le monde protobyzantin (IVe VIIe s.): le cas de la ceramique, in K. Belke, F. Hild, J. Koder, and P. Soustal

(eds.), Byzanz als Raum: Zu Methoden und Inhalten der historischen Geographie des Ostlichen
Mit
telmeerraumes, Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse,
Denkschriften 283/Tabula Imperii Byzantini 7 (Vienna, 2000), pp. 181208; M. McCormick,
Origins of the European Economy: Communications and Commerce, ad 300900 (Cambridge, 2001),
pp. 1003; and C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400800
(Oxford, 2005), pp. 70820.

90

Flight and communications

economy. Pottery kilns in Africa Proconsularis and Byzacena manufactured a fine, red tableware known as African red slip ware or ARS, and
North Africa was also particularly well known as a major producer of
olive oil.92 When this oil was shipped in ceramic amphorae, both it
and ARS can be traced archaeologically, and as such they provide our
best evidence for the direction and extent of Africas overseas exchange.
However, late antique textual sources indicate that the North African
provinces were also important exporters of grain, clothes, and textiles,
and furthermore shipped sponges, figs, cumin, salt, wild animals, and
wood abroad.93 Mauretania was a major source of slaves.94 In addition, in the fourth century, Carthage was said to have had a remarkable
silver-working district.95 Silver has a tendency to be melted down and
reworked, and none of the other goods from this list is easily traceable in
the archaeological record. Thus, at a remove of 1,500 years, we cannot
now quantify their output or know how important they were to late
antique African prosperity. The same is true of the slave trade, though
the case of Maria shows that people including Romans continued
to be exported from Africa in the Vandal period, in this case to Syria.
Given the general easternization of Mediterranean demand for human
chattel in the fifth and sixth centuries, it seems likely that the unfortunate
victims of Vandal raids on Sicily, Italy, the Balkans, and even the Atlantic
coast of Spain suffered a similar fate.96 We would do well to remember,
92

93

94
95
96

Expositio totius mundi et gentium 61, ed. J. Rouge, SC 124 (Paris, 1966), p. 200. Basic studies include
J. W. Hayes, Late Roman Pottery (London, 1972), pp. 13299 with J. W. Hayes, A Supplement
to Late Roman Pottery (London, 1980), pp. 484523; and D. J. Mattingly, Oil for Export? A
Comparison of Libyan, Spanish and Tunisian Olive Oil Production in the Roman Empire,
Journal of Roman Archaeology 1 (1988), 3356.
Edictum Diocletiani et collegarum de pretiis rerum venalium 19.24, ed. T. Mommsen in CIL 3:1942
(rugs) and ibid., 19.39, 19.42, 19.49, 19.56, and 19.61, p. 1943 (clothes); Expositio totius mundi
601, pp. 2002 (grain, clothes, animals); Fl. Vegetius Renatus, Digestorum artis mulomedicinae libri
2.13.8, 2.34.1, 2.80.3, ed. E. Lommatzsch (Leipzig, 1903), pp. 110, 129, and 171 (sponges); ibid.,
2.48.3 and 3.28.15, pp. 141 and 274 (figs); ibid., 1.42.4, p. 68 (cumin); ibid., 3.24, p. 269 (salt); and
CTh 13.5.10 (ad 364), p. 750 (wood); see also Vegetius Renatus, Digestorum artis mulomedicinae
3.14.2, p. 263 (Punic wax) and Val. Nov. 13.1 (ad 445), p. 95 (salt, alum, flax). R. M. Haywood,
Roman Africa, in T. Frank (ed.), An Economic Survey of Ancient Rome, vol. 4 (Baltimore, Md.,
1938), pp. 523 and 11618; A. H. M. Jones, The Later Roman Empire, 284602: A Social, Economic
and Administrative Survey, 2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986), 1:698 and 2:84950;
Panella, Merci e scambi, p. 631; A. Wilson, Timgad and Textile Production, in D. J. Mattingly
and J. Salmon (eds.), Economies beyond Agriculture in the Classical World, Leicester-Nottingham
Studies in Ancient Society 9 (London, 2001), pp. 27196.
Expositio totius mundi 61, p. 200.
Augustine, Confessions 6.9.14, ed. J. J. ODonnell, 3 vols. (Oxford, 1992), 1:66; Expositio totius
mundi 61, p. 202.
Captive-taking as a component of Vandal raids is specifically mentioned by Prosper, Chronicon
s.a. 455, p. 484 (Rome); Vict. Vit. 1.245, p. 12 (Rome); Malchus, frag. 5, p. 410 (Nicopolis);
Priscus, frags. 39.1, p. 342 (Sicily and Italy) and ibid., 53.1, p. 360 (Roman territory); Hydatius,

91

Staying Roman
then, that just as with the movement of refugees and other travellers,
there are important limits to what we can see of the circulation of goods
between Africa and the rest of the Mediterranean.
Even so, it is clear that by the fourth century, Africa was the
most important supplier of grain and olive oil to the city of Rome,
and ARS had achieved widespread distribution throughout the entire
Mediterranean. 97 African grain probably reached Constantinople as
well.98 By the mid fifth century (perhaps beginning slightly earlier),
that situation had changed in the East, where imports of ARS seem
to have dropped significantly.99 The later fourth and early fifth centuries had seen the emergence of a number of locally manufactured
red tablewares, produced most notably in Egypt, Cyprus, and above all
at Phocaea in the eastern Aegean (modern Foca, Turkey); in the Vandal period, it was these wares rather than ARS that were to dominate
eastern markets, the latter two in particular being widely exported in
considerable quantities.100 Imports of African ceramics seem to have
fallen off most precipitously in the greater Aegean basin (the heartland
of Phocaean red slip ware, or Late Roman C), although ARS still made
its way to the region in at least small quantities, and lamps of African
manufacture are consistently present in fifth- and sixth-century layers
on the island of Thasos.101 The same basic pattern of dwindling African

97

98

99
100
101

Chronicon 123 (ad 445), p. 96 (Gallaecia); Proc. BV 1.5.22, 1:334 (Sicily and Italy) and ibid.,
1.22.1718, 1:4067 (Peloponnese); and Vict. Tonn. s.a. 455.15, pp. 78 (Rome). Proc. BV
1.8.28, 1:350 refers in turn to Vandals being taken as captives by their Moorish enemies. Vandal
raiding is mentioned without explicit reference to slaving at Hydatius, Chronicon 77 (ad 424/5),
p. 88 (Balearics); Priscus, frag. 10, p. 242 (coastal regions) and ibid., 31.1 and 38.12, pp. 334
and 340 (Italy and Sicily); Sidonius Apollinaris, Carmen 5.38890, 1:43 (Campania); and Proc.
BV 1.5.235, 1:3345 (Balkans, Italy and Sicily) and ibid., 1.7.26, 1:344 (Roman territory).
Easternization of demand: K. Harper, The New Coordinates of the Slave Trade in Late
Antiquity: Africa, the Mediterranean, and the Near East, Paper delivered to the 124th Annual
Meeting of the American Historical Association, 9 Jan. 2010 (San Diego, Calif.) and now K.
Harper, Slavery in the Late Roman World, ad 275425 (Cambridge, 2011).
Hayes, Late Roman Pottery, p. 423; Panella, Merci e scambi, pp. 62441; and Wickham, Framing,
pp. 70811. See also J. T. Pena, The Mobilization of State Olive Oil in Roman Africa: The
Evidence of Late 4th c. Ostraca from Carthage, in J. H. Humphrey (ed.), Carthage Papers: The
Early Colonys Economy, Water Supply, a Public Bath, and the Mobilization of State Olive Oil, Journal
of Roman Archaeology Supplementary Series 28 (Portsmouth, RI, 1998), pp. 117238.
Panella, Merci e scambi, p. 638; M. McCormick, Bateaux de vie, bateaux de mort: maladie,
commerce, transports annonaires et le passage e conomique du Bas-empire au moyen a ge, in
Morfologie sociali e culturali in Europa fra tarda antichit`a e alto Medioevo, Settimane di studio del

Centro italiano di studi sullalto Medioevo 45 (Spoleto, 1998), pp. 778; Bonifay, Etudes
sur la
ceramique, p. 480.
Panella, Merci e scambi, pp. 6445 and Hayes, Late Roman Pottery, p. 423.
Panella, Merci e scambi, pp. 63940 and 6445; Hayes, Late Roman Pottery, pp. 323401 and
41723; Wickham, Framing, pp. 71415.
Sodini, Productions et e changes, p. 188; Panella, Merci e scambi, p. 645; Hayes, Late Roman
Pottery, pp. 41718; C. Abadie-Reynal and J.-P. Sodini, La Ceramique paleochretienne de Thasos

92

Flight and communications

imports holds in some parts of the Levant, though here micro-regional
diversity makes generalization more difficult, even in the interior.102 In
Egypt, African tableware continued to be imported in important quantities in Alexandria and even along the Nile River basin throughout the
Vandal period.103 The finest of these ceramics, produced at Sidi Marzouk
Tounsi in inland Byzacena (about fifty-five kilometres west of Kairouan),
may have been transported east along overland caravan routes; but in general the workshops that supplied Egypt with its African fine wares attest
to the continued importance of seaborne connections between the Nile
Delta and the region around Carthage.104 Though African amphorae are
rare in the East, these too made their way to Egypt in the fifth and sixth
centuries, in this case laden with olive oil.105 In exchange, amphorae
from the eastern Mediterranean, most of them probably carrying wine,
were imported into Carthage over the fifth and sixth centuries.106
In the West, African imports became rarer in inland Italy after the mid
fifth century and, though ARS and amphora-borne products continued
to reach major Italian cities like Rome and Naples, even there such goods
enjoyed a dwindling share of a shrinking market.107 The population
of Rome dropped precipitously in the fifth and sixth centuries, and
with it much of the citys demand for foodstuffs and tableware.108 Local
products (which were presumably cheaper) and eastern imports came to
command an ever greater share of Italian markets for both of these kinds of

102
103
104
105

106

107
108

francaise dAth`enes Etudes

thasiennes 13 (Athens, 1992),
(Aliki, Delkos, fouilles anciennes), Ecole
pp. 7983 and see also ibid., pp. 8790; L. Anselmino and C. Pavolini, Terra sigillata: lucerne,
Enciclopedia dellarte antica, classica e orientale: Atlante delle forme ceramiche, vol. 1 (Rome, 1981),
pp. 199201.
Sodini, Productions et e changes, pp. 18890; Hayes, Late Roman Pottery, pp. 41920.

Sodini, Productions et e changes, p. 190; Bonifay, Etudes

sur la ceramique, pp. 4546; Hayes, Late
Roman Pottery, pp. 4201.

Bonifay, Etudes
sur la ceramique, pp. 4546.

Panella, Merci e scambi, p. 645; Bonifay, Etudes

sur la ceramique, p. 456; Sodini, Productions
et e changes, pp. 1913. On olive oil in late antique Egypt, see Bagnall, Egypt in Late Antiquity,
pp. 301, 87, and 322.
S. Kingsley, The Economic Impact of the Palestinian Wine Trade in Late Antiquity, in
S. Kingsley and M. Decker (eds.), Economy and Exchange in the East Mediterranean during Late
Antiquity, Proceedings of a Conference at Somerville College, Oxford, 29th May, 1999 (Oxford,
2001), pp. 545 and 57; P. Pentz, From Roman Proconsularis to Islamic Ifrqiyah, GOTARC Serie
B. Gothenberg Archaeological Theses 22 (Copenhagen, 2002), p. 36; Sodini, Productions et
e changes, p. 193; C. Panella, Le anfore tardoantiche: centri di produzione e mercati preferenziali, in A. Giardina (ed.), Societ`a Romana e imperio tardoantico, III, Le merci. Gli insediamenti
(Rome, 1986), pp. 26672. See also J. A. Riley, The Pottery from the Cisterns 1977.1, 1977.2
and 1977.3, in J. H. Humphrey (ed.), Excavations at Carthage 1977, Conducted by the University of
Michigan, vol. 6 (Ann Arbor, Mich., 1981), pp. 85124, esp. 121.
Wickham, Framing, p. 711; Pentz, From Roman Proconsularis, pp. 1514; Reynolds, Trade, p. 113;
Panella, Merci e scambi, pp. 6412 and 644.

J. Durliat, De la ville antique a` la ville byzantine: le probl`eme des subsistances, Collection de lEcole
francaise de Rome 136 (Rome, 1990), pp. 11023 and fig. 1, p. 117.

93

Staying Roman
goods.109 Elsewhere the trends were slightly different. Most significantly,
agricultural exports from Africa Proconsularis to eastern Spain rose in the
Vandal period.110 As in Italy, here too ARS distribution was increasingly
limited to the coasts, but the Mediterranean littoral and the Balearics
remained well-supplied with African fine wares. Indeed, Alicante and
the Vinalopo Valley in south-eastern Spain, and Baelo Claudia (less than
twenty kilometres north-west of Tarifa) on the Atlantic appear to have
been favoured markets for ARS at different points in the fifth and sixth
centuries.111 African lamps from northern workshops similarly continued
to reach southern Spain and Ibiza.112 ARS was shipped to southern Gaul
as well, though locally produced ceramics seem to have met much of the
demand for fine tableware in this region. Imports of African fine wares
in Marseilles dropped appreciably in the second half of the fifth century;
the vessels that were traded testify for the most part to the regions
connections with production centres in Byzacena rather than the region
around Carthage. Much the same was true of amphora imports and
even lamps. However, the late fifth century did see some importation
of ceramics from Provence to the eastern African littoral, north of the
Gulf of Gab`es.113 By the second quarter of the sixth century, African
amphorae commanded a steadily increasing proportion of the Marseilles
market, though in the changeable world of the late antique economy
they now arrived (like contemporary African fine wares) from Africa
Proconsularis.114
To a considerable extent, then, the archaeological picture complements what we can see of individual movements in the written sources.
People and trade goods followed similar paths to and from Africa. This
was probably the case throughout the Mediterranean, and in this regard

109

110

111
112
113
114

Reynolds, Trade, pp. 11415; P. Arthur, Naples: Notes on the Economy of a Dark Age City, in
C. Malone and S. Stoddart (eds.), Papers in Italian Archaeology IV, The Cambridge Conference,
vol. 4, Classical and Medieval Archaeology, BAR International Series 246 (Oxford, 1985),
pp. 24759; P. Arthur, Naples: From Roman Town to City-State, Archaeological Monographs
of the British School at Rome 12 (London, 2002), p. 128; Hayes, Late Roman Pottery, p. 416.
S. J. Keay, Late Roman Amphorae in the Western Mediterranean. A Typology and Economic Study:
The Catalan Evidence, 2 vols., BAR International Series 196 (Oxford, 1984), 2:4207; Reynolds,
Trade, pp. 11314.
Reynolds, Trade, pp. 11415 and Hayes, Late Roman Pottery, p. 415.

sur la ceramique, p. 457.

Bonifay, Etudes
Ibid., pp. 452 and 457; Sodini, Productions et e changes, p. 188; Hayes, Late Roman Pottery,
pp. 4024 and 41516.
M. Bonifay, Observations sur les amphores tardives a` Marseille dapr`es les fouilles de la Bourse
(19801984), Revue archeologique de Narbonnaise 19 (1986), pp. 2978; M. Bonifay and D. Pietri,
Amphores du Ve au VIIe s. a` Marseille: nouvelles donnees sur la typologie et le contenu,

Journal of Roman Archaeology 8 (1995), p. 116; Bonifay, Etudes

sur la ceramique, p. 457.

94

Flight and communications

there are few surprises in the patterns that we observe in Gaul, the Levant, and even the Aegean. Fulgentius plans to travel to Alexandria and
the small but significant Egyptian presence in Vandal Carthage similarly
confirm what we see in the archaeological record of these two regions
economic ties. Yet there are some intriguing discrepancies in the pictures
painted by the archaeological and textual sources as well. For example,
if the Vandal-era surge in exports from Africa Proconsularis to eastern
Spain resulted in the increased circulation of individuals between these
two territories, it did not leave much of a trace in the written sources.
To be sure, Sebastian comes arrived in Carthage from Barcelona, and it is
possible that the passengers on Cassiodorus diverted Gothic grain ship
followed a similar route; but the confessors of Tipasa fled west from Mauretania Caesariensis (which enjoyed its own contacts with Spain) rather
than from Proconsularis.115 Conversely, despite the weakened economic
connections between the regions, the continued presence of Africans in
Italy and above all in Rome and Naples is a further testament to the
strength of the areas sustained social and ecclesiastical ties in the fifth
and sixth centuries. Finally, though Vandal-era ARS and amphorae are
both found in Constantinople, their relative scarcity there would probably not in its own right suggest the importance of the eastern capital as
a destination for African travellers in the Vandal period.116
3. internal connections
The circulation of both goods and travellers can also help illuminate the
related questions of how integrated Africa was as a region in the Vandal
period, and of how far into the interior Africas Mediterranean contacts
reached. Thus, for example, the movements of African ceramic exports
and of some of our trans-Mediterranean travellers both trace a thin web
of connections across the North African hinterland to the sea. Fulgentius
journey overseas to Rome began in the highlands of northern Byzacena,
from which he and his companion Redemptus are said to have travelled
to Carthage a trip of around 200 kilometres along the major Roman
roads and then to have boarded a ship to Syracuse.117 Other travellers
would appear to have made comparable treks across the African countryside, for towns like Zabi in Mauretania Sitifensis (modern Bechilga,
Algeria), Macomades in Numidia (modern Mrikeb Thala, Algeria), and
115
116
117

On ties between Mauretania and south-eastern Spain, see Reynolds, Trade, pp. 1356.

Bonifay, Etudes
sur la ceramique, p. 481; Sodini, Productions et e changes, p. 191; Hayes, Late
Roman Pottery, p. 418.
V. Fulg. 8, p. 47. For the distance, see n. to Table 2.5.

95

Staying Roman
Table 2.5. Travel from the African interior to the nearest port, c.439c.5331
Name

Date

Place of departure

Nearest port

Distance

Cyprian
Fulgentius
Redemptus
Gaudiosus
Pardulius
Possessor
Justin
Quintianus
Rusticus
Sacconius
Valerian

435/57
before 500

Thuburbo Maius? (Proc.)

Medidi (Byz.)

Pupput
Carthage2

60 km
200 km

?
487
before 517

Abitina (Proc.)
Macomedes (Num.)
Zabi (M. Sitif.)

Carthage
Rusicade
Saldae

75 km
140 km
170 km

484/489?
487
484/93
448?/451

Urusi (Proc.)
Tipasa in Numidia? (Num.)
Uzalis (Proc.)
Bassianensis = Fundus
Bassianus? (Proc.)

Hadrumetum
Hippo Regius
Utica
Hippo
Diarrhytus

110 km
90 km
20 km
25 km

Byz. = Byzacena
M. Sitif. = Mauretania Sitifensis
Num. = Numidia
Proc. = Africa Proconsularis
1
Distances are rough approximations, rounded up to the nearest 5 km, following the
Roman road networks as reconstructed in Richard J. A. Talbert (ed.), Barrington Atlas
of the Greek and Roman World (Princeton, NJ, 2000), maps 314.
2
In this instance only we are explicitly told that Carthage was the embarkation point.

Urusi in Africa Proconsularis (modern Henchir Sougda, Tunisia) are

similarly remote from their nearest ports (see Table 2.5).118 We have
already encountered the exceptionally high-quality ARS produced at
Sidi Marzouk Tounsi, which seems to have been intended primarily for
export; similarly, the pottery of the northern workshops of El Mahrine
and Oudhna were transported to Carthage and distributed from there
across the Mediterranean.119 The inclusion of cockleshell earrings in the
dowry of a certain Januarilla, a local notable from the remote Djebel
Mrata region of western Byzacena, further hints at connections direct
or indirect between the interior and the coastal zone, as we would only
expect given the extent to which the economy of that area seems to have
been geared towards the export of olive oil.120
118

119
120

In general, see above, Tables 2.12.4. Macomades and Tipasa in Numidia: PCBE 1:815, s.n.
Pardalius and ibid., p. 1015, s.n. Rusticus 9 (see also ibid., s.n. Rusticus 10), respectively. For
the identification of Bassianensis with the Fundus Bassianus, see ibid., p. 1137, s.n. Valerianus 2.

Wickham, Framing, p. 721; Bonifay, Etudes

sur la ceramique, pp. 4546.
Tablettes Albertini: actes prives de lepoque vandale (fin du Ve si`ecle), ed. C. Courtois, L. Leschi,
C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952), act 1, 1:215: quemae (sc. chamae) auriculariae.

96

Flight and communications

In general, however, the internal communications of Vandal Africa
seem to have resembled an interlocking patchwork quilt of small worlds.
Something of the imaginative landscape of one corner of the African
countryside is revealed in the remarkable collection of legal documents
to which Januarillas dowry belongs, all of them dating to the late fifth
century and known collectively as the Albertini Tablets. These deeds
reveal to us a small rural community centred on a farm known as the
Fundus Tuletianos, in the hill country of what would then have been the
pre-desert fringes of the Vandal kingdom. The land here was divided into
small plots given over to the cultivation of fruit, nut, and above all olive
trees, concentrated along wadis whose seasonal floodwaters were carefully
managed through a system of cross-bed walls, wells, sluice works, and
terraces. The agricultural regime and settlement patterns of this region
were thus similar to those revealed by field survey in the area of Cillium
and Thelepte, some seventy-five kilometres to the north-east.121
The nearest large town, Gafsa (sixty-five kilometres to the east), appears
to have been too far off to figure as a conceptual element in the local
topography, but the Fundus Tuletianos was nevertheless reasonably well
connected to the outside world. The documents mention at least four
roads, none of them leading to places that we can now identify, but
including a camel route which may perhaps hint at longer-distance
contacts.122 Yet the economic transactions that we see in these deeds
written by and for the members of this community appear to have been
fairly small-scale and local. Most of the sales that they record involve
parcels of land in Tuletianos itself. Only two acts were written up on
other nearby fundi (Magula and Gemiones), and at least one of these was
under the dominium of the same landlord as Tuletianos. Indeed, all three
properties may well have been part of the same estate.123 Only with the
121

122

123

Tablettes Albertini, passim; J. Percival, Culturae Mancianae: Field Patterns in the Albertini Tablets,
in B. Levick (ed.), The Ancient Historian and his Materials: Essays in Honour of C. E. Stevens
on his Seventieth Birthday (Farnborough, 1975), pp. 21327; D. J. Mattingly, Olive Cultivation
and the Albertini Tablets, in A. Mastino (ed.), LAfrica romana: Atti del VI convegno di studio,
Pubblicazioni del Dipartimento di storia dellUniversit`a di Sassari 14 (Sassari, 1989), pp. 40315;
and R. B. Hitchner, Historical Text and Archaeological Context in Roman North Africa:
The Albertini Tablets and the Kasserine Survey, in D. Small (ed.), Methods in the Mediterranean:
Historical and Archaeological Views on Texts and Archaeology, Mnemosyne suppl. 135 (Leiden, 1995),
pp. 12442.
Tablettes Albertini, acts 3, 9, 18, 1:219, 246, and 271 (Magula); ibid., acts 3 and 12, 1:219 and
257 (Buresa); and ibid., act 21, 1:278 (Lismul and camel route). On the significance of camel
transport in late antiquity, see R. Bulliet, The Camel and the Wheel (Cambridge, Mass., 1975);
R. Bagnall, The Camel, the Wagon and the Donkey in Later Roman Egypt, Bulletin of the
American Society of Papyrologists 22 (1985), 16; and McCormick, Origins, pp. 757.
Tablettes Albertini, act 4, 1:226 (Magula); ibid., act 29, 1:293 (Gemiones). The purchaser in both
of these acts was the same: Geminius Felix. The sellers do not seem to appear as witnesses or
neighbours in other documents, perhaps indicating that they were not residents of Tuletianos.

97

Staying Roman
sale of a young slave boy by two residents of a place called Capprariana
to one of the notables of Tuletianos do we get a real sense of interaction
between individuals from distinct communities.124
The smallness of the world revealed to us in the Albertini Tablets
is real, though our sense of the limited geographical horizons of this
community is doubtless intensified by the fact that sales of land are by
nature local in character. Nevertheless, important lines of contact also
bound Africa together, on both a micro-regional and kingdom-wide
level. Thus, for example, the political realities of Vandal rule ensured that
Carthage continued to be an important destination for both goods and
individuals in the fifth and sixth centuries. The Vandal court exercised a
draw on those ambitious for advancement, such as the proconsul Victorianus, who was originally from Hadrumetum.125 Ecclesiastical councils in
484 and 525 gathered bishops from across the African countryside in the
metropolis, and in Hunerics reign the Catholic abbot Liberatus and his
monks were similarly ordered to the royal city from their monastery near
Gafsa.126 Under normal circumstances office-holders in the civil administration drew provisions and wages (annonae et stipendia), and combined
with complaints about the Vandal kings extortionate taxation of their
Romano-African subjects, this would seem to indicate that a system for
extracting agricultural and material wealth and redistributing it to the
capital was in place down to the early sixth century.127 The presence in
Carthage of amphorae of type Keay 8B similarly suggests that olive oil
was still reaching the metropolis from southern Byzacena in the Vandal
period.128 Yet the same citys supply of fine red slip tableware was much
more local in character, produced mainly at the nearby workshops of El
Mahrine and Oudhna, and at the unidentified Atelier X.129
By and large, Africas coastal sites continued to be well supplied with
imports from up and down the Mediterranean littoral, but as little as fifteen kilometres inland the economic exchanges that we can most easily
trace took place over much shorter distances.130 In terms of ARS, for
example, the trend in Africa was towards the local production of vessels
intended primarily for distribution on a micro-regional basis. This was
124
126
127

128
129

125 Vict. Vit. 3.27, p. 85.

Tablettes Albertini, act 2, 1:217.
Vict. Vit. 2.52, p. 44; Notitia, pp. 11734; Concilium Carthaginense a. 525, pp. 25582; Passio
septem monachorum, ed. M. Petschenig in CSEL 7:10814.
Annonae et stipendia: Vict. Vit. 2.10, p. 27. Taxation and tax officials: Vict. Vit. 1.22 and 2.2,
pp. 10 and 24; AL 3367, pp. 2678; V. Fulg. 1 and 14, pp. 13 and 73; Proc. BV 1.5.15, 1.10.267,
and 2.3.26, 1:3334, 1:359, and 1:432; Liebeschuetz, Gens into Regnum, pp. 756; Wickham,
Framing, pp. 8992.

Keay, Late Roman Amphorae, 1:1269; Bonifay, Etudes

sur la ceramique, p. 451.
130 Ibid., pp. 4512.

Bonifay, Etudes
sur la ceramique, pp. 49 and 451.

98

Flight and communications

the case, for example, in the Cillium-Thelepte area, in Setif (class. Sitifis) and Belezma (class. Diana Veteranorum), around the rural atelier of
Sidi Zahruni, and in Jerba and Tripolitania.131 Fine wares from Pherdai
Maius (mod. Sidi Khalifa) similarly supplied the neighbouring Segermes
Valley.132 As in classical antiquity, it also seems likely that periodic markets continued to connect villages in the late antique countryside to
one another, to the urban economy, and to nomadic and transhumant
pastoralists. Landlords may have held markets on their estates in an effort
to strengthen their hold on their tenants, but even here itinerant outsiders
played an essential role in supplying the needs of settled communities.
Rural markets were usually located at the meeting-points of two or more
zones of complementary production, though Brent Shaw has suggested
that in general these attracted participation over an effective radius of
only about ten kilometres (rarely more than thirty).133 Thus in Byzacena
the widespread production of olive oil for the export economy may have
diminished possibilities for intraregional exchange, but even there olives
are unlikely ever to have been a true monoculture.134 About half of the
cultivators visible in the Albertini Tablets tended figs as well as olive
trees, at least two raised siteciae (apparently pistachios), and one also grew
almonds.135 Significantly, nuts and figs figure among the taxable goods
on the imperial-era customs lists from Zarai and Lambaesis, alongside
slaves, livestock, clothes, textiles, leather, glue, sponges, wine, garum (fish
sauce), dates, resin, pitch, alum, and iron.136 Strikingly similar to Africas
exports in late antiquity, these are probably at least some of the goods that
we should look for in the interior of the Vandal kingdom; but here again
131

132
133

134
135

136

E. Fentress, A. At Kaci, and N. Bounssair, Prospections dans le Belezma: Rapport Preliminaire,

in Actes du colloque international sur lhistoire de Setif, Bulletin darcheologie algerienne, Sup. 7
(Algiers, 1993), pp. 11112; L. Neuru, The Pottery of the Kasserine Survey, Antiquites africaines
26 (1990), p. 256; J. Dore, The UNESCO Libyan Valleys Archaeological Survey Pottery, in
G. W. W. Barker, D. D. Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The
UNESCO Libyan Valleys Archaeological Survey, 2 vols. (Paris and London, 1996), 2:3215; Hayes,

Late Roman Pottery, pp. 3009; Pentz, From Roman Proconsularis, pp. 1435; Bonifay, Etudes
sur
la ceramique, pp. 537; Wickham, Framing, p. 721; A. Leone, Changing Townscapes in North Africa
from Late Antiquity to the Arab Conquest, Studi storici sulla Tarda Antichit`a 28 (Bari, 2007),
pp. 1289.
Wickham, Framing, p. 721.
B. Shaw, Rural Markets in North Africa and the Political Economy of the Roman Empire,
Antiquites africaines 17 (1981), pp. 3783; see also L. de Ligt, Fairs and Markets in the Roman
Empire: Economic and Social Aspects of Periodic Trade in a Pre-Industrial Society, Dutch Monograms
on Ancient History and Archaeology 11 (Amsterdam, 1993), pp. 15596.
Wickham, Framing, pp. 6989 and 722.
Figs: Tablettes Albertini, acts 3, 4, 1012, 16, 24, and 26, 1:21827, 24958, 2689, 2823, and
288; presumably also ibid., act 6, 1:2336, a caprificus. Siteciae: ibid., acts 4 and 6, 1:2237 and
2336. Almonds: ibid., acts 4 and 19, 1:2237 and 274 (the same cultivator).
CIL 8.4508 (Zarai); R. Cagnat, A New Roman Customs List, JRS 4 (1914), pp. 1426
(Lambaesis).

99

Staying Roman
the difficulty of tracing such products archaeologically means that we do
not know to what extent they continued to circulate in late antiquity.
However, we do have a few indications that both commodities and
people could circulate over considerable distances, even in the interior.
The best-documented traveller within the territory of the Vandal kingdom was the monk Fulgentius, whom we have already encountered
on his journey from northern Byzacena to Carthage and then on to
Rome, and who is also depicted by his biographer as moving restlessly
over the North African countryside in search of spiritual perfection (see
Table 2.6). Fulgentius was eventually ordained bishop of the port city of
Ruspe in Byzacena, and thereafter many of his travels were along Africas
eastern coast, and even further afield: from Ruspe to Carthage and then
into exile in Sardinia; to Iunca, to attend a church council; to the island of
Circina, in a futile attempt at withdrawal and retirement.137 Yet Fulgentius was originally from the inland city of Thelepte, and much of his early
travel, before he became bishop, criss-crossed the African interior. His
original escape to a monastery near Praesidium Diolele, his later flight
from Moorish raids to the region of Sicca in Africa Proconsularis (mod.
El Kef, Tunisia), and his subsequent journey from Sicca to Medidi (mod.
Henchir Medded, Tunisia) all would have involved travel over land.138
The same will have been true of Fulgentius later trips from Ruspe to
Sufes and Furnos Maios, though in both of these cases the bishop may
have moved through a combination of sea and land travel.139 Commercial
goods could also traverse the African hinterland, as would seem to have
been the case, for example, with olive oil, wine, or garum: in the fifth
century, Keay 8B amphorae were imported from southern Byzacena not
just into Carthage, but also into the Cillium-Thelepte region.140 Similarly, for the most part, the kinds of wood on which the Albertini Tablets
were written mainly cedar, though also maple, willow, poplar, and
almond had to be imported into the Tuletianos region from the forests
of the Aur`es Mountains in modern-day Algeria.141 Even ARS continued

137
138
140
141

V. Fulg. 14, 1718, 201, and 268, pp. 737, 8793, 99107, and 12335.
139 V. Fulg. 27, pp. 1313.
V. Fulg. 3 and 58, pp. 213 and 3347.

Bonifay, Etudes
sur la ceramique, p. 451; Neuru, Pottery of the Kasserine Survey, p. 259; and see
above, n. 128.
M. L. Saccardy, conservateur des Eaux et Forets, chef du Service de la Conservation des Sols au
G. G. de lAlgerie in a note to the editors of the Albertini Tablets, quoted by Courtois, Tablettes
Albertini, p. 8 n. 1. The climate of the Maghrib does not seem to have changed radically between
the fifth century and today: see, e.g., D. Davis, Resurrecting the Granary of Rome: Environmental History and French Colonial Expansion in North Africa (Athens, Ohio, 2007), pp. 115 and
D. D. Gilbertson, Explanations: Environment as Agency, in Barker, Gilbertson, Jones, and
Mattingly, Farming the Desert, 1:291317.

100

Flight and communications

Table 2.6. The travels of Fulgentius of Ruspe, c.484c.532
Itinerary

Terrain

Reason for
movement

Date

Thelepte near Praesidium Diolele

(c. 3)1
Near Praesidium Diolele nearby
monastery of Felix (c. 5)
Monastery of Felix near Sicca
(cc. 56)
Near Sicca Medidi (cc. 78)
Medidi Carthage Syracuse (c. 8)
Syracuse island off Sicily Rome
(c. 9)
Rome Sardinia Medidi (c. 10)
Medidi estate of Silvestrio in
Byzacena (cc. 1011)
Byzacena monastery near Iunca
(c. 12)
Monastery near Iunca monastery
in Byzacena (c. 13)
Monastery in Byzacena Ruspe
(c. 14)
Ruspe Carthage Cagliari
(cc. 1718)
Cagliari Carthage (c. 20)
Carthage Cagliari (c. 21)

land

entered monastery

c.484

changed monasteries

land

Moorish raids

land
land sea
sea
sea land
land

Arian persecution
travel to Egypt
advised not to go to
Egypt
return to monastery
founded monastery

? sea (?)

withdrawal and rest

sea (?) ?

return to monastery

ordained bishop

508

? sea

exile to Sardinia

508/509

Cagliari Carthage Ruspe

(cc. 267)
Ruspe Iunca (c. 27)
Ruspe Sufes (c. 27)
Ruspe Furnos Maios (c. 27)
Ruspe Chilmi, island of Circina
(c. 28)
Chilimi Ruspe (c. 28)

sea
sea

recall from exile

second exile to
Sardinia
sea ?
second recall from
exile
sea?
council
sea? land council
sea? land preaching
sea
retirement
sea

c.500

516/517
518/519
523
523
after 523
c.532

recall from retirement

Chapter numbers refer to Ferrandus, Vita S. Fulgentii episcopi Ruspensis, ed. G.-G.
Lapeyre, in Vie de Saint Fulgence de Ruspe (Paris, 1929).

to make its way some distance from El Mahrine up the Mejerda river
valley and into the highlands around Dougga in the Vandal period.142
142

A. Ciotola, Il materiale ceramico rinvenuto nella ricognizione attorno a Dougga (campagne

19941996): una sintesi preliminare, in M. de Vos, Rus Africum: Terra, acqua, olio nellAfrica
settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000), pp. 63 5;
Wickham, Framing, p. 721.

101

Staying Roman
Exile within the territory of the Vandal kingdom also figures as a persistent theme in accounts of the Arian persecution of Africas Nicene
population, and in the later fifth century most exiles seem to have
been sent inland. Long before being banished to Albi, for example,
Bishop Eugenius of Carthage was sent into exile in the desert at Turris
Tamalleni.143 Under Huneric, a similar fate befell between four and five
thousand members of the Nicene clergy, who were gathered together
in Sicca and Lares in western Africa Proconsularis, and then handed
over to Moorish escorts and sent further inland to the region of the
Chott el Hodna in Mauretania Sitifensis.144 A handful of inscriptions
provide further evidence for the movement of North African bishops
into exile: over the course of the Vandal century at least three ended up
in Madauros, in the central Numidian highlands, and the son of another
died and was buried there too, seemingly in the first year of Gelimers
reign.145
The Vandal kings sought to maintain tight control over the coastal
regions of their kingdom too, including not only the African littoral but
also the Mediterranean islands, and in the later fifth and sixth centuries it
would seem that these were increasingly the areas where deportees were
sent.146 Such was the case, for example, with an anonymous bishop who
died in 495 and was celebrated in an inscription from Mouzaaville, near
Tipasa in Mauretania Caesariensis, as having been tested through many
exiles and found a worthy champion of the Catholic faith.147 Much the
same was true of another exile commemorated 100 kilometres down the
coast at Cartennae (mod. Ten`es, Algeria).148 Huneric was also said to
have sent Nicene courtiers who refused to convert to Arianism to the
fields around Utica to cut sod.149 The same king exiled a number of
Nicene bishops to Corsica to hew wood, and in the early sixth century

143
144
145

146
147

148
149

Notitia, Proc. 1, p. 117; Gennadius, De viris inlustribus 98, pp. 956; Vict. Tonn. s.a. 479.50,
p. 16; Laterculus regum Wandalorum 5 (Augiensis liber), p. 458.
Vict. Vit. 2.2637, pp. 338, and Vict. Tonn. Chron. s.a. 479.50, p. 16.
Bishops: ILAlg. 1.2761 = Handley, Dying on Foreign Shores, no. 369; ILCV 1601A = ILAlg
1.2759 = Handley, Dying on Foreign Shores, no. 372 (probably Vandal rather than Byzantine in
date, as in ILCV); ILCV 1601B = ILAlg. 1.2760 = Handley, Dying on Foreign Shores, no. 370.
Bishops son: ILCV 4452 = ILAlg. 1.2758bis = Handley, Dying on Foreign Shores, no. 371; on
bishops children, see also Vict. Vit. 2.30 and 3.24, pp. 35 and 83.
Control: see, e.g., Vict. Vit. 3.2930, pp. 867; Proc. BV 1.10.256, 1.11.224, and 1.24.1,
1:359, 1:3634, and 1:410.
CIL 8.9286 = ILCV 1102 = Handley, Dying on Foreign Shores, no. 421. Whether spiritually or
physically dead in 484, the Reparatus of Tipasa mentioned in Notitia, M. Caes. 99, p. 131, is
unlikely to be this bishop (as in ILCV); see also Moderan, Notitia provinciarum, pp. 1825.
AE (1967), 212, no. 651 = Handley, Dying on Foreign Shores, no. 425.
Vict. Vit. 2.10, p. 27.

102

Flight and communications

Thrasamund deported another sixty or more bishops to Sardinia.150 The
most famous of these, again, was Fulgentius of Rupse, and during his
exile there the island became an important nexus of communications in
the western Mediterranean: as we will see, the bishop engaged in an active
exchange of letters and in some cases even books with correspondents in
Sardinia, Africa, and Italy (see below, section 4.1); a penitent was said to
have travelled to Sardinia in search of the bishops forgiveness; and we are
told that Fulgentius received visitors from Carthage who carried news
between Cagliari and the Vandal capital.151
In short, then, we see three principal directions in which the Vandal
kingdom was bound together as an entity. First, political, fiscal, economic,
and personal exchange linked the towns, villages, and farms of the African
interior to the coastal zone, above all to the metropolis of Carthage, which
continued to be both the seat of political power and a major hub of
Mediterranean communications in the fifth and sixth centuries. Second,
the maintenance of seaborne contacts ensured that the African littoral,
as well as the islands of Sardinia and Corsica, remained well connected
throughout the Vandal century. Third, the African interior appears to
have been made up of overlapping networks of small, micro-regional
communities. Archaeologically traceable goods do not generally seem
to have reached very far inland, but these micro-regions were probably
linked to one another fairly regularly over short distances, and at least
occasionally over longer ones, both by markets and by the circulation
of individuals. However, after c.500 it also becomes harder to see the far
peripheries of the Vandal kingdom a point to which I will return (see
below, Chapter 5) and even exiles were increasingly relegated to the
coastal zone and to the Mediterranean islands.
4. cultural contacts
A final way to trace Africas connections across the late antique Mediterranean is by following its cultural exchanges. This is clearly the case,
for example, with the letters and books that circulated between African
writers and their overseas social and intellectual acquaintances. However,
it is also tempting to look at the diffusion of distinctively African saints
cults and personal names; and while such forms of evidence can only
150

151

Corsica: Vict. Vit. 3.20, p. 81; see also Notitia, Proc. 911, 1315, 1718, 223, 26, 2830, 323,
3540, 42, pp. 11718, and ibid., Num. 68, p. 121; perhaps also ibid., Proc. 27, 24, 27, 41, and
446, pp. 11719, and ibid., Byz. 20, 72, and 99, pp. 124, 126, and 127, mentioned simply as
in exile. Sardinia: V. Fulg. 1718, pp. 8791, and Vict. Tonn. s.a. 497.78, p. 24 (who gives the
figure as 120); for the island as a place of exile, see also Vict. Vit. 2.23, p. 32.
V. Fulg. 1819, pp. 917.

103

Staying Roman
rarely be attributed specifically to the Vandal period, they are nevertheless revealing of the general directions of Africas sustained, long-term
contacts with the rest of the Mediterranean world.
4.1. Letters and books
The only major letter collection to survive from the Vandal period in
Africa is that of Bishop Fulgentius of Ruspe.152 Beyond Fulgentius circle of contacts, we catch only fragmentary glimpses of the social ties
that Africans maintained within the Vandal kingdom and across the
Mediterranean. In Africa itself, these include the exchange of compliments between Sigisteus comes and Parthemius presbyter, the circulation of books among the African literary elite, and a pastoral letter said
to have been sent by the exiled bishop of Carthage, Eugenius, to his
flock.153 Beyond Africa, what scant evidence we have suggests that contacts focused primarily on Italy. In the early sixth century, for example, a
community of Africans wrote to Pope Symmachus (ad 498514) asking
for a secondary relic (benedictio) of the Milanese martyrs, Sts Nazarius
and Romanus, which they were granted.154 Symmachus was also said to
have sent money and clothes every year to the Catholic bishops in exile
in the Vandal kingdom, both in Africa itself and in Sardinia.155 As we
have seen, Bishop Possessor of Zabi, living in exile in Constantinople,
exchanged letters with Pope Hormisdas, and sent at least one of his missives to Rome in the care of his deacon Justin.156 To the confusion of
modern observers, a sentence from Cassiodorus Formula of the Prefecture
of the Police and Fire Brigades of the City of Rome was included in an undated
inscription from An Beda, Algeria, hinting at further contacts between
Africa and Italy.157
152

153

154

155
156
157

S. Stevens, The Circle of Bishop Fulgentius, Traditio 38 (1982), pp. 32741; S. R. Graham,
The Dissemination of North African Christian and Intellectual Culture in Late Antiquity,
Ph.D. diss., University of California, Los Angeles (2005), pp. 3154.
Sigisteus: Epistola ad Parthemium, PLS 3:4478 and Parthemii presbyteri rescriptum ad Sigisteum,
PLS 3:448. Literary elite: see below, Chapter 3.1.1. Eugenius: Gregory of Tours, Historiae 2.3,
pp. 412.
Symmachus, Ep. 11, ed. Thiel in Epistolae Romanorum pontificum, pp. 7089 = Ennodius, Ep.
2.14, p. 68. Stevens, Circle of Bishop Fulgentius, p. 338 takes this community to be the African
bishops in exile in Sardinia; S. A. H. Kennell, Magnus Felix Ennodius: A Gentleman of the Church
(Ann Arbor, Mich., 2000), p. 169 sees it as a community in exile in Italy. On the term benedictio,
see J. McCulloh, The Cult of Relics in the Letters and Dialogues of Pope Gregory the Great:
A Lexicographical Study, Traditio 32 (1976), pp. 16974.
Liber Pontificalis 53.11, ed. Mommsen, p. 125; Stevens, Circle of Bishop Fulgentius, p. 338.
See above, n. 49.
CIL 8.2297 (An-Beda) = Cass. Var. 7.7.2, p. 268; Moorhead, Theoderic, p. 64 n. 133. See,
however, Diesner, Auswirkungen der Religionspolitik, p. 5.

104

Flight and communications

Much more dramatic are the social contacts that Fulgentius of Ruspe
himself developed and maintained with a number of members of the
highest Italo-Roman elite. The young widow Galla, for example, whom
Fulgentius consoled on the death of her husband and encouraged to
pursue the spiritual life, belonged to the prominent aristocratic family
of the Aurelii. Her father, Q. Aurelius Memmius Symmachus, had been
consul in 485, and her grandfather, husband, and father-in-law had each
held the same exalted office.158 Fulgentius referred to another of his
correspondents, Proba, as Gallas sister, though the relationship could
well be spiritual rather than familial. Either way, Proba, too, came from
a distinguished family, for her grandfather and ancestors had also been
consuls.159 The senator Theodore, whose conversion to the spiritual life
Fulgentius praises, was a member of the illustrious gens Decia and had
been a close adviser of Pope Gelasius. He had also served as Praetorian
Prefect of Italy under Theoderic and then, in 505, he too had held
the consulship.160 To judge from her name, Venantia, whom Fulgentius
wrote about the forgiveness of sins, may also have been related to the
Decii.161 A certain Stephania also exchanged letters with Fulgentius
during his Sardinian exile, though these have now been lost. Here again
we have nothing more than the name to go on; but given both its
rarity in the fifth and early sixth centuries and the elevated circles in
which Fulgentius moved, it does not seem unreasonable to suppose that
Stephania too may have been the sister of a consul and a member of
another eminent aristocratic house, the gens Anicia, which also produced
the emperor Olybrius and the philosopher Boethius.162
Fulgentius epistolary contacts thus linked him to some of the most elevated circles of Ostrogothic Italys civilian aristocracy. To these should be
added his Italian ecclesiastical correspondents, including Abbot Eugippius
of Castellum Lucullanum in Naples, as well as a certain frater Romulus and some monks who had been kindly received in Rome by the

158
159
160

161
162

Fulg. Ep. 2.32, 1:208; Gregory I, Dialogi 4.14, 3:548; PLRE 2:491, s.n. Galla 5; Stevens,
Circle of Bishop Fulgentius, p. 334.
Fulg. Ep. 2.31, 1:208; see also Fulg. Epp. 34, 1:21235. The familial relationship is accepted
by PLRE 2:907, s.n. Proba 1, but rejected by Stevens, Circle of Bishop Fulgentius, p. 334.
Fulg. Ep. 6, 1:2404; Gelasius, Ep. 41, p. 454; PLRE 2:10978, s.n. Theodorus 62, and ibid.,
p. 1324, stemma 26; Stevens, Circle of Bishop Fulgentius, p. 335; and see also J. Moorhead,
The Decii under Theoderic, Historia: Zeitschrift fur alte Geschichte 33 (1984), pp. 10715.
Fulg. Ep. 7, 1:24454; PLRE 2:1152, s.n. Venantia, and see also ibid., p. 1324, stemma 26;
Stevens, Circle of Bishop Fulgentius, p. 336.
Fulgentius of Ruspe, Contra sermonem Fastidiosi Ariani 10, ed. J. Fraipont in CCSL 91:296; PLRE
2:1028, s.n. Stephania; Stevens, Circle of Bishop Fulgentius, p. 335.

105

Staying Roman
senator Theodore.163 A number of these individuals may well have been
personally known to Fulgentius, acquaintances from his Italian sojourn
around the year 500, though the bishop had never met either Theodore
or Venantia in the flesh. Fulgentius had been urged to write to Theodore
by Romulus and the monks; Venantias conduct and Christian zeal
were recommended to the bishop of Ruspe by his disciple Junillus.164 Far
more important to the maintenance of these ties than personal acquaintance were the correspondents shared interest in both literary pursuits
and the monastic life, their common aristocratic background, and in
some cases (though probably not in others) a certain sympathy in political outlook.165
Of course, Fulgentius also maintained epistolary ties with African and
Sardinian correspondents.166 The Carthaginian deacon Ferrandus seems
to have exchanged letters with the bishop with some frequency, and
indeed it later fell to the deacon to complete at least some of the correspondence that Fulgentius left unfinished on his death.167 Ferrandus
also refers to a letter that the bishop of Ruspe had written to his colleague John of Tharsensis (an unidentified African see) but which has
subsequently been lost.168 The bishop of Ruspe is said to have written
to his coreligionists in Carthage while in exile in Sardinia, and after his
eventual return to Byzacena most of Fulgentius letters on moral and
theological questions were probably written to Africans.169
Finally, a handful of exchanges turned Fulgentius attention eastward.
We have already seen that the bishop provided Peter, who hoped to travel
to the Holy Land, with a guide to maintaining the purity of his faith
among eastern schismatics (see above, section 1.3). Earlier, while still in
exile in Sardinia, two communications from a group of Scythian monks
in Constantinople had drawn Fulgentius and his fellow bishops into the
163

164
165
166
167
168
169

Fulg. Epp. 5 and 6.1, 1:23540; Stevens, Circle of Bishop Fulgentius, p. 331. Romuluss name
suggests some intriguing possibilities: Moorhead, Decii under Theoderic, pp. 11112, but see
also PCBE 2/2:191619, s.nn. Romulus 19.
Fulg. Epp. 6.1 and 7.12, 1:240 and 245.
Stevens, Circle of Bishop Fulgentius, pp. 3319, but for Theodore (and perhaps Venantia) see
also Moorhead, Decii under Theoderic.
V. Fulg. 25, p. 119.
Fulg. Epp. 1114 and 18, 1:357444 and 2:61924; Ferrandus, Ep. 7.2, PL 67:929C; Stevens,
Circle of Bishop Fulgentius, pp. 3367.
Fulg. Ep. 13.3, 1:386.
Sardinian exile: V. Fulg. 25, p. 119. Fulgentius of Ruspe, Ad Euthymium de remissione peccatorum,
ed. J. Fraipont, CCSL 91A:649707; Fulgentius, Ad Monimum, ed. Fraipont in CCSL 91:164;
and probably Fulg. Ep. 1, 1:18997 also belong to this period. Return to Africa: Fulg. Epp. 8, 9,
10, 1:255356; and Fulgentius, Contra Fabianum, ed. Fraipont, CCSL 91A:763866. The Felix
to whom Fulgentius wrote his Liber de Trinitate ad Felicem, ed. Fraipont, CCSL 91A:63346 and
the Peter to whom he wrote De fide ad Petrum were presumably both Africans too. On the
chronology of Fulgentius works, see Fraipont, Introduction, in CCSL 91: pp. vivii.

106

Flight and communications

debate on Semi-Pelagianism that rippled through the Mediterranean in
the early sixth century. The first of these letters came after a delegation of
the Scythians had travelled to Rome to consult Pope Hormisdas on the
subject of the Incarnation; in c.520 they also sent the African exiles a statement of their views on Christology, the will, grace, and predestination,
which ended with a condemnation of the views of Pelagius, Caelestius,
and Julian of Eclanum, as well as the late fifth-century Gallic bishop
Faustus of Riez.170 After Hormisdas dismissed the Scythians as troublemakers, John, the archimandrite of their monastery in Constantinople,
appealed to Fulgentius and the African bishops once again, this time
sending a copy of what the eastern monks felt to be a misguided treatise
written by Faustus on the subject of grace. The Africans responded to
each of these letters with statements upholding Augustinian theology,
one of them Fulgentius treatise On the Truth of Predestination and Grace,
which he dedicated to John the archimandrite.171 The same series of
exchanges would seem to indicate that exile did not always sever the ties
of communication among Africans: at least, the text of the letter that
Hormisdas sent to Possessor of Zabi in Constantinople concerning the
Scythian monks and their questions about grace and free will was later
cited in one of Fulgentius letters on the same subject.172 Finally, towards
the end of his life, Fulgentius received a letter from a certain Reginus
that mentioned further theological divisions in the East, this time with
respect to the question of whether Christs body was incorruptible from
the moment of his conception, or only after the Resurrection.173
As Fulgentius interactions with the Scythian monks show, in the
fifth and sixth centuries manuscripts as well as letters circulated between
the Vandal kingdom and other parts of the Mediterranean. While in
exile in Sardinia Fulgentius similarly requested copies of some books
that he needed but that were unavailable to him from his correspondent Eugippius of Castellum Lucullanum.174 Such exchanges built on
ties that had developed during the imperial period, above all between
Africa and southern Italy. Thus, for example, while Augustine was still a
priest, Paulinus of Nola had received a copy of the African theologians
early anti-Manichaean writings from his childhood friend Alypius, then
170

171

172
173
174

Fulg. Ep. 16, 2:55162; R. H. Weaver, Divine Grace and Human Agency: A Study of the SemiPelagian Controversy, Patristic Monograph Series 15 (Macon, Ga., 1996), pp. 1812 and see also
ibid., pp. 16580.
Fulg. Epp. 15 and 17, 2:44757 and 563615, and Fulgentius, De veritate praedestinationis et
gratiae, ed. J. Fraipont, CCSL 91A:458548; V. Fulg. 25, pp. 11920; Stevens, Circle of Bishop
Fulgentius, p. 339; Weaver, Divine Grace, pp. 18096.
Hormisdas, Ep. 124.5, pp. 9301, cited in Fulg. Ep. 15.18, 2:456.
Fulg. Ep. 18, 2:61924.
Fulg. Ep. 5.12, 1:240; see also (in Africa) Fulg. Ep. 13.3, 1:3867.

107

Staying Roman
bishop of Thagaste.175 A copy of the first four works that Augustine
composed as bishop of Hippo likewise seems to have been prepared for
an interested reader around 396, though we do not know where it was
sent before eventually ending up in the Picard monastery of Corbie.176 If
the earliest extant manuscript of Augustines City of God was not written
in Africa itself, it was probably copied in the Bay of Naples in the early
fifth century.177 This was also where Eugippius edited Augustines Literal Interpretation of Genesis and perhaps also his City of God in advance of
compiling a florilegium of excerpts of the great theologians corpus around
500.178 Eugippius himself assumed that Augustines works were available
above all in Rome, and indeed Jean-Paul Bouhot has argued the bishops
library was transferred there from Hippo towards the beginning of Leo Is
pontificate (ad 44061), shortly after the entente between Geiseric and
Valentinian III.179 If so, the subsequent loss or destruction of the earliest
Augustinian manuscripts copied there is all the more poignant.180 Augustines Confessions further survive in a sixth-century manuscript, as do the
theologians commentaries on the psalms, which were copied in Spain,
and his Harmony of the Gospels, copied in the Eastern Mediterranean,
perhaps even in Constantinople.181
A handful of texts written in the Vandal era itself also seem to have
travelled abroad fairly quickly. Martianus Capellas Marriage of Philology
and Mercury, probably written in the late fifth century, had reached Italy by
the early sixth, where it was read by Boethius and probably Ennodius. By
534, the work had been so widely copied that its text had become corrupt
175

176
177

178

179

180
181

Paulinus of Nola, Epp. 34, ed. W. Hartel, CSEL 29 (1894), pp. 1324; D. Trout, Paulinus
of Nola: Life, Letters, and Poems, Transformation of the Classical Heritage 27 (Berkeley, Calif.,
1999), pp. 2026.
CLA 11.1613; W. Green, A Fourth Century Manuscript of Saint Augustine?, Revue Benedictine
69 (1959), pp. 1917.
CLA 4.491; M. Gorman, The Manuscript Traditions of St Augustines Major Works, in
V. Grossi (ed.), Atti del Congresso internazionale su S. Agostino nel XVI Centenario della Conversione,
3 vols., Studia Ephemeridis Augustinianum 24 (Rome, 1987), 1:383, repr. in M. Gorman, The
Manuscript Traditions of the Works of St Augustine, Millennio Medievale 27 (Florence, 2001) as
essay 14, pp. 31546; Graham, Dissemination, pp. 1534.
Eugippius, Excerpta ex operibus s. Augustini, ed. P. Knoll, CSEL 9/1 (Vienna, 1885); M. Gorman,
Chapter Headings for St Augustines De Genesi ad litteram, Revue des etudes augustiniennes 26
(1980), pp. 88104 and M. Gorman, A Survey of the Oldest Manuscripts of St Augustines De
ciuitate dei, JThS n.s. 33 (1982), pp. 398410, repr. in M. Gorman, Manuscript Traditions as essay
2, pp. 4460 and essay 6, pp. 17890, respectively; Graham, Dissemination, p. 146.
Eugippius, Epistula ad Probam virginem, ed. Knoll, in Excerpta, pp. 12; J.-P. Bouhot, La Transmission dHippon a` Rome des uvres de saint Augustin, in D. Nebbiai-Dalla Guarda and J.-F.
Genest (eds.), Du copiste au collectionneur: melanges dhistoire des textes et des biblioth`eques en lhonneur
dAndre Vernet, Bibliologia Elementa ad Librorum Studia Pertinentia 18 (Turnhout, 1998),
pp. 2333.
Gorman, Manuscript Traditions, 1:402.
CLA 4.420a, 5.587, and 6.777; Gorman, Manuscript Traditions, 1:384.

108

Flight and communications

and required emendation.182 By c.525, an anti-Arian treatise known as
Against Varimadus, apparently written in Africa at some point in the mid
fifth century, had found its way into the hands of Caesarius of Arles, who
drew on it to compose his own Brief against the Heretics.183 Fulgentius
of Ruspe himself sent a copy of his To Monimus to Eugippius; this text,
along with others dating to Fulgentius Sardinian exile, is also included
in one of the earliest extant manuscripts of the bishops works, written
in southern Gaul in the sixth or seventh century.184 A copy of the Rule
of Faith that Fulgentius wrote to Peter similarly survives in a late sixthcentury manuscript from Italy, perhaps from Cassiodorus monastery at
Vivarium.185 The library of this same monastery apparently included
an apocalyptic work by the late fifth-century African bishop Vigilius of
Thapsa and a treatise by the contemporary or near-contemporary medical
writer Caelius Aurelianus.186 The modern editors of the late fifth-century
poet Dractonius works have detected his influence on Avitus of Vienne,
Ennodius disciple Arator, and Venantius Fortunatus, suggesting that the
Carthaginian advocates poems had spread from Africa to northern Italy
and the Rhone valley within only a few decades of their composition.187
They were certainly popular by the early seventh century. Not only were
Dracontius Satisfactio and De laudibus Dei known to the Spanish bishops
Isidore of Seville and Eugenius II of Toledo, in 630, the epitaph of Abbot
Vincentius of Leon even quoted a line from the poet.188 The Irish monk
Columbanus, who founded the monastery of Bobbio in northern Italy,
also incorporated verses from the Satisfactio and the De laudibus Dei into
one of his compositions.189 Curiously, Columbanus may also have carried
a fourth- or fifth-century African copy of the gospels of Matthew and
182

183

184
185
186

187
188

189

Martianus Capella, De nuptiis Philologiae et Mercurii, ed. J. Willis (Leipzig, 1983), p. 28 app. crit.;
D. Shanzer, A Philosophical and Literary Commentary on Martianus Capellas De Nuptiis Philologiae
et Mercurii Book 1, University of California Publications in Classical Studies 32 (Berkeley, Calif.,
1986), pp. 828; Graham, Dissemination, pp. 1445.
Pseudo-Vigilius of Thapsus, Contra Varimadum, ed. B. Schwank, in Florilegia Biblica Africana
saec. V, CCSL 90 (Turnhout, 1961), pp. 1134, with pp. viiviii and xv; Caesarius of Arles,
Breviarium adversus haereticos, ed. Morin, in Caesarii Arelatensis Opera omnia, 2:182208; Graham,
Dissemination, p. 188.
Fulg. Ep. 5.12, 1:239; CLA 1.104a, see also ibid., 104b; Graham, Dissemination, pp. 1902.
CLA 11.1614.
Cassiodorus, Institutiones 1.9.2, ed. R. A. B. Mynors (Oxford, 1937), p. 33 (Vigilius) and ibid.,
1.31.2, p. 79 (Caelius Aurelianus). On Caelius Aurelianus, see Merrills and Miles, Vandals,
p. 218.
C. Moussy, Introduction to Dracontius, uvres, 1:1001.
ILCV 1645 (Leon, ad 630) = Dracontius, Laudes Dei 1.611, ed. C. Moussy and C. Camus, in
uvres, 1:183: Omnibus hic mos est de flammis tollere flammas; Moussy, Introduction to
Dracontius, uvres, 1:1056; Graham, Dissemination, pp. 1804. On other African texts in
Spain, see ibid., pp. 17780.
Moussy, Introduction to Dracontius, uvres, 1:102; Graham, Dissemination, pp. 1657.

109

Staying Roman
Mark, though unfortunately it is impossible to know when or how this
manuscript crossed the Mediterranean.190
Like personal contacts, then, the circulation of books and letters was
particularly strong between the Vandal kingdom and Italy. Rome once
again figures prominently in this exchange, as does the Bay of Naples
and probably even Ravenna, where in the later sixth century Venantius Fortunatus may have encountered the poetry of Dracontius. Constantinople too emerges as an important node of trans-Mediterranean
intellectual interaction, as does southern Gaul (and especially the Rhone
valley), where the writings of Fulgentius, Dracontius, and the anonymous
author of the Against Varimadus all travelled fairly quickly. By contrast,
clear evidence for the spread of Vandal-era texts to Spain seems to be
somewhat later, an interesting literary echo of the difficulty in tracing
African travellers to the region in this period.
4.2. Saints cults
Since at least the seventeenth century, scholars have assumed that the
diffusion of African saints cults across the Mediterranean was a result of
the translation of the bodies of the holy dead by Catholic refugees fleeing
the Vandals and the Arian persecution.191 If this were the case, we would
have yet another kind of evidence with which to trace the movement of
Africans abroad in the Vandal period. Yet both the textual and the artistic
evidence for the diffusion of these cults suggests that they spread as the
result of a process which was older, slower, and longer than the political
history of fifth-century Africa might seem to imply.192
No contemporary accounts survive of the movement of African martyrs remains in the face of the Vandal invasion. Indeed, we have plausible
accounts of the posthumous translations of only two African saints. The
first of these was Augustine: according to Bede, the great theologians
body was moved twice, for before being brought to Pavia by the Lombard king Liutprand (ad 71244) the bishops remains had first been
190
191

192

CLA 4.465; see also ibid., 4.458 and 464, and in general on the role of Bobbio in preserving
African texts and manuscripts, Graham, Dissemination, 1607.
See, e.g., T. Ruinart, Historia persecutionis Vandalicae (Paris, 1694), pp. 5802; S. A. Morcelli,
Africa christiana, 3 vols. (Brescia, 181617), 3:241; de Rossi, Roma sotterranea, 2:2215; J. S.
Northcote and W. R. Brownlow, Rome souterraine, trans. P. Allard (2nd edn; Paris, 1874), p.
245 n. 2 and J. S. Northcote and W. R. Brownlow, Roma sotterranea (2nd edn; London, 1879),
p. 333; Lanzoni, Diocesi, 2:1101 and see in general ibid., 2:1093103; Fasola, Catacombe di S.
Gennaro, p. 158; and Y. Duval, Loca sanctorum Africae: le culte des martyrs en Afrique du IVe au VIIe

si`ecle, 2 vols., Collection de lEcole

francaise de Rome (Paris, 1982), 1:30 no. 13 (Thibiuca) and
2:72930.
Conant, Europe and the African Cult of Saints.

110

Flight and communications

removed from Hippo to Sardinia on account of the barbarians (propter
barbaros).193 As Michael McCormick has shown, however, this probably refers to the late seventh-century Arab invasion rather than to the
earlier Vandal one.194 The second, and by far the better documented,
translation was that of Cyprian from Carthage to Arles and then Lyons
under Charlemagnes aegis in 801.195 Other non-African texts also lay
claim to the bodies of African saints, but for the most part these accounts
date to the eighth century or later, and by and large they re-imagine
their subjects either as local holy men and women or as travellers who
left Africa prior to their martyrdom.196 In the central Middle Ages, the
transition from foreign to local saint could be quite rapid indeed, and so,
if the translation of African saints to the northern Mediterranean ever
in fact took place, on the whole it may have been a phenomenon of
the Byzantine period (or later) rather than the Vandal era.197 Indeed,
193
194

195

196

197

Bede, De temporum ratione 66.593, ed. C. Jones, CCSL 123B (Turnhout, 1977), p. 535.
McCormick, Origins, pp. 297 n. 41, 508, and 865 no. 100; R. Rowland, Jr. The Sojourn of
the Body of St Augustine in Sardinia, in J. Schnaubelt and F. Van Fleteren (eds.), Augustine
in Iconography: History and Legend (New York, 1999), pp. 18998. However, see also Graham,
Dissemination, pp. 7984.
Florus of Lyons, Carmina 13 (Rector magnificus) and 14 (Hac locuples), ed. E. Dummler, MGH
Poet. 2 (Berlin, 1884), pp. 5446; Florus, Martyrologium, in H. Quentin (ed.), Les Martyrologes
historiques du moyen a ge: etude sur la formation du Martyrologe romain (2nd edn; Paris, 1908), p. 348;
Ado of Vienne, Chronicon s.a. 807, ed. G. H. Pertz, in MGH Scriptores 2 (Hanover, 1829),
p. 320; Ado, Martyrologium, in Quentin, Martyrologes historiques, pp. 50714; and see also Annales
Regni Francorum s.aa. 8012, ed. G. H. Pertz and F. Kurze, MGH SRG 6 (Hanover, 1895;
repr. 1950), pp. 11617. On this translation, see McCormick, Origins, p. 891 no. 257 and ibid.,
pp. 8901 nos. 2546, pp. 8901, but also C. Courtois, Reliques carthaginoises et legende
carolingienne, Revue de lhistoire des religions 129 (1945), pp. 5783.
P. Chiesa, Un testo agiografico africano ad Aquileia: gli Acta di Gallonio e dei martiri di Timida
Regia, Analecta Bollandiana 114 (1996), pp. 24168 (Gallonius); B. de Gaiffier, S. Marcel de
Tanger ou de Leon? e volution dune legende, Analecta Bollandiana 61 (1943), pp. 11639 and
(on the date of the manuscript) B. de Gaiffier, Les Notices hispaniques dans le martyrologe
dUsuard, Analecta Bollandiana 55 (1937), p. 271 n. 2 (Marcellus of Tingi); P. Chiesa, Pellegrino
martire in urbe Bolitana e Pellegrino di Ancona: unaltra agiografia africana ad Aquileia?, Analecta
Bollandiana 116 (1998), pp. 2556 (Peregrinus); Peter the Subdeacon, Passio sanctae Restitutae,
ed. E. DAngelo in Pietro Suddiacono napoletano: Lopera agiografica (Florence, 2002), pp. 18699
(Restituta); Usuard, Martyrologium, ed. J. Dubois, in Le Martyrologe dUsuard, Subsidia Hagiographica 40 (Brussels, 1965), p. 249 (Siriacus and Paula); George the Monk, Chronicon breve
4.200.7, PG 110:716A-B, with Conant, Europe and the African Cult of Saints, p. 27 (Terentius
and Africanus); AASS Sept. 1 (Antwerp, 1746; repr. Brussels, 1970), pp. 13841 and Translatio
duodecim martyrum, ed. Waitz in MGH SRL, pp. 5746 (Twelve Brothers of Hadrumetum); J. de
Guibert, Saint Victor de Cesaree, Analecta Bollandiana 24 (1905), 25764 (Victor of Caesarea).
For a seemingly earlier instance of this phenomenon, see H. Delehaye, La Passion de S. Felix
de Thibiuca, Analecta Bollandiana 39 (1921), 24176.
Rapid transition: e.g., Victor of Caesarea was re-imagined as a Spanish saint over the course of
the eleventh century: J. Vives, El supuesto Pasionario hispanico de San Millan de la Cogolla,
Hispania Sacra 12 (1959), p. 453; Passio S. Victoris martyris Caesareae 3, ed. Bollandists, in Catalogus
codicum hagiographicorum latinorum antiquiorum saeculo XVI qui asservantur in Bibliotheca Nationali
Parisiensi, 3 vols., Subsidia hagiographica 2 (Brussels, 188993), 3:504; A. Fabrega Grau, Pasionario

111

Staying Roman
this is very much what we would expect, given that in late antiquity the
African church, like its Roman counterpart, seems to have been reluctant
to dismember or disinter the bodies of local saints, preferring instead to
rely on contact relics to spread their sacred power.198
Moreover, the patterns in the diffusion of African saints cults do not
correspond well with what we can see of the movements of refugees
from the Vandals. We are able to trace only two main routes by which
African saints cults spread into the rest of the Mediterranean, the first
leading from Carthage to Italy, and the second leading from Carthage and
Mauretania to Spain. In Europe, these cults were mainly concentrated
in Rome, Naples, southern Italy, and the Spanish Levant, especially the
south-east. Their diffusion in Gaul and the eastern Mediterranean was
considerably more limited. Indeed, most of the African cults that became
established in the Greek-speaking world may well have spread there by
way of southern Italy; those that took root in late antique and early
medieval Gaul appear to have arrived from Italy or Spain.199 To be sure,
the cult of the early third-century Carthaginian martyr Perpetua does
seem to have enjoyed an early and direct eastward diffusion, and in some
cases Africans who themselves moved to Gaul or the East came to be
venerated locally as saints, as was the case with Eugenius of Carthage
at Albi and the confessors of Tipasa at Constantinople.200 In general,
though, there is far less evidence for the direct movement of cults from
Africa to Gaul and especially to Constantinople (and the eastern Mediterranean generally) than we would expect from the traceable circulation of
individuals in the Vandal period.
Finally, though the majority of transplanted African saints cults first
appear in our sources over the course of the seventh, eighth, and ninth
centuries, the movement of at least a handful clearly preceded the Vandal
invasion of Africa. By 354, both Perpetua and the martyred third-century
bishop of Carthage, Cyprian, were commemorated in the Roman

198

199
200

hispanico (siglos VIIXI), 2 vols., Monumenta hispaniae sacra, Serie liturgica 6 (Madrid, 19535),
1:2278; de Guibert, Saint Victor de Cesaree, pp. 25760 and 2624.
H. Delehaye, Les Origines du culte des martyrs, Subsidia Hagiographica 20 (2nd edn; Brussels,
1933), pp. 503; McCulloh, Cult of Relics; Duval, Loca sanctorum, 2:549; Conant, Europe and
the African Cult of Saints, pp. 423.
Conant, Europe and the African Cult of Saints, pp. 434.
Passio s. Polyeucti, ed. B. Aube, in Polyeucte dans lhistoire (Paris, 1882), p. 77; Breviarium Syriacum
s.d. 7 March, ed. and (Greek) trans. L. Duchesne, AASS Nov. 2/1 (Brussels, 1894), p. liv;
Gregory of Tours, Historiae 2.3, pp. 445 and Gregory, In gloria martyrum 57, pp. 5278; Synaxarium ecclesiae Constantinopolitanae e codice Sirmondiano nunc Berolinensi adiectis synaxariis selectis s.d.
8 Dec., ed. H. Delehaye, AASS Nov. Propylaeum (Brussels, 1902; repr. 1985), cols. 2879 and
see above, previous n.

112

Flight and communications

ecclesiastical calendar.201 By c.400 Cyprians feast-day was also celebrated
in north-eastern Spain, and in the early fifth century he was mentioned
in one of the poems of Paulinus of Nola.202 Similarly, a fourth-century
sarcophagus from Quintanabureba in northern Spain depicts one of Perpetuas visions, and at the same time the Greek version of her passio was
known in the East.203 The later universality of Cyprian and Perpetuas
veneration in the medieval West gives a deceptive air of inevitability
to the early spread of their cults, though in late antiquity this was by
no means guaranteed. Indeed, in the Greek-speaking world, Cyprian of
Carthage was quickly conflated with the eponymous bishop of Antioch,
with the result that from the late fourth century onwards the African
martyr appears to have enjoyed no significant cult (independent of that
of Cyprian of Antioch) in the eastern Mediterranean.204 Moreover, by
c.429 a handful of other African saints cults were already on the move as
well. Thus, for example, in late fourth-century Verona, the local bishop
Zeno (himself a Mauretanian by birth) delivered a sermon to his flock
on St Arcadius of Caesarea; by c.400, the Mauretanian martyr Cassian of
Tingi was familiar to audiences in north-eastern Spain; and by the early
fifth century a group of martyrs from Utica known as the Massa Candida
may already have enjoyed a nascent cult in southern Italy.205
These early traces of foreign devotion to African saints are significant, because they suggest that the later diffusion of other cults into the
Mediterranean was simply the continuation of a process that was already
under way by the mid fourth century at the latest. Indeed, this diffusion
seems to have been as much the consequence of sustained, mundane
contact across the Mediterranean as it was the result of the escape of
individuals or groups from the horrors of war and persecution. In this
201

202

203

204

205

Depositio martyrum s.d. non. Martias, ed. R. Valentini and G. Zucchetti, in Codice Topografico
della Citt`a di Roma, 4 vols., Fonti per la Storia dItalia 81, 88, 90, and 91 (Rome, 194053), 2:18
(Perpetua) and ibid., s.d. xviii Kal. Oct., p. 26 (Cyprian).
Prudentius, Peristefanon 11.2378, ed. M. Cunningham, CCSL 126 (Turnhout, 1966), pp. 3778
(dies sollemnis); see also ibid., 13, pp. 3825 (passion) and ibid., 4.1718, p. 286; Paulinus of Nola,
Carmen 19.1413, ed. W. Hartel, CSEL 30 (Vienna, 1894), p. 123.
H. Schlunk, Zu den fruhchristlichen Sarkophagen aus der Bureba (Prov. Burgos), Madrider
Mitteilungen 6 (1965), 13966 at pp. 14566; H. Schlunk and T. Hauschild, Die Denkmaler der
fruhchristlichen und westgotischen Zeit (Mainz am Rhein, 1978), pp. 1413, and pl. 35; and Passio
sanctarum Perpetuae et Felicitatis 4.37, ed. C. van Beek (Nijmegen, 1936), pp. 1214 (Latin) and
1315 (Greek).
H. Delehaye, Cyprien dAntioche et Cyprien de Carthage, Analecta Bollandiana 39 (1921),
31432; thus, e.g., the entry for Cyprian of Carthage in F. Halkin (ed.), Bibliotheca hagiographica
graeca, 3 vols., Subsidia Hagiographica 8a (Brussels, 1957; 3rd edn; Wetteren, 1986), 1:140 refers
the reader back to that for Cyprian of Antioch, ibid., 1:13740, nos. 45261c.
Zeno of Verona, Tractatus 1.39, ed. B. Lofstedt, CCSL 22 (Turnhout, 1971), pp. 10710;
Prudentius, Peristefanon, 4.458, p. 287; Paulinus of Nola, Carmen 19.1448, p. 123.

113

Staying Roman
respect, it is doubtless significant that the veneration of African saints
was most highly concentrated in the regions closest to Africa. Farther
afield, especially in Gaul and the East, factors including divergent attitudes towards the relics of the saints, regional chauvinism, and suspicions
of African orthodoxy probably inhibited the diffusion of African saints
cults; yet a saints acceptance in Italy, and in some cases Spain, seems to
have helped allay such anxieties.206 We can only suspect that the flight of
refugees not only from the Vandals, but from the Moors, Muslims, and
possibly even the Byzantines as well must have accelerated the diffusion
of African cults in a series of periodic bursts; but such flight is unlikely
ever to have been the primary motor behind the spread of these cults.
4.3. Personal names207
Over thirty years ago, in a stimulating and important study of the struggles
between the faction of the Sardinian deacon, Symmachus, and that of his
rival, the archpriest Laurentius, for control of the Roman church in the
wake of the contested papal election of 498, P. A. B. Llewellyn suggested
that Symmachus support came from a group clearly defined but not
integrated into Roman society, the Catholic refugees from Arian Vandal
persecution in North Africa.208 Llewellyn quickly supplemented his
argument with a preliminary social analysis of the Roman clergy during
the Laurentian schism.209 Llewellyns analysis relied heavily, though by
no means exclusively, on the evidence of personal names to suggest
the African origins of many of Symmachus supporters. His conclusions
suggested both the presence of a substantial African refugee population
in late fifth-century Rome, particularly in the district of Trastevere,
and a methodology with which to approach the question of large-scale
human migration in the late antique Mediterranean. Since 1977, the
206
207

208
209

Conant, Europe and the African Cult of Saints, 3645.

I acknowledge with gratitude the many useful comments, conversations, and suggestions of
Christopher P. Jones on the use of onomastic evidence (including the bibliography cited in
this note), from which the following discussion benefited immensely. Thanks also to Michael
McCormick for his many beneficial observations. The shortcomings of the discussion are of
course my own. In addition to the material cited below, see esp. R. Bagnall, Religious Conversion and Onomastic Change in Early Byzantine Egypt, Bulletin of the American Society of Papyrologists 19 (1982), 10524 with R. Bagnall, Conversion and Onomastics: A Reply, Zeitschrift fur
Papyrologie und Epigraphik 69 (1987), 24350; L. Robert, Noms indig`enes dans lAsie-Mineure grecoromaine, Biblioth`eque archeologique et historique de lInstitut francais darcheologie dIstanbul
13 (Paris, 1963); and R. Syme, Tacitus, 2 vols. (Oxford, 1958).
P. A. B. Llewellyn, The Roman Church During the Laurentian Schism: Priests and Senators,
Church History 45 (1976), p. 418.
P. A. B. Llewellyn, The Roman Clergy During the Laurentian Schism (498506): A Preliminary
Analysis, Ancient Society 8 (1977), 24575.

114

Flight and communications

continued publication of epigraphic corpora (particularly for Rome)
and the proliferation of late antique prosopographical compendia have
revolutionized our ability to use onomastic evidence in an increasingly
systematic way. While this is not the place to revisit Llewellyns arguments
about Symmachus and his supporters, we must nevertheless ask what (if
anything) names can tell us about the circulation of individuals in late
antiquity.
In seeking to provide a preliminary answer to this question we can use
the prosopographies as a point of departure, though given these reference
tools necessarily restrictive criteria for inclusion, we must later broaden
our investigation to examine a wider social spectrum. An examination
of the first two volumes of the Prosopographie chretienne du Bas-empire
(PCBE) indicates that a handful of names attested five or more times (and
thus with some frequency) among the clergy and religious of fourthand early fifth-century Africa were far less common in a comparable
stratum of contemporary Italian society. Indeed, the African attestations
of nine of these names also account for 80 per cent or more of the
combined attestations in the PCBE volumes for both Italy and Africa (see
Table 2.7). Thus we see seven individuals named Adeodatus in Christian
Africa before the Vandal capture of Carthage, but only one in Italy; we
see sixty-four Victors in Africa before 439, but only four in the Italian
provinces; and so on.
Changing our focus to the highest secular officials in the late empire so
as to get a comparative sample across regions, three of these names (Adeodatus, Benenatus, Cyprian) are simply not attested among the late third-,
fourth-, and early fifth-century individuals included in the Prosopography
of the Later Roman Empire (PLRE). In the remaining cases, one is struck
immediately by three impressions. The first is that at this elite level too,
the nine names we have identified were also attested in Africa or among
Africans in a plurality and, in one instance, an outright majority of
cases (see Table 2.8). The next two impressions, however, both give us
pause as to the distinctive Africanness of the names in question. Indeed,
the first of these is that the PCBE notwithstanding there may, in fact,
have been a certain affinity between the Latin African and Italian namestocks. Among the secular elite, the contrast between the frequency of
the names Donatus, Maximianus, Victor, and Victorinus in Africa and
Italy is not nearly so stark as we might have thought from looking at the
PCBE alone. Finally, as a general trend, the more attestations we have of a
name, the greater its geographic distribution throughout the late Roman
Mediterranean. Among the secular elite, only one name (Cresconius) is
attested in Africa an outright majority of times. The other five names
each appear from the PLRE to have enjoyed a greater frequency in Africa
115

Staying Roman
Table 2.7. Comparison of nine names in PCBE 12 (Africa and
Italy) before the Vandal capture of Carthage, c.300439
Name

Africa

Adeodatus
Benenatus
Cresconius
Cyprian
Donatus
Maximianus
Peregrinus
Victor
Victorinus

7
7
34
7
64
12
5
64
12

(88%)1
(100%)
(97%)
(88%)1
(93%)
(92%)
(100%)
(96%)
(80%)

Italy

Total

1 (13%)1

1 (3%)
1 (13%)1
5 (7%)
1 (8%)

3 (4%)
3 (20%)

8
7
35
8
69
13
5
67
15

Table based on the material in Prosopographie chretienne du Bas-empire, vol. 1,

Prosopographie de lAfrique chretienne (303533), ed. A. Mandouze (Paris, 1982);
vol. 2, Prosopographie de lItalie chretienne (313604), ed. C. Pietri and L. Pietri
(Rome, 19992000). I have discounted individuals who cannot be more
precisely dated than by century, except for those dating to IV/V s. or earlier,
which I have counted as Roman-era. Only names that occur five or more
times in Africa are included; similarly, only names where Africans account
for 80 per cent or more of the combined attestations from Italy and Africa.
1

Percentages total more than 100 per cent because of small sample size and
rounding.

than in any other single region of the empire, but none of these names
can be said to have been uniquely African.
Continuing our comparison into the Vandal period (ad 439533), it is
difficult to speak of the movement of any of our nine names in response
to the fall of Carthage. However, the PLRE does seem to indicate that
the affinity between the Latin name-stock of Italy and the pre-Vandal
Africa already hinted at in the late third to early fifth century was reinforced in the wake of the Vandal conquest of Africa Proconsularis (see
Table 2.9). The absolute number of attestations throughout the late
Roman world drops fivefold in the later fifth and sixth centuries (from
68 to 13), and with this drop much of these names pan-Mediterranean
(even African) distribution is lost. Italy, however, now saw a handful of
high officials named Adeodatus, Benenatus, and Cyprian, and continued
to see members of the secular elite with the more widely attested names
of Maximianus and Victor, even as attestations of these same names in
the PLRE diminished or disappeared in the rest of the Mediterranean.
Returning to the PCBE, we see that this same trend seems to be
borne out among the fifth- and sixth-century Italian clergy and religious
116

Flight and communications

Table 2.8. Comparison of nine names in PLRE 12 before the Vandal
capture of Carthage, c.260c.439
Name
Africa Italy Spain Gaul W. Cp. Illyr. Syria Egypt E. Total
Majority from Africa
Cresconius
4
1
5
Plurality from Africa
Donatus
4
Maximianus
5
Peregrinus
2
Victor
5
Victorinus
4

2
2
3
3

1
1

2
1
4

1
3

2
2

2
1
1
1

2
2

9
12
4
20
15

Adeodatus Benenatus Cyprian: not attested among the late Roman secular elite, either
in Africa or abroad
Cp. = Constantinople
E. = unlocalized East
Egypt = Egypt and Cyrenaica
Illyr. = Illyricum
W. = unlocalized West
Table is based on the material in the Prosopography of the Later Roman Empire (Cambridge,
197192), vol. 1 (ad 260395) and vol. 2 (ad 395527).

Table 2.9. Comparison of nine names in PLRE 23 in the Vandal period,

c.439533
Name

Italy

Adeodatus
Benenatus
Cyprian
Maximianus
Victor

1
1
2
2
3

Constantinople

Thrace

Palestine

Egypt

East

Total

1
1
3
3
6

1
1
1

Cresconius Donatus Peregrinus Victorinus: not attested

Table based on the material in the Prosopography of the Later Roman Empire (Cambridge,
197192), vol. 2 (ad 395527) and vol. 3 (ad 527641).

as well. Considering the nine names with which we began, all are attested
in Christian Italy after the Vandal capture of Carthage even those that
the PCBE does not show us in Italy before the completion of the Vandal
conquest of Africa. Indeed, after 439 we seem to see a surge in attestations
of the name Adeodatus in Italy, as well as the first incidence in the PCBE
117

Staying Roman
Table 2.10. Comparison of nine names in
PCBE 2 (Italy) before and after the
Vandal capture of Carthage
Name

pre-Vandal
(300439)

Vandal era
(439533)

Adeodatus
Benenatus
Cresconius
Cyprian
Donatus
Maximianus
Peregrinus
Victor
Victorinus

1
1
5
1

3
3

10
3
1
4
2
3
2
7
3

Table based on the material in Prosopographie

chretienne du Bas-empire, vol. 2, Prosopographie
de lItalie chretienne (313604), ed. C. Pietri and
L. Pietri (Rome, 19992000).

of Italians named Benenatus and Peregrinus (see Table 2.10). Again, given
the highly selective nature of the prosopographies, we cannot push this
evidence too far. The consistent association of Italian names with names
attested in pre-Vandal Africa seems suggestive, but suggestive of what?
The situation calls for closer study.
Perhaps the most convenient way to cast a wider social net than that
offered by the PLRE and PCBE is to turn to the indexes of cognomina
for the various volumes of the monumental Corpus Inscriptionum Latinarum (CIL). We have seen that most of the names with which we
began Cyprian, Donatus, Maximianus, Peregrinus, Victor, Victorinus, even Cresconius seem to have enjoyed at least limited distribution
beyond Africa and Italy in the fourth and fifth centuries. The names most
likely to be productive of significant results, therefore, are Adeodatus and
Benenatus, which so far we have encountered only in Africa and Italy.
Both names are appealing test cases for other reasons as well. Iiro
Kajanto long ago commented on the particularly high proportion in
the name-stock of Latin-speaking North Africa both of good-omen
names and of names that indicated that a child was a wished-for gift.210
210

I. Kajanto, Peculiarities of Latin Nomenclature in North Africa, Philologus 108 (1964),

pp. 31011.

118

Flight and communications

Adeodatus (given by God) and Benenatus (well-born or even welldestined) both fit this description. Kajanto also notes that In Latin,
new compound names were extremely rare and sentence-names, typical
of Semitic nomenclature, were of course unknown.211 Kajanto stresses
that names like Adeodatus and Benenatus were not necessarily Latinized
versions of Punic names they could equally well be the Greek names
Theodotos and Eugenes but he also suggests that they do seem to have
been translations of names that were ultimately non-Latin in origin.212
By ad 439, of course, Carthage and its environs had been integrated
into the Roman state for the better part of six centuries, and we might
reasonably anticipate finding names of North African origin distributed
throughout the Mediterranean empire. Surprisingly, however, a search
of the indexes to the CIL would seem to indicate that this was not the
case with the name Adeodatus (or its feminine counterpart, Adeodata).
The CIL indexes reveal that the name was reasonably common in Latin
African epigraphy (sixteen instances in 28,085 numbered inscriptions,
or 0.06 per cent), and even enjoyed some frequency in Rome (five
instances in 41,434 numbered inscriptions, or 0.01 per cent) as well as in
the rest of Italy (Milan, Ravenna, Nola, Sardinia, Cisalpine Gaul: seven
instances in 31,977 inscriptions, or 0.02 per cent; see Map 2.2). But the
name does not appear in the index of cognomina to any of the CIL volumes for Spain, Gaul, Britain, Illyricum, or the Greek-speaking East (see
Table 2.11). Of course, we must treat even this result with some caution; like all indexes, those to the CIL contain some omissions, and
in any case, though vast in scope, the CIL is not comprehensive. For
example, the first ten volumes of the new series of Inscriptiones Christianae urbis Romae (ICVR n.s.) reveal no fewer than sixty-seven attestations of the names Adeodatus and Adeodata in the city of Rome itself
(from among 27,668 numbered inscriptions, or 0.24 per cent of this
total), though at least two of these duplicate inscriptions in the CIL (see
Table 2.12).213 The numbers are dramatically higher both in absolute
terms and as a proportion of the total inscriptions in the ICVR n.s. than
in the CIL. However, given both the Christian focus of the ICVR and
the consistently Christian associations of the name Adeodatus, the fact
that the name should enjoy a higher frequency in proportion to the
overall number of Christian inscriptions than in proportion to Christian
and non-Christian inscriptions together is not surprising. The absence of
211
213

212 Ibid., pp. 31112.

Ibid., p. 311.
ICVR n.s. 292 = CIL 6.37278; ICVR n.s. 6524 = CIL 6.33900. Also ICVR n.s. 27381 =
CIL 6.34728b, the latter of which is not indexed in the CIL and therefore not included in the
statistics above.

119

Staying Roman

Map 2.2.

Latin African names in Mediterranean inscriptions (CIL only)

the names Adeodatus and Adeodata from the CIL indexes to provinces
other than Italy and Africa, on the other hand, is striking.
A similar picture emerges when we consider the names Benenatus
and Benenata. Here too the indexes to the CIL seem to indicate that
the greatest proportion of inscriptions containing these names came from
120

Flight and communications

Table 2.11. Adeodatus/Adeodata: comparison between provinces (CIL only)

Inscriptions with Adeodatus/-a

Total of numbered inscriptions
Percentage of inscriptions with
Adeodatus/-a

Africa

Rome

Italy

Spain, Gaul, Britain,

Illyricum, and the East

16
28,085
0.06

5
41,434
0.01

7
31,977
0.02

0
44,426
0

Table 2.12. Adeodatus/Adeodata: Rome (ICVR n.s. only)

Rome
Inscriptions containing Adeodatus/-a
Total of numbered inscriptions
Percentage of inscriptions containing Adeodatus/-a

67
27,668
0.24

Table 2.13. Benenatus/Benenata: comparison between provinces (CIL only)

Inscriptions with Benenatus/-a

Total of numbered inscriptions
Percentage of inscriptions with
Benenatus/-a

Africa

Rome

Italy

Gaul

Spain, Britain,
Illyricum, East

16
28,085
0.06

0
41,434
0

2
31,977
>0.01

2
19,137
0.01

0
25,289
0

Africa, with two attestations each in Italy (Milan and Cremona) and Gallia
Narbonensis (Arles and Marseilles) (see Table 2.13 and Map 2.2). Again,
the names are absent from the CIL indexes to Spain, Britain, Illyricum,
and the East, but also (surprisingly) from the index of cognomina attested
in Rome itself. Though the names were, in fact, present in the Eternal
City in antiquity, they do not appear from the epigraphic evidence to
have been nearly as frequent there as Adeodatus and Adeodata: even
the ICVR n.s. indexes reference only a handful of individuals named
Benenatus and Benenata from inscriptions down to the seventh century
(see Table 2.14). The comparison between the CIL and the ICVR n.s.
is thus instructive for methodological reasons: the absence of the name
Benenatus from the index to CIL 6 (the Roman volume) proved, on
closer inspection, to be a false negative, and in the case of the name
121

Staying Roman
Table 2.14. Benenatus/Benanata: Rome
(ICVR n.s. only)
Rome
Inscriptions containing Benenatus/-a
Total of numbered inscriptions
Percentage of inscriptions containing Benenatus/-a

8
27,668
0.03

Adeodatus, the index to CIL 6 contained only about 7 per cent as many
attestations as the indexes to the ICVR. The near-total absence of these
names from the CIL indexes to the inscriptions from the other western
provinces by no means indicates the absence of the names themselves,
even from the epigraphic record. However, the CIL indexes probably do
give us at least a first-order approximation of these names comparative
frequency between regions in antiquity.
As with the cults of African saints, however, the appearance of the
names Adeodatus and Benenatus (and their feminine equivalents) in
the Italian epigraphic record did not coincide with the Vandal invasion
(see Table 2.15). Of the seven Italian inscriptions that are not from Rome
and that contain the names Adeodatus or Adeodata, only two are dated.
Both are from the sixth century: one from Cant`u (class. Canturium) in
Cisalpine Gaul (ad 525), the other from Ravenna (ad 595). From the
city of Rome, on the other hand, we have at least twelve datable inscriptions bearing the same names, and these come from both before and
after the Vandal capture of Carthage. Two are simply post-Diocletianic,
but another six date to the years between ad 366 and 408. Five more
post-date the fall of Carthage: one comes from the pontificate of Leo I
(ad 44061), two more from the late fifth century, and one from ad
538. Taken together, the epigraphic evidence seems to indicate a more
or less continuous presence in Italy of individuals named Adeodatus or
Adeodata from at least the fourth to at least the late sixth century. Much
the same is true of the names Benenatus and Benenata. Only two of the
non-African inscriptions containing these names are datable, and both
were erected in Rome before the Vandal capture of Carthage: one in 367
and the other in 425. As with the name Adeodatus, the name Benenatus
was already attested in the Italian epigraphic record in the fourth and
early fifth centuries. Moreover, in neither case does the frequency of
attestations per decade (at least as reflected in the CIL and the ICVR
n.s.) seem to indicate any correlation with the presumed timing of the
flight of refugees from the Vandal kingdom (see Figure 2.4).
122

Flight and communications

Table 2.15. Adeodatus/-a and Benenatus/-a beyond Africa:
dated inscriptions
No.

Date

Location

Reference

Rome
Rome
Rome

CIL 6.31893a
CIL 6.31898
ICVR n.s. 1934

Post-Diocletianic
Post-Diocletianic
366, 371, 374,
377, or 380
388

Rome

5
6
7
8
9
10
11

391
394, 396, or 402
395
408
44061
474
496

Rome
Rome
Rome
Rome
Rome
Rome
Rome

12
13
14

525
538
596

Cant`u
Rome
Ravenna

ICVR n.s. 3206

= 20718
ICVR n.s. 1446
ICVR n.s. 9580
ICVR n.s. 15354
ICVR n.s. 26680
ICVR n.s. 4783
ICVR n.s. 4926
CIL 6.37278
= ICVR n.s. 292
CIL 5.5683
ICVR n.s. 997
CIL 11.300

Benenatus/-a
16
17

367
425

Rome
Rome

Adeodatus/-a
1
2
3

ICVR n.s. 940

CIL 5.6278
= ICVR n.s. 3228

Number of attestations

0
3509 3709 3909 41019 4309 4509 4709 4909 51019 5309 5509 5709 5909

Date (earliest possible)

Fig. 2.4.

The names Adeodatus and Benenatus in dated Italian inscriptions,

350599 (CIL and ICVR n.s.)

123

Staying Roman
6

Number

5
4
3
2
1
0
41019 4209 4309 4409 4509 4609 4709 4809 4909 5009 51019 5209 5304

First visible

Fig. 2.5. The names Adeodatus and Benenatus, 410534 in PCBE 2 (Italy)
Figures are based on the materials in Prosopographie chretienne du Bas-empire
(Paris, 19822000), vol. 2, Prosopographie de lItalie chretienne (313604).
Individuals are listed by the earliest date they are attested in Italy; for epitaphs
this is typically their date of death.

This very observation, however, raises a problem. Though they do not

all appear in the epigraphic record, PCBE 2 nevertheless indicates that a
significant number of men named Adeodatus and at least one named
Benenatus were beginning to become bishops, archdeacons, and priests
in fifth- and sixth-century Italy. Moreover, the evidence of PCBE 2 also
suggests that, at least from the perspective of the fourth century, it was
something new for men of these names to hold such positions within
the Italian church (see Figure 2.5 and Table 2.8). For example, the first
Italian bishop named Adeodatus who is included in PCBE 2 held the
see of Nomentum (mod. Mentana), in the Roman archdiocese, some
time between ad 401/17 and 465.214 The middle decades of the fifth
century saw the second and third episcopal Adeodati in Italy, one with
his see at Velletri (again a diocese subject to Rome), the other at the
Neapolitan see of Cumae.215 In this same period we observe the first
eponymous archdeacon of the Roman church to appear in PCBE 2; also
the first Roman priest named Benenatus.216 In 525, as we have seen,
an eighty-five-year-old priest named Adeodatus died and was buried in
Cant`u in Cisalpine Gaul.217
214
215
216
217

ICVR n.s. 22985 = PCBE 2/1:17, s.n. Adeodatus 2.

Hilarus, Ep. 15.1, ed. Thiel in Epistolae Romanorum pontificum p. 160 = PCBE 2/1:18, s.nn.
Adeodatus 56.
ICVR n.s. 4783 and 4926 = PCBE 2/1:18, s.n. Adeodatus 4; PCBE 2/1:292, s.n.
Benenatus 1.
PCBE 2/1:17, s.n. Adeodatus 3 = CIL 5.5683.

124

Flight and communications

Perhaps most arresting, however, are the six Italian clerics named
Adeodatus whom we see for the first time in the 490s and 500s. All of
them were active in Rome and its environs, where (as we have seen) the
name was already well attested in the fourth century. Of these late fifthand early sixth-century Adeodati, one was the bishop of Castel di Guido
(class. Lorium, subject to Rome) and another was the bishop of Formia
(class. Formium, also in Latium), while the remaining four were priests
of the Roman church.218 The consistent presence of individuals named
Adeodatus in Rome from at least the mid fourth century onwards means
that these clerics may simply have been locals who entered the ranks
of the Roman clergy in the late fifth century. Indeed, the biography
of the African Pope Gelasius (ad 4926) in the Liber pontificalis twice
notes that the pontiff enlarged the clergy.219 This was not simply a pious
convention of late fifth-century papal biographers. The same claim would
be made with some frequency of the later seventh-century Popes (when
eastern refugees from the Islamic invasions poured into Rome),220 but
Gelasius was the first to be so commemorated by the Liber pontificalis.
By contemporary standards, Gelasius did indeed ordain an astonishing
number of bishops: sixty-seven in the course of his four-year pontificate,
or roughly seventeen per year a rate nearly twice that recorded for any
of Gelasius predecessors for over a century.221 In such a situation we
218
219
220

221

PCBE 2/1:1924, s.nn. Adeodatus 712, and the sources cited there.
Liber Pontificalis 51.2, ed. Mommsen, p. 116 (clerum ampliavit) and ibid., 51.6, p. 117 (Sub
huius episcopatu clerus crevit).
Liber Pontificalis 80.2, ed. Mommsen, p. 192 (Donus, ad 6768); ibid., 81.17, p. 193 (Agatho,
ad 67881); and ibid., 83.4, p. 204 (Benedict II, ad 6845). The Epitoma Cononiana, ed.
Mommsen in MGH Gestorum Pontificum Romanorum 1:258 makes the same claim for Gelasius successor Symmachus (ad 498514).
The Liber Pontificalis lists the ordinations of the late fourth- and fifth-century Popes as follows
(the per annum calculations are mine and have been rounded to the nearest unit):

Siricius (38499)
Anastasius I (399401)
Innocent I (40117)
Zosimus (41718)
Boniface I (41822)
Celestine I (42232)
Xystus III (43240)
Leo I (44061)
Hilarus (4618)
Simplicius (46883)
Felix II (III) (48392)
Gelasius (4926)

Bishops

(per year)

Priests

(per year)

Deacons

(per year)

32
11
54
8
36
46
52
185
22
88
31
67

2
6
3
8
9
5
7
9
3
6
3
17

31
9
30
10
13
32
28
81
25
58
28
32

2
5
2
10
3
3
4
4
4
4
3
8

16
5
12
3
3
12
12
31
6
11
5
2

1
3
1
3
1
1
2
1
1
1
1
1

125

Staying Roman
would expect to see higher numbers of men of all names appearing in
the ranks of the Roman clergy.
But the association of the name Adeodatus with Africa may suggest
an alternative explanation: Gelasius may well have enlarged the clergy
at least in part by drawing on a pool of his own fellow expatriates living in Rome.222 To be sure, we do not see most of these clerics for the
first time until the pontificate of Gelasius eventual successor Symmachus
(ad 498514); but this does not necessarily preclude their having been
Gelasian ordinations.223 Given the current state of our knowledge, however, this must remain only a suggestion. Nothing within our sources
epigraphic or textual suggests that the bishops and priests named Adeodatus whom we see in the 490s and 500s were in fact Africans, and, as
I have mentioned, the presence of Adeodati in Rome from at least the
middle of the fourth century makes such a conclusion far from certain
when based on the name evidence alone.
This preliminary examination of the onomastic data suggests four
overarching conclusions. The first of these is methodological: the prosopographies contain a wealth of onomastic evidence, but to be truly
revealing this data requires systematic scrutiny in light of the CIL and
other epigraphic compendia. But the epigraphic corpora, too, must be
supplemented with reference to the broader source base upon which
the more socially restricted prosopographies draw. Moreover, all of these
tools must be used with an awareness of their limitations, which inevitably
affect the picture that they present to us. Second, the apparent emergence
of at least the Latin names Adeodatus and Benenatus in the ranks of the
fifth- and sixth-century Italian clergy was not necessarily connected to
the flight of refugees from the Vandal kingdom. Such flight may have
been a contributing factor, but individuals bearing both of these names
had already appeared in the Italian epigraphic record at least seventy
years before the Vandal invasion of Africa, and at least eighty years before
the Vandal capture of Carthage. Third, apart from the appearance of
two individuals named Benenatus in the southern Gallic port cities of
Marseilles and Arles, the names Adeodatus and Benenatus seem from a
preliminary inspection to have been largely borne by Italians and Africans
in antiquity. An equally preliminary analysis of the name Quodvultdeus
yields the same results: apart from a single attestation in Trier (frequently
an imperial residence in the third and fourth centuries), the CIL indexes
222
223

See also Llewellyn, Roman Church, passim and Llewellyn, Roman Clergy, esp. pp. 25963.
Only one of the ecclesiastics most likely to have been Gelasian appointments was visible during
Gelasius pontificate: PCBE 2/1:19, s.n. Adeodatus 7.

126

Flight and communications

give no indication that the name entered the epigraphic record of any
region other than Italy and Africa. Given that the CIL is far from exhaustive, these results demand closer inspection through the national corpora
of inscriptions. As a preliminary finding, however, the apparently Italian and African associations of the names Adeodatus, Benenatus, and
Quodvultdeus seem to be a further indication of the generally close ties
between these two regions in Roman antiquity. We should therefore not
be surprised that nine of the twenty-one groups of Africans whose movements we can trace in the western Mediterranean between 439 and 533
passed through Rome, and that at least another seven found themselves
elsewhere in Italy or Sicily. For Africans of Proconsular and Byzacenan
origins (at least) Italy must have seemed a natural destination. Finally,
if the comparative infrequency of Latin names with particularly African
connotations in the name-stocks of the other provinces holds up under
closer inspection, the onomastic evidence may give us a hint as to the way
the different regions of the Roman empire were integrated into a single
political and cultural entity. It is no surprise that the province of Africa
was bound to the imperial centre like a spoke in a wheel. What is slightly
more surprising and, again, what bears further investigation in another
study is what seems on first inspection to be a relative lack of direct
onomastic interpenetration between the provinces, and the comparable
dearth of large-scale human migration between the various regions of
the empire that this might imply.
When taken together, the evidence for the movement of individuals,
trade goods, letters, books, saints cults, and personal names present a
strikingly coherent picture. They highlight the increasing importance of
Africans social and intellectual connections to the East in the Vandal
era, where they focused on the new centre of political and ecclesiastical
power in the late antique Mediterranean, Constantinople. However, we
also see Africans travelling or intending to travel to Macedonia, Anatolia,
Syria, Palestine, and Egypt; and a number of Egyptians also travelled to
Carthage. This movement of individuals accompanied at least a measured
circulation of oil, wine, pottery, and perhaps other, less easily traceable,
commodities between these regions. Ideas about grace, free will, and the
incorruptibility of Christs body flowed back and forth between the West
and Constantinople as well, but the cults of African saints do not seem
to have been embraced there with any great enthusiasm. Travel to the
imperial capital specifically was probably facilitated by the persistence
there of Latin as a spoken language into the sixth century, though at
least one traveller who set out for the East Fulgentius was said
127

Staying Roman
by an African acquaintance to have spoken flawless Greek.224 Perhaps
some of the others were Greek speakers as well, given that they were
predominantly bishops and senatorial aristocrats.
It is still in the West, though, that we are best able to trace the movement of Africans in the Vandal period. Despite weakening economic ties,
Africas social, cultural, and intellectual connections to Italy remained
particularly strong. Every class of evidence that we have considered suggests this was the case. It was in Italy that the books and letters of African
authors most quickly circulated; it was in Italy that we most frequently
find names that also had particularly African connotations; it was in Italy
that the cults of numerous African saints took root; it was to Italy that
three-quarters of our western travellers moved. Ties to southern Gaul
continued, and are reflected in the movement of ceramics, individuals,
books, and even the name Benenatus; but on the whole these contacts
seem to have been less important than Africas links to Italy. The connections of both southern Gaul and Italy appear to have been strongest
with the more easterly African provinces of Proconsularis, Byzacena, and
Numidia. Spain, by contrast, seems to have maintained ties with both
these regions and Mauretania, as the movement of goods, individuals,
texts, and saints cults appear to indicate.
In late antiquity, information like people followed shipping routes.
Our understanding of the importance of overseas trade to the African
economy in the Vandal period ensures that today few, if any, scholars
would agree with E. A. Thompsons assessment in 1976 that news did
not readily travel outwards from the interior of the grim kingdom of
the Vandals.225 Africa remained remarkably well integrated into the
larger Mediterranean world in the fifth and sixth centuries. Those for
whom movement away from Africa constituted a rejection of the Vandals
and their claims to legitimacy were able to spread the gospel of Vandal
brutality very widely indeed. Gregory of Tours separated from the
Vandals both in time and in space received only a jumbled account of
their reign in Africa, but he understood enough to know that the Vandals
had been Arian oppressors and that in witnessing to the faith Eugenius
of Carthage had been of comparable stature to Cyprian.226 In the East,
dispossessed African landowners were said to have had Justinians ear,
and to have convinced the emperor to reconquer their province for the
224
225

226

V. Fulg. 1, pp. 1113.

Thompson, End of Roman Spain, Part I, p. 11. On overseas trade and the African economy,
see, in addition to the material cited above, M. G. Fulford, Carthage: Overseas Trade and the
Political Economy, c.ad 400700, Reading Medieval Studies 6 (1980), 6680.
Gregory of Tours, Historiae 2.23, pp. 3945 and Gregory, In gloria martyrum 57, pp. 5278;
Cyprian is the only other African saint included in this work, here ibid., 93, p. 550.

128

Flight and communications

empire. The physical presence of the confessors of Tipasa at the imperial
palace cannot have hurt these expatriates cause. The victims of Vandal
persecution par excellence, the Tipasitans must have brought home in a
very tangible way the horrors of Vandal rule in Africa. A symbol of the
oppression of their western province, the confessors would become both
a rallying cry and a justification for the reconquest.227
However, not every African who travelled abroad shook the dust of
his natal province from his feet forever. We have no way of knowing
whether Bishop Aurelius of Hadrumetum, say, returned from the heady
environment of Constantinople and Chalcedon to shoulder once again
the responsibilities of pastoral care, or how long Bishops Victor, Donatus,
Rusticus, and Pardulius spent in Rome. On the other hand, two of
Gordians sons returned to Byzacena in the reign of Geiseric to reclaim
the family estate, and Fulgentius own sojourn in Italy was ultimately
a fleeting one. Hard evidence of the impact these journeys had on the
individuals who took them is by its very nature difficult to come by.
These difficulties are only compounded by the straitjacket of late antique
literary convention, which by and large did not allow for profound
levels of introspection. But Fulgentius, at least, maintained the ties he
established in Rome through a vigorous correspondence and literary
exchange after his return to Africa. The time Fulgentius and others
like him spent abroad must have served to reinforce the notion that they
were part of something larger, something decidedly ecumenical, that they
shared with other Christians and other Romans beyond the frontiers of
Africa at the very least in the major port cities and centres of power of
the late antique Mediterranean.
227

CJ 1.27.1.4 (ad 534), p. 77.

129

Chapter 3

THE OLD RULING CLASS UNDER

T H E VA N D A L S

The Vandal conquest did not sever Africa from the rest of the Mediterranean world; but it was also no longer obvious what it meant to be
Roman in the region after the collapse there of imperial power. Indeed,
to some contemporary observers it appeared that under the Vandal regime
Romanness itself was under attack. One of the loudest Romano-African
voices to emerge from the Vandal kingdom was raised in a passionate cry
of anger and denunciation. Writing in the late fifth century, the ecclesiastical historian Victor of Vita presents a vision of barbarian-Roman
interaction that is dark, hostile, and irredeemably violent. In a much-cited
passage, he wrote:
The few of you who love the barbarians and praise them at length to your own
condemnation, consider their name and understand their customs. Now could
they who own the very word of ferocity, cruelty, and terror be called by any
other proper name, unless they be called barbarians? With however many gifts
you warm them, with however much subservience you mollify them, they do
not know anything other than to envy Romans. And as much as it restrains their
will, they always desire to obscure the splendour and nobility of the Roman
name. They do not desire that any Roman at all should live, and where they
are known to have spared their subjects until now, they spare them to be used
as their slaves; for they have never loved any Roman.1

In Victors thought world, the Vandals were a savage people hell-bent

on the destruction of all things Roman. Victors own community the
Romano-African population of Nicene confession was the New Israel;
1

Vict. Vit. 3.62, pp. 1023: Nonnulli qui barbaros diligitis et eos in condemnationem uestram
aliquando laudatis, discutite nomen et intellegite mores. Numquid alio proprio nomine uocitari
poterant, nisi ut barbari dicerentur, ferocitatis utique, crudelitatis et terroris uocabulum possidentes? Quos quantiscumque muneribus foueris, quantiscumque delinieris obsequiis, illi aliud
nesciunt nisi inuidere Romanis. Et quantum ad eorum adtinet uoluntatem, semper cupiunt splendorem et genus Romani nominis nubulare; nec ullum Romanorum omnino desiderant uiuere,
et ubi adhuc noscuntur parcere subiectis, ad utendum seruitiis illorum parcunt; nam nullum
dilexerunt aliquando Romanum.

130

The old ruling class

and, like the Israelites of old, under the heretical Vandals, Africans were
languishing in a period of Babylonian Captivity.2
Yet even behind Victors fraught vision it is possible to see an increasing acceptance of the Vandal regime on the part of the Romano-African
majority. It was, after all, Romans love and praise of, and even service to,
the new ruling class that so deeply angered and frightened the historian.
Victors attempt to disrupt that growing accord drew on the centuries-old
Mediterranean ideology of the barbarian;3 but critically, it did so in the
context of a society where Romanness itself was undergoing a process
of redefinition. Indeed, in the wake of the Vandal conquest of Africa,
we see the emergence there of three major competing interpretations of
what it meant to be Roman, defined in terms of politics, high culture,
and religion. These three definitions existed simultaneously, and they
overlapped and informed one another in important ways; but Victor was
not completely wrong in insisting that the Vandal kings could in fact
be hostile to the Romanness of their subjects, depending on how that
Romanness was defined. This chapter will thus consider two major questions: first, the conditions under which cultural, political, and religious
accommodation was or was not possible between Romano-Africans and
Vandals; and second, the role that the redefinition of Romanness played
in these processes of reconciliation.
1. law, property, and culture
It is probably no accident that the greatest testament to the cultural
symbiosis between the Vandals and their Romano-African subjects
the poetry of Luxorius dates to the final decades of Vandal rule in
Africa. In the fifth century, the violence of the Vandal conquest deeply
shook those at the top of late Roman African society. Though certainly
shaded with rumour and hyperbole, the contemporary sources paint a
consistent picture of the horrors of war: killings, enslavement, rape, the
flight of refugees, pillage, the torching of buildings, and the extortion of
wealth through torture.4 These experiences were not quickly forgotten:
2
3

Vict. Vit. 3.6470, pp. 1037.

T. Howe, Vandalen, Barbaren, und Arianer bei Victor von Vita, Studien zur Alten Geschichte 7
(Frankfurt am Main, 2007), pp. 183217 and 30218; and in general Y. A. Dauge, Le Barbare:
recherches sur la conception romaine de la barbarie et de la civilisation, Collection Latomus 176 (Brussels,
1981).
Augustine, Ep. 228, CSEL 57:48496; Augustine, Sermones 34445, PLS 3:417840; Possidius,
Vita Augustini 28.413, pp. 14856; Capreolus of Carthage, Epistula ad synodum Ephesinum, in
ACOec. 1/2:645; Leo I, Ep. 12.8 and 12.11, PL 54:653 and 655; Quodvultdeus of Carthage,
De tempore barbarico 2.5, ed. R. Braun, in Opera Quoduultdeo Carthaginiensi episcopo tributa,
CCSL 60 (Turnhout, 1976), pp. 4768. F. M. Clover, Carthage in the Age of Augustine, in

131

Staying Roman
fifty years after the fact, Victor of Vita could still recount the atrocities
of the initial Vandal occupation.5 Yet with time reconciliation was also
possible. Critically for the long-term stability of Vandal rule in Africa, in
the later fifth and sixth centuries our sources testify to an unmistakable
and growing rapprochement between Vandals and Romano-Africans.
Culturally, as we have seen, the two groups quickly came to look very
much alike so much so that it is extremely difficult for us, a millennium
and a half on, to distinguish between them with any degree of real
certainty (see above, Chapter 1.3). But this was not just a question of
the politics of acculturation. The legal protections that the Vandal kings
afforded their Romano-African subjects, together with the continuing
prosperity of the North African countryside, ensured that elite RomanoAfricans were themselves able to maintain (at least to a degree) the lifestyle
to which the new ruling class also aspired. Changes in the fabric of urban
life developed naturally out of trends already visible in the late Roman
period; and while the capture of Carthage and the initial decades of
Vandal rule may have represented a caesura in North African cultural
and intellectual life, here too we quickly see clear signs of revival and
continuity with the classical tradition.

1.1. Rapprochement
Unlike Victor of Vita, most secular African literati seem not to have
wanted to reflect on the ruptures with the past brought about by the Vandal conquest. In concentrating on themes of stability and permanence,
though, such authors were not entirely deluding themselves. Thus, for
example, Africas cities appear to have remained central to elite social and
cultural life in the Vandal period. To be sure, as was the case elsewhere in
the Mediterranean, the African cityscape underwent a profound transformation in late antiquity. To modern observers perhaps the most striking
of the changes visible in the Vandal period is the forums loss of its civic
function. This was not universal, for the forum did survive as a focus of
urban activity at Sbetla (class. Sufetula); but across Africa the old civic
centre of the classical town was for the most part either neglected or

J. H. Humphrey (ed.), Excavations at Carthage 1976, Conducted by the University of Michigan, vol. 4
(Ann Arbor, Mich., 1978), p. 14; R. Gonzalez Salinero, Poder y conflicto religioso en el norte de Africa:
Quodvultdeus de Cartago y los vandalos, Graeco-Romanae Religionis Electa Collectio (GREC) 10
(Madrid, 2002), pp. 815; B. Ward-Perkins, The Fall of Rome and the End of Civilization (Oxford,
2005), p. 13; G. Berndt, Konflikt und Anpassung: Studien zu Migration und Ethnogenese der Vandalen,
Historische Studien 489 (Husum, 2007), pp. 1846.
Vict. Vit. 1.312, pp. 37.

132

The old ruling class

newly equipped with olive presses for the production of oil.6 Theatres
were also abandoned, and some bath complexes ceased to function and
were reused either for poor housing or for industrial purposes.7 Unoccupied parts of the city could also be given over to burials.8 Of course,
the repurposing of buildings indicates a shift in the cultural landscape
of the late antique city, as does the fact that churches increasingly came
to serve as important focuses both of monumental architecture and of
urban activity.9 Yet it bears emphasis that none of these developments
was new to the Vandal period: in Africa, all of them had their roots
in fourth-century trends.10 The reuse of derelict structures, moreover,
suggests a certain economic vitality to fifth- and sixth-century urban
life; and, indeed, in late antiquity, Africas cities were clearly centres of
economic production.11 The regions urban centres continued to sustain
relatively large, dense, and socially differentiated populations in the fifth
and sixth centuries. Many circuses, amphitheatres, and bath complexes
were kept up.12 Some formerly wealthy houses were abandoned or subdivided to accommodate larger numbers of inhabitants or altered living
arrangements, but others both in Carthage and in the towns of the
African hinterland were maintained and even refurbished.13
Africas towns and cities also seem to have enjoyed a thriving intellectual culture. Much as Isidore of Seville would later write an encomium
to his native Spain, the sixth-century African poet Florentius praised
6

7
8

10
11

12
13

P. Pentz, From Roman Proconsularis to Islamic Ifrqiyah, GOTARC Serie B. Gothenberg Archaeological Theses 22 (Copenhagen, 2002), pp. 513; C. Wickham, Framing the Early Middle Ages:
Europe and the Mediterranean, 400800 (Oxford, 2005), pp. 6367; A. Leone, Changing Townscapes
in North Africa from Late Antiquity to the Arab Conquest, Studi storici sulla Tarda Antichit`a 28 (Bari,
2007), pp. 1357 and 159.
Leone, Changing Townscapes, pp. 13741 and 159; see also above, Chapter 1.3.
S. Stevens, Sepultures tardives intra-muros a` Carthage, in P. Trousset (ed.), LAfrique du
Nord antique et medievale: monuments funeraires, institutions autochtones (Paris, 1995), pp. 20717;
A. Leone, Linumazione in spazio urbano a Cartagine tra V e VII secolo d. C, Antiquite
tardive 10 (2002), 23348; Leone, Changing Townscapes, pp. 1478 and 160.
Pentz, From Africa Proconsularis, p. 44; Wickham, Framing, pp. 6378; Leone, Changing Townscapes,
pp. 14854; see also, e.g., S. Stevens, Bir el Knissia at Carthage: A Re-Discovered Cemetery Church.
Report No. 1, Journal of Roman Archaeology Supplementary Series 7 (Ann Arbor, Mich., 1993),
esp. pp. 14 and 3034, a church whose first phase was seemingly built in the Vandal era.
See in general Pentz, From Africa Proconsularis, pp. 2975 and Leone, Changing Townscapes,
pp. 45128.
In addition to the material cited above, see, e.g., D. J. Mattingly, D. Stone, L. Stirling, and
N. Ben Lazreg, Leptiminus (Tunisia): A Producer City?, in D. J. Mattingly and J. Salmon
(eds.), Economies beyond Agriculture in the Classical World (London, 2000), pp. 6689; Pentz, From

Africa Proconsularis, pp. 489; M. Bonifay, Etudes

sur la ceramique romaine tardive dAfrique, BAR
International Series 1301 (Oxford, 2004), pp. 537.
Leone, Changing Townscapes, pp. 1401, but see also ibid., pp. 13740; see also above,
Chapter 1.3.
Leone, Changing Townscapes, pp. 1458 and 1612.

133

Staying Roman
the Vandal capital in glowing terms. Among the poets acclamations
was the epithet, Carthage in studies, Carthage in teachers richly distinguished (Carthago studiis, Carthago ornata magistris).14 Florentius was
not just whistling in the wind. The classical school survived in Africa
throughout the Vandal period, and in our sources we are able to glimpse
a remarkable number of grammarians who will have instructed their
students in the fundamental principles of both language and morality
within the robust but pliant framework of the classical tradition.15 Strikingly, less prestigious schools of letters also seem to have continued to
serve rural communities. At least, the basics of utilitarian literacy continued to be available to some of the male smallholders on the Fundus
Tuletianos, many of whom were able to draw up legal documents or witness them with an autograph signature.16 Those who moved beyond an
elementary education could pursue a medical or a legal career.17 Indeed,
Victor of Vita, who displays such an arresting fascination with torture
and its effects on the body, may himself have been medically trained.18 A
professional notariate also appears to have survived in the Vandal kingdom, and Gelimers secretary Boniface was one of the kings most trusted
courtiers.19
14
15

16

17
18
19

AL 371, l. 32, p. 288.

Dracontius, Romulea 1 titulus and 3 titulus, ed. Bouquet and Wolff, in uvres, 4 vols. (Paris,
198595), 3:134 app. crit. (Praefatio Dracontii discipuli ad grammaticum Felicianum) and 3:143
app. crit. (incipit praefatio ad Felicianum grammaticum); AL 289, p. 240; V. Fulg. 1, p. 13; and
R. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity, Transformation
of the Classical Heritage 11 (Berkeley, Calif., 1988), pp. 2834, 3423, 3468, and 360, nos.
58 (Faustus), 59 (Felicianus), 124 (Pomerius), 126 (Priscian), and 138 (Speciosus); see also ibid.,
pp. 3978 and 41517, nos. 204 (Coronatus) and 235 (Luxorius). Probably also AL 373, pp.
28990 and Pompeius, Commentum artis Donati, ed. H. Keil, in Grammatici latini, 8 vols. (Leipzig,
1867), 5:205 (si interroges verbi causa de Mauro, aut siqui me interroget iste homo cuias est?,
nostras est, id est Maurus), with Kaster, Guardians of Language, pp. 24950 and 3436, nos. 23
(Calbulus) and 125 (Pompeius). On the date of ibid., p. 250, no. 24 (Calcidius), see now G. Hays,
The Date and Identity of the Mythographer Fulgentius, Journal of Medieval Latin 13 (2003),
pp. 2414, which places Fulgentius the Mythographer in the Byzantine period. Possibly also
Kaster, Guardians of Language, pp. 3856, no. 189 (Astyagius) and A. Lurius Geminius, a doctor
and teacher of rhetoric from fourth- to sixth-century Mactar: Revue archeologique, 6th ser. 42
(1953), 180, no. 49. See also in general A. Merrills and R. Miles, The Vandals (Chichester, 2010),
pp. 21319.
J. Conant, Literacy and Private Documentation in Vandal North Africa: The Case of the
Albertini Tablets, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New Perspectives on Late
Antique North Africa (Aldershot, 2004), pp.199224.
Vict. Vit. 3.24 and 3.50, pp. 83 and 96; AL 297 and 304, pp. 2445 and 24950; and Merrills
and Miles, Vandals, pp. 21819.
D. Shanzer, Intentions and Audiences: History, Hagiography, Martyrdom, and Confession in
Victor of Vitas Historia Persecutionis, in Merrills, Vandals, Romans, and Berbers, p. 278.
Boniface: Proc. BV 2.4.33, 1:437. Notariate: Vict. Vit. 2.3, 2.41, and 3.19, pp. 25, 41, and 81; see
also AL 248 and 375, pp. 1868 and p. 291 (each perhaps Byzantine in date). On the functions
of the notarii, see G. W. Bowersock, P. Brown, and O. Grabar (eds.), Late Antiquity: A Guide to
the Postclassical World (Cambridge, Mass., 1999), pp. 61112 s.v.

134

The old ruling class

Furthermore, in the later fifth and early sixth centuries, small circles
of Romano-African aristocrats seem to have exchanged, reviewed, and
critically evaluated one anothers literary efforts.20 It was on the suggestion of his friend, the grammaticus Faustus about whom nothing else is
known that Luxorius compiled his book of epigrams.21 The poet also
corresponded with a certain Coronatus, who dedicated his own work on
the liberal arts to Luxorius.22 In his dedication, Coronatus stresses that
he respected Luxorius learning as well as his defence of the good and
his condemnation of the foolish and useless.23 Indeed, Coronatus and
Luxorius seem to have shared similar tastes. Three of Coronatus own
poems are preserved in the codex Salmasianus, which also contains Luxorius epigrams. Coronatus verses concern classical themes like Medeas
murder of her children and the Virgilian line (Aeneid 3.315), For my
part I live and lead life through all strange things (uiuo equidem uitamque
extrema per omnia duco).24 One cannot but imagine that the line must have
struck a chord for many Romano-African aristocrats. The poet Blossius
Aemilius Dracontius may have been one member of a larger literary circle as well. It probably included Martianus Capella, for in their poetry
the two authors imitate one another.25 Friedrich Vollmer attributed
the anonymous Aegritudo perdicae to the milieu of Dracontius teacher
Felicianus.26 On the evidence of apparent borrowings between the poets,
Pierre Langlois has argued convincingly that Reposianus was another
contemporary or near-contemporary of Dracontius. By extension, the
verses of the poet Regianus, one of which seems to make reference to

20

21
22
23
24
25

26

In addition to the poets discussed here, see also perhaps Citherius rhetor, the author of a poem
from a lost anthology, probably African and resembling the Codex Salmasianus (PLRE 2:298,
s.n. Citherius): Anthologia Latina 484b, ed. Riese, 1/2:9. Also the Christian poet Cresconius
(PLRE 2:329, s.n. Cresconius 3) and a number of named poets whose works are preserved in
the codex Salmasianus: Avitus (AL 16, p. 49), Bonosius (AL 274, p. 200), Lindinus (AL 15, p.
48), Octavian (AL 7, p. 31), Ponnanus (AL 268, pp. 1967), Symphosius scholasticus (author of a
book of riddles: AL 281, pp. 20234), Tuccianus (AL 2712, p. 198), and Vincentius (AL 273,
pp. 1989); note, however, that Modestinus (AL 267, p. 196) may be earlier: M. Schanz and
C. Hosius, Geschichte der romischen Literatur bis zum Gesetzgebungswerk des Kaisers Justinian, 4 vols.
(Munich, 190720), 3:458, esp. 47 and ibid., 4/2:6976.
AL 282, pp. 2356.
H. Keil, De grammaticis quibusdam latinis infimae aetatis commentatio (Erlangen, 1868), p. 4; repr. in
M. Rosenblum, Luxorius: A Latin Poet among the Vandals (New York, 1961), p. 259.
Keil, De grammaticis, p. 4.
AL 214, 218, and 220, pp. 15960 and 1623.
D. Shanzer, A Philosophical and Literary Commentary on Martianus Capellas De Nuptiis Philologiae
et Mercurii, Book 1, University of California Publications in Classical Studies 32 (Berkeley, Calif.,
1986), pp. 1721.
Aegritudo perdicae, ed. E. Baehrens and F. Vollmer, in Poetae latini minores, 6 vols. (Leipzig, 1881
1911), 5:23850, with ibid., p. viii.

135

Staying Roman
Reposianus work, may date from the Vandal or early Byzantine period as
well.27
The air of normality that pervades such literary sources is perhaps the
single most important factor colouring our sense of a growing rapprochement between the Vandal and Romano-African communities in the later
fifth and sixth centuries. By the time the satirist Luxorius cast his critical
eye over elite African society, he was able to see a world where Vandal and
Romano-African aristocrats had come to share the same tastes, the same
interests, the same pastimes. For the men and women of Luxorius generation, the religious reign of terror that the Vandal king Huneric briefly
inflicted on Africas Nicene population (below, section 3) was probably
at its most vivid a recollection from their parents childhood. For many
it would have been even more remote, a thing of their grandparents
day. The intervening generations had seen the still-closer interweaving
of the Vandal nobility and the Romano-African elite. Literature, artwork, gardens, dinner parties, and especially the hunt filled the time of
these leisured communities (see above, Chapter 1.3).28 The inhabitants
of Vandal Carthage including men of Luxorius own social standing
delighted in watching the chariot races and the arena hunters.29 In many
of his poems, Luxorius derides the sexuality and sexual mores of those
who surrounded him.30 The satirist bemoaned the jealousy of a covetous neighbour, and mocked drunkards and poor sports, a teacher
who beat his students, and the poetic pretensions of a man whom
Luxorius himself considered to be uneducated.31 Luxorius vision of
Vandal Carthage was cantankerous, even vitriolic. His perspective was
shaped by his classical literary models, and above all Martial; but the
city that he depicts is still recognizably real, and was a world into which
both Vandals and Romano-Africans appear to have been reasonably well
integrated.
27

28

29
30

31

Reposianus: AL 247, pp. 17786. Regianus: AL 2646 (esp. AL 266), p. 195; so too the anonymous AL 193, p. 145. P. Langlois, Peut-on dater Reposianus (Anth. Lat. 253, Riese)?, Revue de
philologie de litterature et dhistoire anciennes 47 (1973), pp. 30914.
In addition to the poems cited above, see (on the hunt) AL 295 and 302, pp. 243 and 2489. See
also L. Dossey, The Last Days of Vandal Africa: An Arian Commentary on Job and its Historical
Context, JThS n.s. 54 (2003), pp. 11618.
AL 288, 301, 319, 3223, 331, 3489, and 368, pp. 240, 248, 258, 260, 2645, 2735, and 285.
Sexually active elders: AL 292, 296, 304, and 3389, pp. 242, 244, 24950, and 2689; sodomites:
AL 297, 316, and 331, pp. 2445, 256, and 2645; a hermaphroditic woman, a man who
prostituted his wife, a man who loved ugly girls: AL 312, 317, and 324, pp. 254, 257, and 261;
the ugly: AL 305, 353, and 3567, pp. 2501, 2767, and 27980; a beautiful woman devoted to
chastity: AL 359, p. 280.
Neighbour: AL 309, p. 252. Drunkards: AL 292, 298, 306, and 358, pp. 242, 245, 251, and 280.
Poor sports: AL 301 and 328, pp. 248 and 263. Teacher: AL 289, p. 240. Uneducated: AL 311,
p. 253.

136

The old ruling class

1.2. Material prosperity
It mattered deeply to this process of integration that the late Roman elite
lifestyle be accessible to both Vandals and Romano-Africans. Geiserics
early, high-profile confiscations had provided him with the wealth with
which to reward his followers and establish his own royal magnificence;
thereafter, however, it was very much in the Vandal kings interests to
facilitate the process of inter-communal reconciliation and integration
by guaranteeing the rights of all their subjects Vandals and RomanoAfricans alike to own and manage property.
Under normal circumstances, these rights do seem generally to have
been assured. Even according to Victor of Vitas hostile testimony, a number of Romano-Africans in government service were wealthy propertyowners.32 In these cases attachment to the Vandal militia may well have
enabled them to amass and protect at least a portion of their assets;
but, as elsewhere in the West, not every landowner had been subject to
expropriation in the course of the initial barbarian occupation of Africa.
Procopius concedes, for example, that the owners of estates which Geiseric considered marginal were allowed to retain possession of their lands
after the fall of Carthage in 439.33 This had presumably been the situation on the Fundus Tuletianos, in the highlands of western Byzacena,
whose proprietor in the 490s was a certain Flavius Geminius Catullinus,
flamen perpetuus.34 Moreover, the original arrangements that had accommodated Vandal settlement in Africa could later be subject to argument
and appeal. In Geiserics reign, for example, Fulgentius of Ruspes father
and uncle were able to recover a family estate near Thelepte that their
father had lost less than thirty years earlier when he was banished from
Africa.35 Similarly, as an advocate, the poet Dracontius had disputed in
the law courts of late fifth-century Carthage for and against the return of
property to those deprived of their patrimonies. Dracontius had furthermore arranged legacies, and indeed the Vandal kings Romano-African
subjects probably usually enjoyed the right to give, bequeath, and inherit
property.36 Fulgentius certainly gained control of his familys landholdings
32
33
34

35
36

Vict. Vit. 1.4850, 2.23, and 3.27, pp. 212, 32, and 85.
Proc. BV 1.5.15, 1:333. See also above, Chapter 1.3.
Tablettes Albertini: actes prives de lepoque vandale (fin du Ve si`ecle), ed. C. Courtois, L. Leschi,
C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952), acts 3, 4, 614, and 1724, 1:218, 223, 234,
238, 242, 246, 249, 253, 257, 259, 262, 270, 271, 274, 275, 278, 279, 281, and 283.
V. Fulg. 1, p. 11.
Dracontius, Laudes Dei, 3:6547, 2:48; see also Dracontius, Romulea 5 subscriptio, 3:160 app.
crit. (Blossius Emilius Dracontius uir clarissimus et togatus). See also Vict. Vit. 3.9, p. 75 and
Quodvultdeus, Adversus quinque haereses 6.17, ed. R. Braun in CCSL 60:282.

137

Staying Roman
after his fathers death.37 When Fulgentius himself entered a monastery,
he in turn ceded the family estate to his mother, so that if his younger
brother Claudius proved to be a dutiful son, she could later give it to
him.38 Later, in the early sixth century, Fulgentius received a gift of land
from a certain Sylvestrius on which the ascetic was able to establish a new
monastery.39 Further down the social spectrum, the peasant cultivators
from the Fundus Tuletianos revealed to us in the Albertini Tablets held
their land under Mancian tenures. If these still functioned as they had in
the classical period, as seems likely, the tenants thereby enjoyed guaranteed leaseholds for as long as they continued to tend their property, in
exchange for the payment in rent of about a third of their crops and the
rendition of a fixed amount of labour service.40
The effort that the members of the Tuletianos community expended
to document their sales in writing suggests that in the Vandal kingdom
property rights and obligations probably continued to be protected under
the law. Huneric conceived of himself as the font of justice (fons iustitiae), and we have already seen the importance that he placed on the
continuity of late Roman legal form in his reign: whether communicating the conditions of religious toleration, summoning bishops to attend
a council at Carthage, or establishing a system of fines, punishments,
and confiscations for adherence to the Nicene version of Christianity,
the Vandal king issued written rulings that scrupulously observed the
strictures of Roman law.41 Both provincial governors and Arian bishops
probably exercised a judicial function in the Vandal kingdom.42 We also
hear of a handful of Africans who chose to pursue legal careers in the fifth
and sixth centuries, including not only the advocate Dracontius but also
two others whose sexual practices were the target of Luxorius abuse.43
37
40

41

42

43

38 V. Fulg. 5, p. 29.
39 V. Fulg. 1011, pp. 5961.
V. Fulg. 1, p. 13.
Tablettes Albertini, acts 4, 914, 17, 19, 20, 23, and 24, 1:223, 246, 249, 253, 257, 259, 262, 270,
274, 275, 281, and 283; J. Kolendo, Le Colonat en Afrique sous le Haut-empire, Annales Litteraires de
lUniversite de Besancon 447 (2nd edn; Paris, 1991), pp. 4774 and D. P. Kehoe, The Economics of
Agriculture on Roman Imperial Estates in North Africa, Hyptomnemata: Untersuchungen zur Antike
und zu ihrem Nachleben 89 (Gottingen, 1988), pp. 2870; but also C. R. Whittaker, Land and
Labour in North Africa, Klio 60 (1978), 33162, esp. 3601 and J. Percival, Culturae Mancianae:
Field Patterns in the Albertini Tablets, in B. Levick (ed.), The Ancient Historian and his Materials:
Essays in Honour of C. E. Stevens on his Seventieth Birthday (Farnborough, 1975), pp. 21316.
Vict. Vit. 2.34, 2.39, and 3.314, pp. 25, 39, and 728; M. E. Gil Egea, Africa en tiempos de los
vandalos: continuidad y mutaciones de las estructuras sociopolticas romanas, Memorias del Seminario de
Historia Antigua 7 (Alcala de Henares, 1998), pp. 31820.
Vict. Vit. 3.9 and 3.1113, pp. 7678; Quodvultdeus, Adversus quinque haereses 6.27, p. 284;
V. Fulg. 7, pp. 435; Gil Egea, Africa, pp. 2835 and 2879; and Gonzalez Salinero, Poder y
conflicto religioso, pp. 956.
AL 290 and 335, pp. 241 and 2667; Gil Egea, Africa, pp. 28991; Merrills and Miles, Vandals,
pp. 21819.

138

The old ruling class

Taken together, these factors probably imply that something resembling
the late Roman legal system continued to function in Vandal Africa.
Yet we also seem to glimpse a certain degree of cynicism with respect
to the personal nature of justice in both the ecclesiastical and the secular
sources for the Vandal period. In contrast to the other kingdoms of the
barbarian West, no law code survives from Vandal Africa; nor were the
Vandal kings praised as great lawgivers in the surviving court poetry.
Indeed, in his polemical account of Vandal rule, Victor of Vita was at
pains to depict a kingdom in which arbitrariness and cruelty rather than
judicial procedure held sway.44 The satirist Luxorius similarly censured
the royal official Eutychus (perhaps Gelimers secretary Boniface) for
capriciously and violently seizing wealth and property on the excuse that
it was to belong to the king.45 Interestingly, when petitioning for an
honor, Thrasamunds poet Felix was also quite explicit that he did not
want a position in the law courts:
I do not desire to become acquainted with legal fasces, nor do I beg that the
proud laws pay me back; I do not want a sad court of justice, since battles harass
the peace and blind chance sinks into fraternal hatreds; and it is irksome to hear
the conflicts and disputes of the togas.46

To be sure, it is doubtless significant that in protesting against Eutychus abuses, Luxorius could appeal to his readers notions of temporal
order: the poet condemned the official as no better than an enemy or a
brigand.47 And of course pessimism in the face of perceived injustice was
nothing new in the late Roman world.48 But it would seem that in the
fifth and sixth centuries, at least some Romano-Africans in a position
to complain about it could find themselves on the wrong side of a legal
system in which they no longer enjoyed all the advantages.
For the moment, though, property holdings continued to provide privileged members of Romano-African society with considerable wealth. A handful of Latin ostraka and (probably) the Albertini
Tablets indicate the interest that estate owners took in the management
of their assets.49 Simultaneously, the disappearance of the imperial taxspine linking Africa to Rome appears to have resulted in local producers
44
45
46

47
48
49

Gil Egea, Africa, pp. 2878.

AL 3367, pp. 2678. For the identification with Boniface, see Rosenblum, Luxorius, p. 219,
seemingly accepted by PLRE 2:447, s.n. Eutychus 2.
AL 248, ll. 1923, p. 187: non ego litigeros cupio cognoscere fasces, / nec mihi reddantur iura
superba precor; / triste forum nolo, vexant quod proelia pacis / fraternisque odiis alea caeca
subit; / conflictus audire piget rixasque togarum.
AL 336, p. 267.
K. Uhalde, Expectations of Justice in the Age of Augustine (Philadelphia, Pa., 2007).
Bonnal and Fevrier, Ostraka de Bir Trouch, pp. 23949; CIL 8.22646.20; Tablettes Albertini, esp.
act 33, 1:299; and perhaps A. Merlin, in BCTH (1912), cclix, no. 1. Wickham, Framing, p. 266.

139

Staying Roman
and merchants gaining a freer hand in the running of the regions export
economy.50 At least olive oil no longer seems to have been sent to coastal
collection-points for bottling; instead it was put into amphorae primarily
on rural installations.51 The period of Vandal ascendancy in Africa also
saw the introduction of new amphora and fine ware types, developments
suggestive of a degree of economic dynamism in the region.52
This vitality appears to have been particularly characteristic of areas
with comparatively easy access to the Mediterranean littoral. Thus, for
example, the Vandal conquest seems not to have affected the production of ARS at El Mahrine.53 As we have seen, the rural economy also
continued to flourish around nearby Carthage and Segermes; ceramics
production on the inland farms and villas in the coastal zone of the Sahel
and around the Sebkhet Sidi el Hani (a salt lake east of Kairouan) suggests that in eastern Byzacena too estates remained prosperous.54 Further
inland, by contrast, the economic strategies that had ensured affluence
in the late imperial period were gradually becoming less viable. Of the
eight rural kilns producing ARS in the interior of Byzacena in c.400, six
appear to have ceased production by the mid fifth century. In the early
sixth century, a seventh kiln, on the villa at Sidi Marzouk Tounsi, was no
longer producing the exceptionally high-quality fine wares that it had
once exported to Egypt and elsewhere. Only the kiln at Chougafiya,
further to the north-east, appears to have remained active longer.55 Similarly, to judge from the contraction of diagnostic pottery on rural sites
around Dougga (in the highlands of southern Africa Proconsularis), it
would seem that over the Vandal century olive cultivation became less
profitable in this area than it had been in the imperial period.56 The same
50

51

52

53

54
55
56

C. Panella, Merci e scambi nel Mediterraneo tardoantico, in A. Momigliano and A. Schiavone (eds.), Storia di Roma, 4 vols. (Turin, 1988), 3/2:6423 and Leone, Changing Townscapes,
pp. 1334.
D. P. S. Peacock, F. Bejaoui, and N. Ben Lazreg, Roman Amphora Production in the Sahel
Region of Tunisia, in Amphores romaines et histoire economique: dix ans de recherche, Collection de

lEcole
francaise de Rome 114 (Rome, 1989), p. 200; but see also D. P. S. Peacock, F. Bejaoui,
and N. Ben Lazreg, Roman Pottery Production in Central Tunisia, Journal of Roman Archaeology
3 (1990), 5984. See further Leone, Changing Townscapes, pp. 133, 142, and 15960.
C. Panella, Le anfore di Cartagine: nuovi elementi per la ricostruzione dei flussi commerciali
del Mediterraneo in et`a imperiale romana, Opus: Rivista internazionale per la storia economica e
sociale dellantichit`a 2 (1983), pp. 568 and Keay, Late Roman Amphorae, 2:4203.
M. Mackensen, Die spatantiken Sigillata- und Lampentopfereien von El Mahrine (Nordtunesien):
Studien zur Nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts, Munchner Beitrage zur Vorund Fruhgeschichte 50, 2 vols. (Munich, 1993), 1:48791.
Peacock, et al., Roman Amphora Production, pp. 1838 and 199200 and Peacock, et al.,
Roman Pottery Production, esp. p. 82.
Peacock, et al., Roman Pottery Production, pp. 6683.
M. de Vos, Rus Africum: Terra, acqua, olio nellAfrica settentrionale: Scavo e ricognizione nei dintorni di
Dougga (Alto Tell tunisino) (Trent, 2000), pp. 201.

140

The old ruling class

was true some 150 kilometres to the south-west, in the countryside of
the Cillium-Thelepte region.57 Even so, by the standards of the day, it
was still possible to do well and make a name for oneself in this very same
district. At least Fulgentius of Ruspes sixth-century biographer tells us
that the young mans diligent administration of his family estate near
Thelepte contributed to his appointment as a local procurator, presumably on account of the wealth that the property generated.58 References
to new olives and new figs in the Albertini Tablets would seem to indicate
that further south too farmers were at least replacing old unproductive
trees, and possibly even expanding the area under cultivation.59 Here as
well rural prosperity will have translated into social status, for the wealth
generated by his estates in this region probably contributed to the
landowner Flavius Geminius Catullinus ability to secure his position
as a civic priest.
The process of VandalRoman integration may well have been a long,
slow, and at times painful one, if for no other reason than that on a
material level it was predicated on the dispossession of Romans of at least
a portion of their wealth. Yet Romano-Africans continued to control
property under the Vandal regime, and that property probably continued
to be protected under the law. Though we have some indications that the
old elite felt themselves to be in a legally disadvantaged position, control
of property nevertheless meant access to affluence and social status. Vandal
Africa retained its cultural and intellectual vitality, and already when
the advocate Dracontius set about acquiring the essentials of a Roman
education at some point in the second half of the fifth century
his teacher, Felicianus grammaticus, encouraged barbarians and RomanoAfricans to mingle in the auditorium.60 But it is doubtless significant
that such cultural accommodation as took place did so on Roman terms.
In his panegyric to Hilderic, Luxorius only alludes as briefly as possible
to the fact that the king was a Vandal. Despite extensive vaunting of
Hilderics Roman ancestors, the poet never mentions the kings descent
from the Hasding royal house. His kings barbarian lineage was probably
not problematic to Luxorius in and of itself. African poets typically had
no difficulty acknowledging that the subjects of their praise were Vandals;
certainly Thrasamunds encomiasts made open reference to the fact that
the monarch was a Hasding. Rather, everything that these Latin poets
57
59
60

58 V. Fulg. 1, p. 13.
Hitchner, Kasserine Archeological Survey 19821986, pp. 741.
Tablettes Albertini, acts 10, 11, and 24, 1:249, 253, and 283 (new figs); ibid., act 24, 1:283
(new olives).
Dracontius, Romulea, 1:14, 3:134.

141

Staying Roman
found praiseworthy about their Vandal kings fit neatly within the late
Roman ideal of good rulership. Significantly, the same was true of the
criticism that Romano-Africans such as Dracontius levelled at Vandal
kings who were felt to have fallen short of the Roman imperial mark.61
Of course we also do not know what distinctively Vandal acclaim
would have sounded like. Unless the Latin names mask Vandal identities,
no laudatory praise of a Vandal king survives from the hand of a Vandal
author. Given the rapid pace of Vandal acculturation to the norms of
late Roman society, however, even if such tribute were still extant it
might not sound so very different from the adoring verse gathered into
the Latin Anthology. Even so, insofar as Vandals remained beyond the
pale of Roman culture, they seem to have caused their Romano-African
subjects some discomfort. We have already seen (above, Chapter 1.4)
that when the Latin Anthology was assembled, cultured readers despaired
at the uncouth calls for food and drink that the Goths made in their
Germanic tongue at dinner parties. The insight that Luxorius and the
other secular poets of the Vandal kingdom provide us into their society is
thus not simply that as time passed Romans became more familiar with
Vandals and therefore feared and hated them less, but rather that with
the passage of time Vandals became more and more like Romans, and
therefore less frightening and hateful to them. This reality, combined
with the fact that Vandals themselves apparently enjoyed the privilege of
tax-free property allotments (presumably held in exchange for military
service), will in the long term probably have created a strong incentive
for ambitious Romano-Africans to try to go Vandal.62 And, indeed,
this is in part precisely what the polemic of Victor of Vita seems to have
tried to prevent.
2. politics
Politically too the Vandal kings appear to have been remarkably successful
at winning the widespread acceptance of the Romano-African population. This appears to have been the case especially in the wake of the
imperialVandal treaties of 442 and 474, contracting and acknowledging
Eudocias marriage to Huneric, and the legitimate integration of the
Vandal royal family into the late Roman Mediterranean that these acts
implied. We are able to see local acknowledgement of the legitimacy of
Vandal rule in three particular areas: the service of Romano-Africans in
61
62

For Dracontius, see below, n. 64; see also Luxorius satires on Eutychus, AL 3367, pp. 2678
mentioned above.
Proc. BV 1.5.14, 1:333; Innes, Land, 6674; and see also above Chapter 1.34.

142

The old ruling class

the new administration, the composition of panegyrics in honour of the
Vandal kings, and the dating of inscriptions and manuscripts according
to those same kings regnal years.
2.1. Romano-African service in the Vandal administration
To be sure, not all Romano-Africans seem to have been entirely comfortable with the Vandal kings exercise of power in the saeculum. Individuals
like Fulgentius of Ruspe and the poet Dracontius represent important
secular voices of dissent in the late fifth century, the one seeking refuge
from the Vandal regime in the kingdom of God, the other in a foreign earthly kingdom. Fulgentius began his public life as a procurator in
the Vandal administration, but soon laid down his office and became a
monk.63 Dracontius, an advocate in the law courts of Carthage, wrote a
poem in praise of a ruler whom he did not know (dominus mihi ignotus),
presumably an act prefatory to emigration from the Vandal kingdom.
Unfortunately for Dracontius, though, his sedition was discovered, and
the poet found himself in prison for the affront that he had caused the
king Gunthamund (ad 48496). In a new poem, the former lawyer
pleaded his own case, emphasizing to the Vandal king the importance of
royal magnanimity. The appeal fell on deaf ears. Only the intercession
of the brothers Victorianus and Rufinianus secured Dracontius eventual
release from prison and the restoration of his fortunes.64
Yet these cases also serve to illustrate the fact that Romano-Africans
continued to move in court circles, wield political influence, and pursue
office in the Vandal kingdom. Indeed, far from limiting the ability of
African provincials to attain influential posts, the Vandal regime seems to
have generated new opportunities for ambitious members of the regional
elite. By 439, when the Vandals captured Carthage, the African senatorial
order had faded from prominence in the administration of the Roman
empire as a whole for the better part of a century, though they had
been slightly more active in the administration of their own region.65
Opportunities continued to exist under the Vandals. Indeed, the establishment of the Vandal kingdom had involved the creation of at least a
63
64

65

V. Fulg. 13, pp. 1323.

Dracontius, Satisfactio, passim and Dracontius, Romulea 6, 4:15. F. M. Clover, The Symbiosis of
Romans and Vandals in Africa, in E. Chrysos and A. Schwarcz (eds.), Das Reich und die Barbaren,

Veroffentlichungen des Instituts fur Osterreichische

Geschichtsforschung 29 (Vienna, 1989), pp.
626. See also A. H. Merrills, The Perils of Panegyric: The Lost Poem of Dracontius and its
Consequences, in Merrills, Vandals, Romans, and Berbers, pp. 14562
M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Spatantike, Frankfurter althistorische Studien 7 (Kallmunz, 1973), pp. 2340 and Warmington, North African Provinces,
pp. 1068.

143

Staying Roman
handful of new offices; and while some of these, like praepositus regni,
may have been in the hands of individuals who identified as Vandals,
others were clearly open to Romans.66 Thus in Geiserics reign, his sons
Huneric and Theoderic had local stewards (procuratores domus) named
Saturus and Felix, and in the late fifth century the office of proconsul
Carthaginis was held by at least two Romano-Africans named Victorianus and Pacideius.67 Until their fall from favour in 437, Geiseric was
similarly advised by four Hispano-Romans who had accompanied the
Vandals on their crossing to Africa from Spain.68 Later, as we have seen,
the same king welcomed the Roman public enemy Sebastian comes to
Carthage, where he served as the kings adviser until his own fall from
grace (see above, Chapter 2). All of these positions will have involved a
certain degree of social prominence and proximity to the king, but we
hear too of Romano-Africans in less exalted posts, such as chief pantomime and torturer.69 Indeed, as we have seen, Victor of Vita indicates
that by Hunerics day a large number of Africans had accepted offices at
the royal court.70 Royal patronage perhaps extended into the provinces
as well, for by the late fifth century local procuratores like Fulgentius of
Ruspe may also have been centrally appointed.71
However, the Romano-Africans who rose to eminence in the Vandal
kingdom seem for the most part not to have been drawn from the ranks
of what had been the leading families in the region on the eve of the
conquest. The local households that had produced Africas proconsuls
and legates in the late fourth and early fifth centuries have left no discernable trace in the sources for the Vandal period.72 In their place we
find two new groups, the first of which consisted of individuals who
may have been drawn either from cadet branches of prominent families or from families whose political fortunes had been eclipsed by the
later fourth century. Thus, for example, at some point in the mid fifth
century Geiseric used a Romano-African named Proculus as his emissary to gather the books and liturgical vessels of the Nicene churches
in Africa Proconsularis. This Proculus may well have been a member
of the gens Aradia, one of the most distinguished families involved in
the early fourth-century administration of Roman Africa. At least, the
66
67

68
69
71

Praepositus regni: Vict. Vit. 2.15 and 2.43, pp. 29 and 41; C. Courtois, Les Vandales et lAfrique
(Paris, 1955), pp. 2523; and Gil Egea, Africa, pp. 2769.
Procurator domus: Vict. Vit. 1.45, p. 19 (Felix) and ibid., 1.48, p. 21 (Saturus). Proconsul Carthaginis:
Vict. Vit. 3.27, p. 85 (Victorinianus) and Dracontius, Romulea 5 subscriptio, 3:160 app. crit.
(apud proconsulem Pacideium).
Prosper, Chronicon s.a. 437, pp. 4756.
70 Vict. Vit. 2.8, p. 27.
Vict. Vit. 1.47 and 3.347, pp. 201 and 8991.
72 For these families, see Overbeck, Senatsadel, pp. 2333.
Wickham, Framing, p. 89.

144

The old ruling class

Vandal kings agent shared his given name with an early fourth-century
proconsul, L. Aradius Valerius Proculus, who had also served as the governor of Byzacena and was later to be one of the few prominent Africans
in the imperial government of Constantine and his successors. The proconsuls of Africa in 319 and again in 340 had also been named Proculus,
the first perhaps the uncle and the second perhaps the son of L. Aradius
Valerius Proculus. Another man from the same family, Q. Aradius Rufinus Valerius Proculus, served as the governor of Byzacena in 321.73 These
individuals were apparently African in origin, but over the course of the
later fourth century the main branch of the family seems to have been
absorbed into the aristocracy of the city of Rome.74 Geiserics emissary
might thus have been drawn from a less illustrious, collateral line. Similarly, the procurator Fulgentius of Ruspe may well have belonged to the
family that produced the Gordian emperors in the third century; but if
so the family had since receded into relative obscurity, notwithstanding
the place that Fulgentius grandfather had enjoyed on the Carthaginian
curia.75 The unfortunate Dracontius likewise bore the same cognomen as
a mid fourth-century vicar of Africa and a mid fourth-century magister
privatae rei Africae (see Table 3.1), to whom the poet could conceivably
have been related, though between them no other eponymous individuals are attested in the highest ranks of the Romano-African elite for over
a hundred years.76
Alongside these figures, drawn perhaps from families of faded eminence, we also find a host of new names: Pacideius, Victorianus, Saturus,
Boniface. These individuals probably represent a second group, this one
consisting of political parvenus, Romano-Africans from less illustrious
elite families who sensed an opportunity in the Vandal administration,
where they sought positions. This is probably the case, for example,
with the sixth-century poet Luxorius, who held the grade of vir clarissimus et spectabilis but whose name had not been an eminent one in
Roman Africa. It occurs in the epigraphic record from Hadrumetum,
Dougga, and Auzia (the latter in Mauretania Caesariensis), but it was
unknown amongst the more exalted ranks of the late imperial senatorial

73

74
75
76

Vict. Vit. 1.39, p. 17; Overbeck, Senatsadel, pp. 245, 30, 325, and 389; PLRE 1:745, s.nn.
Proculus 34, and ibid., pp. 7479, s.nn. Proculus 1112, and see also ibid., p. 1147, stemma
30.
Overbeck, Senatsadel, pp. 3940; see also PLRE 1:1147, stemma 30 and ibid., p. 1142, stemma
20. Another Proculus (PLRE 2:923, s.n. Proculus 1) was Praetorian Prefect of Africa in 423.
Procurator: V. Fulg. 1, p. 13. Gordiani: Stevens, Circle of Bishop Fulgentius, p. 333. Gordianus
was the name of Fulgentius grandfather; his father and brother were both named Claudius.
PLRE 1:2712, s.nn. Dracontius 34.

145

Staying Roman
Table 3.1. Romano-African families in the late Roman and
Vandal administration
Name
Proculus
Proculus
L. Aradius Valerius
Proculus

Q. Aradius Rufinus
Valerius Proculus
Proculus
Proculus
Dracontius
Domitius Dracontius
Antonius Dracontius
Blossius Aemilius
Dracontius

Office

Date

Proconsul of Africa
legatus of Numidia
praeses of Byzacena
consularis of Europa and Thracia
consularis of Sicily
Proconsul and acting Prefect of Africa
Prefect of the City (Rome)
consul
Prefect of the City (Rome) II
praeses of Byzacena

319
before 333
before c.324
c.325/8
c.330
before 333
3378
340
3512
321

Proconsul of Africa
emissary of Geiseric

340
439/77 (after 457?)

magister privatae rei Africae

Vicar of Africa
advocatus

3201
3647
484/96

aristocracy.77 Some of the other Vandal-era poets may perhaps have

reached the senatorial grade of vir clarissimus through twenty years of service as grammarians, but for the most part they too probably owed their
position and social standing to their willingness and ability to turn their
literary talents to the praise of the new Vandal kings (see below, section
2.2).78 Thus, while what remained in Africa of the old aristocracy was
probably not excluded from office out of hand, the Vandal regime probably offered greater prospects to new men of Roman background. There
were winners as well as losers among the Romans of Vandal North Africa.
2.2. Panegyric
Dracontius notwithstanding, the Vandal kings did not lack RomanoAfrican encomiasts. Regardless of the sincerity of their praise, in writing
77

78

CIL 8.26506 (Dougga), 8.22975 (Hadrumetum), and 8.20793 (Auzia); Rosenblum, Luxorius,
pp. 367. He was probably not a career grammarian: Kaster, Guardians of Language, pp. 41517,
no. 235. Apart from the poet, the name is absent from PLRE 13 and PCBE 1.
AL 2015 and 248, pp. 1503 and 1868 (Felix); AL 214, 218 and 220, pp. 15960 and 1623
(Coronatus); CTh 6.21.1 (ad 425), p. 268; and Miles, Anthologia Latina, p. 309. However,
Kaster, Guardians of Language, pp. 3978, no. 204 is sceptical in the case of Coronatus.

146

The old ruling class

panegyrics to the barbarian rulers of Africa, these poets actively cast
their own lot in with that of the Vandals. The verses of a certain Florentius, for example, celebrated the sixth-century ruler Thrasamund in
the highest terms: In him piety, wisdom, proper behaviour, strength,
beauty, dignity, and a cultivated, mature intellect are in harmony.79 The
poet Felix celebrated Thrasamunds dedication of a new bath complex
at the Carthaginian suburb of Alianae in a cycle of five poems.80 The
pieces are a fireworks display of Felixs poetic prowess, especially the
fifth epigram, which includes an acrostic, a mesostic, and a telestic,
which, taken together, read: Thrasamund, brightening, renews all vota
(Thrasamundus / cunta innovat / vota serenans).81 After being released from
prison, even Dracontius himself is said to have written a panegyric praising Thrasamund.82 An earlier poet by the name of Cato penned a fairly
innocuous celebration of what sounds like the reclamation of coastal land
under Thrasamunds uncle Huneric.83 This praise of these two kings is
particularly remarkable because they are remembered in the ecclesiastical
sources as the arch-persecutors of the African church and of the secular
aristocracy of Carthage (see below, section 3).
Hilderic who officially ended the persecution of the Nicene church
received his encomium from the pen of Luxorius. The poet lampooned
a royal official for his rapaciousness, mocked a royal eunuch for wearing an inappropriate headband, and derided an informer who would
trump up false charges against those who had not asked him to be a
grooms attendant at their weddings.84 But Luxorius also wrote a moving
lament mourning the untimely death of the Vandal prince Hoageis
young daughter Damira while her father was on campaign, praised
Hilderics construction of a new audience hall at Anclae and, as discussed above (Chapter 1.3), lauded the kings sagacity in reference to his
Roman grandfather Valentinian III and the house of Theodosius.85 In
offering such praise to the Vandal kings and their families, poets like Luxorius, Florentius, Felix, Cato, and even Dracontius crossed a meaningful
79
80
81
82
83
84
85

AL 371, ll. 56, p. 287: in quo concordant pietas, prudentia, mores, / virtus, forma, decus,
animus cultusque virilis.
AL 2015, pp. 1503.
AL 205, p. 153; Schanz and Hosius, Geschichte der romischen Literatur 4/2:712.
Baehrens and Vollmer, Poetae latini minores, 5:237; Schanz and Hosius, Geschichte der romischen
Literatur 4/2:59 and 61.
AL 382, p. 295.
AL 3367, pp. 2678 (Eutychus); AL 293, p. 242 (eunuch); and AL 332, p. 265 (informer).
AL 340, pp. 26970 (Damira); AL 194, p. 145 (Anclae); and AL 206, p. 154 (praise). On the name
Damira, which may be Germanic, see M. Schonfeld, Worterbuch der altgermanischen Personen- und

Volkernamen nach der Uberlieferung

des klassischen Altertums bearbeitet (Heidelberg, 1911; repr. 1965),
p. 70.

147

Staying Roman
line between grudging acquiescence to the cold, hard fact of the Vandal
presence on the one hand and enthusiastic co-operation with the Vandal
regime on the other.
2.3. The dating of inscriptions and manuscripts
Further evidence of the political acceptance of the Vandal regime comes
from a very specific and technical area: the dating of inscriptions. The
production of inscriptions was, for the most part, an elite phenomenon
in the Roman world. Moreover, dating an inscription was an act with
political implications, for it spoke in very permanent terms of worldly
loyalties. The dates they employ therefore give us some insight into
elite attitudes not only towards time, but also towards the shifting face
of political rule. To assess the changing Romano-African outlook with
respect to the Vandals, then, we must consider the politics of reckoning
time in late antiquity, the changes in western dating systems with the rise
of the barbarian successor kingdoms to the Roman empire, and more
specifically the systems that developed in Africa under Vandal rule.
For those who measured time according to the succession of the
Roman consuls, refusal to acknowledge one or the other of these officials
was a sign of hostility.86 Their acceptance, on the other hand, implied
recognition of the political legitimacy both of the consuls themselves
and of the rulers who appointed them. In 346, for example, the western
emperor and self-appointed consul Constans refused to acknowledge his
own joint consulate with his brother, the eastern emperor Constantius II.
Thirteen Italian and African inscriptions from that year are dated simply
to the post-consulate of Amantius and Albinus, the consuls of 345, a
practice not followed in the East.87 Similarly, the consuls proclaimed by
the western usurper Magnentius (ad 3513) were not recognized in the
eastern empire, though a number of inscriptions attest their acceptance in
Magnentius own territory.88 Stilicho, later the power behind the western
throne, refused to recognize the eastern consul in ad 399, 400, 404, and
405, and western inscriptions dated according to the consulate list only
one consul for each of those years.89
Outside Italy, however, the epigraphic use of consular dating does
not appear to have been particularly widespread in late antiquity. Of the
2,462 surviving inscriptions with consular dates from the late antique
86
87
89

R. Bagnall, A. Cameron, S. Schwartz, and K. Worp, Consuls of the Later Roman Empire, Philological Monographs of the American Philological Association 36 (Atlanta, Ga., 1987), pp. 246.
88 Ibid., pp. 23641, s.aa. 3513.
Ibid., pp. 2267, s.a. 346.
Ibid., pp. 3325 and 3425, s.aa. 399400 and 4045.

148

The old ruling class

Mediterranean (ad 284541) analysed by Bagnall, Cameron, Swartz, and
Worp, some 2,138 (87 per cent) come from Italy, 1,682 of these from
Rome itself. Only 324 (13 per cent) come from the rest of Western
Europe, the Danubian provinces, Africa, and the East combined.90 This
observation is, however, somewhat mitigated by the fact that Italy was
immensely productive of inscriptions in general, certainly more so than
the provinces of the Roman West, and thus accounts to an uncertain
degree for a greater proportion of surviving inscriptions overall.
Beyond Rome and Italy, a variety of local eras were used to measure
the passage of time. In many provinces, including Spain and Mauretania,
years were reckoned according to provincial eras that commemorated
these regions integration into the Roman empire. Beirut, Tyre, and
Sidon still used municipal eras in late antiquity. The Seleucid era, widely
employed in the East, commenced with the first satrapal year of Seleucus I
in 311 bc. Jews and Christians alike marked years from the creation of
the world, though opinions differed as to precisely when the event had
taken place. Into the Byzantine period, chroniclers continued to make
use of the venerable four-year cycle of Olympiads, although documents
were rarely dated according to this standard. The indiction, a tax cycle
introduced by Diocletian and calculated since ad 312 in fifteen-year
increments, also came to be widely used and under Justinian was mandated for the dating of documents.91 The significance of dating systems
was thus both political and cultural in late antiquity. Ones choice of
how to express the chronology of events could link one in different ways
to the Roman empire and its administrative structures, to a communal
past (including that of the long-defunct Seleucid kingdom), or to sacred
history.
To a greater or lesser degree, the different methods of measuring
time current in the late Roman West all outlived direct imperial rule
itself. In Italy, consular dates continued to be the norm, even after the
deposition of Romulus Augustulus. Between ad 476 and 541, Rome
alone produced at least 244 inscriptions dated according to the consular
90
91

Ibid., pp. 5860.

The classic works on chronography in the ancient world are E. J. Bickerman, Chronology of the
Ancient World (2nd edn; Ithaca, NY, 1980), here esp. pp. 709 and A. Samuel, Greek and Roman
Chronology: Calendars and Years in Classical Antiquity, Handbuch der Altertumswissenschaft 1/7
(Munich, 1972), here esp. pp. 2458. See also H. Seyrig, Antiquites syriennes, Syria: revue dart
oriental et darcheologie 39 (1962), pp. 424 on the survival of the urban eras of Beirut, Tyre, and
Sidon into late antiquity; R. Bagnall and K. Worp, Regnal Formulas in Byzantine Egypt, Bulletin
of the American Society of Papyrologists suppl. 2 (Missoula, Mont., 1979); and R. Bagnall and
K. Worp, The Chronological Systems of Byzantine Egypt (2nd edn; Leiden, 2004). Just. Nov. 47
(ad 537), pp. 2835 mandated the use of imperial regnal year, consular year, indiction, month,
and day to date legal documents.

149

Staying Roman
year, and the rest of Italy produced another 168.92 In the same period
Italian inscriptions of an official nature were occasionally dated to the
reigns of the Ostrogothic kings Theodoric and Athalaric, but like
contemporary imperial inscriptions dated by reign such inscriptions
seem by and large to have avoided using specific regnal years.93 The
survival of inscriptions from late Roman Spain employing consular dates
is patchy, though we have examples from Tarraconensis into the early
sixth century.94 Much more widely used there in late antiquity was the
provincial era.95 Even so, barbarian kings left their mark on the epigraphic
practices of the Spanish provinces as well. Regnal dating may have been in
use as early as 451 in the Visigothic kingdom, and was certainly employed
by the late fifth century when an inscription from Braga commemorated
the completion of a church in era 523 [= ad 485], in the reign of the
most serene king Veremund (era DXXIII regnante serenissimo Veremundu
rex).96 By 496, at least one of the Visigoths Gallic subjects dated his
epitaph according to the years of Alaric; but in Frankish Gaul, the use
of the regnal years of barbarian kings to date inscriptions appears for the
most part to have been a phenomenon of the sixth and later centuries.97
By contrast, the use of consular dates in Gaul was largely confined to
the territory of the Burgundian kingdom, where the practice continued
even after the Frankish conquest of the region in 534.98
In late Roman Africa Proconsularis, Byzacena, and Numidia perhaps
throughout North Africa two dating systems existed side by side during
the last decades of Roman rule (see Table 3.2). Official inscriptions were
typically dated to the reign of the current emperor or emperors, often

92
93

94
95
96
97

98

Bagnall, et al., Consuls, p. 60.

ILCV 93 and 113; see also ibid., 37ah and probably ibid., 225 (Theoderic); and ibid., 38ad
(Athalaric). Specific dates were not unknown: ILCV 39 = CIL 5.6418: + dn. Atalaricus rex +
/ gloriosissimus has / sedis spectaculi anno / regni sui tertio fieri / feliciter precepet +.
Bagnall, et al., Consuls, p. 206, s.a. 336 (Lusitania); ibid., p. 232, s.a. 349 (Baetica); and ibid.,
pp. 238, 308, 378, 444, 476, and 540, s.aa. 352, 387, 422, 455, 471, and 503 (Tarraconensis).
ILCV 3:2735, index 6E.
ILCV 1721; M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain,
ad 300750, BAR International Series 1135 (Oxford, 2003), pp. 1235.
ILCV 1216 = CIL 12.2700 = E. F. Le Blant (ed.), Inscriptions chretiennes de la Gaule anterieures
au VIIIe si`ecle, 2 vols. (Paris, 185665), 2:2067, no. 482 (Viviers): hic requiis/cet in pace / iac.
Domno/lus qui ui/xit annus / XXXVIIII et / dees III, obiit / III k. Maias / XII reg. dom/ni
Alarici; Handley, Death, Society, and Culture, pp. 1256; and see also ILCV 3:27980, indexes 6I
and 6K.
M. Handley, Inscribing Time and Identity in the Kingdom of Burgundy, in S. Mitchell and
G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London, 2000), pp. 83102.

150

The old ruling class

Table 3.2. Consular and Imperial dating systems
in late Roman Africa, c.40754

Africa Proconsularis
Byzacena
Mauretania Caesariensis
Mauretania Sitifensis
Numidia
Tripolitania
Uncertain Algeria
Total

Consular

Imperial

Total

1
1

1
2

1
6

1
1

6
1

1
3
1
1
13

also with reference to the sitting proconsul of Africa.99 When they were
dated at all, private inscriptions such as epitaphs seem to have employed
consular dates.100 As in Spain, however, consular dates seem never to have
been widely employed in late Roman Africa. Surprisingly, of the 2,462
late antique consular inscriptions mentioned above, only eighteen (0.73
per cent) come from Latin North Africa, including the Mauretanias. Six
of these belong to the fifth century; none unambiguously dates to the
sixth.101 As in Spain, the provincial era continued to provide an alternative
to consular dating in late Roman Mauretania, but in Numidia, Africa
Proconsularis, and Byzacena the system had already fallen out of fashion
early in the third century.102 Thus even in the imperial period late antique
epitaphs and other private inscriptions from these provinces were for the
most part simply undated.
With the Vandal conquest of Carthage, both imperial and consular
dating systems fell into disuse in the eastern portion of the late Roman
diocese of Africa. In their place we see the slow evolution of a number of
new dating systems in a limited number of inscriptions. The vast majority
of inscriptions from Vandal Africa continued to be undated, but nine
99

100

101
102

Imperial dates: AE (1974), 197, no. 698; CIL 8.1358 + p. 938, 8.7017 + p. 1847, 8.7018,
8.24069, 8.25837, and see also CIL 8.970 (= 12449); R. Cagnat and A. Merlin, Inscriptions
latines dAfrique (Tripolitaine, Tunisie, Maroc) (Paris, 1923), p. 81, no. 276 and p. 93, no. 314;
ILAlg. 1.263, 1.2108, and 1.3055; and J. M. Reynolds and J. B. Ward-Perkins, Inscriptions of
Roman Tripolitania (Rome, 1952), p. 135, no. 480.
Revue archeologique, 6th ser. 42 (1953), p. 178, no. 39 (ad 409); CIL 8.11127 (ad 427); Ennabli,
Inscriptions chretiennes de Sainte-Monique, pp. 1712, no. 46 (ad 439); AE (1967), 200 and 2089,
nos. 595 and 640 (both ad 452; note that the latter also uses an anno provinciae date); Revue
archeologique, 5th ser. 20 (1924), p. 387, no. 58 (ad 454).
Bagnall, et al., Consuls, pp. 5860; see also Handley, Death, Society and Culture, p. 130.
Clover, Felix Karthago, pp. 1011.

151

Staying Roman
reckon time according to the regnal years of five different Vandal kings,
and a tenth according to those of an unnamed dominus rex, presumably
also a Vandal since the inscription is from Carthage (see Table 3.3).103
The earliest of these may have appeared already in mid fifth century
Byzacena. At Sbetla, a thirty-eight-year-old presbyter named Vitalis was
buried and commemorated with an epitaph that dated his birth to the
twenty-eighth year of Geiserics reign (natus anno XXVIII regis Gesiric).104
Since the word natus may refer to Vitalis birth either to eternal or to
temporal life, it is not entirely clear whether the inscription was erected
in ad 467 or 505; but the earlier date seems the more likely. If so, it is a
remarkable testament to the rapid acceptance of the political legitimacy
of the Vandal regime, even outside Carthage, the strongest centre of
Vandal power. A second inscription, also from Byzacena, dates to the
reign of the third Vandal king, Gunthamund (ad 48496), while the
remaining epigraphic examples of the use of Vandal regnal dates were
commissioned in or after the reign of Thrasamund (ad 496523); that
is to say, in the third or fourth generation of Vandal rule in Africa. The
last known example comes from 531 two years before the Byzantine
reconquest in the first year of the lord king Gelimer (anno primo domini
regis Gelimer).105
Another eleven funerary inscriptions are dated to the year of Carthage
(anno Karthaginis) or simply the year (anno) (see Table 3.4).106 Of these
two formulations, anno Karthaginis dates seem to be distinctively Vandal;
103

104
105
106

Regnal dates (numbers refer to Table 3.3): (1) CIL 8.25357a = ILCV 1406 = Courtois, Vandales,
p. 380, no. 113; (2) A. Merlin and P. Monceaux, in Comptes rendus de seances de lAcademie des
Inscriptions et Belles-Lettres (1914), p. 483; (3) F. Bejaoui, Les Vandales en Afrique: Temoignages
archeologiques. Les recentes decouvertes en Tunisie, in G. Berndt and R. Steinacher (eds.), Das

Reich der Vandalen und seine (Vor-) Geschichten, Osterreichische

Akademie der Wissenschaften,
Philosophisch-historische Klasse, Denkschriften 366/Forschungen zur Geschichte des Mittelalters 13 (Vienna, 2008), p. 202; (4) CIL 8.2013 = CIL 8.16516 = ILCV 1385 = Courtois, Vandales, p. 378, no. 96; (5) CIL 8.11649 = ILCV 3104A = Courtois, Vandales, p. 378,

no. 98 = N. Duval, Recherches archeologiques a` Hadra, 2 vols., Collection de lEcole

francaise
de Rome 18/12 (Rome, 197581), 1:2813, no. 419; (6) Bejaoui, Vandales en Afrique,
p. 201; (7) Bejaoui, Vandales en Afrique, pp. 2001; (8) CIL 10516 = CIL 8.11528 = ILCV
388 = Courtois, Vandales, p. 379, no. 103 = Duval, Hadra 1:2737, no. 413; (9) CIL 8.23053u
= ILCV 2683 adn. = Courtois, Vandales, p. 379, no. 104; (10) ILCV 4452 = Courtois, Vandales,
p. 379, no. 106. Undated inscriptions: see, e.g., Courtois, Vandales, pp. 3848.
Merlin and Monceaux, Comptes rendus de seances de lAcademie des Inscriptions et Belles-Lettres
(1914), p. 483.
ILCV 4452 = Courtois, Vandales, p. 379, no. 106.
Anno and anno Karthaginis dates (numbers refer to Table 3.4): (1) BCTH (19301), p. 253,
no. 13; (2) CIL 8.28044 = Courtois, Vandales, p. 369, no. 22; (3) Courtois, Vandales, p. 370,
no. 30; (4) ibid., p. 371, no. 37; (5) ILCV 3139 = Courtois, Vandales, p. 373, no. 56; (6) Courtois,
Vandales, p. 375, no. 67; (7) ibid., p. 375, no. 70; (8) ILCV 1601A = Courtois, Vandales, p. 379,
no. 107; (9) ILAlg. 2761; (10) Revue archeologique, 6th ser. 48 (1956), p. 200, no. 125; (11) AE
(1967), 200, no. 596.

152

Table 3.3. Vandal regnal dates in African inscriptions

No. City

Vandal Date

Date

Deceased

Age

domini regis
1. Carthage

d(ie) / []s d(omi)n(i) regis

439533

[]ndilu(s)

60

12 Sept. 467;
inscription
= ad 505?

Vitalis presbyter

38

acceptatus / est die XIIII kalendas / octobres anno /

septimo d(omi)ni n(ostri) re/gis Gonthamun/di

18 Sept. 491

Lucilianus diaconus

nat]us est ann(o) VII do(mi)n(i) n(ostri) re[gis

Thra/sa]mundi, III non(as) februari[as / et recessit
ann(o) XII, VII ka[lend(as)] / augustas
d(ie) VIII k(a)l(endas) martias, an(n)o XIIII
d(o)m(ini) r(e)g(is) T(hra)s(a)m(undi)
deposita sub / die XIII k(a)l(enda)s febr/arias ann(o)
XXII / d(omini) n(ostri) regis Thra/samundi
Anno bicesimo vi dom/ni regis Tasamund

3 Feb. 50326
July 508

Ge . . .

22 Feb. 510

Festa

17 Feb. 517

Fortunatiana

521

Geiseric (ad 43977) or Thrasamund (ad 496523)

2. Sbetla, Byzacena
natus anno XXVIII / regis Gesiric, pridie idus /
septembres
Gunthamund (ad 48496)
3. El Ounaissia, Byzacena
Thrasamund (ad 496523)
4. Tebessa

5.

Hadra (class. Ammaedara)

6.

El Erg, Byzacena

7.

El Gousset, Byzacena

Hildiric (ad 52330)

8. Hadra
9.

Uppenna (mod. Hr.

Chigarnia), Byzacena

Gelimer (ad 53034)

10. Madauros

16

VIII / id(us) decem/bres, anno IIII d(omi)n(i) regis / 6 Dec. 526

Ildirix
di]e XV / a[] dec/em[bres, a]nn(o) V (?) / [Hild]iricis 527 (?)

Astius Mustelus, flamen

perpetuus christianus
Quadratilla

72

anno primo dom(i)n(i) regis Geli(mer) / XI

k(a)l(endas) febr(uarias)

Desiderius (buried by
Respectus episcopus (?))

26

Region is Africa Proconsularis, except where otherwise noted

22 Jan. 531

...

Table 3.4. Anno and Anno Karthaginensis dates in African inscriptions

No.

City

Anno
1.
Madauros (mod.
Mdaourouch)
2.
Aquae Caesaris
(mod. Youks)
3.
Chott Manzel-Yayia
4.

5.
Leptis Minor (mod.
Lemta), Byzacena
6.
Tubernuc (mod. An
Tebornok)
7.
Hippo Regius

Anno Karthaginis
8.
Madauros
9.
10.
11.

Madauros
Hippo Regius,
region
Cuicul (mod.
Djemila), Numidia

Vandal date

Date

Deceased

Age

III nonas iulias anno tertio

4 July 442, 480, 487,

499, 526, or 533
28 April 446, 484, 491,
503, or 530
451, 496, 508, or 539
453, 510, or 541
26 June 468 or 556

Abedeu
Honorata

4 months
14 days
14

Felicitas
Gaudentius
Billatica

18

27 Nov. 471 or 560

Margarita

82

11 Sept. 474 or 562

Ermengon
(buried by
Ingomar)

5 Apr. 446, 484, 491,

503, or 530
23 July 445, 483, 490,
502, or 529
30 Oct. 459 or 516

Donatianus
presbyter
Fl. Anu . . .

96

Iobius, vir
clarissimus
C . . . . . . .na

51

IIII k(a)l(endas) maias, / anno VII

anno XII
anno XIIII
die VI / k(a)l(endas) iuli/as, an/no
XX/VIIII
die V k(a)l(enda)s / decemb(res), ann(o)
XXXIII
die III idus septe/mbres . . . ann(o) XXXV

die / nonas apriles, an(no) VII

Kathag(i)n(is)
Anno VI K(arthagini)s, die VIII
k(a)[l(en)d(a)s] augustus
die III idus nob/emb(ris) anno XX
Kartag(ine)
die II Kal(endas) Mar(tias) a/n(no) XXIIII
K(arthaginis)

Region is Africa Proconsularis, except where otherwise noted

27 Feb. 463 or 520

The old ruling class

at least, three ostraka from Bir Trouch, Algeria, use them interchangeably
with Gunthamunds regnal years (e.g., year nine of Carthage of our
lord king Gunthamund).107 Simple anno dates, by contrast, may well
be Byzantine. A number of Justinians copper coins struck at Carthage
bear an anno formula incorporating the emperors regnal year. Moreover,
Justinians law of ad 537 mandating the use of consular dates also required
the incorporation of the emperors regnal year in the dating of documents.
Thus, since the emperor issued the law ten years into his reign, any
anno date greater than ten could conceivably belong to Justinians reign
(Table 3.4, nos. 37).
The inscriptions that record anno and anno Karthaginis dates can nevertheless usefully be discussed in aggregate. If they are not Byzantine,
three certainly belong to the reign of Geiseric. These are dated to the
years 29, 33, and 35. Geiseric is the only Vandal king to have reigned
more than twenty-nine years, and they would therefore date to ad 468,
472, and 474, respectively. Another two are dated to the years 20 and 24
of Carthage, and must therefore belong to the reign either of Geiseric
or of Thrasamund, both of whom held the Vandal throne for more than
twenty-four years. To these should probably be added another epitaph
belonging to the year (anno) 14, unless the date is Byzantine. A seventh inscription is dated to the year 12, implying Geiseric, Thrasamund,
Gunthamund, or Justinian; an eighth, to the third year, which could
refer to any of the Vandal kings, though probably not to the Byzantine
emperor. Two remaining inscriptions date to the year 7 and to the year 7
of Carthage, which only excludes Gelimer and, again, probably Justinian.
At first blush, it might seem that use of the Carthaginian year would
have allowed Romano-Africans to date inscriptions without explicitly
acknowledging the legitimacy of their new Vandal rulers. Using a Vandal
regnal date, by contrast, would presumably have been more problematic
from an imperial (or loyalist) point of view, for it did imply a fairly
unambiguous recognition of legitimacy that the anno Karthaginis system
did not. However, the official nature of the anno Karthaginis dating system
is confirmed by the existence of Vandal pseudo-imperial silver coinage
struck in the years 4 and 5 of Carthage.108 Moreover, from at least the
107
108

Bonnal and Fevrier, Ostraka de Bir Trouch, pp. 23942 and 2445, nos. 12 and 4; the
quotation is from p. 241, no. 1: annu nonu cartaginis domni nostri regis Guntamundi.
W. Wroth, Catalogue of the Coins of the Vandals, Ostrogoths and Lombards and of the Empires of
Thessalonica, Nicaea and Trebizond in the British Museum (Oxford, 1911), p. 5, nos. 12; W. Hahn,
Moneta Imperii Byzantini: Rekonstruktion des Prageaufbaues auf synoptisch-tabellarischer Grundlage,
vol. 1, Von Anastasius I. bis Justinianus I. (491565) einschliesslich der ostgotischen und vandalischen

Pragungen, Veroffentlichungen der Numismatischen Kommission 1/Osterreichische

Akademie
der Wissenschaften, Philosophisch-Historische Klasse, Denkschriften 109 (Vienna, 1973),
pl. 42.1. See also Vict. Vit. 2.39, p. 39; see further ibid., 3.4 and 3.12, pp. 73 and 77.

155

Staying Roman
later fifth century, Africans seem to have made unhesitating use of Vandal
regnal years to date documents: four or five of the Bir Trouch ostraka and
at least twenty-two of the Albertini Tablets are dated according to the
year of Gunthamunds reign in which they were written.109 The authors
of historical and computational works also quickly adopted Vandal regnal
dates, though the significance of this fact is perhaps lessened somewhat
by the general concern of such authors to give the most accurate date
possible. Even so, the anonymous Libellus de computo paschali, composed
in ad 455, already made reference to the tenth and sixteenth years of
king Geiseric (anno decimo regis Geiserici and anno sextodecimo regis), and
the 463 edition of the Donatist Liber genealogus also employed Geiserics
regnal years to date events.110
Indeed, in actual practice neither the use of anno Karthaginis dates
nor the use of Vandal regnal dates on the part of Romano-Africans
appears necessarily to have implied agreement with or endorsement of
Vandal politics or policies. If the Vandals did deflect secularized late
Roman emperor-worship towards themselves, it is perhaps not surprising
to find a flamen perpetuus (responsible for the maintenance of such cultic
veneration) commemorated in an epitaph dated to the fourth year of
the Vandal king Hilderics reign. But both anno Karthaginis and Vandal
regnal dates were also used by people who clearly had reasons to oppose
the Vandal regime. Though his historical concerns perhaps diminish the
ideological implications of the evidence, Victor of Vita who hated
the Vandals as much as anyone nevertheless made occasional use of
regnal dates to order his narrative. The historian writes, for example,
that Geiserics capture of Rome occurred in the fifteenth year of his
reign,111 which (Victor tells us) itself lasted for thirty-seven years and
three months.112 Victors History further indicates that Huneric reigned
for seven years and ten months before dying the horrible death that befit
a heretical persecutor.113 The anonymous Nicene author of the Passio
septem monachorum also made reference to Hunerics regnal year, though
again there was no love lost between this author and the Vandals.114
109

110

111
112
113
114

Bonnal and Fevrier, Ostraka de Bir Trouch, pp. 23946, nos. 14 and probably no. 5; Tablettes
Albertini, acts 14, 614, 1624, 1:215, 217, 218, 223, 234, 238, 242, 246, 249, 253, 257, 259,
262, 26971, 274, 275, 278, 279, 281, 283.
De ratione paschae 1.2, 2.4, and 2.8, ed. B. Krusch in Studien zur christlich-mittelalterlichen Chronologie, 2 vols. (Leipzig, 1880), 1:279, 287, and 289 (anno decimo) and ibid., 1.2 (twice) and 1.5,
pp. 280 and 281 (anno sextodecimo). Liber genealogus 428, 499, and 628c, pp. 181, 188, and 196.
Vict. Vit. 1.24, p. 12: quinto decimo regni sui anno Geisericus caperet Romam.
Vict. Vit. 1.51, p. 23.
Vict. Vit. 3.71, p. 107, although this passage was probably not written by Victor himself:
C. Courtois, Victor de Vita et son uvre: etude critique (Algiers, 1954), p. 16.
Passio septem monachorum 2, p. 108.

156

The old ruling class

The same was true of the author of the Notitia provinciarum et civitatum
Africae.115 The widespread acceptance and use of the new Vandal-era
dating systems, however, is perhaps better illustrated by an inscription
from Madauros, in the eastern Algerian highlands. Donatianus presbyter
had been banished there on account of the universal faith (pro fide
catholica) and had died in exile at the age of ninety-six on the Nones
of April, in the year 7 of Carthage.116 Similarly, the Vitalis presbyter
whose birth was dated to the twenty-eighth year of Geiserics reign
perhaps adhered to the Nicene rather than the Arian faith, given the
invocation of the Father, Son, and Holy Spirit with which his epitaph
begins. Finally, a manuscript of Hilary of Poitiers a text whose concerns
were theological rather than historical in nature, and a codex which was
moreover produced for Fulgentius of Ruspe (or someone close to him)
while he was in exile in Sardinia for his hard-line Chalcedonian stance is
also dated by a colophon to the fourteenth year of King Thrasamund.117
There is, of course, a difference between dating a manuscript and
dating an inscription. The more public nature of the latter is one part of
it; the greater permanence, another.118 Indeed, Romano-Africans may
have come to terms with the likely continuity of a Vandal presence
remarkably quickly. A second inscription from Madauros, this one dated
to the third year (anno tertio), may be as early as 442 (see Table 3.4,
no. 1) the same year that Valentinian III and Geiseric concluded a peace
treaty recognizing Vandal claims to Africa Proconsularis, Byzacena, and
portions of Numidia and Tripolitania. But the priest Vitalis (or, more
properly, whoever set up his epitaph) may well have been among the
first Romano-Africans to stop hedging his bets and begin using Vandal
regnal dates in a permanent medium. Having probably seen the Vandals
115

116

117

118

Notitia, p. 117. Note that the date given in the Notitia for Hunerics council at Carthage (anno
sexto regis Hunerici) does not correspond to the dates in Victor of Vitas account, including those
in Hunerics law of 484 (quoted at Vict. Vit. 3.314, pp. 728), which place it in the kings
eighth year (ibid., 3.12, p. 77).
Donatianus: ILCV 1601A = Courtois, Vandales, p. 379, no. 107:  + ! / Donatianus
pr(e)sb(yter), / in exilio pro fide ca/t(h)olica hic aput col(oniam) Mad(auros) / relegatus,
recessit die / nonas apriles, an(no) VII Kartha/g(i)n(is) (?). uixit annis XCVI.
A. Wilmart, LOdyssee du manuscrit de San Pietro qui renferme les uvres de saint Hilaire,
in L. W. Jones (ed.), Classical and Mediaeval Studies in Honor of Edward Kennard Rand (New York,
1938), p. 301, of Vatican City, Archivio della Basilica di S. Pietro, MS Basilicanus D.182, fol. 288
(CLA 1.1a): Contuli in nomine d(omi)ni ihu xpi aput karalis constitutus anno quarto decimo
trasamund(i) regis.
On the epigraphic habit in late antiquity, see R. MacMullen, The Epigraphic Habit in the
Roman Empire, American Journal of Philology 103 (1982), pp. 23346, but also E. A. Meyer,
Explaining the Epigraphic Habit in the Roman Empire: The Evidence of Epitaphs, Journal
of Roman Studies 80 (1990), pp. 7496; Handley, Death, Society and Culture; N. Everett, Literacy
in Lombard Italy, c. 568774, Cambridge Studies in Medieval Life and Thought, 4th ser. 53
(Cambridge, 2003), pp. 23576.

157

Staying Roman
cross into Africa, capture Numidia and then Carthage, and fend off an
unsuccessful West Roman attempt to reconquer Africa in 461, all within
his lifetime, it is certainly conceivable that Vitalis could have believed the
Vandal presence would not be simply a transient one. That impression
would only have been reinforced by the events of the later fifth century,
which saw one last failed east Roman attempt at reconquest in 468, and
the eventual emergence of a peace between the Vandal kingdom and the
Roman state that would last into the sixth century. The Vandals, it would
seem, were there to stay.
Insofar as it is possible to tell with such a limited data set, over
time more inscriptions seem to have been dated according to the
kings regnal year. Setting aside for a moment the question of Vitalis
presbyters epitaph, Gunthamunds twelve-year reign produced at least
one dated inscription; Thrasamunds twenty-seven years, another two;
Hilderics seven years, two; and Gelimers three years, one more. Looked
at from this perspective, the trend is positive. Each successive generation
after the Vandal conquest seems to have been more and more willing to
erect permanent monuments commemorating the passing of loved ones
within the time-frame established by the temporal rule of the Vandal
kings. If Vitalis epitaph does date to the reign of Geiseric, this process
may already have begun with an inscription commemorating a member
of the very generation that witnessed the Vandal seizure of Africa. At
least for the moment, however, in the middle of the Vandal century
sits a great epigraphic black hole. Hunerics reign, spanning the seven
years from 477 to 484, has so far yielded no dated inscriptions despite
the fact that this same kings reign has produced considerable literary
evidence of the contemporary Romano-African use of regnal dates.
The absence underscores the difficulty of extrapolating from such a
small number of data points. It may argue in favour of a sixth-century date
for the Vitalis epitaph, which would square nicely with the other dated
inscriptions from the Vandal kingdom. Epigraphic reconciliation with
the permanence of the Vandal presence, then, would have been deferred
to the reign of Gunthamund and the third generation of Africans to
grow to adulthood under the Vandal regime. On the other hand, events
also seem to have taken a decidedly nasty turn for Romano-Africans of
Nicene confession in the last years of Hunerics reign, and African elites
had a brutal reminder of the transience of worldly things. The violence
of Hunerics persecution may have cut short a process of accommodation
already under way in the later years of his fathers reign. By the end of
the century, that process had started again. The worst of the persecution
was over Thrasamund chastised the Nicene clergy with exile, not
with tortures (see below, section 3) and the rapprochement attested in
158

The old ruling class

our secular literary sources was under way. And while many RomanoAfricans may have disliked the Vandals or been unhappy with their rule,
the old regional elite seem to have come to the conclusion that the new
was to be a permanent feature of the African landscape.
Politically, Romano-Africans seem to have come to accept the realities of
Vandal rule remarkably quickly. This was true of Nicene bishops as well
as secular office-holders, old aristocratic families as well as parvenus, exiles
as well as favoured courtiers. The inhabitants of the later Roman empire,
after all, had gained long experience of forced changes in loyalty in the
course of the civil wars of the fourth century. Within the first generation
of the Vandal conquest, Romano-Africans were similarly co-operating
with the new regime; by the late fifth century they were celebrating their
new kings accomplishments in poems of praise. After another decade at
most, Romano-Africans were also dating inscriptions according to the
regnal years of Vandal kings a still more public statement of belief in the
permanence of the new order. Politics in the Vandal kingdom certainly
had its discontents: Fulgentius of Ruspe left public office for monastic
withdrawal and contemplation of the divine, while Dracontius sought
the patronage of a foreign king. By and large, however, the marriage
alliance and peace treaties that bound the African kingdom to what was
left of the Roman state, together with the Vandal kings emphatic stress
on the continuities between late Roman and Vandal power, seem to
have been sufficient to ensure the political legitimacy of Vandal rule in
the eyes of their Romano-African subjects. All in all, opposition to the
Vandal regime seems not to have been expressed in political terms, but
rather in terms of confession.
3. confessional boundaries and social integration
In Roman minds there was a strong association between Vandals and
Arianism, a form of Christian belief named after the fourth-century
Alexandrian presbyter Arius that confessed Christs existential emanation
from (and thus subordination to) God the Father.119 It is not clear when
or how the Vandals as a people were converted to Arianism, though
it seems to have happened in Spain or perhaps Gaul, probably under
119

See, e.g., Prosper, Chronicon s.a. 437, p. 475; Hydatius, Chronicon 79 (ad 428) and 110
(ad 439), pp. 8890 and 94; Proc. BV 1.8.4 and 2.9.14, 1:345 and 1:458; Vict. Tonn. s.a. 466.30,
p. 11; Gregory I, Dialogi 3.32.1, 2:390; S. Costanza, Uuandali-Arriani e Romani-Catholici
nella Historia persecutionis Africanae provinciae di Vittore di Vita. Una controversia per luso
del latino nel concilio cartaginese del 484, in Oikoumene: Studi paleocristiani pubblicati in onore del
Concilio Ecumenico Vaticano II (Catania, 1964), pp. 2246; and Howe, Vandalen, pp. 15682.

159

Staying Roman
the influence of the Visigoths, who had themselves been converted to
Arianism under imperial influence in the fourth century.120 In fact we
might dispute the label attached to this version of Christianity: the only
surviving Vandal statement of belief professes acceptance of the Christological formula endorsed at the twin councils of Rimini and Seleucia,
held in 359 under the auspices of the emperor Constantius II, which
maintained that the Son is like (homoios) the Father according to the
Scriptures.121 Strictly speaking, this homoian formulation is unrelated
to Arius original tenets; but from the later fourth century onwards this
was nevertheless the creed that hard-line adherents of Nicaea who
insisted on the homoousia (consubstantiality or existential sameness) of
the Father and the Son understood and condemned as Arian.122 Yet
we should not be misled by Nicene rhetoric into believing that after
its condemnation at the First Council of Constantinople in 381 this
homoian Arianism was a dead letter among Romans. It remained strong
in the eastern imperial capital itself throughout the fourth century, and
there is no reason to assume that Arianism disappeared in Italy in the
fifth century either.123 The confession appears to have been particularly
well-established in the Balkans in this same period. Certainly the Illyrian
Arian bishop Maximinus appealed to the profession of faith articulated
at Rimini and Seleucia in his debate with Augustine in Hippo Regius
in 427/8, and this was presumably also the formulation accepted by the
small Arian congregation that had gathered in that same city by the early
fifth century.124 On the eve of the Vandal invasion of Africa, Arians could
be Romans too.

120

121

122
123

124

Merrills and Miles, Vandals, pp. 1789. The history of Arianism in the barbarian West is
immensely complicated and desperately in need of synthetic examination, a gap soon to be
filled by Yitzhak Hens forthcoming study on the topic.
Vict. Vit. 3.5, pp. 734. On the councils of Rimini and Seleucia, see T. D. Barnes, Athanasius
and Constantius: Theology and Politics in the Constantinian Empire (Cambridge, Mass., 1993),
pp. 14151 and D. H. Williams, Ambrose of Milan and the End of the Nicene-Arian Conflicts
(Oxford, 1995; repr. 2002), pp. 1137. On Vandal Arianism, however, see also Y. Moderan,
Une guerre de religion: les deux e glises dAfrique a` lepoque vandale, Antiquite tardive 11
(2003), 2144, esp. 306.
See in general R. P. C. Hanson, The Search for the Christian Doctrine of God: The Arian Controversy,
318381 (Edinburgh, 1988).
Socrates, Historia Ecclesiastica 6.8, PG 67:6889; Sozomen, Historia Ecclesiastica 7.57 and 8.8, PG
67:142432 and 15367; N. McLynn, From Palladius to Maximinus: Passing the Arian Torch,
Journal of Early Christian Studies 4 (1996), p. 484; P. Amory, People and Identity in Ostrogothic Italy,
489554, Cambridge Studies in Medieval Life and Thought, 4th ser. 33 (Cambridge, 1997),
pp. 2467.
Collatio cum Maximino Arianorum episcopo 1.2, PL 42, col. 710; Augustine, In Iohannis evangelium
tractatus 40.7, ed. R. Willems, CCSL 36 (Turnhout, 1954), p. 354; Augustine, Sermo Morin
Guelferbytanus 17.4, PLS 2, col. 584; and McLynn, From Palladius to Maximinus, pp. 4867.

160

The old ruling class

This is an issue that we will touch on again, but for now it points a
way through the tangles of Victor of Vitas polemical language to a space
where religious accommodation between Vandals and Romans (like the
poets of the Latin Anthology) may have been possible. Of course, the
secular poets wrote from a different point of view from that of Victor.
The manuscripts are unanimous that Victor was a bishop, though he does
not yet appear to have held that office at the time he was writing, if he
ever held it. He was, however, well-versed in the Christian scriptures,
perhaps as a priest of the Carthaginian church, and to all appearances an
eyewitness to much of what he describes.125 His text was revised in 488
or 489, but Victor seems to have been writing during the worst phase
of Arian-Nicene conflict in Africa, just before the death of Huneric in
December 484.126 The horrors of violence are most vividly conveyed
by its witnesses, and no other late antique author records the sufferings
endured by the Nicene church at the hands of the Vandal regime in such
glowing detail as Victor. But Victor also wrote perhaps two generations
before the poet Luxorius and his circle. The historians perceptions of
the Vandals were coloured by experiences that were inaccessibly remote
to the poet. The disparities between the two authors visions of late
antique North African society were undoubtedly affected by real change
over time, by their differing religious and secular points of view, and
quite possibly by a conscious distortion of the facts on one or both of
their parts. The real question, then, is not so much why Victor and
Luxorius present such different pictures of North African society under
the Vandals, but rather what the Vandals sought to achieve through
their religious policies, what methods the barbarians employed, and how
Africans as a whole reacted to the persecution.
3.1. Methods
The techniques of religious coercion deployed by the Vandal kings were
closely modelled on imperial methods for the suppression of heresy; only
now it was homoian Arianism that was considered orthodox and Nicene
consubstantialism that was deemed heterodox.127 Committed adherents
of Nicaea experienced and denounced the Vandal Arianizing policy as
a persecution, but brutal as it undoubtedly could be it does not
for the most part seem to have been a bloodbath. We do occasionally
125
126
127

Courtois, Victor de Vita, pp. 511; Shanzer, Intentions and Audiences, pp. 2728; Howe,
Vandalen, pp. 61119.
Courtois, Victor de Vita, pp. 1617 and Howe, Vandalen, pp. 3860.
See in general E. Fournier, Victor of Vita and the Vandal Persecution: Interpreting Exile in
Late Antiquity, Ph.D. diss., University of California, Santa Barbara (2008).

161

Staying Roman
hear in our sources about executions for specifically religious reasons,
but they are extremely rare.128 Indeed, for the most part the Vandal kings
seem to have sought to break the Nicene ecclesiastical hierarchy through
exile, the redistribution of church property to the Arian establishment,
and the elimination of bishoprics through attrition. But the Vandal kings
also sought to create a court (and perhaps an army) that was devotedly
Arian. To that end, they forbade Trinitarians from holding office, and
on occasion threatened the recalcitrant with loss of property and status, public humiliation, and even execution in an effort to bring them
around.
Even during the most intense phase of the persecution, the tool most
commonly wielded against the Nicene bishops appears to have been
internal exile. Banishment from the territories under Vandal control
seems to have been the exception rather than the rule, though, as we
have seen, this was to be the fate of two bishops of Carthage, Quodvultdeus and Eugenius (see above, Chapter 2.1.3). Yet Huneric, remembered
in the Nicene sources as the worst of the persecutors, is never said to have
banished bishops from the Vandal kingdom. His large-scale deportations
focused on the kingdoms desert fringes and Mediterranean islands. We
have seen this already in the kings exile of over four thousand bishops,
deacons, monks, and lay people to the Chott el Hodna region in Mauretania Sitifensis; if Victor of Vita has not simply restructured his narrative
to make one event look like two, in the spring of 484 Huneric again sent
the Nicene episcopate into exile, this time both on the African mainland
and in Corsica.129 In the early sixth century, Thrasamund, too, exiled
sixty or more Nicene bishops to Sardinia, including Fulgentius of Ruspe;
and after the death of Bishop Deogratias of Carthage in 457, Geiseric
was said to have exiled all of the junior clergy from the metropolis.130
These exiles could last widely variable lengths of time. Hunerics deportation of Trinitarian bishops in 484 appears to have been the shortest,
for the king died in that same year and his successor Gunthamund was
said immediately to have recalled the exiles upon assuming the throne in
December.131 By contrast, the Carthaginian priests and deacons exiled by
128

129

130
131

Prosper, Chronicon s.a. 437, pp. 4756 with Antoninus Honoratus, Epistola consolatoria ad Arcadium, PL 50:56770; Vict. Vit. 3.41, p. 92 and perhaps ibid., 3.267, pp. 845; Passio septem
monachorum 14, p. 113; Shanzer, Intentions and Audiences, pp. 2824.
Vict. Vit. 2.2637, 3.1520, 3.348, pp. 338, 7881, and 8991; see also Vict. Tonn. s.a.
479.50, p. 16, whose fifth-century chronology is singularly unreliable: the exile to Chott el
Hodna probably took place after Eugenius of Carthages ordination in 480/1.
Thrasamund: V. Fulg. 1718, pp. 8791 and Vict. Tonn. s.a. 497.78, p. 24. Geiseric: Vict. Vit.
1.51, p. 22.
Exiles: Vict. Vit. 3.34 and 3.435, pp. 89 and 934. Recall: Vict. Tonn. s.a. 479.512, p. 16,
followed by Isidore of Seville, Historia Vandalorum 80, ed. and trans. C. Rodrguez Alonso,

162

The old ruling class

Geiseric were only recalled in 474, seventeen years after Bishop Deogratias death, through the intervention of the East Roman emperor Zeno.132
The bishops sent by Thrasamund to Sardinia similarly seem to have spent
about fifteen years in exile. As a group they too were allowed to return
to Africa only on the kings death and the accession of his successor
Hilderic, though Fulgentius of Ruspe was briefly recalled to Carthage to
answer Thrasamunds questions about the Nicene confession.133 Indeed,
exile could also be an individual affair. This would seem to have been
the experience of Eugenius of Carthage during his first exile, when
he was sent by Huneric to the desert frontier of Byzacena; the bishops
relegated to Mouzaaville and Cartennae along the Mauretanian coast,
and perhaps those consigned to Madauros in Numidia, were probably
dispatched singly as well.134 Bishops who saw fit to express their opposition to the Vandal regime through homiletic references to Pharaoh,
Nebuchadnezzar, or Holofernes were also sent into exile (see below,
section 3.2).135 Similarly, when Bishop Valerius of Avensa refused to
comply with Geiserics order requisitioning the liturgical vessels and
libraries of the Nicene churches of Africa Proconsularis, that bishop too
was driven from his city.136
The case of Valerius highlights another aspect of the Vandals Arianizing policy: the transfer of Nicene church property to the Arian ecclesiastical establishment. By 437, when the Vandals were still confined to
Numidia and Mauretania Sitifensis, Geiseric had already requisitioned a
number of Nicene basilicas in these provinces.137 The Vandal king seized
numerous churches from the Nicene community of Carthage, too, after
the city fell to his army in 439. These included the most important
cult sites in the African metropolis: the cathedral, the suburban Basilica
Maiorum (which housed the relics of Sts Perpetua and Felicitas), and two
other suburban churches important to the cult of St Cyprian, one built on

132
133
134

135
136
137

Las Historias de los godos, vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica y traduccion,
Fuentes y estudios de historia leonesa 13 (Leon, 1975), p. 302. Proc. BV 1.8.7, 1:346 indicates
that Gunthamund continued and intensified the persecution.
Vict. Vit. 1.51, p. 22.
V. Fulg. 201 and 257, pp. 99107 and 1217; Vict. Tonn. s.a. 523.106, p. 34; PCBE 1:50713,
s.n. Fulgentius 1, at pp. 51011.
Eugenius: Vict. Vit. 3.34 and 3.434, pp. 89 and 934 with Courtois, Victor de Vita, p. 48 n.
190; see also Notitia, Proc. 1, p. 117. Other bishops: see above, Chapter 2.3, nn. 143, 145, and
1478.
Vict. Vit. 1.22, pp. 1011.
Vict. Vit. 1.3940, pp. 1718. Under Huneric, exile was not always far from a bishops see:
V. Fulg. 3, p. 21.
Prosper, Chronicon s.a. 437, p. 475: Gisiricus rex Wandalorum, intra habitationis suae limites
volens catholicam fidem Arriana impietate subvertere.

163

Staying Roman
the place where he was martyred and the other where he was buried.138
These churches remained in Arian hands until the conquering Byzantine
armies approached Carthage in September of 533, at which point the
Arian priests fled their preparations for the annual festival in honour of St
Cyprian, and Nicene priests took control of the basilicas once more.139
Something similar may have happened in Hadra (class. Ammaedara), in
western Byzacena. There, the basilica of Melleus would appear to have
been given over to Arian use under the Vandal regime, and returned to
the Trinitarians after the Byzantine reconquest. At least an inscription in
the church commemorates the eternal rest of one Victorinus, in peace,
a bishop (Victorinus episcopus in pace), to which a later hand added: of
the Vandals (Vandalorum).140 Indeed, by the 530s, a considerable amount
of land and goods throughout Africa must have been in the hands of the
Arian ecclesiastical establishment, for, in the course of his reign, Huneric
too is said to have confiscated churches and properties belonging to the
Nicene community, and to have given them to the Arians.141
Despite the exiles and the confiscations, however, most Vandal kings
seem to have been generally content to let bishoprics become vacant
gradually, as their bishops died of natural causes. Once a Nicene see fell
empty, for whatever reason, the king would simply forbid the ordination of a new bishop. Carthage in particular suffered from this policy.
After Geiseric banished Quodvultdeus, the metropolis was deprived of a
Nicene bishop until an embassy from the West Roman emperor Valentinian III secured the ordination (on 24 October 454) of Deogratias.142
After Deogratias death, the see was allowed to remain vacant for another
138

139
140
141
142

Prosper, Chronicon s.a. 439, p. 477; Vict. Vit. 1.9 and 1.1416, pp. 56 and 78; see also V. Fulg.
1, p. 11, a private house given to Arian priests. On the basilicas dedicated to St Cyprian at
Carthage, see Y. Duval, Loca sanctorum Africae: le culte des martyrs en Afrique du IVe au VIIe si`ecle,

2 vols., Collection de lEcole

francaise de Rome (Paris, 1982), 2:6757 and L. Ennabli, Inscriptions
chretiennes de Sainte-Monique, pp. 1216. Vict. Vit. 1.9, p. 5 may refer to one basilica housing
relics of the Scillitan martyrs, Sts Perpetua and Felicitas, and St Celerina, not three separate
basilicas: Courtois, Victor de Vita, pp. 423, who, however, postulates a founder named Celerina.
St Celerina was an African martyr (Martyrologium Hieronymianum 3 Feb., ed. H. Quentin and
H. Delehaye, AASS Nov. Propylaeum (Brussels, 1931), p. 76), but her name does not appear in
the Kalendarium carthaginiense, PL 13:121930, nor is she mentioned in Duval, Loca sanctorum.
The cults of the Scillitani and of Perpetua and Felicitas, on the other hand, appear to have
been associated in at least one other place in Carthage: Duval, Loca sanctorum, 1:710, no. 3
(Carthage); see also ibid., 2:6912. The identification of the basilica of the Scillitans with that
of the basilica Celerinae is accepted by L. Ennabli, Carthage: une metropole chretienne du IVe a` la fin
du VIIe si`ecle (Paris, 1997), pp. 324.
Proc. BV 1.21.1725, 1:4034, of one of the suburban churches of St Cyprian.
Duval, Hadra 1:878, no. 58, first announced in N. Duval, Rapport preliminaire sur les travaux
effectues a` Hadra en sept.-oct. 1967, Africa 34 (196970), p. 204.
Vict. Vit. 3.2, p. 72, and Vict. Tonn. s.a. 466.30, p. 11.
Vict. Vit. 1.24, p. 11 with Prosper Tiro add. s.a. 454, p. 490; PCBE 1:2713, s.n. Deogratias 1,
at p. 271.

164

The old ruling class

twenty-four years, until the election of Eugenius again a concession to
a Roman emperor, this time Zeno.143 Carthage once more languished
without a Nicene bishop for perhaps eighteen years after Eugenius death
in 505, before Boniface was elected in 523, in the reign of Hilderic.144
The sees of exiled provincial bishops were allowed to remain vacant for
long periods as well.145 After the death of Deogratias, Geiseric explicitly forbade the ordination of Nicene bishops for vacant sees in Africa
Proconsularis.146 Thrasamund was later to reimpose the ban and extend
it to the rest of the kingdom. In the sixth century, however, the sitting bishops chose to defy the king and brought down his wrath upon
the ecclesiastical hierarchy of the kingdom. Fulgentius still an abbot,
though apparently also already chosen by some to be bishop of Ruspe
went into hiding for a time, while the primate of Byzacena, who had performed most of the illicit ordinations, was arrested, brought to Carthage,
and exiled to Sardinia.147
The Vandal kings thus seem generally to have followed a remarkably
patient policy with respect to the Nicene bishops. This was at least in part
because of the canniness of Geiserics Arian patriarch, Jucundus, about
the importance of not creating new martyrs.148 But many Nicene bishops
were presumably drawn from the same families as the secular elite, and the
pragmatic need not entirely to alienate these bishops brothers, cousins,
and friends presumably also played a role in the formation of Vandal
policy. When Fulgentius of Ruspe, a Nicene monk from a prominent
Romano-African family, was set upon and beaten by the toughs of an
Arian priest named Felix, the story made it all the way to Carthage. There
an Arian bishop, a friend of Fulgentius family, offered to aid the monk
should he decide to lodge a complaint against Felix. Fulgentius, we are
told, turned the offer down out of Christian love, but especially since it
may scandalize many little ones if I, a Catholic and a monk, though still
a sinner, demand the trial of an Arian bishop.149 Even so, the protection
offered by social connections continued to matter. Indeed, one of the few
executions of Nicene Romano-Africans that Victor of Vita attributes to
the reign of Geiseric involved escaped slaves. These slaves had fled their
143
144
145
147
148
149

Vict. Vit. 2.2, p. 24, with ibid., 2.6, p. 26.

Vict. Tonn. s.a. 505.86, p. 27 (death of Eugenius) and ibid., s.a. 523.106, p. 34 (election of
Boniface).
146 Vict. Vit. 1.29, p. 13.
Vict. Vit. 1.23, p. 11.
V. Fulg. 1314 and 17, pp. 6975 and 879. See also Isidore, Historia Vandalorum 81, pp. 3024,
who again follows Vict. Tonn. s.a. 497.78, p. 24 (exile of 120 Nicene bishops to Sardinia).
Vict. Vit. 1.44, p. 19.
V. Fulg. 67, pp. 3545. The quotation is ibid., 7, p. 45: maxime quia multos parvulos scandalizare
poterit, si episcopi ariani judicium, qualiscumque peccator, tamen catholicus et monachus
quaeram.

165

Staying Roman
Vandal master, been recaptured, and were exiled to Moorish territory
beyond the borders of the Vandal kingdom, where they attempted to
spread Nicene Christianity among their Moorish hosts. Only then were
they put to death.150 Nicene missionary activity was probably anathema
in the Vandal kingdom; but, more importantly, as slaves these missionaries
presumably lacked the kind of social support system that could be invoked
to protect members of the elite like Fulgentius.
In short, then, the grand vision of the Vandal kings seems to have been
that by exerting slow but consistent pressure on the Nicene ecclesiastical
hierarchy the Trinitarian hold on Africa (initially only Africa Proconsularis) would eventually simply fade away. In theory, for as long as their
sees were empty, the absence of Nicene bishops in the Vandal kingdom
would mean an end to Nicene baptisms as well as to the ordination of
new priests and deacons. The faithful would have no choice but to turn
to the Arian church. Over the long term, this policy of attrition could
perhaps have met with success. By the time Victor of Vita came to write
his History of the Persecution in the 480s, he claimed that only three bishops
from the proconsular province were still alive from before the time of
Geiserics initial decree: Quintianus of Urusi (who, as we have seen, had
fled to Macedonia), Paul of Sinnari, and Vincent of Zigga.151 However,
the Nicene community also engaged in secret ordinations, and the policy
seems not to have been consistently enforced anyway.
Vandal efforts to secure the conversion to Arianism of prominent
members of North African society also focused on the secular elite,
particularly those who served in the royal palace. At some unspecified time perhaps in the late 450s or 460s Geiseric decreed that only
Arians could hold offices at the Vandal court.152 The decree was reissued at some point after Huneric became king, though in general such
decrees seem not to have been consistently enforced; and though we are
told that many Trinitarians left royal service to preserve the integrity of
their faith, others certainly continued to serve at the palace.153 At least
sporadically, though, the Vandal kings do appear to have attempted to
compel the conversion of Nicene courtiers through a combination of the
threat of execution on the one hand and calculated social humiliation
on the other. At these moments it seems to have particularly mattered
to the Vandal kings that their close advisers become Arians. Prosper
of Aquitaine records that already in 437 (before the Vandal capture of
150
151
152
153

Vict. Vit. 1.358, pp. 1517.

Vict. Vit. 1.29, p. 13. For Vincents see, see Notitia, Proc. 41, p. 118.
Vict. Vit. 1.43, pp. 1819.
Decree: Vict. Vit. 2.23, p. 32. Continued service: see below, this paragraph.

166

The old ruling class

Carthage) Geiseric had four of his Hispano-Roman counsellors executed
for refusing to convert to Arianism.154 That religion was at least one of
the issues at stake seems to be confirmed by a letter that Bishop Honoratus of Constantina in Numidia (class. Cirta; modern Constantine) wrote
to one of the condemned Spaniards before his execution, comforting
him with the thought that the entire chorus of your martyr predecessors is with you; the martyrs await and defend you, and extend [you] a
crown. . . . Be assured: If you die, you will be a martyr.155 Later, Geiseric also sought the conversion of the Roman public enemy Sebastian
comes, whom he had embraced as an adviser, though in this case the king
was not willing to press the issue to the point of execution.156 Similarly,
Huneric sought to effect the conversion of his proconsul Carthaginis Victorianus. On refusing the kings demand, Victorianus was said to have
been brutally punished and, Victor of Vita tells us, accepted the crown
of martyrdom; but the ambiguity of Victors language makes it impossible to say with any real certainty whether the official was in fact put
to death.157 Equally important to Geiserics policy was the use of social
degradation. A Nicene courtier with the Germanic name Armogast, for
example, was tortured and then condemned to manual labour, first as
a ditch digger in Byzacena and later, as if to greater disgrace (quasi
ad maiorem [sic] obprobrium), as a cowherd outside Carthage, where his
humiliation would be more visible.158 Under Huneric, at least some of
the Nicene courtiers who refused to convert were sent to labour in the
fields outside Utica.159 Earlier, in Geiserics reign, the procurator domus of
prince Hunerics household, Saturus, was deprived of his property and
reduced to beggary. Saturus was also threatened with the sale of his slaves,
the enslavement of his children, and the marriage of his wife to a camel
driver, but Victor does not tell us whether or not Geiseric acted upon
these threats.160
The Vandal kings sought to trap Nicene flies not only with vinegar, but
also with honey, in particular through the lure of property, privilege, and
courtly office. Fulgentius of Ruspes biographer wrote that Thrasamund
now compelled the Catholics through terrors, now attracted [them] with
154
155

156
157
158
159

Prosper, Chronicon s.a. 437, pp. 4756.

Antoninus Honoratus, Epistola consolatoria, col. 567BC: Tecum est omnis chorus martyrum
praecessorum tuorum; exspectant te martyres et defendunt, et coronam extendunt . . . Si mortuus fueris, certus esto, martyr eris.
Vict. Vit. 1.1921, pp. 910. See also Vict. Vit. 1.47, pp. 201, the case of Geiserics chief
pantomime Mascula.
Vict. Vit. 3.27, p. 85 and Shanzer, Intentions and Audiences, pp. 2834, who nonetheless reads
Victorianus as a martyr in the strong sense of the word.
Vict. Vit. 1.436, pp. 1920; the quotation is ibid., 1.44, p. 19.
160 Vict. Vit. 1.4850, pp. 212.
Vict. Vit. 2.1011, pp. 278.

167

Staying Roman
promises.161 Procopius, too, was aware that Thrasamund attempted to
persuade secular aristocrats to convert with the enticement of wealth,
honours, and offices.162 However, loss of property and status continued
to await those who refused to convert. As we have seen, the assets of
office-holders such as Saturus were not safe from Geiserics attempts
to secure the conversion to Arianism of the members of his and his
sons households. Huneric later decreed that Nicene courtiers were to
receive neither their provisions nor their wages (neque annonas neque
stipendia).163 Some were even deprived of their possessions and exiled to
Sicily or Sardinia.164 Perhaps significantly, it seems to have been in the
reign of Huneric that Fulgentius of Ruspe abandoned his position in
the civil administration of the kingdom, though he did so to become a
Nicene monk hardly a wholly safe class of men in Hunerics Africa.165
Fulgentius himself indicates that, under Thrasamund, Trinitarians were
not allowed to serve as soldiers and that Nicene ship captains whose role
in the dissemination of information rendered them a politically suspect
group were also deprived of their ships.166
In general, though, the Vandal kings religious policy with respect to
the army and ship captains is difficult to detect. The bias of our sources
ensures that we can most easily see the kings efforts to secure lay converts to Arianism at the court and in the administration. However, we
seldom hear of Romano-African aristocrats who confined themselves to
private life and nevertheless ran afoul of the Vandal persecution. Indeed,
the only real exception to this rule is the period between June and
December of 484. In a law of 25 February of that year, Huneric turned
against Africas Nicene population the laws through which the Roman
emperors had sought to bring earlier heterodox communities to heel.
Hunerics edict cited imperial legislation against heretics that closed their
churches, forbade their baptisms and ordinations, and ordered the burning of their books. Lay heretics generally had been denied the right to
bequeath or inherit property, while office-holders were stripped of the
privileges of their rank, and fined; if they still refused to renounce their
dissident beliefs, they could be beaten, deprived of their property, and
driven into exile. Overseers and leaseholders on imperial estates found to
be harbouring heretics were to suffer the penalty of the deviants themselves. From June of 484 until the Hunerics death in December of that
same year, this treatment was to be directed in turn against any Nicene
161
162
165
166

V. Fulg. 20, p. 99: Catholicos nunc terroribus cogebat, nunc promissionibus invitabat.
163 Vict. Vit. 2.10, p. 27.
164 Vict. Vit. 2.23, p. 32.
Proc. BV 1.8.910, 1:3467.
V. Fulg. 3, pp. 213. On the persecution of monks, see the Passio septem monachorum, pp. 10814.
Fulgentius of Ruspe, Abecedarium ll. 24753, p. 884.

168

The old ruling class

Christian who refused to convert to Arianism.167 Beatings, torture, dismemberment, and desert exile were the order of the day. This was to be
persecution in earnest. Even during its most violent phase, the victims
we hear about were still drawn primarily from the ranks of the African
secular and ecclesiastical elite: the proconsul of Carthage and a number
of secular aristocrats, the wife of one of the kings cellarers, two doctors,
two merchants, a bishops daughter, and seven monks, as well as the usual
complement of bishops and members of the Carthaginian clergy.168 But
now the victims included women and children as well as prominent men,
from the provinces as well as from the capital. At least, Victor of Vita tells
of the persecution of seemingly secular individuals centred in the cities
of Thuburbo Maius and Culusi in Proconsularis, as well as Thambaia in
Byzacena.169
As far as we can tell, however, the terrors of these seven months
were the exception rather than the rule. From Geiserics reign to that of
Thrasamund, the essential tactics of the Vandal Arianizing policy seem
to have remained the same: awards awaited those who would convert,
while punishments were meted out against those who would not. In
general, it seems to have been technically illegal for those of Nicene
confession to hold office in the Vandal kingdom; under Thrasamund,
at least, the same was true of serving in the army and operating as a
shipper. In practical terms these laws were not always an impediment
to Nicene service in the Vandal administration, though religion did
sporadically emerge as an issue between the Vandal kings and their leading
courtiers. Punishments for non-conversion could include loss of property
and status, exile, humiliation, and even torture, but rarely execution. By
contrast, the lure of wealth and office awaited those who were willing
to convert. In 484, similar lures were apparently used to secure the
conversion of Nicene monks, and presumably nuns and clergy as well.170
167

168

169
170

Vict. Vit. 3.314, pp. 728; cf. esp. CTh 16.5.52 (ad 412), pp. 8723, but also CTh 16.5.40 (ad
407), 16.5.43 (ad 408), 16.5.45 (ad 408), 16.5.46 (ad 409), 16.5.54 (ad 414), 16.5.65 (ad 428),
16.6.4 (ad 405), pp. 86770, 8734, 8789, and 8812, and Constitutio Sirmondiana 12 (ad 407),
in CTh, pp. 91617; Fournier, Victor of Vita, p. 75 n. 3. See also Vict. Tonn. s.a. 479.50,
p. 16 (per totam Africam) and Proc. BV 1.8.34, 1:345, who first mentions specifically religious
persecution under Huneric.
Proconsul of Carthage: Vict. Vit. 3.27, p. 85; secular aristocrats: ibid., 3.256, pp. 835; cellarers
wife: ibid., 3.33, pp. 889; doctors: ibid., 3.24 and 3.501, pp. 83 and 967; merchants: ibid.,
3.41, p. 92; bishops daughter: ibid., 3.24, p. 83; bishops: ibid., 3.34 and 3.436, pp. 89 and 935;
Carthaginian clergy, esp. the deacon Muritta: ibid., 3.348, pp. 8991; monks: ibid., 3.41, p. 92
and Passio septem monachorum, pp. 10814. See also Vict. Vit. 3.49, p. 96, the forced rebaptism
of an aristocratic boy.
Thuburbo Maius: Vict. Vit. 3.25, pp. 834. Culusi: Vict. Vit. 3.26, pp. 845. Thambaia: Vict.
Vit. 3.28, p. 86. For the locations, see Courtois, Victor de Vita, pp. 4650, with map on p. 45.
Passio septem monachorum 8, p. 110.

169

Staying Roman
In general, the pressures applied to the ecclesiastical hierarchy were also
much the same: the confiscation of church property, exile, loss of status,
humiliation. At least one group of Nicene monks is said to have been
executed,171 though to judge from our sources the ultimate sanction was
seldom applied to bishops of the Nicene church. We understand less well
the pressures exerted on the masses to secure their conversion (though
see below, section 3.2), but the ultimate goal of the Vandal rulers seems
to have been the creation of a theologically pure kingdom, devoted to
the Arian confession of Christianity.
It is perhaps worth noting as an aside that the Arianizing pressures
applied by the Vandal regime were not in the end sustainable. Hilderics
cessation of this approach upon his succession to the throne in 523
marked a dramatic departure from the policies of his predecessors, and
one which Thrasamund seems to have anticipated and attempted to forestall. According to Victor of Tonnena, Thrasamund had Hilderic swear
not to reopen the Nicene churches, allow the celebration of the Nicene
liturgy, or restore their privileges to African Trinitarians. Thrasamunds
efforts were in vain. Not only did Hilderic recall the exiled bishops from
Sardinia, he called for the ordination of a Nicene bishop for Carthage
and, indeed, the ordination of bishops throughout the North African
provinces.172
3.2. New light from the African hinterland
Our understanding of the African response to the Vandal Arianizing
policy is filtered entirely through the eyes of the Nicene clergy and religious who were its victims. No lay African voice mentions persecution,
and indeed the secular literary works that survive from the Vandal kingdom are strikingly uninterested in the institutional church or questions of
theology.173 Even so, Victor of Vitas perspective is only one among many.
Considering the African response to the Vandal Arianizing policy more
broadly raises a host of questions about the attitude of Nicene authors
(other than Victor) towards the Vandals, how these attitudes changed
over time, and the success of pro-Arian tactics in winning converts. This,
in turn, forces us to rethink why Africa produced such a staggering volume of anti-Arian religious literature in the fifth and sixth centuries, and
also to reflect on the political implications of this literary corpus. Finally,
broadening our scope to consider anonymous sermons from the Vandal
171
172
173

Passio septem monachorum 14, p. 113.

V. Fulg. 25, p. 121; Vict. Tonn. s.a. 523.106, p. 34. The career of the newly ordained Bishop
Boniface of Carthage is discussed in PCBE 1:15961, s.n. Bonfatius 26.
Hays, Romuleis Libicisque Litteris, pp. 1312 and Miles, Anthologia Latina, passim.

170

The old ruling class

period allows us to begin to understand the non-elite experience of the
persecution in the African hinterland.
Victor was not alone in his hostility towards the Vandals. As F. M.
Clover has observed, in 438, the Donatist editors of the Liber genealogus
demonstrated that numeric values of the Greek letters of Geiserics name
($) equalled 666, the number of Antichrist.174 From the safe
distance of his Campanian exile, Quodvultdeus of Carthage too seems
to have viewed the Vandal occupation of Africa and the Arian efforts
to overthrow Nicene Christianity as a sign of the apocalypse.175 Over
time these apocalyptic visions of Vandal rule seem to have given way to
the perception that we find in Victor of Vitas history, of Arian Vandal
ascendancy as a period of captivity for the Nicene church. The sufferings
of the Israelites in Egypt and Babylon seem particularly to have resonated
with fifth-century African Trinitarians. As we have seen, Geiseric exiled
bishops who made reference to Pharaoh and Nebuchadnezzar in their
sermons. Quodvultdeus of Carthage was among their number.176 The
anonymous African bishop whose homilies have been wrongly attributed
to Fulgentius of Ruspe mentioned the same figures in his sermons to
an audience in the hinterland of the Vandal kingdom, though we have
no way of knowing whether this preacher was later exiled.177 Another
anonymous African bishop sought to fortify his congregation with the
thought that if this temporal captivity of the barbarian host is so cruel and
bitter, when it is escaped by flight or bought off with money, or at worst
ended by death, of what sort shall be that eternal captivity which is not
ended by death, but suffers torture in eternity among the damned?178
The worldly sufferings of Nicene Christians under the Vandal regime
might be terrible to endure, but they were nothing in comparison to the
eternal torment incurred by apostasy and conversion to Arianism.
174
175

176
177

178

Liber genealogus 616 and 618, pp. 1945 (for the date, see ibid., 628b, p. 196); Clover, Carthage
and the Vandals, p. 4.
Quodvultdeus, Liber promissionum et praedictorum Dei 4.8.1516, ed. R. Braun in CCSL 60:200
1; see also ibid., 5.7, pp. 1945. On Quodvultdeus, see R. Gonzalez Salinero, Poder y conflicto
religioso en el norte de Africa: Quodvultdeus de Cartago y los vandalos, Graeco-Romanae Religionis
Electa Collectio (GREC) 10 (Madrid, 2002).
Quodvultdeus: De cataclysmo 3.1224 and 5.39, ed. R. Braun in CCSL 60:41112 and 41415
(Pharaoh) and De tempore barbarico 1.3.18, ed. Braun in CCSL 60:427 (Nebuchadnezzar).
Pseudo-Fulgentius, Sermones, PL 65:855954, esp. Sermo 8, col. 868 (Nebuchadnezzar) and
Sermones 1315 and 78, cols. 8745, 8778, and 950 (Pharaoh). For the identification as African
of these and all other anonymous sermons discussed here, see Dossey, Christians and Romans,
pp. 36678.
Pseudo-Augustine, De navitate domini 1 (sermo Mai 117), PLS 2, col. 1222: Et si ista temporalis
captivitas hostium barbarorum tam crudelis et amara est, quae aut per fugam labitur, aut pecunia
redimitur, aut postremo finitur morte, qualis erit illa aeterna captivitas, quae nec morte finitur,
sed in aeternum apud inferos cruciat? See also pseudo-Augustine, De natali domini (Homilia
Vindobonense 5a), ed. H. Barre, PLS 4, col. 1911.

171

Staying Roman
Indeed, throughout the Vandal period, Nicene authors seem to have
been anxiously aware of the threat that Arian Christianity posed to the
Nicene church in Africa. In one of the sermons he preached before his
exile from Carthage, Quodvultdeus warned his congregation, Beware
the Arian pestilence, o most beloved; do not let them separate you from
Christ by promising worldly things, do not let them despoil you of the
faith for the sake of a tunic.179 As early as the 440s or 450s, however,
even Quodvultdeus had to concede that (through trickery and the lure of
temporal power) the Arians had succeeded in winning many converts.180
Already by 488 Pope Felix II was worried about the situation in Africa and
established conditions for the readmission into communion of apostate
bishops, priests, deacons, clerics, monks, and nuns, as well as members of
the laity181 a sure sign of Hunerics achievement in his reign of terror of
484. This was not just an abstract concern. Elpidoforus, one of Hunerics
torturers, was a former Nicene Christian who had received baptism at
the Carthaginian basilica of Faustus before converting to Arianism.182
A certain Teucharius, once a Nicene lector, had converted as well, and
thus was able to save twelve of his former pupils from exile in the
persecution of 484.183 In the early sixth century the Arian preacher Fastidiosus was said previously to have been a Nicene monk and priest.184
Gregory of Tours alleges that even the African Arian bishop Revocatus
was an apostate Trinitarian.185 Indeed, Yves Moderan has convincingly
argued that nearly 20 per cent of the African episcopate converted to
the Arian confession in the persecution of 484.186 Thrasamunds court
poet Felix asked for a position in the church, raising at least the possibility that he too was a convert.187 In any case, as Leslie Dossey has
shown, Romano-African conversion to Arianism probably continued
into the sixth century, and even after the Byzantine reconquest of Africa
the reconciliation of the lapsed posed a problem for Pope Agapetus
(ad 5356).188
179

180
182

183
185
186
188

Quodvultdeus, De tempore barbarico 1.8.7, p. 436: Cauete, dilectissimi, arrianam pestem; non
uos separent a Christo terrena promittendo, propter tunicam non uos exspolient fide. See also
pseudo-Fulgentius, Sermo de Simbolo, ed. G. Morin, PLS 3, col. 1371.
181 Felix II (III), Ep. 13, pp. 25966.
Quodvultdeus, Liber promissionum 4.5.7, p. 194.
Vict. Vit. 3.347, pp. 8991; the basilica of Faustus seems to have become the Nicene cathedral
church of Carthage following the requisition by the Vandals of the basilica Restituta: see, e.g.,
Prosper Tiro add. s.a. 454, p. 490 and Vict. Vit. 1.25, 2.18, and 2.4750, pp. 12, 301, and 424.
184 Fulgentius of Ruspe, Contra sermonem Fastidiosi 1.1, pp. 2834.
Vict. Vit. 3.39, pp. 912.
Gregory of Tours, Historiae 2.3, p. 45.
187 AL 248, pp. 1868.
Moderan, Notitia provinciarum, pp. 17281.
Collectio Avellana 857, ed. O. Gunther, in Epistulae imperatorum, pontificum, aliorum inde ab a.
ccclxvii usque ad a. dliii datae Avellana quae dicitur collectio, CSEL 35/1 (Vienna, 1895), pp. 32833;
Dossey, Last Days of Vandal Africa, pp. 11112.

172

The old ruling class

In fact, the overwhelming impression left by the Christian literature
of Vandal Africa is that Trinitarians and Arians were engaged in incessant
debate with one another about the nature of the faith. Very little evidence
survives to portray the Arian side of that debate, but African Trinitarians occasionally cast their works as imaginary dialogues in which they
confounded their Arian opponents.189 Cerealis of Castellum Ripensis
placed his Trinitarian libellus in the dramatic context of the council that
Huneric convened at Carthage in 484 to debate the faith. There, Cerealis
writes, he met an Arian bishop named Maximinus. In a brief opening
dialogue, this Maximinus poses a series of questions that structure the
substance of Cerealis book.190 Two pseudo-Augustinian works imagined
past debates between the great bishop of Hippo and different Arian opponents, while Vigilius of Thapsa pulled the three great heresiarchs Arius,
Sabellius, and Photinus out of time altogether, and had them debate
the orthodox Athanasius.191 Such exchanges, moreover, were not simply
rhetorical fictions composed by and exchanged among African bishops.
The preacher of an anonymous Vandal-era sermon took as his point of
departure a conversation he had with an Arian.192 Another such sermon
explicitly rebuts Arian objections to Nicene theology.193 Quodvultdeus
of Carthage preached two long sermons against religious dissenters of all
stripes, including Arians, pagans, and Jews.194 These preachers clearly felt
that the danger apostasy posed to their own congregations was real; the
fact that their sermons were written down and collected is a testament
to the persistent significance of that danger.
The same might be said of a handful of anonymous texts that envisaged questions or objections which Arians might be expected to pose
to Trinitarians and which were duly dispelled through reference both
189

190
191

192
193
194

For the Arian side, see Dossey, Last Days of Vandal Africa and the African Arian homilies in
the late fifth- or early sixth-century Verona, Biblioteca Capitulare, MS LI (49) (CLA 4.504):
Collectio Arriana Veronensis, ed. R. Gryson, in Scripta Arriana Latina, CCSL 87 (Turnhout, 1982),
pp. 1145, here De sollemnitatibus, pp. 4792; see also the Arian Contra Iudaeos and Contra
paganos, ibid., pp. 93117 and 11840, perhaps by the same author. On the African origins of
these sermons, see Dossey, Christians and Romans, p. 370. On the perceived threat of Arianism
among African writers in general in this period, see Howe, Vandalen, pp. 14753.
Cerealis of Castellum Ripensis, Libellus contra Maximinum Arianum, PL 58:757; see also Notitia,
M. Caes. 119, p. 131. On this assembly, see Vict. Vit. 2.383.15, pp. 3879.
Collatio beati Augustini cum Pascentio ariano, cols. 115662; Contra Felicianum Arianum de unitate
Trinitatis, PL 62:33352; Vigilius of Thapsa, Contra Arianos, Sabellianos, Photinianos dialogus, PL
62:179238, esp. col. 180C.
Pseudo-Augustine, Sermo 245.2, PL 39:2196: Sed dicit mihi haereticus: Ergo si unum sunt,
omnes sunt incarnati.
Pseudo-Augustine, Sermo 246, PL 39:2198200.
Quodvultdeus, Contra Iudaeos, paganos et Arrianos, ed. R. Braun in CCSL 60:22558 and Quodvultdeus, Adversus quinque haereses, pp. 261301.

173

Staying Roman
to Scripture and to reason. Other manuscripts similarly collected biblical testimonies backing Nicene doctrine on the Trinity, the equality of
the Father and the Son, and the deity of Christ and the Holy Spirit.195
Like Augustine had against the Donatists, Fulgentius of Ruspe wrote a
rhyming ABC against the Arians.196 These works resemble nothing so
much as handbooks for the faithful to use in their own debates with the
Arians. And use such handbooks the faithful probably did. In the 520s,
Fulgentius received a letter from a young man named Donatus who had
studied secular letters but not much scripture, and who therefore did
not know how to respond when a heretic told him that the Father was
greater than the Son; Donatus wanted to be able to answer the challenge
in the future.197 Earlier in the century, a certain Felix similarly wanted
to know how to defend Trinitarian doctrine against the errors of the
heretics, and wrote insistently to Fulgentius about the question.198 A
third of the bishops followers sent him an Arian sermon that attacked
Nicene ideas about the Trinity, and asked Fulgentius for a refutation.199
For the most part these requests presuppose ongoing, probably informal, debates in the Vandal kingdom across confessional lines, and
among clergy and laity alike about the proper character of Christian
confession.
The sheer volume of the extant anti-Arian literature from fifth- and
sixth-century Africa is extraordinary. Still more such work was once
written but no longer survives: according to Gennadius of Marseilles,
writing in the 490s, the fifth-century African bishops Asclepius, Victor
of Cartenna, and Voconius of Castellanus also wrote treatises against the
Arians, as did Eugenius of Carthage.200 What is most surprising about the
anti-Arian literature of Vandal Africa, however, is its remarkably public
nature. Fulgentius of Rupse dedicated to Thrasamund not one but two
treatises aimed at effecting the kings conversion to the Nicene faith.201
Bishop Eugenius of Carthage produced an exposition of the Nicene faith
195

196
197
198
199
200
201

Pseudo-Vigilius of Thapsa, Contra Varimadum, ed. B. Schwank, CCSL 90:1134 and pseudoVigilius, Solutiones obiectionum Arianorum, PL 62:46972; pseudo-Augustine, Solutiones diuersarum
quaestionum ab haereticis obiectarum, ed. Schwank, CCSL 90:141223; also the anonymous Testimonia de Patre et Filio et Spiritu Sancto, ed. D. de Bruyne, CCSL 90:22733 and Liber de Trinitate,
ed. J. Fraipont, CCSL 90:23960.
Fulgentius of Ruspe, Abecedarium and Shanzer, Intentions and Audiences, pp. 2856.
Fulgentius, Ep. 8, 1:25773.
Fulgentius of Ruspe, Liber de Trinitate ad Felicem 1.1, p. 633.
Fulgentius, Ep. 9, 1:27780; Fastidiosus, Sermo, ed. J. Fraipont in CCSL 91:2803; and Fulgentius
of Ruspe, Contra sermonem Fastidiosi, pp. 283308.
Gennadius, De viris inlustribus, 74, 789 and 98, pp. 878 and 956; see also ibid., 97, p. 95
(Cerealis).
Fulgentius of Ruspe, Dicta regis Trasamundi et contra ea responsionum liber unus, ed. J. Fraipont in
CCSL 91:6594 and Fulgentius, Ad Trasamundum regem libri iii, ed. Fraipont in CCSL 91:95185.

174

The old ruling class

at the behest of Huneric.202 Victor of Cartenna is said to have sent his
long book against the Arians to Geiseric.203
With the transfer of power to barbarian rulers, the earlier flood of
anti-Arian treatises seems to have slowed to a trickle elsewhere in the
western Mediterranean. In the long fourth century, no fewer than nine
Nicene authors writing in Italy, Gaul, Africa, and Spain had all sought
to confound the followers of Arius, many of them in works that still
survive.204 All of these regions experienced some form of Arian domination thereafter, in the fifth and sixth centuries. Yet anti-Arian material
seems not to have featured prominently in the literary output of fifthand sixth-century Italy.205 For Spain, Isidore of Seville mentions that
his brother, Bishop Leander (ad 579c.600), had written against the
Arians.206 So had Severus of Malaga, a contemporary of Leander, who
wrote a polemical treatise against Bishop Vincent of Saragossa, himself a
convert from the Nicene faith to Arianism.207 Neither of these treatises
survives, but the 580s were heady days for Trinitarians in Visigothic Spain.
Over the course of that decade, Leander accomplished the conversion
of two Visigothic princes: first Hermenegild, who led an unsuccessful rebellion against his father, Leovigild; then Hermenegilds younger
brother Reccared, who succeeded Leovigild and officially converted the
kingdom to the Nicene faith.208 At least some of the anti-Arian literature of fifth- and sixth-century Gaul seems to have been produced
202

203
204

205
206

207

208

Gennadius, De viris inlustribus 98, p. 95; perhaps the same Liber fidei catholicae preserved in Vict.
Vit. 2.56101, pp. 4671, where it is attributed to the Numidian bishops Januarius of Zattara
and Villaticus of Casae Medianae, and to the Byzacenan bishops Boniface of Foratiana and
Boniface of Gatiana.
Gennadius, De viris inlustribus 78, p. 88.
Italy: Eusebius of Vercelli (CPL no. 1056); Lucifer of Cagliari (CPL no. 11213); and Faustinus
presbyter (CPL no. 120); also Marius Victorinus, an African (CPL nos. 956) and Pelagius, a
Briton or perhaps an Irishman (CPL no. 748a). Gaul: Hilary of Poitiers (CPL nos. 4345 and
462) and Phoebadius of Agen (CPL no. 473). Africa: Augustine (CPL no. 699700 and 702).
To these Gennadius adds the mid fourth-century Spanish bishop Audentius of Toledo, whose
work has been lost: De viris inlustribus 14, p. 66; see also ibid., 16, p. 67 (Faustinus presbyter).
CPL, index nominum et operum, p. 751, s.v. Ariana et Antiariana.
Isidore of Seville, De viris inlustribus 28, ed. C. Codoner Merino, in El De viris illustribus de Isidoro
de Sevilla: Estudio y edicion crtica, Theses et Studia Philologica Salmanticensia 12 (Salamanca,
1964), pp. 14950.
Isidore, De viris inlustribus 30, ed. Codoner Merino, p. 151. This libellus does not survive, but a
letter written by Severus and Licinianus of Cartagena to the deacon Epiphanius on angels and
souls does: Licinianus of Cartagena, Ep. 2, ed. J. Madoz, in Liciniano de Cartagena y sus Cartas:
edicion crtica y estudio historico (Madrid, 1948), pp. 97124.
Gregory I, Dialogi 3.31, 2:38490; Paul the Deacon, Historia Langobardorum 3.21, ed. L. Capo,
in Paolo Diacono, Storia dei Longobardi (Milan, 1992), p. 152; see also Isidore, De viris inlustribus
28, ed. Codoner Merino, pp. 14950, and Greg. Ep. 9.229, 2:80511. See further Gregory of
Tours, Historiae 5.38, p. 244, who indicates that Hermenegild was converted by his Frankish
wife Ingundis.

175

Staying Roman
in a similar context. Bishop Avitus of Vienne (ad 490518) wrote an
anti-Arian treatise in the form of a dialogue with the Burgundian king
Gundobad.209 Gundobad himself remained an Arian, though he wavered
in his faith and is said to have inclined towards the Nicene confession.210
His son Sigismund converted to Nicene Christianity during his fathers
lifetime.211 Earlier, another bishop in southern Gaul, Faustus of Riez
(ad 461c.493), had written against the Arians and been exiled by the
Visigothic king Euric.212 Finally, as we have seen, at some point during
the turbulent early sixth century during which his city was controlled
in turns by Visigoths, Ostrogoths, Burgundians, and Franks Caesarius of Arles (c. ad 50243) also wrote an anti-Arian brief (see above,
Chapter 2.4.1). Collectively, the evidence of these five Gallic and Spanish
treatises suggests that under Arian kings debate of the Christological question may have been taken up by Nicene theologians primarily when they
thought a royal conversion was at stake. And, indeed, though Fulgentius
of Ruspes biographer dismisses Thrasamunds questioning as dissimilation, it sounds as if the Vandal king may very well have been sincere in
his desire to evaluate between the Nicene and Arian interpretations of
the faith, but that the arguments of the Trinitarians may genuinely have
failed to convince him.213
Public statements in opposition to the Arian theology of their Vandal
kings drew Nicene African bishops into a realm where their political
loyalties were in question. After the fall of Carthage, Quodvultdeus
had blamed the barbarian victory and the sufferings of the Africans on
their own failure to care for the poor.214 The solution Quodvultdeus
proposed was spiritual: the embracing of Christian ideals and a turn to
other-worldliness.215 Quodvultdeus also spoke of Christ as our David,
our king (Dauid noster, rex noster).216 The usage was not radical. Ambrose
had already made it clear that Christ was the Christians king in the fourth
century.217 But, at a critical moment in the acceptance or rejection of
209
210
211
212

213
215
217

Avitus of Vienne, Dialogi cum Gundobado rege vel librorum contra Arrianos reliquiae, ed. R. Peiper,
MGH AA 6/2 (Berlin, 1883), pp. 115.
Gregory of Tours, Historiae 2.34, pp. 812.
Avitus of Vienne, Epistula 29 (ad 514/516), ed. Peiper in MGH AA 6/2:59.
Faustus: Gennadius, De viris inlustribus 86, p. 91 (anti-Arian treatise, now lost). For a brief
overview of Faustus career, including his exile, see F. L. Cross and E. A. Livingstone (eds.),
The Oxford Dictionary of the Christian Church (3rd edn; Oxford, 1997), p. 601, s.n. Faustus of
Riez, St.
214 Quodvultdeus, De tempore barbarico 2, pp. 47386.
V. Fulg. 20, p. 99.
216 Ibid., 2.14.26, p. 486.
See esp. ibid., 2.7.4, p. 479.
Ambrose of Milan, De bono mortis 2.7, ed. C. Schenkl, in Sancti Ambrosii Opera, CSEL 32/1
(Leipzig, 1897), p. 707.

176

The old ruling class

a new temporal king, Quodvultdeus words made an unmistakable and
powerful statement that the Nicene Christians true loyalties were not
to this world, but to Gods heavenly kingdom. Over the course of the
Vandal century, other preachers and writers would hammer much the
same point home, but with a decidedly more worldly political twist.
In 484, after his forced rebaptism, Bishop Habetdeum is said to have
declared to his Arian tormenters that he had drawn up in his heart a
record (gesta) of the violence he had suffered, and sent it to my emperor
to be read.218 As the angels had written this account, we can safely take
Habetdeums emperor to be God; but it is likely that there is an element
of double entendre in Victor of Vitas reporting of the incident all
the more so if Victor did in fact write partly for an eastern, imperial
audience.
Even without such an audience, Huneric may well already have been
suspicious of the political loyalties of the Nicene bishops. As the Vandal
kings sporadic religious pressure on their closest advisers probably also
suggests, conversion to Arianism seems to have been something of a
litmus test of political loyalty to the Vandal regime, and one which
Nicene bishops by definition failed. Moreover, an air of paranoia seems
to have descended on Hunerics kingship around 479. Huneric seems to
have suspected a plot against him among the family and supporters of his
eldest brother, Theoderic.219 The king also became desperate to ensure
that, upon his death, his son Hilderic would immediately accede to the
Vandal throne.220 Yet Geiseric had decreed that the eldest of his own
male descendants should always succeed, placing Hunerics brother and
nephews before Hilderic in the succession.221 The causal relationship
between the question of the royal succession and the suspected plot
against Huneric is not entirely clear. Andy Merrills and Richard Miles
have recently and plausibly suggested that the succession crisis was a
reaction to the conspiracy, though the hostile testimony of Victor of Vita
implies that the reverse was the case.222 Either way, Huneric began a purge
of his court. Theoderics wife and the couples eldest son, evidently the
nexus of the plot, were tried and executed. Theoderic himself was sent
into exile, as (later) were his daughters and younger son. The eldest son
of Genton, another of Hunerics brothers, was also exiled together with
218
219
220
221
222

Vict. Vit. 3.46, pp. 945: lectitanda imperatori meo transmisi.

Vict. Vit. 2.1213, p. 28; Merrills and Miles, Vandals, p. 75.
Vict. Vit. 2.12, p. 28; see also ibid., 3.1920, pp. 801.
Vict. Vit. 2.13, p. 28; Proc. BV 1.7.29, 1:3445; Jordanes, Getica 33.169, p. 72.
Merrills and Miles, Vandals, pp. 756.

177

Staying Roman
his wife.223 The king furthermore ordered the execution of the retired
official Heldica Geiserics former praepositus regni and his family.224
Finally, Huneric had the Arian patriarch of Carthage, Jucundus, publicly
burned to death, and a number of Arian priests and deacons burned or
condemned to the beasts.225 Jucundus, at least, had been a supporter of
Theoderic.226 The purge, however, did not bring about a change in the
law of succession, and after their defeat at the council of Carthage five
years later, in 484, Huneric mooted the issue with the Nicene episcopate.
From Victors account of the episode, though, it is difficult to believe
that Huneric genuinely sought the backing of the Nicene church in
this regard. Not surprisingly, those bishops who refused to swear to a
document endorsing Hilderics succession were banished to Corsica and
condemned to hard labour, on the grounds that they did not want the
kings son to succeed. Yet those who did swear apparently the vast
majority were also exiled, this time to the African hinterland where
they were reduced to the status of coloni and given fields to cultivate,
on the grounds that they had violated the scriptural injunction against
oath-taking (Matt. 5:347).227 These are not the actions of a king casting
about for a base of support.228 These are the actions of a king already
old, embittered, and quite possibly sick who had come to view the
Nicene episcopate as inherently disloyal, and who saw confirmation of
their faithlessness in their every action. Those bishops who refused to
swear had shown their seditiousness openly; but those who had agreed
to swear were deceivers too, for, though proving their infidelity through
an obstinate and criminal adherence to the Nicene confession, they
dissembled allegiance to Hunerics house and the succession of his son.
From the kings point of view, they were damned either way.
This concern with political loyalties does not seem to have been unique
to the spring of 484. Victors text also contains the implication that earlier,
in the reign of Geiseric, the reception of the foreign monk John by
Bishop Felix of Hadrumetum was seen as a subversive act, for the bishop
was sent into exile as a result.229 Fulgentius of Ruspe openly informed
Thrasamund that the sempiternal King of Kings was more to be feared
223

224
226
227
228
229

Vict. Vit. 2.1214, pp. 289. For the difficulties to Hilderics immediate succession posed by
Theoderic and Gentons male heirs, see Courtois, Vandales, pp. 399404 and ibid., p. 390,
Tableau genealogique.
225 Vict. Vit. 2.13 and 2.16, pp. 28 and 30.
Vict. Vit. 2.1516, pp. 2930.
Vict. Vit. 1.44 and 2.13, pp. 19 and 28.
Vict. Vit. 3.1720, pp. 7981; Moderan, Notitia provinciarum, p. 181.
See, however, Merrills, Perils of Panegyric, pp. 1545 and Merrills and Miles, Vandals, p. 75;
see also Howe, Vandalen, pp. 2768.
Vict. Vit. 1.23, p. 11; Schmidt, Geschichte der Wandalen, p. 93 suggests that John was probably an
agent of the East Roman court.

178

The old ruling class

than temporal monarchs.230 Pseudo-Fulgentius, who refers to Christ as
our emperor (imperator noster), encouraged his congregation with the
thought that no usurper could dominate the Christian who adhered
to the narrow path of the true faith.231 The bishop also preached that
justice was nowhere secure against the hatred of the king, and prayed
for the visitation of the Holy Spirit in terms that were decidedly hostile
to the Vandal regime: If only he be the blessed visitor coming to us,
and the Gothic and barbarian guest now not be with us.232 Another
anonymous African preacher virtually encouraged his congregation to
become martyrs in opposition to the Vandals Arianizing policy with the
reminder that even though a man might lose the world, Christ receives
those who die for him.233
The thought may have been very welcome to the bishops parishioners. At least some Arians seem to have eyed the Nicene community
with suspicion right up until the final days of Vandal rule.234 Moreover,
the conditions of fifth- and sixth-century Africa allowed for sporadic
outbreaks of inter-communal violence, or at least localized Arian violence aimed against Nicene Christians. According to Victor of Vita, in
Hunerics day the persecution was under the direction of local Arian
bishops. Victor tells us that Arians forced their way into Nicene services
in the provincial towns of Tunuzuda, Gales, and Vicus Ammoniae, where
they disrupted the communion. At Regia, an armed band of Arians under
the leadership of the Arian priest Anduit were said to have attacked and
killed Trinitarians as they celebrated Easter.235 Others were reported as
having roamed the kingdom, bursting into the homes of Trinitarians
whom they would forcibly rebaptize as Arians.236 Victor also claims that
the rebaptized were given passports which allowed them to travel within
the kingdom.237 Perhaps because they lacked such documents, Fulgentius
of Ruspe and his companion were set upon and beaten by the minions of
Felix of Sicca, probably early in the reign of Thrasamund.238 We do not
really know how far the Vandal kings actively promoted such violence,
but it is difficult to believe that they strenuously opposed it.

230
231

232
233
234
236
237

Fulgentius of Ruspe, Ad Trasamundum 1.2.2, p. 99.

Pseudo-Fulgentius, Sermo 7, cols. 8667; the text reads Qui sedet in via, qua Christus transit,
non ei poteest domiuari [sic] tyrannus (ibid., col. 867A), but is clearly corrupt and should
probably read non ei potest dominare tyrannus.
Pseudo-Fulgentius, Sermo 80, cols. 9523: Utinam ipse sit hospes nobis superveniens benedictus,
et jam non sit nobis hospes Gothus et Barbarus. Justice: pseudo-Fulgentius, Sermo 13, col. 874.
Pseudo-Augustine, De occisione infantium 2.4 (sermo Mai 151), PLS 2, col. 1249.
235 Vict. Vit. 1.412, p. 18.
Dossey, Last Days of Vandal Africa, pp. 10910.
Vict. Vit. 3.48, pp. 956; see also ibid., 3.467, pp. 945.
238 V. Fulg. 67, pp. 3543.
Vict. Vit. 3.47, p. 95.

179

Staying Roman
3.3. Politics, ethnicity, and communities of faith
The decentralized nature of such aggression suggests that something
other than high politics was at stake in the Vandals Arianizing policy.
So too do the anonymous sermons delivered to audiences outside the
metropolis and its court circles, in the provinces of the kingdom, where
the lived experience of Vandal power could be represented by preachers
to their flocks as a persecution in which the whole Nicene community
suffered. Once again in Africas troubled sectarian history the tables had
turned, and it was not only bishops and high-ranking office-holders that
were affected by the changes. The Vandal kings religious policies set up
a system in which Arian communities in general enjoyed a politically and
socially privileged position, and non-Arian communities all non-Arian
communities, including not only Nicene Christians, but also Donatists,
Jews, Manicheans, and pagans were comparably disadvantaged.
By the late fifth century, though, the social frontiers between those
populations were looking increasingly unstable. To judge from Victors
account, two boundaries in particular, both of them deeply important to
him between Nicene and Arian, and between Roman and barbarian
had begun to crumble after nearly five decades of Vandal rule in Africa.
Indeed, Hunerics persecution of the Nicene Christians in 484 probably
could not have taken the particular form it did without considerable
integration between the Romano-African and Vandal communities, on
both political and cultural levels. As Peter Heather has rightly observed,
Victors reaction to the accommodation that he witnessed all around him
was to try to fortify or even re-establish these deteriorating communal
boundaries.239 But the Vandal kings Arianizing policy also betrays the
anxieties of a succession of barbarian rulers about a different aspect of
the same process that bothered Victor.
There is an unmistakable element of fear on the part of Huneric
himself the widowed husband of a Nicene wife that by the late fifth
century his Vandal subjects had begun to fall prey to the heretical influences of the Trinitarian confession. The kings fears were not completely
unfounded. As inter-communal boundaries began to erode in Africa, the
Vandal community had indeed witnessed at least some conversion to the
Nicene faith.240 Moreover, in Victor of Vitas account, the persecution
under Huneric had a very clear proximate cause: the refusal of Bishop
Eugenius, the Nicene metropolitan of Carthage, to ban Vandals from
239
240

Heather, Barbarian in Late Antiquity, p. 248; now see also Howe, Vandalen, pp. 15682 and
30218.
Vict. Vit. 3.38, p. 91; Howe, Vandalen, pp. 16876.

180

The old ruling class

entering Nicene churches.241 Until then, Victor tells us, Huneric had
shown himself mild toward everyone.242 Thereafter, the Vandal king
unleashed a reign of terror upon the Trinitarian community, which, as
we have seen, began with the scalping of anyone caught attending Nicene
services who looked like a Vandal (see above, Chapter 1.4). The move
was in some sense a logical extension of Hunerics earlier efforts to limit
contact between the Arian and Nicene communities. Even before the
situation descended into brutality and violence, Huneric had attempted
to reduce fraternization by forbidding Arian and Nicene Christians to
eat with one another.243 As we have seen, Huneric reiterated his fathers
decree that only Arians would be allowed to serve at court, and he
also issued an edict forbidding Nicene priests to celebrate the liturgy
in the sortes Vandalorum, lest they lead the inhabitants into Trinitarian
error.244
The kings edict was quite specific in its use of the term sortes Vandalorum, and given the official nature of the document the phrase was
presumably used in a technical sense, probably to refer to the lands and
estates that Geiseric had given to his warriors in the wake of their conquest of Africa.245 Indeed, the early attempts of Geiseric and Huneric to
suppress the Nicene church seem to have focused specifically on Africa
Proconsularis, where the sortes Vandalorum were probably for the most part
located.246 Yet Victor of Vita perceived Hunerics efforts as an attempt to
rid the entire kingdom of the Nicene church.247 In the changed circumstances of 484, when Victor was writing his text, the histrionic historian
may not have been far off the mark. Certainly Hunerics law of 25 February of that year demanded the immediate conversion of all the peoples
subject to his authority (universi populi nostro regno subiecti).248 The specific
social categories that the king mentions high-ranking aristocrats, senators, town council members, businessmen, commoners, circumcelliones
were drawn from earlier imperial legislation and thus might not represent
the social realities of the Vandal kingdom, but the intent of the law

241
242
243
244
245

246
247

Vict. Vit. 2.89, p. 27; see also ibid., 2.11, p. 28.

Vict. Vit. 2.12, p. 28: sese iamdudum omnibus lenem ostenderat. For important limitations on
Victors definition of omnes, see also Vict. Vit. 2.12, p. 24.
Vict. Vit. 2.46, p. 42.
Court: Vict. Vit. 2.23, p. 32. Liturgy: Vict. Vit. 2.39, p. 39 (an edict of Huneric).
Vict. Vit. 1.13, p. 7 with Proc. BV 1.5.1114, 1:333 (the ) K
)!, or lots of the
Vandals); however, see also Schwarcz, Settlement of the Vandals, p. 56, who reads sortes
vandalorum as referring to the kingdom as a whole.

Y. Moderan, LEtablissement
territorial des Vandales en Afrique, Antiquite tardive 10 (2002),
pp. 10710.
248 Vict. Vit. 3.314, pp. 728.
Vict. Vit. 2.40, pp. 3940.

181

Staying Roman
is clear: the legally mandated turn of the entire populace to the Arian
faith.249
Thus, central as Arianism seems to have been to the royal construction
of Vandal identity in the fifth and sixth centuries, it is important to stress
that the confession did not serve as an ethnic badge in the Vandal kingdom, at least not in the sense of being an exclusively Germanic religion
jealously guarded against incursions from the Romano-African majority.
Nor do we have any evidence to suggest that the Vandal kings themselves believed that religious conversion alone would turn a RomanoAfrican into a Vandal. To be sure, in his surviving laws Huneric refers
to your religion and your bishops on the one hand, and our bishops
or the bishops of our religion on the other. But the question turns
on what the king meant by the adjectives yours and ours; and the
distinction he draws is not between Romans and Vandals, but rather
between the misguided faith of the homoousians and the pristine faith
and true religion of the Arians.250 Moreover, the catholicity of their
ministry seems to have mattered deeply to the Arian ecclesiastical hierarchy in the Vandal kingdom.251 Whatever role the Vandalic language
may have played in Arian liturgical celebrations, the regions Arian bishops preached to their congregations in Latin.252 More to the point, at
the council of Carthage in 484 the representatives of the Arian episcopate were upset when their Nicene counterparts tried to arrogate to
themselves the appellation catholic.253 From the Arian point of view,
it was their church that represented the universal faith, and the sixthcentury preacher Fastidiosus, at least, unabashedly applied the adjective
catholica to his own Arian confession.254 Indeed, in the late fifth century
what Huneric seems to have feared was the erosion not primarily of
ethnic or cultural barriers between Vandals and Romano-Africans, but
249
250

251
252

253
254

Vict. Vit. 3.10, p. 76; cf. CTh 16.5.52 (ad 412), pp. 8723.
Religio uestra: Vict. Vit. 2.3, p. 25; sacerdotes uestri: ibid., 2.39, p. 39. Nostrae religionis episcopi: ibid.,
2.4, p. 25; episcopi nostri: ibid., 2.39 (twice) and 3.5, pp. 39 and 73; sacerdotes nostri: ibid., 3.14,
p. 78. Omousiani: ibid., 2.39 and 3.12, pp. 39 and 77; fides omousianorum: ibid., 2.39, p. 39; omousiani
sacerdotes: ibid., 3.4, p. 73; omousion: ibid., 3.5 and 3.12, pp. 73 and 77. Integra fides: ibid., 2.39,
p. 39; integra regula Christianae fidei: ibid., 2.39, p. 39; integra regula fidei: ibid., 3.4, p. 73. Uera
religio: ibid., 3.12, p. 77. Our bishops as ueris maiestatis diuinae cultores: ibid., 3.14, p. 78.
Howe, Vandalen, pp. 26876, who sees this ecumenism as new to Hunerics reign.
Liturgical use of Vandalic: Collatio beati Augustini cum Pascentio, col. 1162 and Tiefenbach,
Wandalische Domine miserere. Latin preaching: see, e.g., the African Arian homilies in the
late fifth- or early sixth-century MS Verona Bibl. Capit., LI (49) (De sollemnitatibus, ed. R.
Gryson, CCSL 87:4792; see also above, n. 189) and Fastidiosus, Sermo, ed. J. Fraipont in CCSL
91:2803.
Vict. Vit. 3.1, p. 72 and Shanzer, Intentions and Audiences, p. 285.
Fastidiosus, Sermo cc. 3 and 5, pp. 2812; see also Fulg. Ep. 9.4, 1:279. My thanks to Robin
Whelan for kindly alerting me to these passages.

182

The old ruling class

rather of the confessional boundaries between the Arian and Nicene faith
communities.
If the apostasy of Vandals specifically mattered to Huneric, as it seems
to have done, this probably had to do with the kings understanding of
his peoples providential history and his sacral conception of his own
rulership. Hunerics surviving laws reveal a court ideology that held
that the Vandal monarch was guided in his decisions by divine judgment, and indeed that the African provinces had been granted to him
by God.255 These concepts were apparently not unique to Huneric and
were moreover understood by the Vandals Romano-African subjects.
At least, in seeking to make amends to the jilted king Gunthamund, the
poet Dracontius conceded that transgression against the sovereign was
transgression against the deity.256 Such notions also seem to have made
an impression on foreigners: in his Getica, Jordanes indicates that Geiseric
was said to have accepted his authority from God himself.257 Procopius
too probably echoes a widespread perception that the first Vandal king
saw himself as the agent of divine will in an anecdote about Geiserics
seaborne raiding. Asked by his pilot where to go, the king was once said
to have replied, Obviously against those with whom God is angry.258
Indeed, the stunning rise of the Vandals as a people from insignificance to
dominance in the western Mediterranean over the course of thirty years
or so apparently seemed inexplicable to late antique audiences. Even
more so the barbarians ability to fight off repeated imperial attempts
at reconquest. Licking their wounds, Romans later reassured themselves
that these events were the result of treachery.259 The Vandal kings, by
contrast, may well have seen in their history the miraculous hand of God.
If so, then in a very real sense it probably mattered to Geiseric and his
successors that they keep themselves and their people in Gods special
favour.
Like the Roman emperors, the Vandal kings certainly seem to
have taken seriously what they saw as their responsibilities within the
255
256
258
259

Vict. Vit. 2.39, 3.3, and 3.14, pp. 39, 73, and 78; Gil Egea, Africa, p. 317.
257 Jordanes, Getica 33.169, p. 72.
Dracontius, Satisfactio 41 and 1078, 2:178 and 181.
Proc. BV 1.5.245, 1:335: ) ; L  + M.
Prosper, Chronicon s.aa. 427 and 439, pp. 472 and 477; Hydatius, Chronicon 107 (ad 439), p. 94;
Priscus, frag. 53.1, p. 362 = Theophanes, Chronographia AM 5961, 1:11516; Proc. BV 1.3.14
36 and 1.6.126, 1:3204 and 1:33540; Jordanes, Getica 33.167 and 33.172, pp. 72 and 74, and
see also ibid., 36.184 and 47.244, pp. 789 and 100; Jordanes, Romana 330, p. 42; John Malalas,
Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn, 1831),
pp. 3723; John of Antioch, frag. 196, ed. K. Muller, in Fragmenta historicorum Graecorum, 5 vols.
(Paris, 184170), 4:613; and Theophanes, Chronographia AM 5931, 1:935 (which is stitched
together from Procopius account). See also above, Chapter 1.1 n. 41. Salvian, De gubernatione
Dei 7.13.56, p. 470 offers a moral interpretation. On the rise of the Vandals, see Merrills and
Miles, Vandals, pp. 4155 and 11124.

183

Staying Roman
Christian church. Huneric, at least, seems to have seen himself as the
defender of the wider Arian community throughout the Mediterranean.
In his negotiations with the Roman emperor Zeno, one of the conditions
that the king placed on the ordination of Eugenius as bishop of Carthage,
and indeed on the religious toleration of Nicene practice in general in
the African kingdom, was the right of Arians in Constantinople and the
East to worship with equal freedom.260 Later, Hunerics advisers were
able to persuade him not to confiscate the goods of deceased African
Nicene bishops or impose a payment of 500 solidi for the ordination of
their successors out of fear of reprisals against Arians living in Thrace
and other imperial territories.261 The later, sixth-century pilgrim Theodosius reference to a monastery in Memphis, Egypt, that adhered to
the religion of the Vandals (religio Wandalorum) could conceivably also
hint at continued African connections to Arian communities elsewhere
in the Mediterranean.262 Within their own territory, the Vandal kings
promotion of an Arian court, their provisions for the material support of
the Arian ecclesiastical establishment out of confiscated Nicene property,
and if we can believe the Gallic moralist Salvian on this point their
enforcement of public decorum all factored in to their defence of the
church.263 So did their suppression of heresy, including not only Nicene
Christianity, but also Manichaeism and probably Donatism.264
Ultimately, though, the question of why the unity of the faith mattered
so much to the Vandal kings remains an open one. In some ways it was a
very Roman obsession, cast in the mould of Constantine and Theodosius.
More than that, religious identity was inherently political in the Vandal
kingdom. In changing the confession of Africas Roman population, the
Vandal kings would also have changed the fundamental identity of the vast
majority of their subjects. The successful Arianization of Africa would
thus also have erected a permanent barrier to imperial reconquest, which
by and large was predicated on a large Nicene population discontent
with the rule of a small Arian elite. Through their confiscation and
redistribution of wealth, the Vandal kings ensured the loyalty of that elite.
Moreover, the Vandal kings themselves appear to have been fired by an
260
262
263
264

261 Vict. Vit. 2.234, pp. 323.

Vict. Vit. 2.4, p. 25.
Theodosius, De situ terrae sanctae 14, ed. P. Geyer, CCSL 175 (Turnhout, 1965), p. 120.
Salvian, De gubernatione Dei 7.21.897.22.100, pp. 494524 (morality), and in general Gil Egea,
Africa, pp. 31518; see also Lambert, Barbarians in Salvians De Gubernatione Dei, pp. 10912.
Vict. Vit. 2.12, p. 24; Liber genealogus 616 and 618, pp. 1945; and Clover, Carthage and the
Vandals, p. 4. See also Fastidiosus, Sermo c. 2, p. 281. Perhaps also Judaism and paganism: Contra
Judaeos and Contra paganos, ed. R. Gryson, CCSL 87:93117 and 11841, respectively (and see
above, n. 189).

184

The old ruling class

Arian religious fervour unparalleled elsewhere in the barbarian West. In
addition to a sense of election at the hands of God, they may perhaps
have brought with them to Africa the zeal of the newly converted.
The fifth-century Spanish chronicler Hydatius repeats a rumour that
Geiseric had once been a Nicene Christian, and that he apostatized and
became an Arian before leading his people to Africa.265 The anecdote
was picked up and repeated by Isidore of Seville more than a century
later, but it is difficult to know how far to credit it.266 True or not, the
story stayed in Spain. Victor of Vita, who would almost certainly have
mentioned Geiserics alleged apostasy if he had been aware of it, is silent
on the subject. On the other hand, the kings of the Vandals, witnessing
the remarkable military successes of Arian Goths at the expense of the
Nicene Roman empire in the later fourth and fifth centuries, may perhaps
genuinely have come to believe that God was angry with the Trinitarians
for their heretical insistence on the consubstantiality of the Father and
the Son.267 In any case, the Vandal kings seem to have sought to establish
Arianism as the only form of Christianity within the territory of their
kingdom and above all in their court, and to that end they sought the
conversion of the Romano-African majority.
Confronted with this ambition, Africans of Nicene confession faced
two options: they could either acquiesce, or they could refuse to convert.
The hierarchy of the Nicene church organized an active resistance to the
Vandals Arianizing policy, exhorting, cajoling, and generally encouraging the faithful to stay the course. Their efforts were not completely in
vain. At least a handful of secular office-holders were willing to undergo
exile, humiliations, and even torture rather than abjure their confession.
But many Africans did convert to Arianism, including not only ambitious members of the secular elite but also at least some members of the
secular and regular clergy. In many cases these conversions were no doubt
sincere; but the overwhelming impression from the Nicene literature of
fifth- and sixth-century Africa is that wealth, honour, and advancement
or simply an escape from punishment could await those willing to
embrace the Arian confession. Of course, between resistance to and
enthusiastic acceptance of the Vandal Arianizing policy there doubtless lay an extensive middle ground populated by crypto-Nicenes, by
those on whom conversion rested as an uneasy burden, and by Nicene
Africans willing to confine themselves to private life in order to avoid the
265

266

Hydatius, Chronicon 79 (ad 428), pp. 8890. This is one of only two instances both, curiously,
connected with the Vandal king Geiseric where Hydatius admits that he is reporting a rumour:
Thompson, End of Roman Spain, Part I, p. 12.
267 Cf. Proc. BV 1.5.245, 1:335.
Isidore, Historia Vandalorum 74, p. 294.

185

Staying Roman
conversionary pressures associated with public office. Nor was conversion
a one-way phenomenon: we hear of a few Nicene Vandals over the course
of the fifth and sixth centuries, and the Nicene church may perhaps
have harboured dreams of a royal conversion as well. However, whether
through conversion to Arianism on the part of Romano-Africans or
through the increasing tolerance of the Vandal ruling elite towards their
Nicene subjects, we can only imagine that on the eve of the Byzantine
reconquest Africans were increasingly reconciled to Vandal rule not just
politically and culturally, but also religiously.

4. the perils of romanness

A. H. M. Jones once observed that No one who has read the letters,
poems, speeches, and histories which they wrote can doubt that the
literate upper classes of the empire regarded themselves as Romans, as
was only natural, seeing that they all shared the same cultural tradition.268
Jones is surely right, and not just for the empire, but for the immediate
post-imperial period as well. But, though they almost certainly continued
to think of themselves as Romans, to the best of my knowledge the
secular Vandal-era Romano-African literati never unambiguously referred
to themselves as Romans in any of their extant works. The reason for this,
I suspect, was that (the Vandal kings emphatic stress on their imperial
connections notwithstanding) the continued existence of the Roman
empire, in however truncated a state, probably ensured that the word
Roman was a politically loaded term in the Vandal kingdom. Indeed,
in the court circles in which our secular African authors moved, the
term probably carried dangerous political implications of affiliation to
the empire or, perhaps even more specifically, loyalty to the emperor.
The potential consequences of such foreign loyalties are illustrated in the
case of the poet Dracontius, who as we have seen was imprisoned by
Gunthamund for writing a poem of praise to a foreign ruler.
Their apparent reluctance to call themselves Romans leaves somewhat ambiguous the question of how secular African authors did imagine their identities. Two possibilities suggest themselves: the Vandalera poet Luxorius may perhaps have referred to his milieu as Punic,
while the mythographer Fabius Planciades Fulgentius, writing in the

268

A. H. M. Jones, Were the Ancient Heresies National or Social Movements in Disguise?, JThS
n.s. 10 (1959), p. 295.

186

The old ruling class

wake of the Byzantine reconquest, perhaps shortly after 550, probably saw himself as Libyan.269 Yet neither case provides unequivocal
evidence for how Africans self-identified in the Vandal period. There
are problems with Fulgentius chronology, and the possible reference
to a Punic girl (Poenica) in one of Luxorius poems is an emendation
on the part of the works modern editor for the manuscripts pontica
(Pontic). Plausible though this reconstruction may be, it does not present
a solid foundation on which to build an analysis of sixth-century African
identity.
In any case, it is questionable whether either a Punic or a Libyan identity would have been understood in late antique North African circles as
being in opposition to a Roman one. Fulgentius references to Libyans
and their language occur within a lipogrammatic work that prohibited the use of the letter A, rendering the adjective Libycus preferable
to Latinus. Yet, at least as far as idiom went, Fulgentius seems to have
understood the two words as being synonymous.270 Punic, by contrast,
was clearly a different language from Latin; but Punic culture enjoys only
the most shadowy of existences in the sources for the Vandal period,
consisting effectively of the artistic celebration of Carthage and its foundation legend discussed above (Chapter 1.3) and perhaps the borrowing
of a few words from a local vernacular into the Latin of the Albertini
Tablets.271 However, the Punic language had continued to be spoken in
the Numidian countryside in Augustines day, and (though Latin was the
common tongue in the city) it even enjoyed some currency in Hippo
Regius.272 The frequent use that Augustine made of Punic to elucidate
the meaning of scripture leaves no doubt that it was a Semitic language
closely related to Biblical Hebrew.273 Over sixty inscriptions in this same
269
270
271

272

273

Fulgentius Mythographer, De aetatibus mundi et hominis praef. ed. Helm in Opera, p. 130 and AL
324, l. 4, p. 261: ut tibi non placeat Poenica sed Garamas.
Hays, Romuleis Libicisque Litteris, pp. 1045.

V. Vaa nanen, Etude

sur le texte et la langue des Tablettes Albertini, Suomalaisen Tiedeakatemian
toimituksia, ser. B, 141/2 (Helsinki, 1965), pp. 489, s.nn. gemio and maforsenum; J. N. Adams,
Bilingualism and the Latin Language (Cambridge, 2003), pp. 4545; and J. N. Adams, The Regional
Diversification of Latin 200 bcad 600 (Cambridge, 2007), pp. 5568 and 561. On the distinctive
features of the Latin of the Albertini Tablets in general, see ibid., pp. 54962.
Countryside: Augustine, Epp. 66.2 and 108.5.14, CSEL 34:236 and 628; ibid., 209.23, CSEL
57:348; ibid., 20*.3.3 and 20*.21.1, ed. J. Divjak, CSEL 88 (Vienna, 1981), pp. 96 and 105;
Augustine, Epistolae ad Romanos inchoata expositio 13, ed. J. Divjak, CSEL 84 (Vienna, 1971),
pp. 1612; see also Augustine, De haeresibus 87, ed. R. Vander Plaetse and C. Beukers, CCSL
46 (Turnhout, 1969), pp. 33940. Hippo: Augustine, Sermones 113.2.2 and 167.3.4, PL 38:648
and 910.
The classic study is W. Green, Augustines Use of Punic, in W. Fischel (ed.), Semitic and
Oriental Studies: A Volume Presented to William Popper, Professor of Semitic Languages, Emeritus,
on the Occasion of his Seventy-Fifth Birthday, October 29, 1949, University of California Publications in Semitic Philology 11 (Berkeley, Calif., 1951), pp. 17990; see also C. Krahmalkov,

187

Staying Roman
language, but written in the Latin script, survive from third- and fourthcentury Tripolitania. Numerous Latino-Punic ostraka have also been
found at Bu Njem, Silin, and Wadi el-Amud, all in Libya, and at Henchir
Khanefi (south-west of Gab`es) in Tunisia.274 Late Roman authors furthermore refer to the existence of Punic books, and Augustine implies
that in his day written Christian devotional literature was composed in
Punic.275
To be sure, the bishop certainly also encountered lingering elite scorn
for Christian Punic culture on the part both of a pagan RomanoAfrican grammarian named Maximus of Madauros and of the Christian
Italo-Roman polemicist Julian of Eclanum.276 But Augustine strongly
defended the Punic language against Maximus, whose derision may have
scored him points elsewhere but seems genuinely to have offended the
bishop of Hippo.277 Against Julian, Augustine argued forcefully for the
catholicity of Punic Christianity, pointing out repeatedly that Cyprian
already honoured throughout the West as a martyr and doctor of the

274

275

276

277

A Phoenician-Punic Grammar, Handbook of Oriental Studies/Handbuch der Orientalistik, Section One, The Near and Middle East 54 (Leiden, 2001), pp. 5 and 1415, though his assessment
that Neo-Punic was the native tongue of . . . Augustine (ibid., p. 14) may perhaps be somewhat
too optimistic. See also F. Millar, Local Cultures in the Roman Empire: Libyan, Punic and
Latin in Roman Africa, JRS 58 (1968), 12634; D. J. Mattingly, Tripolitania (London, 1995),
pp. 1626; C. Lepelley, Temoignages de saint Augustin sur lampleur et les limites de lusage
de la langue punique dans lAfrique de son temps, in C. Briand-Ponsart (ed.), Identites et culture
dans lAlgerie antique, Publications de lUniversite de Rouen 377 (Rouen, 2005), pp. 12753;
and in general Bowersock, Brown, and Grabar, Late Antiquity, pp. 6567, s.v. Punic; but also
P. Brown, Christianity and Local Culture in Late Roman Africa, JRS 58 (1968), 8595 and
the bibliography cited there.
A good recent overview is provided by K. Jongeling and R. Kerr (eds.), Late Punic Epigraphy:
An Introduction to the Study of Neo-Punic and Latino-Punic Inscriptions (Tubingen, 2005), pp. 105
6. The most important collections include R. G. Goodchild, La necropoli Romano-Libica
di Bir ed-Dreder, Quaderni di Archeologia della Libia 3 (1954), 91107; F. Vattioni, Glosse
Puniche, Augustinianum 16 (1976), 53655; and the material cited in Reynolds and WardPerkins, Inscriptions of Roman Tripolitania, pp. 1013. See also R. Marichal, Les Ostraca de Bu
Njem, Supplements de Libya Antiqua 7 (Tripoli, 1992) and P. Berger and P. Gauckler, in BCTH
(1902), p. clxxvi.
Books: Solinus, Collectanea rerum memorabilium 32.2, ed. T. Mommsen (Berlin, 1895; repr. 1958),
p. 138; Ammianus Marcellinus, Res gestae 22.15.8, ed. W. Seyfarth, in Ammiani Marcellini Rerum
gestarum libri qui supersunt, 2 vols. (Leipzig, 1978), 1:283; Augustine, Ep. 17.2, CSEL 34:41;
and for the classical period Sallust, De bello Iugurthino 17.7, ed. L. D. Reynolds, in C. Sallusti
Crispi Catilina, Iugurtha, Historiarum fragmenta selecta (Oxford, 1991), p. 69 and Pliny, Naturalis
historia 18.223, ed. L. Ian and C. Mayhoff, 6 vols. (Stuttgart, 18921909; repr. 196770),
3:1478. Devotional literature: Augustine, Enarratio in Psalmum 118.32.8, 40:1776 with Lepelley,
Temoignages, p. 132.
Augustine, Epp. 16.2 and 17.2, CSEL 34:378 and 412; Augustine, Contra secundam Juliani
responsionem 1.7, 1.72, 1.73, 2.19, 3.78, 3.199, 5.11, and 6.18, PL 45:1053, 1097, 1148, 1280,
1333, 1440, and 1541; see also ibid., 1.48, col. 1069. See further, for the second century, Apuleius,
Apologia 98, ed. R. Helm in Apulei Platonici Madaurensis Opera quae supersunt, 3 vols. (Leipzig,
195591), 2/1:109.
Augustine, Ep. 17.2, CSEL 34:412. Thanks to Brian Stock for this reference.

188

The old ruling class

Church was himself Punic.278 Indeed here, as elsewhere, the bishop
used the word Punic interchangeably with African, including elite,
Latin-speaking, Roman Africans like himself.279
This does not seem simply to have been a rhetorical strategy on Augustines part. As early as the third century, a Punic background had not
prevented Septimius Severus from attaining the purple.280 The fourthcentury catacombs from Sirte in Tripolitania contain twelve short Christian epitaphs in late Punic, alongside similar inscriptions in Greek and
Latin. Regardless of the language of commemoration, the deceased are
consistently identified only by name and age, as was becoming increasingly common in Christian funerary epigraphy in general.281 Moreover,
in late antiquity, Roman authors both from Africa and from abroad were
willing to concede that Punic was not a completely alien language. The
fifth-century African expatriate Arnobius the Younger contrasted Latin
and Punic to the barbaric languages spoken in the African interior.282
Cassiodorus indicated that the pronunciation of vowels after the letter
N worked the same way in foreign (peregrinus) and Punic words as it
did in Latin.283 In a third-century opinion, later gathered into Justinians
Digest, the jurist Ulpian indicated that wills written in Punic were legally
binding under Roman law.284 Punic was not Latin, but neither was it
a barbarian language. In short, then, by late antiquity, self-identifying
as Punic or Libyan was not necessarily in conflict with a larger sense
of being culturally Roman. Then as now, identity was not an exclusive
proposition: one could be both.
There is no reason to suppose that this situation changed in the wake
of the Vandal invasion. If Luxorius did think of himself and his fellow
Africans as Punic, then this may have been synonymous in his mind
with being Roman. Indeed, we have two indications that secular African
authors did continue to conceive of their identities in Roman terms
under the Vandal regime. First, like Augustine before them, such authors
278
279
280

281
282
283
284

Augustine, Contra Iulianum 3.17.32, PL 44:719 and Augustine, Contra secundam Juliani responsionem 1.7, 1.72, 1.106, 6.6, 6.18, and 6.23, cols. 1053, 1097, 11201, 1511, 1542, and 1557.
See, e.g., Augustine, In epistolam Joannis ad Parthos 2.3, PL 25:1991: punicam, id est, Afram.
Severus 15.7, ed. E. Hohl in Scriptores Historiae Augustae, 2 vols. (Leipzig, 195565), 1:148; Epitome
de Caesaribus 20.8, ed. F. Pichlmayr in Sexti Aurelii Victoris Liber de Caesaribus (Leipzig, 1966),
pp. 1556; and Statius, Silvae 4.5.456, ed. A. Marastoni (Leipzig, 1961), p. 92.
R. Bartoccini, Scavi e rinvenimenti in Tripolitania negli anni 19261927, Africa Italiana 2
(1928), 187200 with Jongeling and Kerr, Late Punic Epigraphy, pp. 7174.
Arnobius Junior, Commentarii in Psalmos 104, ed. K.-D. Daur, CCSL 25 (Turnhout, 1990),
p. 159.
Cassiodorus, De orthographia 9, ed. H. Keil in Grammatici latini, 8 vols. (Leipzig, 185580), 7:201.
Digesta 32.11, ed. T. Mommsen, in Corpus Iuris Civilis, 3 vols. (7th edn; Berlin, 1895), 1:443;
but see also Digesta 45.1.1.6, p. 721.

189

Staying Roman
seem to have understood the city of Rome and its empire, their remote
historical or mythological past, and cultural institutions like the poetry
of Virgil as having defined Romanness.285 Romano-Africans might not
have been part of the empire any more, but they continued to be educated
in the classical tradition and their own individual histories were still
inextricably linked to the ancient history of the Roman empire. The poet
Dracontius wrote of Romano-Africans as the descendants of Romulus,
and the Latin literature of Vandal Africa is permeated with a clear sense
later made explicit by the Byzantine author Procopius that Africans
were Romans by descent.286 Second, Latin poets from the Vandals North
African kingdom seem to have been capable of praising members of the
new ruling class only insofar as they could, like Africans themselves,
function within the late Roman cultural matrix. By the 520s and 530s,
when Luxorius published the poems of his Liber epigrammaton, four or five
generations of Vandals and Romano-Africans had grown to adulthood
alongside one another in a society where (from the surviving evidence)
both sides seem to have acknowledged the supremacy of artistic forms
and modes of expression ultimately derived from classical antiquity. As
a result, we do not really know how long distinctively Vandal cultural
traits survived in Africa indeed, we scarcely know what kind of traits to
look for because, like most of us, the Romano-Africans who produced
the extant literary sources viewed the world through cultural lenses that
preconditioned what they were able to see. And amid the uncertainty
of a world drifting slowly apart in subtle yet meaningful ways, what the
Romano-African elite for the most part wanted to see were signs of
stability and continuity.
Yet if their silence on this point is significant (and it may not be),
Romano-Africans who moved in political circles and who probably
felt culturally Roman may not have been entirely comfortable giving
expression to their Romanness. This, in turn, left all of the rhetorical
cards in the hands of Nicene polemicists; for it was to be the ecclesiastical
authors Quodvultdeus of Carthage, Fulgentius of Ruspe, and Victor of
Vita who were to allow for perhaps the broadest concept of Romanness
in the Vandal period. Though Fulgentius of Ruspe usually used the
285

286

Romans of Rome: AL 243, p. 176. Empire: AL 422, p. 324. Ancient and mythological past:
Dracontius, Satisfactio 183, 2:185 (Titus); Dracontius, Laudes Dei 3.1467, 3.3223, 3.419, 3.456,
2:23, 32, 36, and 38 (ancient Romans). Virgil: AL 258, p. 192. Military: AL 390, p. 304. Roman
youth: AL 438, p. 332. See also (though probably later in date) Fulgentius Mythographer,
Mitologiae 1.praef., 1.20, and 3.5, pp. 8, 31, and 64; Fulgentius Mythographer, Expositio Virgilianae
continentiae, ed. Helm, in Opera, p. 105; Fulgentius Mythographer, Expositio sermonum antiquorum
5 and 49, ed. Helm, pp. 113 and 124; Fulgentius Mythographer, De aetatibus mundi, praef. and
11, pp. 131 and 167. This usage is consistent with that of Augustine in De civitate Dei.
Dracontius, Romulea 1.14, 3:134; Proc. BV 1.16.3, 1:382.

190

The old ruling class

word Roman in the same way as his secular contemporaries, the bishop
nevertheless seemed to allow for the possibility that Africans could also
be Romans when he wrote to Thrasamund:
Until now it has been considered a rare thing that the mind of a barbarian king,
continually occupied with the numerous cares of the kingdom, be so inflamed
with the burning delight of acquiring wisdom, since only some kind of leisured
man or a Roman is accustomed to make efforts of this sort, which are always
wearying.287

Quodvultdeus usage too was usually in line with that of late antique
secular African authors; but on one occasion he used the phrase the
Roman faith (fides romana) in apposition to the true universal faith (vera
fides catholicae), implying that, for him, Roman meant Nicene or, from
his theological point of view, orthodox.288 Similarly, Fulgentius of Ruspe
once quoted with approbation a letter of Pope Hormisdas (ad 51423)
that used the same terms interchangeably of the Nicene church.289 But
the clearest declaration that Nicene Africans living under Vandal domination were still Romans is to be found in Victor of Vitas History of the
Persecution. Though he also used the term to refer specifically to the city
of Rome, Victor unabashedly applied the word Roman to his fellow
287

288

289

Fulgentius of Ruspe, Ad Trasamundum 1.2.2, p. 99: rarum hactenus habeatur barbari regis
animum numerosis regni curis iugiter occupatum tam feruenti cognoscendae sapientiae delectatione flammari, cum huiuscemodi semper infatigabiles nisus non nisi uel otiosus quis habere
soleat uel Romanus. In Fulgentius corpus, the word Romanus typically refers specifically to
the city of Rome or its church, in the majority of cases in reference to Pauls Epistle to the
Romans: Dicta regis Trasamundi, p. 92; Ad Thrasamundum 3.27.1, p. 171; Epp. 3.4, 14.18, 17.5,
and 17.21 (twice), 1:213, 1:407, 2:567, and 2:580; De veritate praedestinationis et gratiae dei 2.43,
p. 519; Ad Euthymium 1.10.1, p. 656; Contra Fabianum fr. 34.16, p. 842. Fulgentius Ep. 6.2, 1:240
refers to the populus Romanus (presumably the empires inhabitants); ibid., 6.3, 1:241 to the
(western) Romanus consul.
Quodvultdeus, De accedentibus ad gratiam 2.13.6, ed. R. Braun in CCSL 60:470: Non crederis
ueram fidem tenere catholicae, quae fidem non doces esse seruandam romanam. See also
Quodvultdeus, Adversus quinque hereses 5.5, p. 277 (Pauls Epistle to the Romans); Quodvultdeus,
De tempore barbarico 2.3.6, p. 475 (Roman history); Quodvultdeus, De quattuor virtutibus caritatis
12.3, ed. Braun in CCSL 60:375 (a quotation from John 11:48); Quodvultdeus, Liber promissionum
2.34.74, p. 140 (Roman empire); ibid., 2.40.91 and 2.40.92, p. 154 (ancient Romans); and ibid.,
2.40.92 (bis), p. 155 (the apostle Paul, a Roman citizen). See too the Augustinian or pseudoAugustinian Sermo 381 (De Natali apostolorum Petri et Pauli), PL 39:16834: Petri et Pauli
apostolorum dies, in quo triumphalem coronam, devicto diabolo, meruerunt, quantum fides
Romana testatur, hodiernus est. Apart from this sermon, however, Augustine himself only ever
uses the phrase fides Romana when quoting Pelagius: Augustine, De gratia Christi et de peccato
originali 1.43.47, ed. C. Urba and J. Zycha, CSEL 42 (Vienna, 1902), p. 159; Augustine, De
nuptiis et concupiscentia 1.35.40, ed. Urba and Zycha, CSEL 42:252; and Contra Iulianum 1.7.30,
col. 661. Indeed, down until the early fifth century only Jerome and, apparently, Pelagius use
the term fides Romana unambiguously in the sense of Catholic faith, apart from quotations of
or clear references to one of these two authors.
Fulg. Ep. 15.18, 2:456.

191

Staying Roman
Romano-Africans, despite the fact that they were no longer subjects of
the empire.290 Three-quarters of a century earlier, writing before the
full implications of the collapse of the Danube and Rhineland frontiers
were to become clear, Augustine had maintained that the distinction
between the different peoples of the Roman empire or, for that matter,
between Romans and non-Romans was less important than an individuals acceptance or rejection of the truth of Christian revelation.291
For Augustine, Christianity, not Romanness, was to be the touchstone
of identity. Victor partially embraced that vision, but he also twisted it to
suit his own needs. Like Augustine, Victor perceived the world as being
separated into two kinds of human society. Only to Victor, those societies were two camps divided by implacable hatred and animosity: one
Roman and Nicene, the other barbarian and Arian.292 For in Victors
mind what unified Romans throughout the Mediterranean world, what
made them a single people and distinguished them from everyone else,
was their adherence to Nicene Christianity.
Thus Victors vision of what it meant to be Roman was more expansive than that of most of his contemporaries: it had little to do with
existing political boundaries, but was defined rather in terms of faith and
culture. If anything, in the assertion that Romano-Africans were ultimately part of the same society as those still living under imperial rule,
the polemical objectives of Victors text demanded that he minimize the
cultural importance of the new borders that had emerged over the course
of the fifth century. Victor understood very clearly that his interests lay
with the other Nicene Christians of the Mediterranean world at large,
and his History is nothing if not an appeal to a Nicene, Roman sense of
commonality and of mutual distinction from the barbarians. But Victors
vision was also narrower than that of the secular authors who were to
follow him in quieter times; for when he looked around himself, he did
not want to see the commonalities that Vandals and Romans had come
to share. When regarding the Vandals, Victor perceived only a hostile
hoard set upon the destruction of the Roman people. This was precisely
because he like other Nicene writers had redefined Romanness to
be synonymous with a theological orthodoxy that three generations of
Vandal kings unquestionably did find inimical.
290

291
292

Vict. Vit. 2.43, p. 41 clearly refers specifically to the papacy and the city of Rome; Vict. Vit.
1.44, p. 19 uses the term Romani clearly in reference to Romano-Africans. Vict. Vit. 1.37,
p. 16 (civitas Romana) is more ambiguous. Vict. Vit. 3.62, p. 108, cited above, also seems to be
ecumenical in its implications. See also on this point Costanza, Uuandali-Arriani, pp. 2268
and 2347.
Augustine, De civitate Dei, 14.1, 48:414.
On this point, see also Howe, Vandalen, pp. 32833.

192

The old ruling class

The Nicene ecclesiastical establishment and their mouthpieces were, I
suspect, by no means oblivious to the cultural and even political implications of their equation of Nicene with Roman and Arian with barbarian.
This is certainly true if, as Christian Courtois once argued, Victor was
writing at least in part with an eastern, imperial audience in mind.293
But it was also true within Africa itself. Victor probably does not address
and certainly does not need an eastern audience when he denounces
those who praise the barbarians or carps upon those barbarians envy
of Romans, their desire to obscure the splendour and nobility of the
Roman name, their use of their subjects as slaves, and, in short, their
hatred of Romans.294 This is powerful rhetoric that plays heavily on
the cultural politics of the Mediterranean ideology of the barbarian, and
perhaps even on the apparent reluctance in elite Romano-African circles
unambiguously to proclaim their own Romanness. Victors text was a
desperate appeal to those members of the Vandal-era elite who found
that, after all, their lives were somewhat hollow without the adjective
Roman. To these readers Victor insisted that staying Roman mattered,
and while confronting them with this predicament he simultaneously
offered them a solution. One could stay Roman, he contended, by
clinging faithfully to the Nicene confession.
The Romano-African response to the Vandal presence was, of course,
both complex and varied. In aggregate we see meaningful change over
time, from the apocalyptic fears of Bishop Quodvultdeus of Carthage and
the anonymous author of the Liber genealogus to Victor of Vitas perception of Vandal rule as a period of Babylonian Captivity for the African
church to a guarded optimism under the benignly tolerant Hilderic.
Ecclesiastical and secular sources also give us different perspectives on the
Vandal kingdom; though not, it would seem, because they were targeted
differently by the kings Arianizing policy. Rather, the threat that royally
sponsored Arianism posed to Nicene Christianity in Africa was a burning concern to ecclesiastical authors, while to all appearances that threat
mattered less (if at all) to the secular poets and literati whose works define
for us the lay African experience of Vandal power. Put another way,
the terms on which Romano-African accommodation with the Vandals
was possible were acceptable to Luxorius and his milieu, and simultaneously inimical to men like Victor of Vita. High culture remained
essentially late Roman in character, as did the grammar and vocabulary of
293
294

Courtois, Victor de Vita, pp. 1722.

Shanzer, Intentions and Audiences, pp. 27990 and Howe, Vandalen, pp. 33456; see also
Costanza, Uuandali-Arriani, p. 237.

193

Staying Roman
power though now these operated within the context of an autonomous
Vandal kingdom. This fact in itself does not seem to have posed a
tremendous obstacle to VandalAfrican accommodation. The Vandals
seem generally to have won the struggle for political legitimacy quite
quickly, at least if we measure success in terms of Romano-African service in the Vandal administration, the production of Latin verse laudatory
of Vandal kings, or the dating of inscriptions according to Vandal regnal
years. Moreover, as the Vandals came to look more and more like their
Romano-African subjects, those same subjects seem increasingly to have
felt that they could live and do business with barbarians of such culture.
In religious terms, accommodation seems for the most part only to have
been possible through Romano-African conversion to Arianism, though
it is important to stress that we have no real indications that either the
Vandal kings or Roman converts like the preacher Fastidiosus regarded
conversion alone as making an individual any less Roman. The Vandal
Arianizing policy seems to have been remarkably successful chillingly
so from the point of view of the Nicene ecclesiastical hierarchy, which
strenuously sought to turn back the tide of Arian conversions through the
production of a voluminous anti-Arian polemical literature and a rhetorical association of Romanness with the Nicene confession and barbarism
with Arianism. These facts, combined with Victors deeply held conviction that the Vandal kings were willing to resort to unspeakable violence
in order to secure the conversion of the Romano-African majority, go a
long way in explaining the intensity of Victors hatred of the Vandals.
Both social perspective and change over time certainly mattered in
terms of forming the shifting attitudes of individual Africans to the
Vandals from one generation to the next. Nicene Christians who refused
to change their confession even in the face of strong external pressure
clearly defined themselves to a meaningful degree in terms of their
faith. On the face of it, it makes sense that we would find a greater
percentage of such individuals among the clergy than among the laity,
and, as a corollary, a greater ideological opposition to Vandal rule. It is
worth bearing in mind, however, that the fault lines of African society
were never this clear-cut: some members of the Nicene clergy certainly
converted to Arianism under the Vandals, while at least a handful of
secular office-holders were willing to undergo exile, humiliation, torture,
and even execution to avoid apostatizing their faith. At the same time,
Thrasamund does not appear to have unleashed anything like the horrors
on African society that Huneric used to compel the conversion of his
Nicene subjects in 484. Two generations after the worst violence of
Hunerics reign, the Arianizing push to which that king had been so
committed stalled and died under his son Hilderic. It is not unreasonable
194

The old ruling class

to suppose an inverse relationship between the level of Vandal violence
towards the Nicene majority and the degree to which that majority was
reconciled to Vandal rule.
But Romano-African responses to the Vandals were complex on an
individual level as well. Fulgentius of Ruspe was unambiguously opposed
to Thrasamunds religious policy, which the bishop attempted to counter
with great effort and at considerable personal cost. Yet in his Sardinian
exile the same Fulgentius, or someone very close to him, also dated a
codex of Hilary of Poitiers Trinitarian writings according to the year of
Thrasamunds reign in which the manuscript was produced. Fulgentius
and his circle, then, would seem to have accepted Thrasamunds political
legitimacy while opposing his theological policy. Indeed, this fact is
probably implied by the very energy Fulgentius expended in attempting
to secure the Vandal kings conversion to the Nicene faith. To Fulgentius,
Thrasamunds right as a Vandal to rule his African kingdom was not at
stake. The only question was whether this worldly king would rule
according to Gods eternal law or against it.
The answer to that question, of course, depended on ones confessional
point of view; and the evidence suggests that many Romano-Africans
were quite willing to change their confession. The unexpressed subtext
virtually omnipresent in Vandal-era North African ecclesiastical literature was that persecution could be ended (on a personal level) not just
through bribes or flight, but through acquiescence and conversion. The
violence of the persecution also seems to have diminished in proportion to its success. By Fulgentius generation, there were almost certainly
more Romano-African converts to Arianism than there had been even
in Victors day. Like the Arian bishop who offered to protest Fulgentius beating at the hands of some Arian thugs, these converts probably
retained their ties across confessional boundaries, and the more influential
among them presumably blanched at the extension of physical violence
to their friends and relatives. The diminishing virulence of the Vandal
kings aggressive Arianism simply set the stage for the ever-closer interweaving of the Vandal and Romano-African communities. By the early
sixth century, the Vandal kingdom was probably well along the path
simultaneously being trodden in Gaul, Spain, and Italy of barbarianRoman reconciliation. Whatever the epithets meant on the eve of the
Byzantine reconquest, by the early sixth century Romans and Vandals
could live side by side in relative peace, because even in terms of
that touchstone of late antique identity, religious belief and practice
they were in the slow, sometimes painful process of becoming a single
people.
195

Chapter 4

N EW R O M E, N EW R O M A N S

Conquests breed cultural ambiguities. When eastern troops recaptured

North Africa for the Byzantine empire in ad 5334, nearly a century of
maturing VandalAfrican accommodation was cut short, and the political, cultural, and religious orientation of the Romano-African elite was
challenged once again. This time, however, that elite faced a group of
putative liberators who laid claim to the same Roman cultural heritage
as did the Romano-Africans themselves. Indeed, as far as the Byzantine
historian Procopius was concerned, the Romans (3 -.! ") were the
eastern forces sent by Justinian to recover the territory of the Vandal
kingdom for the empire; not Africans of Nicene confession like Victor
of Vita who had bewailed their provinces subjugation to the Vandals.
The Africans (or Libyans, 3 B
*) were, of course, Romans by origin
(-.! "  8+), but this is a distinction that cut both ways.1 In
the context of Procopius History of the Wars, it serves to highlight the
commonalities between the local population and the invading army: their
shared history and a certain affinity between Romano-African culture
and the culture of the eastern Mediterranean. But Procopius usage also
contains the implication, intended or not, that (from the Byzantine point
of view) the Latin-speaking population of Africa no longer participated
fully in the Roman world if, indeed, they continued to do so at all. To
Procopius and his Greek-speaking audience, Romanness and the empire
shared coterminous boundaries.
Even so, from an imperial point of view, Africa had not so much
been conquered as reintegrated into the respublica after a protracted barbarian interregnum. Justinians rescript outlining how the region was to
be governed places particular emphasis on the liberation of Africa from
1

Proc. BV 1.16.3, 1:382. Interestingly, the later Arab chronicler Ibn c Abd al-Hakam maintained a
similar distinction between three cultural groups whom the Muslims encountered on their first
raids into the seventh-century Maghrib: Rum (Romans or Byzantines), Afariq or Africans, and
Berbers or Moors. Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, ed. and trans. A. Gateau, in
Conquete de lAfrique du nord et de lEspagne (2nd edn; Algiers, 1947).

196

New Rome, New Romans

the Vandal yoke; but it also points out that, in facilitating the territorys liberation, God had chosen to restore (restituere) these provinces
to imperial control.2 The challenge was thus one not of making Africa
Roman, but of making Africa Roman again. Therefore, in this chapter
I will investigate how the Byzantine emperors sought to re-Romanize
their recaptured prize. This examination will focus on the mechanics of
the sixth- and seventh-century imperial administration of Africa and the
provinces reintegration into the larger political and military structures of
the Byzantine or East Roman empire. First of all it is necessary to look
briefly at the theoretical organization of the African provincial administration, and consider what the sources allow us to see of its practical
workings. Then I will examine the primary agents through whom the
emperors in Constantinople secured their distant province: Africas civil
officials and military officers. I will investigate their regional origins,
career patterns, terms of appointment, and connections to the imperial
centre before finally evaluating the extent to which representatives of
the Byzantine state infiltrated sixth- and seventh-century African society. The new ruling elite of Byzantine Africa was, I argue, recruited
from throughout the empire including Africa though at its highest
levels this elite was primarily drawn from the borderlands of the eastern
empire. While Byzantine civil administrators and military officers probably never constituted much more than a relatively thin stratum at the
very top of African society, it was here that power concentrated, and the
intimate connections to the emperor enjoyed by the regions most elevated officials played a vital role in re-assimilating Africa into the sixthand seventh-century Byzantine empire.
1. the administration of byzantine africa: structure,
sources, and source problems
In his rescript of 534, the Byzantine emperor Justinian laid the groundwork for the reorganization of Africa as a diocese of the Byzantine
Empire. The region as a whole was to be under the civil administration of
a Praetorian Prefect headquartered at Carthage. Africa was to be divided
into seven provinces: Proconsularis (now called Zeugi Carthago, perhaps
in an echo of the Vandal kings Carthage-centred ideology), Byzacena,
and Tripolitania, all under the governorship of senatorial governors or
consulares; and Numidia, the two Mauretanias, and Sardinia, under the
authority of praesides, who since the end of the fourth century were also
of senatorial rank. The prefect was to have a staff of 396 civil servants
2

CJ 1.27.1.8 (ad 534), p. 77.

197

Staying Roman
and soldiers under his command; each of the provincial governors, fifty
administrators. These were to include bureaucrats, registrars of public
documents, scribes, tax-collectors, couriers, heralds, grooms, standardbearers, overseers of public works, surveyors, chartularii. In addition, five
doctors and two grammatici were to be on the public payroll.3
Justinian divided military command in the new diocese between the
duces of Tripolitania, Byzacena, Numidia, Mauretania, and Sardinia. They
were each assigned a staff consisting of a legal advisor (assessor), a clerk,
several officers, and various non-commissioned officers: some forty-three
men in all. Septem (mod. Ceuta), on the Strait of Gibraltar, was to be
defended by a tribune in command of a garrison of soldiers. In addition,
limitanei of local origin were to be recruited to see to the defence of the
Byzantine frontier.4 In practice, all these military officers were under the
supreme command of a general referred to by our sources (with some
variants) as the magister militum Africae or strategos Libyes.5 The difficulties
of administering such a far-flung province also ensured that even in
Justinians lifetime civil and military command were, at least occasionally,
unified in the same individual. Later, towards the end of the sixth century,
this arrangement was formalized in the creation of the post of exarch,
a new official who outranked rather than replaced the prefect and the
magister militum Africae.
The sources for Byzantine Africa are comparatively abundant, and
include letters, ecclesiastical tracts, chronicles, histories, poetry, and legal
sources, as well as archaeological, numismatic, sigillistic, and epigraphic
data.6 These sources preserve a remarkable amount of information concerning the North African elite in the Byzantine period, which is nevertheless limited in two important ways. The first of these is chronological.
Although our sources cover the entire period from the disembarkation
of Belisarius expeditionary army at Caput Vada (mod. Ras Kapudia) in
3

4
5

CJ 1.27.1 (ad 534), pp. 779. The most recent full-length synthesis of the history of Byzantine
Africa is still C. Diehl, LAfrique byzantine: histoire de la domination byzantine en Afrique (533709)
(Paris, 1896), here pp. 97137 and 483502 (civil and military organization of the province); on
Byzantine Africa in general, see also Averil Cameron, Gelimers Laughter: The Case of Byzantine
Africa, in F. M. Clover and R. S. Humphreys (eds.), Tradition and Innovation in Late Antiquity
(Madison, Wis., 1989), pp. 17190.
CJ 1.27.2 (ad 534), pp. 7981.
When precisely this title first began to be used is a matter of some debate, though the title clearly
post-dates Apr. 534; Belisarius was in Africa under exceptional circumstances and is addressed in
CJ 1.27.2 (ad 534), p. 79 as magister militum per Orientem. On this question, see C. Zuckerman,
La Haute Hierarchie militaire en Afrique byzantine, Antiquite tardive 10 (2002), 16975.
For an excellent discussion of the literary evidence, see Averil Cameron, Byzantine Africa
The Literary Evidence, in J. H. Humphrey (ed.), Excavations at Carthage 1978, Conducted by the
University of Michigan, vol. 7 (Ann Arbor, Mich., 1982), pp. 2962.

198

New Rome, New Romans

533 to the final Islamic capture of Carthage in 698, our richest evidence
for the provincial administration of Byzantine Africa comes from the first
two decades of the occupation. Collectively, the historian Procopius, who
had accompanied Belisarius expeditionary force, and the poet Corippus,
who was himself a native Romano-African, describe local society down
to c.550 with an immediacy and detail that is unparalleled at any point in
the later Byzantine period. The discussion that follows will range from
the second quarter of the sixth century to the final years of the seventh,
but the nature of the sources is such that the majority of the evidence is
inevitably drawn from the 530s and 540s.
Second, Byzantine Africa was a militarized society. The sources provide us with considerably more information about the military officers
who saw to the defence of the Byzantine prefecture than they do about
the civil officials who administered it. In both instances, though, it is
easiest to see the highest ranks of the administration: the magistri militum, Praetorian Prefects, and exarchs of Africa. Below this exalted level,
considerably more evidence survives for the careers (even the names) of
important military commanders than for high-ranking civilian administrators. For example, while it is possible to identify perhaps nine provincial duces from Byzantine Africa, we are unable to reconstruct the career
of even a single provincial governor. What the sources do reveal on the
administrative side is something of Byzantine Africas fiscal organization:
its commerciarii, apo eparchontes, and a dioecetes provinciarum (dignitaries who
controlled trade and tolls along the frontiers, supervised state workshops,
and were responsible for the taxation of the province). We can also see
a handful of sacellarii, administrative officials whose precise duties in the
sixth and seventh centuries are not well understood, though they were
at least partially financial. The limitations in terms of how far the sources
illuminate the civilian administration of Byzantine Africa are not unrelated to these same sources chronological strengths: in addition to being
early, our two richest informants (Procopius and Corippus) were also
both primarily concerned with military affairs and the military administration of the African provinces. The discussion that follows naturally
deals with both the civilian and the military officials of Byzantine Africa,
but the majority of the evidence necessarily concerns high-ranking military commanders.
2. origins
In a stimulating and important essay, Averil Cameron has suggested that
the Byzantines, presenting themselves as the restorers of Roman rule,
199

Staying Roman
probably seemed unconvincingly eastern to the Roman Africans.7 I
will return to the question of how unconvincing Byzantine claims to
Romanness appeared to sixth-century Romano-Africans; for now, however, Camerons suggestion raises the important question of how long the
administration of Africa remained in the hands of easterners. To answer
this question, in the discussion that follows I will consider not only where
the high officials who ran the province originated over the two centuries
of Byzantine rule in North Africa, but whether military officers and
civil servants were recruited from different regions of the empire, and
whether these patterns of recruitment changed over time. I will also
examine the evidence for local participation in the administration of the
North African provinces, and, finally, consider what the onomastic evidence suggests about the composition of the Afro-Byzantine elite. An
appreciation of the Afro-Byzantine elites regional origins is not only
central to our understanding of the strategies employed by Constantinopolitan emperors to re-Romanize the reconquered North African
provinces; it also casts a unique light on the broader administrative structures and policies of the Roman empire as a whole in the age of Justinian
and his successors.
2.1. Individuals
In the sixth and seventh centuries, nothing guaranteed that a highranking imperial official would be of local origins anywhere in the
Byzantine empire. Indeed, until the reign of Justin II (ad 56578) imperial policy did not normally allow the appointment of an individual to the
governorship of his native province.8 To be sure, Justinian had already
relaxed this rule in Italy in 554, when he allowed bishops and notables to elect their governors from among the regional elite.9 Even in
Italy, though, the men who enjoyed supreme command over imperial
armies were consistently easterners.10 The same seems generally to have
been true of the regions highest civil office, the Praetorian Prefecture,
notwithstanding the fact that the first two Byzantine prefects were both
7
8

9
10

Cameron, Byzantine Africa, p. 31; see also Y. Moderan, Les Maures et lAfrique romaine (IVe VIIe
si`ecle), Biblioth`eque des e coles francaises dAth`enes et de Rome 314 (Rome, 2003), p. 584.
CJ 1.41.1 (n.d.) and 9.29.3 (ad 385), pp. 86 and 385; Justin II, Novella 5.1 (ad 569), ed. C. E.
Zachariae von Lingenthal, in Jus Graeco-Romanum, vol. 3 (Leipzig, 1857), pp. 1112 = Just.
Nov. 149.1, p. 724. A. H. M. Jones, The Later Roman Empire, 284602: A Social, Economic and
Administrative Survey, 2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986), 1:389.
Just. Nov. App. 7.12 = Constitutio pragmatica 12 (ad 554), pp. 8001 and Jones, Later Roman
Empire, 1:389.
T. S. Brown, Gentlemen and Officers: Imperial Administration and Aristocratic Power in Byzantine Italy
ad 554800 (London, 1984), pp. 645.

200

New Rome, New Romans

Italians: Fidelis (ad 5378) was a native of Milan; Reparatus (ad 5389),
a native of Rome.11 Both had held high office under the Ostrogothic
king Athalaric, but in the late 530s both became actively engaged in
the imperial effort to reconquer Italy.12 Yet, after the death of Reparatus, easterners like Athanasius (a future prefect of Africa) begin to be
appointed to the Italian prefecture. The contemporary administration of
the diocese of Egypt, however, presents an instructive counter-example,
for there the emperors seem to have shown a distinct preference for
officials of Egyptian origins. Of the twenty or so governors of Egyptian provinces whose regional origins we can trace between the reigns
of Justinian and Heraclius (ad 61041), at least fourteen seem themselves to have been Egyptians.13 After Justinians reforms of 539, these
men combined civil and military command in their provinces, and in
practical terms they were no longer subject to an official (other than
the emperor) with authority over the whole diocese.14 While they do
not generally seem to have been posted to their home provinces men
from the western Delta or the lower Nile valley serving in the Thebaid,
for example there appear to have been surprisingly few non-Egyptians
among them.15 The striking difference in the administration of Egypt
(long a province of the Roman empire) and Italy (which in the 530s was
11

12
13

14

15

Fidelis: Proc. BG 1.14.5 and 2.12.278, 2:76 and 2:203; his appointment as Praetorian Prefect
is mentioned at BG 1.20.20, 2:104. Reparatus was the brother of Pope Vigilius (Proc. BG
1.26.2, 2:127) and therefore a Roman by birth (Liber Pontificalis 61.1, ed. Mommsen, p. 148).
His appointment as Praetorian Prefect is mentioned at BG 2.21.40, 2:246. For their careers, see
PLRE 2:46970, s.n. Fidelis and ibid., pp. 93940, s.n. Reparatus 1. On the Praetorian Prefects
of Italy in general, however, see Brown, Gentlemen and Officers, p. 26 and Jones, Later Roman
Empire, 1:292.
Brown, Gentlemen and Officers, p. 32.
PLRE 3:968, s.n. Apion 3; ibid., pp. 11819, s.n. Aristomachus 2; ibid., pp. 2623, s.n.
Callinicus 4; ibid., p. 423, s.n. Dorotheus 7; ibid., pp. 4378, s.n. Elias 3; ibid., pp. 5823, s.n.
Hephaestus; ibid., p. 642, s.n. Ioannes 31 Laxarion; ibid., p. 664, s.n. Ioannes 59; ibid., p. 704,
s.n. Ioannes 247; ibid., p. 981, s.n. Paulus 26; ibid., p. 1011, s.n. Petrus 56; ibid., p. 1021, s.n.
Philiades; ibid., pp. 11212, s.n. Senuthius 1; and ibid., pp. 13723, s.n. Victor 4. See also
ibid., pp. 4089, s.n. Domentianus; ibid., pp. 7334, s.n. Iulianus 12; and ibid., p. 1294, s.n.
Theodosius 15. See further perhaps ibid., p. 105, s.n. Archelaus 2 and also below, next n.
A good recent synopsis of the administrative reorganization of Egypt in late antiquity is provided
by B. Palme, The Imperial Presence: Government and Army, in R. Bagnall (ed.), Egypt in
the Byzantine World, 300700 (Cambridge, 2007), pp. 2459; see also Jones, Later Roman Empire,
1:281. The most recent synthesis of the Byzantine military administration of Egypt is still J.

Maspero, Organisation militaire de lEgypte

byzantine, Biblioth`eque de lEcole
des hautes e tudes,
Sciences historiques et philologiques 201 (Paris, 1912), here pp. 729; on the governors regional
origins, see ibid., p. 83.
Service in native province: PLRE 3:664, s.n. Ioannes 59, and ibid., pp. 13723, s.n. Victor 4;
see also PLRE 3:105, s.n. Archelaus 2, and ibid., pp. 7334, s.n. Iulianus 12. Non-Egyptians:
see PLRE 2:67781, s.n. Liberius 3, and PLRE 3:92830, s.n. Narses 2; ibid., pp. 9403, s.n.
Nicetas 7; ibid., pp. 10856, s.n. Rhodon; and ibid., p. 1418, s.n. Zeno 1. See also perhaps
PLRE 3:7504, s.n. Iustinus 4.

201

Staying Roman
Table 4.1. Supreme commanders of the Byzantine forces in Africa:
regional origins
Name

Title

Dates

Region of origin

Belisarius

MVM per Orientem

5334

Solomon

magister militum (?)

Germanus
Sergius
Areobindus
Artabanes
John Troglita
Marcian
Theoctistus
Amabilis
Gennadius

magister militum praesentalis

magister militum Africae
magister militum Africae
magister militum Africae
magister militum Africae
MVM ()1)
magister militum Africae
magister militiae Africae
(1) magister militum Africae
(2) exarchus Africae
exarchus Africae
exarchus Africae

(1) 5346;
(2) 53944
5369
5445
545
5456
546551/2
563564/5
570
571
(1) 57785
(2) 5918
602?10
6457

Germania
(Balkans)
Mesopotamia

Heraclius
Gregory

Illyricum (?)
Mesopotamia
Constantinople (?)
Armenia
Thrace (?)
Illyricum?
?
?
?
Armenia
Armenia?

MVM = magister utriusque militiae

For references to individuals not specifically mentioned in the text, see PLRE 23 fastes.

contested between the Byzantines and the Goths) calls for another point
of comparison.
In Byzantine Africa, nearly all of the highest-ranking civil and military officials whose regional origins are known to us were themselves
recruited from what were then the most militarily active frontier regions
of the empire: the Persian and Balkan borderlands (see Table 4.1). Procopius tells us that Belisarius, the general in command of the initial reconquest, was from the province of Germania. To the sixth-century historian, the term referred to the northern Balkan region between Illyricum
and Thrace.16 John Troglita, magister militum Africae in the later 540s, seems
to have come from the Balkans as well. Procopius implies that he was from
Thrace itself, while his cognomen may derive from Trogilos (C:)),
a district in Macedonia.17 Solomon and his nephew Sergius who
16

17

Proc. BV 1.11.21, 1:363: 9  A  K)1  $ 

, N J E  / 'I)) E
G ". C. Jirecek identified this Germania with what is now the territory of western
Bulgaria: Archaologische Fragmente aus Bulgarien, Archaeologisch-epigraphische Mittheilungen aus
Oesterreich-Ungarn 10 (1886), pp. 712. The term Germania typically referred to the Rhineland
provinces: see, e.g., OCD, pp. 6334, s.v. Germania.
PLRE 3:644, s.n. Ioannes qui et Troglita 36. Trogilos was also the name of a district in Sicily.

202

New Rome, New Romans

successively combined the offices of Praetorian Prefect and magister militum to rule as civil and military governors of Africa were from the
Byzantine province of Mesopotamia.18 Indeed, Solomon was from the
region of Dara, an urban military stronghold commanding the very border between the Byzantine and Persian empires.19 Artabanes was from
Armenia, a region which in the sixth century was still contested between
Byzantium and Persia.20 The early seventh-century exarch of Africa,
Heraclius, was also of Armenian descent.21 To judge from his name, the
later exarch and would-be usurper Gregory may have been a relative of
Heraclius, and was thus perhaps an Armenian as well.22
We only know the regional origins of two Praetorian Prefects who
served after Solomon and Sergius, but the same pattern of frontier
recruitment holds for both of them (see Table 4.2). A late sixth- or
early seventh-century prefect named George came from the region of
Apamea in Byzantine Syria. Though not particularly close to the PersianByzantine border, Apamea was nevertheless twice sacked by invading
Persian armies in the course of the sixth century, once in 540 and again
in 573.23 Of course, Africa was itself one of the empires militarily active
frontiers, and in 600 we hear of an African holding the provinces highest civil office: on being elevated to the Praetorian Prefecture, in that
year, Innocent received a letter from Gregory the Great in which the
18

19
20

21

22

23

Proc. BV 1.11.9, 1:362; Zachariah of Mitylene, Chronicle 9.2, p. 223 (from the fortress of
Edribath[?]); Theophylact Simocatta, Historiae 2.3.1213, ed. C. de Boor with corrections by P.
Wirth (Stuttgart, 1972), p. 75. It is not clear that Solomon was Armenian, as in D. Pringle, The
Defence of Byzantine Africa from Justinian to the Arab Conquest: An Account of the Military History and
Archaeology of the African Provinces in the Sixth and Seventh Centuries, BAR International Series 99
(Oxford, 2001), p. 22.
Proc. BV 1.11.9, 1:362. On the foundation of Dara, see Proc. BP 1.10.1316, 1:478 and, in
general, ODB 1:588, s.n. Dara.
Proc. BV 2.24.2, 1:530 and Proc. BG 3.32.1, 2:433; see also Proc. BV 2.27.16, 1:542 and Proc.
BP 2.3.25, 1:157. Artabanes was a member of the Armenian royal house of the Arsacids, on
whom see ODB 1:186, s.n. Arsacids.
Thus PLRE 3:584, s.n. Heraclius 3, on the evidence of Theophylact Simocatta, Historiae 3.1.1,
pp. 10910, that Heraclius was ordered  & O # ) !)+"  2 P 

 ; accepted as a probability by W. Kaegi, Heraclius, Emperor of Byzantium (Cambridge,
2003), p. 21.
PLRE 3:554, s.n. Gregorius 19, and the literature cited there. There is a slight preponderance of
eastern Gregories in the sixth- and seventh-century Mediterranean. In addition to our Gregory:
East: PLRE 3:5467, s.nn. Gregoria 13, and ibid., pp. 54954, s.nn. Gregorius 4, 7, 1013,
1718 and 20. Greek seals: ibid., pp. 5525, s.nn. Gregorius 89 and 216. Italy: ibid., pp.
54753, s.nn. Gregorius 1, 56, and 15. Gaul: ibid., pp. 5489, s.n. Gregorius 3. The Vita S.
Gregorii Agrigentini, PG 98:549716, which records the voyage of another Gregory from Sicily
to Carthage and thence to Tripoli, is a ninth-century forgery.
George: John Moschus, Pratum spirituale 196, PG 87/3:3080D. Apamea: Proc. BP 2.11.238,
1:198203 and John of Ephesus, Historiae ecclesiasticae pars tertia 6.6, trans. E. W. Brooks, CSCO
(Scr. Syr.) ser. 3, 3 (Louvain, 1936), pp. 2212 with ODB 1:127, s.n. Apameia.

203

Staying Roman
Table 4.2. Praetorian prefects of Byzantine Africa: regional origins
Name

Dates

Regional origins

Archelaus
Solomon

534
(1) 5346
(2) 53944
5369
5445
545c.549/50
552
556/61
558
563
(1) 563/5
(2) 574?8
570
mid to late sixth century
582
582/602
589
594
600
late sixth/early seventh
century
627
c.640
6412
Sept. 641/Jan. 642

East (Constantinople?)
Mesopotamia

Symmachus
Sergius
Athanasius
Paul
Boethius
John 1
John 2 Rhogathinus (?)
Thomas
Theodore 1
Menas
Theodore 2
John 3
Anonymous
Pantaleon
Innocent
George 1
Gregory
George 2
George 3 (= George 2?)
Marinus

East (Constantinople?)
Mesopotamia
East (Constantinople?)
?
Italy?
?
?
?
?
Egypt?
?
?
?
?
Africa
Syria
Armenia?
?
?
?

For references to individuals not specifically mentioned in the text, see PLRE 23 fastes.

pontiff refers to the prefect as Augustine of Hippos patriota or fellowcountryman.24

Regardless of the prefects regions of origin, the extreme strategic and
economic importance of Africa to the sixth-century imperial endeavour in the West is underscored by the fact that the leading generals and
governors of Africa often had intimate connections to the highest circles
of power in Constantinople. I will return to this point below, but for
now a handful of examples will serve to illuminate the general point.
The magister militum Africae Areobindus was a senator from an aristocratic family and was, moreover, married to the niece of the emperor
24

Greg. Ep. 10.16, 2:845; see also ibid., 11.7, 2:869, again to Innocent.

204

New Rome, New Romans

Justinian.25 Similarly, the magistri militum Germanus and Marcian were
both cousins of the same emperor.26 Like Areobindus, the Praetorian
Prefect Symmachus was a member of the Constantinopolitan senate.27
Though nothing else is known about him, one cannot but imagine that
he may have had some connection to the western Symmachi, one of
whom owned a house in Constantinople that was destroyed in the Nika
riot of 532.28 Archelaus had held two critical appointments as Praetorian
Prefect (first of Illyricum, then of Oriens) before becoming the supply officer for Belisarius Vandal expedition and subsequently the first
Byzantine prefect of Africa.29 Athanasius had served as an envoy from
Justinian to the Goths and subsequently as Praetorian Prefect of Italy in
the 530s before assuming the African prefecture.30 Irrespective of their
regions of personal origin, then, all of these men must have had direct,
personal connections to the emperor in Constantinople. The Praetorian
Prefects and supreme military commanders of Byzantine Africa seem
often to have been of frontier origins: they were natives of the Balkans,
Mesopotamia, Syria, Armenia, even Africa. More importantly, though,
they were frequently men who had gone from the frontier to enjoy close
relations with the emperor in Constantinople and access to the highest
circles of imperial power.
The subordinate officers of the Byzantine army in Africa also seem
to have been overwhelmingly eastern in origin, at least throughout Justinians reign. The commanders of Belisarius expeditionary army, and
thus the new military elite in the first several years of Byzantine rule
in Africa, were drawn from a number of provinces throughout the
early sixth-century empire. Once again, however, the majority appear
to have been men of Balkan origins, supplemented by two Egyptians,
the Mesopotamian Solomon, and a number of federated barbarian commanders. Thracians appear to have been by far the best represented
regional group among the commanders of Belisarius army. Procopius
25

26
28
29

30

Areobindus: Proc. BV 2.24.1 and 2.24.3, 1:52930. One of his consular diptychs is preserved
in the Louvre: see C. Giroire and D. Roger, Roman Art from the Louvre (New York, 2007),
pp. 667 and J. Hubert, J. Porcher, and W. F. Volbach, The Carolingian Renaissance (New York,
1970), p. 237. He was presumably also related to the eponymous consuls of 434 and 506, and
therefore to Ardabur (cos. 447) and Aspar (cos. 434): see PLRE 3:1079, s.n. Areobindus 2, at
p. 107 and the references cited there. On the importance of proximity to the emperor in general,
see Jones, Later Roman Empire, 1:390.
27 Proc. BV 2.16.2, 1:497.
See below, n. 110. Cf. Brown, Gentlemen and Officers, p. 64.
Chronicon Paschale s.a. 532, ed. L. Dindorf, 2 vols., Corpus Scriptorum Historiae Byzantinae
1516 (Bonn, 1832), 1:623 with PLRE 3:1212, s.n. Symmachus 1.
Proc. BV 1.11.17, 1:363 and CJ 1.27.1 (ad 534), p. 77 with PLRE 2:1334, s.n. Archelaus 5.
The fact is strongly suggestive of the great expectations Justinian entertained for a reconquered
Africa.
PLRE 3:1424, s.n. Athanasius 1.

205

Staying Roman
tells us that, with the exception of Solomon and the Hun Agan, almost
all of Belisarius commanders came from the region of Thrace ( /
 J
 !
).31 The commander-in-chief of Belisarius regular
infantry, John, came from Dyrrachium (class. Epidamnus) in the province
of Illyricum and was therefore not a Thracian, though Procopius could
have considered him as coming from the regions toward Thrace, broadly
interpreted.32 Based on their names, it seems likely that the infantry
commanders Zadus and Sarapis were among the non-Thracians as well:
Zadus is perhaps the Arabic name Sacd or Zayd, while Sarapis was presumably from Egypt.33 Calonymus, the commander of Belisarius fleet,
was also an Egyptian.34 Although the name Cyprian is itself eastern in
origin, the strong connections of the name with Africa raise at least the
possibility that the Cyprian who commanded a detachment of foederati
in Belisarius army may have been part of the African diaspora in the
eastern Mediterranean (see above, Chapter 2.1). Pharas, the commander
of the 400 allied Heruls who accompanied the expedition, was himself a
Herul, and similarly the commanders of the 600 allied Huns were apparently Huns.35 However, there is no particular reason to doubt that three
of Belisarius six infantry commanders, at least seven of his nine commanders of foederati, and three of his four cavalry commanders did in fact
come from Thrace. Indeed, Rufinus, a cavalry commander and Belisarius standard-bearer, is explicitly called a Thracian later in Procopius
history.36
For our purposes, the composition of Belisarius command corps is
primarily important because these were the same men who became the
earliest officers of the Byzantine occupying army in Africa. Solomon,
who succeeded to supreme command of both civil and military affairs
in Africa upon the recall of Belisarius and Archelaus to Constantinople
in 534, had commanded a detachment of the foederati during the Vandal
campaign.37 Marcellus, another of the foederati commanders and one of
the presumed Thracians, was appointed dux Numidiae.38 Cyril, also a
commander of foederati, was sent to regain Sardinia and Corsica, then
stationed in Numidia as well.39 Indeed, Numidia was the largest and
31
33

34
35
36
37
38
39

32 Proc. BV 1.11.8, 1:362.

Proc. BV 1.11.10, 1:362.
So PLRE 3:141415, s.n. Zadus and ibid., p. 1114, s.n. Sarapis; I. Shahd, Byzantium and the
Arabs in the Sixth Century, 2 vols. (Washington, DC, 19952009), 1/1:180 is more cautious, but
see also ibid., pp. 1812.
Proc. BV 1.11.14, 1:362.
Proc. BV 1.11.11 and 2.4.29, 1:362 and 1:436 (Pharas); Proc. BG 4.19.67, 2:5856 (Sinnion).
Proc. BV 2.10.3, 1:459.
Proc. BV 1.11.5, 1:361; PLRE 3:116777, s.n. Solomon 1, at p. 1169.
Proc. BV 2.15.51, 1:495; PLRE 3:814, s.n. Marcellus 2.
Proc. BV 2.5.24 and 2.15.50, 1:439 and 1:495; PLRE 3:3712, s.n. Cyrillus 2.

206

New Rome, New Romans

most heavily fortified region of Byzantine Africa.40 No fewer than seven
of Belisarius original twenty-two commanders were stationed there in
the first two years after the fall of Carthage. In addition to Marcellus
and Cyril, the foederati commanders Althias and Valerian, the infantry
commanders Sarapis and Terentius, and the cavalry commander Barbatus
were all given commands in Numidia in the first years of the Byzantine
occupation.41 With the single exception of Sarapis, who as we have seen
was probably an Egyptian, all of these men are among the presumed
Thracians in Belisarius army. Although our sources are less clear on
this point, an eighth commander, the Herul Pharas, seems to have been
posted to Numidia as well.42
The only officer of the original twenty-two who seems to have
remained in Africa Proconsularis was Martin, yet another of the Thracian foederati commanders. When, in the spring of 536, the Byzantine
army revolted under the leadership of a certain Stotzas, Martin was
in Carthage. The rebellion threatened to swamp Byzantine control of
Africa, and Martin was quickly sent to Numidia to secure the loyalty
of the troops there not yet in revolt.43 Byzacena, Tripolitania, and the
few Byzantine outposts in Mauretania also seem to have received fewer
troops than Numidia. John Troglita perhaps another commander of
foederati, perhaps also the John sent by Belisarius to recapture the city of
Caesarea in Mauretania44 served as dux probably either of Byzacena
or of Tripolitania at least until the eruption of Stotzas rebellion.45 The
cavalry commanders Agan (a Hun) and Rufinus (a Thracian) also served
in Byzacena.46 John, one of Belisarius bodyguards, was sent to occupy
Septem, controlling the Strait of Gibraltar, and may have remained there
as tribunus in command of a small detachment of troops.47
After the recall of Belisarius to Constantinople in 534, we begin to see a
displacement of commanders of Balkan origins in favour of men from the
other frontiers of the sixth-century Byzantine empire. Thracians do not
appear to have been deliberately purged by the imperial administration:
the Thracian Himerius was dux Byzacenae in 544 when a revolt of the
40
41
42
43
44
45
46
47

Pringle, Defence, pp. 1024.

Proc. BV 2.13.2, 1:475 (Althias), ibid., 2.14.40, 1:488 (Valerian), and ibid., 2.15.501, 1:495
(Marcellus, Cyril, Barbatus, Sarapis, and Terentius).
Jordanes, Romana 369, p. 48; but see also Proc. BV 2.15.501, 1:495.
Proc. BV 2.14.3740, 1:4878.
Proc. BV 2.5.5, 1:439; on his career in general, see PLRE 3:6449, s.n. Ioannes qui et Troglita
36.
PLRE 3:645, s.n. Ioannes 36: Ioannes seems to have been dux of coasts and of territory adjacent
to Antalas, suggesting either Byzacena or Tripolitania.
Proc. BV 2.10.311, 1:45960.
Proc. BV 2.5.6, 1:43940, and PLRE 3:635, s.n. Ioannes 12.

207

Staying Roman
Moors of that province began under the chieftain Antalas, and a certain
Peter (another Thracian who had been one of Solomons bodyguards)
was still in Africa as late as 546.48 Even so, as Thracians were killed in
the line of duty or were transferred to other posts, we see the rise to
prominence of men from the empires other peripheries particularly
the Eastern Prefecture paralleling the generally eastward shift in the
regional origins of the prefects, magistri militum, and exarchs of Africa.
These newly prominent easterners included Cyrus and Sergius, nephews
of Solomon and therefore presumably also natives of Mesopotamia, who
were appointed to the command of Libya in 544 (see below, section 3.3).
A young Phoenician by the name of Severianus, who came from Emesa
in Syria, commanded a cavalry unit under Himerius dux Byzacenae in 544.
Although captured by the Moors, Severianus later escaped and returned
to Carthage, after which he disappears from sight.49 In 539, two brothers
of Lazic descent, Rufinus and Leontius, were among the commanders
sent to Africa with Solomon at the start of his second governorship.50
One of John Troglitas soldiers, a man by the name of Ornus, was said to
have come from Persia itself.51
From the mid sixth century onwards, Armenians enjoyed a certain
prominence among the subordinate officers in the military administration of Africa. In 545, the future magister militum Africae Artabanes was
sent there in command of a detachment of fellow Armenians, including
his brother John and his cousin (8>) Gregory.52 Artabanes foreignness as an Armenian must have been striking to the Romano-African
elite; at least, he is consistently referred to as the Armenian (Armenius)
by the African poet Corippus, who does not otherwise emerge from his
great epic as a man who was especially interested in his subjects regional
origins.53 Nevertheless, Corippus also reveals the presence of perhaps
two more Armenians in John Troglitas army in the 540s. The Gregory
who in the winter of 546 or 547 fought under Johns command may
plausibly though not certainly be associated with Artabanes cousin
48

49
50

51
52
53

Peter: Proc. BV 2.28.3, 1:545; see also ibid., 2.28.2433, 1:54850, where he assists Artabanes in
dispatching Guntharis bodyguards. Himerius: ibid., 2.23.317, 1:5257; Coripp. Ioh. 4.865, pp.
668. Contrast the purge of Lycians from the eastern imperial administration in 3925 discussed
by C. Kelly, Ruling the Later Roman Empire, Revealing Antiquity 15 (Cambridge, Mass., 2004),
pp. 489.
Proc. BV 2.23.69 and 2.23.17, 1:5267.
Proc. BV 2.19.1 and 2.20.19, 1:508 and 1:515. They are mentioned as being the sons of Zaunas
(PLRE 2:1196, s.n. Zaunas); therefore the grandsons of Pharesmanes and of Lazic origins (PLRE
2:8723, s.n. Pharesmanes 3).
Coripp. Ioh. 5.24851, p. 101. The name does not appear in F. Justi, Iranisches Namenbuch
(Marburg, 1895; repr. Hildesheim, 1963).
Proc. BV 2.24.2 and 2.27.10, 1:530 and 1:541.
Coripp. Ioh. 4.236, 4.361, and 4.367, pp. 75, 80, and 81.

208

New Rome, New Romans

of the same name.54 If the two Gregories are indeed one, he probably
fought alongside another of his relatives. At least a certain Arsaces who,
from his name, was presumably also a member of the royal Armenian
house of the Arsacids fought in John Troglitas army, too.55 Of course,
Armenians played an important role in the Byzantine army in general;
and, indeed, from the age of Justinian onwards it was military service, far
more than any other career, that brought Armenians into the Byzantine
ruling class.56 We hear, for example, of a Persarmenian defector named
Narses serving in the Thebaid, perhaps as dux, in the 530s or 540s.57 Yet
the role of non-native soldiers in the military administration of Egypt in
general seems to have been very limited.58 In Byzantine Italy, on the other
hand, T. S. Brown has shown that Armenian officers were quite common
in the sixth century, but in the seventh appear only to have been present
as commanders of expeditionary armies.59 Thus, both of these provinces
present contrasts with Africa, for we still hear of at least two Armenians
serving as regular officers there in the seventh century. At least John, a
dux commemorated in an inscription from Timgad (class. Thamugadi)
in Numidia, seems to be called Armenus, and was therefore presumably an Armenian.60 Similarly, as Constantin Zuckerman has shown, the
future Armenian prince Nerseh Kamsarakan served as dux Tripolitaniae in
the 650s.61
At least in the sixth century, though, the military administration of
Africa was not completely dominated by men from the eastern frontier. Three of John Troglitas officers bore the Germanic names Fronimuth, Geisirith, and Sinduit.62 Towards the end of the sixth century, a
Suevic commander named Droctulf sought to serve in Africa under the
54

55
56

57
58
60
61
62

The identification is made by PLRE 3:5478, s.n. Gregorius 2. Artabanes Armenian cousin:
Proc. BV 2.27.1019, 2.28.710, and 2.28.1416, 1:5413 and 1:5458. The military commander
under John Troglita: Coripp. Ioh. 4.4878, p. 85: third in line, furious Gregory gleamed with
a pillaged spear, a polished shield and Iberian darts (tertius inde furens rapta Gregorius hasta /
atque leui clipeo telis fulgebat Iberis). See also PLRE 3:547, s.n. Gregorius 2, which reads
this passage as meaning that Gregory was positioned among Iberian (presumably Armenian?)
troops.
Coripp. Ioh. 5.254, p. 101.
N. Garsoan, The Problem of Armenian Integration into the Byzantine Empire, in H. Ahrweiler
and A. Laiou (eds.), Studies on the Internal Diaspora of the Byzantine Empire (Washington, DC,
1998), pp. 616.
Proc. BP 1.19.37, 1:106; PLRE 3:92830, s.n. Narses 2.
59 Brown, Gentlemen and Officers, p. 66.
Palme, Imperial Presence, p. 258.
CIL 8.2389 + 17822 = ILCV 1832 = Pringle, Defence, p. 338, no. 57: In temporibus Constantini
imperatoris Bel. Gregorio patricio / Ioannes dux de Tigisi offeret domum Dei + Armenus.
Zuckerman, Haute hierarchie militaire, pp. 1745.
Fronimuth: Coripp. Ioh. 4.525, 5.446, 6.518, and 8.377, pp. 87, 109, 133, and 179. Geisirith:
2.188, 4.489, 5.326, 6.522, 8.372, and 8.475, pp. 34, 85, 105, 134, 179, and 183. Sinduit: 6.522
and 8.374, pp. 134 and 179. The others included Gregory, Putzintulus, and Tarasis. On their

209

Staying Roman
command of the exarch Gennadius.63 A Cappadocian named Theodore
was one of two commanders of the army which Justinian sent to reinforce Solomon in 534; the other was the son-in-law of Belisarius wife
Antonina.64 Theodores closest friend in Africa, Asclepiades, was from
Palestine; though his precise position is not clear, he does seem somehow
to have been associated with the Byzantine administration of Africa.65
Then too our best source for the events of the first decade or so of the
Byzantine occupation of the region, Belisarius assessor Procopius, was
also from Palestine.66
Finally, we also hear of a handful of Romano-Africans in positions of
command over the Byzantine army in North Africa. Pudentius, a native
of Tripolitania, seems to have held some kind of appointment in the new
military administration of Africa.67 Two others are mentioned only in
Corippus account of John Troglitas wars against the Moors. One was
a certain John, John Troglitas envoy to the Moorish chieftains Ifisdaias
and Cusina; the other was the tribune Liberatus Caecilides.68 Indeed, it
seems likely that by the 540s the Byzantine army had moved towards the
regular recruitment of locals as regimental officers in Africa. Certainly
this was common practice in contemporary Egypt, where tribunes were
generally drawn from among the notables of the towns where they and
their troops were stationed.69 Yet Corippus implies that there were few
Africans among the leading officers of John Troglitas army: the origins of
the war against the Moors in which they were engaged were obscure to
the Byzantine commanders, and they called upon Liberatus to explain the
situation to them.70 This is a literary device to be sure, but a significant
one nonetheless. Indeed, the men who held the most important military
commands in Africa seem overwhelmingly to have been foreigners
men of Balkan, eastern, and occasionally even Italian origins. This was
true not only in the initial years of the Byzantine occupation, but
throughout the ensuing decade or more as well. Even so, Africans like

63
64
66

67
68
69

identification as Johns commanders of the field army, see PLRE 3:5478, 1071, and 1216, s.nn.
Gregorius 2, Putzintulus, and Tarasis.
Greg. Ep. 9.9, 2:570; see also CIL 11.319 and Paul the Deacon, Historia Langobardorum 3.1819,
pp. 14650.
65 Proc. BV 2.18.3, 1:5056.
Proc. BV 2.8.24, 1:455; see also Proc. BG 2.7.15, 2:182.
Proc. BP 1.1.1, 1:4; Proc. Anecd. 11.25, 3:74, and (on Procopius in general) Averil Cameron,
Procopius and the Sixth Century, Transformation of the Classical Heritage 10 (Berkeley, Calif., 1985)
and Kaldellis, Procopius of Caesarea. Procopius was in Carthage in 536 when the army rebelled
against Solomon (Proc. BV 2.14.3941, 1:488) but need not have been there continuously from
533: see PLRE 3:10606, s.n. Procopius 2, at p. 1061.
Proc. BV 2.5.10, 2.21.3, and 2.21.1315, 1:440 and 1:51819; native of Tripolitania: ibid., 1.10.22,
1:359.
John: Coripp. Ioh. 7.24261, pp. 1534. Liberatus Caecilides: Coripp. Ioh. 3.4751, pp. 489.
70 Coripp. Ioh. 3.4151, pp. 489.
Maspero, Organisation militaire, p. 95.

210

New Rome, New Romans

Pudentius and Liberatus assisted the Byzantine endeavour in Africa from
the very start, and, within two or three generations of the reconquest,
the region had produced its own Praetorian Prefect.
2.2. Names
The presence in Africa of high officials from the Balkan and Persian
frontiers represents a certain Byzantinizing of African society. The same
trend is visible in the sixth- and seventh-century developments in the
African name stock. Regardless of their regional origins, the new elite
of sixth- and seventh-century Africa seem by and large to have participated in a broadly Byzantine onomastic culture also focused on the
eastern Mediterranean and the newly reconquered western provinces
of Italy and southern Spain. This is particularly true of names like
John, Theodore, Peter, Stephen, Paul, Thomas, and Gregory, popular
throughout the empire, and borne too by sixth- and seventh-century
magistri militum, exarchs, and Praetorian Prefects of Africa.71 But it is also
the case with, for example, the name George, derived from the Greek
word georgios (peasant) and popular among the intellectuals and state
functionaries of the fifth- and sixth-century East.72 So too other Greek
names like Gennadius, Leontius, Photinus, and Cyril, and even Latin
ones like Marinus and Julian, by which generals, commerciarii, sacellarii,
and minor civil officials in Africa were called in the sixth and seventh
centuries.73 It is important to note, however, that all of these names had
been known in late Roman and Vandal Africa, and indeed the names
Theodore, Leontius, Peter, and especially Paul had enjoyed considerable
popularity there in the fourth, fifth, and early sixth centuries.74 Moreover, in Byzantine Africa, eastern names were presumably chosen at least
on occasion for reasons of fashion, just as they were in contemporary
Italy.75 The names of the Afro-Byzantine elite thus do not necessarily
speak of an eastern presence in the province in the sixth and seventh
71

72

73
74
75

On the first five of these names, see in general ODB 2:10423, s.n. John and ODB 3:1604,
1636, 1953, and 2039, s.nn. Paul, Peter, Stephen, and Theodore. On the last two, see PLRE
23 s.nn. Thomas and Gregorius. On the role of the regional name-stock in Italo-Byzantine
identity, see M. McCormick, The Imperial Edge: Italo-Byzantine Identity, Movement and
Integration, ad 650950, in Ahrweiler and Laiou, Internal Diaspora, pp. 201.
On the name in general, see ODB 2:834, s.n. George. At least one George was probably a
native of Byzantine Africa: lector in the church of Sila in Byzantine Numidia, son of Tiberius and
Capria, who was 24 when he died: AE (1969/70), 21011, no. 703.
On these names, see PLRE 23 s.nn.
PCBE 1 lists seven Theodores (pp. 11079), seven Leontii and one Leontia (pp. 6324), nine
Peters (pp. 8704), and seventeen Pauls (pp. 83948) for the pre-Byzantine period.
On names in Byzantine Italy, see Brown, Gentlemen and Officers, p. 68.

211

Staying Roman
centuries; but they do contribute to the generally Byzantine feel of the
elite African name-stock.
At the same time, some of the names of high officials and military
commanders do appear to have been fairly new to Africa in the Byzantine period. Thus, for example, a sixth-century Praetorian Prefect bore
the eastern name Menas, which appears to have been unattested in the
name-stock of pre-Byzantine Africa.76 In 594, another Praetorian Prefect had the equally eastern name Pantaleon. In this case the name may
have had some earlier currency in Africa a late fifth- or early sixthcentury inscription from Hadra in Byzacena commemorates the martyrs
Pantaleon, Julian, and [their] companions but in general the African
cult of St Pantaleon seems to have developed after the Byzantine reconquest, and to have focused on the Nicomedian martyr of that name.77
Moreover, both in the case of Pantaleon and in that of Menas, the sixthcentury prefects appear to represent the unique attestation of each name
among the late antique North African elite. The same is true of the name
Theoctistus, borne by a later sixth-century magister militum Africae.78 Similarly, none of the names of the military commanders in Tripolitania who
remained loyal to the emperor Phocas in the face of Heraclius rebellion
in 609 Mardius, Ecclesiarius, Isidore seems to recur in the prosopography of late antique North Africa.79 A sacellarius who served two terms
during the reign of the emperor Constans II (ad 64168) and a magister
militum in Mauretania Caesariensis were both named Maurice, a name
that seems to have been new to Africa but which is attested throughout the Byzantine world in the sixth and seventh centuries.80 The name
Sergius is likewise unattested in Africa before the Byzantine reconquest.
Indeed, in the fifth and early sixth centuries the name Sergius is attested
76
77

78
79

80

CIL 8.22655, no. 1 = 15.712; see also CIL 10.8072, no. 7.

Praetorian Prefect: Greg. Ep. 4.32, 1:2512. Martyr: Duval, Loca sanctorum, 1:1213 no. 56
(Hadra) and 2:6656; see also C. Courtois, Sur un carreau de terre cuite representant saint
Pantaleon, Karthago 3 (19512), pp. 20913. The quotation is from Y. Duval, Loca sanctorum

francaise
Africae: le culte des martyrs en Afrique du IVe au VIIe si`ecle, 2 vols., Collection de lEcole
de Rome (Paris, 1982), 1:121 no 56: Hic habentur / memorie sa(n)c(tor)um / Pantaleonti /
Iuliani e(t) comitu(m).
John of Biclar, Chronicon s.a. 569.11, ed. R. Collins, CCSL 173A (Turnhout, 2001), p. 62.
John of Nikiu, The Chronicle of John, Bishop of Nikiu, trans. R. H. Charles (London, 1916),
109.23, p. 176. The names Ecclesiarius and Mardius occur only once in PLRE 3:434 and 826,
respectively. The name Isidore seems to have had imperial (and especially eastern) connotations:
PLRE 3:7236, s.nn. Isidorus 113.
C. Morrisson and W. Seibt, Sceaux de commerciaires byzantins du VIIe si`ecle trouves a`
Carthage, Revue numismatique, 6th ser. 24 (1982), p. 237, nos. 1920; Pringle, Defence, p. 334,
no. 48 = ILCV 234a. On the popularity of the name in Italy: PLRE 3:8545, s.n. Mauricius
2, and ibid., pp. 8612, s.nn. Mauricius 89 (3 of 10 examples), with PBE 1, s.nn. Maurikios
14 and 78 (6 of 8 seventh-century examples); see also ibid., s.n. Maurikios 6 (Sicily).

212

New Rome, New Romans

exclusively in the eastern Mediterranean; in the later sixth and seventh
centuries, it is attested in Italy as well, while six of the known seals of
seventh-century commerciarii found at Carthage were issued by men of
that name.81
A number of names may have had more specific regional
connotations.82 For example, the name Boethius (Praetorian Prefect
of Africa in ad 556/61) is strongly evocative of an Italian connection.
Indeed, though a prominent sixth-century bishop from Byzacena active
many years before the Byzantine reconquest was also called Boethius,
J. R. Martindale suggests that the sixth-century prefect may even
have been the son of the Italian philosopher.83 The Byzantine magister
militum in Byzacena named Pompeianus may have been from a family
that was western in origins, though the prevalence of the name Pompeius
in the Illyrian house of Anastasius also suggests intriguing possibilities.84
Gaudiosus, the magister militum Africae in 591, may also have been from a
western family.85
In the absence of statistically sound samples across different regions,
none of this evidence can be taken as anything other than suggestive.
The frequency with which these names are apparently attested, East
or West, is no certain guarantee of any particular individuals regional
origins. Put another way, the new popularity of eastern names is not
necessarily indicative of eastern immigration to Africa.86 In none of
81

82

83
84

85

86

CIL 8.22656, no. 25 (seal, from Carthage); Morrisson and Seibt, Sceaux de commerciaires,
pp. 229, 23031, and 2334, nos. 56, 1011, and 16. Fifth and early sixth centuries: PLRE 2:994
5, s.nn. Sergius 19. Later sixth and seventh century, eastern connotations: PLRE 3:112335,
s.nn. Sergius 14, 612, 15, and 3843; see also Greek seals and weights: ibid., pp. 11317, s.nn.
Sergius 1823, 25, 2837, and 4554. Italy: ibid., pp. 1128, 11301, and 1137, s.nn. Sergius 5,
16 and 55; see also Latin seals: ibid., pp. 11312, s.nn. Sergius 17, 24 and 27.
Among those borne by individuals of securely attested geographic origins are, for example,
Procopius, the name both of the patron saint of Caesarea and of the historian from the same city:
Proc. BP 1.1.1, 1:4 with ODB 3:1731, s.n. Prokopios (Q), saint.
PCBE 1:146, s.n. Boethos (bishop) and PLRE 3:2367, s.n. Boethius 1 (prefect). On the name
Boethius, see PLRE 23, s.n.
Pompeianus: sixth- or seventh-century magister militum: CIL 8.23230 = ILCV 233 = Pringle,
Defence, p. 336, no. 53. Conceivably from a family of Italian or even African origins? See PLRE
1:71214, s.nn. Pompeianus 111; also the proconsul of Africa ad 4001: PLRE 2:8978, s.n.
Pompeianus 2; the sponsor of games in Rome: Q. Aurelius Symmachus, Epistula 5.65, ed.
O. Seeck, MGH AA 6/1 (Berlin, 1883), p. 142 = PLRE 2:897, s.n. Pompeianus 1; and the
Africans in PCBE 1:881, s.nn. Pompeianus 1 and 3. Pompeiana: Greg. Ep. 1.46, 1.61, 3.36,
11.13, and 14.2, 1:60, 1:72, 1:182, 2:879, and 2:10678. However, see also A. D. E. Cameron,
The House of Anastasius, Greek, Roman, and Byzantine Studies 19 (1978), pp. 25976, esp.
25963 and stemma p. 274 on the name Pompeius in the Anastasian house.
Greg. Ep. 1.74, 1:823. On the name, which seems to have had western connotations, see PLRE
2:496, s.n. Gaudiosa; PLRE 3:505, s.n. Gaudiosus; and PBE 1, s.n. Gaudiosos 15; see also
Kajanto, Onomastic Studies, pp. 657.
Cf. Ostrogothic Italy, where it would seem that names could be chosen for reasons of profession
or confession: P. Amory, People and Identity in Ostrogothic Italy, 489554, Cambridge Studies in

213

Staying Roman
these cases is it possible to prove that the apparent westward spread of a
name even such a distinctively eastern name as Sergius represents the
movement of individuals rather than a change in the fashions of western
naming patterns. It is not even entirely clear that this apparent westward
spread is not simply an illusion created either by the prosopographies
(which are, after all, highly selective in their criteria for inclusion) or
by the sources themselves. The important point, however, is that the
names attested among the elite of Byzantine Africa are ones which, in
the sixth and seventh centuries, spoke of an onomastic culture that was
broadly Byzantine in character. Names like Sergius, Photinus, Gennadius,
Leontius, even George, appear to have been most popular within the
eastern provinces of the empire; they enjoyed some popularity in Italy,
Africa, and southern Spain, the western peripheries of imperial control;
but beyond the edge of empire, in Gaul and northern Spain, we hear
very little of these names, when we hear of them at all.
The onomastic evidence, of course, can really only be used fruitfully in
conjunction with the copious hard data that survive for particular individuals. As we have seen, at least a handful of Africans were consistently
involved in the civil and military administration of their natal province,
even occasionally at the highest levels. By and large, however, the highestranking civil servants and military commanders in Africa were themselves
eastern in origin. In this specific case, the pan-Byzantine character of the
elite name-stock and the pan-Byzantine origins of the provinces ruling
class appear to mesh seamlessly.
3. patterns of appointment
The security of Carthage, remote from Constantinople and beset by
slow communications with the imperial capital, was probably always a
concern to the Byzantine emperors. According to Procopius somewhat
hostile testimony, Justinians Praetorian Prefect of the East, John the
Cappadocian, estimated that it would take nearly five months (140 days)
to make the journey by land. If the estimate is exaggerated in either
direction, it has probably been stretched, not shortened, for John was
trying to dissuade Justinian from venturing the reconquest of Africa in
the first place.87 According to Procopius himself, who was present on
the voyage, Belisarius expeditionary force completed the sea passage to
Caput Vada in three months, hugging the shore and occasionally delayed

87

Medieval Life and Thought, 4th ser. 33 (Cambridge, 1997), pp. 8791, 97102, and 26372; for
Vandal Africa, see above, Chapter 1.3.
Proc. BV 1.10.14, 1:3578.

214

New Rome, New Romans

Table 4.3. Constantinople to Africa: late ancient itineraries
Date Itinerary

Nautical
miles
Comments

533 ConstantinopleHeraclea
1,400
(class. Perinthus, mod.
Marmaraereglisi)Abydus
(mod. C
anakkale)Sigeum (mod.
Yenisehir)MaleaCaenopolis
(class. Taenarum, mod. Cape
Matapan)MethoneZacynthus
Sicily near Mt. EtnaCaucana
(near Santa Croce Camerina,
Sicily)Gaulus (mod. Gozo)
MaltaCaput Vada
534 LesbosPeleponneseAfrica (near
1,200
Numidia and Mauretania?)
546 ConstantinopleAbydusSigeum 1,300
AegeanAdriaticSicily
CaucanaCaput Vada

Belisarius/Procopius: Delays at
Heraclea (5 days), Abydus
(4 days), Methone (unspecified
length); strong winds
AbydusSigeum and Malta
Caput Vada; gentle winds
SigeumMalea and
ZacynthusSicily (passage
between which thus took
16 days, apparently longer
than usual)
Vandal deportee mutineers
John Troglita

along the way (see Table 4.3).88 The movement of a fleet could take longer
than that of an individual ship or even a smaller convoy.89 Belisarius was
clearly worried, for example, that the Vandals would already have heard
of the movements of his expedition and have laid a naval ambush for the
Byzantine flotilla by the time they reached Sicily.90 However, the coastal
route detailed by Procopius linking the Aegean and Africa may have been
a fairly common one in late antiquity: in describing the journey of John
Troglita from Constantinople to Africa in 546, the poet Corippus gives
an itinerary that is strikingly like the one described by Procopius, while
the historian himself indicates that in 534 a shipload of Vandal warriors
who had been deported after the Byzantine reconquest mutinied against
their captors and followed a similar route back to Africa.91 Passage by
88

89
91

Proc. BV 1.15.31, 1:380; for his account of the voyage see ibid., 1.12.11.14.17, 1:36576. The
distance of 1,400 nautical miles for this journey in Table 4.3 is a rough estimate based on an
itinerary that sticks close to the coast. On the warships that made up this fleet, see in general
Pryor and Jeffreys, Age of the Dromwn, pp. 12361.
90 Proc. BV 1.14.3 and 1.14.810, 1:3734.
McCormick, Origins, pp. 482 and 491.
Coripp. Ioh. 1.159371, pp. 918; Proc. BV 2.14.18, 1:485. The distances of 1,300 and 1,200
nautical miles (respectively) for these journeys in Table 4.3 are rough estimates based on itineraries
that stay close to the shore. A direct route, over deep water, would be shorter (about 1,000 nautical
miles); but on the geographical, meteorological, and technological conditions favouring coastal
routes in general in the medieval Mediterranean, see J. Pryor, Geography, Technology, and War:
Studies in the Maritime History of the Mediterranean, 6491571 (Cambridge, 1988), pp. 12101;

215

Staying Roman
way of Crete may have been typical of the reverse journey, from Africa
to Constantinople.92 If staying close to land, then, the voyage from the
imperial metropolis to the African hinterland will thus have covered some
1,300 nautical miles or more. At a plausible average speed of 1.7 knots,
a late antique ship could have covered this distance in about 770 hours
of sailing: 64 days if putting into shore at night, 32 if sailing around
the clock. Other than that of Procopius, no contemporary estimate
survives of the time required to make the sea passage between Africa and
Constantinople, but the contemporary and near-contemporary evidence
for Mediterranean voyages of comparable distance suggests that such a
journey could well take from one and a half to three months.93
Given the amount of time that it could thus take for those in Constantinople to learn of developments in Africa and to respond, the dangers of disloyalty and administrative or military incompetence to imperial
control of the province must have been particularly acute. These dangers
were repeatedly demonstrated over the course of the first thirty years
of Byzantine rule in Africa. The rebel officers Stotzas and Guntharis
are both called tyrannus in the sources, implying that each claimed the
imperial title.94 In the late 530s, a certain Maximinus, one of the officers in the bodyguard of Theodore the Cappadocian, conspired with
the disgruntled soldiers to establish a tyranny ( 
) as well.95 In
the 540s, the generals under the Praetorian Prefect and magister militum
Africae Sergius were so disgusted with their commanders immaturity
and ineptitude that they refused to do anything to check the raids of the
Moors in Byzacena.96 In 563, John Rhogathinus (probably the prefect

92

93
94

95
96

P. Horden and N. Purcell, The Corrupting Sea: A Study of Mediterranean History (Oxford, 2000),
pp. 13343; and M. McCormick, Origins of the European Economy: Communications and Commerce,
ad 300900 (Cambridge, 2001), pp. 928.
M. McCormick, Bateaux de vie, bateaux de mort: maladie, commerce, transports annonaires et
le passage e conomique du Bas-empire au moyen a ge, in Morfologie sociali e culturali in Europa fra
tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sullalto Medioevo
45 (Spoleto, 1998), pp. 778; see also Pryor, Geography, pp. 945.
Speed of travel: McCormick, Bateaux de vie, p. 102. Comparable voyages: McCormick, Origins,
pp. 481500.
Stotzas: Proc. BV 2.15.1, 1:489; Marcellinus comes, Chronicon addit. s.a. 545.2, p. 107; Vict.
Tonn. s.aa. 541.129, 543.131, and 545.134, pp. 41 and 445; and Agathias of Myrina, Historiae
1.prooem.25, ed. R. Keydell in Agathiae Myrinaei historiarum libri quinque (Berlin, 1967), p. 8.
Guntharis: Proc. BV 2.25.28, 2.28.29, 2.28.34, 2.28.41, 1:536 and 1:54951; Marcellinus comes,
Chronicon addit. s.a. 547.6, p. 108; Jordanes, Romana 384, p. 51; Vict. Tonn. s.a. 546.136, p. 45;
Epistolae aevi Merowingici collectae 4 (Letter of the Milanese clergy, ad 552), ed. W. Gundlach,
MGH Epist. 3 (Berlin, 1892), p. 439 l. 40; Coripp. Ioh. 4.22235, pp. 745; Agathias, Historiae
1.prooem.25, p. 8.
Proc. BV 2.18.118, 1:5058.
Proc. BV 2.22.16, 1:5223. Eventually, John son of Sisiniolus moved against the Moors: ibid.,
2.23.132, 1:5259.

216

New Rome, New Romans

of Africa) provoked a rebellion among the Byzantines Moorish allies
by murdering an important Moorish leader of proven loyalty.97 At least
five times in three decades, disloyalty or incompetence had threatened to
wreck the Byzantine venture in Africa.
Though clearly not always successful, the appointment of capable,
trustworthy individuals to the governance and command of the African
provinces must therefore have been of prime importance to the central
administration. Five main questions will therefore govern the discussion that follows: First, how were appointees selected? Second, what
guaranteed their loyalty to the empire (or the emperor), and what guaranteed their competence? Third, how deep into the provincial administration did metropolitan appointments reach? Put another way, with
which appointments did the emperor concern himself personally, and
which did he leave to his provincial officers and administrators? Fourth,
how long could a provincial official expect to remain in his post before
being recalled to the capital or reassigned to a different post? And
finally a question whose answer must be deferred to the end of this
chapter what can these patterns of appointment tell us about the importance the imperial administration placed on the security of the African
provinces?
3.1. Military governors: exarchs and magistri militum Africae
An extraordinary wealth of information survives from which to answer
the first of these questions for the military governors of Africa. It reveals
four distinct trends in the pattern of their appointments. First, most of
the magistri militum and exarchs of Africa had already acquired considerable military experience, especially on the empires Persian frontier,
before being elevated to the command of the provincial forces in Africa.
Second, very close, personal connections to the emperor could be even
more important than military competence as a rule of advancement.
Third, while the supreme commander of the Byzantine forces in Africa
ultimately held his office at the pleasure of the emperor, in times of
crisis either a close association with the outgoing commander or the
acclamation of the army could lead to a field promotion which would
subsequently be confirmed from the metropolis. Finally, after the initial reconquest, local experience of Africa itself also seems to have been
particularly privileged in an exarch or magister militum Africae.
97

John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn,
1831), pp. 4956; Theophanes, Chronographia AM 6055, 1:2389. On his office, see PLRE 3:670,
s.n. Ioannes qui et Rogathinus 75.

217

Staying Roman
Table 4.4. Supreme commanders of the Byzantine forces in Africa:
previous careers
Name

Dates

Previous career

Belisarius

5334

Solomon

(1) 5346;
(2) 53944

Germanus

5369

Sergius
Areobindus
Artabanes

5445
545
5456

John
Troglita

546551/2

Marcian
Theoctistus

563564/5
570

Amabilis
Gennadius
Heraclius

571
(1) 57785;
(2) 5918
602?-610

Bodyguard of Justinian
dux Mesopotamiae
magister militum per Orientem
magister militum per Orientem II
notarius to Felicissimus dux Mesopotamiae
domesticus of Belisarius
commander of foederati (under Belisarius)
magister militum per Thracias
magister militum praesentalis, ex consule, patricius
dux limitis Tripolitaniae provinciae
member of the senate; patricius
rebel in Armenia; served under the Persian Great
king Chosroes I; defected to Byzantines
army commander (under Belisarius)
dux of Tripolitania or Byzacena
dux Mesopotamiae
?
dux in Lebanon?
magister militum per Numidiam?
?
(magister militum Africae)

Peter
Gregory

637
6457

general ()in the Persian wars; military

governor of Armenia
magister militum per Numidiam
Praetorian Prefect of Africa?

With one exception, all of the supreme commanders of the Byzantine forces in Africa whose earlier careers we can trace had served in
a military capacity elsewhere in the empire (see Table 4.4).98 In the
case of Justinians cousin Germanus this was in Thrace, where he had
commanded Byzantine forces between 518 and 527.99 However, magistri
militum Africae usually acquired their military experience on the Persian
frontier, the most militarily active of the empires eastern borders, rather
than in Thrace or in Italy. Early in Justinians reign, from 527 to 529,
98
99

See also Jones, Later Roman Empire, 1:386.

Proc. BG 3.40.56, 2:476; PLRE 2:5057, s.n. Germanus 4. On Thrace in the sixth century,

see P. Soustal, Thrakien (Thrake, Rodope und Haimimontos), Osterreichische

Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften 221/Tabula Imperii Byzantini 6
(Vienna, 1991), pp. 6973.

218

New Rome, New Romans

Belisarius had served as dux Mesopotamiae, and by 533 he was serving
for the second time as magister militum per Orientem.100 In the 540s, John
Troglita like Belisarius before him had commanded the Byzantine
troops in the province of Mesopotamia, though Procopius and the sixthcentury North African poet Flavius Cresconius Corippus give varying
evaluations of Johns performance at the battle of Nisibis in the Persian
Wars.101 Solomon had similarly served in Mesopotamia (under Belisarius), and was later one of the commanders of the foederati during the
Vandal campaign.102 The seventh-century exarch Heraclius had served
as a general in the Persian wars and as one of three military governors
of Armenia.103 Artabanes certainly did not lack military experience, but
he also had a complicated relationship with Justinians empire. Artabanes
had killed two Byzantine governors of Armenia and served the Persian
Great King Chosroes I in his wars against the Byzantines before deserting
to the Roman side with his brother and a band of Armenians.104 In the
very same year he was sent to Africa, doubtless in part because it was as
far as possible from his native Armenia. In the West, Artabanes proved
his loyalty to Justinian by assassinating the usurper Guntharis, who had
deposed and murdered Areobindus, and threatened imperial control of
Africa.105 The situation in Africa thus paralleled that of the few Italian
exarchs whose careers we can tentatively reconstruct and who were for
the most part also eastern commanders, administrators, and court officials though military service may perhaps have been held at a somewhat
greater premium in the empires southern Mediterranean provinces.106

100
101
102

103
104

105

106

PLRE 3:181224, s.n. Fl. Belisarius 1, at pp. 1827.

Battle of Nisibis: Coripp. Ioh. 1.5667, pp. 56; Proc. BP 2.18.1623, 1:2267. Service in
Mesopotamia: PLRE 3:6456, s.n. Ioannes 36.
Zachariah of Mitylene, Chronicle 9.2, p. 223 (accompanied Belisarius to Mesopotamia); Proc.
BV 1.11.56, 1:361 (commander of foederati on Vandal campaign). See also PLRE 3:116777,
s.n. Solomon 1, at pp. 11689.
Kaegi, Heraclius, pp. 215; see further PLRE 3:5846, s.n. Heraclius 3, and the sources cited
there.
Killed two Byzantine governors: Proc. BV 2.27.17, 1:542 with Proc. BP 2.3.67 and 2.3.25,
1:154, and 1:157. However, see also Proc. BP 2.3.27, 1:157 which reports an alternative account
in which a certain Solomon is said to have killed Sittas. Time with the Persians and defection:
Proc. BV 2.24.2 and 2.27.17, 1:530 and 1:542. On the date of their defection, see PLRE
3:12530, s.n. Artabanes 2, at p. 125.
Guntharis coup: Proc. BV 2.25.12.26.33, 1:532540; Vict. Tonn. s.a. 546.136, p. 45; and see
also Marcellinus comes, Chronicon addit. s.a. 547.6, p. 108. Artabanes counter-coup: Proc. BV
2.27.910 and 2.28.141, 1:541 and 1:54551; Marcellinus comes, Chronicon addit. s.a. 547.6, p.
108; Jordanes, Romana 384, p. 51; Vict. Tonn. s.a. 546.136, p. 45. See also Coripp. Ioh. 4.23242,
p. 75, who claims that the Praetorian Prefect Athanasius was actually responsible for the plot
and that Artabanes simply put it into action.
Brown, Gentlemen and Officers, pp. 645.

219

Staying Roman
It is perhaps no surprise to find tested men of extensive military
training and experience in positions of high command in the Byzantine
West, but it is nonetheless worth pointing out that this does not seem
to have been the norm in contemporary Egypt. There, as we have seen,
the provincial duces combined military and civil authority after 539 as the
highest-ranking local officials; but until the military disturbances of the
seventh century (at least) the men who held the ducate in Egypt rarely
seem to have been career soldiers. Nor was the post typically a steppingstone to higher military command, either locally or elsewhere in the
empire (see Table 4.5). Indeed, in Egypt the office of dux appears to have
been primarily conceived as a civil one with police powers.107
Of course, in the outwardly secure and economically critical province
of Egypt the emperors probably feared the threat of internal unrest or
rebellion far more than they did that of external invasion. Byzantine
Africa, by contrast, was a region geared for war. As such, it demanded
battle-hardened men. It was unusual for Africa to be the first command of
a magister militum. Indeed, the only supreme commander of the Byzantine
forces in the newly reconquered African provinces whom we know to
have had no direct experience of warfare when he was sent to the West
was Areobindus. In his total lack of a previous military career, Areobindus
seems to have been the exception rather than the rule. Procopius remarks
upon the senators inexperience in military affairs no fewer than three
times, and notes that when Areobindus finally did find himself in a
pitched battle, he fled, horrified by the brutality of men killing one
another.108
Areobindus appointment nevertheless proves that a personal connection to the emperor could, at least on rare occasions, trump even military
experience: as we have seen, the senator was married to Justinians niece,
Praejecta (see above, section 2.1). To be sure, Areobindus was appointed
in the midst of Antalas revolt, when many of the empires own soldiers
were joining the Byzantines one-time Moorish allies to make common
cause against the incumbent military governor, and when concern about
the stability and continued loyalty of Africa and its armies must have been
especially high.109 Even so, a significant number of the military governors of Africa were either family members or close, long-term, trusted
107
108

109

Maspero, Organisation militaire, p. 83.

Proc. BV 2.24.1, 1:52930 (! A )
!  E  ), ibid., 2.25.25, 1:535
(8
!  )
     #), and especially ibid., 2.26.16, 1:538: P* A
 E D    ,R ( 1 ! + # +1  0 0
M) )1  / 8)1  :   :  ;0. See also, however,
PLRE 3:8213, Marcianus 7, whose previous career is not clear.
Proc. BV 2.22.12.24.4, 1:52230. See also Jones, Later Roman Empire, 1:386.

220

New Rome, New Romans

Table 4.5. Duces of Egypt, c.538641: previous and subsequent careers
Name

Dates

Previous career

Alexandria
Hephaestus
Liberius 3

?545/6
538/9c.542

see below, Thebaid

PPO of Italy, c.493500
PPO of Gaul, 51034
(both under
Ostrogoths)

Nicetas 7

?61017

Led military expedition

to Libya and Egypt,
609
praefectus Augustalis, 538
?MVM in Egypt, 6401

Rhodon
5389
Theodorus 166 6412
Arcadia
Domentianus
Marcianus 6

6401
549/50

Thebaid
Apion 3

54850

Military commander in
Italy and Spain
(despite frets about
his lack of military
experience)
Patrician and comes
excubitorum 61213
(?)

? patrician
vir gloriosissimus,
honorary consul
consul, 539
comes domesticorum,
539
patrician, 547/8

Aristomachus 2 578/82

MVM and pagarch of

Arsinoe, 556

Urban Prefect of
Constantinople and
Curator domus
Augustae, after 582
MVM, ex consule, and
patrician (all
honorific); praefectus
Iustini; ? dux et
augustilis Thebaidis
(bis), 5668

Athanasius 3

before c.567/8

Cyrus 8

vir gloriosissimus,
referendarius
vir gloriosissimus

Hephaestus

mid sixth
century
mid sixth/
seventh
century
543/5

Ioannes 25

537

vir gloriosissimus

Gabrielius 3

Subsequent career

advocate in Alexandria

dux et praefectus
Alexandriae, ?545/6
(551?)
Praetorian Prefect of the
East, 5512
(cont.)

221

Staying Roman
Table 4.5. (cont.)
Name

Dates

Previous career

Ioannes 59

held previous office

Iulianus 19

mid sixth
century
578

Narses 2

535

MVM and praefectus

Iustinianorum (both
honorific)
Fought for Persians
against Romans, 527
Defected to Romans,
530

Senuthius 1
Theodorus 35

sixth century
577

patrician
decurio (?)

Subsequent career

?MVM or comes rei

militaris in Italy,
53840
?MVM or comes rei
militaris (East), 543

MVM = magister utriusque militiae

PPO = Praetorian Prefect
Names are listed above as given in PLRE 3 except for Liberius 3, who appears in PLRE
2. This list does not include duces whose previous or subsequent careers are unknown.

associates of the emperors they served. As mentioned above, two more of

Justinians magistri militum were relatives of the emperor. Germanus was
Justinians cousin, while Marcian was Germanus nephew and therefore
also a cousin of the emperor.110 Similarly, Gregory, the prefect of Africa
in 627, may have been a relative of the emperor Heraclius, though this is
far from certain.111 Although not a family member, Belisarius had long
been a trusted associate of Justinian, having served as an officer in his
bodyguard even before the Illyrian became emperor.112 Belisarius was
one of only two generals who had remained actively loyal to Justinian
during the fateful Nika revolt; it was Belisarius who led the attack on the
rebellious crowd gathered in the Hippodrome and thus turned the tide of
the uprising in Justinians favour.113 Once in Africa, however, even Belisarius loyalty was open to question. Procopius claims that the generals
sub-commanders wrote to Justinian accusing the general of plotting to
110

111
113

Germanus was the nephew of Justin I and therefore a cousin of Justinian; he is called Justinians
8> at, e.g., Proc. BV 2.16.1 and 2.23.23, 1:497 and 1:528. See further PLRE 2:505, s.n.
Germanus 4. Germanus wife and Marcians mother were sisters; he too is called an 8> of
Justinian: Malalas, Chronographia, p. 496; Theophanes, Chronographia AM 6055, 1:239 and PLRE
3:821, s.n. Marcianus 7.
112 Proc. BP 1.12.201, 1:58.
PLRE 3:553, s.n. Gregorius 16.
Proc. BP 1.24.4054, 1:1303; Malalas, Chronographia, p. 476; Chronicon Paschale s.a. 532, 1:621;
Theophanes, Chronographia AM 6024, 1:1816, esp. p. 185; Cedrenus, Compendium Historiarum
1:647; and Nicephorus Callistus, Historia Ecclesiastica 17.10, PG 147:244C-D.

222

New Rome, New Romans

set himself up as the ruler of an independent African kingdom. Justinian
perhaps tested Belisarius loyalty by giving him the option of remaining
in Africa or returning to Constantinople; but the general chose to return,
in triumph, to the imperial metropolis.114
Although the other magistri militum and exarchs lacked such intimate
connections to the emperors they served, at least two of them enjoyed
close relations with their predecessors. Solomon had been an associate
of Belisarius for at least six years by the time of the Vandal campaign,
by which point he had become the generals domesticus.115 Indeed, it was
Solomon whom Belisarius sent to Constantinople to announce the initial
victory over the Vandals in 533, and Procopius seems to indicate that
Belisarius hand-picked Solomon to succeed him upon his own departure
for the imperial capital.116 Solomon must have enjoyed the emperors
confidence as well, for Justinian entrusted him with the critical mission of
testing Belisarius loyalty by presenting him with the choice of remaining
in Africa or returning to Constantinople.117 Sergius, too, enjoyed close
relations with his predecessor, for as we have seen he was Solomons
nephew. Procopius indicates that the young mans appointment was a
product of Justinians respect for his then-deceased uncle.118 Similarly,
Artabanes gained the African command after avenging the death of the
previous magister militum. This, in turn, would seem to suggest that a
particularly close relationship with ones predecessor could lead to a
promotion should circumstances demand that a trusted magister militum
Africae be replaced.
The manner in which the dux Numidiae Guntharis went about staging
his coup in the 540s is also revealing of patterns of promotion to supreme
command of the Byzantine armies in Africa. While still feigning loyalty
to the sitting magister militum Africae Areobindus, Guntharis hoped to have
the general killed in battle with the Moors and thus himself be compelled

114

115
116
117
118

Proc. BV 2.8.18, 1:4523. However, see also Proc. Anecd. 18.9, 3:113, which seems overly
cynical in light of the fact that Belisarius exceptionally for the time was granted a formal
triumph upon his return; not the sort of treatment one would expect an accused traitor to
receive.
Proc. BV 1.11.56, 1:361 and Zachariah of Mitylene, Chronicle 9.2, p. 223 (accompanied Belisarius to Mesopotamia).
Proc. BV 1.24.19, 1:412 (sent to Justinian) and ibid., 2.8.23, 1:455 (Belisarius successor).
Proc. BV 2.8.4, 1:452.
Nephew of Solomon: Proc. BV 2.21.1, 2.21.16, 2.21.19, 2.22.1, 2.22.9, 1:51720 and 1:5223,
and Proc. BG 3.27.2, 2:417. Appointed civil and military governor of Africa: Marcellinus comes,
Chronicon addit. s.a. 541.3, p. 106; see also Proc. BV 2.22.1, 1:522. Together with Areobindus,
he is called B*0 , i.e., magister militum Africae: Proc. BV 2.24.4, 1:530; see also
Zuckerman, Haute hierarchie militaire, p. 171. Respect for Solomon: Proc. BV 2.22.11,
1:5234.

223

Staying Roman
by the army to assume the supreme command.119 In the end, things did
not go as Guntharis had planned and he had to seize power openly; if his
plots had borne fruit, however, it would seem that Guntharis believed the
field promotion would subsequently have been regularized by Justinian.
The fact that all of these men had also served in Africa, even at this
early date, highlights a further important trend in the appointment of the
regions military governors: local experience seems to have been held
at a premium. Although Procopius claims Sergius was both young and
green, he had been the dux of Tripolitania for perhaps somewhat over
a year before assuming supreme command of the Byzantine armies in
Africa, during which time he was actively involved in warfare with the
Moors.120 Similarly, as we have seen, John Troglita had not only served
in the Vandal campaign, he remained in Africa as a provincial dux and
fought against the rebel Stotzas in 536 and 537 (see above, section 2.1).
As with Johns service on the Persian frontier, Procopius and Corippus
give us very different appraisals of the commanders performance in this
campaign: the poet was later to claim that Scalae Veteres [the site of
an important battle] watched you with remarkable love (te Cellas Vatari
miro spectabat amore); the historian, that Johns forces were routed and lost
their battle standards mirus amor indeed!121 But then Procopius does not
seem to have been very enthusiastic about John Troglita: the History of
the Wars account of Johns subsequent tenure as supreme commander of
the Byzantine forces in Africa is cursory; the historians final assessment
of conditions in Africa under John is positive but far from exuberant;
and the generals final victory in battle over the Moors in particular
is referred to as unexpected ( #  ).122 Yet the general
had long experience in Africa, and this doubtless among many other
factors appears to have recommended him to Justinian. Gennadius too
had served in Africa, in his case as supreme commander, before being
appointed exarch of the province. His term as magister militum Africae
began in 577; the epigraphic evidence suggests that he remained in that
position through the reign of Tiberius Constantine (ad 57882), and
that he was still in office on 6 May 585.123 He may well have continued
119
120

121
122
123

Proc. BV 2.25.22, 1:535 (scheme) with ibid., 2.25.1, 1:532 (dux Numidiae).
Proc. BV 2.21.1316 and 2.21.19, 1:51920. A young man at the time of his appointment: BV
2.22.2, 1:522. On the specific post that Sergius held in Tripolitania and the date at which he
was appointed to it (Apr. 543/Apr. 544), see PLRE 3:11248, s.n. Sergius 4, at p. 1125.
Coripp. Ioh. 3.318, p. 59; Proc. BV 2.17.6 and 2.17.1317, 1:5013, and see also a little below,
ibid., 2.17.19, 1:503 where Johns troops are conspicuously not mentioned.
Proc. BV 2.28.4552, 1:5512; see also Proc. BG 4.17.202, 2:579.
John of Biclar, Chronicon, s.a. 577.47, ed. Collins, p. 69; CIL 8.2245 + 17671 = ILCV 795 = J.

Durliat, Les Dedicaces douvrages de defense dans lAfrique byzantine, Collection de lEcole
francaise
de Rome 49 (Rome, 1981), no. 28, pp. 6771 = Pringle, Defence, p. 329, no. 33; Durliat,

224

New Rome, New Romans

in the post until he was made exarch, by 591 (see below). The seventhcentury exarch Gregory may have served in Africa in a civil capacity as
Praetorian Prefect twenty years before holding the exarchate, though this
is by no means certain.124 The patrician John, sent to recover Carthage
from the Muslims in 698, may not have had experience in Africa, but he
had previously served against the Arabs, and had proven himself a skilled
general.125
Service in Numidia specifically may have been seen as a stepping-stone
to supreme command of the Byzantine forces in Africa. As we have
seen, Numidia was the most heavily fortified of the Byzantine provinces
of Africa, and perhaps the most militarily active as well (see above,
section 2.1). In addition to Guntharis, the exarch Peter appears to have
served as a general in Numidia in 633/4 before becoming civil and military governor of Africa in 637.126 Alfred Merlin has also proposed that
the magister militum Africae Theoctistus, mentioned by John of Biclar as
having been killed by the Moors in 569, should perhaps be identified
with the Theoctistus magister militum per Numidiam named on a sixth- or
seventh-century lead seal. Like John Troglita, Merlin suggests, Theoctistus may have commanded troops in one of the four military districts
of Africa before becoming supreme commander of the Byzantine forces
in the province.127 This same Theoctistus may also have served as one
of two commanders of the troops stationed in Lebanon who twice took

124

125

126

127

Dedicaces, no. 29, pp. 717 = Pringle, Defence, p. 328, no. 31; and Pringle, Defence, p. 334,
no. 49. See also CIL 8.12035 = ILCV 793 = Durliat, Dedicaces, no. 30, pp. 778 = Pringle,
Defence, p. 330, no. 36.
Patrician and exarch: Maximus Confessor, Disputatio cum Pyrrho, PG 91:288A and 353A and
Theophanes, Chronographia AM 6138, 1:343. Gregory may perhaps be the prefect of the same
name mentioned by Pope Honorius I in ad 627: Ep. 9, PL 80:478; see PLRE 3:553, s.n.
Gregorius 16, for the identification.
Nicephorus of Constantinople, Breviarium 41, ed. and trans. C. Mango, in Nikephoros Patriarch of
Constantinople Short History, Corpus Fontium Historiae Byzantinae 13 (Washington, DC, 1990),
p. 98; Theophanes, Chronographia AM 6190, 1:370; PBE 1, s.n. Ioannes 7.
Relatio factae motionis inter domnum Maximum monachum et socium eius coram principibus in secretario,
ed. P. Allen and B. Neil, in Scripta saeculi VII vitam Maximi Confessoris illustrantia, Corpus Christianorum Series Graeca 39 (Turnhout, 1999), p. 15, ll. 2837; V. Laurent, Une Effigie inedite
de Saint Augustin sur la sceau du duc byzantin de Numidie Pierre, Cahiers de Byrsa 2 (1952),
pp. 8795; N. Duval, Nouvelles recherches darcheologie et depigraphie chretiennes a` Sufetula

(Byzac`ene), Melanges darcheologie et dhistoire de lEcole

francaise de Rome 68 (1956), pp. 2846,
no. 9; V. Laurent, Les Sceaux byzantins du Medaillier Vatican, Medagliere della Biblioteca vaticana
1 (Vatican City, 1962), pp. 857, no. 92; Y. Duval, Le Patrice Pierre, exarque dAfrique?,
Antiquites africaines 5 (1971), pp. 20914; Pringle, Defence, p. 336, no. 54 (see also ibid, pp. 378,
no. 55); and in general PLRE 3:1013, s.n. Petrus 70. On Peter, see also Maximus Confessor,
Epistolae, PG 91:363649, here Epp. 1314, cols. 50944; Diffloratio ex epistola s. Maximi scripta
ad Petrum illustrem, in Maximus, Opuscula Theologica et Polemica, PG 91:1416; and Maximus,
Brevis enarratio christiani paschatis, PG 19:121780.
John of Biclar, Chronicon, s.a. 569.11, ed. Collins, p. 62; A. Merlin, in BCTH (1925),
pp. cclivcclv.

225

Staying Roman
part in Byzantine incursions into the Persian empire in the early 540s.128
This reconstruction of Theoctistus career, while speculative, seems reasonable, especially given the fact that the name is uncommon and the
career span and trajectory are both plausible.
3.2. Praetorian prefects
It is also possible to trace something of the previous careers of three
Praetorian Prefects of Byzantine Africa. As we have seen (above,
section 2.1), before becoming the first Byzantine prefect of Africa
Archelaus had held both the Illyrian and the Oriental prefecture critical
appointments which provide us with important insight into Justinians
expectations from the newly reconquered African provinces. Similarly,
Athanasius had already served as an imperial envoy to the Goths and
then as Prefect of Italy before holding the African prefecture. Both of
these men must have been known to Justinian personally, and as mentioned above each certainly would have had access to the highest circles
of power in the imperial metropolis. Their considerable experience with
respect to diplomacy and logistical questions must also have stood them
in good stead. The poet Corippus was later to praise the prefect Thomas
(c. 563/565) for having exercised skilful diplomacy when he too held
office in Africa. By 566, however, Thomas seems to have retired. J. R.
Martindale suggests that Thomas came out of retirement in c.571 to serve
a second term as prefect of Africa, though the name Thomas was not
uncommon in the later sixth and seventh centuries, and the two prefects
cannot necessarily be taken to be the same individual.129 However, if
Martindale is correct in reconstructing Thomas career, the man must
have brought no small degree of personal continuity to the administration of Africa. Having presumably been first appointed under Justinian,
inscriptions record that Thomas was in office again in the reigns of
both Justin II (ad 56578) and Tiberius II (ad 57882).130 As with the
regions military governors, then, formidable experience and access to
the highest circles of power seem to have been the defining characteristics of sixth-century appointees to the Praetorian Prefecture of Byzantine
Africa.
128

129
130

Proc. BP 2.8.2, 1:184 (relieved Antioch in 540); ibid., 2.8.1719, 1:1867 (abandoned Antioch);
ibid., 2.16.1719, 1:2234 (persuaded to participate in the invasion of Persia in 541); ibid.,
2.19.334, 1:236 (eager to return to Lebanon); and ibid., 2.24.13, 1:262 (accompanied the
invasion of Persia in 543). The association between the two is made in PLRE 3:12267, s.n.
Theoctistus 2.
PLRE 3:131719, s.n. Thomas 15. PLRE 3 lists 36 Thomases for the period ad 527641.
Durliat, Dedicaces, nos. 25 and 278, pp. 5962 and 6471 = Pringle, Defence, pp. 327 and 329,
nos. 29 and 323 (respectively).

226

New Rome, New Romans

Table 4.6. Sixth- and seventh-century commanders in Africa:
previous careers
Name

Position

Previous career

Domnicus

commander-in-chief of
infantry (5369)
dux Byzacenae (5456)
dux Numidiae (545)
magister militum Byzacenae
(sixthseventh centuries)
magister militum Byzacenae
(sixthseventh centuries)
magister militum (Carthage,
sixthseventh centuries)
applied to serve under
Gennadius in Africa (598)

comes domesticorum (?)

Marcentius
Guntharis
John 1
John 2
(=John 1?)
John 3
Droctulf

commander in Italy
bodyguard of Solomon
cubicularius, imperialis spatharius
cubicularius, imperialis spatharius
cubicularius, imperialis spatharius
dux (under the Lombards);
commander in Italy (under the
Byzantines); sub-commander
(S1) in Thrace

3.3. Subordinate officers

A. H. M. Jones long ago observed that for lower appointments the emperors probably had to rely on the recommendations of their ministers.131
This observation, however, raises a number of important questions, all
connected to the second main question with which this section began:
How did Byzantine generals come by their key subordinates? How were
these men selected, and to what extent if any did Constantinople
interfere in their appointment? These questions have important implications for our understanding of the operation and promotional patterns
of the late Roman and early Byzantine army not only in Africa but
throughout the empire.
The offices of a number of the commanders in Africa hint at a pattern of at least selective imperial appointments (see Table 4.6). Two of
the generals named on the lead seals from Carthage and Byzacena were
eunuchs who had served as imperial bodyguards (cubicularii, imperiales
spatharii) before being posted to Africa.132 An officer named Theodore
who served under Solomon during his first governorship of Africa held
131
132

Jones, Later Roman Empire, 1:391; see also Maspero, Organisation militaire, p. 95, of Egypt specifically.
G. Zacos and A. Veglery, Byzantine Lead Seals, 2 vols. (Basle, 197285), 1/3:1643, no. 2885 and
P. Monceaux, Enquete sur lepigraphie chretienne dAfrique, Revue archeologique, 4th ser.
2 (1903), 75, no. 15. See also ODB 3:19356, s.v. Spatharios; ODB 2:1154, s.v. Koubikoularios; and M. McCormick, Emperor and Court, in A. Cameron, B. Ward-Perkins, and

227

Staying Roman
the office of comes excubitorum, or commander of a select corps of imperial guards.133 The senator Domnicus accompanied Justinians cousin
Germanus to Africa to take command of the Byzantine infantry there
in 536. At the time of his appointment, Domnicus may already have
been comes domesticorum, another commander of troops attached to the
imperial household.134 At least on occasion, then, it would seem that
the emperor posted intimates of proven loyalty in Africa not only to
serve as magistri militum Africae, but also as subordinate officers.
The same, of course, was true elsewhere in the empire as, for example, in the Thebaid, along Egypts southern frontier135 but Jones is
surely right that this cannot have been the norm. Although our sources
speak of the emperor making the appointment of duces, for example, one
can only imagine that as in Byzantine Italy these appointments, along
with those of other subordinate military officers, were in fact generally
in the hands of the magister militum Africae and later the exarch.136 This
certainly seems to have been the case under Solomon. Allegedly as the
result of an accident which befell him when he was a child, Solomon was
himself a eunuch.137 While he had no children of his own, towards the
end of his governorship (as we have seen) he seems to have extended his
patronage to the three sons of his brother Bacchus: Cyrus, Sergius, and
the younger Solomon. The younger Solomon took part in his uncles
final campaign against the Moors, though it is not clear in what capacity;
but Procopius tells us that in 544 Justinian gave Cyrus and Sergius command of Libya. Sergius was made governor of Tripolitania, while Cyrus
was perhaps appointed dux of Libya Pentapolis (the latter in the diocese
of Egypt, and thus not under Solomons command).138 The appointment
was ultimately in the hands of the emperor, but the candidates were surely
put forward by their uncle, the Praetorian Prefect and magister militum
Africae. Similarly, loyalties to the man more immediately responsible for

133
134

135
136
137
138

M. Whitby (eds.), Late Antiquity: Empire and Successors, ad 425600, Cambridge Ancient History
14 (Cambridge, 2000), pp. 13563, esp. 1512. Cf. Brown, Gentlemen and Officers, pp. 656.
Proc. BV 2.12.17 and 2.14.35, 1:472 and 1:487 with PLRE 3:1248, s.n. Theodorus 9, and
ODB 1:6467, s.v. Domestikos ton Exkoubiton.
Proc. BV 2.16.2, 1:497; thus PLRE 3:41516, s.n. Domnicus 3. John of Dyrrachium, Domnicus predecessor, had died of disease. Domnicus was recalled to Constantinople with Germanus
and Symmachus in 539: Proc. BV 2.19.1, 1:508.
Egypt: PLRE 3:968, s.n. Apion 3, who was consul and comes domesticorum before becoming
dux Thabaidis; see also ibid., pp. 12567, s.n. Theodorus 35.
Contemporary Italy: Brown, Gentlemen and Officers, p. 68. African appointments: see the sources
cited in this paragraph.
Proc. BV 1.11.6, 1:361.
Younger Solomon: Proc. BV 2.21.19 and 2.22.1220, 1:520 and 1:5245. Cyrus and Sergius:
BV 2.21.1, 1:51718 with PLRE 3:374 and 1125, s.nn. Cyrus 3 and Sergius 4.

228

New Rome, New Romans

Table 4.7. Early commanders in Africa: previous careers
Name

Position

Previous career

Agan

commander in
Byzacena (534)

Cyril

commander in
Numidia (536)
dux Numidiae (536)

served at the battle of Dara (530); officer

of Belisarius bodyguard; cavalry
commander 5334
served at the battle of Dara (530);
commander of foederati 5334
served at the battle of Dara (530);
commander of foederati 5334
served in Persian wars under Sittas;
commander of foederati 5334
served at the battle of Dara (530);
commander of allied Heruls 5334
member of Belisarius household; cavalry
commander, bandifer 5334

Marcellus
Martin
Pharas
Rufinus

commander at
Carthage (536)
commander in
Numidia (536)
commander in
Byzacena (534)

his appointment may perhaps help to explain why a certain Ksl, apparently dux Tripolitaniae in 609, supported the rebellion of the two Heraclii
against Phocas.139 Perhaps most tellingly, however, in the autumn of 598,
the Suevic commander Droctulf decided that he wanted to serve under
the exarch Gennadius in Africa. In pursuit of this ambition, he secured
for himself a letter of commendation from Pope Gregory the Great, not
to the emperor but to Gennadius himself.140
In this respect, it is doubtless significant that no fewer than six of
Belisarius original twenty-two commanders had served with him in the
Persian wars or were members of his household (see Table 4.7). As we
have seen, both statements apply to his eventual successor Solomon. But
Marcellus, who was to become the first Byzantine dux Numidiae, had
fought with Belisarius on the eastern frontier too, serving under the
general at the battle of Dara in 530, where Belisarius dealt the Persian
army a resounding defeat.141 The same was true of Cyril and Pharas,
who were also given commands in Numidia, and Agan, who served in
Byzacena after the reconquest.142 Agan, moreover, had been a member
139

140
141
142

John of Nikiu, Chronicle, 109.24, p. 176 with PLRE 3:762, s.n. Ksl. See, however, Maspero,
Organisation militaire, p. 8, who argues that Tripolitania was attached to the diocese of Egypt in
the late sixth century, and note that Leontius, apparently dux Libyae, who had been appointed
to the province of Mareotis by Phocas, also supported Heraclius and he was outside the exarchs
jurisdiction: John of Nikiu, Chronicle, 107.4 and 1213, pp. 1678 (the quotation is from 107.4,
p. 167) with PLRE 3:77980, s.n. Leontius 28.
Greg. Ep. 9.9, 2:570.
Solomon: PLRE 3:116777, s.n. Solomon 1, at pp. 11689. Marcellus: Proc. BP 1.13.21, 1:62.
Agan: Proc. BP 1.13.20, 1.14.39, and 1.14.44, 1:62 and 1:712; Cyril: Proc. BP 1.13.21, 1:62;
Pharas: Proc. BP 1.13.1927, 1.14.323, and 1.14.39, 1:623 and 1:701.

229

Staying Roman
of Belisarius household. So had Rufinus, who also commanded troops
in Byzacena in the 530s.143
This pattern of appointing trusted associates to key subordinate commands presumably continued under subsequent magistri militum. The
rebel Guntharis, for example, had been one of Solomons bodyguards
before being appointed dux Numidiae.144 But, as the case of Droctulf
shows, not all subordinate appointees can have been personally known
to their commanding officer. Here too it is possible to glimpse something of the pattern of recommendations that might lead to an appointment. Two more of the original commanders on the Vandal campaign
were veterans of the Persian wars, although they seem not to have served
under Belisarius: Martin, who appears to have remained at Carthage after
the reconquest, and Dorotheus, who died before Belisarius expedition
reached Africa. Both had fought on the eastern front under the Byzantine
general Sittas, Dorotheus apparently holding the rank of magister militum
per Armeniam.145 Sittas himself must have been a close associate not only
of Belisarius but of the emperor as well. Like Belisarius, Sittas had served
as an officer in Justinians bodyguard before the future emperor ascended
the throne, and the commander was later favoured enough to marry the
sister of the empress Theodora. Moreover, in the early days of Justinians
reign, Sittas and Belisarius had campaigned together in Persarmenia.146
Previous to the Vandal war Martin and Dorotheus may not have fought
under Belisarius himself, but they presumably came recommended by a
trusted source. Belisarius himself would seem later to have exercised his
own influence on his subordinates behalf. At least a certain Marcentius,
who held the office of dux Byzacenae during the rebellion of Guntharis
in 5456, had previously served in Italy under Belisarius command.147
Here again the personal connections towards the top of the military
hierarchy are pronounced: Marcentius must have been an appointee of
Sergius (the nephew of Belisarius former domesticus) or Areobindus (the
husband of Justinians niece).148
The earliest subordinate officers in Africa were thus by and large
close associates of their commander. I suspect that the large number
143
144
145
146
147
148

Proc. BV 1.11.7 and 2.10.4, 1:361 and 1:459.

Proc. BV 2.19.6, 1:509 (bodyguard) and ibid., 2.25.1, 1:532 (dux Numidiae).
Dorotheus: Proc. BP 1.15.317, 1:747; Malalas, Chronographia, pp. 469 and 4723; and PLRE
3:4201, s.n. Dorotheus 2. Martin: Proc. BP 1.21.271.22.2, 1:11415.
On his career in general, see PLRE 3:11603, s.n. Sittas 1.
Proc. BV 2.27.46 and 2.27.31, 1:541 and 1:544 (Byzacena); Proc. BG 2.5.1, 2:170 (Italy). PLRE
3:81819, s.n. Marcentius.
Marcentius predecessor, Himerius, would seem still to have been in office when Solomon
was killed in battle with the Moors: Proc. BV 2.23.317, 1:5257 (Himerius); see also ibid.,
2.21.268, 1:5212 (death of Solomon).

230

New Rome, New Romans

of Thracians in Belisarius command corps was another aspect of the
same phenomenon. They were a known quantity to the Balkan general;
he could trust them.149 To be sure, Thrace was still one of the best
recruiting grounds for the Byzantine army in the sixth century, and
Belisarius later raised troops there for the Justinianic campaigns in Italy
on several occasions.150 But with the very notable exception of John
Troglita, few new Thracian officers were given commissions in North
Africa after Belisarius departure from the region in 534. It thus seems
reasonable to suppose that the generals personal connections to Thrace
had something to do with the fact that his officers as well as his soldiers
were recruited from there.151 As we have seen, Artabanes later travelled to
Africa together with his own band of Armenians, while Gregory, Arsaces,
and perhaps Ornus seem to have formed an Armenian group within the
army of John Troglita. Heraclius may have stacked the administration
with his own relatives, including the exarch Gregory and the earlier
Praetorian Prefect of the same name (who may in fact have been the same
person). Family ties, personal connections, and shared regional origins
seem to have mattered at all levels and at all times in the appointment of
subordinate officers to positions of command in Byzantine Africa.
3.4. Terms of appointment
This brings us to the third of our major questions: how long did these
appointments last? In answering this question, we will take into account
the evidence for the average term served by the magistri militum, exarchs,
and Praetorian Prefects of Byzantine Africa, and ask whether there
were significant differences in the length of the appointments of civilian
and military officers. We must also confront the important question of
whether the recall of subordinate officers was connected to that of the
magistri militum Africae who had appointed them. This in turn raises some
final questions as to the length of time such subordinate officers remained
in Africa, especially in comparison to the highest-ranking officials.
Africa remained a militarily active frontier throughout the Byzantine
period, and a number of the officers sent there were killed in the line of
duty. As we shall see, Solomon was killed fighting the Moors. Similarly,
the magistri militum Africae Theoctistus and Amabilis and the Praetorian
149
150
151

Cf. the provincial cliques discussed by Kelly, Ruling the Later Roman Empire, pp. 445, 489, and
1734.
Proc. BG 3.10.13, 2:3367; see also ibid., 3.12.4 and 3.39.1617, 2:347 and 2:4734; Jones,
Later Roman Empire, 1:660. On sixth-century Thrace, see above, n. 99.
Note that Belisarius adoptive son was also a Thracian, and accompanied the army during the
Vandal campaign: Proc. Anecd. 1.1520, 3:79.

231

Staying Roman
Prefect Theodore were killed in the three successive years between 568
and 570 by the Moorish king Garmules, before the king himself was
killed by Gennadius in 577.152 While commanding a detachment of cavalry in Byzacena in 534, Rufinus and Agan also engaged in skirmishes
with the Moors, by whom they were eventually captured and killed.153
Pudentius, too, died in battle with the Moors.154 The seventh-century
exarch Gregory, though in rebellion against the empire, was probably
killed by an Arab raiding party in 647.155 The general John son of Sisiniolus mortally wounded the rebel Stotzas, but was himself killed in the
same battle.156 Artabanes brother, also named John, died in the same
encounter.157
Not all such deaths were in battle. Areobindus, of course, was assassinated during Guntharis coup, and Guntharis himself was killed in
Artabanes counter-coup. Six of the earliest commanders in Numidia
Marcellus, Barbatus, Cyril, Sarapis, Terentius, and Pharas were executed by Stotzas in the rebellion of 536.158 Probably in the same year
John of Dyrrachium, the commander-in-chief of the Byzantine infantry
in Africa, died of illness.159 John Troglitas brother Pappus was also said to
have died early on in the course of the Byzantine occupation of Africa,
apparently of natural causes.160 Though we do not know what caused
the death of the general Maurice, at some point in the sixth or seventh
century he was buried in the basilica of Rusguniae, at the age of fiftyfive.161 As we have seen, Dorotheus, one of Belisarius commanders of

152
153
155

156
157
158
159
160

John of Biclar, Chronicon s.aa. 568.8, 569.11, 570.16, and 577.47, ed. Collins, pp. 612 and 69.
154 Proc. BV 2.21.15, 1:519.
Proc. BV 2.10.311, 1:45960; see also ibid., 2.11.22, 1:466.
Patricius (and exarch) by 645: PLRE 3:554, s.n. Gregorius 19. His rebellion and death in
647: Theophanes, Chronographia AM 6139, 1:343; Ahmad ibn-Yahya al-Baladhur, Kitab futuh
al-buldan, ed. M. J. de Goeje, in Liber expugnationis regionum (Leiden, 1866), p. 227; pseudoFredegarius scholasticus, Chronicae 4.81, ed. B. Krusch, MGH SRM 2 (Hanover, 1888), p. 162;
Agapius of Menbij (Mabbug), Kitab al-c Unwan, ed. and trans. A. Vasiliev as Kitab al-c Unvan:
histoire universelle ecrite par Agapius (Mahboub) de Menbidj, PO 5, 7, 8, 11 (Paris, 191015), 8:479;
Michael the Syrian, Chronicle 11.10, ed. and trans. J. B. Chabot in Chronique de Michel le Syrien,
Patriarche Jacobite dAntioche (11661199), 4 vols. (Paris, 18991910), 2:440; Gregorius Abu alFaraj Bar Hebraeus, Chronography, ed. and trans. E. A. W. Budge, 2 vols. (Oxford, 1932), 1:97;
Chronicon ad annum Christi 1234 pertinens c.126, trans. J.-B. Chabot, CSCO (Scr. Syr.) ser. 3, 14
(Paris, 1920), pp. 2034, although according to Agapius and the Syriac sources Gregory escaped
with his life and returned to Constantinople where he made peace with the emperor, Constans
II; see also PLRE 3:554, s.n. Gregorius 19.
Proc. BV 2.24.914, 1:5312; Marcellinus comes, Chronicon addit. s.a. 545.2, p. 107; Coripp. Ioh.
4.103200, pp. 703.
Proc. BV 2.24.15, 1:532.
First five: Proc. BV 2.15.509, 1:4956. Pharas: Jordanes, Romana 369, p. 48.
Proc. BV 2.16.2, 1:497; see also (on his appointment) ibid., 1.11.8, 1:362.
161 Pringle, Defence, p. 334, no. 48 = ILCV 234a.
Coripp. Ioh. 1.390403, pp. 1819.

232

New Rome, New Romans

the foederati, died before the Byzantine army even reached Africa, as did
five hundred of Belisarius soldiers.162
Civil and military officials in Byzantine Africa appear to have served
at the pleasure of the emperor rather than for a set term, but half of
the reconquered provinces first eight magistri militum remained at their
post for only about one year.163 As we have seen, Belisarius chose to
return to Constantinople almost immediately after the completion of the
reconquest.164 He probably spent less than twelve months in Africa: his
fleet had landed there in late June of 533 and the general was recalled
shortly after informing Justinian of Gelimers capture in March of the
following year.165 Similarly, Sergius administration of the North African
provinces began a period of rapid succession to the supreme military
command of the troops stationed there. None of Sergius, Areobindus,
or Artabanes retained the post for much more than a year. We have seen
that Areobindus was assassinated. Sergius was recalled from Africa and
sent to Italy;166 Artabanes desired to marry Justinians niece, whom he
had sent back to Constantinople, and therefore secured his own recall
to the capital.167 On the other hand, Germanus served for three years
as commander of the field army in Africa, returning to Constantinople
only when he was recalled by Justinian so as to secure the eastern frontier
of the Byzantine empire against the Persians.168 John Troglita retained
the supreme command for at least five or six years, from 546 to 551/2.169
Solomon served two terms as military commander of Africa and, though
the first lasted only two years, the second lasted six. Both ended in
disaster. In 536, Solomon was forced to flee Carthage in the face of a
mutiny amongst his troops; and in 544 he was killed in battle by the
Moors.170 Even so, there is no indication that Justinian was unhappy
162
163

164
165

166

167
168
169

170

Dorotheus: Proc. BV 1.14.14, 1:375. 500 soldiers: ibid., 1.13.20, 1:371.

On terms of office, see (for Egypt) Maspero, Organisation militaire, p. 84 and (in general) Jones,
Later Roman Empire, 1:3812 and Kelly, Ruling the Later Roman Empire, pp. 3743, 8993, and
1945.
Proc. BV 2.8.18, 1:452453.
Proc. BV 1.12.12, 1:365 (departure in Mar. 533) and ibid., 1.15.31, 1:380 (arrival in Africa
three months later); see also ibid., 1.21.1725, 1:4024 (entry into Carthage in mid September
533). Recall shortly after late March 534: ibid., 2.7.117 and 2.8.18, 1:44851 and 1:4523.
Proc. BV 2.22.2, 1:522 (misrule); ibid., 2.24.46, 1:530 (shared command of Africa with Areobindus); ibid., 2.24.78, 1:5301 (refused to co-operate with Areobindus); ibid., 2.24.16, 1:532
(sent to Italy).
Proc. BV 2.28.44, 1:551 and Proc. BG 3.31.24, 2:431.
Proc. BV 2.16.1, 1:497 (appointment); ibid., 2.19.1, 1:508 (recall ad 539/540); Proc. BP 2.6.910,
1:1745 (sent to Antioch).
Proc. BV 2.28.45, 1:551; Proc. BG 4.17.202 and 4.24.337, 2:579 and 2:6223; Marcellinus
comes, Chronicon addit. s.a. 547.6, p. 108; Jordanes, Romana 385, pp. 512; and Coripp. Ioh., esp.
ibid., 1.4853, p. 5.
Proc. BV 2.14.3741 and 2.21.268, 1:4878 and 1:5212.

233

Staying Roman
with Solomons administration of Africa and were it not for his untimely
death the generals second governorship might well have lasted even
longer. Thus it would seem that magistri militum considered particularly
successful or trustworthy by the emperor could remain in power for three
to six years, and possibly more.
Successful exarchs could serve for even longer. Gennadius was already
exarch of Africa by 591 and he held the post until 598 at least.171 Gennadius would thus appear to have spent seven years or more as civil and
military governor of the province, and possibly if his appointments as
magister militum Africae and exarch were continuous over twenty years of
his professional life in Africa in service to the empire. Similarly, Heraclius
was said to have been appointed to the exarchate by the emperor Maurice (ad 582602). The Armenian was still in office in 608 or 609 when
he and his son rebelled against the usurper Phocas.172 Indeed, Heraclius
tenure as exarch appears to have lasted a minimum of eight years: he
seems to have died in office at Carthage some time around 610.173 Thus,
both exarchs whose careers it is possible to trace over time remained in
office for at least seven years comparable to the terms served by their
Italian colleagues, and against the prevailing patterns of supreme military
command in Africa, a formidable tenure indeed.174
Praetorian Prefects seem to have served terms comparable in length
to those of the magistri militum Africae, perhaps three years on average.
In the initial years of the reconquest, the two officials were consistently
replaced at the same time, and (as has already been mentioned) in the
cases of Solomon and Sergius the two offices were combined in a single
individual. With the prefecture of Athanasius, appointed in 545, the
offices were separated once again, and the officials themselves began to
be assigned at different times and for terms of different lengths. Apart
from Solomon, there is no clear case of an individual retaining the
prefecture for more than three years. Even so, in at least two instances a
longer tenure is likely. Athanasius seems still to have been in office when
Corippus composed his Iohannis (c.549/550), though he had definitely
been replaced by 552.175 His prefecture thus probably lasted between
four and six years. Thomas was probably in office at least from c.574 to
171
172
174

175

Greg. Ep. 1.59, 1.723, 4.7, 6.62, 7.3, and 9.9, 1:701, 1:802, 1:223, 1:4367, 1:4456, and
2:570; see also ibid., 6.64, 7.2, and 9.11, 1:440, 1:444, and 2:572.
173 John of Nikiu, Chronicle, 110.13, p. 178.
Nicephorus, Breviarium 1, ed. Mango, p. 34.
Italian colleagues: PLRE 3:11646, s.n. Smaragdus 2 (two terms, each perhaps 5 years); ibid., pp.
10923, s.n. Romanus 7 (died in office after perhaps 7 years); ibid., pp. 2645, s.n. Callinicus
10 (perhaps 6 years); ibid., pp. 4356, s.n. Eleutherius (at least three years); and ibid., pp. 719
21, s.n. Isaacius 8 (18 years). The average term was thus slightly over 7 years, with a median of
56 years and mode of 5 years.
PLRE 3:1424, s.n. Athanasius 1.

234

New Rome, New Romans

578; though, as we have seen, he may have held an earlier term as well
(see above, section 3.2).176 Moreover, J. R. Martindale has suggested
that, though the sources only clearly illuminate the final months of his
administration, the prefect George (recalled in 642) may conceivably
have been in office as early as 633/4, in which case he would have
retained the prefecture for perhaps as long as nine years.177 On the other
hand, Symmachus was prefect for only three years178 and, as we have
seen, Archelaus remained in Africa for at most one year. Once again,
these terms are comparable with the observable patterns from Italy,
where Praetorian Prefects seem to have served an average of two to three
years.179
We are seldom able to see the reasons behind the imperial withdrawal
of long-serving officials, and so Georges removal from the African prefecture is particularly instructive. Early in 642, George was recalled to
Constantinople under the cloud of what Maximus Confessor refers to
as slander ( ;
).180 The year 641 was an unsettled one for
the empire, witnessing the death of the emperor Heraclius; the joint
accession of his sons, the half-brothers Constantine and Heraclonas; the
subsequent death (under suspicious circumstances) of Constantine; the
brief sole rule of Heraclonas, allegedly as a tool of his mother Martina;
the association of the young Constans II with his uncle Heraclonas rule;
and, finally (by January 642 at the latest), the deposition, mutilation,
and banishment of Heraclonas and Martina. Amidst such dizzying shifts
in the centre of power, it would not be at all surprising if George did
somehow run foul of those who backed the eventual winner, Constans II
(ad 64168). But, in 641, George is also said to have received a letter
on behalf of some Monophysite nuns that purported to be from the
empress Martina. The prefect rejected the communication as a forgery,
but one wonders whether George was quite right to have dismissed
it out of hand.181 Indeed, two lead seals struck late in 641 with the
images of Heraclonas and Constans II on the obverse bear on the reverse
the inscription + Marini prefecti et commerciariu (+ Of Marinus, prefect

176
178
179

180
181

177 As suggested by PLRE 3:5212, s.n. Georgius 50.

PLRE 3:131719, s.n. Thomas 15.
Proc. BV 2.16.2, 1:497 (appointment) and ibid., 2.19.1, 1:508 (recall).
PLRE 2:46970, s.n. Fidelis (died in office after perhaps slightly over 1 year); ibid., pp. 93940,
s.n. Reparatus 1 (died in office after roughly 1 year); PLRE 3:1424, s.n. Athanasius 1 (at least
1 year, perhaps 3); ibid., p. 90, s.n. Antiochus 2 (at least 2 years); ibid., p. 797, s.n. Longinus 5
(at least 8 years); and ibid., pp. 51516, s.n. Georgius 11 (more than 2 years).
Maximus Confessor, Ep. 16, PG 91:576D. On Maximus relationship with George, see also
ibid., 1, PG 91:36492.
Maximus Confessor, Ep. 12, PG 91:460509; on this incident see also ibid., 18, PG 91:5849.

235

Staying Roman
and commerciarius).182 In the opinion of Cecile Morrisson and Werner
Seibt the term prefectus is here an office, not simply a title; in which
case Marinus may have succeeded the beleaguered George as prefect of
Africa before Martinas fall from power in December 641 or January 642.
Georges recall may thus have been intimately connected to the problem
of communications with the imperial capital, the vicissitudes of power
in the metropolis, and the anxieties this raised in Constantinople about
the loyalties of imperial officials in Africa.
Commerciarii, fiscal officials whose fortunes were intimately linked with
the shifting currents of power in Constantinople, seem to have retained
their offices only for a short period, perhaps even a matter of months.183
Subordinate military officers, on the other hand, appear generally to have
continued to serve under successive military governors. Marcentius dux
Byzacenae, for example, served under both Areobindus and John Troglita,
and presumably retained his command throughout the intervening coup
and counter-coup.184 The general Ildiger was sent to Africa to reinforce
Solomon and remained there to fight the Moors under Germanus.185 As
we have seen, when Belisarius was recalled to Constantinople in 534, his
commanders remained behind to serve under Solomon. The term that
subordinate officers spent in Africa, therefore, does not seem to have
been limited by realignments in the political centre.
Even when confronted with a turnover in the senior command, then,
provincial duces and their subordinates presumably provided a degree of
stability and continuity to the military administration. This does not
necessarily mean, however, that officers of such rank remained in Africa
for longer terms than their superiors. The general John, son of Sisiniolus,
had spent six years in Africa before he died in battle but this term is
comparable to that of Solomons second governorship.186 Theodore the
Cappadocian remained in Africa from 534 until at least 536 or 537, when,
despite the rivalry between himself and Solomon, he consistently refused
to join the revolt of the Byzantine forces in Africa against Solomons
command.187 Again, though, Justinians cousin Germanus would seem
to have remained in Africa as magister militum for a similar length of time.
182

183
184
185
186
187

Morrisson and Seibt, Sceaux de commerciaires, pp. 2323, nos. 1314. Thus no. 13; the reverse
inscription of no. 14 is more properly: + MARINI / PREFCTI / ET COMMER/CIARIV
(i.e., lacks the second E in prefecti).
Ibid., esp. p. 240 and table p. 239.
Proc. BV 2.27.56 and 2.27.31, 1:541 and 1:544; Coripp. Ioh. 4.53240 and 5.4479, pp. 87 and
109. For the identification, see PLRE 3:81819, s.n. Marcentius.
Proc. BV 2.8.24, 2.15.49, and 2.17.56, 1:455, 1:495, and 1:501.
He was sent to Africa in 539: Proc. BV 2.19.1, 1:508. For his death in 545, see above, n. 156.
Proc. BV 2.8.24, 1:455 (sent to Africa); ibid., 2.14.3241, 1:4878 (mutiny of the army); ibid.,
2.15.6, 1:489 (refused to surrender Carthage to the rebels); ibid., 2.15.49, 1:495 (with Ildiger,

236

New Rome, New Romans

Only among regimental commanders and Justinians native RomanoAfrican allies do we find a group of men who demonstrably served in
Africa for ten years or more. If the tribune Liberatus did indeed fight in
the Vandal campaign, as Corippus implies, he would have served in the
Byzantine army in Africa for fifteen years by 548, when we last hear of
him.188 The Tripolitanian commander Pudentius similarly worked with
the Byzantines for ten years before his death.189 Pudentius and Liberatus
were both Africans, and may therefore have been special cases; but even
so, as in Egypt, in Africa tribunes like Liberatus probably served out their
long terms of service in the same place, unless by chance their regiments
happened to be redeployed.190 The sixth-century tribune Ziper who
to judge by his name was probably not an African had been stationed
at Rusguniae (mod. Tamentfoust) in Mauretania Caesariensis with his
unit, the numerus Felicium Justinianorum, for twelve years at the time that
he died.191
Indeed, much more than in Byzantine Italy, high-ranking commanders in Africa would seem to have been regularly and rapidly re-assigned
to distant posts. To be sure, senior officers served in Italy for an average of
only perhaps five years, but many of them remained in the region for well
over a decade (see Table 4.8). By contrast, a number of the commanders
sent to Africa in 5334 had already been removed to Italy itself within
three years. This was the case with Martin, Valerian, and Cyprian, all of
whom had served in the Vandal campaign, and also Ildiger, who had been
sent to Africa to reinforce Solomon.192 Domnicus, sent to Africa in 536
with Germanus, was recalled with him in 539.193 By 541, John Troglita,
then dux of Byzacena or Tripolitania, had ended his tour in Africa and
was serving on the empires easternmost frontier as dux Mesopotamiae.194

188
189
190
191
192

193
194

left by Belisarius in command of Carthage); ibid., 2.17.6 and 2.17.19, 1:501 and 1:503 (battle of
Scalae Veteres); ibid., 2.18.34, 1:5056 (informed by Asclepiades about the plot of Maximinus).
Service against Vandals: Coripp. Ioh. 7.385, p. 158. Last appearance: ibid., 7.374498, pp. 15862;
and in general PLRE 3:7901, s.n. Liberatus.
Proc. BV 2.21.15, 1:519 (death) and for his career see above, n. 67.
Egypt: Maspero, Organisation militaire, pp. 8891; see also (on Africa) ibid., p. 96 n. 8.
CIL 8.9248 = ILCV 442 = Pringle, Defence, p. 333, no. 45. The name is apparently Thracian:
V. Besevliev, Untersuchungen u ber die Personennamen bei den Thrakern (Amsterdam, 1970), p. 80.
Martin and Valerian: Proc. BV 2.19.2, 1:508 and Proc. BG 1.24.1820, 2:120; Marcellinus comes,
Chronicon addit. s.a. 537.2, p. 105. Cyprian: Proc. BG 1.23.1920, 2:116 (Rome); ibid., 2.23.2,
2.24.18, and 2.27.26, 2:251, 2:25960, and 2:2734 (Fiesole); ibid., 3.5.4, 2:318 (Florence); ibid.,
3.6.8, 3.12.1820, and 4.33.10, 2:322, 2:349, and 2:663 (Perugia); and ibid., 3.23.6, 3.25.21,
and 4.33.10, 2:400, 2:411, and 2:663 (death); see also the discussion in PLRE 3:36870, s.n.
Cyprianus. Ildiger: BG 2.7.15, 2:182.
Proc. BV 2.16.2 and 2.19.1, 1:497 and 1:508.
Corippus and Procopius both describe Johns exploits in Mesopotamia, Procopius typically in a
much less flattering light: Proc. BP 2.18.1623, 1:22931; cf. Coripp. Ioh. 1.5698, pp. 57. John
in Mesopotamia in 541: Proc. BP 2.14.12, 1:215. Stationed in Africa until 540/1: Coripp. Ioh.

237

Table 4.8. Subordinate commanders in Italy: terms of appointment

Name

Position

Date

Term

Aratius

?MVM or comes
rei militaris
MVM vacans
MVM
?MVM (vacans)

53840/549

211 years

53546/550
5935
53748

1115 years
2 + years
11 years

Constantinus 3 ?MVM vacans

Cyprianus
?commander of
foederati
MVM vacans
Francio 1
MVM

5357/538
53740

3 years
8 years (total)

killed by
his troops
executed
murdered

Fulcarius
Herodianus 1

553
53540

approx.
20 years
>1 year
5 years

killed in action
defected to Goths

5425
537/8540
5357/546
53749
53544
53740
5989
5912
53840/543

3 years
2 years
210 years
12 years
9 + years
3 years
11/2 + years
1 + year
26 years

5357

2 + years

5359/543
5389
5523
55366

48 years
>1 year
1 year
13 years

5357
5445

410 years
(total)

magister militum
magister militum
patrician
MVM
magister militum
infantry
commander

53640
54754/555
559
54750
5912
5357

31/2 years
7 + years
1 + year
3 years
1 + year
2 + years

mean: 5 years

mode: 2 years median: 3 years

Bessas
Castus
Conon 1

Ildiger
Innocentius 1
Ioannes 46
Magnus 1
Martinus 2
Maurentius 3
Mauricius 2
Narses 2
Paulus 4
Peranius
Philemuth
Sindual
Valentinus 1

Valerianus 1

Verus
Vitalianus 3
Vrsicinus 1
Averages

MVM (vacans)
infantry
commander
commander
?MVM vacans
cavalry commander
MVM (vacans)
cavalry commander
MVM
MVM
magister militum
?MVM or comes
rei militaris
infantry
commander
?MVM vacans
Herul commander
?MVM vacans
Herul commander,
MVM
cavalry commander

5405
c.568c.588

Comments

died on campaign
executed
perhaps two
tours; killed in
action

killed in action

MVM = magister utriusque militiae

Names given as cited in PLRE 3, q.v. for references; for Bessas see also PLRE 2.

New Rome, New Romans

In short, within less than ten years of the reconquest, all of the officers of Belisarius army whose careers we can trace and who had not
already died had been transferred to other posts in Italy or on the eastern
front.

4. byzantines and africans

The new elite of Byzantine Africa consisted of men of pan-imperial origins who would seem more often than not to have been non-Africans. At
the highest levels, these men were bound to the emperor through close,
personal ties: they were his relatives by blood or marriage, his bodyguards,
his attendants. They were the men he knew and trusted. They were also
men who had a proven record in warfare usually on the critical Persian front in diplomacy, or in administration. Among the slightly less
exalted ranks of the duces and subordinate officers, patronage networks
appear for the most part to have extended from the military governors of
Africa rather than from the emperors themselves. Otherwise, however,
the same observations seem generally to hold: by and large subordinates,
too, appear to have been long-term, trusted associates of their superiors.
Given the extent to which these military officers and civilian administrators probably formed a small and very closely knit community, then, we
are forced to ask what effect these men actually had in re-Romanizing
Byzantine Africa.
The discussion that follows will approach this difficult question from
four main directions. First, I will attempt to establish as well as possible
how many individuals and especially how many non-Africans seem to
have been involved in the direction of Byzantine Africa. Second, I will ask
what long-term connections, if any, these officials established with Africa
(especially with respect to acquiring property and establishing families)
and whether any structural impediments hindered the establishment of
such ties. Third, I will consider the languages that were spoken by the
new Byzantine elite, asking particularly whether there was a conscious
effort to appoint Latin-speakers to the new administration of Africa, and
exploring the evidence for sixth-century Africans abilities to understand
Greek. Finally, I will examine inscriptions and epitaphs from rural Africa
for evidence of a Byzantine presence beyond the cities and administrative
centres of the African prefecture.

3.2934, p. 48 implying a departure after the defeat of Iaudas in 540 and before the disturbances
under Antalas and his brother in 543. See also PLRE 3:645, s.n. Ioannes 36.

239

Staying Roman
4.1. Numbers
The imperial administration in Africa unquestionably represented a
thin stratum at the very top of provincial society. As we have seen
(above, section 1), the Praetorian Prefects officium consisted of 396 civil
servants and soldiers. Each of the six provincial governors was allowed
fifty administrators, while each of the five duces was assigned a staff of
forty-three aids, clerks, and subordinate officers. Adding to these the five
doctors and two grammatici also on the public payroll, and also the tribune
of Septem, the number of personnel involved in the administration of
Byzantine Africa comes to 931, not including regular soldiers. Even if we
allow for expansion over time the posts of magister militum Africae and
exarch, for example, were both created after Justinians rescript of 534
the total number of office-holders and staff must have been less than
two thousand. These figures compare well to those from contemporary
Egypt, and as in the East, in Africa the proportion of officiales to the
overall population must have been miniscule.195 Yet even so, on the face
of it, it seems unlikely that these individuals were ever exclusively eastern
in their origins.
Similarly, although the Byzantine army in Africa was repeatedly reinforced with eastern recruits over the course of the sixth century, locals
must have been begun to be recruited into its ranks almost immediately.
After the initial invasion of 5334, we know of at least six instances over
the course of the sixth century when magistri militum were accompanied to Africa by substantial numbers of reinforcements, five of them
connected to the disturbances of the 530s and 540s.196 Yet Solomons
general overhaul of the army in Africa in 53940, at the beginning of
his second governorship, included the weeding out of subversive elements and the recruiting of new soldiers into the army.197 Moreover,
in his rescript concerning the military organization of Byzantine Africa,
Justinian ordered Belisarius to enlist a suitable body of local men and
195
196

197

Palme, Imperial Presence, p. 251.

(1) Theodore the Cappadocian and Ildiger (ad 534): Proc. BV 2.8.24, 1:455. (2) Germanus
(ad 536): Proc. BV 2.16.1, 1:497; see also Marcellinus comes, Chronicon addit. s.a. 536.2, p. 104.
(3) Solomon (ad 539): Proc. BV 2.19.1, 1:508; see also Marcellinus comes, Chronicon addit. s.a.
539.5, p. 106. (4) Areobindus (ad 545): Proc. BV 2.24.1, 1:529; see also Marcellinus comes,
Chronicon addit. s.a. 546.3, p. 107; Vict. Tonn. s.a. 546.136, p. 45; and Coripp. Ioh. 4.825,
p. 69. (5) John Troglita (ad 546): Coripp. Ioh.; see also Proc. BV 2.28.45, 1:551; Marcellinus
comes, Chronicon addit. s.a. 547.6, p. 108; Jordanes, Romana 385, pp. 512. (6) Marcian (ad 563):
Theophanes, Chronographia AM 6055, 1:239; see also Malalas, Chronographia, p. 496, which seem
to have been the last troops noted as having been sent West before the bid to reconquer Africa
from the Muslims in 697/8. Bis Electi in Africa: CIL 8.17414; Jones, Later Roman Empire, 1:655.
Pringle, Defence, p. 27.

240

New Rome, New Romans

former soldiers (presumably of the Vandal kingdom) to serve as limitanei.
These men were both to defend the forts along the Moorish frontier and
to cultivate the land, with the express purpose of serving as the nuclei
for further settlement.198 Denys Pringle has argued that the land survey
undertaken in Africa between 534 and 539 would have led not only
to the reimposition in the newly reconquered provinces of the Roman
tax system, but also to conscription based on landholding.199 Whatever
recruitment did take place would appear to have been a success, for sufficient troops were raised in Numidia during Justinians reign to post a
regiment of them in Egypt.200 In Egypt itself, by contrast, local soldiers
served locally: there is no evidence to suggest that that regions recruits
served abroad in the Byzantine period.201 Even at its very strongest, then,
the number of easterners sent to govern and defend Africa for the empire
can only have represented a minute fraction of the overall population.
4.2. Family life
In Byzantine Italy, eastern troops were rapidly assimilated into local Italian society.202 The same seems to have been true in Africa. In the wake
of the reconquest, at least some of the soldiers of the Byzantine army had
begun to put down roots. Enlisted men had been allowed to coerce the
captive daughters and wives of defeated and exiled Vandal warriors into
legal marriages, a policy which Solomon was later to regret and even
(apparently) reverse. What exactly it meant to be a Vandal four or five
generations after the initial conquest and how precisely the Byzantines
distinguished Vandals from Romans in the newly reconquered African
provinces are questions whose answers remain unclear to us, but apparently these Vandal women were possessed of a certain amount of property. By 536, the soldiers had begun to lay claim to the landed estates
of their Vandal wives, and Solomons refusal to grant his soldiers these
estates was one of the factors that contributed to the mutiny in the army
198
199

200
201
202

CJ 1.27.2.8 (ad 534), pp. 7980; Jones, Later Roman Empire, 1:663.
Pringle, Defence, p. 67. Census: Proc. BV 2.8.25, 1:455. Taxes were collected in Africa by spring
534: CJ 1.27.2.18 (ad 534), p. 80. Such revenue was apparently abundant by c.540: Proc. BV
2.19.4: /  B*0  
  1!    &. On the roles of
conscription and volunteerism in the sixth-century army, see A. D. Lee, War in Late Antiquity:
A Social History (Oxford, 2007), pp. 7985, but also, e.g., Jones, Later Roman Empire, 1:61419

and J. Haldon, Recruitment and Conscription in the Byzantine Army c.550950, Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse, Sitzungsberichte 357 (Vienna,
1979), pp. 208.
Jones, Later Roman Empire, 1:655. Once in Egypt the unit presumably began to recruit locally.
Palme, Imperial Presence, p. 262.
Brown, Gentlemen and Officers, p. 67; cf. Egypt, where soldiers were for the most part locals:
Palme, Imperial Presence, pp. 2601.

241

Staying Roman
(Stotzas rebellion) in that same year.203 Once the mutiny had been put
down, Solomon deported the last remaining Vandal elements in Africa,
and not least all of their women presumably the new wives of his
veteran soldiers.204
By contrast, higher-ranking officers appear to have remained largely
unassimilated in Byzantine Africa, probably by imperial design.205 In
the early seventh century we hear of a local taxeotes (or bailiff acting
for a magistrate) who tried to escape the plague in Carthage by fleeing
to his suburban estate, where he molested one of his female peasants.206
Unlike this official, however, we never hear of military officers becoming
landowners in the North African provinces; though this may simply be
a product of the fact that the region has thus far yielded no sources
comparable to the Italian papyri. Apart from the tribune Liberatus, who
was himself an African, we also do not hear of officers marrying local
women. Indeed, both Belisarius and Areobindus brought their wives
with them to Africa.
The presence in Africa of Belisarius wife Antonina and Areobindus
wife Praejecta, however, may have been the exception rather than the
rule. Though John Troglita was certainly married, we do not hear of
his wifes presence in Africa.207 When Artabanes was sent to Africa as
a subordinate officer he too was married, though he and his wife were
estranged. Procopius indicates that she remained at home in Armenia,
though when her husband rose to the supreme military command in
Africa she made her way to Constantinople. In this case absence did
not make the heart grow fonder. While in the West, Artabanes conceived a desire to marry the widowed Praejecta, although in the event
the marriage never took place. Indeed, Procopius tells us that Artabanes
anonymous wife came to Constantinople specifically to complain to
Theodora about the conduct of her husband.208 One wonders, however,
if that was all there was to the move. In the seventh century, the emperor
Phocas perhaps already apprehensive as to the loyalty of his distant
203

204
205

206
207
208

Proc. BV 2.14.810, 1:4834; W. Kaegi, Byzantine Military Unrest, 471843: An Interpretation

(Amsterdam, 1981), pp. 4750. On soldiers marriages in general, see Lee, War in Late Antiquity,
pp. 14953.
Proc. BV 2.19.3, 1:508: /  T1  E  " U1.
On the struggle elsewhere in the empire between the landed aristocracy and the emperor over
the fiscal and administrative structures of provincial government, see P. Sarris, Economy and
Society in the Age of Justinian (Cambridge, 2006).
Anastasius the Sinaite, Per tn n Sin gwn patrwn 40, ed. F. Nau, Le Texte grec des
recits du moine Anastase sur les saints p`eres du Sina, Oriens Christianus 2 (1902), p. 83.
Coripp. Ioh. 1.197202, p. 11.
Proc. BG 3.31.13, 2:432; see also Proc. BV 2.28.434, 1:551 and Proc. BG 3.31.216, 2:4313.
Guntharis himself had hoped to regularize his rule by marrying Praejecta: Proc. BV 2.27.202,
1:543 and Jordanes, Romana 384, p. 51.

242

New Rome, New Romans

exarch appears to have held the women of Heraclius family in Constantinople as security for good behaviour. Heraclius was accompanied
to Africa by an extensive male family network, including his son, also
named Heraclius; the elder Heraclius brother Gregoras, who probably
held the position of magister militum Africae; and Gregoras son Nicetas.209
Nevertheless, the exarchs wife remained in the imperial capital, as did the
fiancee of the younger Heraclius, Eudocia Fabia, though she was herself a
member of a leading Romano-African family. In 610, when the Heraclii
rebelled, Eudocia Fabia was imprisoned by the emperor Phocas along
with her future mother-in-law, the elder Heraclius wife Epiphania.210
If Phocas precautions in this respect were common practice, Justinian
would seem to have made an exception in the cases of Praejecta (his own
niece) and of Antonina (the wife of one of his closest and most trusted
associates). However, for male relatives and possibly even daughters to
accompany a Byzantine officer to Africa does not seem to have been at all
exceptional. Belisarius and Antonina, for example, were joined in Africa
by the generals adoptive son.211 As magister militum Africae, John Troglita
too was accompanied by his young son Peter.212 Writing 200 years after
the fact, the Arab historian Ibn c Abd al-Hakam tells us that the daughter
of the exarch Gregory was among the captives taken in the wake of the
Byzantine rebels defeat and death at the hands of the invading Muslims
in 647.213 Some years before his own death, the general Maurice buried
209

210

211

212
213

Gregoras the brother of Heraclius: Nicephorus, Breviarium 1, ed. Mango, p. 34. Magister militum:
PLRE 3:546, s.n. Gregoras 3. On Heraclius, Gregoras, and Nicetas, see also: John of Antioch, frag. 218e, ed. Muller, in Fragmenta historicorum Graecorum, 5:37; Histoire nestorienne inedite
(Chronique de Seert) 2.82, ed. A. Scher, PO 4, 5, 7, and 13 (Paris, 190818), 13:5267; Agapius,
Kitab al-c Unwan, 8:449; Cedrenus, Compendium Historiarum 1:71112; Bar Hebraeus, Chronography, 1:87; Michael the Syrian, Chron. 10.25, 2:378; Chronicon 1234 c. 90, p. 177; Nicephorus
Callistus, Historia Ecclesiastica 18.55, PG 147:445. The later rebel exarch Gregory may well have
been the son of this Nicetas: PLRE 3:554, s.n. Gregorius 19. On the younger Heraclius, his
family, and their African sojourn generally, see Kaegi, Heraclius, pp. 2537.
John of Antioch, frag. 218f, ed. Muller, Fragmenta historicorum Graecorum, 5:38; Theophanes,
Chronographia AM 6102, 1:298; Zonaras, Epitomae 14.14, 3:303; John of Nikiu, Chronicle, 106.2,
p. 167.
Antonina (Belisarius wife): Proc. BV 1.12.2, 1.13.24, 1.19.11, and 1.20.1, 1:365, 1:372, 1:393,
1:396. Praejecta (Areobindus wife): BV 2.24.3, 2.26.18, 2.27.20, and 2.28.43, 1:530, 1:538,
1:543, and 1:551. Belisarius adoptive son: Proc. Anecd. 1.1520, 3:79.
Coripp. Ioh. 1.197207, p. 11.
Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, p. 46. As told in the Arabic sources for the
conquest of Africa the presence of Gregorys daughter has something of a fairy-tale quality to
it. For a brief introduction to the Arabic sources for the seventh century, see M. Brett, The
Arab Conquest and the Rise of Islam in North Africa, in J. D. Fage and R. Oliver (eds.),
The Cambridge History of Africa, 8 vols. (Cambridge, 197586), 2:4905; but also V. Christides,
Byzantine Libya and the March of the Arabs towards the West of North Africa, BAR International
Series 851 (Oxford, 2000), p. 42 and W. Kaegi, Muslim Expansion and Byzantine Collapse in
North Africa (Cambridge, 2010), pp. 2940. For a fuller discussion of the early Arabic historical

243

Staying Roman
two of his daughters, Patricia and Constantina the latter of whom was
only three years old in the same church in Rusguniae where he was
himself later to be interred.214 If he or his family were not themselves
from Africa, Maurice too would thus seem to have brought them with
him to the province. Brothers and cousins also seem to have been sent to
Africa with some frequency. As we have seen, Artabanes went to Africa
with his brother John and his cousin Gregory.215 John Troglita similarly
accompanied his brother Pappus to Africa in 533 when the two were
subordinate officers in Belisarius army; and the Lazic brothers Leontius
and Rufinus served together under Solomon.216 The presence of such
extensive male family networks brothers, sons, cousins, nephews
doubtless lent the officer corps of Byzantine Africa a greater sense of
solidarity. The fact that the highest-ranking officers wives seem typically
to have remained in Constantinople, on the other hand, probably tended
to serve as a check on the centrifugal forces this could otherwise inspire,
keeping the social focus of their distant husbands firmly on the imperial
capital and supplying de facto hostages for their continued loyalty.
4.3. Language
This generally eastern focus of the new elite of Byzantine Africa is further
illustrated in their almost certain preference for Greek over Latin.217 Of
course, we have seen a number of Latin-speakers among the new elite of
Byzantine Africa. Droctulf the Sueve would unquestionably have spoken
the language fluently, as would the Praetorian Prefect Boethius if he
was, in fact, the son of the philosopher. The same may well have been
true of the prefect Symmachus, who may have been drawn from the
same family; presumably also of Germanus and Marcian who, as cousins
of Justinian, were probably Illyrians. Gregory the Great assumed that
Gennadius could understand Latin, and indeed this was probably a safe
assumption of any officer in the late Roman army.
However, based on the regional origins of those whose personal histories we can trace, many members of the upper reaches of the Byzantine
administration of Africa both civil and military would for the most

214
215
216
217

tradition, see A. Noth and L. Conrad, The Early Arabic Historical Tradition: A Source-Critical
Study, trans. M. Bonner, Studies in Late Antiquity and Early Islam 3 (2nd edn; Princeton, NJ,
1994).
Constantina: Pringle, Defence, p. 333, no. 46 = ILCV 234b. Patricia: Pringle, Defence, p. 334,
no. 47 = ILCV 234c.
Proc. BV 2.24.2 and 2.27.10, 1:530 and 1:541.
John and Pappus: Coripp. Ioh. 1.375404, pp. 1819. Leontius and Rufinus: Proc. BV 2.19.1
and 2.20.19, 1:508 and 1:515.
For the role of Greek in Italo-Byzantine identity, see McCormick, Imperial Edge, pp. 223.

244

New Rome, New Romans

part have spoken Greek, probably in preference to Latin. As we have
seen, the vast majority of those about whose geographical origins it
is possible to speak with any certainty came from the Greek-speaking
provinces of the empire (Thrace, the Levant, Egypt) or from even further east (Armenia, Lazica), and Vassilios Christides has recently suggested
that the Greek or bilingual GreekLatin inscriptions dedicating imperial
fortifications that survive from Byzantine Africa were probably aimed
primarily at just such troops. The suggestion is rendered all the more
plausible by the fact that all but one of these inscriptions date to the
first decade of the imperial occupation, when the number of eastern
soldiers stationed in Africa would have been at its highest.218 An undated
Byzantine-era Greek inscription from Tabarka reserving a horse for use
by an imperial courier presumably had a similarly official audience.219 Of
course, we also hear of Armenian soldiers in Africa speaking Armenian
with one another, quite conscious of the fact that they could not be
understood by those around them.220
Carthage naturally retained its fiercely cosmopolitan character
throughout the Vandal period. Eastern charioteers, monks, arena hunters
and merchants all took part in the bustling life of the fifth- and sixthcentury metropolis. The name of the citys rectangular harbour, Mandrakion, is a Greek word meaning little square, while the name for the
palace prison, Ankon, is Greek for the corner. Procopius records a childrens chant playing on the names of the letters of the Greek alphabet,
that Gamma shall pursue beta, and contrariwise, beta shall itself pursue
gamma.221 This is clearly not just a Latin saying that had been translated
into Greek; Greek was its original language.
Beyond Carthage, however, it is difficult to tell how widely Greek
was spoken by the local population in the sixth century. A bilingual
218

219
221

Bilingual inscriptions: Durliat, Dedicaces, no. 2, pp. 911 = Pringle, Defence, pp. 3223, no. 15
(Bordj Hellal); Durliat, Dedicaces, no. 6, pp. 1821 = Pringle, Defence, p. 319, no. 5 (Madaura,
mod. Mdaourouch); and perhaps Durliat, Dedicaces, B, p. 90 = Pringle, Defence, p. 320, no. 8
(Churisa, mod. La Kessera). Greek inscriptions: Durliat, Dedicaces, no. 10, pp. 256 = Pringle,
Defence, p. 321, no. 12 (Thagura, mod. Taoura); Durliat, Dedicaces, no. 14, pp. 357 = Pringle,
Defence, pp. 3201, no. 9 (Sufes, mod. Sbiba); and perhaps Durliat, Dedicaces, no. 34, pp. 856
= Pringle, Defence, p. 322, no. 13 (Thagura), which may be a copy of the first inscription from
Thagura. Durliat, Dedicaces, no. 26, pp. 624 = Pringle, Defence, p. 328, no. 30 (Sidi Gherib), a
bilingual inscription, dates to the joint reign of Justin II and Tiberius II (ad 5748). Christides,
Byzantine Libya, p. 12; see also Durliat, Dedicaces, p. 107 n. 43. For the presence of eastern armies
in Africa, see above, n. 196.
220 Proc. BV 2.28.16, 1:548.
Pringle, Defence, p. 230 and p. 338, no. 56.
Proc. BV 1.21.14, 1:402:  1  :  *, / 1)   * : 
1 . F. M. Clover, Carthage and the Vandals, in J. H. Humphrey (ed.), Excavations
at Carthage 1978, Conducted by the University of Michigan, vol. 7 (Ann Arbor, Mich., 1982),
pp. 1011.

245

Staying Roman
GreekLatin inscription dedicating a fortification and dating to the joint
reign of Justin II and Tiberius II (ad 5748) may speak to a certain facility
with the language along the Mediterranean littoral.222 The Vandal-era
Arian scholars Fabianus and pseudo-Origen both understood Greek, as
did the Catholic Fulgentius of Ruspe.223 Indeed, Ferrandus (a deacon
from the metropolis) praised Fulgentius for speaking the language like a
native; but this was a result of the careful upbringing the future bishop
received at the hands of his mother. She may well have come from the
eastern Mediterranean herself, or have been part of the fifth-century
Greek-speaking diaspora in the city of Rome, where she had presumably
married Fulgentius father. Her name Mariana sounds eastern,224
and Ferrandus notes that she was particularly concerned that Fulgentius,
who would be living among Africans, learn Greek before he learned
Latin, so that he could pronounce Greek speech more easily, retaining
the aspiration just as if [he had been] raised there.225 She was wise to have
done so. Procopius gives us some insight into the kind of ridicule that
could greet Africans who spoke Greek haltingly, or with an accent, in
the scorn which he heaps upon the African-born Quaestor of the Sacred
Palace, Junillus. In Procopius estimation, Junillus was desperately underqualified for the office of Quaestor. Not being a lawyer (2!), the
historian complained with vitriol, Junillus was ignorant of the law; and
moreover he could function only in one of the capitals two languages:
Latin. Junillus had never learned Greek, he couldnt speak the language,
and even his subordinates laughed at him when he tried.226
4.4. The countryside
The limited numbers of the easterners involved in the imperial administration of Africa, combined with their generally eastern social focus
(at least at the highest levels), also raises the important question of how
deeply the non-African elite penetrated provincial society in the sixth
222

223
224

225
226

Durliat, Dedicaces, no. 26, pp. 624 = Pringle, Defence, p. 328, no. 30 (Sidi Gherib). See also
in general J. Desanges, Quelques considerations sur lusage du grec dans les ports de lAfrique
romaine, Graeco-Arabica 6 (1995), pp. 2736.
Dossey, Last Days of Vandal Africa, pp. 11316.
Eastern connotations: see PLRE 1:559, s.n. Marianus 2; PLRE 2:772, s.n. Marianus 2; and
PLRE 3:82930, s.nn. Marianus 12 and 45. See also ibid., p. 829, s.n. Marianus 3. Not
exclusively eastern: PLRE 2:722, s.nn. Marianus 1 and 3.
V. Fulg. 1, p. 11: quo facilius posset, victurus inter Afros, locutionem graecam servatis aspirationibus tanquam ibi nutritus exprimere.
Proc. Anecd. 20.1720, 3:1278; see, however, M. Maas, Exegesis and Empire in the Early Byzantine
Mediterranean, Studien und Texte zu Antike und Christentum 17 (Tubingen, 2003), pp. 1115,
esp. 1213, and below, Chapter 6, n. 149.

246

New Rome, New Romans

and seventh centuries. Beyond the higher ranks of the civil and military
administration, the answer is probably: not far. As we have seen, the
military commanders and civil servants on whom the smooth running
of Byzantine Africa relied whether transmarine or local in their origins formed a fairly restricted stratum of society which was by its very
nature focused on Constantinople and Carthage. The imperial presence
was certainly real and felt elsewhere than the provincial metropolis; but
beyond Carthage there is very little hard evidence of significant non-local
settlement in Africa in the Byzantine period.
As in Byzantine Italy and Egypt,227 minor officials, lower military
officers, and soldiers were for the most part probably recruited locally
throughout most of the sixth and seventh centuries. In Africa, however,
the evidence is primarily epigraphic, and here again it is only possible to
draw tentative conclusions as to the regional origins of the individuals so
commemorated on the basis of their names. In the inscriptions from the
hinterland of Byzantine Africa, Greek or otherwise eastern names do not
seem to have been particularly prevalent. As we have seen, the general
Maurice and the tribune Ziper, both of whom served at Rusguniae in
Mauretania Caesariensis, may well have been non-Africans. The tribune
Trajan, who died at the age of forty and was buried in the Basilica of
Sts Silvanus and Fortunatus at Sbetla, was apparently not a native of
that city (peregrinus) and, indeed, his name is unusual for a late antique
North African.228 The magister militum Crescens, too, was not a native of
Sbetla, though he was buried in the same church.229 One of the soldiers
in the numerus stationed at Hippo Regius was called Buraido, perhaps an
indigenous African name, but more likely the Thracian Borades.230
Perhaps the most fully documented epigraphic corpus from Byzantine Africa, however, comes from Hadra in Byzacena, near the modern
border between Tunisia and Algeria. The vast majority of the names
recorded in the churches of Byzantine Hadra speak of a singularly
227

228

229
230

Italy: Brown, Gentlemen and Officers, pp. 689. Egypt: J. Keenan, Evidence for the Byzantine
Army in the Syene Papyri, Bulletin of the American Society of Papyrologists 27 (1990), p. 146;
Bagnall, Egypt in Late Antiquity, p. 176; and Palme, Imperial Presence, pp. 2601.

francaise de Rome
N. Duval, Inscriptions byzantines de Sbetla (Tunisie) III, Melanges de lEcole
(Antiquite) 83 (1971), pp. 4313 = AE (1971), 174, no. 495 = Pringle, Defence, pp. 3356, no.
52 = Handley, Dying on Foreign Shores, no. 390. The name seems otherwise to be unattested in
late antique North Africa.
Duval, Inscriptions byzantines de Sbetla (Tunisie) III, pp. 42831 = AE (1971), 1734, no.
494 = Pringle, Defence, p. 335, no. 51 = Handley, Dying on Foreign Shores, no. 389.

D. Detschew, Die thrakischen Sprachreste, Osterreichische

Akademie der Wissenschaften,
Philosophisch-historische Klasse, Schriften der Balkankommission, Linguistische Abteilung 14
(Vienna, 1957), p. 80; Besevliev, Personennamen, p. 79; and PLRE 3:2456, s.n. Boraides (a
cousin of Justinian); but see also K. Jongeling, North-African Names from Latin Sources (Leiden,
1994), p. 28.

247

Staying Roman
Christian, Latin, African identity. The citys name-stock included only
a handful of Germanic names Hildiger, Guitifrida and, indeed, surprisingly few indigenous ones. After Latin names, Greek ones were the
most popular here in the sixth and seventh centuries: Antiochia, Diotimus, Dynamius, Gennadius, Poemenius, Polibius, and Theodore, possibly Evodia, Theophilus, and Zosimus as well.231 All together, they
represent perhaps 8 per cent of the citys overall name-stock (ten of 124
names), with the Greek Dynamius tied with the Latin Innocent as the
most popular name attested locally in the epigraphic record. However, a
handful of these very same Greek names including Dynamius were
already known in Roman and Vandal Africa, and in the absence of further data these sixth- and seventh-century attestations cannot be taken as
indicative of an eastern presence.232
Even if easterners did settle in Hadra, they seem not to have greatly
affected the existing structures of power. Certainly, the citys most distinguished citizens and the wielders of secular authority continued to bear
Latin names. Only two Byzantine-era inscriptions record individuals of
the rank of illustris, and there is no reason to suppose that either of these
was not himself an African.233 The Byzantine-era urbis defensor or defensor
civitatis Mustelus, also apparently known as Bellator, bore a local name
231

232

233

francaise de Rome
N. Duval, Recherches archeologiques a` Hadra, 2 vols., Collection de lEcole
18/12 (Rome, 197581), 1:41316. The subsequent publication of numerous supplements to
this catalogue in Duval, Recherches archeologiques a` Hadra II, pp. 21524; F. Baratte, F. Bejaoui,

and Z. Ben Abdallah (eds.), Recherches archeologiques a` Hadra. Miscellanea 2, Collection de lEcole
francaise de Rome 17/2 (Rome, 1999), pp. 1437, 21725, and 233; and F. Baratte, F. Bejaoui,

and Z. Ben Abdallah (eds.), Recherches archeologiques a` Hadra III, Collection de lEcole
francaise
de Rome 18/3 (Rome, 2009), pp. 13155 has not substantially changed the situation.
PCBE 1:333, s.n. Dynamius. The name Dynamius in Byzantine Hadra: Duval, Hadra 1:489,
578, 967, 10910, and 2778, nos. 23, 31, 66, 77, and 414; ibid., pp. 2879, no. 424, which also
mentions the name Dynamius, appears to have been Vandal-era. The name Innocent: Duval,
Hadra 1:1034, 11213, 25960, 284, and 2912, nos. 72, 79, 403, 420, and 426. The names
Donatus and Felix were probably equally popular, although they are less securely attested in the
extant epigraphic record; no other name occurs five independent times: see the list of names
in Duval, Hadra 1:4013. The name Theodore (Duval, Hadra 1:501 and 645, nos. 24 and
38) was popular throughout the late Roman world; for Evodia (Duval, Hadra 1:12930, no.
97: Ivuza) and Evodius, see Notitia, Byz. 6, p. 124 and PCBE 1:36674, s.nn. Evodius 12; for
Diotimus (Duval, Hadra 1:367, no. 13), see the earlier proconsul Diotimus, CTh 16.5.39 (ad
407) and 16.11.2 (ad 405), pp. 867 and 905; for Gennadius (Duval, Hadra 1:122, no. 88), see
PCBE 1:534, s.nn. Gennadius 12. Less well attested in Africa are Theophilus: Duval, Hadra
1:21415, no. 203 (tentative: Teauf?il?[i]); see also PLRE 3:13079, s.nn. Theophilus 15.
Zosimus: Duval, Hadra 1:129, no. 96 (tentative: Zosiu); see also PLRE 3:1421, s.n. Zosimus.
Antiochia: Duval, Hadra 1:2323, no. 304; see also PLRE 3:901, s.nn. Antiochus 16, most
of whom are attested only in the sigillographic record. See also Duval, Hadra 1:778 and 125,
nos. 501 (Poemenius) and 91 (Polibius), neither of whose names appears in PLRE 3.
Duval, Hadra 1:1935 and 2568, nos. 200A (Marcellus) and 402 (Silvanianus). Two earlier
inscriptions, presumably Vandal-era, also record one Albucius inlustris (ibid., pp. 534, no. 26)
and Astius Vindicianus vir clarissimus et flamen perpetuus (ibid., pp. 2545, no. 401).

248

New Rome, New Romans

and may also have held the title of magister or perhaps a magistracy.234
Similarly, the only military officer known from Byzantine Hadra is an
optio with the Latin name Maurianus.235
A similar pattern emerges in what is now eastern Algeria, where
the citizens of An el-Ksar constructed a castrum in their town at some
point in the reign of Tiberius II Constantine (ad 57882). Again, for
the most part their names were thoroughly Latin and African: Donatus
and Donatius, Victor and Victorianus, Januarius, Saturninus, Cresconius,
Felix. One had a Germanic name, Guntharith, and another bore the
ambiguous name Gudulus, which could be either Germanic or indigenous. Phocas magister is the only one of their number to be distinguished
by a Greek or otherwise eastern-sounding name. Even so, nothing within
the inscription seems to indicate that he was himself anything other than
local in his origins. The individuals commemorated in the inscription
seem for the most part to have been Africans recruited into the Byzantine army as limitanei, perhaps serving under a tribune named Flavius
T . . . and trained by one Dominicus, who would appear to have been
the garrisons drill-instructor (campiductor). Phocas also seems to have
been a member of the local community, though exactly what kind of
magister he was is not entirely clear. Based on the position of his name
within the inscription, Denys Pringle has suggested that Phocas duties
were not military. Pringle suggests instead that Phocas was a magister fundi,
an official connected with the organization of the (presumably imperial)
estate at An el-Ksar, who would have supervised the construction of the
castrum and set up the inscription recording the act.236
A handful of other communities demonstrate the same trends. As at An
el-Ksar, at Ksar Lemsa (near Furnos Maius) in Africa Proconsularis local
notables appear to have worked in cooperation with the imperial administration to ensure the regions defence. Three brothers named Maximianus, Stephen, and Mellosus presumably local landowners built
234

235
236

Duval, Hadra 1:1558, nos. 1212; see also ibid., pp. 1401, no. 109, perhaps the tomb of his
wife. On the title, see the discussion ibid., pp. 4489. For the local currency of the name, see
Astius Mustelus, flamen perpetuus christianus: ibid., pp. 2737, no. 413.
Duval, Hadra 1:64, no. 37, with the discussion ibid., p. 449.
Durliat, Dedicaces, no. 29, pp. 717 = Pringle, Defence, p. 328, no. 31, and see the discussions
at Durliat, Dedicaces, pp. 757 (who suggests that Phocas was the matre-duvre) and Pringle,
Defence, pp. 747. On the name Guntharith, see M. Schonfeld, Worterbuch der altgermanischen

Personen- und Volkernamen nach der Uberlieferung

des klassischen Altertums bearbeitet (Heidelberg,
1911; repr. 1965), p. 119. On Gudulus, see ibid., p. 115 s.n. Gudullus but also Jongeling,
North-African Names, pp. xxivxxv. On the function and status of the campiductor, see P. Rance,
Campidoctores Vicarii vel Tribuni: The Senior Regimental Officers of the Late Roman Army and
the Rise of the Campidoctor, in A. Lewin and P. Pellegrini (eds.), The Late Roman Army in the
Near East from Diocletian to the Arab Conquest, BAR International Series 1717 (Oxford, 2007),
pp. 395409.

249

Staying Roman
a tower there at some point in the late sixth or early seventh century, in
the reign of Maurice (ad 582602) and under the auspices of the magister militum or exarch Gennadius and the Praetorian Prefect John.237 A
tribune with the quintessentially Romano-African name of Victor seems
somehow to have been associated with the fortification of Mascula (mod.
Khenchela) in the Aur`es Mountains.238 A certain Masticiana, perhaps a
local possessor fundi, financed the building of fortifications at Henchir
Bou Sboa, also in the Aur`es.239 Unfortunately, it is not clear whether the
three known primicerii of Byzantine Africa served in the civil, military,
or ecclesiastical administration of the province, but they were possessed
of a mixed bag of Latin, Greek, and biblical names. Thus, for example,
the primicerius Cosmas, who died at age sixty and was commemorated in
a sixth- or seventh-century epitaph in the basilica of Servus in Sbetla,
bore a Greek name.240 The primicerii John and Donatus, on the other
hand, both of whom were buried in Tebessa in the Byzantine period,
bore a biblical and a Latin name, respectively.241 Again, however, nothing
within their epitaphs would seem to indicate that any of them were of
non-African origins.
Indeed, beyond Carthage and the provincial capitals, there is little clear
indication of a substantial non-African elite presence in Byzantine North
Africa. The effects of the Byzantine occupation on the structure of
African society are therefore difficult to gauge. Certainly some integration between easterners and Africans took place. A number of the military
officers and civil servants appointed by the emperors in Constantinople
served in Africa for three to six years, or even longer. Byzantine soldiers
took Vandal wives and began to establish themselves as landowners in
Africa. At the same time, Africans were recruited into the Byzantine
army both as limitanei and as regular soldiers usually, no doubt, to
serve in Africa, but at least on occasion to serve in other parts of the
Mediterranean. Between the first uneasy decades of Byzantine rule in
sixth-century North Africa and the final confrontation with the invading
seventh-century Muslim armies, there is little evidence for the large-scale
237

238
239
240
241

CIL 8.12035 = ILCV 793 = Durliat, Dedicaces, no. 30, pp. 778 = Pringle, Defence, p. 330, no.
36 (Ksar Lemsa). On local management of public works, including fortifications, see CJ 1.4.26
(ad 530), pp. 424. On defensive works in Africa in general, see Durliat, Dedicaces, pp. 93114,
esp. 11213.
Durliat, Dedicaces, no. 28, pp. 6771 = Pringle, Defence, p. 329, no. 33, with ibid., p. 218.
Durliat, Dedicaces, no. 32, pp. 803 = Pringle, Defence, p. 331, no. 39.
Cosmas: Duval, Inscriptions byzantines de Sbetla (Tunisie) III, pp. 43941 = AE (1971), 175,
no. 499 = Pringle, Defence, p. 335, no. 50.
John: Pringle, Defence, p. 338, no. 58; Donatus: CIL 8.10637 = ILCV 488A = ILAlg. 1.3433 =
Pringle, Defence, p. 339, no. 59.

250

New Rome, New Romans

arrival of fresh troops from the East. One can only imagine that by the
late sixth and seventh centuries the African provinces would have relied
primarily on local recruitment to provide for their own defence. The
construction of local fortifications also seems to have been carried out
by individuals with Latin names who were themselves probably local
African notables, frequently working in conjunction with the imperial
administration.242
The representatives of that administration seem to have been men of
pan-imperial origins. Lesser officials were doubtless recruited locally; the
names and languages of higher-ranking officers and civil servants by and
large point to participation in a broadly Byzantine culture shared by the
empires ruling elite regardless of their region of origin. However, the
fact that imperial appointments to high rank favoured personal associates
of the emperor and the fact that we rarely hear of Africans who made
it in Constantinople and the eastern provinces further confirm our
deduction that the highest-ranking civil servants and military officers in
Africa probably came most often from the East, and this throughout the
Byzantine period. Indeed, in the instances where we know the regional
origins of military governors, prefects, and subordinate officers with any
degree of certainty, they were often men from the imperial borderlands:
Illyrians, Thracians, Mesopotamians, Armenians, Syrians. Only rarely
were they Africans. Augustines patriotae men like the Praetorian Prefect Innocent seem to have been the exception rather than the rule
among the highest strata of the civil and military administration of Byzantine Africa. Nevertheless, their presence is significant: the governing of
this western prefecture was an empire-wide endeavour. Perhaps more
importantly, however, the officials who ran Africa for the empire were
also men with access to the highest circles of power in Constantinople:
members of the emperors family, his former bodyguards, commanders
of palace troops, high-ranking senators, and others who had gained the
autarchs special trust. The military officers were also men who for the
most part had been tried and hardened in the Persian wars, on a frontier
deemed critical by the imperial administration. Taken together, these
last two facts point towards the extreme importance a series of sixthand seventh-century Constantinopolitan emperors attached to regaining
and maintaining control over the African provinces; a control that would
nevertheless be contested repeatedly by the Moors.
242

On the organization of the Byzantine military administration in Africa generally, see Pringle,
Defence, pp. 55120.

251

Chapter 5

T H E M O O R I S H A L T E R N A T IVE

The Vandals were not the only people to threaten imperial control of
North Africa in the fifth and sixth centuries. After the assassination of the
western Roman emperor Valentinian III, in 455, the imperial territories
of the Mauretanian and Numidian interior came to be dominated by
independent African kingdoms that were thought of as Moorish by the
inhabitants of Carthage and by Roman populations throughout the rest
of the Mediterranean. In the maintenance and extension of their own
power in North Africa, both the Vandals and their Byzantine successors
were forced to contend with these Moorish kingdoms and, as presented
in our literary sources, the struggle was to have a devastating effect on
the populations of the African borderlands.
This was so much the case that, to the African poet Flavius Cresconius Corippus (writing after the fact), the Moors provided the ultimate
validation of the Byzantine reconquest of 5334. In the poets great epic,
the Iohannis, the fundamental benefit that the restoration of an imperial
presence in Africa brought was salvation from the harassment, not of
Vandals but of Moors. The Vandals had been unable to restrain Moorish violence and bring peace to Africa; the armies of Byzantium did
precisely that. One gets the sense from Corippus that there were in fact
two Africas: one, a world of serene tranquillity, cultivated like the gardens
that Procopius describes with such wonder and whose fruits so plentifully
adorn the mosaics of this time and region; the other, a dark and terrifying place, existing on the fringes of the world which Christianity and
Roman civilization had surrounded and enclosed, and which Byzantine
arms protected against all comers. Indeed, to judge from the Iohannis,
Corippus hated and feared the Moors. In the poets mind, these nonRoman Africans seemed to be wild, untamed savages (feri) living beyond
the pale of civilization.1 But then, Corippus was an imperial apologist,
1

Coripp. Ioh. 1.1, 1.54, 1.254, 2.61, 2.109, 2.237, 4.51, 4.104, etc., pp. 3, 5, 13, 28, 30, 36, 68, 70
(feri). The word has slightly Virgilian overtones: Augustus poet used the word of Aeneas himself
at Aeneidos 4.466, ed. R. A. B. Mynors, in P. Vergili Maronis Opera (Oxford, 1969), p. 190 and

252

The Moorish alternative

and his evidence must be treated with great caution. Indeed, Moorish
barbarism was a key element of sixth-century Byzantine propaganda in
Africa, and so the cultured Latin poet had good reason to depict Moors
in as dark a light as possible.
Yet the anxiety of Byzantine imperial ideologues and their need to
de-legitimate the Moors nevertheless force us to ask whether the African
interior had become somehow un-Roman in late antiquity. We must
approach the question from two directions. First, we need to understand
what our two richest and most detailed textual sources Corippus himself, and the Byzantine historian Procopius tell us about Moors and
Moorish lack of Romanness, how this information functioned within the
context of Corippus and Procopius thoughts on non-Roman populations in general, and these accounts ties to Byzantine imperial ideology.
Second, we must evaluate how far the same process of cultural and political accommodation between Romans and non-Romans which had been
developing elsewhere in Africa and the western Mediterranean had progressed in fifth- and sixth-century Mauretania and Numidia, and the
pre-desert regions of Byzacena and even Tripolitania. Understanding the
answers to these questions will in turn help us to assess the rhetorical
strategies and cultural assumptions that informed our sources discussion
of the de-Romanization of Moorish Africa.
1. moors and the rhetoric of barbarism
Our understanding of Moorishimperial relations in the Byzantine
period is governed by two central aspects of imperial ideology. First, in
the sixth century, the court at Constantinople began to espouse a newly
militaristic mentality of which, for us, Justinian is the supreme symbol.2
The Illyrian emperor and his successors appear to have had neither the
inclination nor the patience for the slow, steady process of combining

of Jupiter ibid., 2.326, p. 137 as well as ibid., 6.49, p. 228 in the sense of cruel or savage; and
ibid., 5.818 and 10.12, pp. 225 and 333 in the sense of wild or untamed the latter with
regard to Carthage. See also Aeneidos 2.51 and 7.489, pp. 128 and 271, where Virgil uses the
word substantivally (of a horse and a stag, respectively). Claudian, Corippus other great model,
also used the word: Carmen 5.458, ed. J. B. Hall, in Claudiani Carmina (Leipzig, 1985), p. 48.
However, Corippus usage is closer to that of Julius Caesar, Bellum Gallicum 1.31.5, 1.33.4, 1.47.3,
2.4.8, 2.15.5, and 4.10.4, ed. W. Hering, vol. 1 of Commentarii rerum gestarum (Leipzig, 1987),
pp. 13, 15, 22, 27, 32, and 55, and Sallust, De bello Iugurthino 80.1, p. 118 (genus hominum ferum
incultumque). On feroces in a late Roman military context in general, see G. Halsall, Gender and
the End of Empire, Journal of Medieval and Early Modern Studies 34 (2004), pp. 223; my thanks
to Guy Halsall for an enlightening conversation on this topic. For Corippus attitude towards the
Vandals, and for the Vandals inability to restrain the Moors, see below, sect. 1.
W. Goffart, Rome, Constantinople, and the Barbarians, Romes Fall and After (London, 1989),
pp. 132, esp. 225.

253

Staying Roman
forceful diplomacy with the politics of conciliation that seems to have
characterized the Roman empires original expansion into the Maghribi
hinterland.3 In the wake of his armys easy victories over the Vandals,
Justinian envisioned the reconquest of all the territory in Africa that
had once belonged to the Roman empire. A rescript addressed by the
emperor to his general Belisarius on 13 April 534 speaks in imperative
terms about the need for Roman arms to recapture those parts of the
African provinces that had been lost to Vandal and Moorish invasions.
Those enemies who remained in the region were to be driven out, and
the fortresses that they had occupied were to be invested immediately
with Roman garrisons.4 These soldiers and their officers were commanded to be bold and fierce (audaces et feroces) to the empires enemies,
and to keep in constant training so as to be able to resist all comers.5 The
Roman army was to be the key to the future peace and security of Africa,
lest licence be given to enemies to invade and lay waste the places that
our subjects possess.6 Realities on the ground may well have been quite
different; but as far as the emperor was concerned, accommodation was
out, military assertiveness was in.
Despite the fact that the Byzantine reconquest of Africa had initially
been launched against the Vandal kingdom, it seems unlikely that the
hostile forces whom Justinian envisaged threatening the tranquillity of
the empires newly reconquered western provinces were ever predominantly Vandals. Belisarius kept the emperor well informed about developments in Africa, and by April 534 the court in Constantinople would
have learned of the imperial armys decisive victories over the Vandals in
the battles of Ad Decimum and Tricamarum, of the triumphal entry of
the Byzantine expeditionary force into Carthage itself, and of Belisarius (largely successful) efforts to recover the far-flung coastal cities and
Mediterranean islands that had once belonged to the Vandal kingdom.7
News might not yet have reached Justinian of the capture of Gelimer, the
last Vandal king, in March 534; but the writing was already on the wall.8
Moreover, Vandal resistance to the empire had more or less completely
collapsed over the course of the previous winter. At least in Procopius
account we hear of Vandal suppliants fleeing to local churches for refuge
3

4
5
6
7

Mattingly, War and Peace, p. 53; on the first Byzantine-Moorish contact, see Y. Moderan, Les
Maures et lAfrique romaine (IVe VIIe si`ecle), Biblioth`eque des e coles francaises dAth`enes et de
Rome 314 (Rome, 2003), pp. 58593.
CJ 1.27.2.4 (ad 534), p. 79.
CJ 1.27.2.9 (ad 534), p. 80 (military training) and ibid., 1.27.2.11 (ad 534), p. 80 (audaces et
feroces).
CJ 1.27.2.4b (ad 534), p. 79: ne detur hostibus licentia incurrendi aut devastandi loca, quae nostri
subiecti possident.
8 Proc. BV 2.7.17, 1:451.
Proc. BV 1.24.19 and 2.5.25, 1:412 and 1:443.

254

The Moorish alternative

in Carthage, its rural hinterland, and in Hippo Regius.9 The Byzantine military authorities were initially willing to let rank-and-file soldiers
keep the Vandal women and children that they had enslaved in the
wake of their military victories and, as we have seen, even to force their
female captives into coerced marriages.10 Vandal fighting men, on the
other hand, were disarmed, sent under guard to Carthage, and eventually
deported to Constantinople before being sent to fight for the emperor
along the Persian frontier.11 Though the ships bearing these exiles to the
East did not set sail until the summer of 534, plans for cleansing Africa of
the Vandal threat were clearly already being laid over the previous winter.
Thus despite the fact that most of the Moorish rulers of Mauretania,
Numidia, and Byzacena had accepted imperial authority as early as the
autumn of 53312 Justinians rescript would seem to have set the stage
for a new age of conflict not primarily with Vandals, but rather with
Moors.
The second aspect of imperial ideology that shapes our understanding of Moorishimperial relations in the Byzantine period is perhaps a
corollary of the emperors desire to resurrect and defend old frontiers,
real or imagined; for imperial apologists seem to have felt that the perceived cultural dividing line between Romans and barbarians, too, was
under assault in Africa, and also needed to be resurrected. Here again
Moors rather than Vandals were seen as the greater threat. In the Vandal period, African authors seldom discussed the barbarians who lived
across the provinces old Roman frontier.13 Under the Byzantine regime,
by contrast, Moors and Moorish culture were a central concern of both
the Romano-African poet Corippus and the eastern historian Procopius.
Both were hostile witnesses. The two were exact contemporaries, writing
towards the middle of the sixth century, fifteen years or so after the fall of
the Vandal kingdom and the eruption of hostilities between Byzantines
(now masters of Carthage) and Moors. Publically, at least, both of our
authors were strongly supportive of the Byzantine venture in Africa.14
9
10
11

12
13

14

Proc. BV 1.20.1, 2.4.1012, and 2.4.32, 1:396, 1:434, and 1:437.

Proc. BV 2.3.24, 1:431 (enslaving of women and children); ibid., 2.4.3, 1:4323 (beauty of these
slaves); ibid., 2.14.8, 1:483 (enslaved women married to their captors).
Proc. BV 2.4.1012, 1:434 (disarmed and sent to Carthage); ibid., 2.5.1, 1:439 (readied for the trip
to Constantinople); ibid., 2.9.1, 1:455 (arrive in Constantinople); and ibid., 2.14.1718, 1:4845
(organized into cavalry units). The empire had long demanded military service from defeated
enemies: A. D. Lee, War in Late Antiquity: A Social History (Oxford, 2007), pp. 814.
Proc. BV 1.25.38, 1:41213.
See, however, Vict. Vit. 1.25, 1.358, 2.4, 2.28, 2.323, 2.36, and 3.68, pp. 12, 1617, 25, 34,
367, 38, and 105; Passio septem monachorum 5, p. 109; and Fulg. Ep. 11.2, 1:360 (Ferrandus to
Fulgentius). See also V. Fulg. 5 and 28, pp. 33 and 141 (probably early Byzantine in composition).
See, however, Proc. Anecd. 6.25 and 18.59, 3:42 and 11213; M. Cesa, La politica di Giustiniano
verso loccidente nel giudizio di Procopio, Athenaeum, Studi Periodici di Letteratura e Storia

255

Staying Roman
Procopius, whose History of the Wars provides us with our single most
detailed record of the period from 533 to c.547, was the secretary and
legal advisor (1 or assessor) of Belisarius.15 Corippus epic work
in honour of John Troglita, Belisarius eventual successor as the leading
military officer in Africa, complements and extends Procopius account,
providing an invaluable African perspective on the events of the 540s;
but the perspective is still that of one who welcomed and rejoiced in the
Byzantine occupation of Africa.16 Given the hostilities between Byzantines and Moors, it is hardly surprising that these authors depict the latter
as barbaric savages on whom the refinements of Roman civilization were
lost. However, both authors also draw freely on ancient ethnographic
stereotypes when describing Moorish society. Before considering what
Procopius and Corippus have to say about Moors specifically, therefore,
we would do well to consider their differing visions of barbarian peoples
in general.
By far the fuller account of such peoples to emerge from either of
these sources is to be found in Procopius. Like most members of the
late Roman literary elite, Procopius was convinced of his own cultural
superiority over the peoples whose lands surrounded the empire.17 In his
thought, a sharp line divided the Roman world and the poverty, violence,
and faithlessness of the barbarians. Barbarians languished in a world of
scarcity. From the territory around the Upper Nile to the date-palm
groves of the Arabian peninsula to the imposing heights of the Caucasus
and even to the forests and mountains of Scandinavia, barbarian lands
were typically worthless. At best they might produce marginal tribute; at
worst as in the case of the Lazi, who lived in the shadows of the Caucasus such lands produced neither salt nor wheat nor wine nor any good
thing.18 The Lazi lived off millet () ), which offended Procopius
Mediterranean tastes, and otherwise supplemented their diet by purchasing foodstuffs from the Byzantine empire in exchange for hides and
slaves.19 The Tzani, another Caucasian people, took a more direct route:

15
16

17
18

19

dellAntichit`a n.s. 59 (1981), pp. 389409; and A. Kaldellis, Procopius of Caesarea: Tyranny, History,
and Philosophy at the End of Antiquity (Philadelphia, Pa., 2004), pp. 1724.
Proc. BV 1.14.3, 1:373.
Averil Cameron, Corippus Iohannis: Epic of Byzantine Africa, in F. Cairns, F. Williams, and
S. Cairns (eds.), Papers of the Liverpool Latin Seminar, Fourth Volume, 1983 (Liverpool, 1984),
pp. 16780.
See also, however, Kaldellis, Procopius of Caesarea, p. 92.
Proc. BP 1.19.13, 1:102 (Arabia); ibid., 1.19.29, 1045 (Upper Nile); Proc. BG 2.15.1619,
2:217 (Scandinavia: Scrithiphini); Proc. BP 1.15.21, 1:77 and much more expansively Proc. Aed.
3.6.25, 4:956 (Caucasus: Tzani), and Proc. BP 2.28.27, 1:286: !/ A 0! <) A 
B%   
,  & V " V W V  D)) 8+ ;0 (Lazi;
see also Proc. BP 1.12.15, 1:58).
Proc. BP 1.12.17 and 2.28.28, 1:58 and 1:286.

256

The Moorish alternative

they lived on what they could steal from the neighbouring Byzantines
and Armenians.20 Such raiding and plundering were characteristic of
barbarian lands in Procopius mind. Those same lands that suffered from
such a paucity of resources the Upper Nile, Arabia, the Caucasus
were laid waste by incessant raiding and plundering.21 The barbarian
world could seem to be trapped in an endless cycle of violence: the
Heruls, for example, had attacked and plundered their neighbours until
their power was broken by the Lombards; then the Gepids began to raid
the Heruls, carrying off their cattle and other property and raping Herul
women.22
Barbarian raiding seems to have been linked in Procopius mind to
another barbarian trait: faithlessness. They might swear oaths or sign
treaties, they might be paid off with gold or other tribute, but such measures could never stop them from attacking and plundering the empire.23
Barbarians in general and Franks above all simply could not be relied
upon to keep their word.24 Even Heruls were faithless and unstable in
Procopius estimation, despite the fact that they provided countless troops
for Justinians armies, including the force that recaptured Africa for the
empire.25 All of this prompted Procopius to conclude that attempting to
reach an understanding with barbarians was useless, for there is no way
for any of the barbarians to keep faith with the Romans, except though
fear of avenging soldiers.26 Military might alone could keep barbarians
in line.27

20
21
22
23

24
25

26
27

Proc. BP 1.15.213, 1:778.

Proc. BP 1.19.11, 1:102 (Arabia); ibid., 1.19.2933, 1:1046 (Upper Egypt); ibid., 2.28.278,
1:286 (slaving in the Caucasus).
Proc. BG 2.14.827, 2:20912.
Proc. BP 1.19.2933, 1:1046 (Nobatae and Blemyes); ibid., 1.15.205, 1:778; and see also Proc.
Aed. 3.6.28, 4:957 (Tzani). Y. A. Dauge, Le Barbare: recherches sur la conception romaine de la
barbarie et de la civilisation, Collection Latomus 176 (Brussels, 1981), pp. 4334 sees claims of
barbarian faithlessness functioning as an element of their perceived vanitas.
Proc. BG 2.25.2, 2:261:    + #   
 ;): 8+:!
U1!; see also Proc. BG 2.22.10, 2:248.
Proc. BG 2.14.35 and 2.14.41, 2:21314; see also Proc. BV 2.4.30, 1:436, where Procopius claims,
For a Herul man not to give himself up to faithlessness and drunkenness, but to lay claim to
excellence, is hard to do and worthy of great approval (D A X @ ) &  8
 
/ + 8"+, 8)) 8 "+, )  / 
 ))# D). For
the Herul presence in the sixth-century Byzantine army, see, e.g., Proc. BP 1.13.19, 1:62 (Battle
of Daras); ibid., 2.21.4, 1:2434; Proc. BV 1.11.11, 1:362 (Heruls in the African expedition of ad
5334); Proc. BG 2.13.18 and 3.27.3, 2:208 and 2:417; see also Proc. BG 3.33.13, 2:444.
Proc. BP 1.19.33, 1:1056: Y ! D **1  < 
 & :+
&  -.! 
  
 Y &  E 8  ! !E.
In this respect, Procopius seems to have shared in the renewedly militaristic outlook of sixthcentury Constantinople, on which see Goffart, Rome, esp. pp. 226. See also G. Ladner, On
Roman Attitudes toward Barbarians in Late Antiquity, Viator 7 (1976), pp. 126.

257

Staying Roman
It is important to stress from the outset, however, that even in Procopius mind, not all barbarian peoples were such uncouth, uncivilized
thugs. The nature of Procopius historical endeavour divided the peoples
living beyond the frontiers of the Byzantine empire into two categories
of unequal importance to that authors narrative. Persians and Ostrogoths,
and to a lesser extent Vandals, were more important to Procopius than
Arabs, Lazi, Franks, and the other barbarian peoples who pepper the
historians account like so many ethnographic curiosities. One gets the
sense that Procopius made more of an effort to understand these major
barbarians and was therefore less willing to explain their actions through
reference to simple barbarism and lack of cultivation. As we have seen,
he found Vandals to be in some ways comfortingly familiar, even perhaps
excessively Roman in their tastes and indulgences.28 Procopius also had
a considerable amount of respect and sympathy for Persians, even if he
did find their officials insufferably arrogant, their way of life excessively
rigid, and their laws incomprehensible.29 As Anthony Kaldellis has rightly
emphasized, Procopius attributed to the Ostrogothic king Theodoric the
startling virtue of protecting Italy from barbarians; and Averil Cameron
has even argued that towards the end of his History of the Wars the sympathies of the disillusioned historian are more with Theodorics eventual
successor Totila than with the Byzantines themselves.30
The other, minor barbarian peoples received less attention from Procopius. He did not understand them as well as he did Persians and
Ostrogoths, he generalized more freely about their national character,
and was typically more willing to discuss them in terms of ethnographic
tropes. While Persians, Ostrogoths, and Vandals shared a predominantly
urban civilization, the minor barbarians tended to be (though were by
no means universally) nomads and montagnards.31 Procopius response to
28
29
30

31

Decadence, of course, is also a trope in the description of barbarians: Dauge, Barbare, pp. 4334.
See also above, Chapter 1.3.
Proc. BP 1.11.33 and 2.28.256, 1:545 and 1:286.
Proc. BG 1.1.27, 2:8 (  **1! E 
! & : 8;)E ;0)) and
Kaldellis, Procopius of Caesarea, pp. 15960. Cameron, Procopius, p. 201, citing particularly Proc.
BG 4.31.1819, 2:6534.
On the anxiety that nomadism caused in Greco-Romans, see B. Shaw, Eaters of Flesh, Drinkers
of Milk: The Ancient Mediterranean Ideology of the Pastoral Nomad, Ancient Society 13/14
(1982/3), pp. 531 and for North Africa specifically, B. Shaw, Fear and Loathing: The Nomad
Menace and Roman Africa, in C. M. Wells (ed.), Roman Africa/LAfrique romaine: The 1980
Governor-General Vanier Lectures, Revue de lUniversite dOttawa 52 (Ottawa, 1982), pp. 2546.
On the role of nomadism in the Roman conception of barbarians generally, see Dauge, Barbare,
esp. pp. 6206. See also B. Shaw, Autonomy and Tribute: Mountain and Plain in Mauretania
Tingitana, in P. Baudel (ed.), Desert et montagne: hommage a` Jean Dresch, Revue de lOccident
Musulman et de la Mediterranee 412 (Aix-en-Provence, 1986), pp. 6689 on Romans and
montagnards in North Africa.

258

The Moorish alternative

these minor barbarians ranges from disgust to admiration, often (though
again not universally) in proportion to the similarity of their lifestyle to
that of the cultivated Byzantine historian himself. Unlike other Huns, for
example, Procopius tells us that Ephthalites lived a settled, non-nomadic
lifestyle, that they did not engage in the kind of border-raids into Roman
territory that in the historians mind seem to have characterized most barbarians, that they were ruled by one king under a lawful constitution, and
that no less than Romans and Persians, their interactions both among
themselves and with their neighbours were characterized by justice and
righteousness.32 Thus, Ephthalite Huns receive very favourable treatment
in Procopius History. But so too do Slavs, who were said to lead a hardy
life not given to material comfort, and perhaps as a result of this preserved a simplicity of character that also won Procopius respect. The old
Roman virtue of simplicity seems to have gone a long way with Procopius. Though he felt that they were intemperate drinkers, unwashed,
and ugly to boot, even nomadic Huns won a good word from the historian on account of their simple character.33 On the opposite extreme
were peoples such as the Heruls, whose men the historian disdainfully
remarked were said to bugger donkeys.34
The poet Corippus, by contrast, leaves us few clues as to his overall
attitudes towards barbarians. As an educated African who had been raised
and pursued a career as a grammaticus somewhere in the rural provinces of
the Vandal kingdom, Corippus early perceptions of barbarians must have
been formed primarily by Moors, and of course by Vandals themselves.
But Vandals barely enter into Corippus epic account of mid sixthcentury Africa, and when they do so the poet highlights their weakness
and inability to check the rising power of the lawless Moors.35 Even so,
Vandals emerge from Corippus work just as they do from Procopius
as seeming quite familiarly Roman. At the very least they seem to have
fought much like Romans did: Vandal cavalry regiments were organized
under duces and had mounted standard bearers to direct their troops,
32
33

34

35

Proc. BP 1.3.17, 1:1011. The quotation is from Proc. BP 1.3.5, 1:11: -.! 
!  / QE
  Z.
Huns: also called Massagetes: Proc. BV 1.11.9, 1:362 and Proc. Anecd. 7.10, 3:445; ugly: Proc.
BP 1.3.4, 1:11; unwashed: Proc. BG 3.14.28, 2:358; intemperate drinkers: Proc. BV 1.12.8, 1:366;
simple character: Proc. BG 3.14.28, 2:358 (another ancient trope).
Proc. BG 2.14.36, 2:213. C. R. Whittaker summarizes the ethnocentric stereotypes Romans
had of their neighbours: Barbarians were natural slaves, animals, faithless, dishonest, treasonable,
arrogant, drunken sots (G. W. Bowersock, P. Brown, and O. Grabar (eds.), Late Antiquity:
A Guide to the Postclassical World (Cambridge, Mass., 1999), pp. 3345, s.v. Barbarian; the
quotation is from p. 335). Neither Procopius nor Corippus seems to have strayed too widely
from this perception of non-Romans.
Coripp. Ioh. 3.178264, pp. 547.

259

Staying Roman
while at the same time Vandal warriors were accompanied into the field
by attendants who brought water to their thirsty masters.36 The Vandal
pursuit of warfare comes across as being (from a Roman point of view)
all very civilized and totally inadequate to deal with the Moorish threat
to Vandal control of the kingdom.
In his later life, however, Corippus travelled to the imperial court at
Constantinople, where, in 565, he witnessed an Avar embassy to the
newly acclaimed emperor Justin II. Corippus description of the event
gives us another data point from which we can begin to deduce his attitude towards barbarians writ large; and in Corippus account Avars appear
as anything but civilized. Much like Procopius barbarians, Corippus
Avars inhabited a cold, barren, unproductive wasteland of frozen rivers
and wintry snows.37 The Avar people themselves had long, snake-like
hair and huge limbs, and a rough look about them; indeed, they reminded
the poet of nothing so much as Hyrcanian tigers.38 Like all barbarians,
Avars were fierce, and their ambassador to Justins court boasted of their
northern conquests.39 But in the Greco-Roman mind barbarian ferocity
was often an illusion, and in Corippus account the Avars too proved
to be pitiable cowards at heart. In fact, they had fallen upon and conquered the peoples of the north only because the Avars themselves had
been unable to defend their own homeland, from which they had been
driven as refugees. They were certainly no match for the highly trained
Byzantine army.40
In the sixth century, militarism and the rhetorical image of the barbarian were two sides of the same coin. Procopius and Corippus shared a
view of the peoples living across the frontiers of the empire that had been
broadly typical of Roman authors for centuries.41 This tradition saw nonRomans as fierce, warlike, and untamed inhabitants of the wastelands that
36

37

38
39

40
41

Cavalry: Coripp. Ioh. 3.2434 and 3.2535, pp. 567. Duces: ibid., 3.242, p. 56. Standard-bearers:
ibid., 3.2368, p. 56; see also ibid., 3.198, 3.222, and 3.2423, pp. 546. Attendants: ibid., 3.2301,
p. 56.
Flavius Cresconius Corippus, In laudem Iustini Augusti minoris 3.28199, ed. S. Ant`es, in Eloge de
lempereur Justin II (Paris, 1981), pp. 656. See also Dauge, Barbare, pp. 46771 and 6024 and
R. Pallas-Brown, East Roman Perceptions of the Avars in the Mid- and Late-Sixth Century,
in S. Mitchell and G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London, 2000),
pp. 30929, esp. 31415.
Corippus, In laudem Iustini, praef. 45, 3.24554 and 3.2603, pp. 13, 63, and 634. On the trope
of unkempt Avar (and barbarian) hair, see Pallas-Brown, East Roman Perceptions, pp. 31517.
Corippus, In laudem Iustini 3.27180, pp. 645. That barbarian peoples such as the Franks,
Goths, Lombards, and Gepids were fierce is the implication, at least, of Corippus, In laudem
Iustini, praef. 1015 and 1.2547, pp. 13 and 27. On barbarian feritas, ferocia, and belli furor in the
Roman thought world in general, see Dauge, Barbare, pp. 42831.
Corippus, In laudem Iustini 3.31722, p. 67. Vanitas is another essential element of the barbarian
in Roman thought: Dauge, Barbare, pp. 4334.
See in general Dauge, Barbare and, on specific points, the material cited above.

260

The Moorish alternative

lay beyond the Roman world, who were inexorably and eternally constrained by poverty, military necessity, or the lack of a moral compass to
attack the empire, to lay waste its frontiers, and to despoil its population.
Fifth-century emperors may have thought that they could do business
with the barbarians, but that had merely led to the loss of the West.
Reflecting on the history of the previous century and on the recent
fall of the Vandal kingdom to his own armies, Justinian observed that
our predecessors did not deserve this favour of God; to them not only
was it not allowed to free Africa, but they also saw Rome itself taken
by these same Vandals and all of the imperial regalia carried off from
there to Africa.42 Understanding the Roman empire and its place in the
cosmic order meant understanding that relations between Romans and
barbarians would always be violent. Barbarian faithlessness was such that
treaties could not restrain them and tribute could not mollify them. Yet
barbarians were also ultimately cowards. They could be contained, their
fierceness harnessed, and as federates they could be turned to the empires
advantage. To do so, however, neither the emperor nor his armies could
waiver in their assertiveness or falter in the strength of their arms.
2. the moors through roman eyes
Part of the sixth-century imperial agenda, then, was to draw clear lines
that could sharply define who was Roman and who was not. As we
have seen, in Procopius account it mattered to the Byzantine authorities
that Africans be included in the myths of kinship and shared descent
that bound Romans together as a single people: Africans were Romans
by origin (-.! "  8+).43 But it also mattered that Moors
be excluded from these same stories of collective ancestry. Procopius
calls the Moors a Phoenician people ([ +), which on
the face of it could potentially have served to emphasize commonalities
between them and at least some of the peoples under imperial control.44
Unlike Romans, though, in Procopius vision, the Moors were on the
wrong side both of history and of God. In the historians telling, Moors
were descendants of the peoples displaced from ancient Phoenicia by the
Hebrews after the biblical exodus; the Moors arrival in Africa was said to
have preceded that of the next wave of Phoenician emigrants Dido and
42

43
44

CJ 1.27.1.6 (ad 534), p. 77: Quod beneficium dei antecessores nostri non meruerunt, quibus non
solum Africam liberare non licuit, sed et ipsam Romam viderunt ab eisdem Vvandalis captam et
omnia imperialia ornamenta in Africam exinde translata.
Proc. BV 1.16.3, 1:382. On myths of shared descent in general, see G. Schopflin, Nations, Identity,
Power (New York, 2000), pp. 7998, esp. 97.
Proc. Aed. 6.3.9, 4:176.

261

Staying Roman
her companions and the foundation of Punic Carthage.45 However
unconvincing this account of Moorish origins may seem to modern
observers, to Procopius, the Moors were a people apart.46 In seeking
to demonstrate that a vast cultural gulf separated them from Romans,
Procopius and Corippus as well as a number of other secular poets and
ecclesiastical writers from both Africa and the rest of the late antique
Mediterranean brought together a wide array of cultural prejudices
and chauvinism, which they wielded like weapons. Moors and Romans,
these authors insisted, could be distinguished by the whole range of
traditional markers of ethnic identity: language, diet, dress and battlegear, military tactics, marriage customs, religion, and even the perceived
darkness of their skin.47
As a Phoenician people, Procopius tells us that amongst themselves
Moors spoke neither Greek nor Latin, but what he refers to as the
Phoenician tongue ( [
! ;!2).48 Given his understanding of
their collective history, there is every reason to believe that by this the
historian meant a language related to the Punic spoken by the ancient
Carthaginians, though the name-patterns of sixth-century Moorish leaders might lead us to suspect that a late antique ancestor of the modern
Berber dialects was (or was also) spoken in the North African interior
(see below, section 3). Yet Procopius may not have been completely off
his mark. Names can endure longer than spoken languages, and, as we
have seen, there is ample evidence for the survival of Punic in parts
of Africa well into late antiquity (see above, Chapter 3.4). Indeed, in
the mid fifth century the African expatriate Arnobius the Younger had
similarly observed that the indigenous North African Garamantes spoke
Punic, as distinct from the barbaric languages of the interior.49 I have
also argued that use of the Punic language or, indeed, a Punic identity
more generally, was probably not in itself considered un-Roman, at least
not in a late antique North African context. Of course, the same may
not have been true for Procopius Greek-speaking audience; but it is
worth pointing out that the linguistic similarities between Punic and
Syriac were well understood in late antiquity, and indeed an addition to
45
46
47

48

Proc. BV 2.10.1327, 1:4602.

On the origins and pre-history of the indigenous North African population, see M. Brett and
E. Fentress, The Berbers (Oxford, 1996), pp. 1041.
On markers of ethnic identity in general in late antiquity, see W. Pohl, Telling the Difference:
Signs of Ethnic Identity, in W. Pohl and H. Reimitz (eds.), Strategies of Distinction: The Construction
of Ethnic Communities, 300800, Transformation of the Roman World 2 (Leiden, 1998), pp. 1769;
the contrasts between Procopius and Corippus evidence are examined by Moderan, Maures,
pp. 3542.
49 Arnobius Junior, Commentarium in Psalmum 104, p. 159
Proc. BV 2.10.20, 1:461.

262

The Moorish alternative

one of the manuscripts of Zachariah of Mitylenes Syriac chronicle even
boldly notes that sixth-century Africans spoke Syriac and Latin.50 Yet
Procopius gives us no indication as to whether he thought Moors spoke
a language that could be understood by contemporary Levantines like
Severianus (a Byzantine cavalry commander in Africa who was a native
of Emesa in Syria) or Rhodon (the governor of Alexandria), each of
whom the historian calls Phoenician ([").51 Procopius was simply
not interested in drawing parallels between Moors and imperial officials
or, for that matter, Romans in general.
As with the other barbarian peoples whom he describes in his History,
in Procopius mind the barbarousness of the Moors seems to have been
closely linked to their marginality and impoverishment. If there had been
something familiar to the historian about the luxury-loving Vandals and
their excessive indulgence in baths, good food, fine clothes, beautiful
gardens, and pleasurable entertainments, the same was not true of the
Moors. For of all the peoples of which we know, Procopius wrote, that
of the Moors is the most hardy.
The Moors live in stifling huts in winter and in summer and at every other time,
and they are not driven out from these either by snows or by the heat of the
sun or by any other constraining nuisance. And they sleep upon the ground, the
fortunate ones, if it should happen thus, laying a sheepskin under themselves.
Their custom is not to change their clothes with the seasons, but they wear a
thick cloak and a rough tunic all the time. And they have neither bread nor
wine nor any other good thing, but they eat their grain, either spelt or barley,
neither boiling it nor grinding it into meal, nor at all differently than the other
animals.52

As far as Procopius was concerned, the Moors did not live like Romans,
they did not dress like Romans, they did not eat like Romans, and they
were not even properly human.
Even so, one is left with the impression that the cultural differences that
most struck Procopius between Moors and Romans were due as much to
the perceived poverty of the Moors as to anything else. Procopius most
implausible claim here that Moors ate their grain raw is an interesting
variation on Greco-Roman ethnographic tropes about nomads, who
were often said to eat the uncooked or partly cooked flesh of animals.53
However, like the claim that Moors did not eat bread, it was probably
50

51
52

Zachariah of Mitylene, Chronicle 12.7, p. 326. On the similarities between Syriac and Punic,
see, e.g., Jerome, Commentarius in Ionam 4.6, ed. Y.-M. Duval, SC 323 (Paris, 1985), p. 298 and
Augustine, In Iohannis evangelium tractatus 15.27, p. 162.
Severianus: Proc. BV 2.23.6, 1:526. Rhodon: Proc. Anecd. 27.3, 3:166.
53 Shaw, Eaters of Flesh, esp. pp. 1213.
Proc. BV 2.6.5 and 2.6.1013, 1:4435.

263

Staying Roman
never meant to be taken literally. In this passage, Procopius is primarily
drawing a contrast between the respective abilities of Vandals and Moors
to endure hardship. Moreover, only a few sentences on, the historian
contradicts his own testimony by recounting the story of a Moorish
woman who ground some grain and made a barley cake ( %). She
then threw it onto the hot ashes of the hearth to cook; for among
the Moors, Procopius now writes, the custom is thus for loaves to be
baked.54 The historians point was thus not so much that Moors did not
eat bread, but rather that they did not eat the panis siligneus (bread-wheat
bread) long preferred by the Roman upper classes.55 The issue was the
Moors grain of choice. Earlier, in the late fifth century, Victor of Vita
had reacted with a similar level of disgust to the barley with which their
Moorish guards fed the Nicene clergy exiled by the Vandals to the Chott
el Hodna region of Mauretania Sitifensis. The Romano-Africans, Victor
tells us, came to places of desolation in which, having been billeted,
they received barley to eat, like beasts of burden.56 Critically in the
arid regions of Mauretania, Numidia, Byzacena, and Tripolitania under
Moorish domination, barley requires less water to grow than does wheat.
It also requires less labour to cultivate, and was consequently cheaper than
wheat in antiquity. For both Procopius and Victor of Vita, barley was
feed for animals; but it was also the food of slaves and the poor.57
Much the same may have been true of Moorish dress. To be sure, the
pseudo-Augustinian author of the Questions of the Old and New Testament
informs us that it is the custom of the Moors that women wear earrings
even in [their] noses58 a practice which was presumably not in fashion
among the women of late antique Roman Africa and in at least two
cases Corippus tells us that Moorish men wore their hair long, this being
a common trope in Roman descriptions of barbarians, and one which
Corippus himself was later to use when describing the Avars.59 Coiffure
54
55
56
57
58
59

Proc. BV 2.7.3, 1:448: 7!     \  

 G D  ]+; see also Coripp.
Ioh. 5.4323: cecidit lectusque lapisque / quo Cererem frangit.
R. Sallares, Ecology, in W. Scheidel, I. Morris, and R. Saller (eds.), Cambridge Economic History
of the Greco-Roman World (Cambridge, 2007), p. 32.
Vict. Vit. 2.37, p. 38: ad solitudinis loca perueniunt, in quibus collocati hordeum ad uescendum
ut iumenta accipiunt.
Sallares, Ecology, p. 31.
Pseudo-Augustine, Quaestiones Veteris et Novi Testamenti 115, PL 35:2349: Mos Maurorum est ut
inaures etiam in naribus habeant feminae.
Coripp. Ioh. 5.113 and 8.193, pp. 96 and 172. For Corippus description of the Avars long hair
and the trope of long barbarian hair in general, see above n. 38. According to Procopius, Anecd.
7.10, 3:445, the Huns wore their hair completely cut off as far as the temples in the front, while
for no reason they allow that at the back to hang down to the greatest length (E A  
;) E   + D  G 1;  8    =+ 8 +
;
 /  ) / ^!). In the same passage, however, Procopius tells us that

264

The Moorish alternative

and ornamentation thus may have marked out Moors from Romans;
but it is difficult to understand how the tunic and cloak (tribonion) that
Procopius says Moors wore year-round could have done so. At least,
such garments do not sound strikingly different from the tunic and
cloak (paenula) that the late Roman and Vandal-era men of Tabarka are
depicted as wearing in that northern port citys remarkable collection
of funerary mosaics.60 There the tunic and cloak appear to have been a
sign of humbleness and modesty, and much the same seems to have been
true in Augustines Hippo.61 Of course at the time there may well have
been perceptible differences between Procopius Moorish tribonion and
the Tabarkan paenula that have since become indistinguishable to us. But
context may also have made the meaning, and what Procopius read as
an indicator of barbarism in the mountains of Numidia may well have
been understood as a symbol of Christian humility in the cities of the
Mediterranean coast.
Both Procopius and Corippus also agreed that Moors did not fight
like Romans. In part this was a question of battle-gear. Corippus is
explicit that Moors and Byzantines were distinguishable by their dress and
weapons.62 Procopius indicates that many Moors went into battle lightly
clad and without armour ( 
).63 This seems to agree with Corippus
more detailed descriptions, including that of the Barcaei warriors whom
John Troglita encountered in Byzacena and who, according to the poet,
wore a linen mantle that was knotted about the head, a cloak that hung
from the shoulders, and sandals or shoes of some sort, as well as armbracelets to which they attached a shield, and a sword slung over the

60
61
62
63

this had become quite the fashion throughout the Byzantine empire as a result of the influence
of the circus factions.
See J. Downs, The Christian Tomb Mosaics from Tabarka: Status and Identity in a North African
Roman Town, Ph.D. diss., University of Michigan (2007), pp. 1947.
Augustine, Sermo 356.13, PL 39:157980; Downs, Christian Tomb Mosaics, 1956.
Coripp. Ioh. 8.18895, p. 172, esp. l. 188: habitus gentes discernit utrasque (its dress distinguishes
each of the two tribes).
Proc. BV 2.11.26, 1:466; see also Proc. BV 1.19.7, 1:392:  / \  
. See also Procopius description of Slavic and Herul warriors, of whom, however, he does not use the word
 
: Proc. BG 3.14.256, 2:358 (Slavs) and Proc. BP 2.25.278, 1:2667 (Heruls). Procopius
use of   to mean without armour in a military context: see Proc. BP 2.8.28 and 2.16.2,
1:188 and 1:221; Proc. BG 1.29.25, 3.21.5, and 3.30.22, 2:143, 2:390, and 2:430, and esp. ibid.,
2.2.22 and 4.35.29, 2:157 and 2:676; see also Proc. Anecd. 7.28, 3:478. Comparable use in a
non-military context: Proc. BP 1.25.42, 1:141; Proc. Anecd. 9.20, 3:5960; and perhaps Proc. BP
1.25.40, 1:141; but see also ibid., 1.7.18, 1:33 and Proc. Anecd. 9.14, 3:58. When using  
to mean naked, Procopius tended use the intensifying adverb 1, completely: Proc.
Anecd. 9.20, 3:60 and Proc. BV 1.23.11, 1:408. The latter seems to imply that Byzantine soldiers
slept entirely without clothes. Persians too slept naked: Proc. BP 2.30.42, 1:302. See further
Bowersock, Brown and Grabar, Late Antiquity, pp. 61516, s.v. Nudity.

265

Staying Roman
back.64 Different individuals also seem to have been set apart by their
distinctive dress: Corippus refers to the Moorish chieftains Varinnus and
Alacanza as being feathered (pinnatus), presumably meaning that they
wore some kind of plumed battle dress into combat.65
In part, however, what these Byzantine authors felt distinguished
Moors from Romans on the battlefield was a question of tactics. We
hear again and again, both in Procopius History of the Wars and especially
in Corippus Iohannis, of the Moors practice of circling their camels
and other livestock to create a strongly defensible barrier around their
palisaded camp.66 The Austur were said to have thrown up walls and
trenches within whose protective rings they placed their herds.67 The
Moorish chieftain Ierna was said to have done something similar, surrounding his camp with eight rows of camels, various traps and snares, a
ring of herd animals, and a barricade of pitchforks, rocks and sharpened
stakes.68 Later, the Byzantine commander John Troglita cut through the
legs of a camel to reach the Moorish camp.69 On another campaign, the
Moorish chieftains Carcasan and Antalas quickly dug trenches and circled
their camels to defend themselves against a Byzantine attack.70 According
to Corippus, Moors were also quite willing to lay waste the countryside
and engage in delaying tactics so as to weaken the Byzantine army by
starvation, though it is worth pointing out that this was also a thoroughly
Roman tactic.71 Moors were apparently skilled at guerrilla warfare, and
they seem to have preferred fighting in the mountains to fighting on the
open plains.72 We also hear of Moorish troops leading hostile forces into
ambushes, or lying in wait for their enemies along steep river banks and
in oak or olive groves, where they enjoyed a tactical advantage over the
imperial troops.73 Again, however, there is some question as to how far
such tactics would have served to distinguish Moors from Romans; at
64

65

66
67
69
71
72
73

Coripp. Ioh. 2.12537, pp. 312; J. Desanges, Catalogue des tribus africaines de lantiquite classique a`
louest du Nil (Dakar, 1962), p. 150, s.v. Barcaei, Barkitae, locates the Barcaei in the province of
Cyrenaica, in the region south of the Pentapolis.
Coripp. Ioh. 5.264, p. 102 (of Misantas); ibid., 7.419 and 7.510, pp. 159 and 163 (of Varinnus);
and ibid., 8.543, p. 185 (of Alacanza); see also Virgil, Aeneidos 9.473, p. 321 (of Fama), the only
instance of the word cited in H. H. Warwick, A Vergil Concordance (Minneapolis, Minn., 1975),
p. 642, and in general Thesaurus Linguae Latinae (Leipzig, 1900), 10/1, fasc. 7, cols. 10946, s.v.
pennatus.
Proc. BV 1.8.256 and 2.11.1719, 1:350 and 1:465. For references in Corippus Iohannis, see
below.
68 Coripp. Ioh. 4.597619, pp. 8990.
Coripp. Ioh. 2.919, p. 30.
70 Coripp. Ioh. 8.3640, p. 166.
Coripp. Ioh. 5.42138, pp. 1089.
Coripp. Ioh. 7.3049, p. 155. On such tactics in Byzantine warfare, see J. Haldon, Warfare, State
and Society in the Byzantine World, 5651204 (London, 1999), pp. 378.
Proc. BV 2.12.4, 1:471 and Coripp. Ioh. 8.3840, p. 166.
Coripp. Ioh. 1.52263 and 2.515, pp. 235 and 28; see also ibid., 8.173, p. 171, where Corippus
seems to indicate that the imperial forces did not like fighting in forests.

266

The Moorish alternative

least, Byzantine soldiers were themselves not above ambushing hostile
Moors.74
On the question of Moorish paganism, too, we have reason to doubt
the testimony of our principal sources. Augustine was aware of the failure of Christian missionaries to evangelize the pacified Moors who lived
under prefects appointed by the Romans beyond the empires frontier.
Though Augustine tells us that these prefects had begun to become
Christians, in the minds of other late Roman authors, their people had
not.75 In his On the Government of God, Salvian of Marseilles asks, Is
the sacrosanct law brought into reproach because of the savage customs
of the Moors? Salvian clearly expected his readers to answer the question in the negative, because the Moors, as non-believers, were not
subject to the law of Christ.76 According to Procopius, Moors practised polygamy, and in Corippus account the Moors of Byzacena and
Tripolitania were inveterate pagans who had no respect for the sanctity of Christian holy days and continued to perform animal and even
human sacrifice.77 Paganism certainly survived in the North African
countryside into the sixth century, including the pre-desert wadi valleys
of Tripolitania, which had by this time been integrated into the Moorish
Laguatan confederation (plu. Ilaguas); and while it is perhaps conceivable that the rites practised there at the temple at Ghirza involved severed
heads, Corippus claim that Moors practised human sacrifice may simply have been intended to signal that, from the poets point of view,
they did not follow accepted social norms.78 Certainly, human sacrifice
was one of Procopius favourite ethnographic bogeymen, and he would
probably have flung the accusation at the Moors if he thought it would
stick.79
Indeed, despite the hostile testimony of Corippus, there is every reason to believe that in the Byzantine period Christianity was substantially
74
76
77

78

79

75 Augustine, Ep. 199.12.46, CSEL 57:2845.

Proc. BV 2.10.5, 1:459.
Salvian, De gubernatione Dei 4.17.82, p. 296: Numquid propter Maurorum efferos mores lex
sacrosancta culpatur?
Polygamy: Proc. BV 2.11.13, 1:464; see also ibid., 2.10.11, 1:460. Paganism: Coripp. Ioh. 2.10911,
3.815, 5.494502, 6.14590, 8.30017, pp. 30, 50, 111, 11820, and 1767. Animal sacrifice:
Coripp. Ioh. 3.86140 and 8.30017, pp. 502 and 1767. Human sacrifice: ibid., 8.3079,
pp. 1767. Moors attack Byzantines on a feast day: ibid., 8.2547, pp. 174475.
Thus Pallas-Brown, East Roman Perceptions, p. 312 of the Avars. On Moorish paganism, see
O. Brogan and D. J. Smith, Ghirza: A Libyan Settlement in the Roman Period (Tripoli, 1984), p.
88 and pl. 125; D. J. Mattingly, Explanations: People as Agency, in G. W. W. Barker, D. D.
Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The UNESCO Libyan Valleys
Archaeological Survey, 2 vols. (Paris and London, 1996), 1:3378. See also Cameron, Byzantine
Africa, p. 40.
Proc. BP 1.19.36, 1:106 (Blemyes); Proc. BG 2.14.1, 1:208 (Heruls; but see also ibid., 2.14.334,
2:213); ibid., 2.15.245, 2:218 (Thule); and ibid., 2.25.910, 2:262 (Franks). See also Proc. BP
1.19.15, 1:102, where he accuses Saracens of cannibalism.

267

Staying Roman
widespread among many of the peoples that late Roman authors thought
of as Moorish. This was apparently less true in the pre-desert Tripolitanian valleys, but the Gebel area to the north is peppered with Christian
sites, and, indeed, even in the pre-desert Ghirza is the only known pagan
cult centre.80 Moreover, by the middle of the sixth century, the conversion of the Moors of Tripolitania and Cyrenaica to Christianity was a
major policy objective of the Byzantine imperial administration.81 We
also hear of earlier efforts at evangelization: in the late fifth century Victor of Vita wrote of the efforts of a handful of exiled Romano-African
Catholics to convert the Moors of Caprapicta, a region that cannot
be precisely localized but was perhaps the part of south-western Tunisia
north of the Chott el-Djerid, between the cities of Gafsa and Negrine,
in what would at the time have been the province of Byzacena.82 This is
the same region to which Yves Moderan has localized the sixth-century
Moorish chieftain Cusina, who may himself have been a Christian.83
The same is true of Antalas, a contemporary chieftain whose territory
lay to the north of Cusinas, in the hill country west of Thelepte.84 A
good number of the people of these regions would certainly have been
Christians. Indeed, Moderan has identified archaeological evidence of
Christianization at no fewer than 183 sites in these chieftains territories, in and around the highlands of the eastern Aur`es Mountains. These
include basilicas, chapels, inscriptions, crosses and other artistic motifs.85
Such evidence extends further west as well. The late-fifth-century Djedar
(monumental tombs in the region of Frenda, in west-central Algeria) similarly employ Christian elements in their decoration, including a cross,
doves, a peacock or dove sitting in a basin, and a figure with a cross
and red mitre.86 In the latter part of the fifth century, the Moorish prefect Iugmena began building a church near Thanaramusa Castra (mod.
Berrouaghia) in the highlands of central Algeria which was subsequently

80

81
82
83
86

Mattingly, Explanations, pp. 3378; J. B. Ward-Perkins and R. G. Goodchild, The Christian

Antiquities of Tripolitania, Archaeologia: Miscellaneous Tracts Relating to Antiquity 95 (1953),
pp. 3556; A. di Vita, La diffusione del Cristianesimo nellinterno della Tripolitania attraverso
i monumenti e sue sopravvivenze nella Tripolitania araba, Quaderni di Archeologia della Libia 5
(1967), pp. 12142; Brogan and Smith, Ghirza, pp. 808; Brett and Fentress, Berbers, p. 66.
Proc. Aed. 6.2.1820, 6.3.911, and 6.4.12, 4:1745, 4:1778, and 4:181; John of Biclar, Chronicon
s.a. 568.7, ed. Collins, p. 61.
Vict. Vit. 1.358, pp. 1617; Courtois, Victor de Vita, pp. 378.
84 Ibid., pp. 31634.
85 Ibid., pp. 53440 and map 17.
Moderan, Maures, pp. 33846.
G. Camps, Rex gentium Maurorum et Romanorum: recherches sur les royaumes de Mauretanie des
VIe et VIIe si`ecles, Antiquites africaines 20 (1984), pp. 2018. The standard study of the Djedar is
F. Kadra, Les Djedars, monuments funeraires berb`eres de la region de Frenda (Wilaya de Tiaret, Algerie)
(Aix-en-Provence, 1974).

268

The Moorish alternative

completed, in 474, under the auspices of one Zabenses.87 At some point
in the sixth century, the Moorish king of the Ucutamani was commemorated in an inscription that began with the formula In the peace + of
Christ and that called the king the servant of God.88 A sixth-century
inscription from Arris in the western Aur`es commemorating the local
Moorish leader, Masties, is also Christian in character.89 At least in the
west, then, Corippus claims of intractable Moorish paganism appear to
be somewhat disingenuous.
Finally, Moors were also typically thought of as darker-skinned than the
Byzantines and Romano-Africans living along the Mediterranean coast.
Procopius says as much, calling the Moors dark-skinned or swarthy
( )) as a people.90 Indeed, by the time Isidore of Seville came
to write his Etymologies, the word Maurus or Moor had become an
adjective in Latin, for the Greeks call black mauron.91 In Isidores day,
Moors were black by definition. The Carthaginian deacon Ferrandus
once wrote to his mentor, Fulgentius of Ruspe, concerning the slave
of a certain religious man, in age a youth, in colour an Ethiopian, from
(I believe) the furthest regions of the barbarian province, where mens dry
limbs are darkened by the heat of the suns fire.92 Despite the reference to
Ethiopia, the youth was probably a Moor. Describing the captive Moorish women paraded through the streets of Carthage after John Troglitas
victory over the Moorish leader Antalas confederation, Corippus tells us
that not all the captives were of one colour;93 but on the same occasion
and on several others the poet also refers to Moors, both individually
and collectively, as being black or dark (niger), and even goes so far as
to liken one Moorish woman and her children to a raven and its chicks.94
87
88
89

90
91

92

93
94

C. Courtois, Les Vandales et lAfrique (Paris, 1955), p. 375, no. 68 and PLRE 2:634, s.n. Iugmena,
both citing Revue archeologique, 5th ser. 24 (1926), p. 323, no. 60.
CIL 8.8379 = 20216 and Camps Rex Maurorum, pp. 199200: (In pa)ce+ (Christi) in monti
Mux . . . / (Dei s)erbus et rex gentis Ucutaman(orum) . . .
See esp. P. Morizot, Pour une nouvelle lecture de lelogium de Masties, Antiquites africaines 25
(1989), p. 274, with fig. 1, p. 265; see also Courtois, Vandales, p. 382, no. 132. The cross which
Morizots new reading places in the second line of the inscription only reinforces the Christian
nature of the text, which is apparent with or without the cross.
Proc. BV 2.13.29, 1:479.
Isidore of Seville, Etymologiae 12.1.55, ed. W. M. Lindsay, in Isidori Hispalensis episcopi Etymologiarum sive originum libri XX, 2 vols. (Oxford, 1911), n.p.: Mauron niger est; nigrum enim Graeci
# vocant. See also Isidore, Etymologiae, 9.2.122: Mauri ob colorem a Graecis vocentur.
Graeci enim nigrum # vocant.
Fulg. Ep. 11.2, 1:360 (Ferrandus to Fulgentius): Religiosi cuiusdam uiri famulus aetate adolescens, colore Aethiops, ex ultimis credo barbarae prouinciae partibus, ubi sicca hominum
membra solis ignei calore fuscantur.
Coripp. Ioh. 6.92, p. 116.
Raven metaphor: Coripp. Ioh. 6.936, p. 116. Other examples: ibid., 1.2446, 2.137, 7.426, and
8.594, pp. 1213, 32, 160, and 187. On the raven metaphor, see also AL 172 (De Aegyptio),

269

Staying Roman
This sensitivity to skin pigmentation seems not to have been restricted
to inhabitants of the coastal regions of Africa and foreigners from the
Byzantine East or Visigothic West. At least, one of the late fifth-century
wooden tablets found near Djebel Mrata on the south-western fringes of
the Vandal kingdom, near the present-day border between Algeria and
Tunisia, records the sale of a six-year-old slave boy named Fortinis who
was white in colour (coloris candidum).95
Procopius makes reference to the dark skin of the Moors without further comment, but earlier in his History of the Wars, the historian remarks
that Ephthalite Huns are the only Huns who have white bodies and are
not unseemly in appearance a statement which would seem to indicate that Procopius generally found people whom he considered darkskinned to be ugly.96 A similar prejudice seems to have been expressed
by the sixth-century Romano-African poet Luxorius. Myrro, writes
Luxorius, loves hideous and foul girls; those with ordinary or beautiful
faces whom he sees, he fears . . . so that a Pontic [=Punic?] girl does not
please you, but a Garamantian does.97 A problem with the text makes
it difficult to know what kind of women the poet found beautiful, but
in Luxorius eyes and he clearly expected his audience to agree with
him on this point Moorish ones were ugly.98 That this may in part
have had to do with the poets prejudices about skin colour is suggested
by another epigram in which Luxorius exclaims to the Egyptian arena
hunter Olympius, Your darkened form and black colour in no way harm
you, and then proceeds to expand upon a number of other good things

95
96

97

98

pp. 1212. See, however, J. Desanges, The Iconography of the Black in Ancient North Africa,
in J. Vercoutter, et al. (ed.), The Image of the Black in Western Art, vol. 1, From the Pharaohs to
the Fall of the Roman Empire (Cambridge, Mass., 1976), pp. 24668, cited with approbation in
F. Snowden, Jr., Before Color Prejudice: The Ancient View of Blacks (Cambridge, Mass., 1983),
p. 9. Whether or not any of these populations would have been considered black by modern
American or European standards is irrelevant.
Tablettes Albertini: actes prives de lepoque vandale (fin du Ve si`ecle), ed. C. Courtois, L. Leschi,
C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952), act 2, 1:217.
Proc. BP 1.3.4, 1:11:  A (V! _ ) 
  :  /  D ;  =>
,
. That beauty and fairness were connected in Procopus mind seems to be indicated by Proc.
BV 1.2.4, 1:311: all of them [the Gothic peoples] have white bodies and golden hair, and are tall
and handsome in appearance () 
  <  : 1 , /    +/,
 2  / 8+/  =>). One need not agree with the prejudices of Procopius and his
society to recognize them.
AL 324, p. 261: Diligit informes et foedas Myrro p<u>ellas; / quas <medio> aut pulchro viderit
ore, timet. / iudicium hoc quale est oculorum, Myrro, fatere, / ut tibi non placeat Poenica sed
Garamas?, The Codex Salmasianus (Paris BN Lat. 10318) reads pontica for Poenica, perhaps
meaning a coastal girl rather than a Punic one; the sense would be much the same in either
case.
On the Garamantes, see Desanges, Catalogue des tribus, pp. 936. Luxorius prejudices seem to
have been widely accepted by his society, but they need not have been universal: Snowden, Before
Color Prejudice, p. 63; see also ibid., pp. 759.

270

The Moorish alternative

that are black, such as ebony, murex, elephants, and pepper.99 The resulting statement may well be a laudatory multiple image of blackness,100
but the implication is nonetheless that Olympius dark skin might normally be considered a disadvantage, an attitude certainly also represented
in two anonymous poems gathered with those of Luxorius into the Latin
Anthology.101
Luxorius younger contemporary Corippus leaves us in no doubt about
the repugnance and fascination with which he and, it would seem, other
Romano-Africans viewed their southern and western neighbours a
repugnance which, at least to some extent, seems to have been associated
in the poets mind with the very darkness of the Moors. When crossing
from Constantinople to Africa, John Troglita was said to have had a
vision of a fallen angel which Corippus describes as having appeared to
the general with a face like a Moors in darkness, and frightful in its black
colour.102 However, John, we are told, did not fear the expressions of
a man imitating savage forms.103 Later, Corippus describes the meeting
of John Troglitas emissary with the Moorish chieftain Antalas and his
gathered warriors in similarly infernal terms: black faces filled up the
tents just as they say Dis, provoking a battle with the gods, once held
a council and a thousand monsters came through the spacious ways.104
This perception of the Moors as in some sense the earthly embodiment of the cosmic forces of evil was not unique to Corippus. An
anonymous poem included in the Latin Anthology discusses a Moorish
slave in strikingly similar terms:
The dregs of the Garamantes have now advanced to our region, and the houseslave Niger [= black] rejoices in his pitch-black body. If the voice discharged
from his lips did not make him sound human, the gristly demon would terrify
living men. Hadrumetum, may the fearful infernal regions carry off the monster!
The house of Dis should have him as a guardian.105
99
100
101

102
103
104
105

AL 348, pp. 2734: nil tibi forma nocet nigro fu<s>cata colore.
Snowden, Before Color Prejudice, p. 8.
AL 173 and 179, pp. 122 and 1245; L. Thompson, Blacks and Romans (Norman, Okla., 1989),
pp. 313. However, see also Snowden, Before Color Prejudice, pp. 779, with AL 288, 348, and
349, pp. 240 and 2735, which nevertheless do not imply that Luxorius let alone his society
at large was free of colour prejudice as such.
Coripp. Ioh. 1.2446, pp. 1213: cognata tenebris / Maura uidebatur facies nigroque colore /
horrida.
Coripp. Ioh. 1.2545, p. 13: nec timuit tamen ipse feras imitantia formas / ora uiri.
Coripp. Ioh. 4.3214, p. 79: nigrae facies tentoria complent: / ut quondam Ditem moturum
proelia diuis / concilium fecisse ferunt et mille per amplas / monstra uias uenisse . . . .
AL 173, p. 122: Faex Garamantum iam nostrum processit ad axem / et piceo gaudet corpore
verna Niger, / quem nisi vox hominem labris emissa sonaret, / terreret vivos horrida larva viros. /
dira, Adramentue, tu<u>m rapiant sibi Tartara monstrum; / custodem hunc Ditis debet habere
domus.

271

Staying Roman
Quodvultdeus of Carthage had a positively apocalyptic vision of these
southern tribesmen: Gog and Magog, as some say, Goths and Moors,
Geats and Massageats, through the savagery of whom the devil already
lays waste the Church and will take an even greater vengeance then.106
Writing a half-century or so after Corippus, Gregory the Great tells the
story of a young boy tormented by evil demons (maligni spiritus) who,
when asked by his father what frightened him so, replied, Moorish men
are coming, who want to take me away.107 The anxieties that Moors
raised in western Roman populations seem to have run very deep indeed.
Corippus also refers to the foul head (taetrum . . . caput) of Moors, or
at least of Moorish men, in general.108 Nor were Moorish women free
from his censure. Corippus applies the adjective frightful (horrida) to the
dark-skinned Moorish woman whom he had compared to a raven, horrida
being one of Corippus more common epithets for things Moorish.109
This woman and her children, and perhaps the other Moorish captives
who were paraded through the streets of Carthage, were, moreover, gaped
and gawked at by the inhabitants of the metropolis: Fathers and mothers
enjoyed showing their little children [these] horrible faces.110 Corippus
himself seems clearly to have found Moors physically repulsive, while
by his own account the Carthaginians as a whole found these Moorish
captives a fascinatingly exotic spectacle.
To Procopius and Corippus, then, Moors were in a very real sense
defined by their non-Romanness. Their vision is not necessarily a reflection of the realities of life in the borderlands of Byzantine North Africa.
Indeed, as we have seen, we have plenty of grounds on which to doubt
their interpretations of Moorish society. Yet these authors visions also
do not appear to have been entirely ideologically driven: many of the
elements pre-date the Byzantine reconquest of Africa, and Procopius and
Corippus both express a view of the Moors which seems to have been
shared by other authors throughout the late Roman and early Byzantine
world. To these authors, Moorish difference was expressed both in the
immediately apparent markings of cultural identity (dress and coiffure,
language, perceived physical differences) and in social practice (religion,
106

107
108
110

Quodvultdeus, Liber promissionum 4.13.22, p. 207: Gog et Magog, ut quidam dixerunt, Gotos
et Mauros, Getas et Massagetas, per quorum saeuitiam ipse iam diabolus ecclesiam uastat et tunc
amplius persequetur. On the connections between Goths and Geats, see Isidore, Etymologiae,
9.2.118. On the association of the Goths with Gog and Magog, see H. Wolfram, History of the
Goths, trans. T. J. Dunlap (Berkeley, Calif., 1988), pp. 289.
Gregory I, Dialogi 4.19.3, 724: Mauri homines uenerunt, qui me tollere uolunt.
109 Coripp. Ioh. 6.93, p. 116.
Coripp. Ioh. 2.1356, p. 32.
Coripp. Ioh. 6.978, pp. 11617: horribiles uultus paruis ostendere natis / dum patres matresque
libet.

272

The Moorish alternative

diet, marriage customs, and battle tactics). Their ferocity and faithlessness as well as the marginal nature of the territory they inhabited further
marked Moors out as singularly non-Roman. It was a foreignness which
left both Procopius and Corippus cold. Neither author writes of Moorish society with much sympathy or understanding. To Procopius, Moors
were not even truly human: they could satisfactorily be classed with the
other animals a taxonomy into which, for all his Roman chauvinism,
the historian never classifies any of the other barbarian peoples whom
he discusses in any of his extant corpus.111 To Corippus and, indeed, to
a number of other writers from the late antique West, the Moors were
even worse: they were positively demonic.
3. moors and mauretanians
Our evidence for how Moors viewed themselves is rather more limited.
There is no Moorish Procopius to chronicle the military struggles of the
mid sixth century from a Moorish point of view, no Moorish Corippus
to render those struggles into an epic battle of cosmic proportions. Nor
has the traumatic history of Algeria where most of the western Moorish
kingdoms were based lent itself to modern archaeological exploration.
What does survive from fifth- and sixth-century north-western Africa is
a remarkable wealth of epigraphic evidence. The very fact of erecting an
inscription speaks volumes, and augmented by the literary accounts and
a few archaeological field surveys, these inscriptions seem to indicate that
the social processes under way in the African interior in the later fifth
and sixth centuries were fundamentally very similar to those taking place
along and across the other borders of the western Roman empire at the
time. Before turning to this evidence, though, we must consider what
Latin-speakers meant by the word Maurus (Moor) in late antiquity, and
how this changed as imperial control of the West collapsed in the fifth
century not the transition from noun to adjective which we find in
Isidores Etymologies, but an earlier and more revealing shift in meaning.
In the late Roman period, Mauri had generally been thought of
as a people (a gens or populus).112 Indeed, in 320, a certain Victor, a
111

112

The nearest parallels are Proc. BG 2.15.16, where Procopius says that, of the thirteen tribes of
Thule ancient Scandinavia only the Scrithiphini lead something of a savage life (+:
 *&  ); and Procopius reference to the Centaurs, the half-men, half-horses of
Greek mythology who, Procopius says at Aed. 4.3.12, 1:114 were foolishly said to have had
the commingled nature of two creatures (%! ;0   " 0 ). For Romans
to speak of barbarians as animals or animal-like was, however, nothing new: Dauge, Barbare,
pp. 6049.
See, e.g., Liber generationis 1.1967.64, ed. Mommsen in MGH AA 9:107; Chronica Alexandrina
1667.64, ed. Mommsen in MGH AA 9:107; Servius, Commentarius in Vergilii Aeneidos 6.60,

273

Staying Roman
grammarian from Constantina in Numidia, even testified in a legal proceeding that his family was of Moorish blood (de sanguine Mauro).113
Yet the term Maurus seems to have had ethnic connotations only in a
very broad sense, for it was the general Latin word for the indigenous
inhabitants of the western Maghrib.114 At the end of the fourth century,
the word Maurus could still be deployed by (non-African) historians and
poets alike to evoke a lack of Romanness; but by Augustines day the term
also referred perhaps even primarily, at least in an African context
to the Roman inhabitants of the Mauretanian provinces, rather than to
the desert tribesmen and montagnards who lived beyond the frontiers
of Roman power.115 These peoples Augustine tended to call either Afri
barbari or simply barbari, (African) barbarians.116 But the appearance of
the Vandals in Africa changed this situation in two important ways. First,
the local entrenchment of the new arrivals rendered Augustines terminology ambiguous. Afri barbari could now refer to either of two groups
who were, from a Roman point of view, equally unsavoury: Vandals or
indigenous non-Roman populations.117 Second, the Vandal presence in

113
114

115

116

117

ed. G. Thilo in Servii grammatici qui feruntur in Vergilii carmina commentarii, ed. G. Thilo and H.
Hager, 3 vols. (Leipzig, 18811902), 2:15; Pompeius, Commentum artis Donati, p. 205; Claudian,
In Gildonem ll. 956, ed. Hall, in Claudianus Carmina, p. 111; see also Ammianus Marcellinus,
Res gestae 26.4.5, 2:9, ibid., 30.7.10, 2:152 (gentes Mauricae), and ibid., 29.5.2, 2:114 (nationes
Mauricae) and Dio Cassius, Historiae Romanae 78.11.1, ed. U. P. Boissevain in Cassii Dionis
Cocceiani Historiarum Romanarum quae supersunt, 5 vols. (Berlin, 18951931), 3:413 ( 
\#). See in general Moderan, Maures, pp. 41744.
Gesta apud Zenophilum, ed. C. Ziwsa as Appendix 1 in Optatus of Mileu, Libri VII, CSEL 26
(Vienna, 1893), p. 185.
Y. Moderan, Les Maures de lAfrique romaine dans lantiquite tardive, Revue des etudes latines
82 (2004), pp. 25968; cf. the indigenous term Mazices, presumably a cognate of the modern
Berber self-epithet Imazighen, on which see Brett and Fentress, Berbers, pp. 56.
Non-Romans: see, e.g., Ammianus Marcellinus, Res gestae 22.7.10, 1:259 and Claudian, In
Gildonem, esp. ll. 18991 and 4324, pp. 114 and 124; see also Expositio totius mundi 60, p. 200.
Mauretanians: Augustine, Epp. 59.1 and 93.10.43, CSEL 34:219 and 487; Augustine, De utilitate
credendi 7.17, ed. J. Zycha, CSEL 25/1 (Vienna, 1891), p. 21; Augustine, De unitate ecclesiae
(= Epistula ad catholicos de secta Donatistarum) 3.6, ed. M. Petschenig, CSEL 52 (Vienna, 1909),
p. 237; and Augustine, Contra Cresconium 4.58.69, ed. Petschenig, CSEL 52:568. This usage was
not unique to Augustine: Ammianus Marcellinus, Res gestae 27.6.14, 1:44 and Registri ecclesiae
Carthaginensis excerpta 72, p. 202. See also P.-A. Fevrier, Differences et conflits: Maures et
Barbares, in Approches du Maghreb romain: pouvoirs, differences et conflits, 2 vols. (Aix-en-Provence,
198990), 2:1413; Moderan, Les Maures de lAfrique, pp. 2678; and J. Moralee, Maximinus
Thrax and the Politics of Race in Late Antiquity, Greece & Rome 55 (2008), pp. 5582, esp.
645 and 723.
Afri barbari: Augustine, Ep. 220.7, CSEL 57:436. Barbari: Augustine, De civitate Dei 16.6, 48:507;
Augustine, De unitate ecclesiae 15.37, p. 280; Augustine, Ep. 199.12.46, CSEL 57:284; see also
Liber generationis 1.197.67, p. 107.
See, e.g., Vandals: Quodvultdeus, De cataclysmo 6.17, p. 419; Quodvultdeus, De tempore barbarico
1.4.13 and 1.7.32, pp. 429 and 435; Quodvultdeus, De tempore barbarico 2.5.11, 2.5.13, 2.6.8,
and 2.6.11, pp. 4778; Fulgentius of Ruspe, Ad Trasamundum 1.2.2, p. 99; pseudo-Fulgentius,
Sermones 46, 64, and 80, cols. 912, 936, and 953; see also ibid., 78, col. 950 (of the Pharaonic

274

The Moorish alternative

Africa set in motion or in some places perhaps accelerated a reconfiguration of power in the interior that ultimately led to the ascendancy
there of local warlords and strongmen. As far as writers in Carthage
and elsewhere in the Mediterranean were concerned, by the later fifth
century the word Maurus had lost forever the Roman connotations that
it had gained by Augustines day. To these authors, Mauri wherever
they lived were first and foremost Moors, and only incidentally (if
at all) inhabitants of the old imperial provinces of Mauretania Sitifensis,
Mauretania Caesariensis, or Mauretania Tingitana.118
Given both the vast geographical area involved and the diversity of
local circumstances, the Moorification of the African hinterland seems
to have played out differently in contexts ranging from the Mauretanias
in the west to Tripolitania in the east.119 In Numidia retroceded to the
empire by Geiseric in the treaty of 442 powerful, affluent landowners appear to have raised private armies through which they sought to
augment their status and power. The western emperor Valentinian III
legitimated the maintenance of such armed bands for the purpose of
making raids into Vandal territory, but charged the dux Numidiae with
preventing attacks by irregular troops within the province itself.120 This
in turn would seem to suggest that border warfare at least occasionally degenerated into internal violence as local magnates vied for wealth
and control in the region.121 The situation is unlikely to have changed
much as the Vandals re-incorporated Numidia and part of Mauretania

118

119

120
121

Egyptians, often a code-word for the Vandals in the fifth-century African sermons). Moors and
Vandals interchangeably: V. Fulg. 5, p. 33 and V. Fulg. 21, p. 103; Vict. Vit. 1.25, p. 12. Moors:
Fulg. Ep. 11.2, 1:360 (Ferrandus to Fulgentius); Vict. Tonn. s.a. 523.106, p. 34 (see also, in a
non-African context, ibid., s.aa. 512.93 and 514.95, pp. 30 and 32). More broadly: Facundus of
Hermiane, Pro defensione trium capitulorum 12.4.12 and 12.5.9, ed. J.-M. Clement and R. Vander
Plaetse, in Opera omnia, CCSL 90A (Turnhout, 1974), pp. 392 and 395.
See, e.g., Salvian of Marseilles, De guernatione Dei 4.17.82, p. 296; Quodvultdeus, Liber promissionum 4.13.22, p. 207; Sidonius Apollinaris, Carmen 5.389, 1:43; Vict. Vit. 1.25, 1.35, 1.38, 2.4,
2.28, 2.323, 2.36, and 3.68, pp. 12, 16, 17, 25, 34, 36, 38, and 105; Passio septem monachorum 5,
p. 109; Dracontius, Satisfactio 214, 2:186; Boethius, De differentiis topicis 2, PL 64:1189c, quoted
in pseudo-Cassiodorus, De artibus ac disciplinis liberalium litterarum 3, PL 70:1184c; CJ 1.27.2.4
(ad 534), p. 79; Marcellinus comes, Chronicon addit. s.aa. 537.3 and 543.3, pp. 105 and 107;
Jordanes, Getica 33.172, p. 74; John of Biclar, Chronicon, s.aa. 568.8, 569.11, 570.16, and 577.47,
ed. Collins, pp. 612 and 69; and Isidore, Etymologiae 9.2.118, 9.2.122, 19.23.7, and probably
18.12.5: [S]cetra scutum loreum sine ligno, quo utuntur Afri et Mauri. Coripp. Ioh. 2.29, p. 27
calls Antalas the Maurorum princeps, leader of the Moors, though Antalas was based in Byzacena,
not Mauretania.
Contra G. Camps, De Masuna a` Koceila: les destinees de la Mauretanie aux VIe au VIIe
si`ecles, Bulletin archeologique du Comite des Travaux Historiques et Scientifiques n.s. 19B (1985),
pp. 30725, who sees post-Roman Mauretania as a single, unified kingdom.
Val. Nov. 13.14 (ad 445), p. 97.
On the evolution of landlords into warlords generally, see C. R. Whittaker, Frontiers of the Roman
Empire: A Social and Economic Study (Baltimore, Md., 1994), pp. 24378, esp. 2468 and 2629.

275

Staying Roman
into their kingdom after Valentinians death in 455. At least, according
to Corippus, the career of Antalas a chieftain of humble origins from
the olive-growing region of south-western Byzacena seems to have
followed a similar trajectory, beginning with the theft of a ram at night,
followed by the gathering of a band of adherents, larger-scale sheepand cattle-rustling, the ambushing of Vandal troops in local wadi valleys,
and finally culminating with the pillaging and torching of villages on the
plains in full-blown insurrection against Vandal authority.122
Further to the west, in Mauretania Caesariensis, we hear of another
kind of violence. In Mouzaaville, Algeria (slightly over twenty kilometres
south-east of Tipasa as the crow flies, and perhaps a Vandal holding in
the late fifth century), an anonymous bishop was killed in the war of
the Moors (occisus est in bello Maurorum) and buried on 10 May 495.123
Deeper inland, in Altava, another inscription speaks of a local citizen who
died by the sword at the hands of barbarians (glaudio p[eriit] a barbaros).124
Given the date August 429 this inscription could well refer to the
Vandal invasion; but the violence in Altava seems to have been both older
and longer lasting than the Vandals passage through the region. In the
spring of 419, and then again in the spring of 449, two more residents
of the same city were killed: first, an old man by the name of Januarius,
and then a youth called Ispiacus Cerealis, the latter specifically cut down
by a sword as well.125 The evidence is suggestive rather than conclusive,
but it seems to speak of a breakdown of public order in Altava in the first
half of the fifth century.
By the opening decade of the sixth century, order had been restored
in that city by a strongman named Masuna who claimed the title king of
the Moorish and Roman peoples [or people]: rex gent(ium) [or gent(is)]
Maur(orum) et Romanor(um).126 His choice of titulature is striking for a
number of reasons, among them the fact that the use of dual ethnic titles
was extremely rare among barbarian kings in late antiquity, the only other
122
123
124
125

126

Coripp. Ioh. 3.15688, pp. 534; Moderan, Maures, pp. 31534, esp. 3267.
CIL 8.9286 = ILCV 1102 = Courtois, Vandales, p. 378, no. 94.
J. Marcillet-Jaubert, Les Inscriptions dAltava, Publications des Annales de la Faculte des Lettres
n.s. 65 (Aix-en-Provence, 1968), pp. 1012, no. 147 = Courtois, Vandales, p. 367, no. 1.
Marcillet-Jaubert, Inscriptions dAltava, pp. 945, no. 133 = CIL 8.9865 + p. 2059 = ILCV 2053
(Januarius); Marcillet-Jaubert, Inscriptions dAltava, p. 111, no. 165 = CIL 8.9866 with p. 975
= ILCV 2054 = Courtois, Vandales, p. 381, no. 120 (Ispiacus Cerealis).
Marcillet-Jaubert, Inscriptions dAltava, pp. 1267, no. 194 = CIL 8.9835 + pp. 975 and 2059
= ILCV 42 = Courtois, Vandales, p. 378, no. 95. Masuna may be identical with the Moorish
leader Massonas mentioned by Proc. BV 2.13.19, 1:4778: J. Carcopino, Un empereur
maure inconnu dapr`es une inscription latine recemment decouverte dans lAur`es, Revue des
etudes anciennes 46 (1944), p. 110 n. 3; rejected by Courtois, Vandales, pp. 3345 and Camps,
Rex Maurorum, p. 198. PLRE 3:850, s.n. Massonas concedes only the possibility of the
identification.

276

The Moorish alternative

instance before Charlemagnes eighth-century conquest of the Lombard
kingdom being the Hasding title rex Vandalorum et Alanorum, king of
the Vandals and Alans.127 In the case of Mauretania Caesariensis, Alan
Rushworth has recently and convincingly argued that as imperial authority collapsed it was replaced in Altava and elsewhere in the pre-Saharan
frontier zone by dual states in which the chieftains of formerly federated barbarian gentiles now exercised power over mixed populations of
settled Romano-African provincials and pastoralist Berber tribesmen.128
Certainly the Djedar of west-central Algeria presumably the tombs of
the ruling dynasty of one such kingdom are significantly sited along
the old imperial frontier where the complementary production-zones
of well-watered Tell and high steppe meet. Moreover, these monuments seem to represent a form of indigenous, pre-Saharan funerary
architecture adapted to a Christian, Roman provincial setting.129 Developments similar to those along the Mauretanian frontier may have been
under way in the Belezma region of southern Numidia, too, where postRoman tumuli of pre-Saharan type have been found in the area of Diana
Veteranorum.130 Further to the east, the inhabitants of the pre-desert
zone in Tripolitania appear to have been integrated into the Laguatan
confederation of Moorish tribes by the sixth century at the latest; but
here the UNESCO Libyan Valleys Survey found no evidence for the
appearance of a new warrior elite. Rather, between the third century
and the Islamic period local notables seem to have taken ever increasing responsibility for their own protection and that of their dependants,
with the result that neither Vandals nor Byzantines seem ever to have
controlled the region directly.131
How and why local communities across the interior went
Moorish indeed, what this even meant in practical terms thus seems
127

128

129
130

131

A. Gillett, Was Ethnicity Politicized in the Earliest Medieval Kingdoms?, in A. Gillett (ed.),
On Barbarian Identity: Critical Approaches to Ethnicity in the Early Middle Ages, Studies in the Early
Middle Ages 4 (Turnhout, 2002), esp. pp. 10910.
A. Rushworth, From Arzuges to Rustamids: State Formation and Regional Identity in the
Pre-Saharan Zone, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New Perspectives on Late
Antique North Africa (Aldershot, 2004), pp. 7798 and A. Rushworth, From Periphery to Core
in Late Antique Mauretania, in G. Fincham, G. Harrison, R. Holland, and L. Revell (eds.),
TRAC 99, Proceedings of the Ninth Annual Theoretical Roman Archaeology Conference, Durham 1999
(Oxford, 2000), pp. 90103; see also Camps, Rex Maurorum, pp. 2018.
Rushworth, From Arzuges to Rustamids, pp. 826.
E. Fentress, A. At Kaci, and N. Bounssair, Prospections dans le Belezma: Rapport Preliminaire,
in Actes du colloque international sur lhistoire de Setif, Bulletin darcheologie algerienne, Sup. 7
(Algiers, 1993), pp. 11011.
Mattingly, Explanations, esp. 32636; on the Laguatan, see also D. J. Mattingly, The Laguatan:
A Libyan Tribal Confederation in the Late Roman Empire, Libyan Studies 14 (1983), pp. 96108
and Moderan, Maures, pp. 123310.

277

Staying Roman
to have varied considerably from place to place. Yet two features of
Moorish rule are immediately striking. The first is how typical it seems
to have been for the new leaders of Mauretania, Numidia, and southwestern Byzacena to lay claim to Roman forms of power.132 As we have
seen, Masuna claimed to be king not only of Moors but also of Romans;
but the title rex Romanorum (king of the Romans) was widely used in
fourth-, fifth-, sixth-, seventh-, and even eighth-century textual sources
to refer to the emperor, including in panegyrics delivered in the imperial
presence.133 Masties, a Moorish leader buried at Arris in the western
Aur`es Mountains, seems explicitly to have claimed the title of imperator, and before that he certainly styled himself dux.134 The dedicatory
inscription of the earliest of the fifth-century Djedar seems to mention
another dux.135 As we have seen, in the sixth century, the anonymous
Moorish ruler of the Ucutamani claimed for himself the title of rex gentis,
while the church near Thanaramusa Castra in Algeria was begun late in
the fifth century by the Moorish prefect Iugmena and completed by
his presumed successor Zabenses (see above, section 2). In Altava, the
Masuna inscription similarly commemorates a prefect named Masgiven
as well as a procurator named Iider.136 Slightly to the north, at Albulae
(mod. An Temouchent), a fifth- or sixth-century inscription mentions
yet another prefect, this one named Safar.137
The second striking feature of Moorish rule is that, despite their
occasionally troubled relations, most of the Moorish leaders that we hear
about in the sixth century seem to have acknowledged the fundamental
legitimacy of the Byzantine empire as such and to have accepted its
presence in Africa. In a remarkable passage in his History of the Wars,
Procopius himself informs us that, like the Lazic kings of the Caucasus

132
133

134

135
136

137

In addition to the evidence cited below, see in general Camps, Rex Maurorum.
S. Fanning, Emperors and Empires in Fifth-Century Gaul, in J. Drinkwater and H. Elton
(eds.), Fifth Century Gaul: A Crisis of Identity? (Cambridge, 1992), pp. 28897, esp. 2915 and
Hays, Date and Identity, pp. 2434.
Carcopino, Un empereur maure, p. 95 = Revue archeologique, 6th ser. 24 (1945), p. 169,
no. 97 = Courtois, Vandales, p. 382, no. 132; but see also P.-A. Fevrier, Masuna et Masties,
Antiquites africaines 24 (1988), pp. 13347; Morizot, Nouvelle lecture, p. 274; and P. Morizot,

Byzantines, Coll`ege de
Masties fut-il jamais imperator?, in XXe Congr`es international des Etudes
France Sorbonne, 1925 aout 2001, Pre-Actes, vol. 2, Tables rondes (Paris, 2001), p. 262.
Camps, Rex Maurorum, pp. 2045.
Marcillet-Jaubert, Inscriptions dAltava, pp. 1267, no. 194 = CIL 8.9835 + pp. 975 and 2059 =
ILCV 42 = Courtois, Vandales, p. 378, no. 95 and pl. VI. Iider is a variation on the well-attested
name Iader, probably of Berber origin: K. Jongeling, North-African Names from Latin Sources
(Leiden, 1994), p. 59.
CIL 8.9800 = Courtois, Vandales, p. 388, no. 173. The name of the prefect seems clear, although
the reading of this inscription is particularly difficult.

278

The Moorish alternative

and to a lesser extent like the Franks in Gaul,138 Moorish chieftains
received the symbols of their authority from the Byzantines, just as they
had done from the Romans and the Vandals before them:
For those who ruled the Moors in Mauretania and Numidia and Byzacena,
sending ambassadors to Belisarius, said that [they] were slaves of the emperor
and promised to be [his] allies. And there were those who supplied their children
as hostages, asking by word that the symbols of rule be sent to them by him
according to the ancient law. For it was a law that no one ruled the Moors, even
if he were hostile to the Romans, until the emperor of the Romans gave him
the tokens of rule. Although they had already received these from the Vandals,
they did not think they held the kingdom securely. Now, these symbols are a
silver rod covered with gold leaf and a silver cap, not covering the whole head,
but held up like a crown, with silver bands on every side, and a certain white
cloak gathered into a golden brooch at the right shoulder in the fashion of a
Thessalian mantle. So Belisarius sent these things to them and presented each
one of them with much money.139

Procopius pointedly does not mention the Moors of Tripolitania in this

passage, but in a later one the historian indicates that the leaders of the
Laguatan confederation, too, requested customary gifts and symbols to
confirm the peace, this time from the Byzantine dux Tripolitaniae.140 The
Ilaguas lived beyond the frontiers of imperial control and in this case the
exchange may have been a diplomatic one; but in the western regions
of Africa, in the Byzantine period the Roman state and its representatives continued to play a critical role in legitimating Moorish power.
This included the granting of titles: in keeping with the administrative
traditions of the late empire, many Moorish leaders appear to have been
given the title of praefectus gentis.141 Moreover, some chieftains most
prominently Antalas, Cusina, and Iaudas ruled territory that was inside
the network of fortifications that Justinian and his successors erected in
the frontier zone of Byzantine Africa.142 Cusina in particular had very
138

139
141
142

Proc. BP 2.15.2, 1:215. Procopius also claims that the Frankish kings sought imperial approval
to secure their title to the Gallic throne: Proc. BG 3.33.4, 2:442. For one such example,
which pre-dates Procopius by some fifty years or more, see M. McCormick, Clovis at Tours:
Byzantine Public Ritual and the Origins of Medieval Ruler Symbolism, in E. Chrysos and A.

Schwarcz (eds.), Das Reich und die Barbaren, Veroffentlichungen des Instituts fur Osterreichische
Geschichtsforschung 29 (Vienna, 1989), pp. 15580.
140 Proc. BV 2.21.2, 1:518.
Proc. BV 1.25.38, 1:41213.
See, e.g., Coripp. Ioh. 4.545, p. 87 (Ifisdaias); see also the praefecti mentioned ibid., 5.193, p. 99
(Bruten) and ibid., 7.279, p. 154 (Bezina) and in general Brett and Fentress, Berbers, pp. 635.
See in general Moderan, Maures, pp. 313415. Moderan argues compellingly for the important
distinction in Corippus mind between these allied interior Moors and the exterior Moors of
Tripolitania, whom the poet regarded as more dangerous: ibid., pp. 63119; contrast Cameron,
Byzantine Africa, p. 41. On the fortifications of Byzantine Africa, see D. Pringle, The Defence
of Byzantine Africa from Justinian to the Arab Conquest: An Account of the Military History and

279

Staying Roman
close relations with the imperial authorities for most of his career.143
Justinians regnal year seems to have been used to date documents in
his territory, and in 548 he even received the title of magister militum
and command over Roman soldiers.144 Though the Arabic sources for
seventh-century Africa are problematic, according to the ninth-century
historian Ibn Abd al-Hakam, in the early 680s, the North African leader
Kusayla (Kocela) ibn Lamzam similarly led a great band of Romans and
Berbers (
) against the Muslim general c Uqba ibn
c
Nafi , whom they encountered at Tehouda, just to the north of the Sidi
Okba oasis at the base of the western Aur`es Mountains.145 Such coordinated military action, the widespread use of Roman titles and offices
among Moorish rulers, these same leaders acceptance of the legitimacy
of the Byzantine presence in Africa, Masunas earlier claim to be king of
the Moors and Romans, even the imperator Masties claim that he never
swore falsely nor broke faith either with Romans or with Moors (nunquam periuraui neque fidem fregi neque de Romanos neque de Mauros)146 all
of this seems to speak of a certain Romanized political identity among
Moorish rulers which we might not initially suspect from the literary
sources alone.
But the fact that Moorish kings and chieftains should express their
power in Roman terms should, perhaps, not surprise us. For a start,
along the frontiers of Roman Africa, whether an individual was a Roman

143
144

145

146

Archaeology of the African Provinces in the Sixth and Seventh Centuries, BAR International Series 99
(Oxford, 2001), pp. 171339 and 52341, maps 37.
Moderan, Maures, pp. 34650.
Documents: E. Albertini, Ostrakon byzantin de Negrine (Numidie), Cinquantenaire de la Faculte
des Lettres dAlger (18811931) (Algiers, 1932), p. 54 and, for the localization of Cusinas territory,
Moderan, Maures, pp. 33846. Magister militum: Coripp. Ioh. 8.26871, p. 175; PLRE 3:3678,
s.n. Cutzinas; Moderan, Maures, p. 349.
Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, ed. and trans. A. Gateau, in Conquete de lAfrique
du nord et de lEspagne (2nd edn; Algiers, 1947), p. 70. On his career, see Y. Moderan, Kusayla,
lAfrique et les Arabes, in C. Briand-Ponsart (ed.), Identites et culture dans lAlgerie antique,
Publications de lUniversite de Rouen 377 (Mont-Saint-Aignan, 2005), pp. 42357.
Morizot, Nouvelle lecture, p. 274 = Courtois, Vandales, p. 382, no. 132. Masties is perhaps
identical with Procopius Mastigas/Mastinas (Proc. BV 2.20.31 and 2.13.19, 1:517 and 1:478,
respectively): Morizot, Nouvelle lecture, pp. 2823 and Camps, Rex Maurorum, p. 198;
rejected by J. Carcopino, Encore Masties, lempereur maure inconnu, Revue africaine 100
(1956), pp. 33948 and implicitly by PLRE 2:734, s.n. Masties and PLRE 3:851, s.n. Mastigas.
The series of silver coins with the bust and title of Justinian on the obverse and a monogram on
the reverse attributed to Mastinas by P. Grierson, Matasuntha or Mastinas: A Reattribution,
Numismatic Chronicle (1959), pp. 11930, has long been shown by C. Morrisson, Catalogue des
monnaies byzantines de la Biblioth`eque Nationale, 2 vols. (Paris, 1970), 1:103 to be an African
issue of Justinian. Vartaia himself has consistently been identified with Procopius Ortaas (Proc.
BV 2.13.19, 2.13.289, and 2.17.8, 1:4779 and 1:5012): Carcopino, Un empereur maure,
pp. 11112 and Carcopino, Encore Masties, p. 341; Camps, Rex Maurorum, p. 198; Morizot,
Novelle lecture, p. 280; PLRE 3:9578, s.n. Ortaias.

280

The Moorish alternative

or a Moor probably depended very much on who was asking. One of
the poems in the Latin Anthology, reflecting the literary prejudices and
cultural elitism of the sixth-century Carthaginian upper crust, ruthlessly
mocks the poetic pretensions of a Tripolitanian provincial. In the lemma
of his critics riposte, the Tripolitanian is referred to as an Arzugitanus
poeta, a poet of the Arzuges Moors, or at least an inhabitant of Arzugitana,
the region bordering the Chott el-Djerid on the pre-desert limes.147 As
Averil Cameron has observed, The man was hardly really a Berber,
but to our author might as well have been.148 But the disdain of the
cosmopolitan sophisticate for the verses of this rustic poet hints at some
of the complexities of cultural identity in the borderlands of the remote
south and west complexities which are more fully illuminated in the
Albertini Tablets.
The first of these, dated 17 September 493, records the dowry of a
wealthy woman named Geminia Januarilla, who lived along the southwesternmost fringes of the Vandal kingdom. The language of the document is Latin; the act of her endowment properly observes the forms
of Roman law; her name, and the name of her husband Julian, are
thoroughly Roman. Christian Courtois and Susan Raven have both
remarked upon Januarillas wholly Berber taste in clothing,149 but the
goods with which the young bride was endowed do not seem to have
been remarkably out of place for a woman of notable status from the
late Roman countryside. They consisted primarily of clothes, jewellery,
bedding, and other furnishings: a wide-sleeved tunic (dalmatica), a mafors
or garment which covers the head and shoulders, a kerchief (orarium),
ribbons, sandals, slippers, bracelets, rings, cockle-shell earrings, linen,
cushions, ox-hides, and some kind of loom for weaving wool.150 Again,
a comparison with the clothing depicted in the late Roman- and Vandalera tomb mosaics from Tabarka is illuminating, for in that coastal city
matrons are typically shown wearing a long tunic and a veil. In Tabarka
147
148
149

150

AL 120, pp. 99100. Moderan, Maures, pp. 36474; Desanges, Catalogue des tribus, pp. 7880,
s.v. Arzuges; Cameron, Byzantine Africa, pp. 312 and esp. 32 n. 13.
Cameron, Byzantine Africa, p. 32 n. 13.
S. Raven, Rome in Africa (3rd edn; London, 1993), p. 200. The same point is made by Courtois,
Les Hommes et les choses, Tablettes Albertini, p. 207: Si lon juge par son trousseau de noces, la
petite Geminia Ianuarilla . . . devait ressembler davantage aux e ternelles e pousees de la Kabylie
ou de lAur`es qu`a une jeune Romaine de province.
Tablettes Albertini, act 1, 1:215. OCD, pp. 4978 and 148990, s.vv. dress, and textile production (respectively), and U. Scharf, Straenkleidung der romischen Frau, Geschichte und ihre
Hilfswissenschaften ser. 3, 585 (Frankfurt am Main, 1994), pp. 5561 (dalmatica) and 7682
(mafors); on the latter, see also Le Martyre de Pionios, pretre de Smyrne, ed. and trans. L. Robert,
with G. W. Bowersock and C. P. Jones (Washington, DC, 1994), p. 94 n. 6: une esp`ece de
tunique longue avec capuchin. On dowries in general in late antiquity, see Bowersock, Brown,
and Grabar, Late Antiquity, p. 419 s.v. Dowry.

281

Staying Roman
the latter was probably a symbol of a womans married status, which
in turn may suggest that Januarillas dowry included not just clothes in
general, but specifically the kind of clothing that was felt to be appropriate to a Roman matron.151 Indeed, if anything, it is the ox-hides
presumably intended to be used as mattresses or seat-covers, and as such
reminiscent of Procopius claims that wealthy Moors slept on sheepskin
bedding that seem most strikingly to bespeak a rural taste in material
goods closer to that of the tribeswomen of the African interior than to
the refined elegance of the Roman and Vandal ladies of Carthage.152
However, rather than viewing Januarillas predilections and physical possessions as either wholly Berber or wholly Roman, it is perhaps more
satisfying to see them as reflective of a frontier taste in material goods,
with all the blending and overlapping of cultural traditions that implies.
Nor was Januarilla a poor woman. Her jewellery alone would have
spoken of her elevated social status, and the cockle-shell earrings in particular must have been quite precious, as they would have had to be
imported a considerable distance from the coast.153 Moreover, comparing the value of the goods with which she was endowed to the prices
found in the tablets as a whole, Januarillas dowry was worth about 750
olive trees.154 Though the value of the goods involved may have been
artificially inflated, the new bride must have come from a family that, for
the region, controlled almost unimaginable wealth. Her family must also
have enjoyed elite status, at least locally. The man who held dominium
over the estate around which most of the Albertini Tablets focus was
named Flavius Geminius Catulinus, flamen perpetuus, and almost without
exception it was members of the gens Geminia Januarillas family
who were recorded in these documents as buying up local property, both
human and landed.
Like Januarillas dowry, the deeds of sale observe the forms of Roman
law.155 The inhabitants of this region, whose lives we glimpse briefly in
151
152

153
154

155

Downs, Christian Tomb Mosaics, pp. 2013.

The comparative evidence of the Cairo geniza suggest that the constituent elements of dowries
remained largely unchanged over time, and typically consisted of jewellery, clothing, bedding,
and household goods: S. D. Goitein, A Mediterranean Society: The Jewish Communities of the Arab
World as Portrayed in the Documents of the Cairo Geniza, 6 vols. (Berkeley, Calif., 196793), esp.
3:125 and 3:128. Thanks to Jennifer Ball for a very useful conversation about clothing and
dowries in the late Roman and Byzantine period.
See in general A. Stout, Jewelry as a Symbol of Status in the Roman Empire, in J. L. Sebesta
and L. Bonfante (eds.), The World of Roman Costume (Madison, Wis., 1994), pp. 77100.
Courtois, Le Probl`eme des prix, Tablettes Albertini, pp. 2035. On this same problem, see also
P. Grierson, The Tablettes Albertini and the Value of the Solidus in the Fifth and Sixth Centuries
ad JRS 49 (1959), pp. 7380. In many parts of the medieval Islamic world it was customary to
inflate highly the value of the goods listed in a dowry: Goitein, Mediterranean Society, 3:1268.
H. Wessel, Das Recht der Tablettes Albertini (Berlin, 2003).

282

The Moorish alternative

these documents, generally bore Latin names, often with a distinctively
African flavour, from Geminius Catulinus himself to Geminius Felix to
the smallholders Adeodata, Benenatus, Cresconius, Donatianus, Quodvultdeus, Saturninus, and Victor. Moreover, Leslie Dossey has convincingly argued that these remarkable deeds of sale reveal a rural community
in which the estate was becoming a city for its tenants.156 But in the
Tuletianos community, as in the heartland of the Vandal kingdom, we
must also be attentive to potential distinctions between legal form and
social practices. Thus, for example, C. R. Whittaker has argued that
the transactions in which these individuals engaged, while conducted
in Roman legal terms, may in fact have been governed by local, customary practice and customary property relations, and moreover that
the social structure of the Fundus Tuletianos, as revealed in the Albertini
Tablets, may well be a closed, familial group or groups, presided over
by . . . some sort of cad or chief.157 Along the Moorish frontier, as elsewhere in the late antique North African countryside, Roman and local
identities were probably not mutually exclusive.
In any case, the pre-desert fringes of the Roman world had never presented an insurmountable barrier to cultural inter-penetration. Inscriptions from the portus of Zarai and Lambaesis and from the market at Casae
reveal that in the imperial period the nomadic pastoralists living in the
desert regions beyond the area of immediate Roman control summered
their herds in the Algerian and Tunisian Tell, within the area regulated
by a string of Roman fortifications. The nomads would provide harvest
labour for the sedentary farmers of the region, and the two groups would
exchange food and produce.158 As mentioned above, Augustine tells us
that the Romans appointed prefects to rule the Moors living along the
Roman frontier.159 And a fourth-century tomb from Ghirza in Libya
depicts a local chief seated on a stool similar to the sella curulis used by
Roman magistrates.160 As with the Vandals and in the early Germanic
156
157
158

159
160

L. Dossey, Peasant and Empire in Christian North Africa, Transformation of the Classical Heritage
47 (Berkeley, Calif., 2010), pp. 11920.
C. R. Whittaker, Land and Labour in North Africa, Klio 60 (1978), pp. 35661; all three
quotations are from p. 360.
Ibid., pp. 335 and 34450, with reference to CIL 8.4508 and R. Cagnat, A New Roman
Customs List, JRS 4 (1914), pp. 1426; see also E. W. Bovill, The Golden Trade of the Moors:
West African Kingdoms in the Fourteenth Century (London, 1958), p. 11 and in general H. Elton,
Defining Romans, Barbarians, and the Roman Frontier, in R. Mathisen and H. Sivan (eds.),
Shifting Frontiers in Late Antiquity (Aldershot, 1996), pp. 12635, esp. 134.
Augustine, Ep. 199.12.46, CSEL 57:2845.
Brogan and Smith, Ghirza, pp. 153 and 231 and pl. 78. O. Wanscher, Sella Curulis: The Folding
Stool, an Ancient Symbol of Dignity (Copenhagen, 1980), pp. 12190 discusses the curule chair in
the Roman and late Roman world; ibid., pp. 1912 and 20420 and P. E. Schramm, Herrschaftszeichen und Staatssymbolik: Beitrage zu ihrer Geschichte vom dritten bis zum sechzehnten Jahrhundert,

283

Staying Roman
kingdoms of Europe, Roman symbols provided a vocabulary of power
for the peoples living along the frontiers of Roman Africa, regardless
of whether or not they saw themselves as subject to Roman political
control.
But Moorish cultural identity is only part of the equation. The
Romano-African inhabitants of what had long been the provincial hinterland form the other part; for whatever the vicissitudes of power in
the region, local populations remained largely the same. At the top of
provincial society was a highly Romanized element, and for the fifth
century through the seventh century their lives are attested almost exclusively through a handful of inscriptions, augmented by a few scattered
references in the textual sources. Even when used to complement one
another, of course, the textual and epigraphic evidence is likely to show
any given population at its most Roman. But the fact that elite urban
populations in western Numidia and the Mauretanias continued to identify strongly enough with the social and cultural forms of the Roman
empire to produce such evidence, even at much diminished levels, is
surely significant. These were the very populations in control of whom
the Moorish kings of the time would have found themselves, and with
whom they would have found it necessary to reach a measure of accommodation if they were to be able to rule. In late antique Numidia and
Mauretania, the adoption of the Roman vocabulary of power made a
conscious and public statement about continuity aimed precisely at this
Romano-African population.
In the thirteen years between 442 and 455 that is to say, between
the conclusion of a treaty between the Vandal kingdom and the western
empire returning Numidia and Mauretania to imperial control on the
one hand and the death of the emperor Valentinian III on the other
the retroceded provinces remained integrated into the empire as a whole.
Though the use of consular dates was rare in Africa (see above, Chapter 3.2.3), at least three inscriptions from mid fifth-century Numidia
recognized the succession of consuls and the authority of the Roman
emperors.161 Tax revenues continued to be collected in what was left of

161

Schriften der Monumenta Germaniae historica 13, 3 vols. (Stuttgart, 19546), 1:31720 discuss
the relationship between the sella curulis and the thrones of the early Germanic kingdoms.
Courtois, Vandales, p. 371, no. 35 = AE (1967), 200, no. 595 (Cuicul [mod. Djemila, Algeria],
2 Mar. 452); ILCV 2104 = Courtois, Vandales, p. 370, no. 33 = Y. Duval, Loca sanctorum Africae:

francaise de Rome
le culte des martyrs en Afrique du IVe au VIIe si`ecle, 2 vols., Collection de lEcole
(Paris, 1982), 1:30610 no. 146 (Setif, 3 Aug. 452); Courtois, Vandales, p. 371, no. 40 = Revue
archeologique, 5th ser. 20 (1924), p. 387, no. 58 (Cuicul, 15 Oct. 454). See also Cassius Felix,
De medicina, ed. V. Rose (Leipzig, 1879), p. viii: sub ardebre et asclepio consulibus (Paris BN
Lat. 6114), i.e., under the consuls Ardabur and Calepius (ad 447); PLRE 2:461, s.n. Cassius Felix 13. See also, perhaps, ILAlg. 2.375 (Rusicade): CONSE////ACIVC, reconstructed

284

The Moorish alternative

Roman Africa, although in the spring of 445 a delegation to the emperor
from Numidia and Mauretania Sitifensis secured relief from payments. As
we have seen, Valentinian continued to legislate both for the refugee population and for the remaining African provinces, passing laws that relaxed
restrictions on legal procedures for Africans, exempted them from certain
payments, and regulated the terms under which they had to pay back
loans.162 The mid fifth century furthermore saw the entrenchment of
the Romano-African landed gentry in Numidia and the Mauretanias.
In July of 451, Valentinian III ordered that thirteen thousand centuriae
of deserted land in Numidia be given to the dignitaries and landholders
whom the Vandal invasion had turned into refugees. Similarly, imperial
estates in Mauretania Sitifensis and Caesariensis were to be leased to
those magnates from the proconsular province and Byzacena whom the
Vandals had deprived of their patrimonies.163
It is thus no surprise to find that the dignities of the late Roman social
system continued to carry prestige in the North African provinces, at
least down to the mid fifth century and the loss of Roman control in the
region. In their epitaphs, the Numidian aristocrats Rusticula and Jobina
were celebrated as clarissimae feminae, or women of senatorial status.164
So too the twenty-five-year-old Cypriana, honored as a clarissima et
spectabilis femina in an inscription found on the outskirts of late antique
Setif. Similarly, Romanilla and Emerita, both from Sertei in Mauretania
Sitifensis, were each independently commemorated as honestae feminae.165
Finally, the embassy to Valentinian which secured tax relief for the African
provinces was led by men with the thoroughly Roman names, titles, and

162
163
164

165

as cons(ulatu) E . . . aci, v(iri) c(larissimi) (a revision of CIL 8.19914: consecravit), i.e., perhaps
in the consulate of E(uthar)acus (ad 519): PLRE 2:621, s.n. (I)ob(in)a; not included in
R. Bagnall, A. Cameron, S. Schwartz, and K. Worp, Consuls of the Later Roman Empire, Philological Monographs of the American Philological Association 36 (Atlanta, Ga., 1987), p. 572, s.a.
519. Anno provinciae dates were the norm in Mauretania Sitifensis: e.g., near Setif: CIL 8.20410 =
ILCV 189 = Courtois, Vandales, p. 371, no. 41. Castellum Thib . . . (An Melloul): CIL 8.8708
= ILCV 2966 = Courtois, Vandales, p. 370, no. 25. Guellal: Courtois, Vandales, p. 372, no. 42.
Sertei (Kherbet Guidra): CIL 8.20643 = ILCV 332A = Courtois, Vandales, p. 369, no. 20; and
CIL 8.20644 = ILCV 332B = Courtois, Vandales, p. 373, no. 53. Kherbet-el-Ma-Abiod (near
Bordj-Mamra): ILCV 2069.
Val. Nov. 1.3 (ad 450), 2.3 (ad 443), 6.3 (ad 444), 12 (ad 443), 13 (ad 445), and 34
(ad 451), pp. 747, 789, 845, 937, and 1401.
Val. Nov. 34 (ad 451), pp. 1401.
Courtois, Vandales, p. 371, no. 35 and ILAlg. 2.375, which may perhaps date to ad 519 (see
above, n. 161). See also CIL 8.8345 = ILCV 109c: Fl. / Paulus / u. c. ex trib. / uotum / soluit;
perhaps fifth century in date: PLRE 2:854, s.n. Fl. Paulus 32.
Cypriana: CIL 8.20410 = ILCV 189 = Courtois, Vandales, p. 371, no. 41 (19 Oct. 454).
Romanilla: CIL 8.9898 = ILCV 2862C adn. = Courtois, Vandales, p. 369, no. 20 (6 July 444?).
Emerita: CIL 8.20644 = ILCV 332B = Courtois, Vandales, p. 373, no. 53 (28 July 467).

285

Staying Roman
dignities of Palladius comes vir spectabilis, Maximinus vir laudabilis (a civil
priest), and an unnamed tribunus vir clarissimus.166
In the 440s and 450s, north-western Africa also maintained important
religious ties with the rest of the empire. In c.446, for example, Pope
Leo I (ad 44061) sent the priest Potentius from Rome to Mauretania
Caesariensis to undertake a disciplinary inquiry into the irregularities of
ordination and practice in the province.167 The bishop David presumably a North African with his see at the unidentified Tadamatensis was
later to act as the intermediary between the bishop of Rome and the rest
of the Mauretanian episcopate.168 The same period saw what appears to
have been a new interest in Mauretania Sitifensis in relics of specifically
Roman martyrs.169 In 452, relics of St Laurence were deposited in the
city of Setif itself.170 Within twenty-two years, relics of the same martyr
had been deposited at two other sites within a fifty-kilometre radius,
at Beni Fouda and Kherbet el Ma el Abiod, Algeria. In both instances,
Laurence was now associated with the Roman martyr Hippolytus; at
Kherbet el Ma el Abiod, the saint was linked too with Euphemia of
Chalcedon, a certain Minna (either a local martyr or Menas of Alexandria), and a secondary relic of the True Cross.171 In the later fifth and
sixth centuries, epigraphic attestations of Laurence and Hippolytus cult
continued to spread eastward into Numidia and Byzacena, frequently in
association with other martyrs whose names suggest that in this region
the cult of saints was becoming increasingly international in character.172
With the death of Valentinian in 455, the Vandals retook Numidia and
Mauretania Sitifensis, as well as a small amount of territory in Mauretania
166
167

168
169

170

171
172

Val. Nov. 13 (ad 445), p. 95.

Leo I, Ep. 12.1, PL 54:6467. Leos legate may be the same Potentius mentioned in an
inscription from the basilica of St Salsa in Tipasa, in Mauretania Caesariensis: CIL 8.20914 =
ILCV 1824 (where the association is made in the apparatus) = Courtois, Vandales, p. 382,
no. 131, with PCBE 1:898, s.n. Potentius 2.
Leo I, Ep. 12.9, cols. 6534; see also Notitia, M. Caes. 105, p. 131. The identification is made
by PCBE 1:26970, s.n. David.
See in general Duval, Loca sanctorum, 2:64850 and W. H. C. Frend, From Donatist Opposition
to Byzantine Loyalism: The Cult of Martyrs in North Africa 350650, in Merrills, Vandals,
Romans, and Berbers, pp. 25969 at pp. 2645.
CIL 8.8630 = ILCV 2104 = Courtois, Vandales, p. 370, no. 33 = AE (1967), 2089, no. 640 =
Duval, Loca sanctorum, 1:30610, no. 146 (Setif). See also Duval, Loca sanctorum, 1:31012, no. 147
(Setif), apparently a sixth-century dedication. The cult of Laurence was not new to Africa in
452: Augustine, Sermones 3025, PL 38:1385400; see also ibid., 296.5.6, PL 38:1355.
Duval, Loca sanctorum, 1:2847, no. 135 (Kherbet el Ma el Abiod) and ibid., 1:299301, no. 141
(Beni Fouda).
Duval, Loca sanctorum, 1:1024, no. 49 (Thala); ibid., 1:1714, no. 83 (An Zirara); ibid., 1:259
64, no. 126 (Henchir Akrib); ibid., 1:2679, no. 128 (Henchir Akrib); and ibid., 1:31012, no.
147 (Setif). Sixtus (no. 49) is presumably the bishop of Rome; Julian (nos. 126 and 147) may
be either the Antiochene martyr or a local saint; Nabor (no. 147) may be either the Milanese
martyr or a local saint.

286

The Moorish alternative

Caesariensis, perhaps limited to the major port cities.173 The writ of the
Vandal kings does seem to have run here, at least for a time; but in the predesert, and in some places further into the agricultural highlands, effective
control seems to have remained in Moorish hands. Thus, for example,
we have seen that when the Vandal king Huneric exiled large numbers of
the Nicene clergy to the Chott el Hodna region of Mauretania Sitifensis,
the clerics were handed off to Moorish guards as far east as the cities
of Sicca and Lares in Africa Proconsularis (Chapter 2.3).174 In the same
kings reign, the Moors of the Aur`es Mountains revolted against Vandal
rule, presumably under the dux (and later imperator) Masties.175 Hunerics
son and eventual successor Hilderic seems later to have exercised some
kind of dominion over the city of Setif, but in the early sixth century
a leader named Ortaas ruled Mauretania beyond the Aur`es, seemingly
southern Mauretania Sitifensis.176 This situation does not seem to have
changed substantially with the arrival of the Byzantines. Towns were
fortified and garrisons established in the territory of Mauretania Sitifensis,
Numidia, and Byzacena between the pre-desert and the sea, but Moorish
leaders still wielded considerable influence in the western provinces.177
However it was that the populations of this region imagined their
identities from the mid fifth century onwards, the economic and social
ties that bound them to the wider Mediterranean world seem to have
persisted, though seemingly on a diminished level. Fine red slip tableware
continued to reach the ports of Tipasa, Caesarea, and Septem into at
least the sixth century.178 These cities remained important enough for
the Vandals and then the Byzantines to take them over and, as we have
seen, the movement of refugees and exiles in the fifth and sixth centuries
provides further evidence for their contacts with points as far apart as
Spain and Constantinople (see above, Chapter 2). Further inland, in the
plains and mountain valleys around Caesarea, the cultivation of olives for
oil seems to have become less profitable at the end of antiquity (a pattern
173

174
175
176
177

178

Vict. Vit. 1.13, p. 7; Y. Moderan, Les Fronti`eres mouvantes du royaume vandale, in C. Lepelley
and X. Dupuis (eds.), Fronti`eres et limites geographiques de lAfrique du Nord antique (Paris, 1999),
pp. 2578.
Vict. Vit. 2.268, pp. 334.
Proc. BV 1.8.5, 1:346; Moderan, Maures, pp. 398413. See also (on the Moors of the Aur`es)
ibid., pp. 38398 and (on Moorish rebellions against Vandal authority) ibid., pp. 54161.
Concilium Carthaginense a. 525, p. 260, ll. 2259; Proc. BV 2.13.278, 1:479; Moderan, Fronti`eres
mouvantes, pp. 2589; and Moderan, Maures, pp. 3801.
Y. Duval, La Mauretanie sitifienne a` lepoque byzantine, Latomus 29 (1970), pp. 15761; J.-P.
Laporte, Zabi, Friki: notes sur la Mauretanie et la Numidie de Justinien, Antiquite tardive 10
(2003), pp. 15167. On Moorish leaders and the Byzantine fortifications, see also above, n. 142.
J. W. Hayes, Late Roman Pottery (London, 1972), maps 712; P. Reynolds, Trade in the Western
Mediterranean, ad 400700: The Ceramic Evidence, BAR International Series 604 (Oxford, 1995),
p. 24.

287

Staying Roman
that we have also observed elsewhere in the North African countryside),
but even so a handful of villas and farms in the citys hinterland continued
to import ARS into the sixth and possibly even the seventh centuries,
testifying both to a degree of local prosperity and to a measure of
integration into the urban economy.179 A similar pattern holds slightly
to the west, in the region between the old Roman fortress of Tigava and
the sea.180 East of Caesarea, in the Belezma region of southern Numidia,
fifth-century ARS has been found on the farms and settlements around
Diana Veteranorum, including on two fortified structures with associated
hamlets in the mountains south of the city, though much of this fine
ware was manufactured locally.181 We catch a hint of contact between
inland Numidia and the city of Rome, though, in a metrical inscription
that the priest Probantius erected in the sanctuary of a church dedicated
to Sts Peter and Paul at An Ghorab in the Aur`es Mountains. Probantius
inscription, put up some time in the later fifth or sixth centuries, imitates
one that Pope Sixtus III (ad 43240) placed in St Peter-in-Chains.182
In April 493, a certain Sabas, presbyter of the holy Roman church, was
buried in Manacor on the island of Mallorca. The use of the Mauretanian
provincial era in his epitaph suggests that he was probably an African.183
It becomes harder to trace such long-distance connections into the sixth
and seventh centuries, though they seem to have continued on some
level, for, as Moderan has pointed out, Mauretanian bishops attended
church councils in Carthage in 525 and 646.184
Much further east, in a part of inland Tripolitania that always lay
beyond the control of the Vandal kingdom and its Byzantine successor,
179

180
181
182

183

184

francaise
P. Leveau, Caesarea de Mauretanie: une ville romaine et ses campagnes, Collection de lEcole
de Rome 70 (Rome, 1984), esp. pp. 2945, no. 87 and p. 298, fig. 105; pp. 3367, no. 163 and
p. 338, fig. 143; pp. 3446, no. 176; pp. 35760, no. 188 and p. 359, fig. 169; pp. 3658, no. 196
and p. 367, fig. 179; p. 368, no. 197 and p. 370, fig. 182; pp. 3856, no. 220; and p. 389, no. 224.
A hoard of gold coins, the latest a solidus of Marcian (ad 4507), was also found at Sidi-Amar:
ibid., p. 457.
P. Leveau, Recherches historiques sur une region montagneuse de Mauretanie Cesarienne: des

Tigava Castra a` la mer, Melanges de lEcole

francaise de Rome, Antiquite 89 (1977), pp. 257311.
Fentress, At Kaci, and Bounssair, Prospections dans le Belezma, pp. 10727.
CIL 8.107078 + 17615 = Courtois, Vandales, p. 387, no. 165; PCBE 1:922, s.n. Probantius
2. In antiquity the economy of this region was focused at least in part on the production of
olive oil, presumably for export: P. Morizot, LAur`es et lolivier, Antiquites africaines 29 (1993),
pp. 177240.
J. Vives, Inscripciones cristianas de la Espana romana y visigoda, Monumenta Hispaniae sacra, serie
patrstica 2 (2nd edn; Barcelona, 1969), p. 79, no. 268 = Handley, Dying on Foreign Shores, no.
78. Thanks to Mark Handley for bringing this reference to my attention. See also AE (1996),
2634, no. 814 = Handley, Dying on Foreign Shores, no. 307 (a woman from Cuicul who died
in Sardinia) and D. Noy, Jewish Inscriptions of Western Europe, vol. 1, Italy (Excluding the City of
Rome), Spain and Gaul (Cambridge, 1993), pp. 501, no. 31 = Handley, Dying on Foreign Shores,
no. 292 (a Jewish Mauretanian senior civis in Naples), both fifth century.
Moderan, Maures, p. 388.

288

The Moorish alternative

contact between the Mediterranean littoral and the pre-desert wadi valleys
was also sustained into the seventh century. In general, though, the trend
in this region between the late Roman and the Islamic period was towards
lessened importation of fine wares from the coastal zone, decreasing
investment in the production of olive oil for the export economy, and
the concentration of settlement in fortified farms or gsur. Yet society
here remained both complex and intricately integrated, at least on a local
level. In Wadi Umm el-Kharab, for example, the floodwater farming
system seems to have been under unified management in the sixth and
seventh centuries, with farmers cultivating barley, durum wheat, lentils,
peas, grapes, figs, and pistachios, and herders raising sheep and goats at
levels that supported one of the highest populations ever achieved in the
pre-desert.185
In the far west, too, we have so far only been able to glimpse a
few traces of sustained interregional contacts. On the Atlantic coast of
Mauretania Tingitana, the city of Lixus continued both to produce salt
fish and garum and to export them into the Mediterranean into the
600s.186 Further to the south, in Volubilis, a certain Julia Rogatiana died
in 655 at the age of seventy-six and was commemorated there by her sons
and nephews, who tell us that she originally came from Altava, about
425 kilometres (overland) to the east as the crow flies.187 After the fifth
century, though, we lose sight of the diagnostic pottery through which to
assess the economic contacts of the cities and dispersed rural settlements
of this same region, the Oued Sebou valley.188
Remarkably, though, from the later fifth century onwards, our most
plentiful epigraphic evidence from the African hinterland comes from the
westernmost regions of Mauretania Caesariensis. Julia Rogatianas native
city of Altava preserves perhaps the richest collection of late antique
inscriptions in this region, though the cities of Pomeria (mod. Tlemcen), Numerus Syrorum (mod. Lalla Maghnia), Albulae, Ala Miliaria
(mod. Benian), and the areas around El Guetna and Tiaret are also
185

186
187
188

D. Mattingly, RomanoLibyan Settlement: Typology and Chronology, in Barker and Mattingly, Farming the Desert, 1:11158; Mattingly, RomanoLibyan Settlement: Site Distribution
and Trends, ibid., 1:15990; D. D. Gilbertson and C. O. Hunt, RomanoLibyan Agriculture:
Walls and Floodwater Farming, ibid., 1:191225; M. van der Veen, A. Grant, and G. Barker,
RomanoLibyan Agriculture: Crops and Animals, ibid., 1:22763; G. Barker, D. D. Gilbertson, C. O. Hunt, and D. Mattingly, RomanoLibyan Agriculture: Integrated Models, ibid.,
1:26590. See also I. Sjostrom, Tripolitania in Transition: Late Roman to Early Islamic Settlement
(Aldershot, 1993).
McCormick, Origins, pp. 58 and 634 n. 67.
M. Euzennat and J. Marion (eds.), Inscriptions antiques du Maroc, vol. 2, Inscriptions latines (Paris,
1982), pp. 3534, no. 608 = Handley, Dying on Foreign Shores, no. 427.
R. Rebuffat, Recherches sur le bassin du Sebou (Maroc), Comptes rendus de seances de lAcademie
des Inscriptions et Belles-Lettres (1986), p. 651.

289

Staying Roman
well represented in the surviving epigraphic corpus. This would seem
to have been imperial territory as late as the first quarter of the fifth
century: at least, at some point during the joint reign of Honorius and
Theodosius II (ad 40823), the citizens of Altava erected an inscription
dedicated to the prosperity of their city and the safety of the emperors.189
Thereafter, although they were never reintegrated into the structures of
the empire, and though their funerary inscriptions do not record claims
to Roman dignities, the urban populations of western Mauretania Caesariensis seem to have remained Roman in their cultural outlook. These
city dwellers continued to employ the Roman provincial dating system,
even if only by default; they still used Latin, at least as an epigraphic
language; and their name-stock long retained a distinctively RomanoAfrican tinge, in which Latin names like Donatus, Emeritus, Januarius,
Julius, Reparatus, Rogatus, and Secundus appeared comfortably alongside Punic ones like Saposus and Romano-Berber ones like Aurelia
Cursen, Aurelia Tifalis, Julia Getula, and Valerius Zabullus.190 All of
these trends held into the sixth century and beyond.191
189

190

191

Marcillet-Jaubert, Inscriptions dAltava, pp. 889, no. 122; but see also B. H. Warmington, The
North African Provinces from Diocletian to the Vandal Conquest (Cambridge, 1954), pp. 702 and
map 3; and J. Carcopino, Le Maroc antique (8th edn; Paris, 1943), pp. 23344 who argue for
imperial withdrawal from the region.
See, e.g., Donatus: Courtois, Vandales, p. 369, no. 21; ibid., p. 377, no. 87; CIL 8.21544 =
ILCV 3620 = Courtois, Vandales, p. 378, no. 97; CIL 8.21571 = Courtois, Vandales, p. 381,
no. 117; CIL 8.9869 = ILCV 3666 = Marcillet-Jaubert, Inscriptions dAltava, p. 132, no. 204.
Emeritus/-a: CIL 8.9911 = ILCV 3677 = Courtois, Vandales, p. 374, no. 61; CIL 8.21688
= ILCV 3271 = Courtois, Vandales, p. 374, no. 63; CIL 8.21676 = ILCV 3275 = Courtois,
Vandales, p. 377, no. 88; CIL 8.9949 + p. 2065 = ILCV 3679. Januarius: CIL 8.21689 =
ILCV 3276 = Courtois, Vandales, p. 377, no. 90; CIL 8.9271 = ILCV 1829b = Courtois,
Vandales, p. 381, no. 119b; Julius/-a: Courtois, Vandales, p. 372, no. 43; CIL 8.21782 = ILCV
3667A = Courtois, Vandales, p. 372, no. 47; CIL 8.21675 = ILCV 3272B = Courtois, Vandales,
p. 378, no. 99; Euzennat and Marion, Inscriptions antiques du Maroc 2:31718, no. 506; ibid.,
pp. 3501, no. 603. Reparatus: CIL 8.9709 = ILCV 1105 = Courtois, Vandales, p. 375, no. 72;
CIL 8.21591 = ILCV 2815A adn. = Courtois, Vandales, p. 381, no. 124. Rogatus: Courtois,
Vandales, p. 372, no. 48; CIL 8.9944 = ILCV 3668 = Courtois, Vandales, p. 379, no. 100; CIL
8.21582 = ILCV 3935A = Courtois, Vandales, p. 380, no. 114; see also Rogatiana: Euzennat
and Marion, Inscriptions antiques du Maroc 2:3534, no. 608. Secundus/-a: Courtois, Vandales, p.
369, no. 16; CIL 8.21681 = ILCV 2866A = Courtois, Vandales, p. 374, no. 62; CIL 8.21797
= ILCV 3936 = Courtois, Vandales, p. 376, no. 81; CIL 8.9934 + p. 2065 = ILCV 3662A.
Saposus: CIL 8.9713 = ILCV 2850a = Courtois, Vandales, p. 373, no. 57. Aurelia Cursen: AE
(1985), 279, no. 988. Aurelia Tifalis: CIL 8.21729 = ILCV 2862D = Courtois, Vandales, p. 370,
no. 29 = Marcillet-Jaubert, Inscriptions dAltava, pp. 11112, no. 166. Julia Getula: CIL 8.9930
= ILCV 3670. Valerius Zabullus: CIL 8.21792 = ILCV 3676 = Courtois, Vandales, p. 372, no.
45. On the Punic and Berber names, see Jongeling, North-African Names, p. 39 (Cursen), ibid.,
pp. 48 and 50 (Getula), ibid., p. 128 (Saposus), ibid., p. 143 (Tifzalis), and ibid., pp. 1534
(Zabullus).
The latest inscriptions in Mauretania Caesariensis are all from Pomaria (Tlemcen); the seventhcentury inscriptions are: CIL 8.9953 + p. 2065 = ILCV 3673 (ad 629/30); CIL 8.9949 +
p. 2065 = ILCV 3679 (ad 636); and CIL 8.9935 = ILCV 3675 A adn. (ad 651). In Mauretania

290

The Moorish alternative

The epigraphic evidence also suggests that the urban elite throughout the Mauretanias and Numidia shared a common Christian religious
culture in the late Roman and post-imperial period. Crosses, alphas and
omegas, and Chi-Rho monograms litter the epitaphs.192 In 406, the
parents of a deceased infant buried at Castellum Tingitanum had noted
the young boys deposition in the basilica of Sts Peter and Paul and in
the name of God and Christ.193 Two years later, the magister Umbrius
Felix similarly invoked the name of God and Christ in Tebessa in the
fulfilment of a vow.194 Numerous inscriptions testify to the strength of
the cult of saints in Numidia and the Mauretanias in the fifth, sixth,
and seventh centuries.195 Moreover, bishoprics dotted the countryside
of north-western Africa in remarkable abundance: we know the names
of no fewer than 120 bishops from Mauretania Caesariensis, forty-two
bishops from Mauretania Sitifensis, and another 123 from Numidia in the
year 484 alone.196 Down to the sixth and in some places even the seventh

192

193
194
195

196

only inscriptions from Caesarea and Tipasa were not dated according to the Roman provincial
year. Caesarea in Mauretania: CIL 8.21424 = Courtois, Vandales, p. 387, no. 169; CIL 8.9586
+ p. 1984 = ILCV 1179. Tipasa: S. Lancel, Tipasitana II: une inscription martyrologique de
Tipasa, Bulletin darcheologie algerienne 2 (19667), pp. 2519 = AE (1967), 211, no. 646 = AE
(196970), 219, no. 731 = Duval, Loca sanctorum, 1:37780, no. 178 (Tipasa); CIL 8.20909 =
Courtois, Vandales, p. 387, no. 170; Courtois, Vandales, p. 387, no. 171; CIL 8.20914 = ILCV
1824 = Courtois, Vandales, p. 382, no. 131. If Roman provincial dates were used in CIL 8.9800 =
Courtois, Vandales, p. 388, no. 173, a fragmentary inscription from Albulae (An Temouchent),
they are not preserved; but this is the exception rather than the rule for this city (see below). For
the widespread use of the anno provinciae system in Mauretania, see ILCV 3:2702, index 6D.
Chi-Rho monogram: CIL 8.9869 = ILCV 3666 = Marcillet-Jaubert, Inscriptions dAltava,
p. 132, no. 204; CIL 8.9866 = ILCV 2054 = Courtois, Vandales, p. 381, no. 120 = MarcilletJaubert, Inscriptions dAltava, p. 111, no. 165; CIL 8.21774 = ILCV 2862C adn. = Courtois,
Vandales, p. 369, no. 23 = Marcillet-Jaubert, Inscriptions dAltava, p. 109, no. 161. Alpha and
Omega: Courtois, Vandales, p. 387, no. 171. Chi-Rho monogram flanked by alpha and omega:
CIL 8.20909 = Courtois, Vandales, p. 387, no. 170; CIL 8.20410 = ILCV 189; CIL 8.9715
+ p. 2034 = ILCV 2186; CIL 8.21550 = ILCV 4385 = Courtois, Vandales, p. 376, no. 80.
Crosses: CIL 8.21738 = ILCV 1720 = Courtois, Vandales, p. 379, no. 105 = Marcillet-Jaubert,
Inscriptions dAltava, pp. 12930, no. 199; CIL 8.21729 = ILCV 2862D = Courtois, Vandales,
p. 370, no. 29 = Marcillet-Jaubert, Inscriptions dAltava, pp. 11112, no. 166; CIL 8.9871 = CIL
8.21747 = ILCV 2862B = Courtois, Vandales, p. 367, no. 4 = Marcillet-Jaubert, Inscriptions
dAltava, pp. 1023, no. 148; CIL 8.21745 = ILCV 2862C = Courtois, Vandales, p. 367, no. 6
= Marcillet-Jaubert, Inscriptions dAltava, p. 104, no. 151; CIL 8.9870 + p. 2059 = ILCV 423;
Courtois, Vandales, p. 369, no. 16 = Marcillet-Jaubert, Inscriptions dAltava, p. 143, no. 223.
CIL 8.9715 + p. 2034 = ILCV 2186: aput / [sancto]s apostolos Petru et / [Paulu i]n nom[i]ne
dei et Cri/[sti].
CIL 8.21551 = ILCV 1915.
In addition to the cults of Sts Laurence and Hippolytus discussed above, nn. 16972, see, e.g.,
Duval, Loca sanctorum, 1:18895, nos. 93 and 94A (Constantine); ibid., 1:21521 and 2257, nos.
1067 and 110 (Sila); ibid., 1:23141, no. 112 (Telergma); ibid., 1:2757, no. 132 (Hr. Akrib);
ibid., 2814, no. 134 (Hr. Tarlist); ibid., 1:3379, no. 158 (Mechta el Bir); and ibid., 1:3546, no.
168 (Sidi Ferruch).
Notitia, M. Caes., pp. 12831; ibid., M. Sitif., pp. 1323; and ibid., Num. pp. 11923 (the Notitia
gives the number of bishops from Sitifensis as forty-four, but lists only forty-two; similarly

291

Staying Roman
centuries, then, the epigraphic and textual evidence would seem to suggest that elite urban populations in Numidia and Mauretania remained
highly Roman and Christian in their outlook.
Indeed, it seems likely that, as elsewhere in the sub-Roman West,
ecclesiastical office may to a certain extent have replaced secular office as
the sphere of Romano-African aristocratic activity in Moorish Africa.197
From the second half of the fifth century onwards, inscriptions from
Mauretania Caesariensis commemorate at least five bishops, four priests,
a deacon, and two subdeacons. The names of three of these individuals
a bishop, the deacon, and one of the subdeacons do not survive.
Of the remaining eight, all bore Latin names: the bishops Reparatus,
Potentius, Maximus, and Donatus; the priests Boniface, Donatus, Julius
Capsarius, and Victor; and the subdeacon Julius Lucianus.198 Much the
same seems to be true of the far larger number of names preserved in
the Notitia provinciarum of 484. Here a handful of Punic and otherwise
not obviously Greco-Latin names Burco, Metcun, Tacanus, Mingin,
Uzulus, Gedalius199 form perhaps 4 per cent of the name-stock of
the late fifth-century episcopate of Numidia, Mauretania Sitifensis, and
Mauretania Caesariensis. The fifth-century Numidian and Mauretanian
episcopal onomasticon clearly included distinctively local options; but
the vast majority of the bishops bore Greco-Latin names or names with
long-standing Christian connotations. A handful of these were scriptural

197
198

199

it gives the number of bishops from Numidia as 125, but lists only 123). For archaeological
evidence of rural Christianity, see also Leveau, Caesarea, pp. 24852, no. 12; p. 290, no. 79;
p. 316, no. 133; probably pp. 3446, no. 176; and perhaps p. 349, no. 180 and Fentress, At Kaci,
and Bounssair, Prospections dans le Belezma, p. 109; see also Leveau, Recherches historiques,
pp. 2834.
Cf., e.g., the flight of Gallo-Roman senatorial aristocrats and their ideals into the Merovingian
episcopate, explored by Mathisen, Ecclesiastical Aristocracy, and Mathisen, Roman Aristocrats.
Bp. Reparatus: CIL 8.9709 = ILCV 1105 = Courtois, Vandales, p. 375, no. 72. Bp. Potentius:
CIL 8.20914 = ILCV 1824 = Courtois, Vandales, p. 382, no. 131. Bp. Maximus: MarcilletJaubert, Inscriptions dAltava, pp. 1289, no. 197 = AE (1965), 74, no. 224. Bp. Donatus: Courtois,
Vandales, p. 381, no. 117: [m]emo(ria) sancti ma(r)t(y)r(is) Donati ep(i)s(copi) = CIL 8.21571,
///// EPS /////. Boniface:
which, however, renders the inscription: MEM SANCTI PAR
CIL 8.9731 = ILCV 1180 = Courtois, Vandales, p. 370, no. 32. Donatus presbyter: Courtois,
Vandales, p. 369, no. 21. Julius Capsarius: Marcillet-Jaubert, Inscriptions dAltava, p. 124, no. 190
= CIL 8.21742 = ILCV 1183 C = Courtois, Vandales, p. 378, no. 93. Victor: CIL 8.9586 + p.
1984 = ILCV 1179. Julius Lucianus: CIL 8.21588 = ILCV 1248 = Courtois, Vandales, p. 372,
no. 44.
Burco: Notitia M. Caes. 45, p. 129; Metcun: ibid., M. Caes. 54, p. 129; Tacanus: ibid., M. Caes.
79, p. 130; Mingin: ibid., M. Caes. 93, p. 131; Uzulus: ibid., M. Sitif. 37, p. 133; Gedalius: ibid.,
Num. 110, p. 123. Also Onesimus: Notitia M. Caes. 17, p. 128; Apocorius: ibid., M. Caes. 21,
p. 128; Idonius: ibid., M. Caes. 69, p. 130; Maddanius: ibid., M. Caes. 102, p. 131; Cardelus:
ibid., Num. 122, p. 123. Some of these seem to be Punic, e.g., Burco = Semitic brk (to bless)
and Mingin = Semitic mgn (to give): Jongeling, North-African Names, p. ix.

292

The Moorish alternative

Table 5.1. Secular office-holders in Moorish Africa, fifth to seventh centuries:
the epigraphic evidence
Name

Date

Provenance

Office

Reference

Iidir

508

Altava

CIL 8.9835

Iugmena

474

Julius 1
Julius 2
Masgiven
Masties
Masuna

605
606
508
fifthsixth
centuries
508

Thanaramusa
Castra
Volubilis
Volubilis
Altava
Arris

procurator castrae
Severianae
praefectus

Maximus

508

Altava

Safar

fifthsixth
centuries

Albulae

Altava

princeps
vice prepositus
praefectus
dux then
imperator
rex gentium
Maurorum et
Romanorum
procurator
Altavae
praefectus

Courtois, Vandales
no. 68
IAM 2, no. 603
IAM 2, no. 506
CIL 8.9835
Morizot, 1989,
p. 274
CIL 8.9835

CIL 8.9835
CIL 8.9800

(David, Peter, Stephen, Jacob),200 while another handful might perhaps

indicate the more humble social origins of some of these western bishops: Arator (ploughman), Mensor (land-surveyor), Patera (bowl or
saucer).201 Even these, however, are overwhelmed by the usual complement of Africans named Victor, Felix, Donatus, Benenatus, or Quodvultdeus.
By contrast, a far greater number of the men holding secular and military office in Moorish Africa in the same period seem to have borne
Berber (as opposed to Latin or even Punic) names (see Table 5.1); thus,
the Moorish officials whom we considered earlier in this section: the rex
Maurorum et Romanorum Masuna, the dux and imperator Masties, the prefects Iugmena, Safar, and Masgiven, and the procurator Iidir.202 The
trend is even more pronounced in Corippus Iohannis. The name of the
200
201
202

David: Notitia M. Caes. 105, p. 131; Peter: e.g., ibid., M. Caes. 72 and 75, p. 130; Stephen: ibid.,
M. Caes. 20, p. 128; Jacob: ibid., M. Sitif. 16, p. 128.
Arator: Notitia M. Caes. 48, p. 129; Mensor: ibid., Num. 108, p. 123; Patera: ibid., M. Caes. 97,
p. 131.
Jongeling, North-African Names, pp. 59 (Iider), 85 (Masgiven), 88 (Masties and Masuna), 1478
(Vartaia); see also ibid., p. 70 (Iugmena); on the name Safar (not listed), see also ibid., pp. 1234.

293

Staying Roman
chieftain Zabeas might be Punic; but in general the men who wielded
power and influence in the Moorish world of Corippus epic had Berber
names: Antalas himself and his father Guenfan; Iaudas; Carcasan; the praefectus gentis Ifisdaias and his son Bitipten; the duces Guentan and Iutungun;
the Laguatan chief and high priest Ierna; and so on.203
Of course, Latin names were far from unknown among the new ruling
class of Moorish Africa. The fort begun by Masgiven and Iider at Altava
was completed by a procurator named Maximus.204 An early seventhcentury vice prepositus commemorated in an inscription from Volubilis
was called Julius.205 This man was presumably a member of the local
ruling family; at least, the local princeps at the time was also named
Julius and, indeed, all of the contemporary inscriptions from Volubilis
memorialize members of the gens Julia.206 Kusayla, the name of the late
seventh-century ruler said to have led Roman and Berber forces against
the Muslims, could conceivably be the Latin Caecilius, well attested in the
Roman inscriptions of Morocco.207 In Corippus account, the envoy of
the Moorish chieftain Antalas to the Byzantine general John Troglita was
a Latin-speaker called Maccus, who may perhaps have borne a Roman
name.208 The same might be true of the chieftains Cusina (whose mother
was Roman) and Ceraus.209
As in the Vandal kingdom, it would probably be a mistake to conclude
too quickly that we can tell how an individual self-identified ethnically
on the basis of his or her name alone. Indeed, the appearance of Latin and
indigenous names in the same immediate family had been a feature of
African provincial society for quite some time. We have already encountered the phenomenon in the Romano-Berber names Aurelia Cursen,
Aurelia Tifalis, Julia Getula, and Valerius Zabullus. The fourth-century
African rebel Firmus bore a Latin name, while his brother Gildo and
203

204
205
206
207
208
209

Jongeling, North-African Names, pp. 10 (Antalas), 26 (Bitipten), 31 (Carcasan), 54 (Guenfan and

Guentan), 63 (Iaudas), 65 (Ifisdaias and Ierna), and 71 (Iutungun); see also e.g. ibid., pp. 49
(Gantal), 58 (Hisdreasen), 60 (Ialdas), and 66 (Ilasan). Jongeling suggests that Zabeas (Zaba) is
perhaps the Punic s.o (to keep watch) or s.pn (the north): ibid., pp. xxv and 152.
Marcillet-Jaubert, Inscriptions dAltava, pp. 1267, no. 194 = CIL 8.9835 + pp. 975 and 2059 =
ILCV 42 = Courtois, Vandales, p. 378, no. 95.
Euzennat and Marion, Inscriptions antiques du Maroc 2:317, no. 506.
Ibid., 2:3501 and 3534, nos. 603 (Julius princeps) and 608 (Julia Rogatiana); see also perhaps
ibid., pp. 3578, no. 619.
Camps, Rex Maurorum, p. 218 and Euzennat and Marion, Inscriptions antiques du Maroc 2:438
9; Moderan, Kusayla, pp. 4256, however, is sceptical.
Coripp. Ioh. 1.4627, p. 21 and Jongeling, North-African Names, p. 77; for the name Macus,
see also CIL 8.4966 and CIL 8.11475.
Ceraus: Coripp. Ioh. 6.732, p. 142 and Jongeling, North-African Names, p. 33. Cusina: Coripp.
Ioh. 8.26671, p. 175 and (on his career in general), PLRE 3:3668, s.n. Cutzinas; Jongeling,
North-African Names, p. 38; for the name Cosina, see also CIL 8.12945.

294

The Moorish alternative

father Nuvel bore Moorish ones. Nuvels father, in turn, was named
Saturnius, and his grandfather Florus.210 Hagith Sivan has argued convincingly that by the fourth century marriages between the provincials of Mauretania and their barbarian neighbours were remarkably
widespread.211 Certainly, Augustines father (who bore the Roman name
Patricius) married a woman with the North African name Monica.212
In 434, Julius Victor and his sons Peregrinus and Messorus erected an
inscription to the memory of Victors father and mother, whose names
were Peregrinus and Alurula.213 A somewhat tentative reading of an
early fourth-century inscription from Chott Manzel-Yayia in proconsular Africa seems to indicate that a certain Marcus had a son named
Aurelius Usasamauca.214 It is entirely plausible, then, that bishops called
Victor, Felix, Donatus, and for that matter Burco and Metcun could have
come from families which drew on a mixed Roman and Moorish namestock. Of course, the same is true of secular office-holders with names
like Maximus and Julian on the one hand and Masties and Masgiven on
the other.
If an individuals name is no sure indicator of ethnic identity, though,
it does seem reasonable to suppose that in the high imperial period
and probably in the post-imperial period as well the use of GrecoLatin names in an African provincial context could signal a particular
kind of status, authority, and wealth, whether actual or aspirational. That
individuals of this social profile should pursue careers in the church and
overwhelmingly so speaks volumes of the degree to which Christianity
had probably come to define at least one aspect of Romanness in the
Moorish hinterland. This is all the more the case both on the analogy of
the Nicene episcopate of the Vandal kingdom (into which many of the
Numidian and Mauretanian bishops remained integrated) and given the
subsequent emphasis that the Byzantine imperial administration placed
on the evangelization of Moorish kingdoms.
The inscriptions from Moorish Numidia and Mauretania, then,
seem to speak of important cultural continuities that underlay the
210
211

212
213
214

Camps, Rex Maurorum, p. 193. The precise etymology of Gildos name is unclear to Jongeling,
North-African Names, p. 50.
H. Sivan, Why Not Marry a Barbarian? Marital Frontiers in Late Antiquity (The Example of
CTh 3.14.1), in Mathisen and Sivan, Shifting Frontiers, pp. 13645. More recently, however, D.
Cherry, Frontier and Society in Roman North Africa (Oxford, 1998), pp. 10140 concludes that
there was a low incidence of RomanMoorish marriages along the African frontier from c.50 bc
to c.ad 250, remarking that There is no evidence of any significant measure of cross-cultural
interrelationship in the frontier-zone (pp. 1323).
Monica may be either Punic or Berber: Jongeling, North-African Names, p. 98.
CIL 8.21553 = ILCV 883 = Courtois, Vandales, p. 368, no. 8.
Courtois, Vandales, p. 370, no. 27.

295

Staying Roman
transformation of power in the region. Although Romano-African and
Byzantine authors such as Corippus and Procopius viewed Moors as
quintessential non-Romans, from the middle of the fifth century onwards
we find Moorish rulers in north-western Africa laying claim to Roman
titles and expressing their power in Roman terms. We know very little
about the backgrounds of these men, and the diversity of circumstances
and local conditions across the interior make it difficult to generalize.
Along the fringes of the Vandal kingdom, though, some new leaders
like Antalas appear to have been Romano-African provincials of humble
origins. Further west, monumental tombs of pre-Saharan type like the
Djedar perhaps testify to the presence of indigenous populations from
across the limes now wielding authority within what had once been
Roman territory. Cultural identity, of course, is a malleable concept, and
it is quite possible even probable that the gentiles living in the arid
regions across the empires frontiers had, in the course of centuries of
interaction with Romans, come to see themselves as participating in a
culture which continued beyond the limits of direct imperial control.215
But, whatever their origins, from the second half of the fifth century
rulers who were seen as Moorish from the perspective of Carthage and
Constantinople also found themselves in control of populations which
probably continued to consider themselves Roman or at least who
still used Latin, who measured time from the initial Roman occupation of their province, and who remained Christian and with whom
it was necessary for these Moorish leaders to reach some measure of
accommodation in order to rule.
Irrespective of local circumstance, then, the key point is that the
regions new rulers exercised power in a cultural context that was at once
Roman and African. From the perspective of the wider Mediterranean,
being Roman and being Moorish may have become incompatible over
the course of the fifth century, and Masties assertion that he never broke
faith either with Romans or with Moors probably implies that even in
the African borderlands there were some individuals who considered
themselves one but not the other. But, as in the imperial period, it is
entirely possible that in the interior itself one could also be both. Masuna
appears to have conceived of his power in dual terms, and the leader for
whom the earliest Djedar was built ruled a community in which both
the erection of an indigenous-style tomb and its adornment with a Latin
dedicatory inscription carried important symbolic meaning. As in, say,
contemporary Gaul where we find Gallo-Romans serving in an official
215

On frontiers as regions rather than narrow boundaries, see B. Isaac, The Meaning of the Terms
Limes and Limitanei, JRS 78 (1988), pp. 12547 and Whittaker, Frontiers of the Roman Empire.

296

The Moorish alternative

capacity under the new Visigothic, Burgundian, and Frankish regimes
within a few decades of their establishment216 there were probably
important social continuities in the exercise of power in Moorish Africa,
as men with names like Maximus served alongside those called Iider
and Masgiven in the new ruling class. Moreover, ecclesiastical office was
dominated by men who bore Latin names; and in an age when Romanness and Christianity were so closely linked, this too must have done
much to ensure a kind of stability in Romano-African society under
Moorish rule.
4. strategies of violence
Such continuities, however, raise two final questions. If high culture
was at least partly conceived in Roman terms in the Moorish West, why
were authors throughout the Mediterranean so reluctant to acknowledge
that Romanness? And if power was conceived in Roman terms among
the new Moorish leaders of the African hinterland, why did they so
frequently find themselves in conflict with imperial armies?
To take the second of these questions first: though Procopius tells us
that the Moors of Byzacena and Numidia initially rebelled against Byzantine imperial authority for no reason ( ,
  ), it is perhaps
possible to see political strategies at play in the actions of sixth-century
Moorish rulers.217 Indeed, his dismissive commentary notwithstanding,
Procopius himself appears to have seen in their first insurgence a test
of the empires strength: Belisarius, until then the supreme commander of Byzantine forces in Africa, had been recalled to Constantinople, and yet the imperial troops in the hinterland were few and still
unprepared (])
 . . . /  81 ).218 The Moors, Procopius
implies, sensed weakness; but what kind is not altogether clear. For a start,
what the historian described as a revolt seems by his own account to have
consisted initially not of attacks on the scattered and unready easterners
but rather in raiding and pillaging aimed against the Romano-African
population.219 This same pattern of violence appears also to have been
characteristic of both of the other early Moorish uprisings discussed
by Procopius.220 To be sure, late Roman and early Byzantine military
tactics generally emphasized raiding, counter-raiding, and war by attrition rather than set-piece battles.221 But such actions were usually in
216
218
219
220
221

217 Proc. BV 2.8.9, 1:453.

Mathisen, Roman Aristocrats, pp. 11931.
Proc. BV 2.8.201, 1:4545; the quotation is ibid., 2.8.21, 1:454.
Proc. BV 2.8.202, 1:4545.
Proc. BV 2.12.1 and 2.13.1, 1:470 and 1:475; see also ibid., 2.21.17, 1:520.
Haldon, Warfare, State and Society, pp. 3642.

297

Staying Roman
pursuit of a greater strategic end, and in the case of the Moorish raids
of the mid 530s that end may well have been material concessions from
the imperial administration. Certainly, Procopius purports to relate the
contents of a letter sent by the aggrieved Moorish rulers to Belisarius successor Solomon, recounting their complaints that the Romans
shared nothing good with them, and that as a result they were pressed
by hunger.222 The evidence must be treated with caution, for it is a key
element in the historians guarded indictment of Justinian as a tyrant.223
But it is striking that Cusina, one of the leaders of the early Moorish
raiding parties, was later to receive a subsidy in gold from the Byzantine
governor and become a steadfast imperial ally.224 Similarly, one of the
few Moorish leaders said to have remained aloof from the early conflicts
was the chieftain Antalas, who likewise received a public maintenance
(
; literally, a feeding) from Solomon.225
Moreover, the early raids are not the only instance of politically motivated aggression in sixth-century Africa. Antalas himself later revolted,
apparently both because Solomon took away his maintenance and
because the governor-general had the chieftains brother killed for his
part in some disturbances in Byzacena.226 John Rhogathinos, probably
the Praetorian Prefect of Africa, similarly tried to rescind Cusinas subsidy in 563. In this case the Moorish leader himself was killed, prompting
his sons to rebel.227 A meeting between the dux Tripolitaniae Sergius
and eighty chieftains of the Laguatan confederation in 544 resulted in
the deaths of all but one of the Moorish leaders, shredding local perceptions of Byzantine trustworthiness and leading to retaliatory attacks
against imperial territory.228 In short, then, violence frequently seems to
have been deployed in the North African hinterland with the strategic
goal of forcing the imperial administration to acknowledge and redress
grievances.
Of course, economic strategies probably also underlay Moorish raiding. We may catch a hint of this, too, in the chieftains complaints to
Solomon about being pressed by hunger, though it is apparent in any
case from their actions. Moorish raids were said to involve the killing of
222
224

225
226
227
228

223 Kaldellis, Procopius of Caesarea, pp. 11864, esp. 132.

Proc. BV 2.11.9, 1:464.
Subsidy: John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae
15 (Bonn, 1831), p. 495. Ally: Proc. BV 2.25.1522, 2.27.247, and 2.28.50, 1:5345, 1:5434,
and 1:552; Coripp. Ioh. 3.4068, 4.50914, 5.4502, 6.2678, 6.468, 6.51217, 7.245, 7.2624,
8.1212, 8.2656, 8.371, 8.42841, and 8.465, pp. 63, 86, 109, 123, 131, 133, 153, 154, 170,
175, 179, 1812, and 183.
Proc. BV 2.12.30, 2.21.17, and 2.22.8, 1:475, 1:520, and 1:523.
Proc. BV 2.21.17 and 2.22.8, 1:520 and 1:523; Coripp. Ioh. 2.28 and 4.35875, pp. 27 and 801.
Malalas, Chronographia, pp. 4956; Theophanes, Chronographia AM 6055, 1:2389.
Proc. BV 2.21.522, 1:51821.

298

The Moorish alternative

men, capturing of women and children as slaves, and plundering of the
countryside.229 This kind of brigandage was thus probably aimed primarily at undefended rural populations; at least, Fulgentius the Mythographer
seems to have been able to sit out one such incident in his country house,
which was presumably fortified, while for a time the invaders controlled
the rest of his estate.230 In Justinians reign such raids appear generally to
have focused on the frontier provinces of Numidia and Byzacena, and
imperial fortifications of important settlements and administrative centres are correspondingly most heavily concentrated there; though from
the later sixth century onwards defensive structures also extended into
Africa Proconsularis.231 However, incursions of this sort, presented to us
in our sources as typically Moorish, probably went both ways. At least,
in 544 Laguatan chieftains complained to Sergius about the plundering
of their crops, and after a crushing defeat at the hands of Solomons army,
some Moorish communities were said to have fled Byzacena for fear of
violence at the hands of local Romano-Africans.232 Roman soldiers, too,
regularly took Moorish women and children captive in the wars of the
mid sixth century.233
The launching of slave raids into each others territory may seem to
us to be the ultimate sign of a profound gulf separating Romans and
Moors in Byzantine Africa. But we should remember that in the 440s
it had proved possible for an aristocratic Roman woman captured in
the same basic region to be sold as a slave in imperial territory in the
East (see above, Chapter 2.1.1). Towards the end of his life, Augustine
was troubled to discover that Roman slave catchers were operating in
Numidia, capturing rural Romans and selling them overseas.234 Nor was
this a problem distinctive to Africa. In the 470s, in the twilight years of late
229
230
231

232
233
234

Proc. BV 2.8.22, 2.10.25, and 2.13.1, 1:455, 1:4589, and 1:475.

Fulgentius Mythographer, Mitologiae 1.praef., pp. 56.
Proc. BV 2.8.21, 2.10.2, 2.12.12, 2.13.1, and 2.21.17, 1:4545, 1:4589, 1:470, 1:475, and
1:520; Pringle, Defence, pp. 99109. See also Coripp. Ioh. 3.393400 and 3.44257, pp. 62 and
645.
Proc. BV 2.12.29 and 2.21.5, 1:475 and 1:518.
Proc. BV 2.11.55, 2.12.27, and 2.21.14, 1:470, 1:475, and 1:519; see also ibid., 2.20.239, 1:516
17.
Augustine, Ep. 10*.28, CSEL 88:4651. On this letter, see in general M. Melluso, Alcune
testimonianze in tema di mercati di schiavi nel tardo antico, in M. Garrido-Hory (ed.), Routes et
marches desclaves, 26e colloque du GIREA (Paris, 2002), pp. 34570; J. Szidat, Zum Sklavenhandel
in der Spatantike (Aug. epist. 10*), Historia: Zeitschrift fur Alte Geschichte 34 (1985), pp. 36071;
J. Rouge, Escroquerie et brigandage en Afrique romaine au temps de saint Augustin (Epist.
8* et 10*), in Les lettres de saint Augustin decouvertes par Johannes Divjak: communications presentees
au colloque des 20 et 21 Septembre 1982 (Paris, 1983), pp. 17788; M. Humbert, Enfants a` louer
ou a` vendre: Augustin et lautorite parentale (Epist. 10* et 24*), in Les lettres de saint Augustin
decouvertes par Johannes Divjak. Communications presentees au colloque des 20 et 21 Septembre 1982
(Paris, 1983), pp. 189204.

299

Staying Roman
Roman Gaul, Sidonius Apollinaris learned that his very own negotiator
had bought a freeborn woman who had been captured and sold as a
slave by bandits.235 Imperial legislation including a rescript collected
into Justinians Codex long struggled to prevent Roman soldiers from
keeping as slaves those captive Roman citizens whom they had re-taken
from defeated barbarians.236 In this context, it is perhaps significant that,
while Procopius tells us that Byzantine units sometimes attacked Moorish
raiding parties in an attempt to seize their captives, we do not hear
whether the unfortunate provincials were ever freed.237
Indeed, one generally gets the sense that, though Moorish raiding
could and did serve as a causus belli, small-scale raids escalated into largescale conflicts primarily when bigger political factors were involved, and
above all the death of imperial officers. Thus, for example, Solomon
led an expedition into Byzacena because of Moorish brigandage in the
area; but he was most troubled because the Roman commanders Agan
and Rufinus had been killed in a skirmish with Moorish forces while
trying to take for themselves the raiders captives.238 Though we have
only the barest outlines of an account, something similar seems to have
happened in the 560s and 570s, when, as we have seen, two successive magistri militum Africae and a Praetorian Prefect were all killed in
a war with the Moorish king Garmules, before the general Gennadius
killed the king himself in 577 (see above, Chapter 4.3.4). In this context,
one wonders too about Antalas difficulties in effecting a reconciliation
with Justinian. After his brother was killed, the chieftain led a revolt
which ultimately resulted in the death in action of the governor-general
Solomon.239 Though Antalas apparently retained the generals battlestandards, twice thereafter we hear of the chieftain proffering his submission to the emperor.240 Intent on projecting strength and indomitability,
however, and perhaps personally upset by the death of a trusted representative, Justinian apparently rejected the offer, sending John Troglita to
crush Antalas once and for all.241
It is important to stress, though, that the interactions between Moorish rulers and imperial administration were not simply characterized by
endless cycles of retaliatory violence. Byzantine officials had Moorish
235
236
237
238
239
240
241

Sidonius Apollinaris, Ep. 6.4, 3:1516.

CJ 8.50.12 (ad 293), p. 360; CTh 5.6.2 (ad 409), p. 221.
Proc. BV 2.10.5 and 2.13.217, 1:459 and 1:4757.
Proc. BV 2.10.111, 1:45860; see also, however, ibid., 2.13.119, 1:4758, where Moorish
politics seem to have been in play.
Proc. BV 2.21.1628, 1:51922; Coripp. Ioh. 3.40141, pp. 624.
Submission: Proc. BV 2.22.610 and 2.27.4, 1:523 and 1:541. Battle-standards: ibid., 2.28.46,
1:551.
Proc. BV 2.28.4552, 1:5512 and Coripp. Ioh. passim.

300

The Moorish alternative

allies as well. Indeed, as late as the seventh century, it would appear that
Moors and Byzantines may still have fought side by side, at least on
occasion. As we have seen, Ibn Abd al-Hakam claims that the North
African leader Kusayla led Roman and Berber troops against the invading
Muslims in the 680s (see above, section 3); and John of Nikiu similarly
indicates that the forces sent against the emperor Phocas by the usurper
Heraclius in his rebellion of 609 included a large number of barbarians,
seemingly Moors of the Laguatan confederation.242 Indeed, one of the
aspects that seems to have troubled Justinians administration as it shall
have troubled Phocas was not the nature of the conflict between imperial officials and Moorish leaders in Africa but rather the nature of their
collaboration.
In the 530s and 540s, Moorish rulers both inside and outside the
frontiers of Byzantine Africa repeatedly allied themselves with forces
that threatened imperial control of the region; forces which, moreover,
as often as not included renegade Roman officers and soldiers. First,
as we have seen, in the wake of the Byzantine reconquest, about 400
defeated Vandal warriors sought refuge among the Moors of the Aur`es
Mountains and Mauretania (see above, Chapter 1.3). Then, in April of
536, a rebellion erupted in the ranks of the Byzantine occupying army
which threatened to swamp the imperial endeavour in Africa. For a brief
moment, the rebels even gained the upper hand in Carthage, forcing the
historian Procopius himself, in the company of Solomon, then supreme
commander of the Byzantine armies in Africa, to flee to Sicily.243 The
fortunes of the rebels soon turned, but in 537 they still posed a threat
to imperial control of the newly reconquered African provinces. The
rebellious Byzantine soldiers were joined by the fugitive Vandals and at
least briefly by some Moors as well, as they confronted a loyalist army
commanded by the emperor Justinians own cousin, Germanus.244 When
the rebellion collapsed, the rebel leader Stotzas and some of his followers
fled to Mauretania too, where Stotzas was said to have married a Moorish
princess, and from where he continued to participate in Moorish attacks
on Byzantine territory.245
242
243
244

245

John of Nikiu, Chronicle, 109.25, p. 176 and see also ibid., 109.224, p. 176; Moderan, Maures,
pp. 6503.
Proc. BV 2.14.3741, 1:4878.
Proc. BV 2.17.8, 1:5012. The Moors were, however, unreliable allies: Proc. BV 2.17.912
and 2.17.31, 1:502 and 1:505. On the importance of family connections to the exercise of
late Roman and early Byzantine imperial power, see McCormick, Emperor and Court, esp.
pp. 14851.
Proc. BV 2.17.35, 2.22.5, 2.23.117, 2.23.2631, and 2.24.614, 1:505, 1:5223, 1:5257, 1:529,
and 1:5302.

301

Staying Roman
In 545, a rebellious Byzantine officer allied himself with Moors against
imperial interests once again. In a bid to seize power in Carthage, Guntharis, the commander of Byzantine Numidia, entered into a conspiracy
with the chieftain Antalas, as well as with Stotzas successor John and
the last hold-outs of Stotzas mutineers. Guntharis briefly achieved his
goal of supreme command in Africa by other means and the alliance fell
apart; but for a time the Byzantine officer had agreed to share Africa with
the Moorish chieftain, who was to rule Byzacena with 1,500 Byzantine
soldiers, while Guntharis himself ruled the rest of the imperial provinces
from Carthage as basileus.246 From an African perspective, the deal is perhaps most revealing of Antalas aspirations to Roman power.247 From an
imperial vantage point, on the other hand, it was revealing of a disturbing
trend in African society for chieftains who only a few years before had
proclaimed their submission to the empire and received the symbols of
their authority from the conquering general Belisarius to prove singularly
unreliable allies indeed.
This difference in perspective is probably important to late antique
assessments of the Romanness of the Moorish kingdoms in general.
By the sixth century, it seems likely that in the African hinterland
just as in Carthage and Constantinople Romanness could be defined
in terms of politics, high culture, and religion. Many key signifiers of a
Roman identity may have been shared between and among these regions
as well: particular ways of expressing power; the imagining of oneself,
ones family, and ones community in terms of the longer history of the
empire; the acceptance of Christianity. But, as we also saw in the Vandal
kingdom, in Moorish Africa the devil was still in the details. The rulers of
the Aur`es and of Altava could claim to be emperors, and in a provincial
setting they may even have gained local recognition of the validity of
their claims; but in Constantinople there were ways of legitimating an
imperial accession that were simply not available to African provincials
among them coronation by the senior emperor or, in the event that there
was no senior emperor, by the patriarch.248 From a Byzantine point of
view, an empire that was confined to Numidia or Mauretania would
have seemed like a rebellious province, an empire of shadows. Something
246
247
248

Proc. BV 2.25.10, 1:533; see also ibid., 2.25.13, 1:532. Guntharis plot, its ephemeral success,
and eventual defeat are the subject of ibid., 2.25.12.28.41, 1:53251.
Moderan, Maures, p. 329.
S. MacCormack, Art and Ceremony in Late Antiquity, Transformation of the Classical Heritage
1 (Berkeley, Calif., 1981), pp. 2407. Provincial setting: cf. M. McCormick, Clovis at Tours:
Byzantine Public Ritual and the Origins of Medieval Ruler Symbolism, in E. Chrysos and A.

Schwarcz (eds.), Das Reich und die Barbaren, Veroffentlichungen des Instituts fur Osterreichische
Geschichtsforschung 29 (Vienna, 1989), pp. 15580.

302

The Moorish alternative

similar appears to have played out in the realm of high culture. Just as
the anonymous epigrammatist of the Latin Anthology mocked the literary
pretentions of the Arzugitanus poeta, when cosmopolitan sophisticates like
Corippus and Procopius regarded the African borderlands, what they
seem to have seen was rusticity, poverty, and perhaps even superstition
with which they wanted no part.
The loss to the empire of western Africa and inland Tripolitania in the
fifth century set in motion a series of developments in these regions
which our sixth-century literary sources understood in terms of deRomanization. These sources perspectives are at once political and cultural, and to make sense of the evidence that they present us it is important
to appreciate how these two aspects interact. The Moorish kingdoms
that emerged in the African interior continued to enjoy a fair degree of
practical autonomy visible in the literary sources primarily as Moorish
violence even after the reimposition of Byzantine power in the region.
We know little about the background of Moorish leaders themselves,
but, whatever their origins, like other rulers in the barbarian West they
seem to have sought to legitimate their power to a provincial Roman
audience in part though an appeal to continuity with the imperial past.
To judge from the procurator Maximus, for example, it seems likely that
such appeals at least occasionally worked. Moreover, in the 530s and
540s, the western Moorish kings and chieftains also attracted the alliance
of dissident elements within the occupying Byzantine army further to
threaten imperial control of North Africa. Given the need of imperial
apologists to de-legitimate the new leaders and their claims to Roman
power, it is thus perhaps no surprise to find our sources depicting Moors
as quintessential barbarians.
Yet, as Moderan has pointed out, in the Byzantine period the dividing
line between Romans and Moors was not strictly political, at least not in
the sense that Africans who supported the empire were called Romans
and Africans who were hostile to the empire were called Moors. In
Corippus epic, the chieftain Cusina was unambiguously Moorish, but
also a loyal supporter of the empire.249 Other Moorish leaders also worked
successfully with the empire, and even in their raids on imperial territory
it is possible to see a meaningful political engagement with the Byzantine
administration in Africa. In his constancy and political loyalty, though,
there was something particularly comfortable and familiar to Corippus
about Cusina. He was at heart a Roman, the poet writes, and not
249

Moderan, Maures, p. 418.

303

Staying Roman
far off by blood, distinguished by a mild character and Latin dignity.250
Mild; not savage (ferox) like other Moors, who raised such anxieties in
Roman authors throughout the Mediterranean.
These anxieties should be taken seriously. Not because they are an
accurate reflection of lived realities in the African borderlands; but
because they are revealing of some of the tensions pulling at the Mediterranean world in late antiquity. In voicing his distaste for Byzantine Africas
Moorish neighbours, Corippus (for example) is not simply serving as a
mouthpiece for imperial propaganda; he is drawing on widespread elements of what seem to have been shared late Roman attitudes towards
Afri barbari. Already by the fifth century the purported savagery and
perfidiousness of Moors were proverbial in the late Roman world.251
This perception was surely further reinforced by Sallust, one of the most
studied Latin authors in late antiquity.252 Moreover, in some minds a distinct colour prejudice combined with apocalyptic fears to make African
barbarians seem positively demonic.253 In one way or another, these attitudes tinged the lens through which every author we have considered
here viewed the Moors. Romans probably especially including such
intensely Roman Africans as Corippus, Victor of Vita, and the poets of
the Latin Anthology appear to have been quite genuinely terrified of
the world of barbarism that, to them, Moorishness represented. To these
Africans, the Moorish alternative was no alternative at all.
As we have seen, though, we have plenty of grounds on which to
question such authors reasoning. Our sources sought to explain cultural
difference through ethnography and the rhetoric of barbarism; yet their
need to separate the two communities, to establish firm dividing lines
between Romans and non-Romans, masks a much more complex reality.
In this context it bears repeating that Romanness meant different things
to different people in different places. In late antiquity, the African predesert and indeed much of Numidia and the Mauretanias was a frontier
culture in which whether one was Roman or not probably depended on
250
251
252

253

Coripp. Ioh. 4.51112, p. 86: ille animo Romanus erat, nec sanguine longe, / moribus ornatus
placidis, grauitate Latina.
Fevrier, Differences et conflits, esp. pp. 1436; Bowersock, Brown, and Grabar, Late Antiquity,
p. 569 s.n. Mauri.
See, e.g., M. Manitius, Geschichte der lateinischen Literatur des Mittelalters, 3 vols., Handbuch der
Altertumswissenschaft, 9. Abt., 2 (Munich, 1911), 1:65, 90, 131, 184 n. 3, and 218; and A.-D.
von den Brincken, Studien zur lateinischen Weltchronistik bis in das Zeitalter Ottos von Freising
(Dusseldorf, 1957), pp. 72 and 81. For Sallusts continuing influence in the Middle Ages, see
B. Smalley, Sallust in the Middle Ages, in R. R. Bolgar (ed.), Classical Influences on European
Culture ad 5001500: Proceedings of an International Conference Held at Kings College, Cambridge,
April 1969 (Cambridge, 1971), pp. 16575.
See, however, Snowden, Before Color Prejudice, pp. 825.

304

The Moorish alternative

who was asking. Critically, though, from the vantage point both of Constantinople and of Carthage, Mauri had ceased to be truly Roman: they
had become barbarian Moors rather than Roman Mauretanians. Indeed,
if (as I suspect) by the early sixth century western Romano-Africans
were already making common cause with their new Moorish masters,
it must have seemed to the urbane elite of Vandal and Byzantine Africa,
no less than to the inhabitants of the rest of the Roman Mediterranean,
that even once-Roman Mauretanians had begun to partake in some of
the perfidy of the African barbarians who now ruled those western lands.
Yet Romanness continued to matter in this Moorish society, and, in the
historical memory of medieval Arabic scholars, the final resistance to the
Islamic invasions of the seventh century was thought to have come not
from Byzantines, but from a Berber prince named Kusayla and a Berber
queen known as the Kahina.

305

Chapter 6

THE DILEMMA OF DISSENT

The fragility of the Byzantine reconquest was not a foregone conclusion.

If they rejected Moorish claims to Roman power, Romano-Africans living along the Mediterranean littoral seem generally to have accepted and
identified with the Byzantine empire. This did not always equate to easy
acceptance and identification with imperial policy. Unlike Corippus,
who linked his fortunes so closely with the Byzantine military administration in Africa (and later with the imperial court), the staunchly
conservative African episcopate consistently found itself at odds with
the theological innovations of successive Byzantine emperors. In this,
however, they were not alone; and, indeed, in the seventh century, the
African church found its voice in the eastern monk and theologian Maximus Confessor. Seventh-century Africa also produced two notable rebels
(the exarchs Heraclius and Gregory), one of them successful, the other
not, but both representatives of the Byzantine administration in North
Africa. Far from indicating disaffection with the empire, however, the
forms of dissent in sixth- and seventh-century Africa reveal the extent to
which the provincial elite was integrated into the political, cultural, and
mental structures of the Roman world at large on the eve of its collapse.
1. the legitimation of reconquest
The Byzantine reconquest of Africa once again necessitated the legitimation of political and military power in the region; only this time it was the
actions of the East Roman state rather than the rule of the Vandal kings
that required justification. Our surviving sources provide two distinct
but complementary pictures of how Justinian went about explaining his
African war. Imperial legislation, followed by the accounts of contemporary and near-contemporary historians, emphasizes the dual themes of
the return of Roman power and the reimposition of theological orthodoxy. But this picture is deepened and complicated by Procopius a
first-hand observer and the author of our fullest account of the reconquest who also emphasizes Justinians efforts to defend the invasion
306

The dilemma of dissent

in terms of the Vandal royal succession when addressing a specifically
African audience.
In retrospect, Justinians seizure of the Vandal kingdom looks to us
very much the way the emperor wanted it to look: as a reconquest, a
restoration of the empire in Africa. From the very start Justinian took
an active hand in the effort to shape wider perceptions of his audacious
western venture. As we have seen, the emperors rescript of 534, issued
hard upon the collapse of Vandal rule, emphasized that through Byzantine
intervention Africa had regained its ancient liberty, a euphemism for
direct imperial control (see above, Chapter 4). In Justinians rhetoric,
however, this had not been an end in itself. Rather, the principal reason
that the emperor gave for his invasion of the African kingdom was the
Vandal Arianizing policy: the rebaptism of local Nicene Christians, the
pollution of African churches, the mutilation of the confessors of Tipasa.
God himself had allowed Justinian to recover the region in order to
vindicate the Nicene church and rescue it from the yoke of servitude;
now the autarch invoked the aid of the Virgin too in the hopes that
through her prayers the deity would restore even the most insignificant
fragments of the empire.1 For the next twenty years and more, Justinians
African legislation continued to sound this same basic theme: that Africa
had been redeemed from Vandal servitude through his own vigilance and
with divine assistance.2
Justinian seems to have been phenomenally successful at winning
the propaganda campaign. At least, not a single contemporary or nearcontemporary account of the reconquest viewed the war as an unjustified
attack on a legitimate king. To be sure, most were writing within the
empire itself. Only the Gallic chronicler Marius of Avenches (d. 601)
composed his account outside imperial territory, and he was writing
perhaps fifty years after the event; but his matter-of-fact record accepts
that after ninety-two years Africa was restored to the Roman empire.3
In this, Marius echoes the assumptions and attitudes of Justinians former
cancellarius, Marcellinus comes. In his own chronicle, Marcellinus celebrated the capture of Carthage in terms that leave no doubt that he
viewed the event as a restoration rather than a subjugation of the region
1
2
3

CJ 1.27.1.19 (ad 534), p. 77.

Just. Nov. 36 (ad 535), App. 2 (ad 541), and App. 9 (ad 558), pp. 2434, 7967, and 803.
Marius of Avenches, Chronica s.a. 534, ed. T. Mommsen, MGH AA 11:235: Eo anno Africa
Romano imperio post nonaginta et duos annos per Belesarium patricium restituitur: et Gelimer
rex Vandalorum captivus Constantinopoli exhibetur et Iustiniano Augusto cum uxoribus et
thesauris a supra scripto patricio praesentatur. Marius presumably used the treaty of 442 as the
starting-point of Vandal dominance in Africa rather than their capture of Carthage (439), as was
typical in the Byzantine sources.

307

Staying Roman
to Roman power: almost 100 years after their loss to the empire, the
North African provinces had been reintegrated into the Roman state
through the guidance of Justinian.4 This was certainly the court view.
The reactions of other authors writing in the East were mixed in their
appraisals of Justinians endeavour. Like Marcellinus, Jordanes celebrated
the return of Africa to the freedom of Roman rule (libertas regni Romani)
after a nearly century-long Vandal interregnum.5 John Lydus, too, seems
to include Africa among those parts of the Mediterranean world that
had once been Roman but that had slipped out of imperial control
through the indolence of previous generations.6 John Malalas who was
not particularly interested in the fate of the African kingdom gives
no indication that he saw the Byzantine invasion as in any way illegitimate, though his account of it is oblique in the extreme, mentioning
only the capture of the last Vandal king, Gelimer, and his deportation to
Constantinople.7 Even the continuator of Zachariah of Mitylenes Syriac chronicle, who seems to have been somewhat ambivalent about the
reconquest, acknowledged that Africa had once been a Roman province,
and that Belisarius army re-established it as such.8 Indeed, there can be
little doubt that the war was widely perceived as having been undertaken for precisely the reasons Justinian and his administration projected:
to reimpose imperial rule on a former Roman province and to free
Africans of Nicene confession from the Arian yoke of the Vandals.
Justinians claims to have undertaken the reconquest with divine
assistance also seem to have resonated with contemporary and nearcontemporary observers. Marcellinus comes wrote of the event as a vindication of Africa through Gods will.9 Procopius claimed that an anonymous eastern bishop had come to the palace and told the emperor that,
through a dream-vision, God himself had offered his assurances that Justinian would fight with divine aid should he attempt the reconquest of
Africa.10 Cyril of Scythopolis, writing in the 550s, similarly indicated
that St Sabas came in person to Constantinople and assured Justinian
4

5
6
7
8
9
10

Marcellinus comes, Chronicon s.a. 534, pp. 1034: Carthago quoque civitas eius anno excidionis
suae nonagensimo sexto pulsis devictisque Vandalis et Gelimer rege eorum capto et Constantinopolim misso, quarto Iustiniani principis consulatu, ipsius moderatione recepta est, sua cum
patria firmius, quam dudum fuerat, redintegrata.
Jordanes, Getica 33.172, ed. F. Giunta and A. Grillone, in Iordanis De origine actibusque Getarum,
Fonti per la storia dItalia 117 (Rome, 1991), pp. 734.
John Lydus, De Magistratibus Populi Romani 3.55, ed. R. Wunsch (Leipzig, 1903), p. 144.
John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn,
1831), pp. 4789.
Zachariah of Mitylene, Chronicle 9.17, pp. 2623.
Marcellinus comes, Chronicon s.a. 534, p. 103: Provincia Africa . . . volente deo vindicata est.
Proc. BV 1.10.1821, 1:3589.

308

The dilemma of dissent

that if the emperor supported and defended the Palestinian church, God
would reward the autarch by returning Africa and Italy to the empire and
orthodoxy.11 Most of what the sixth-century ecclesiastical historian Evagrius Scholasticus had to say about Vandal Africa consists of quotations
from Procopius History of the Wars, but Evagrius strung these together in
such a way as to emphasize the violence of Vandal Arianism, the piety of
Justinians motivations in reconquering Africa, the fact that Carthage was
captured at the time of the Cypriana (an annual festival in honour of St
Cyprian), and Cyprians own role in assuring Nicene Africans through
dream-visions that he would avenge himself on their Arian oppressors.12
Gregory the Great, a contemporary of Evagrius, even believed that the
confessors of Tipasa had their tongues ripped out in Justinians own day.
The incident had in fact occurred nearly fifty years earlier, but Gregorys
misunderstanding on this point is probably indicative of the importance
of the story to the emperors justification of his daring undertaking.13
From within Africa itself, Victor of Tonnena (followed by Isidore of
Seville) recorded that Justinians invasion of the Vandal kingdom had been
prompted by a vision of Bishop Laetus, executed by Huneric and now
remembered as a martyr of the persecution.14 Even Pseudo-Zachariah of
Mitylene conceded that God willed this expedition and assisted it.15 In
short, then, Justinian seems to have won widespread acceptance of both
of the essential elements of the imperial line with respect to the invasion
of Africa: that it was a justifiable war of reconquest and that its success
was a result of divine favour.
Like the Vandal kings, in legitimating his newfound control of Africa
Justinian had to address multiple audiences at the same time. I have
suggested above that even after the reconquest not all contemporary
observers were equally enthusiastic about the undertaking; Procopius
makes it clear that earlier, before news of Belisarius easy victories over the
11

12
13
14

15

Cyril of Scythopolis, Vita Sabae, ed. E. Schwartz, in Kyrillos von Skythopolis, Texte und Untersuchungen zur Geschichte der altchristlichen Literatur 49/2 (Leipzig, 1939), pp. 1756; see also
pp. 1789.
Evagrius Scholasticus, Historia Ecclesiastica 4.1416, ed. J. Bidez and L. Parmentier, in The Ecclesiastical History of Evagrius with the Scholia (London, 1898), pp. 1637.
Gregory I, Dialogi 3.32, ed. A. de Vogue , in Diologues, 3 vols., SC 251, 260, 265 (Paris, 197880),
2:3902.
Vict. Tonn. s.a. 534.118, p. 38; Isidore, Historia Vandalorum 834, ed. and trans. C. Rodrguez
Alonso, Las Historias de los godos, vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica
y traduccion, Fuentes y estudios de historia leonesa 13 (Leon, 1975), pp. 3068. On Laetuss
transformation into a martyr, see D. Shanzer, Intentions and Audiences: History, Hagiography,
Martyrdom, and Confession in Victor of Vitas Historia Persecutionis, in A. H. Merrills (ed.),
Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot, 2004),
p. 284.
Zachariah of Mitylene, The Syriac Chronicle Known as that of Zachariah of Mitylene 9.17, trans.
F. J. Hamilton and E. W. Brooks (London, 1899), p. 262.

309

Staying Roman
Vandal army reached Constantinople, the emperors decision to invade
Africa was deeply unpopular in court circles.16 Earlier imperial efforts
to reconquer the province had ended in disaster and defeat: the failure
of the western emperor Majorians expedition against the Vandals in 461
had led to that rulers deposition, while the annihilation of the eastern
imperial fleet sent against Carthage in 468 had forced the expeditionary
commander (Basiliscus) to seek sanctuary in St Sophia upon his return
to Constantinople.17 But success breeds success. The surprising triumph
of Justinians armies seems to have sent a powerful message to doubters
and supporters alike presumably both inside and outside the empire
about the strength of Roman arms, about the emperors determination to
enforce imperial hegemony in the Mediterranean, and apparently about
his special relationship with God in the reimposition of orthodoxy.
Before victory was assured, however, the emperors concerns appear
to have been slightly different. Later avowals of the favour with which
God looked on the Vandal war presumably do echo the propaganda
with which Justinian and his inner circle sought to overcome ministerial
reluctance to attack Africa in advance of the invasion; but, to ensure its
success, the reconquest also had to be sold in Africa. Of all the surviving
accounts only Procopius gives any real sense of how this was done.
Here again the imperial administration addressed two distinct (though
probably at least occasionally overlapping) audiences: supporters of the
Vandal regime on the one hand and those elements of the RomanoAfrican population who still continued to feel affection for or affiliation
with the empire on the other.
To the second of these groups, the Byzantine high command wanted
very much to appear not as conquerors but as restorers of a Roman
presence. This meant convincing Belisarius troops. Procopius indicates
that on at least two separate occasions the general reminded his soldiers
that the African population had once been Roman: the imperial army was
in Africa only to recover for the empire what rightfully belonged to it.18
Moreover, an almost obsessive concern with plundering emerges from

16
17

18

Proc. BV 1.10.217, 1:3558.

Majorian: Fasti Vindobonenses Priores s.a. 461, ed. T. Mommsen in MGH AA 9:274336,
p. 305; Hydatius, Chronicon 205 (ad 461), p. 114; Marcellinus comes, Chronicon s.a. 461.2,
p. 88; Chronica Gallica a. dxi 6335, p. 664; John of Antioch, frag. 203, ed. Muller, in Fragmenta
historicorum Graecorum 4:616; Evagrius, Historia Ecclesiastica 2.7, p. 55; Malalas, Chronographia,
p. 375; Theophanes, Chronographia AM 5955, 1:112. Basiliscus: Proc. BV 1.6.1026, 1:33740.
He was accused of accepting bribes from the Vandals: Priscus, frag. 53.1, p. 362 (= Theophanes,
Chronographia AM 5961, 1:116); Malalas, Chronographia, pp. 3723, Zonaras, Epitomae 14.1, 3:253.
Africans once Romans: Proc. BV 1.16.3 and 1.20.19, 1:382 and 1:399. Recovery of Africa: ibid.,
1.19.5, 1:392.

310

The dilemma of dissent

Procopius account of Belisarius African policy.19 Under late Roman law,
citizens could be compelled to billet soldiers, but not to feed them.20
Belisarius was said sharply to have reprimanded his men for stealing fruit
from the orchards of private Romano-Africans, even going so far as to
inflict corporal punishment on the offenders.21 Only from Vandal royal
estates were the generals troops allowed to eat their fill with impunity.22
In the cities of Sullecthum and Carthage, and presumably in Leptis
Minor (near mod. Monastir) and Hadrumetum, the inhabitants provided
markets for the Byzantine army where Belisarius soldiers bought their
own food.23 Africans were to be treated like any other imperial subjects.
Our sense of how the reconquest was legitimated to supporters of the
Vandal regime comes primarily from a series of letters that Procopius
claims Justinian sent to Gelimer and the Vandal magnates. We must treat
the evidence of this epistolary exchange with caution, for it develops
the discourse on tyranny that underpins the History of the Wars as a
whole.24 Even so, in Procopius account these letters were all scrupulously
concerned with the constitution of Geiseric, the law passed by the first
Vandal king to ensure that his oldest male descendant would rule from
Carthage. The coup in which Gelimer had dethroned and succeeded his
cousin Hilderic was, in the rhetoric of Procopius Justinian, a violation
of that law. An eastern invasion was not a breach of the treaty between
Vandals and Byzantines; the emperors armies were simply the restorers
of right order in the Vandal kingdom. In seeking to undermine the
consensus of the Vandal elite as to the legitimacy of their king, Justinian
was said to have recognized the role of Geiseric and his law in defining
legitimacy in much the same way that the Romano-Africans had for the
better part of a century (see above, Chapter 3.2). Still more remarkable,
though, was the assumption Procopius attributed to Justinian (which
the emperor apparently hoped that the Vandal elite would share) that as
emperor of the Romans he was himself the ultimate arbiter of questions
pertaining strictly to the legitimacy of Vandal succession. In Procopius
retelling, Justinian naturally also promised to bring peace and freedom
to Africa ( ,2 /  ) +
) again, a euphemism for imperial
19
20
21
23
24

In addition to the specific references below, see Proc. BV 1.20.2, 1.20.18, 1.20.224, and 2.4.38,
1:396, 1:399400, and 1:4323.
W. Goffart, Barbarians and Romans, ad 418584: The Techniques of Accommodation (Princeton, NJ,
1980), pp. 427.
22 Proc. BV 1.17.910, 1:3867.
Proc. BV 1.16.18, 1:3813.
Sullecthum: Proc. BV 1.17.6, 1:386; see also ibid., 1.17.8, 1:386 (Leptis and Hadrumetum).
Carthage: ibid., 1.21.910, 1:4012.
See in general A. Kaldellis, Procopius of Caesarea: Tyranny, History, and Philosophy at the End of
Antiquity (Philadelphia, Pa., 2004), pp. 11864.

311

Staying Roman
rule but the question of religion and the fact that Africa used to be a
province of the Roman empire were never broached.25
If we can trust Procopius on this count, Justinians claims to be intervening in African affairs so as to put to rights a crisis of Vandal power
may well have been addressed to Romano-Africans as much as it was
to Vandals. Alone of all our sources for the Byzantine reconquest the
chronicler Victor of Tonnena had lived under the Vandal kings, and perhaps significantly in his account the invasion was discussed in terms of
capture (capere) rather than recovery (recipere or restituere), the phraseology employed by Marcellinus comes and Marius of Avenches.26 To be
sure, Victor viewed Gelimer as a tyrannus, an illegitimate ruler; but the
chronicler never applied that word to the usurpers predecessors, whom
he designated with the more dignified title of king (rex).27 This suggests
to me that even Victor, a Nicene bishop to whom the replacement of
an Arian regime with one of his own confession must have been a welcome development, accepted the fundamental legitimacy of Vandal rule
as such. In Victors mind, what made Gelimer a tyrant was not that he was
a Vandal but that he had seized power from the king who rightfully held
it according to the law of succession Geiseric had established in the fifth
century. Once again it would seem that Vandals and Romano-Africans
shared a set of common beliefs and assumptions.
Ultimately, of course, Procopius History of the Wars is a literary work.
The primary audience of his discussion both of Justinians fastidious concern with the laws of Vandal succession and of Belisarius protestations as
to the Romanness of the Africans was the historians own Greek-speaking
East Roman audience, who were expected to evaluate for themselves the
justness of each actors claims.28 But these concerns nonetheless do seem
to capture something of the uncertain reality of Africa in the 530s, caught
as it was between the Vandal and the imperial regimes. Critically, Vandal
power did not collapse in Africa primarily because of Romano-African
discontent with Vandal rule at least not in the kingdoms heartland. As
the career of the Moorish leader Antalas demonstrates, such discontent
does seem to have ended effective Vandal control over portions (at least)
of the kingdoms pre-desert frontier zone in the sixth century (see above,
25

26
27

28

Proc. BV 1.9.1013 and 1.9.1519, 1:3534 (Justinian to Gelimer); ibid., 1.9.203, 1:3545
(Gelimer to Justinian); ibid., 1.16.1314, 1:384 (Justinian to the Vandals). Peace and freedom:
ibid., 1.16.14, 1:384.
Vict. Tonn. s.a. 534.118, p. 38; see also Isidore, Historia Vandalorum 834, pp. 3068.
Vict. Tonn. s.aa. 531.115, 533.117, and 534.118, pp. 378 (tyrannus); see also ibid., s.aa. 455.1415,
464.28, 466.30, 479.50, 497.78, 523.106, 534.118, 566/7.173, pp. 7, 11, 16, 24, 34, 38, and 54
(rex, of each of the Vandal kings, including Gelimer).
Cf. the speeches analysed by Kaldellis, Procopius of Caesarea, pp. 2932.

312

The dilemma of dissent

Chapter 5.3). The fall of the Vandal kingdom was also immediately preceded by rebellions on the part of the governors of two of the kingdoms
furthest-flung provinces: Sardinia and Tripolitania. Both appealed to Justinian for assistance. In the first instance, though, Procopius depicts the
revolt as a case of personal ambition, for Godas, the rebel governor of
Sardinia, proved to be unwilling to accept the practical constraints either
of Vandal or of Byzantine sovereignty.29 The second revolt may have
been connected to Moorish raiding in Tripolitania, against which Vandal military might appears to have been desperately ineffective.30 In the
central provinces of the Vandal kingdom, however, Procopius account
reveals at least a measure of continued Romano-African support for and
collaboration with the Vandal regime even after the arrival of the imperial
army in Africa and its triumphal entry into Carthage. Most dramatically,
Belisarius had a Carthaginian citizen named Laurus publically impaled
for conspiring with the Vandals, presumably to betray the metropolis to
them again.31 African peasants were also willing to kill a number of the
attendants and slaves who accompanied the same generals expeditionary
force and thus risk Byzantine reprisals in order to secure the bounty
Gelimer offered on the heads of imperial troops.32 In some minds, at
least, the Vandal regime remained the legitimate authority in Africa,
even when it was no longer the ascendant military power.
Even so, by the 530s, the delicate balancing act of legitimating their
rule was becoming more difficult for Vandal kings to maintain. More
than any emperor before him, Justinian seems to have been fired by the
idea of restoring the western provinces to the empire; but, even so, with
the emperors guest-friend Hilderic on the throne the last reigning
scion of the house of Theodosius Africa was probably secure. Indeed,
in the wake of the Byzantine reconquest, it is easy to forget how close
relations between Africa and the empire must have been in the 520s. As
we have seen, Hilderic suspended the Vandal Arianizing policy shortly
after succeeding to the throne, perhaps on his own initiative but perhaps
as a concession to the zealous Justinian, already the power behind the
throne (see above, Chapter 3.3.1). Moreover, at some point, presumably
in the first four years of his reign, Hilderic had even gone so far as to place
the bust and legend of the emperor Justin on the Vandal silver coinage.33

29
30
31
32
33

Proc. BV 1.10.2534, 1:35960.

Proc. BV 1.10.224, 1:359; Malalas, Chronographia, p. 459; and Coripp. Ioh. 3.184264, pp. 547.
Proc. BV 2.1.8, 1:420; see also ibid., 1.20.12 and 2.1.4, 1:396 and 1:419.
Proc. BV 1.23.14, 1:407.
Hahn, Moneta Imperii Byzantini 1, pl. 42.9 (DN IVSTINVS PPAVG) and pl. 42.11 (DN
IVSTINVS PPA). Hilderic also issued silver coins in his own name: ibid., pls. 42.8 and 42.10

313

Staying Roman
Gelimers claim that Hilderic intended to betray the kingdom into Justins
hands must have seemed eminently plausible to a great many members
of the Vandal ruling class.34 The military weakness of Hilderics regime
when confronted with major Moorish raids that ravaged the kingdoms
southern provinces further undermined the kings credibility.35 However,
it was Gelimers subsequent seizure of the royal power in a palace coup
that precipitated the crisis of legitimacy that ultimately led to the downfall
of the Vandal kingdom.
On the face of it, our two best sources for Justinians reaction to this
coup Procopius and Malalas seem to differ entirely over the stance
that Gelimer adopted with respect to the emperor. In Malalas retelling,
the new king sent an ambassador to Constantinople bearing diplomatic
gifts.36 Procopius says nothing about this exchange, claiming instead
to quote the letters from Justinian to Gelimer discussed above. In his
response to these communications the Vandal king revealed himself to
be neither pliant nor predictable. Rather, he was said to have addressed
Justinian as one basileus to another, even having the audacity to place his
own name in the privileged first position in the salutation, relegating the
emperor to second place, a position of rhetorical inferiority.37 The two
accounts can perhaps be reconciled, if the events described by Malalas
preceded those discussed by Procopius; but, in any case, both historians
agree that the emperor was in the end enraged by the Vandal kings
coup.38 Both also indicate that Hilderics supporters travelled to Constantinople to intercede with Justinian on their deposed kings behalf.39 If
the emperor did count Hilderic among his guest-friends, he may in part
have felt the coup to be a personal affront. Gelimer was also a committed Arian, and if his regime showed anything like the signs of growing
self-confidence and autonomy that Procopius attributes to it, Justinian
may have been worried about a resurrection of the Vandal Arianizing
policy.40 As we have seen, highly placed members of the African e migre
community in Constantinople were furthermore said to have had the
emperors ear, which they were filling with descriptions of the wealth
of their province (see above, Chapter 2.1.3); and in any case Justinian
may have been concerned about the continued flow of African grain

34
36
37
38
39
40

(DN HILDIRIX REX). Under Gelimer, silver coins seem to have been stuck only in the name
of the Vandal king: ibid., pl. 42.12 (DN REX GEILAMIR).
35 Coripp. Ioh. 3.178264, pp. 547.
Proc. BV 1.9.8, 1:352.
Malalas, Chronographia, p. 459.
Proc. BV 1.9.924, 1:3525, esp. ibid., 1.9.20, 1:354: K)G $)
 'I  K)".
Proc. BV 1.9.24, 1:355 and Malalas, Chronographia, p. 459.
Proc. BV 2.5.78, 1:440 and Malalas, Chronographia, p. 459.
Committed Arian: Proc. BV 2.9.14, 1:458.

314

The dilemma of dissent

and goods to the markets of the eastern Mediterranean under Gelimer.41
Moreover, to a ruler with ambitions of restoring the empire of Honorius,
Africa was the key to the West. In this sense, the fact that Gelimer had
come to power through a coup was fortuitous for the eastern autarch.
Although the new African king seems to have commanded genuine support among the Vandal elite themselves, in Constantinople, Justinian was
able plausibly to dismiss Gelimers succession to the Vandal throne as
illegitimate.
Potentially just as troubling from Gelimers point of view was the
fact that he was unable to convince any of the other powers in the
western Mediterranean to do anything on his behalf. OstrogothicVandal
relations had soured over the affair of Amalafrida. In addition, Malalas
indicates that the emperor now pressured the Ostrogoths not to recognize
Gelimers succession; for internal political reasons, too, the Gothic queen
Amalasuntha decided to back Justinian.42 While it is harder to know how
the Visigothic court would have reacted had the Vandal envoys arrived
in Spain before the fall of Carthage to imperial forces, the track record
of relations between the two kingdoms was not inspiring (see above,
Chapter 1.2). The rebellions of Tripolitania and Sardinia against Vandal
authority further weakened Gelimers position. All in all, Africa must
have appeared to Justinian as something of an autumn pomegranate ripe
for the plucking. And, indeed, the province fell quickly to the Byzantine
army. Gelimer had been caught off guard; much of his army was away,
recovering Sardinia.43 The Vandal forces were decisively defeated in only
two major encounters, and within a few months imperial control of
North Africa had been re-established.44
Gelimers usurpation of the Vandal throne was not forgotten after the
collapse of his kingdom. By the mid sixth century, however, the role of
the Vandal law of succession in justifying the Byzantine intervention in
African affairs seems only to have interested Procopius, to whom tyranny
was a key intellectual concern. Other authors were more interested in
the role of the emperor in the restoration both of Roman power and
of Nicene orthodoxy, themes that Justinian himself had earlier sounded
in his legislation in order to explain his decision to go to war. Authors
throughout the empire accepted that God and the saints had played
41

42
43
44

Flow of grain: P. Reynolds, Trade in the Western Mediterranean, ad 400700: The Ceramic Evidence,
BAR International Series 604 (Oxford, 1995), p. 114. On Justinians motivations in general, see
also A. Merrills and R. Miles, The Vandals (Chichester, 2010), pp. 22930.
Malalas, Chronographia, pp. 45960; Proc. BV 1.14.56, 1:3734; see also above, Chapter 1.2.
Proc. BV 1.11.22, 1:363; for Godas rebellion in Sardinia, see ibid., 1.10.267, 1:359.
Proc. BV 1.18.11.19.33, 1:38896 (Ad Decimum) and ibid., 2.2.12.3.28, 1:42332
(Tricamarum).

315

Staying Roman
a special role in the fall of the Vandal kingdom, in a war that was
generally acknowledged throughout the Mediterranean as a reconquest
of formerly Roman territory. However, Victor of Tonnenas discussion
of the invasion in terms of capture raises the important question of how
successful Justinian was in terms of legitimating the Byzantine occupation
to a Romano-African audience.
2. african resistance
Writing about fifth-century Africans living before the time of the Vandal
invasion, Christian Courtois observed in 1955 that in an obscure recess
of their souls, there was something that said no to the Empire.45 The
comment was taken up by R. A. Markus nearly twenty years later and
applied to African Christianity as a whole:
From its beginnings until almost the moment of its submergence African Christianity is marked by the same quality of intransigence, a jealous sense of its
independence and a peculiar identity over against the churches across the sea,
against the empire and the secular world as a whole, and against the pseudoChurch which is their creature . . . Dissent, we may say, was its instinctive
posture in society.46

This rejection of imperial and perhaps more importantly, at least as

far as Markus was concerned papal interference in the affairs of the
African church was most strikingly highlighted in the Byzantine period
by the rejection on the part of the African bishops of the condemnation of the Three Chapters.47 The affair focused on the writings of
Theodore of Mopsuestia, Theodoret of Cyrrhus, and Ibas of Edessa,
declared orthodox by the Fourth Ecumenical Council at Chalcedon in
451, but denounced by Monophysites in the sixth century as tainted with
Nestorianism. In an effort to conciliate these Monophysites and secure
their reconciliation with the orthodox church, in ad 5435 Justinian
issued an edict demanding the condemnation of the three theologians
writings throughout the empire; a move that was vehemently opposed
in the western provinces of Africa and Italy. Indeed, opposition to Justinians edict was so strong in Africa that it has led Averil Cameron
45
46

47

Courtois, Vandales, p. 148: dans un obscur recoin des a mes, il y avait quelque chose qui disait
non a` lEmpire.
R. A. Markus, Christianity and Dissent in Roman North Africa: Changing Perspectives in
Recent Work, in D. Baker (ed.), Schism, Heresy and Religious Protest, Studies in Church History
9 (Cambridge, 1972), pp. 345.
R. A. Markus, Reflections on Religious Dissent in North Africa in the Byzantine Period, in
G. J. Cuming (ed.), Papers Read at the Third Winter and Summer Meetings of the Ecclesiastical History
Society, Studies in Church History 3 (Leiden, 1966), pp. 1409.

316

The dilemma of dissent

to conclude that the general effect of the condemnation of the Three
Chapters was to turn the African church into a nationalistic organ and
to cause it to reflect on the dubious benefits which liberation from the
Arian Vandals had brought.48 The Three Chapters Controversy in the
540s and 550s, the re-emergence in our sources of references to Donatism
in the 590s, and the staunchly anti-Monothelete position adopted by the
African church in the 640s all seem, on the face of it, to have placed the
African church (or portions of it) in opposition to the imperial and papal
authorities in the Byzantine period.
Did Africans therefore resent or resist the Byzantine empire? The
terms in which these debates were framed tell against such a conclusion.
The affair of the Three Chapters is the last such controversy for which
anything like a substantial number of sources survive in which several
different Africans themselves expressed their own independent positions.
In the 590s and early 600s, we have only Pope Gregorys perspectives
on and responses to the state of affairs in North Africa, and for the
Monothelete controversy our most important sources cluster around the
person of Maximus Confessor, an eastern monk and sometime sojourner
in seventh-century North Africa. Leaving the Monothelete controversy
aside for a moment, then, throughout the sixth century and into the
seventh, the African church consistently rejected the attempts of the
emperor and the Pope to arrogate to themselves spiritual authority that,
from an African point of view, they did not possess.49 Neither doctrine
nor practice could be dictated by the emperor or the Pope or, for
that matter, by any patriarch. The African position was not, however,
essentially rejectionist. Both the Pope and the emperor could serve as a
court of appeal. But, in their writings and actions, African churchmen
of the sixth century betray a perception that ultimate spiritual authority
lay within the orthodox church as a whole when gathered in council,
whether at a provincial or ecumenical level.
2.1. The Three Chapters
From the very beginning, the most serious concern shared by the African
ecclesiastical hierarchy with respect to the condemnation of the Three
Chapters had been the threat this posed to Chalcedonian orthodoxy.
48

49

Cameron, Byzantine Africa, p. 47. On the controversy over the Three Chapters, see C. Chazelle
and C. Cubitt (eds.), The Crisis of the Oikoumene: The Three Chapters and the Failed Quest for Unity
in the Sixth-century Mediterranean (Turnhout, 2007), esp. Y. Moderan, LAfrique reconquise et
les Trois Chapitres, ibid., pp. 3982; and M. Maas, Exegesis and Empire in the Early Byzantine
Mediterranean, Studien und Texte zu Antike und Christentum 17 (Tubingen, 2003), pp. 4264.
Markus, Religious Dissent in North Africa in the Byzantine Period, p. 146.

317

Staying Roman
Writing in 544 or 545, Bishop Pontianus of Thaenae in Byzacena (mod.
Henchir Thina) praised Justinians piety, justice, and faith, but worried
that the condemnation of the Three Chapters would see the revival of
the Eutychian heresy, an extreme form of Monophysitism that had been
rejected as heretical at Chalcedon; and, worse, that the condemnation
could lead to a clash and persecution for which Justinian would have to
answer on the Day of Judgement. That conflict would be all the more
pointless for the fact that Theodore, Theodoret, and Ibas had all long
since died, and their condemnation could not now serve the purpose
of correction, for they could no longer recant any heretical beliefs they
may have held. As they were now in the hands of the Eternal Judge, all
would-be mortal judges had best leave well enough alone.50
Pontianus was concerned, too, that the works of Theodore, Theodoret,
and Ibas were little known in Africa. Should their works make it to
the western province, the African bishops would be only too happy
to read them and determine if they were heretical: we can consider
statements, not condemn already-dead authors of statements with a precipitous verdict.51 This concern was apparently to remain unaddressed
until c.560, when a Carthaginian deacon named Liberatus compiled a
brief explaining in Latin the essentials of the Nestorian and Eutychian
heresies for the benefit of those who were still uninformed as to the
issues at stake.52 But the African theologians who wrote in defence of
the Three Chapters seem to have understood those issues well enough.
In a letter to the Roman deacons Pelagius and Anatolius written some
time before his own death in 545/6, the Carthaginian deacon Ferrandus was quite clear that what was at stake in the controversy of the
Three Chapters was nothing less than the fate of Chalcedonian Christianity. Ferrandus adopted the essentially conservative position that The
whole Chalcedonian council is true because it is the whole Chalcedonian council: no part of it merits any censure.53 Informing much of
Ferrandus discussion, though never made explicit, is the fact that the
Council of Chalcedon had condemned Nestorianism; and, more than
that, so far from condemning Ibas of Edessa as a Nestorian heretic, the
representatives of the universal church gathered at Chalcedon declared
Ibas orthodox and even went so far as to reinstate him to the see from
50
51
52
53

Pontianus, Epistola ad Justinianum imperatorem de tribus capitulis, PL 67:9958.

Ibid., col. 997: dicta possumus respicere, non auctores dictorum jam mortuos praecipiti condemnatione damnare.
Liberatus, Breuiarium causae Nestorianorum et Eutychianorum, in ACOec. 2/5:98141, esp. ibid., 1,
pp. 989.
Ferrandus, Ep. 6.3, PL 67:923c: Totum concilium Chalcedonense, cum est totum concilium
Chalcedonense, verum est: nulla pars illius habet ullam reprehensionem.

318

The dilemma of dissent

which he had been deposed. The council had made an unambiguous
declaration about Ibas non-Nestorianism, which the condemnation of
the Three Chapters sought to overturn. If Ibas was to be condemned
as a Nestorian 100 years on, Ferrandus reasoned, then any of the other
decisions reached at Chalcedon could be called into question, too, and
the authority of the council would begin to unravel. This was all the
more troubling to Ferrandus because the fathers gathered at Chalcedon
had reached a unanimous decision, thus fulfilling Pauls admonition that
there be no divisions within the fellowship of all believers (1 Cor. 1:10); a
condition that the condemnation of the Three Chapters manifestly failed
to meet.54
It did not end there. Overthrowing Chalcedon opened the door to
overthrowing Nicaea,55 whose definition of the faith Ferrandus and,
perhaps more importantly, his hero, Fulgentius of Ruspe had undergone
exile under the Vandals to defend.56 What shall be lasting, Ferrandus
asked the Roman deacons, if what the Chalcedonian council established
is called into question?57
The spectre of Eutychian Monophysitism and the threat to Chalcedonian orthodoxy were to haunt African objections to the condemnation
of the Three Chapters in the decades that followed. In letter after letter
to the Roman deacon Anatolius, to the priest Eugippius, to Severus
scholasticus in Constantinople Ferrandus defended Christs place within
an indivisible Trinity, his oneness of person, and his two natures and
two substances, human and divine.58 But the most forceful expression
of African opposition to Justinians policy came from the pen of Bishop
Facundus of Hermiane, who wrote to the emperor at great length (twelve
books at 396 pages in the modern Corpus Christianorum edition) to convince him of the error of the condemnation of the Three Chapters.
Though he worked out the theological case in greater detail than any
of his contemporaries, the issues at stake for Facundus in the months
around January 548 were substantially the same as those that concerned
the rest of the African ecclesiastical hierarchy: the dual danger of the
54
56

57
58

55 Ferrand. Ep. 6.7, col. 926.

Ferrand. Ep. 6.5, cols. 9245.
V. Fulg. prologus, p. 9 (Ferrandus) and ibid., 1726, pp. 87123 (Fulgentius). Note, however,
that the attribution of Vita S. Fulgentii to Ferrandus has recently been called into question:
A. Isola (trans.), Vita di San Fulgenzio, Collana di testi patristici 65 (Rome, 1987), pp. 58 and
C. Leyser, A Wall Protecting the City: Conflict and Authority in the Life of Fulgentius of
Ruspe, in A. Camplani and G. Filoramo (eds.), Foundations of Power and Conflicts of Authority
in Late-Antique Monasticism, Proceedings of the International Seminar, Turin, December 24, 2004,
Orientalia Lovaniensia Analecta 157 (Louvain, 2007), p. 177.
Ferrand. Ep. 6.5, col. 925b: Quid erit firmum, si quod statuit Chalcedonense concilium vocatur
in dubium?
Ferrand. Epp. 35, PL 67:889921.

319

Staying Roman
Eutychian and Nestorian heresies on the one hand, and the overturning
of the Council of Chalcedon on the other.59
Indeed, the ultimate confrontation between the emperor and the
African church was precipitated not so much by the latters nationalistic
reflection on the dubious benefits of Byzantine rule as by the two parties
mutual suspicion of the others orthodoxy. On 11 April 548, a reluctant
Pope Vigilius acceded to the condemnation of the Three Chapters in
a statement known as the Iudicatum. All five patriarchs had now come
down on the side of Justinians redefinition of the faith; but, in Africa,
Vigilius was condemned by hard-line Chalcedonians as a prevaricator.60
Although it seems that the Pope had attempted to safeguard the decisions
of Chalcedon even as he condemned the Three Chapters, Ferrandus earlier arguments about the inviolability of the whole council seem to have
carried the day in Africa. In 550, a synod of African bishops excommunicated Vigilius, sending a letter presumably Facundus monumental
In Defence of the Three Chapters, perhaps already in a second edition to
Justinian in defence of the condemned theologians.61
Critically, in terms of assessing the African reaction to the emperors
theological policies, Justinian had already established his good faith with
the local Nicene episcopate in the immediate aftermath of the Byzantine
reconquest. To be sure, the emperor seems initially to have pursued a
policy of religious accommodation in Africa that sought to ensure that
Arian clergy who converted to the Nicene confession would retain their
status in the ecclesiastical hierarchy and similarly limited the pace at which
property would be transferred from the Arian to the Nicene church.
Perhaps not surprisingly, though, this policy proved unpalatable to what
remained of Africas Nicene bishops as they set about re-establishing
their dominance locally. In 535, they rallied the support of the Pope and
protested to the emperor.62 Justinian bowed to western pressure: later
that same year, the emperor wrote to the Praetorian Prefect Solomon,
formally and thoroughly reversing the Arianizing policy of the Vandal
59
60
61

62

Facundus of Hermiane, Pro defensione trium capitulorum, ed. J.-M. Clement and R. Vander Plaetse,
in Opera omnia, CCSL 90A (Turnhout, 1974), pp. 3398.
Vict. Tonn. s.a. 557.157, p. 51 and Facundus of Hermiane, Epistula fidei catholicae in defensione
trium capitulorum 7, ed. Clement and Vander Plaetse, CCSL 90A:420.
Vict. Tonn. s.a. 550.141, p. 46. Victors chronology is, however, notoriously unreliable. He
mentions the letter to Justinian and Facundus Pro defensione separately, though he notes that this
is when Facundus work shone brightly: Eo tempore duodecim libri Facundi Hermianensis
ecclesiae episcopi refulsere (ibid., 550.142, p. 46). Second edn: E. Chrysos, Zur Datierung und
Tendenz der Werke des Facundus von Hermiane, Kleronomia 1 (1969), pp. 31419.
Collectio Avellana 858, pp. 32838; and Just. Nov. 37 (ad 535), pp. 2445, esp. ibid., 37.1,
p. 244. C. Sotinel, Emperors and Popes in the Sixth Century: The Western View, in M. Maas
(ed.),The Cambridge Companion to the Age of Justinian (Cambridge, 2005), p. 277 and Merrills and
Miles, Vandals, pp. 24851.

320

The dilemma of dissent

kings. All of the property that had been confiscated from the Nicene
church over the course of the previous century was to be returned;
heretics (now understood as non-adherents of the Nicene confession)
were to be excluded from public office; and, indeed, Arians, Donatists,
Jews, and pagans were not even to be allowed to celebrate their religious
rites a mandate which Procopius claimed was intensely resented by
Arians serving in the imperial army and thus helped precipitate Stotzas
rebellion in the spring of the following year.63 Over the following decade,
Justinian continued to show his support for the customs of the African
ecclesiastical establishment. In 541, the emperor wrote to the bishops
of Byzacena and their primate Datianus, confirming the privileges of
their church; a year later, Justinian again wrote to Datianus, this time
indicating that he would back whatever privileges the authority of the
councils and tradition granted to the metropolitan of Carthage and the
primates of Numidia and Byzacena.64 These communications hint at
internal tensions within the African ecclesiastical hierarchy; but more
importantly for our purposes they lay the groundwork for the debate
surrounding the Three Chapters. Despite their initial disagreement over
the shape that NiceneArian reconciliation should take, Justinian had
proven himself willing to bend to the bishops demands, and had emerged
from the exchange as a strong supporter of the Nicene church. When
the orthodoxy of the emperors theological policy was later called into
question as it was repeatedly in the mid sixth century the African
episcopate seems to have believed that correct faith could be debated in
familiar terms and from shared assumptions that both sides understood.
Justinians response to the debate over the Three Chapters sought to
safeguard what he saw as an orthodoxy already assured by the unanimity of the Christian patriarchs. On 23 December 551, Justinian issued
an edict specifically answering Facundus arguments.65 But Justinian was
also an enthusiastic persecutor of all those who did not subscribe to
correct Christian doctrine, including not only heretical Christians, but
Jews and Samaritans as well as pagans. Like the Vandal kings, Justinians
efforts to bring the African church to heel targeted their leadership. In
551, the emperor summoned four of the most prominent African bishops
to Constantinople to answer for their actions: Archbishop Reparatus of

63
64
65

Just. Nov. 37 (ad 535), pp. 2445 and Proc. BV 2.14.1121, 1:4845; W. Kaegi, Arianism and
the Byzantine Army in Africa 533546, Traditio 21 (1965), pp. 2353.
Just. Nov. App. 2 (ad 541) and App. 3 (ad 542), pp. 7967.
W. Pewesin, Imperium, Ecclesia universalis, Rom: Der Kampf der afrikanischen Kirche um die Mitte des
6. Jahrhunderts, Geistige Grundlagen romischer Kirchenpolitik, Forschungen zur Kirchen- und
Geistesgeschichte 11 (Stuttgart, 1937), pp. 13941.

321

Staying Roman
Carthage; Firmus, the primate of Numidia; and Primasius of Hadrumetum and Verecundus of Iunca, both from Byzacena. Justinian probably
intended that the then-primate of Byzacena, Boethius, be present rather
than Primasius and Verecundus; but by 551 Boethius was very probably
unable to travel. The bishop must certainly have been extremely old by
the standards of late antiquity, and he was to die in the coming year or so.
Primasius, next in line for the primacy of Byzacena, was probably sent
in Boethius stead.66 The Numidian primate Firmus was brought round:
he helped represent the African church at the Second Council of Constantinople in 553 and was sent back to Africa with the emperors gifts.67
Initially, however, the other three stood firm, and their punishments varied in proportion to their importance. Reparatus was deposed from the
metropolitan see and exiled to Euchaita (mod. Beyozu, Turkey). He was
replaced by the more compliant Primosus, Reparatus representative at
the imperial court (apocrisarius) and a deacon of the Carthaginian church.
Verecundus was imprisoned at the church of St Euphemia in Chalcedon,
where he died. Primasius was confined to the monastery of the Sleepless
Monks at Eirenaion, also on the eastern shore of the Bosporus. By 554,
however, after word of Boethius death reached Constantinople, Primasius was convinced to condemn the Three Chapters in order to assume
the primacy of Byzacena.68
Justinians policy seems to have had the intended effect. Although Firmus was said to have died on his way back to Africa and the council
of bishops in Byzacena condemned Primasius prevarication, by 555,
the councils of Africa Proconsularis and Numidia seem largely to have
been in communion with the new archbishop of Carthage, Primosus.69
Opposition, of course, continued. Primosus was said to have beaten,
imprisoned, and exiled the recalcitrant.70 Facundus of Hermiane, the
mouthpiece of the African church, continued to write in support of
the Three Chapters though he had gone into hiding possibly as early as
66

67

68
69

70

Boethius: see PCBE 1:146, s.n. Boethos, which suggests that the primate may well have been
one of the bishops sent into exile on Sardinia by Thrasamund in 508/9, in which case Boethius
would have been a bishop for at least forty-two years by 551. Primasius as heir apparent: Vict.
Tonn. s.a. 552.145, pp. 478.
ACOec. 4/1:4 (l. 22), 21 (l. 18), 33 (l. 22), 40 (l. 18), and 204 (l. 19); Vict. Tonn. s.a. 552.145,

p. 47; and J. L. Maier, LEpiscopat

de lAfrique romaine, vandale et byzantine, Bibliotheca Helvetica
Romana 11 (Rome, 1973), p. 78.
Vict. Tonn. s.a. 552.145, pp. 478; Vigilius, Constitutum de tribus capitulis = Collectio Avellana
83.311, p. 319.
Vict. Tonn. s.aa. 552.145, pp. 478 (death of Firmus and condemnation of Primasius); ibid.,
554.149 and 555.152, pp. 4950 (Africa Proconsularis and Numidia in communion with Primosus).
Vict. Tonn. s.a. 556.155, p. 51.

322

The dilemma of dissent

550.71 In 553, Felix, the abbot of the African monastery of Gillensis, was
exiled to the Thebaid. Two years later, the bishops Victor of Tonnena and
Theodore of Cebarsussi were also sent into exile in Egypt. Indeed, Victor
claimed that by 555 he had already endured imprisonment throughout the
Mediterranean, from the monastery of Mandracium in Carthage to
the Balearic Islands to Alexandria. In 565, he and Theodore, along with
the African bishops Musicus, Brumasius, Donatus, and Chrysonius, were
summoned to Constantinople where they debated the condemnation of
the Three Chapters with the Patriarch Eutychius before being confined
to separate monasteries throughout the imperial capital.72 But, as R. A.
Markus has observed, one cannot but be struck by the apparent ease with
which so many bishops the primate of Numidia among them allowed
imperial threats, bribes, or pressure to secure their compliance.73 Perhaps
towards 560, the Carthaginian deacon Liberatus wrote that many bishops
had been bribed to condemn the Three Chapters, and that the rest had
fled or been sent into exile.74 But the dissenting bishops form only a
small fraction of the literally hundreds of bishops throughout late antique
North Africa. In the fifth and early sixth centuries, when the Vandal
kings had tried to impose Arianism on the region, African sermons were
full of denunciations of the barbarians and exhortations to resist the pressures to convert. The homiletic literature attributable to the Byzantine
period consists only of a handful of sermons, but one searches them in
vain for a similar approach to imperial policy.75 In fact, by the later sixth
century the issue appears to have been settled, at least as far as Africa
was concerned. In 565, Justin II seems to have rehabilitated hard-line
Chalcedonian bishops exiled for their opposition to the condemnation
of the Three Chapters.76 Three or four years later, Facundus, unbending to the end, published his Letter of the Catholic Faith in Defence of the
Three Chapters, reviling Primasius as a semi-Eutychian and condemning
71

72
73
74
75

76

Facundus notes that he was in hiding in the first chapter of his Contra Mocianum scholasticum, ed.
Clement and Vander Plaetse, CCSL 90A:401. On the date, see Clement and Vander Plaetses
introduction to the edition at p. xii and Chrysos, Zur Datierung, 31922.
Vict. Tonn. s.aa. 553.147, 555.153, 556.156, 564/5.169, and 566/7.173, pp. 4951 and 534; for
the death of Felix of Gillensis at Sinope, Vict. Tonn. s.a. 557.158, p. 51.
Markus, Religious Dissent in North Africa in the Byzantine Period, p. 144.
Liberatus, Breviarium 24.
Pseudo-Augustine, In natali domini V (Caillau-Saint-Yves 1.11), PLS 2:92931; pseudoAugustine, In natali domini VIII (Caillau-Saint-Yves 1.14), PLS 2:9356; pseudo-Augustine,
In natali martyris Vincentii II (Caillau-Saint-Yves 1.48), PLS 2:10057; Sermo de natiuitate domini,
ed. H. Barre, PLS 4:19936; pseudo-Augustine, Sermo 195, PL 39:210710. For the (possible)
attribution to Byzantine Africa, see L. Dossey, Christians and Romans: Aspiration, Assimilation,
and Conflict in the North African Countryside, Ph.D. diss., Harvard University (1998), pp. 369
and 3712.
E. Stein, Histoire du Bas-empire, ed. J.-R. Palanque, 2 vols. (Paris, 194959), 2:681.

323

Staying Roman
Vigilius and his successor Pelagius as prevaricators; but that is the last we
hear from any African on the subject.77 The Three Chapters were still
a live topic in the pontificate of Gregory the Great (ad 590604), but,
when Gregory wrote on the subject, his letters were directed to Italians,
Isaurians, and Egyptians; not, apparently, to Africans.78
2.2. Pope and Emperor
From his earliest defence of the Three Chapters, but hardening and
intensifying over his twenty years or so in hiding, Facundus had argued
that the emperor should leave the business of theology to the bishops.79
In the late sixth century, Gregory himself seems to have encountered a
similar resistance on the part of the Numidian bishops to papal interference in the business of the provincial council.80 At several points in his
pontificate Gregory attempted to intervene in the affairs of the Numidian church to address concerns of bribery, corruption, and other abuses,
as well as the spectre of Donatism.81 In 593, he even went so far as to
ask the exarch Gennadius to reverse the un-canonical decisions of the
77
78

79

80

81

Facundus of Hermiane, Epistula fidei catholicae, esp. c. 7, p. 420; Stein, Histoire du Bas-empire,
2:682.
Greg. Ep. 2.43, 1:1312 (to uniuersis . . . apparently addressing an Italian audience); ibid., 4.2,
4.3, and 4.37, 1:21820 and 1:2579 (to Bishop Constantius of Milan); ibid., 6.65, 1:441 (to the
Isaurian monk and priest Athanasius); ibid., 7.31, 1:493 (to Patriarch Eulogius of Alexandria); see
also ibid., 9.148, 2:7002 (to the monk Secundinus, of unknown provenance).
Facundus, Pro defensione trium capitulorum 12.3, pp. 3819. Markus, Religious Dissent in North
Africa in the Byzantine Period, p. 148, argues that Facundus emphasis on the primacy of
conciliar authority is already hinted at in the writings of Ferrandus.
R. A. Markus, Donatism: The Last Phase, in C. W. Dugmore and C. Duggan (eds.), Papers
Read at the First Winter and Summer Meetings of the Ecclesiastical History Society, Studies in Church
History 1 (Leiden, 1964), pp. 11826 and R. A. Markus, Country Bishops in Byzantine Africa,
in D. Baker (ed.), The Church in Town and Countryside, Studies in Church History 16 (Oxford,
1979), pp. 115.
Greg. Ep. 2.39, 3.478, 4.35, 6.36, 8.14, 12.3, 12.89, 1:1257, 1:1914, 1:2556, 1:41011,
2:5323, 2:971, and 2:97982. Some of the abuses that troubled Gregory would probably have
troubled the Carthaginian church as well: Ferrandus, Breviatio canonum, esp. 33, 85, and 121,
pp. 290, 294, and 297. The problem of the survival of Donatism in sixth-century Africa is
vexed: see, inter alia, W. H. C. Frend, The Donatist Church: A Movement of Protest in Roman North
Africa (Oxford, 1952), pp. 30014; W. H. C. Frend, Donatist and Catholic: The Organization of
Christian Communities in the North African Countryside, in Cristianizzazione ed organizzazione
ecclesiastica delle campagne nellalto medioevo: espansione e resistenze, 2 vols., Settimane di studio del
Centro italiano di studi sullalto medioevo 28 (Spoleto, 1982), 2:60134; Markus, Donatism:
The Last Phase; Markus, Christianity and Dissent; and R. A. Markus, The Problem of
Donatism in the Sixth Century, in Gregorio Magno e il suo tempo: XIX Incontro di studiosi

dellantichit`a cristiana in collaborazione con lEcole

francaise de Rome, Roma, 912 maggio 1990, Studia
ephemeridis Augustinianum 33, 2 vols. (Rome, 1991), 1:15966. It is not entirely true that
concern with Donatism was confined to Numidia in the Byzantine period: Ferrandus, Breviatio
canonum 50, 18991, and 193, pp. 291 and 303.

324

The dilemma of dissent

Numidian council.82 By October of 596, Gregory was confronted with
the complaints of his agent in Numidia, the local bishop Columbus, that
he had incurred the enmity of many of his fellow bishops because of his
frequent correspondence with the Pope.83
The extent to which this indicates that Numidia was falling out of
the Roman orbit in the late sixth century is, however, an open question.
Already in the first year of Gregorys pontificate there are implications that
something had happened to cause a falling-out between Rome and the
Numidian church.84 This falling-out seems not to have been of Gregorys
making. In a letter to the Numidian bishops written in August of 591,
Gregory refers to a request they had made to his predecessor, Pelagius II,
that he confirm all of the ancient customs of Numidia. It would seem that
Pelagius had not yet done so, for Gregory took it upon himself to answer
the request, granting the Numidian bishops what they had asked with
the reservation that the primacy of the ecclesiastical province was not to
be held by former Donatists.85 Gregorys position represented a graceful
withdrawal, but, in the already charged atmosphere of Numidianpapal
relations, Gregorys continued intervention in the affairs of the Numidian church seems to have rubbed a number of the local bishops the
wrong way.
It would, however, probably be misleading to speak of a generalized
African rejection of the Pope. Bishop Dominicus of Carthage seems
to have enjoyed cordial relations with Gregory, and from the time the
Pope heard of Dominicus election (apparently embarrassingly late) in
July 592 until the last years of Gregorys pontificate the two exchanged
a number of letters and gifts.86 In a couple of these Gregory sought
to influence Dominicus actions or policies: in 594, the Pope wrote
to Dominicus praising his zeal in repressing heretics but rejecting the
opinion of an African synod that those who were lax in their search for
the heterodox should be punished with the loss of wealth and dignity; in
597, Gregory asked that Dominicus punish some monks for abandoning
their monastery.87 But Gregory also sought to console Dominicus on
the eruption of plague in North Africa with the comforting thought
that the disease and other signs heralded the imminent end of the world,
and, a year later, the Pope rejoiced that Dominicus had himself recovered
82
84

85
86
87

83 Greg. Ep. 7.2, 1:444.

Greg. Ep. 4.7, 1:223.
Greg. Ep. 1.72, 1:81, esp. ll. 256: si per eam dispersarum ecclesiarum potuerit societas restaurari
(if through it the fellowship of the scattered churches will be able to be re-established), in
reference to allowing Numidian bishops to come to Rome.
Greg. Ep. 1.75, 1:834; Markus, Country Bishops, pp. 34.
Greg. Ep. 2.40, 1:1279 and, in addition to the letters cited below, ibid., 6.19, 1:389; see also
ibid., 10.17, 2:846.
Greg. Ep. 5.3, and 7.32, 1:2689 and 1:4956.

325

Staying Roman
from an illness.88 Gregory thanked Dominicus for his gifts, including
relics of the third-century Carthaginian martyr Agileus, expostulated to
the bishop of Carthage on love, and praised the bishops enthusiastic
temperance.89
Even in Numidia, however, opinion was not united in opposition to
papal interference. The Numidian bishop Columbus had made a special promise to St Peter that somehow bound his interests to those of
the Roman church, and, as we have seen, together with the Roman
chartularius Hilarus, Bishop Columbus acted throughout Gregorys pontificate as the Popes agent in Africa.90 A certain Bishop Victor also seems
to have been willing to work with the Pope. In 594, Gregory encouraged
Victor and Columbus to convene an anti-Donatist council, and four years
later they were to see to the restoration to their rightful bishop of some
disputed churches.91 In the problematic affair of the Numidian bishop
Paul (about which, more below), the Pope seems to have believed that
he could rely on the goodwill or at least the professionalism of Bishops
Adeodatus (probably the primate of Numidia) and Maurentius.92 The
same was true of the later primate of Numidia, Victor.93
Perhaps more telling, however, is the fact that even in Numidia both
the Pope and the emperor could serve as a court of appeals. When the
Carthaginian deacon Ferrandus came to write his Breviatio canonum, in
the first half of the sixth century, the African church recognized the special authority of the Roman church to adjudicate contested cases: An
adjudged bishop may appeal to the apostolic see, if he wishes.94 And,
indeed, Gregorys letters are filled with examples of African bishops,
priests, and deacons who had brought their cases before the papal court.
In February of 598, the Numidian bishop Cresconius personally secured
a ruling from the Pope ordering another bishop, Valentio, to restore the
churches that had been stolen from Cresconius predecessor fifteen years
88
89
90

91
92

93
94

Greg. Ep. 10.20 and 12.1, 2:8501 and 2:9678.

Greg. Ep. 6.63, 8.31 and 12.1, 1:4379, 2:5545, and 2:968.
Oath: Greg. Ep. 3.47, 1:192, where Columbus is also sent the keys of St Peter and a few links
of the chains in which the apostle was imprisoned. Columbus: Greg. Ep. 2.39, 3.47, 3.48, 4.7,
4.35, 6.36, 7.2, 8.1415, 12.3, 12.8, and 12.9, 1:1257, 1:1912, 1:194, 1:223, 1:2556, 1:41011,
1:4445, 2:5324, 2:971, 2:97980, and 2:981. Hilarus: Greg. Ep. 1.735, 1.82, 9.133, 10.16,
12.2, 12.8, 12.9, 1:823, 1:89, 2:683, 2:845, 2:970, 2:980, and 2:982.
Greg. Ep. 4.35 and 8.14, 1:2556 and 2:5323.
Greg. Ep. 8.13, 2:5312. The primate Adeodatus and Gregory seem to have had somewhat rocky
relations: Greg. Ep. 3.48, 1:1934; see also ibid., 3.47, 1:192. However, the name Adeodatus was
popular in late antique Africa, and the association of the bishop and the primate by name alone
is not certain.
Greg. Ep. 12.89, 2:97982.
Ferrandus, Breviatio canonum 59, p. 292: Vt adiudicatus episcopus, ad apostolicam sedem, si
uoluerit, appellet.

326

The dilemma of dissent

before.95 Several years earlier, the priest Adeodatus had travelled from
Byzacena to the papal court to accuse Bishop Quintianus of removing
him from his office because of illness.96 Later, a Numidian by the name
of Donatdeus journeyed to Rome to complain that he had been unjustly
deprived of a deaconate by a bishop named Victor.97 And, in 592, the deacons Constantius and Mustelus of the church of Pudentiana in Numidia
alleged at the papal court that their bishop, Maximian, had been corrupted by Donatist bribes: Behold, Gregory wrote to Columbus, the
wolf now tears the Lords flock to pieces not secretly in the night, but
in the open daylight.98 Two deacons of the Numidian church of Lamigensis brought a similar complaint to the Pope against their own bishop,
Argentius, probably (although not certainly) in person.99 The complaints
that Paulinus, bishop of Tigesis in Numidia (mod. An el-Bordj), beat
his clerics and engaged in simony somehow also reached the Pope.100
So too the protestations of the inhabitants of the diocese of Fausianensis
against the double taxes to which they were being subjected.101 In 598,
Crementius, primate of Byzacena, sent Martin scholasticus to Rome to
represent the primate in a case at the Papal court. Martin, however, seems
never to have travelled further than Sicily, where he tarried to discuss the
affair with Bishop John of Syracuse.102
On the other hand, the canons of the African church sought to prevent
appeal to the emperor on legal matters. Any bishop who sought a public
judgment from the emperor was to be deprived of his office, and, just in
case, no deposed bishop was allowed to appeal his case to the imperial
court.103 Even so, this was not always the way things worked in practice.
In the 540s, the communications between Justinian and Datianus, then
primate of Byzacena, were clearly the result of the bishops initiative.
Indeed, it seems likely that one of the privileges that Datianus and the
Byzacenan episcopate sought to preserve through these exchanges was
the very right to appeal cases to the emperor; at least in the 560s that
specific concession was confirmed by Justin II for the primate and bishops
of Byzacena, even for cases that had already been judged in Africa by
the prefect.104 Nor was this a hollow privilege. In 598, Crementius, then
95
98
99
102
103

104

96 Greg. Ep. 4.13, 1:231.

97 Greg. Ep. 12.3, 2:971.
Greg. Ep. 8.14, 2:5323.
Greg. Ep. 2.39, 1:126: Ecce lupus Dominicum gregem non iam in nocte latenter sed in aperta
luce dilaniat.
100 Greg. Ep. 12.89, 2:97982.
101 Greg. Ep. 11.7, 2:869.
Greg. Ep. 1.82, 1:89.
Greg. Ep. 9.24 and 9.27, 2:584 and 2:5889.
Ferrandus, Breviatio canonum 43, p. 290: Vt quicumque episcopus ab imperatore publicum
iudicium postulauerit, honore proprio priuetur and ibid., 66, p. 293: Vt depositus episcopus
ad imperatorem causam suam non deferat.
D. Feissel, Un Acte de la prefecture dAfrique sur leglise de Byzac`ene au debut du r`egne de
Justin II, Antiquite tardive 11 (2003), pp. 97112.

327

Staying Roman
primate of Byzacena, was accused by his fellow bishops of some crime.
The nature of the crime and indeed almost all of the details surrounding
the affair are unclear to us, but the case was first brought before the
emperor Maurice, who in turn entrusted it to the Pope. However, one
of the parties involved and, again, it is not entirely clear which one
sought to block the transfer of the case to the papal court by bribing the
magister militum Theodore with ten pounds of gold. Though the attempt
did not succeed, four years later, in 602, Gregory remanded the case to
the council of bishops in Byzacena.105
However, the best documented case of an appeal reaching the emperor
was first channelled through the papal court. In August of 591, Gregory
had written to Gennadius, the exarch of Africa, requesting that anyone
who desired to come to Rome from Numidia be allowed to do so
without hindrance. The admonition was phrased in general terms, but
three years later the obstacles Gregory envisioned were to take a very
concrete form. In July of 594, Gregory wrote two letters, one to the
Praetorian Prefect of Africa, Pantaleon, the other to the bishops Victor
and Columbus, urging them to send to Rome a Numidian bishop named
Paul who had been having some trouble with the local Donatists.106 We
hear nothing of Pauls case for another two years, but in August of 596,
the Pope again wrote to the exarch of Africa, reproaching him not only
for not helping Paul but for placing impediments in the bishops way to
prevent his coming to Rome.107 When Paul arrived at the papal court
he had complained that imperial laws against the Donatists were being
neglected in Africa.108 Gennadius had, however, also written to Gregory
about Pauls case, and the Pope had had the letter read out in Pauls
presence. At this point Paul began to equivocate: he responded that he
does not incur the hatred of certain people because he represses Donatists,
but rather endures the ingratitude of many on account of his defence of
the Catholic faith.109 Gennadius had further informed Gregory that Paul
had been excommunicated. Why, the Pope wondered, had he learned this
not from the primate of Pauls province but from the exarch?110 Gregory
seems to have asked the Numidian bishop Columbus for a report on the
whole affair. The bishops letter arrived in October, as did a cancellarius
from Gennadius bureau and three of the (putatively) many accusers from
105
106
107
109

110

Greg. Ep. 9.24, 9.27, and 12.12, 2:584, 2:5889, and 2:9856.
Greg. Ep. 4.32 and 4.35, 1:2512 and 1:2556.
108 Greg. Ep. 6.64, 1:43940.
Greg. Ep. 6.62, 1:4367.
Greg. Ep. 6.62, 1:436: respondit non se quorundam odio, quia Donatistas cohibeat, laborare,
sed magis pro defensione catholicae fidei multorum perhibet ingratitudinem sustinere. See also
Greg. Ep. 6.64, 1:440.
Greg. Ep. 6.62, 1:437.

328

The dilemma of dissent

Pauls church who could speak against the Numidian bishop. But by this
point it was too late. Deciding that a case involving the secular judges was
best decided by the secular authorities, Gregory had already sent Paul on
to the imperial court at Constantinople.111 Within eighteen months, by
February of 598, the case had been remanded to Gregory, who, in turn,
sent Paul back to Numidia, entrusting the bishop to the mercies of the
council that had excommunicated him in the first place.112
In case after case, Gregory sent complaints brought by African clerics
back to the ecclesiastical councils of the provinces in which they originated. This surely had to do not only with Gregorys personal style of
pastoral administration but also with the nature of North African Christianity in the later sixth and early seventh centuries. Markus is certainly
right, at least in this regard: the African church was an essentially conservative institution that instinctively rejected what it saw as encroachments
on its prerogatives from external authorities such as the emperor in Constantinople and the Pope in Rome. It was not the place of the emperor
to define orthodoxy ex cathedra. Nor yet was it the place of the Roman
pontiff to dictate procedural norms to a church whose traditions were
nearly as ancient as those of Rome itself. But this did not equate to a
rejection of the empire nor indeed of the papacy as such.
As we have seen, the first decades of the Byzantine occupation saw a
series of rebellions that threatened imperial control of the newly conquered western province: the revolt of Stotzas in 5368, the rebellion of
the Moors under Antalas in 544, the revolt of Guntharis in 545/6, and the
ongoing Moorish wars of the late 540s. In none of these cases is there even
the faintest hint of an implication that the African church supported the
political objectives of the rebels, even when those objectives involved
as with the revolt of Guntharis the establishment of an independent
kingdom of Africa. Although he went into hiding, Facundus of Hermiane went to tremendous lengths to convince Justinian of the untenable
nature of the imperial position with regards to the condemnation of the
Three Chapters. And, indeed, it bears repeating that Facundus bitter disillusionment with the emperor only emerged after the bishop had gone
into hiding, after the imperial system had so dramatically failed him. The
deacon Ferrandus, on the other hand, never seems to have become disillusioned with the Roman state. Love the empire (respublica) as yourself,
the Carthaginian cleric suggested to a certain Reginus comes in a letter
111
112

Greg. Ep. 6.64 and 7.2, 1:43940 and 1:4445.

Greg. Ep. 8.13 and 8.15, 2:5312 and 2:5334; E. Caspar, Geschichte des Papsttums von den
Anfangen bis zur Hohe der Weltherrschaft, 2 vols. (Tubingen, 19303), 2:445; Markus, Country
Bishops, p. 14.

329

Staying Roman
explaining how both to serve as a general and lead a religious life.113 The
same letters of Gregory the Great that reveal the prickly response of the
Numidian church to the Popes intervention in the provinces internal
affairs also show that both the Roman church and the emperor could
serve as a final court of appeal in the settlement of ecclesiastical disputes,
even in Numidia.
These are not the actions of a church whose instinctive posture with
respect to the empire and the Roman church was dissent, nor are they
indicative of a church acting as a nationalistic organ uncertain as to the
benefits of Byzantine rule. Rather, I would argue, the lengths to which
African theologians went to convince their opponents of the Chalcedonian position, the speed with which the African episcopate came to terms
with the realities of imperial policy, and the frequent appeals of African
churchmen to the emperor and the Pope all betray a church that on a
deep and meaningful level saw itself as fully integrated into the imperial
and ecclesiastical structures of the sixth-century Byzantine empire. To be
sure, the African vision of that empire differed from that of the imperial court and the Roman church in this important regard: that it was
more decentralized and (though I use the term with some apprehension)
federalist in its outlook than either of these two most centralizing of
institutions. Put another way, the African church re-entered the empire
with what we might characterize as a fourth-century attitude towards
the enforcement of church discipline and the proper relations between
secular and ecclesiastical power. But the sixth-century African church
was nevertheless steadfast in what Peter Brown has called, in a different context, a doctrinaire and cocksure belief that what was good for
Numidia and, we may venture, for Africa generally was good for
the Roman Empire.114
3. africa and the mediterranean
It was not just the church that was reintegrated into Byzantine society.
The African elite as a whole remained connected to the larger Mediterranean world throughout the sixth and seventh centuries. The circulation
of books, letters, and information; the maintenance of economic ties; the
movement of individuals; and the diffusion of saints cults all testify to
the endurance of these connections. Africas links abroad were multivalent, but the age-old bond with Rome remained strong socially, and the
113
114

Ferrandus, Ep. 7.910, PL 67:9369 written after the death of Fulgentius of Ruspe, itself around
the time of the Byzantine reconquest.
Brown, Christianity and Local Culture, p. 93.

330

The dilemma of dissent

reconfiguration of power in the sixth-century Mediterranean ensured
that Constantinople similarly formalized its status as the new focus of
secular and ecclesiastical ambition. A handful of Africans sought preferment at the imperial capital, but, insofar as we can tell from the surviving
sources, even within Africa itself the local elite seems to have accepted
and identified with the new regime.
That local elite seems to have remained urban in character throughout the Byzantine and into the early Islamic periods. The geographer
George of Cyprus, writing in the first years of the seventh century,
certainly still conceived of Africa as a network of cities.115 To be sure,
the current state of research strongly suggests that the widespread urban
affluence so visible in Africa in the fourth century was not sustained
into the Byzantine period; but continuity of occupation into the Islamic
Middle Ages is nonetheless archaeologically visible at sites like Carthage,
Sbetla, Bulla Regia, Rougga (class. Baraus), Belalis Maior, and Uchi
Maius.116 Critically, too, a number of major late antique cities including Hadrumetum (mod. Sousse), Vaga (mod. Beja), and Sicca (mod. El
Kef) have remained important into the modern period and thus are
not well understood archaeologically. Much the same is true of towns
like Iunca and Thelepte, which were important enough in c.600 to be
included on George of Cyprus list but which were eventually abandoned, and whose remains are still largely unexplored. The result is that
our image of urbanism in the Byzantine period tends to be dominated by
those cities that proved least capable of negotiating the transformations
of the sixth and seventh centuries.
By the time George was writing, moreover, the idea of the city had
changed radically from that of the classical past this throughout the
115
116

George of Cyprus, Descriptio orbis Romani 63884, ed. H. Gelzer (Leipzig, 1890), pp. 335.
Continuity of settlement into the Islamic period: A. Mahjoubi, Recherches dhistoire et darcheologie
a` Henchir el-Faouar (Tunisie): la cite des Belalitani Maiores, Publications de lUniversite de Tunis,
Faculte des lettres et sciences humaines, ser. 1, Archeologie histoire 12 (Tunis, 1978),
p. 451, with ibid., pp. 345, 37187, and 420; H. Broise and Y. Thebert, Recherches archeologiques

franco-tunisiennes a` Bulla Regia II/1: les thermes Memmiens, Collection de lEcole

francaise de
Rome 28/2 (Rome, 1993), pp. 38797; R. Guery, Survivance de la vie sedentaire pendant les
invasions arabes en Tunisie centrale: lexemple de Rougga, Bulletin archeologique du Comite des
Travaux archeologiques et historiques n.s. 19B (1983), pp. 399410; F. Bejaoui, Nouvelles donnees
archeologiques a` Sbetla, Africa 14 (1996), pp. 389 and 413; and B. Caron and C. Lavoie,
Les Recherches canadiennes dans le quartier de la Rotonde de lOdeon a` Carthage: un
ensemble paleochretien des IVe Ve si`ecles ou une phase doccupation et de construction du
VIIIe si`ecle?, Antiquite tardive 10 (2002), pp. 24961; but see also G. Vitelli, Islamic Carthage:
The Archaeological, Historical and Ceramic Evidence, Dossiers CEDAC 2 (Carthage, 1981). On
urbanism and the African elite in the Byzantine period in general, see C. Wickham, Framing
the Early Middle Ages: Europe and the Mediterranean, 400800 (Oxford, 2005), pp. 6389 and
6434.

331

Staying Roman
Mediterranean.117 In the Byzantine era, developments in Africa progressed naturally along lines that, as we have seen, were already visible
in the Vandal and even late Roman periods, and that have pithily been
characterized by Peter Pentz as the dissolution of the nucleated town
(see above, Chapter 3.1.1).118 Cities continued to be important administrative, religious, and economic centres, but now they came to be
characterized above all by their fortifications, churches, and economic
production.119 Indeed, it would seem that in Procopius mind it was
the wall that made the city.120 Most of the African towns mentioned by
George of Cyprus had some kind of fortification as well.121 Anna Leone
has recently suggested that towns buttressed with city walls retained a
more classical urban layout longer than those that received forts, where
settlement tended to re-nucleate around or near the fortified complex.122
In either case, though, if monumental military architecture in Byzantine
Africa sought to project a message, it was one of power and security
under God and the empire. This, at least, is the impression conveyed by
the dedicatory inscriptions on urban fortifications like those at Cululis
and Tebessa, which emphasize imperial strength against the Moors, or
those at Calama, which speak of conquest in the sign of the cross and
defence under Sts Clement and Vincent.123 Churches, too, remained an
important focus of settlement and urban activity in Byzantine Africa, and
as such probably communicated a similar point about faith and power.124
117

118
119

120
121
122
123

124

Wickham, Framing, pp. 591692; see also J. Haldon, The Idea of the Town in the Byzantine
Empire, in G. P. Brogiolo and B. Ward-Perkins (eds.), The Idea and Ideal of the Town between
Late Antiquity and the Early Middle Ages (Leiden, 1999), pp. 123.
P. Pentz, From Roman Proconsularis to Islamic Ifrqiyah, GOTARC Serie B. Gothenberg Archaeological Theses 22 (Copenhagen, 2002), p. 43, explored more fully, ibid., pp. 4451.
Proc. Aed. 6.3.96.7.16, 4:17685; J. Durliat, Les Dedicaces douvrages de defense dans lAfrique

byzantine, Collection de lEcole

francaise de Rome 49 (Rome, 1981), pp. 10812; D. Pringle,
The Defence of Byzantine Africa from Justinian to the Arab Conquest: An Account of the Military History
and Archaeology of the African Provinces in the Sixth and Seventh Centuries, BAR International Series
99 (Oxford, 2001), pp. 10920; Pentz, From Roman Proconsularis, pp. 2975; A. Leone, Changing
Townscapes in North Africa from Late Antiquity to the Arab Conquest, Studi storici sulla Tarda
Antichit`a 28 (Bari, 2007), pp. 167279.
Proc. Aed. 6.5.13 and 6.6.1316, 4:181 and 4:1823.
George of Cyprus, Descriptio 63884 and 7958, pp. 335 and 41 with Pringle, Defence, pp.
5357, map 6 and the gazetteer references cited there.
Leone, Changing Townscapes, p. 237; see also Wickham, Framing, p. 638 and (for the Islamic
period) Pentz, From Roman Proconsularis, pp. 6772.
Pringle, Defence, p. 319, no. 4 (Cululis); CIL 8.1863 + 16507 = ILCV 806 = Pringle, Defence,
p. 325, no. 23 (Tebessa); CIL 8.5352 = ILCV 791 = Pringle, Defence, p. 323, no. 17 (Calama);
CIL 8.5346 + 17579 and 8.5359 + 17529 = ILCV 1622ab = Pringle, Defence, pp. 3234, no.
18 (Calama). See also Pringle, Defence, p. 109; Pentz, From Roman Proconsularis, pp. 10613; and
Wickham, Framing, p. 638.
Leone, Changing Townscapes, p. 188; Pentz, From Roman Proconsularis, pp. 445; and Wickham,
Framing, pp. 6378.

332

The dilemma of dissent

In his Buildings, Procopius certainly indicates something of the significance of ecclesiastical architecture to Justinians ideology of empire in
the region.125 Moreover, when olive presses were erected in urban areas
in the Byzantine period they were repeatedly sited close to churches,
a fact that may indicate some kind of ecclesiastical supervision of the
production of olive oil, perhaps even in a public capacity.126
These transformations in the urban landscape accompanied a transformation in the character of the urban elite. By the Byzantine period, we
are no longer able to trace the old senatorial aristocracy in Africa.127 The
same is true of the curiae or town councils that had run the provinces
cities under the high empire, and which may already have ceased to
function at some point in the Vandal period.128 In their place, we find
prominence and status accorded to a more informal collection of urban
notables, consisting normally of members of the church hierarchy and
locally stationed military officers generals, tribunes, bishops, and priests,
among others as well as local citizens of wealth and influence.129
Wealth, of course, is a relative concept, and from a long-term perspective these local notables probably experienced diminishing levels of
prosperity as the sixth and especially the seventh centuries wore on. To
be sure, the same difficulties that we encountered in assessing the African
economy under the Vandal regime recur in the Byzantine period. Ceramics tableware, amphorae, lamps continue to provide us with archaeologically traceable data through which to assess some of Africas economic
connections, both external and internal, in the sixth and seventh centuries; but the archaeological evidence probably also continues to give
125
126
127
128
129

Proc. Aed. 6.4.4, 6.5.9, and 6.7.16, 4:177, 4:180, and 4:185.
Leone, Changing Townscapes, pp. 22737; see also ibid., pp. 2207.
M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Spatantike, Frankfurter althistorische Studien 7 (Kallmunz, 1973), pp. 838.
Wickham, Framing, p. 637.
See, e.g., magistri militum: Pringle, Defence, pp. 3336, nos. 468 (Rusguniae), 51 and 53 (Sbetla).
Tribunes: Pringle, Defence, pp. 319, 329, 333, and 3357, nos. 4 (An Djelloula), 33 (Khenchela),
45 (Rusguniae), 52 (Sbetla), and 54c (Thizica). Bishops: N. Duval, Recherches archeologiques a`

Hadra, 2 vols., Collection de lEcole

francaise de Rome 18/12 (Rome, 197581),1:202 and
257, nos. 1 and 3 (Hadra); Y. Duval, Loca sanctorum Africae: le culte des martyrs en Afrique du

IVe au VIIe si`ecle, 2 vols., Collection de lEcole

francaise de Rome (Paris, 1982), 1:448, no. 21
(Hr. Fellous); ibid., 1:13842, no. 64 (Rouis); ibid., 1:25964, no. 126 (Hr. Akrib); ibid., 1:277
81, no. 133 (An Guigba); AE (1967), 189, no. 562 (Bir Baktach); Mahjoubi, Recherches,
p. 420 = AE (1974), 196, no. 696 (Hr. el-Faouar); J. Terry, Christian Tomb Mosaics of
Late Roman, Vandalic and Byzantine Byzacena, Ph.D. diss., University of Missouri, Columbia
(1998), pp. 4445, no. 91 (Sbetla). Priests: AE (1967), 189, no. 562 (Bir Baktach); Duval, Hadra
1:2930 and 612, nos. 6 and 34; Duval, Loca sanctorum, 1:2679 and 277, nos. 128 and 132B
(Hr. Akrib); Terry, Tomb Mosaics, pp. 5225, nos. 1301 (Sidi Abich). Unspecified: Proc. BV
2.23.18, 2.23.23, 2.23.29, 1:5279 (Hadrumetum); ibid., 2.26.6, 1:536 (Carthage); and Coripp.
Ioh. praef. 1, p. 1 (Carthage). In general, see Wickham, Framing, pp. 596602 and see also Leone,
Changing Townscapes, p. 198.

333

Staying Roman
an incomplete picture of the nature and extent of exchange. Already in
the fourth century it would seem that olive oil produced in the African
hinterland could be carried to collection and trans-shipment centres in
animal skins. By the eleventh and twelfth centuries, skins, canvas, and
even wicker baskets were used in preference to earthenware containers in
the transportation of goods across the southern Mediterranean, and this
shift may conceivably already have begun in late antiquity.130 Moreover,
in the Islamic period, as in the late Roman era, without such durable
amphorae the African exports mentioned in the written record would
for the most part be difficult to trace archaeologically: olive oil and oil
products, leather, hides, wax, and honey, as well as coral, cloth, shoes,
and figs.131 In addition to suggesting long-term continuities in terms of
African arboriculture and stock-raising, the similarities between this list
and Africas major commodities in the third and fourth centuries serve
to remind us that throughout the pre-modern era much exchange will
have taken place that we simply cannot see in the material record.
What we can see is that the fine tableware and transport amphorae so
characteristic of the late Roman age reached the African countryside in
progressively diminishing quantities in the Byzantine period. This was
already the trend in the interior in the Vandal period (see above, Chapter
3.1.2); now it extended into the coastal zone as well: on the island of
Jerba, in the area around Segermes, and even in the immediate hinterland
of Carthage itself the number of rural sites where diagnostic pottery has
been found dwindles steadily in the later sixth and seventh centuries.132
130

131

132

In general: Pentz, From Roman Proconsularis, p. 79. Late Roman: J. T. Pena, The Mobilization
of State Olive Oil in Roman Africa: The Evidence of Late 4th c. Ostraca from Carthage, in
J. H. Humphrey (ed.), Carthage Papers: The Early Colonys Economy, Water Supply, a Public Bath,
and the Mobilization of State Olive Oil, Journal of Roman Archaeology Supplementary Series 28
(Portsmouth, RI, 1998), p. 186 and see also ibid., p. 168 and D. J. Mattingly, Oil for Export?
A Comparison of Libyan, Spanish and Tunisian Olive Oil Production in the Roman Empire,
Journal of Roman Archaeology 1 (1988), p. 43. Medieval: S. D. Goitein, A Mediterranean Society:
The Jewish Communities of the Arab World as Portrayed in the Documents of the Cairo Geniza, 6 vols.
(Berkeley, Calif., 196793), 1:3325. Wickham, Framing, p. 717 is sceptical.
Goitein, Mediterranean Society, 1:47 (coral), 1:60 (oil and oil products), 1:112 (leather, hides, gilded
shoes), 1:121 (figs), 1:125 (wax and honey), 1:1534 (coral, wax, felt, hides, leather, shoes), 1:224
(linen made from imported Egyptian flax), and 1:344 (olive oil and soap). Evidence for flax
cultivation in Africa itself in late antiquity is scant but significant: S. Dietz, A Summary of the
Field Project, in S. Dietz, L. Ladjimi Sebaj, H. Ben Hassen, P. rsted, and J. Carlsen (eds.),
Africa Proconsularis: Regional Studies of the Segermes Valley of Northern Tunisia, 3 vols. (Aarhus,
19952000), pp. 7967; van der Veen, Grant, and Barker, RomanoLibyan Agriculture,
p. 246.
J. A. Greene, Une Reconnaissance archeologique dans larri`ere-pays de la Carthage antique,
in A. Ennabli (ed.), Pour sauver Carthage: exploration et conservation de la cite punique, romaine et
byzantine (Tunis, 1992), pp. 1967; Dietz, Summary of the Field Project, p. 782 with figs. 5m
p, p. 785; E. Fentress, A. Drine, and R. Holod, An Island through Time: Jerba Studies, vol. 1, The
Punic and Roman Periods, Journal of Roman Archaeology Supplementary Series 71 (Portsmouth,

334

The dilemma of dissent

The African countryside, it would seem, was becoming less affluent.
Even so, it is worth remembering that in the historical memory of later
Arabic accounts Africa was still prosperous enough in the second half of
the seventh century to make the conquest of the region worthwhile. In
part this had to do with the regions potential as a source of slaves, but
that was not the full story.133 In a much-cited anecdote recounted by
the ninth-century historian Ibn c Abd al-Hakam, in the wake of the first
major Arab incursion into Africa, in 647, local Romano-African notables
paid the leader of the expedition, c Abd Allah ibn Sac ad, a substantial
tribute in coin to leave their province. Astounded at their wealth, c Abd
Allah asked the notables where they got their money, whereupon one
of them dug up an olive pit and showed it to the warrior, telling him
that East Romans had no olives of their own, and so Africans sold them
olive oil.134 We might dismiss the story as nothing but a charming fantasy
except for the fact that over a dozen hoards of seventh-century Byzantine
gold coins have been found in the territory of modern Tunisia, the best
documented of which is a collection of 268 solidi from Rougga that was
almost certainly hidden at the time of the Arab raid of 647.135 Three or so
generations later, an even larger hoard was assembled; unfortunately, in
the twentieth century, it was dispersed on the market, but the collection
apparently consisted of about 700 coins, mainly solidi of Constantine IV
(ad 66885), though also including a coin of Justinian II (ad 68595
and 70511) and two early eighth-century Arab-Byzantine transitional
issues.136 Africas social landscape would thus seem to have contained
its wealthy elements all the way into the early Islamic period. Similarly,

133

134

135

136

RI, 2009), pp. 198200. These trends were not reversed in the Cillium-Thelepte region:
R. B. Hitchner, The Kasserine Archaeological Survey 1987, Antiquites africaines 26 (1990),
pp. 23159, esp. 247. Dougga may have experienced a slight revival in the later sixth century,
but the general trend is towards fewer diagnostic sites: M. de Vos, Rus Africum: Terra, acqua, olio
nellAfrica settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000),
pp. 725.
See, e.g., Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, ed. and trans. A. Gateau, in Conquete
de lAfrique du nord et de lEspagne (2nd edn; Algiers, 1947), pp. 46, 604, 80, and 88; Ahmad ibn
Yahya al-Baladhur, Kitab futuh al-buldan, ed. M. J. de Goeje, in Liber expugnationis regionum (2nd
edn; Leiden, 1968), pp. 22930; E. Savage, A Gateway to Hell, a Gateway to Paradise: The North
African Response to the Arab Conquest, Studies in Late Antiquity and Early Islam 7 (Princeton,
NJ, 1997), pp. 6778.
Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, pp. 468. The gold coins collected in this
expedition were remembered as being worth one and one-quarter a normal dinar: ibid., p. 44.
See also al-Baladhur, Kitab futuh al-buldan, p. 227.
R. Guery, C. Morrisson, and H. Slim, Recherches archeologiques franco-tunisiennes a` Rougga III: le

tresor de monnaies dor byzantine, Collection de lEcole

francaise de Rome 60 (Rome, 1982). Lists
of Byzantine-era coin hoards from North Africa are provided by ibid., interleaf pp. 789 with
fig. 7, p. 79 (map) and Pringle, Defence, pp. 12830.
C. Morrisson, Un Tresor de solidi de Constantin IV de Carthage, Revue numismatique, 6th ser.
22 (1980), pp. 15560.

335

Staying Roman
though the last known monumental building project in Carthage was a
mid seventh-century reworking of the basilica complex at Bir el Knissia,
at this point the church which appears to have been a pilgrimage
centre seems to have been at the peak of its prosperity. The basilica
apparently continued in use until its destruction, perhaps about the time
of the fall of Carthage to the Arabs.137 Down to the Islamic conquest and
beyond, then, it is still possible to see signs of economic success among
the generally downward trends.
The same is true of Africas export economy. The reimposition of the
Roman tax system in Africa meant a revival of annual grain shipments
to Constantinople. With them came a reinvigoration of shipping links
between the Queen of Cities and one of her farthest-flung provinces.138
Fine tableware remained in production at a number of ARS ateliers in
eastern Africa Proconsularis down to at least the early and mid seventh
century, and some of the latest bowls and dishes may even have been manufactured into the eighth century.139 In central Tunisia, Michel Bonifay
has recently characterized the period spanning the end of the sixth and the
first half of the seventh centuries as one of relative economic dynamism
on the basis of ceramics production.140 Moreover, as our understanding of
the chronology and typology of the latest of late ancient ceramics develops, so too does our appreciation for the persistence of Africas economic
connections even if on a reduced level to the rest of the Mediterranean world. Throughout the Byzantine period, and in some places
into the early years of the Islamic era, African fine wares and amphorae
continued to reach major urban centres and imperial military installations
in the West, including Cartagena, Tarragona, Marseilles, San Antonino
di Perti, Rome, and Naples.141 Exports to Catalunya, Provence, and in
137
138
139

140
141

Stevens, Bir el Knissia, pp. 3068.

M. McCormick, Origins of the European Economy: Communications and Commerce, ad 300900
(Cambridge, 2001), pp. 10211.
M. Mackensen and G. Schneider, Production Centres of African Red Slip Ware (3rd7th c.)
in Northern and Central Tunisia: Archaeological Provenance and Reference Groups based on

Chemical Analysis, Journal of Roman Archaeology 15 (2002), pp. 12530 and M. Bonifay, Etudes
sur la ceramique romaine tardive dAfrique, BAR International Series 1301 (Oxford, 2004), pp. 210
and 482; see also ibid., pp. 41017.

Bonifay, Etudes
sur la ceramique, p. 482; see also Reynolds, Trade, p. 119.
P. Arthur, Naples: From Roman Town to City-State, Archaeological Monographs of the British
School at Rome 12 (London, 2002), pp. 110, 113, 1302, and 141; G. Murialdo, Le Anfore
da trasporto, in T. Mannoni and G. Murialdo (eds.), S. Antonino: un insediamento fortificato
nella Liguria bizantina, Collezione di monografie preistoriche ed archeologiche 12, 2 vols. (Bordighera, 2001), 1:2679 and 28995; J. M. Macias i Sole and J. A. Remol`a Vallverdu, Tarraco
visigoda: caracterizacion del material ceramico del siglo VII d. C, in J. M. Gurt and N.
Tena (eds.), V Reunio dArqueologia Cristiana Hisp`anica [= V Reunion de Arqueologa Cristiana
Hispanica], Monografies de la Seccio Hist`orico-Arqueol`ogica 7 (Barcelona, 2000), pp. 4867;
C. Panella and L. Sagu`, Consumo e produzione a Roma tra tardoantico e altomedioevo:

336

The dilemma of dissent

the second half of the seventh century Liguria are particularly significant, for they demonstrate that overseas commerce could be profitable
even outside the redistributive fiscal structures that bound together the
territory of the empire. In the East, African amphorae are rarer but ARS
imports to Constantinople remained strong from the mid sixth to the
mid seventh centuries and even surged between c.650 and c.675, perhaps
mirroring Constans IIs increased financial extractions from Africa.142
Imports of African fine wares into the Aegean basin similarly revived
in the sixth and seventh centuries, and ARS also continued to reach
Levantine cities like Antioch, Beirut, Caesarea, and even Hama in inland
Syria.143 Alexandria too imported fine African tableware in significant
quantities into the seventh century.144 In exchange, eastern amphorae
were sent west to Carthage, though in general these seem to contract
over the seventh century.145 However, the presence of a few fragments
of glazed tableware seemingly of Constantinopolitan manufacture in the
latest levels from the African metropolis would seem to indicate that
connections between Carthage and the imperial capital were maintained
down to the Arab conquest of 698.146

142

143
144
145
146

le merci, i contesti, in Roma nellalto medioevo, 2 vols., Settimane di studio del Centro italiano di studi sullalto medioevo 48 (Spoleto, 2001), 2:791815; D. Whitehouse, et al., The
Schola Praeconum II, Papers of the British School at Rome 53 (1985), pp. 1859; and in general

Wickham, Framing, pp. 7248; Bonifay, Etudes

sur la ceramique, pp. 153, 4578, 482, and 485;
McCormick, Origins, pp. 51112; Reynolds, Trade, pp. 314, 589, and 11920; and C. Panella,
Merci e scambi nel Mediterraneo tardoantico, in A. Momigliano and A. Schiavone (eds.),

Storia di Roma, 4 vols. (Turin, 1988), 3/2:6767. Cf. lamps: Bonifay, Etudes
sur la ceramique,
p. 457 and Panella, Merci e scambi, p. 674. African coinage of the Byzantine period also
enjoyed a widespread diffusion: C. Morrisson, LAtelier de Carthage et la diffusion de la
monnaie frappee dans lAfrique vandale et byzantine (439695), Antiquite tardive 11 (2003),
pp. 7484.
McCormick, Origins, p. 102; J.-P. Sodini, Productions et e changes dans le monde protobyzantin
(IVe VIIe s.): le cas de la ceramique, in K. Belke, F. Hild, J. Koder, and P. Soustal (eds.), Byzanz

als Raum: Zu Methoden und Inhalten der historischen Geographie des Ostlichen
Mittelmeerraumes,

Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften
283/Tabula Imperii Byzantini 7 (Vienna, 2000), pp. 188 and 1913; Reynolds, Trade, pp. 34,
119, and 121; Panella, Merci e scambi, pp. 659 and 675; J. W. Hayes, Excavations at Sarachane in
Istanbul, vol. 2 (Princeton, NJ, 1992), p. 7; and Hayes, Late Roman Pottery, pp. 418 and 4234.

See also Bonifay, Etudes

sur la ceramique, pp. 482 and 485.
Sodini, Productions et e changes, pp. 1889; Pentz, From Roman Proconsularis, p. 156; and
Panella, Merci e scambi, p. 659.
Sodini, Productions et e changes, p. 190.
Pentz, From Roman Proconsularis, pp. 36 and 157; Reynolds, Trade, pp. 75 and 789; Panella,
Merci e scambi, pp. 668 and 6756 and Panella, Le anfore di Cartagine, pp. 589.
J. W. Hayes, Pottery Report 1976, in J. H. Humphrey (ed.), Excavations at Carthage 1976,
Conducted by the University of Michigan, vol. 4 (Ann Arbor, Mich., 1978), pp. 935; J. W. Hayes,
Probl`emes de la ceramique des VIIe` me IXe` me si`ecles a Salamine et a Chypre, in Salamine de
Chypre: histoire et archeologie: etat des recherches: Lyon, 1317 mars 1978 (Paris, 1980), p. 379; Hayes,
Sarachane 2:1218; and Panella, Merci e scambi, pp. 65960.

337

Staying Roman
The African urban elite whose lifestyle such economic exchange supported have not left us many sources through which to understand how
they construed their identities. The evidence of their epitaphs, though,
suggests that into the seventh century they continued to participate in
a commemorative culture that was predominantly Christian and Latin.
Moreover, to judge by their use of indictional dates, their experience of
time was on a meaningful level structured by the rhythms of imperial
fiscal exactions. One inscription from Belalis Maior even dates the burial
of a four-year-old boy named Spesindeum not just to a second indiction
but specifically to the reign of Heraclius, the unusual naming of the
emperor in this context a particularly public statement of support.147
Indeed, insofar as we can tell, Byzantine-era Romano-African urban
notables seem quickly to have come to identify with the empire. Leaving
aside the likelihood that many of the soldiers and officers who served in
the Byzantine army were recruited locally, this was still true of the three
laymen whose writings survive to us from this period. Most dramatically,
the poet Corippus made the transition from provincial grammaticus to
imperial insider with considerable facility, seeking out and apparently
winning the patronage first of the general John Troglita in Carthage
then of a whole array of highly placed court officials in Constantinople, and eventually even delivering a panegyric to Justin II himself at
his imperial coronation.148 No less impressive is the career of Junillus
Africanus. Derided by Procopius as an uneducated buffoon, Junillus was
a lawyer and Christian exegete of talent and ambition whose social circle
included Cassiodorus, Bishop Primasius of Hadrumetum, and Fulgentius of Ruspe. Under Justinian, Junillus rose to the highest-ranking legal
office in the Roman state and became a spokesman for the emperors
vision of imperial theocracy in a way that even Corippus never did.149
147

148
149

Mahjoubi, Recherches, p. 345. For Byzantine-era inscriptions in general, see most conveniently
J. Terry, Christian Tomb Mosaics of Late Roman, Vandalic and Byzantine Byzacena, Ph.D.
diss., University of Missouri, Columbia (1998); L. Ennabli, Les Inscriptions funeraires chretiennes

de la basilique dite de Sainte-Monique a` Carthage, Collection de lEcole

francaise de Rome 25
(Rome, 1975); L. Ennabli, Les Inscriptions funeraires chretiennes de Carthage II: la basilique de Mcidfa,

Collection de lEcole
francaise de Rome 62 (Rome, 1982); L. Ennabli, Les Inscriptions funeraires

chretiennes de Carthage III: Carthage intra et extra muros, Collection de lEcole

francaise de Rome
151 (Rome, 1991); N. Duval, Les Byzantines a` Rusguniae, Bulletin archeologique du Comite
des Travaux archeologiques et historiques n.s. 19B (1983), 34160; Duval, Hadra 1; and Mahjoubi,
Recherches, pp. 344, 3501, 416, and 420.
Corippus, In laudem Iustini Augusti minoris 1.1527, pp. 1617, with notes, pp. 923; Cameron,
Corippus Iohannis, p. 169.
Junillus Africanus, Instituta regularia diuinae legis, ed. H. Kihn, in Theodor von Mopsuestia und
Junilius Africanus als Exegeten: Nebst einer kritischen Textausgabe von des letzteren Instituta regularia
divinae legis (Freiburg im Breisgau, 1880), pp. 465528 = M. Maas, Exegesis and Empire in the
Early Byzantine Mediterranean, Studien und Texte zu Antike und Christentum 17 (Tubingen,

338

The dilemma of dissent

Writing perhaps not long after 550, the mythographer Fabius Planciades Fulgentius also publically welcomed the return of the empire, and
may similarly have hoped for preferment of the sort that his compatriots
received.150
No less than Ferrandus admonition to Reginus comes to devote himself
to the state, the writings of these men reveal a deep level of comfort
with the empire and the recognition that its ambitions and their own
could be furthered simultaneously. Much the same could be said of men
like the African-born Praetorian Prefect Innocent (see above, Chapter
4.2.1) or an African of humble origins named John who, in the late
550s, acted as the provisioner for the Byzantine forces in Armenia.151
Of course, in most of these cases pursuit of wealth and influence also
ultimately meant leaving Africa. Even Innocent may have spent some
time at the imperial court, given the tendency of sixth- and seventhcentury emperors to appoint their intimates to positions of high civil and
military command in the West. But we also find individuals allying their
interests with those of the empire in Africa itself. At least, in 544, the
notables of Hadrumetum were said to have preferred imperial rule to
that of the rebel soldier Stotzas and the Moorish leader Antalas.152 Then,
too, as we have seen, in the latter third of the sixth century, African
notables appear to have taken an increased responsibility for the erection
of local fortifications, acting in cooperation with the imperial authorities
(see above, Chapter 4.4.4). The evidence is admittedly thin, but, as far as
it goes, it seems to speak of the African secular elite accepting and even
embracing the empire.
Despite occasional theological friction, much the same seems to have
been true of the provinces bishops, at least in their capacity as local
notables. We have already seen that in the late sixth and seventh centuries
churches were apparently involved in the production and perhaps even
the fiscal collection of olive oil in urban centres. In the sixth century, a
bishop of Tebessa named Faustinus also seems to have been one of the
prominent local individuals responsible for the management of public
works in his city and its territorium. At least, a certain Masticana erected
a small fort outside the city on the road to Gafsa and dated it to the
time of the vir beatissimus Bishop Faustinus.153 At much the same time,

150
151
152
153

2003), pp. 118234; Proc. Anecd. 20.1720, 3:1278; Fulg. Ep. 7.2, 1:245; Maas, Exegesis and
Empire, pp. 132 and 65115.
Fulgentius Mythographer, Mitologiae 1.praef., p. 5; Hays, Date and Identity, pp. 2434.
Agathias, Historiae 4.21.54.22.6, pp. 14951.
Proc. BV 2.23.1126, 1:5269. On the citys bishop, Primasius, see above, section 2.
Durliat, Dedicaces, no. 32, pp. 803 = Pringle, Defence, p. 331, no. 39. See also CJ 1.4.26
(ad 530), pp. 424.

339

Staying Roman
this same Faustinus and his fellow bishops seem to have been expanding
their spiritual authority a process to which close relations with the
imperial authorities probably also proved useful. Another inscription,
this one from Mechtat er-Rouis (about twenty-eight kilometres southwest of Tebessa), indicates that Faustinus erected a memoria to the African
martyrs Maxima, Donatilla, and Secunda, perhaps in 550. Relics of the
saints were themselves probably not physically located at this site, and the
African episcopate had a long-standing suspicion of memoriae erected to
martyrs in the absence of some material remembrance; but in this case
Bishop Faustinus participation in the dedication probably legitimated
the memorial.154 Even when a material remembrance of the saint was
present, over the sixth century the African episcopate seems to have
imposed a rule on local priests requiring them to solicit the presence of a
bishop when consecrating an altar, and from c.580 onwards the thinning
epigraphic evidence suggests that bishops themselves normally performed
the deposition of relics.155 From Justinian to Heraclius, at least a handful
of the inscriptions recording these depositions were dated according
to the reign of the emperor.156 As they asserted and extended their
combined religious and political role in their own cities, the ability of
some bishops to pursue disputes or secure confirmation of their privileges
at the papal curia and imperial court probably bolstered their standing
locally (see above, section 2). It is thus perhaps understandable why, with
only a few notable exceptions, the African episcopate was generally not
prepared strenuously to oppose the will of the emperor in the mid sixth
century.
The traditional fundamentals of a late Roman education probably also
continued to be available to both the ecclesiastical and the secular elites
in Africa throughout the Byzantine period.157 To be sure, authors like
Corippus, Junillus, and probably Fulgentius the Mythographer, who lived
and wrote in the first decades of the restored imperial regime, will have
been educated under the Vandals. As we have seen, though, Justinian
established two Carthaginian grammarians on the public payroll in 534
(see above, Chapter 4.1). Corippus himself was a provincial schoolteacher,
and, according to Junillus, it was normal among the African educated
154

155
156
157

Duval, Loca sanctorum, 1:13842, no. 64 (Rouis) and Registri ecclesiae Carthaginensis excerpta 83,
pp. 2045; Concilium Carthaginense a. 525, p. 266, ll. 4568; Ferrandus, Breviatio canonum 171,
p. 301.
Duval, Loca sanctorum, 2:5724.
Ibid., 1:11719, no. 54 (Hadra); ibid., 1:21520, no. 106 (Sila); ibid., 1:2319, no. 112 (Telergma);
and ibid., 1:25964 and 2757, nos. 126 and 132 (Hr. Akrib).
P. Riche, Education and Culture in the Barbarian West, from the Sixth through the Eighth Century,
trans. J. J. Contreni (Columbia, SC, 1978), p. 39.

340

The dilemma of dissent

classes to study grammar and rhetoric in succession.158 This system may
still have been in place during the youth of Abbot Hadrian, an African
said to have been expert in both Greek and Latin who settled abroad
in a monastery in Campania and then travelled to England in the late
660s, where together with Theodore of Tarsus he founded the school
of Canterbury.159 Epigraphic evidence suggests that Virgil remained a
staple of Africas Latin curriculum. A Byzantine-era epitaph from Belalis
Maior adapts a line from the Aeneid to a new, commemorative context,
informing the reader, Here lies unworthy Florentius, in a tomb; he
was buried with uncut beard and bitter funeral.160 More elaborate verse
epitaphs also survive from Mactar and Sbetla.161 As in the late Roman
and Vandal periods, the written word was also put to pragmatic uses in
Byzantine Africa, as an ostracon from 5423 recording the assessment
of olive oil on an estate south of the Aur`es testifies.162 In the 660s, we
similarly hear of an African father who travelled to Constantinople to seek
a cure from St Artemios for his sons testicular ailment and who wrote
down on a leaf of parchment (>  1) the directions that he
received to the church where the saints relics worked their miracles.163
This reference to an African among Artemios devotees is indicative
of ties beyond the economy religious, intellectual, administrative, and
social that continued to bind Africa to the rest of the Mediterranean
158

159

160

161

162
163

R. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity, Transformation
of the Classical Heritage 11 (Berkeley, Calif., 1988), pp. 2613, no. 36 (Corippus); Junillus,
Instituta, praef., p. 120: sicut apud nos in mundanis studiis grammatica et rhetorica, ordine
et regulariter traditur. Now see also Fulgentius Mythographer, Expositio sermonum antiquorum,
p. 110 and Kaster, Guardians of Language, p. 250, no. 24 (Calcidius), in the light of Hays, Date
and Identity, pp. 2414. See also Vossing, Schule und Bildung, pp. 6313.
Bede, Historia ecclesiastica gentis Anglorum 4.1, ed. and trans. B. Colgrave and R. A. B. Mynors, in
Bedes Ecclesiastical History of the English People (Oxford, 1969), pp. 32830 implies that unlike
Theodore Hadrian was not a Greek (ne quid ille contrarium ueritati fidei Graecorum more
in ecclesiam cui praeesset introduceret, ibid., p. 330); thus the interpretation of M. Handley,
Disputing the End of African Christianity, in Merrills, Vandals, Roman, and Berbers, pp. 2967
on the question of Hadrians origins is to be preferred to that of B. Bischoff and M. Lapidge
(eds.), Biblical Commentaries from the Canterbury School of Theodore and Hadrian, Cambridge Studies
in Anglo-Saxon England 10 (Cambridge, 1994), pp. 82132, esp. 8492.
Mahjoubi, Recherches, p. 416: +++++/ hic indig/nus Florenti/us iacet in an/tro intonsa /
barba fune/re est mer/sus acerbo / vixit in pace / annis XX II men/ses VIII dies XXV /
d(e)p(o)s(itus) die IV idus / f(e)br(uarias) indict(ione) / II. Cf. Virgil, Aeneidos 6.429, p. 240:
abstulit atra dies et funere mersit acerbo.
Terry, Tomb Mosaics, pp. 4403, nos. 8990 with discussion ibid., pp. 1612; Duval, Inscriptions byzantines de Sbetla (Tunisie) III, pp. 4258. See also perhaps Terry, Tomb Mosaics,
pp. 4589 and 4701, nos. 98 and 104.
E. Albertini, Ostrakon byzantin de Negrine (Numidie), Cinquantenaire de la Faculte des Lettres
dAlger (18811931) (Algiers, 1932), 54.
The Miracles of St Artemios: A Collection of Miracle Stories by an Anonymous Author of Seventh-Century
Byzantium c. 4, ed. and trans. V. S. Crisafulli and J. W. Nesbitt, The Medieval Mediterranean
13 (Leiden, 1997), pp. 824.

341

Staying Roman
world in the Byzantine period. Indeed, though less dramatic than the
tensions between the provincial councils and the Pope or emperor, the
changing complexion of the African cult of saints is equally revealing of
the regions religious and cultural orientation in late antiquity. In part this
had to do with imperial patronage of existing cults in the post-reconquest
period. Thus, for example, Justinian sought to strengthen devotion to
the Virgin, who had also served as a patron to the Vandal kings and to
whom the emperor was said to have dedicated churches at Carthage,
Leptis Magna, and Septem.164 Procopius tells us that the cult of a local
martyr named Prima also received imperial sponsorship.165 The same
seems to have been true of the cult of St Cyprian. At least, as we have
seen, the martyr-bishop played an important role in the legitimation of
the reconquest (see above, section 1), and it is probably also significant
that of the four surviving references to Cyprian in the African epigraphic
record, three date to the second half of the sixth century.166
Perhaps more significantly, depositions of foreign saints relics already
a feature of African cult from the fourth or fifth centuries also became
increasingly common in the Byzantine period. Throughout the fifth
century, Africas strongest and most important devotional ties seem to
have been with Rome, the rest of Italy, and to a lesser extent Spain. The
cults of the apostles Peter and Paul and of the Roman martyrs Laurence
and Hippolytus appear to have been particularly popular.167 These saints
remained the objects of devotion after the Byzantine reconquest, and,
indeed, in September of 601, Gregory the Great even sent keys to the
body of St Peter and links of the apostles chains to the aristocratic African
women Savinella, Columba, and Galla, praising their care for the poor
and claiming that the links from Peters chains would secure the saints
intercession if worn around the neck.168 The martyred Spanish deacon
Vincent of Saragossa also seems to have become one of the most popular

164

165
166

167

168

Proc. Aed. 6.4.4, 4:177 (Leptis Magna); ibid., 6.5.9, 4:180 (Carthage); ibid., 6.7.16, 4:185
(Septem); see also ibid., 6.2.20, 4:175 (Boreium in Cyrenaica). On her cult in late antique
Africa, see A. Delattre, Le Culte de la Sainte Vierge en Afrique dapr`es les monuments archeologiques
(Paris, 1907), pp. 1136; H. Barre, Le Culte marial en Afrique apr`es saint Augustin, Revue des
etudes augustiniennes 13 (1967), pp. 285317; and Duval, Loca sanctorum, 2:61617.
Proc. Aed. 6.5.9, 4:180; on her cult, see also perhaps Duval, Loca sanctorum, 1:1089, no. 51B
(Hadra).
Duval, Loca sanctorum, 1:548, no. 25 (Kelibia); ibid., 1:11720, nos. 545 (Hadra); and in
general ibid., 2:6801; see also ibid., 1:3317, no. 157 (Kherbet Oum el Ahdam), a fourthcentury reference.
Duval, Loca sanctorum, 2:63357 and W. H. C. Frend, From Donatist Opposition to Byzantine
Loyalism: The Cult of Martyrs in North Africa 350650, in Merrills, Vandals, Romans, and
Berbers, pp. 2645.
Greg. Ep. 12.2, 2:96970.

342

The dilemma of dissent

saints in Africa in the Byzantine period.169 From the later fifth century
onwards, though, such western cults were increasingly joined by those of
numerous eastern martyrs, military saints, and healers, including Julian
of Antioch, Theodore the Recruit, Pantaleon of Nicomedia, Menas of
Alexandria, and Isidore of Chios.170 We hear too of soldiers serving in
Africa with particular devotions to the virgin Thecla and the Egyptian
healers Cyril and John.171 The cults of Theodore and Pantaleon were
popular enough that scenes from the saints lives were featured on a
series of decorative terracotta tiles that enjoyed widespread diffusion
in Byzacena and even parts of Africa Proconsularis in the Vandal and
especially the Byzantine periods.172 Above all, though, we find relics
of such overseas saints in sixth- and seventh-century depositions from
across Byzantine North Africa Proconsularis, Byzacena, Numidia, and
Mauretania Sitifensis often together with relics of local martyrs, but
sometimes in collections that are entirely foreign in their composition.173
Indeed, the last dated relic deposition in Africa took place in 636 in the
hill country of central Numidia, about forty kilometres south-west of
Constantina, and involved remembrances of Theodore and Stephen the
Protomartyr (since the translation of his remains to Africa in the fifth
century a local saint), as well as the Pontic saint Phocas and the Egyptians
Victor and Corona.174
In the Byzantine period, Africans similarly continued to participate
in an intellectual culture that still spanned the Mediterranean. This is
visible not only in the personal movements of individuals like Corippus,
Junillus, Primasius, and Hadrian, but also in the circulation of books and
manuscripts. Facundus of Hermianes defence of the Three Chapters
seems to have enjoyed a particularly rapid diffusion. The earliest extant
169

170
171

172

173

Duval, Loca sanctorum, 2:6458 with ibid., 1:67, no. 2 (Carthage); ibid., 1:1089, no. 51B
(Hadra); ibid., 1:13842, no. 64 (Rouis); ibid., 1:17982, nos. 878 (Guelma); ibid., 1:31719,
no. 150 (Mezloug) and ibid., 1:33941, no. 159 (Thamallula); see also ibid., 1:424, no. 19
(Dougga) and ibid., 1:3412, no. 160 (An Zeraba).
Duval, Loca sanctorum, 2:57880 and 65770; Frend, Donatist Opposition, p. 265.
John Moschus, Pratum spirituale 20, PG 87/3:2868A-B; Sophronius of Jerusalem, Miracula sanctorum Cyri et Ioannis c. 46, ed. N. Fernandez Marcos, in Los Thaumata de Sofronio: Contribucion
al estudio de la Incubatio cristiana (Madrid, 1975), pp. 3513.
R. du Coudray La Blanch`ere and P. Gauckler, Catalogue du Musee Alaoui, Catalogue des Musees
et Collections Archeologiques de lAlgerie et de la Tunisie 7 (Paris, 1897), p. 210, no. 13;
P. Gauckler, L. Poinssot, A. Merlin, L. Drappier, and L. Hautcoeur, Catalogue du Musee Alaoui,
Supplement, Catalogue des Musees et Collections Archeologiques de lAlgerie et de la Tunisie
15/1 (Paris, 1910), p. 281, nos. 879 and p. 283, nos. 11112; A. Merlin and R. Lantier, Catalogue
du Musee Alaoui, 2e Supplement, Catalogue des musees et collections archeologiques de lAlgerie
et de la Tunisie 15/2 (Paris, 1922), p. 281, no. 127 and p. 282, no. 137; C. Courtois, Sur un
carreau de terre cuite representant saint Pantaleon, Karthago 3 (19512), pp. 20913; Duval,
Loca sanctorum, 2:6636.
174 Ibid., 1:2319, no. 112 (Telergma).
Duval, Loca sanctorum, 2:57880.

343

Staying Roman
manuscript of the text dates to the second half of the sixth century,
and was probably written in Verona.175 Moreover, within a decade of
its publication in 548, the work was known and cited by authors in
Rome and Constantinople as well as in Africa. The chronicler Victor of
Tonnena mentions the publication of Facundus composition in his entry
for the year 550, at very much the same time that the Roman deacon
Pelagius drew on the African treatise to write his own, similarly entitled
In Defence of the Three Chapters.176 We have seen that Justinians edict
of 551 specifically sought to counter Facundus arguments (see above,
section 2.1), and, during his sojourn in Constantinople from c.540 to
c.554, Cassiodorus too became acquainted with the treatise, which he
mentioned in his own exposition on the psalms.177
The polemical nature of Facundus text may have made it something
of a special case, but the circulation of the few other books that we can
trace followed similar patterns. Cassiodorus exposition on the psalms
also quotes from a lost composition by Primasius of Hadrumetum on
heresy, and later the Italian official-turned-monk would make mention
of the same authors Commentary on the Apocalypse.178 The two men had
overlapped in Constantinople, and it is likely there that Cassiodorus first
learned of Primasius work; but we have also seen that some Vandalera African texts quickly made their way to Cassiodorus monastery
at Vivarium in Calabria (see above, Chapter 2.4.1), and, indeed, after
Fulgentius of Ruspes death his personal copy of Hilary of Poitiers
Trinitarian writings may have ended up in that same monastic library,
after first passing into the possession of Ferrandus of Carthage.179 It is
possible, then, that Primasius exposition on the Apocalypse did not find
175
176

177
178

179

CLA 4.506 and Graham, Dissemination, pp. 1547; see also CLA 1.8, 4.490 and 4.509, other
manuscripts in the same group.
Vict. Tonn. s.a. 550.142, p. 46; Pelagius, In defensione trium capitulorum 3, ed. R. Devreesse, Studi
e testi 57 (Vatican City, 1932), p. 29, at ll. 1012; and, on Facundus influence on this work in
general, L. Abramowski, Die Zitate in der Schrift In Defensione Trium Capitulorum der
romischen Diakons Pelagius, Vigiliae Christianae 10 (1956), pp. 16093.
Cassiodorus, Expositio psalmorum 138.24, ed. M. Adriaen, 2 vols., CCSL 978 (Turnhout, 1958),
at 2:1255; J. J. ODonnell, Cassiodorus (Berkeley, Calif., 1979), pp. 13176, esp. 16972.
Cassiodorus, Expositio psalmorum 118.1.2 and 138.24, 2:1060 and 2:1255; Cassiodorus, Institutiones 1.9.4, p. 33; Graham, Dissemination, p. 148. On Primasius life and work, see J.
Haussleiter, Leben und Werke Bischofs Primasius von Hadrumetum: Eine Untersuchung (Erlangen,
1887).
Vatican City, Archivio della Basilica di S. Pietro, MS Basilicanus D.182 (CLA 1.1a). On the
history of this manuscript, see A. Wilmart, LOdyssee du manuscrit de San Pietro qui renferme
les uvres de saint Hilaire, in L. W. Jones (ed.), Classical and Mediaeval Studies in Honor of
Edward Kennard Rand (New York, 1938), pp. 293305; F. Troncarelli, LOdissea di unodissea:
note sullIlario basilicano (Arch. S. Pietro D 182), Scriptorium 45 (1991), pp. 321; L. Boyle,
The Basilicanus of Hilary Revisited, in E. Condello and G. De Gregorio (eds.), Scribi e
colofoni: le sottoscrizioni di copisti dalle origini allavvento della stampa, Biblioteca del Centro per il

344

The dilemma of dissent

its way into Cassiodorus hands until after his return to Italy. In any case,
the work is revealing in other ways of the intellectual ties that linked
Africa to the rest of the Mediterranean world in the mid sixth century,
for it is heavily indebted to the exegesis of Primasius and Cassiodorus
older contemporary Caesarius of Arles.180 Primasius treatise also shares
one slightly altered line with the apocalyptic commentary of Apringius
of Pax Julia (mod. Beja, Portugal), composed under the Visigothic king
Theudis (ad 53148).181 In addition to testifying to the dissemination of
texts within the empire, then, Primasius composition would thus seem
to demonstrate the continued circulation of manuscripts between Africa
on the one hand and both Gaul and the Iberian Peninsula on the other.
Elsewhere, too, we catch glimpses of similar trans-Mediterranean intellectual ties. Thus, for example, a manuscript of Ferrandus Breviatio
canonum (composed in Carthage before 545) was copied out in the lower
Rhone valley in the sixth or seventh century. In Spain, Isidore of Seville
seems to have been well informed about the works of African authors,
while the mid seventh-century bishop Ildefonsus of Toledo claimed that
the Trinitarian writings of his predecessor Eugenius III would have been
sent to Africa and the East had a storm in the Strait of Gibraltar not prevented the journey.182 At much the same time, the non-resident Patriarch of Antioch, a Monothelete named Macarius, dedicated three books
against Maximus Confessors Dythelete doctrine to the African monk
Lucas, who, according to the Syriac life of Maximus, was the only man
in Africa who could be found to dispute with the orthodox theologian
in support of Monotheletism.183
The flow of letters and information followed similar patterns. The
nature of the relations between the provinces and the imperial centre must

180

181

182

183

collegamento degli studi medievali e umanistici in Umbria 14 (Spoleto, 1995), pp. 93105; and
Graham, Dissemination, pp. 14951.
Primasius of Hadrumetum, Commentarius in Apocalypsin, ed. A. W. Adams, CCSL 92 (Turnhout,
1985), passim, conveniently summarized ibid., Index scriptorum, pp. 3603; cf. Caesarius of
Arles, Expositio de Apocalypsi sancti Iohannis, ed. Morin, in Caesarii Arelatensis Opera omnia,
2:21077.
Primasius, Commentarius in Apocalypsin 1.1, ed. Adams, p. 20, ll. 2758; cf. Apringius of Beja,
Tractatus in Apocalypsin 1.16, ed. M. Ferotin, in Apringius de Beja: son commentaire de lApocalypse,
Biblioth`eque patrologique 1 (Paris, 1900), p. 13.
Munier, Concilia Africae, p. 286 with CLA 5.619; J. Fontaine, Isidore de Seville et la culture classique
dans lEspagne wisigothique, 3 vols. (2nd edn; Paris, 1983), 2:8579; Ildefonsus of Toledo, De
viris illustribus 13, ed. C. Codoner Merino, CCSL 114A (Turnhout, 2007), p. 615; Graham,
Dissemination, pp. 1757, 181, and 1878.
S. Brock, An Early Syriac Life of Maximus the Confessor, Analecta Bollandiana 91 (1973), p. 318
(c. 19), with commentary ibid., p. 326. On this Lucas, see also ACOec.2 2/1:504, 510 (Greek)
and ibid., pp. 505, 511 (Latin).

345

Staying Roman
have required substantial routine communications between Constantinople and Carthage, of which we are now able to catch only glimpses in
either the secular or the ecclesiastical sphere. We have seen, for example,
that the imperial cancellarius Theodore travelled from Constantinople to
Carthage with a letter from the empress Martina ordering the Praetorian Prefect of Africa, George, to release some Monophysite nuns a
letter which George chose to regard as a forgery. The seals of seventhcentury commerciarii are similarly revealing of communications between
the imperial and provincial capitals (see above, Chapter 4.3.4). So too
the evidence of another two sixth- and seventh-century lead seals from
Carthage. The first is that of a private imperial secretary (a secretis) named
George; the second that of the imperial nipsistiarios Theodore, one of the
eunuchs in attendance on the emperors person.184 In late 641 or early
642, Maximus Confessor sent a letter to another palace eunuch, to whom
he recommended the bearer, a vir inlustris named Theocharistus.185 In
550, an imperial messenger conveyed the African bishops defence of
the Three Chapters to Justinian.186 The African church also sent its own
representatives to the court. Justinians novel reversing the Vandal Arianizing policy makes reference to a Carthaginian deacon named Theodore
who had been sent to the imperial capital to secure a judgment in favour
of the return of church property.187 The deacon Evasius played a similar
role in terms of securing the privileges of the church of Byzacena in
568, under Justin II.188 Earlier, in 541, the Byzacenan bishops Restitutus
and Heraclius had sought an audience with Justinian to discuss the same
issue, resulting in the emperors first rescript to Datianus.189 Primasius of
Hadrumetum apparently also travelled to Constantinople on unspecified
business at about the same time.190 The same was later true of Archbishop Fortunius of Carthage, who visited the Queen of Cities in the
640s, during the patriarchate of the Monothelete bishop Paul II.191
Only occasionally are we able to see lateral communications between
or within provinces, but these certainly existed as well. At least fourteen
lead seals from Africa struck in the names of sixth- or seventh-century
184

185
186
188
190
191

George: A. Merlin, in BCTH (1925), xliixlvii, at xlv, no. 5. On the function of the a secretis, see
Proc. BP 2.7.15, 1:180 and Proc. Anecd. 14.4, 3:90, with ODB 1:204, s.v. Asekretis. Theodore:
Merlin, in BCTH (1925), xliv, no. 3. On the function of the nipsistiarios, see ODB 3:1488, s.v.
Nipsistiarios.
Maximus Confessor, Ep. 44, PG 91:644D, with PLRE 3:1225 s.n. Theocharistus 2. For seals
struck in Theocharistuss name, see PBE 1, s.n. Theocharistos 1.
187 Just. Nov. 37.1 (ad 535), p. 244.
Vict. Tonn. s.a. 550.141, p. 46.
189 Just. Nov. App. 2 (ad 541), pp. 7967.
Feissel, Acte de la prefecture, p. 112.
Junillus Africanus, Instituta, praef., p. 118; Maas, Exegesis and Empire, p. 6.
Concilium universale Constantinopolitanum tertium, in ACOec.2 2/2:652 (Greek) and 653 (Latin).

346

The dilemma of dissent

duces and magistri militum would originally have been attached to documents issued and circulated under these officials authority.192 In 633
or 634, Maximus Confessor wrote to the magister militum per Numidiam
Peter when the general was abroad, presumably involved in the defence
of Egypt against the Muslims.193 Earlier, in October 596, the exarch
Gennadius summoned a certain Ruferius comes and some of his fellowcitizens from Corsica to Africa. Gregory the Great wrote a letter on their
behalf, stressing to Gennadius that the Corsicans were already grieved
by the absence of the tribune Anastasius, who was apparently serving
under the exarchs command and whom Gregory asked be returned to
Corsica.194 As we have seen, Gennadius exploits had won him quite a
reputation in the late sixth century, and probably while in Italy but possibly while in Thrace the Suevic commander Droctulf had heard of the
general and secured a commendation to his service from Pope Gregory
(see above, Chapter 4.3.3).195
It is probably no mistake that the latter two cases involved communications directed to Africa through Rome, for, as we have seen, individuals
and letters circulated frequently between the Eternal City and the southern Mediterranean (see above, section 2). Such travel, however, was not
confined to petitioners and the adjudication of disputes. We have seen,
for example, that in 535, the African Nicene episcopate wrote to Pope
John II requesting his advice about how to deal with Arian priests who
now sought to embrace the Trinitarian confession. By the time the letter
arrived, though, John had died and it was his successor Agapetus who
192

193

194
195

A. Delattre and P. Monceaux, in Bull. Soc. Nat. Ant. Fr. (1912), pp. 3312, no. 1; Monceaux,
Enquete sur lepigraphie, p. 75, no. 15; J. Nesbitt and N. Oikonomides (eds.), Catalogue of
Byzantine Seals at Dumbarton Oaks and in the Fogg Museum of Art, vol. 1, Italy, North of the Balkans,
North of the Black Sea (Washington, DC, 1991), no. 6.1 (DO 47.2.1454), p. 35; G. Zacos and
A. Veglery, Byzantine Lead Seals, 2 vols. (Basle, 197285), 1/3:1643, no. 2885, and 1/3:16478,
nos. 2898ab; Corpus Inscriptionum Graecarum, ed. E. Curtius and A. Kirchhoff (Berlin, 1877;
repr. Hildesheim, 1977), 4:411, no. 8990 = Monceaux, Enquete sur lepigraphie, pp. 756,
no. 16; F. Icard, Sceaux et medailles de plomb trouves a` Carthage, BCTH (1917), p. 9, no. 33;
A. Delattre and P. Monceaux, in Bull. Soc. Nat. Ant. Fr. (1908), p. 164, no. 1 = Revue archeologique,
4th ser. 12 (1908), p. 449, no. 184 = ILCV 235b; A. Delattre and P. Monceaux, in Bull. Soc.
Nat. Ant. Fr. (1913), p. 316, no. 1 = ILCV 235a; A. Delattre and P. Monceaux, in Bull. Soc.
Nat. Ant. Fr. (1914), p. 284, nos. 1 and 2; F. Icard, Sceaux et plombs marques decouverts a`
Carthage, BCTH (1927), p. 479, no. 1 and A. Merlin and P. Monceaux, in Bull. Soc. Nat. Ant.
Fr. (1915), p. 300, no. 1; V. Laurent, Les Sceaux byzantins du Medaillier Vatican, Medagliere della
Biblioteca vaticana 1 (Vatican City, 1962), pp. 857, no. 92.
Maximus Confessor, Ep. 13, PG 91:50933; see also ibid., 14, PG 91:53344. On Peter, see in
general PLRE 3:1013, s.n. Petrus 70 and above, Chapter 4.3.1 n. 126. On the date of the letter
and on Peters movements, see P. Sherwood, An Annotated Date-List of the Works of Maximus the
Confessor, Studia Anselmiana 30 (Rome, 1952), pp. 3940.
Greg. Ep. 7.3, 1:4456.
On Droctulf s career in general, including his movements between defecting to the Byzantines
and requesting a post in Africa, see PLRE 3:4257, s.n. Droctulfus 1.

347

Staying Roman
responded to the inquiry.196 Over a century later, in the 640s, a bishop
named Mellosus travelled from Africa to Rome and back again bearing
the greetings of and a letter from Archbishop Victor of Carthage to Pope
Theodore I (ad 6429).197 In 594, an envoy of Bishop Dominicus of
Carthage brought news of an anti-heretical council in Africa Proconsularis to Gregory the Great.198 That same year, Gregory ordered an
African sceptic by the name of Boniface to come to Rome with his
companions in doubt in order to discuss the Christian faith.199
Africas position at the crossroads between the eastern and western
Mediterranean further made it a communications nexus for travellers
headed from one end of the sea to the other. In the 630s, a possibly fictitious Jewish clothes-merchant named Jacob, said to have been a native
of Acre in Syria, travelled from Constantinople to Carthage to sell his
wares.200 There are hints that Jacob or the shipmaster who gave him
passage may have been on his way further west.201 In the sixth or seventh
century, we also hear of a ship bearing architectural fittings from the
East to Marseilles that travelled by way of Africa.202 In the sixth century,
another, possibly bound for Africa itself, sank off Marzameni in eastern
Sicily.203 In 583, after the failed rebellion of her husband, the Visigothic
prince Hermenegild, against his father Leovigild, the Frankish princess
Ingundis fled with her infant son to the Byzantine territory of southern
Spain. Ingundis died in Carthage on the way to Constantinople, but her
young son Athanagild was taken to the imperial city where he was the
object of diplomatic negotiations between the Byzantine and Frankish
courts.204 In 589, Athanagilds uncle, the Frankish king Childebert II, sent
196
197

198
200

201
203

204

Collectio Avellana 857, pp. 32833.

Victor of Carthage, Epistula ad Theodorum papam, ed. R. Riedinger, in Concilium Lateranense
anno 649 celebratum, ACOec.2 1:100, ll. 1217 (Greek) and ibid., p. 101, ll. 1216 (Latin). The acts
of the Lateran Synod of 649 are probably a papal forgery: R. Riedinger, Die Lateranakten von
649: Ein Werk der Byzantiner um Maximos Homologetes, Byzantina 13/1 (1985), pp. 51734.
However, Victors letter is accepted as genuine by Dekkers: CPL no. 874 (with the note to CPL
no. 877).
199 Greg. Ep. 4.41, 1:262.
Greg. Ep. 5.3, 1:2689.
Doctrina Jacobi nuper baptizati 1.3, ed. and trans. G. Dagron and V. Deroche, Juifs et Chretiens
dans lOrient du VIIe si`ecle, Travaux et Memoires 11 (1991), p. 73. In the Slavic version, Jacob
is said to have arrived from Constantinople: ibid., p. 72.
202 Ibid., p. 108 with n. 104.
McCormick, Origins, p. 107.
G. Kapitan, Schiffsfrachten antiker Baugesteine und Architekturteile vor den Kusten Ostsiziliens, Klio 39 (1961), pp. 3003; Kapitan, The Church Wreck off Marzamemi, Archaeology
22 (1969), pp. 12233; Hayes, Late Roman Pottery, p. 265; A. J. Parker, Ancient Shipwrecks of the
Mediterranean and the Roman Provinces, BAR International Series 580 (Oxford, 1992), p. 267.
Gregory of Tours, Historiae 6.40, 8.18, 8.21, and 8.28, pp. 310, 384, 387, and 390; Epistolae
Austrasicae 435, ed. Gundlach in MGH Epist. 3:14951; and Paul the Deacon, Historia Langobardorum 3.212, pp. 1502 who claims that Ingundis died in Sicily. In Constantinople, Athanagild
received letters from his uncle (the Frankish king Childebert II) and grandmother (Brunhild),

348

The dilemma of dissent

Grippo, Bodegiselus, and Evantius on an embassy to the emperor Maurice. In Carthage, they had to gain the clearance of the Praetorian Prefect
to continue their journey to the imperial capital; and while delayed in
the African metropolis, one of Evantius servants killed an African merchant. The urban prefect raised a force to apprehend the murderer, and
in the siege of the Frankish envoys lodgings, Bodegiselus and Evantius
were killed. The incident sparked diplomatic tensions between Childebert and Maurice, despite the emperors promises to punish those who
had killed the Frankish envoys. Maurice even went so far as to send
twelve of the guilty men to Childeberts court in chains, with the offer
to let the Frankish king punish them as he saw fit. Childebert, however,
was not impressed, and Grippo (the only one of the envoys to escape
with his life) offered to return to Carthage in an effort to identify the
real killers.205
Other movements were more lasting. As we have seen, the circulation
of officials sent both to administer the North African provinces and to
defend them for the empire points to a kind of kinetic integration of
the region with the imperial capital that presumably continued into the
seventh century (see above, Chapter 4).206 But the Byzantine period also
witnessed the relocation of a number of individuals and families from the
eastern Mediterranean to Africa. At least occasionally they seem to have
made their way into the interior: in the sixth century or perhaps the
seventh century a seven-year-old girl from Constantinople known both
as Ulpia and Constantia was buried in Constantina in Numidia and commemorated there with a Greek epitaph.207 For the most part, though,
these travellers seem to have settled in the port cities along the Mediterranean littoral. Thus, for example, in 587 in Hippo Regius the grieving
father of a family from Lycia in Asia Minor erected a bilingual Latin
Greek epitaph in remembrance of his ten-year-old son Theodosius.208 A
mid sixth-century inscription from Hadrumetum commemorates a man
named Cyril whose date of death was recorded (in Greek) according to

205
206

207

208

both addressed to [my] sweetest child, Athanagild, the king (dulcissimo nepoti Athanagildo regi):
Epistolae Austrasicae 278, pp. 13940.
Gregory of Tours, Historiae 10.2 and 10.4, pp. 4823 and 4867; Paul the Deacon, Historia
Langobardorum 3.31, pp. 1646.
Kinetic integration: the quotation is from M. McCormick, The Imperial Edge: ItaloByzantine Identity, Movement and Integration, ad 650950, in H. Ahrweiler and A. Laiou
(eds.), Studies on the Internal Diaspora of the Byzantine Empire (Washington, DC, 1998), p. 31.
C. Wessel, A. Ferrua, and C. Carletti (eds.), Inscriptiones Graecae Christianae veteres Occidentis,
Inscriptiones Christianae Italiae Subsidia 1 (Bari, 1989), p. 6, no. 16 = Handley, Dying on
Foreign Shores, no. 396.
Supplementum Epigraphicum Graecum, vol. 9, ed. J. J. E. Hondius (Amsterdam, 1938), p. 111,
no. 872 = Handley, Dying on Foreign Shores, no. 375.

349

Staying Roman
the Egyptian calendar, and an anonymous Latin epitaph from the basilica
of Ste-Monique in Carthage is similarly dated to the Egyptian month
of Mesori.209 Both inscriptions probably testify to the continued presence of an Egyptian community in the eastern coastal zone of Byzantine
Africa. Another two epitaphs similarly attest to a Syrian presence. One,
also from Ste-Monique, commemorates a certain Porphyry who was said
to have been from the cities of Canatha and Bostra (mod. Qanawat and
Bosra, Syria); the second indicates that a family from Apamea buried
their nine-month-old son in the port city of Sullecthum.210
In at least a handful of cases, Africa also served as a place of exile for late
sixth- and seventh-century political undesirables, a fact that indicates the
confidence of Byzantine control over the region. According to Sebeos,
the Armenian nobleman Smbat Bagratuni was exiled to Africa by the
emperor Maurice after leading an abortive revolt against Byzantine rule
in Armenia in 589.211 Smbats son, Varaztirotsc , was similarly exiled to
Africa with his family some time around 635 for conspiring to overthrow the emperor Heraclius. He was later recalled from exile by Heraclius grandson and eventual successor, Constans II.212 After the death of
Constantine III, in 641, the imperial sacellarius Philagrius was banished
to Africa according to the ninth-century Breviarium of Nicephorus,
to the fortress of Septem on the Strait of Gibraltar.213 In all of these
cases, it would appear that the imperial government was attempting to
remove the given individual as far as possible from his previous sphere of
activity.214
Whether in exile or in flight, or simply in search of a more congenial
atmosphere, the movement of eastern monastic and ecclesiastical figures
to and from Africa was also a pronounced feature of the provinces
209

210
211
212
213
214

Terry, Tomb Mosaics, pp. 6323, no. 185 = Handley, Dying on Foreign Shores, no. 384
(Hadrumetum); L. Ennabli, Inscriptions chretiennes de Sainte Monique, p. 228, no. 91B = Handley,
Dying on Foreign Shores, no. 365 (Carthage).
L. Ennabli, Inscriptions chretiennes de Sainte Monique, pp. 2623, no. 117A = Handley, Dying on
Foreign Shores, no. 366 (Carthage); ibid., no. 393 (Sullecthum).
Sebeos, History 20, trans. R. Thomson in The Armenian History Attributed to Sebeos (Liverpool,
1999), pp. 3840.
Sebeos, History 41 and 44, pp. 923 and 107.
Nicephorus, Breviarium 30, p. 80 with ibid., 29, pp. 7980; John of Nikiu, Chronicle, 119.23,
p. 191.
The legend, reported by Eutychius of Alexandria (Sacd ibn Bat.rq), Annales ecclesiasticae
26.266, ed. and trans. M. Breydy, in Das Annalenwerk des Eutychios von Alexandrien. Ausgewahlte
Geschichten und Legenden kompiliert von Sac id ibn Bat.rq um 935 ad, CSCO (Scr. Arab.) 445
(Louvain, 1985), 44:11315, of a brigand who was pardoned, granted a dignity by the emperor
Maurice, and went to live in Constantinople does not relate to Africa, contra PLRE 3:1446, s.n.
Anonymus 106 (an attribution based on L. Cheikhos reading of Ifrqiya, CSCO [Scr. Arab.]
ser. 3, 67 (Beirut, 19069; repr. Leiden, 1954), 6:211, which Breydy, Annalenwerk, 45:94, n. 1
notes is the result of a mistaken transcription of Atraqiya).

350

The dilemma of dissent

relations with the outside world in the seventh century. Early in the
century the Syrian monks Eshac ya and Ishoc had moved from Nisibis
to Hippo Diarrhytus in Africa Proconsularis (mod. Bizerte), where they
founded a monastery. By c.628, when Maximus Confessor first came to
Africa, it boasted eighty-seven monks, though it is not clear whether
they were Africans or immigrants.215 Maximus was himself originally
from Palestine, the son of a Samaritan and a Persian slave girl according
to a near-contemporary, though hostile, Syriac life.216 It was presumably
the joint attack of the Avars and Persians on Constantinople in 626
that precipitated Maximus flight west.217 There Maximus seems to have
met his friend and correspondent Thalassius, a shadowy individual
seemingly also an easterner resident in Africa, who figures prominently
in a seventh-century account of a miraculous vision of hell granted to a
local magistrates bailiff during an outbreak of plague in Carthage in the
reign of Heraclius.218 It also appears to have been in Africa that Maximus
met the Syrian monk and future patriarch of Jerusalem Sophronius and
became his disciple.219 It is just possible that Sophronius own spiritual
father John Moschus spent some time in Africa, too, but this is far from
clear.220 As we have seen, in 641 a fugitive community of Monophysite
nuns from Alexandria as well as other Syrian and Egyptian heretics had all
settled in Africa, presumably fleeing the Persian and Muslim incursions
into their native provinces earlier in the century (see above, Chapter
4.3.4). In a rather curious turn of events, some time around 642, the
deposed Patriarch of Constantinople Pyrrhus too made his way to Africa,
where he disputed with his theological arch-rival Maximus Confessor on
the Monothelete controversy before the exarch Gregory.221
215
216
218

219

220

221

Brock, Syriac Life, p. 317 (c. 19) with commentary ibid., p. 326 and Sherwood, Annotated
Date-List, pp. 56.
217 Sherwood, Annotated Date-List, pp. 2728.
Brock, Syriac Life, p. 314 (c. 1).
Anastasius the Sinaite,  
     40, pp. 837; George the Monk,
Chronicon 4.231, PG 100:8415; Synaxarium Constantinopolitanae s.d. April 29, cols. 63740
(Synax. Select. Mc) and ibid., s.d. Apr. 30, cols. 6412 (Synax. Select. Bb, C and Cg)(BHG
13181318a); Kaegi, Heraclius, pp. 1025. On Thalassius, see also his Centuriae, PG 91:1428
69; Maximus Confessor, Epp. 9, 26, and 402, PG 91:4459, 61617, and 6337; Maximus,
Quaestiones ad Thalassium, ed. C. Laga and C. Steel, 2 vols., CCSG 7 and 22 (Turnhout, 1980

90); M.-T. Disdier, Le Temoignage spiritual de Thalassius le Lybien, Etudes

byzantines 2 (1944),
pp. 79118; and M. van Parys, Un Matre spirituel oublie: Thalassios de Libye, Irenikon 52
(1979), pp. 21440.
Maximus Confessor, Opuscula Theologica et Polemica, PG 91:142A; Maximus, Ep. 13, PG
91:533A; Sherwood, Annotated Date-List, pp. 289; and H. Chadwick, John Moschus and
his Friend Sophronius the Sophist, JThS n.s. 25 (1974), p. 52 n. 4.
See Chadwick, Moschus and Sophronius, pp. 589 and Cameron, Byzantine Reconquest of
N. Africa, p. 160; Moschus may at least have stopped in Africa on his way to Rome, if at no
other time.
Maximus Confessor, Disputatio cum Pyrrho, PG 91:288353.

351

Staying Roman
Africa, far to the west and as yet safe from invasion, was a refuge
for these easterners. But Moorish and, later, Muslim raids ensured that
the movement was not always towards Africa. In the 540s, most of the
notables of Hadrumetum fled to Constantinople in the face of Moorish
raids into Byzacena.222 In the later sixth century, the monk Nanctus
fled to Visigothic Spain, where the Arian king Leovigild (ad 56986)
granted him the land to establish a monastery.223 Some time around 570,
the monk Donatus also fled the hostile activity of the Moors in Africa and
sailed to Spain, along with seventy monks and a large library, where he
established the monastery of Servitanum.224 John Moschus tells the story
of a soldier in the Byzantine army who, overcome by a Moorish warrior
in battle, vowed that should he escape with his life he would become
a monk in Jerusalem. The Moorish warrior miraculously disappeared,
and the dragoon fulfilled his promise.225 Sophronius similarly tells of
an African tribune who went blind and travelled east to bathe in the
pool of Siloam in Jerusalem, where he was cured.226 In the wake of
the first Muslim incursions into Africa, the Syrian monks at Hippo
Diarrhytus were said to have abandoned the North African provinces for
Rome, where they founded a monastery dedicated to St Sabas.227 Abbot
Hadrian may well have left Africa for Campania at this time too.228 To
judge from the numismatic evidence, on the other hand, later flight from
the Muslim conquest appears to have focused on Sardinia and perhaps
southern Gaul.229
The movements of individuals, letters, and books; administrative ties,
both secular and ecclesiastical; the character, composition, and political identity of the local notable class; spiritual devotion and economic
exchange in the sixth and seventh centuries all of these speak of the
successful reintegration of Africa into the structures of the Byzantine
empire. As was the case throughout the Mediterranean world, those
who wielded authority and influence in African society were increasingly members of the military and ecclesiastical elite rather than representatives of the old senatorial aristocracy or even the curial order. But
insofar as we can tell, these individuals seem to have allied their interests
with the restored empire. Predominantly Christian, their attachments to
222
223
224
225
226
227
228
229

Proc. BV 2.23.29, 1:529.

Vitas sanctorum patrum Emeretensium 3.2 and 3.810, ed. and trans. J. Garvin, Studies in Medieval
and Renaissance Latin Language and Literature 19 (Washington, DC, 1946), pp. 156 and 158.
Ildefonsus, De viris illustribus 3, p. 605.
John Moschus, Pratum spirituale 20, PG 87/3:2868AB.
Sophronius, Miracula ss. Cyri et Ioannis 46, pp. 3513.
Brock, Syriac Life, pp. 31819 (c. 24), with commentary ibid., pp. 3289.
Bischoff and Lapidge, Biblical Commentaries, pp. 912, though see also above, n. 159.
McCormick, Origins, pp. 3546.

352

The dilemma of dissent

the cults of particular saints were becoming ever more international in
character and speak of a growing enthusiasm for the holy dead of the
East flourishing alongside a continued attachment to the martyrs of the
city of Rome. In other areas, too, the networks that we can see most
clearly linked Carthage to the two old poles of Rome and Constantinople; but a web of connections spread out from Africa and led, directly or
indirectly, to Sicily, Spain, Gaul, Italy, Lycia, Syria, Palestine, and Egypt.
The communications that bound these regions together were routine
and extraordinary, official and personal, ecclesiastical and economic; the
travellers whose movements we can trace spanned the social elite, and
included office-holders, diplomats, churchmen, merchants, exiles, and
refugees. They seem to indicate by their movements that Africa was
well-integrated into the Byzantine world, for all the provinces problems
of communication with the capital. But, by the later seventh century,
our travellers also indicate one of the consequences of that integration:
Africa had ceased to be a sanctuary from the troubles that beset the East.
4. rome and her enemies
In the seventh century, the Byzantine empire was a society under
pressure.230 Monophysitism continued to be a live issue, but now it was
joined by closely related debates over Monotheletism, which confessed
that Christ had a single will, and Monoenergism, which recognized only
one energy in the incarnate Word. Between 613 and 619, Persian armies
made incursions into Asia Minor and seized control of Antioch, Damascus, Jerusalem, and, finally, Egypt, which was to remain in Persian hands
for the better part of the following decade. Though the empire rallied
under Heraclius, decisively defeating the Persians in 628, the victory was
a fleeting one. The years between 634 and 646 once again saw the loss
of Syria and Egypt, this time to the Arab armies of the Muslim caliphs
c
Umar and c Uthman. By 647, Africa, too, was threatened.231 A century
earlier, Romanness had been defined in political terms by loyalty to the
230

231

See in general J. Haldon, Byzantium in the Seventh Century: The Transformation of a Culture (Cambridge, 1990). On seventh-century Africa, see Averil Cameron, The Byzantine Reconquest of
N. Africa and the Impact of Greek Culture, Graeco-Arabica 5 (1993), pp. 15365, reprinted in
her Changing Cultures in Early Byzantium (Aldershot, 1996); and W. Kaegi, Society and Institutions in Byzantine Africa, in P. Corrias and S. Cosentino (eds.), Ai confini dellimpero. Storia,
arte e archeologia della Sardegna bizantina (Cagliari, 2002), pp. 1528.
On the Islamic invasions, see F. Donner, The Early Islamic Conquests (Princeton, NJ, 1981); W.
Kaegi, Byzantium and the Early Islamic Conquests (Cambridge, 1992); V. Christides, Byzantine
Libya and the March of the Arabs towards the West of North Africa, BAR International Series 851
(Oxford, 2000); and now W. Kaegi, Muslim Expansion and Byzantine Collapse in North Africa
(Cambridge, 2010).

353

Staying Roman
empire.232 Now, under the strain of the Persian campaigns and the early
Islamic conquests, the loyalty of Romes subjects was called sharply into
question.
In 632, Heraclius issued an edict that all Jews be baptized. This forced
conversion was probably inspired at least in part by an apocalyptic vision
of the events of the previous twenty years. But there are also indications
that from an imperial point of view the political loyalty of Jews to
the empire was at stake. Are you the servants of the most clement
emperor?, the Praetorian Prefect George (or Sergius) asked the Jews of
Carthage. Are you not his obedient subjects? Informing them of the
emperors edict, the prefect continued, The emperor desires that you
be baptized. . . . So you are not his faithful subjects, for you do not obey
your master.233 In the seventh century, from a Byzantine point of view,
a Christian must by definition love the emperor.234 However, opposition
to Heraclius edict was naturally widespread within the Mediterranean
Jewish community. This is the context of the Doctrina Jacobi nuper baptizati
and its story about an eponymous eastern clothes-merchant in Carthage
who avoided conversion by passing himself off as a Christian until one
night he stumbled on his way home and cried out, Adonai! Jacob was
then secretly followed to the baths where he was found to be circumcised,
and so was arrested, imprisoned, and forcibly baptized. After witnessing
a miraculous vision, Jacob experienced a conversion of the heart as
well, and began to proselytize the Christian faith among the Jews of the
Levant.235
232

233

234

235

G. Greatrex, Roman Identity in the Sixth Century, in S. Mitchell and G. Greatrex (eds.),
Ethnicity and Culture in Late Antiquity (London, 2000), pp. 26792, esp. 268; see also above,
Chapters 4 and 5.
Doctrina Jacobi 1.2, pp. 713, after the Arabic version of this passage, lost in the Greek. The
Slavic version contains substantially the same account as the Arabic: ibid., pp. 702. On the
North African Jewish community in late antiquity, see: W. H. C. Frend, The Christian Period
in Mediterranean Africa, c.ad 200 to 700, in J. D. Fage and R. Oliver (eds.), The Cambridge
History of Africa, 8 vols. (Cambridge, 197586), 2:4512; the papers in C. Iancu and J.-M. Lass`ere
(eds.), Juifs et judasme en Afrique du Nord dans lantiquite et le haut moyen-age (Montpellier, 1985);
J. Lund, A Synagogue at Carthage? Menorah-Lamps from the Danish Excavations, Journal
of Roman Archaeology 8 (1995), pp. 24562; J.-P. Darmon, Les Mosaques de la synagogue de
Hammam Lif: un reexamen du dossier, in R. Ling (ed.), Fifth International Colloquium on Ancient
Mosaics, Held at Bath, England, on September 512, 1987, 2 vols., Journal of Roman Archaeology
Supplementary Series 9/Colloque international pour letude de la mosaque antique 5 (Ann
Arbor, Mich., 1995), 2:729; and now K. Stern, Inscribing Devotion and Death: Archaeological
Evidence for Jewish Populations of North Africa, Religions in the Graeco-Roman World 161
(Leiden, 2008).
Relatio factae motionis inter domnum Maximum monachum et socium eius coram principibus in secretario,
ed. P. Allen and B. Neil, in Scripta saeculi VII vitam Maximi Confessoris illustrantia, Corpus
Christianorum Series Graeca 39 (Turnhout, 1999), p. 13, ll. 1219.
Doctrina Jacobi, esp. 1.14, pp. 715 (Arabic version) and 704 (Slavic version). Maximus Confessor had opposed the forced conversion of the Jews as well: R. Devreesse, La Fin inedite

354

The dilemma of dissent

It was not, however, the Diaspora Jewish community that was to
pose the most serious internal threat real or imagined to Byzantine control of North Africa in the seventh century. Probably in 646,
Gregory, the exarch of Africa, rebelled against the emperor Constans II.
The precise circumstances surrounding the rebellion are unclear, but
Gregory seems to have proclaimed himself emperor.236 In this, he probably sought to follow in the footsteps of Constans great grandfather
Heraclius, another African exarch, who, over the course of 60810,
launched a successful revolt that ended in the capture of Constantinople,
the execution of the reigning emperor Phocas, and the coronation in
his stead of the exarchs eponymous son, the emperor Heraclius.237 In
the 640s, Gregory may also have presented himself as the champion of
Chalcedonian orthodoxy. Earlier in the decade, Gregory had apparently
been convinced of the Chalcedonian position in the disputation between
Maximus Confessor and Pyrrhus, the deposed patriarch of Constantinople, which was held in Carthage before the exarch and bishops of the
province.238 Eight years later, Maximus was himself accused of having
sent an abbot named Thomas from Rome to Africa to recount to Gregory a dream of Maximus foretelling the exarchs triumph should he
rebel against the emperor.239 But, whatever Maximus relation to the
rebellion, Gregorys insurrection was short lived. In 647, the exarch was
defeated in battle by the invading Muslims. He was probably killed as
well, though the medieval Syriac sources indicate that he escaped with
his life and effected a reconciliation with Constans.240 In either case, his
bid for empire was effectively over.
By June of 654, Maximus Confessors own loyalty to the emperor
was on trial. Indeed, at his arraignment before the Senate in Constantinople, the theologians accusers claimed that Maximus had hated
and betrayed the two successive emperors. As evidence they pointed to
hearsay from the confessors sojourn in the West. John, the former sacellarius of Maximus correspondent, the magister utrius militiae per Numidiam

236
237
238

239
240

dune lettre de Saint Maxime: un bapteme force de Juifs et de Samaritains a` Carthage en 632,
Revue des sciences religieuses 17 (1937), pp. 2535.
Theophanes, Chronographia AM 6139, 1:343:  0 $.
Kaegi, Heraclius, pp. 3757.
Maximus Confessor, Disputatio cum Pyrrho, col. 353A; Kaegi, Byzantine Military Unrest, p. 159;
Cameron, Byzantine Africa, p. 57. On the Christian context of the seventh-century Arab raids
more broadly, see Kaegi, Muslim Expansion, pp. 6991.
Relatio motionis, ed. Allen and Neil, p. 17, ll. 5362.
Theophanes, Chronographia AM 6139, 1:343; Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus,
p. 42; and al-Baladhur, Kitab futuh al-buldan, p. 227. Syriac sources: Michael the Syrian, Chronicle 11.10, 2:4401; Chronicon 1234 c. 126, pp. 2034; Bar Hebraeus, Chronography, 1:97. So
too Agapius, Kitab al-c Unwan 8:479 (writing in Arabic). On the Arab raid, see most recently
Christides, Byzantine Libya, pp. 403 and Kaegi, Muslim Expansion, pp. 11644.

355

Staying Roman
Peter, reported twenty-two-year-old army gossip to the effect that Maximus had written to the general to dissuade him from taking his army to
defend Egypt and Tripolitania against the invading Muslims in 632. God,
Maximus was said to have claimed, was displeased with Heraclius and his
family.241 Sergius Magudas raised the question of Maximus dream vision
foretelling Gregorys triumph over the emperor.242 Maximus denied both
allegations, but his close association with such politically questionable figures as the Praetorian Prefect George (who was recalled from Africa to
Constantinople in 642 to answer what Maximus regarded as trumped-up
charges) cannot have helped his case.243 In any event, the more political actions alleged against Maximus were said merely to be the outward
manifestations of the theologians lack of faith from the point of view
of the imperial court and his consequent hatred of the emperor and his
Church.244
Maximus Confessor had been the single most vocal supporter of Chalcedonian orthodoxy against the Monophysites and the Monotheletes: an
inspiration to the churches of Africa and the West, but a threat to imperial
control of the Church. The central issue in Maximus first trial, indeed,
was the confessors rejection of the notion that the emperor could define
Christian doctrine. Maximus position on this point was exactly parallel to that adopted by the African theologian Facundus of Hermiane a
century earlier. Priests alone could establish orthodox dogma, and the
emperor was not a priest.245 Moreover, the specific doctrinal formulations promulgated under the house of Heraclius over the course of
the seventh century the Nine Chapters of Alexandria (ad 633), the
Ecthesis (ad 638), and the Typos (ad 647) had, in Maximus estimation, abrogated the decisions of the first four ecumenical councils and
removed the patriarchate of Constantinople from communion with the
orthodox.246 Maximus went out of his way to assert that it was not the
emperor himself who was responsible for these heretical formulations:
the Ecthesis had been written by Patriarch Sergius of Constantinople,
not by the emperor Heraclius; the Typos, by highly placed ecclesiastical
hierarchs under Constans II, in whose name it had been issued. The
confessor claimed that he had himself never actively encouraged anyone
to break communion with Constantinople, and strenuously asserted that
his condemnation of the Typos in no way redounded upon Constans.247
But Maximus theology was too dangerous, too divisive for an emperor
241
243
244
246

242 Ibid., p. 17, ll. 5362.

Relatio motionis, ed. Allen and Neil, p. 15, ll. 2843.
Maximus Confessor, Epp. 1, 12, 16, 18, and 445, PG 91:36492, 460509, 5767, 5849, and
6419.
245 Ibid., pp. 217, ll. 112206.
Relatio motionis, ed. Allen and Neil, p. 13, ll. 1226.
247 Ibid., pp. 3943, ll. 34482 and p. 45, ll. 4227.
Ibid., pp. 313, ll. 24964.

356

The dilemma of dissent

desperate to enforce unity on the Church by whatever means necessary.
The confessor was exiled and eventually mutilated, dying at the age of
eighty-five in 662.
To the African bishops, as to their mouthpiece, Maximus, it was the
defence of Chalcedonian orthodoxy that was paramount. Included in
the forged acts of the Lateran Synod of 649 was a letter apparently
genuine from the archbishop of Carthage, Victor, opposing Paul (ad
64153), the Monothelete patriarch of Constantinople.248 The bishops
of Africa Proconsularis seem to have gone one step further, writing to
Paul himself in a further attempt to convince the patriarch to abandon
Monotheletism.249 Finally, at some point in the mid seventh century, the
church of Carthage received a letter from Pope Martin I (ad 64955)
commending their efforts on behalf of the universal faith.250 Orthodoxy
thus seems to have been the concern of the African episcopate to the very
end. These are the last glimpses that we get of the assembly of African
ecclesiastical hierarchs before the Islamic period.251
Neither the rebellion of Gregory nor the intractably Chalcedonian
stance taken by the African church against the Christological definitions
of the emperors saw Africa abandoned by the Byzantine empire. In 662,
Constans II (the same emperor against whom Gregory had rebelled sixteen years earlier) moved the imperial residence from Constantinople to
Syracuse. He was said to have intended to restore the capital of the empire
to the city of Rome,252 and the choice of a western seat must at least in
part have been motivated by a desire to secure the African and Sicilian
grain supply. The decision was not entirely without precedent. Constans
grandfather, the emperor Heraclius, had allegedly considered returning
to Africa in 619 when the Persian conquest of Egypt brought to an end
248

249

250
251

252

Victor, Epistula ad Theodorum, pp. 98103. A letter of similar tenor putatively sent by the African
bishops Columbus, et al., Epistula ad Theodorum papam, ed. R. Riedinger, in ACOec.2 1:6671
and another from Stephen and the bishops of Byzacena, Epistula ad Constantinum Augustum, ed.
Riedinger, in ACOec.2 1:749, have been rejected by H. J. Frede as a forgery: CPL nos. 8756
(with the n. to no. 876).
Gulosus, Probus, and the bishops of Africa Proconsularis, Epistula ad Paulum episcopum Constantinopolitanum, ed. R. Riedinger, in ACOec.2 1:8095; accepted by Dekkers as genuine: CPL
no. 877.
Martin I, Ep. 4, PL 87:14553.
Cresconius Concordia Canonum, attributed by Dekkers (CPL no. 1769) to c.690 on the basis
of common agreement, belongs to the mid sixth century: see K. Zechiel-Eckes, Die Concordia
canonum des Cresconius: Studien und Edition, 2 vols., Freiburger Beitrage zur mittelalterlichen
Geschichte 5 (Frankfurt am Main, 1992) and L. Kery, Canonical Collections of the Early Middle
Ages (c.4001140): A Bibliographical Guide to the Manuscripts and Literature (Washington, DC, 1999),
pp. 337.
Theophanes, Chronographia AM 6153, 1:348. On Constans II, see Kaegi, Muslim Expansion,
pp. 16699.

357

Staying Roman
the provinces annual annona payments to the imperial capital.253 Arab
raids had continued in Africa into the 660s, and were even said once
again to have reached central Byzacena.254 However, the next major
attack seems not to have occurred until about the time of Constans
assassination in Syracuse in 668 and the reversion of the imperial capital
to Constantinople. The raid, in which Muslim forces were said to have
captured 80,000 or even 100,000 prisoners, may have been accompanied
by an attack on Sicily as well.255 It was followed by the establishment of
the new Islamic city of Kairouan in central Byzacena. Only then were
Arab warriors said to have begun remaining year-round in Africa: previously they had returned to Egypt each winter.256 The following thirty
years saw further raids and counter-raids, through the course of which
the empire seems generally to have maintained its grasp over the cities of
the coastal littoral, but gradually to have lost control of the interior.257
As Vassilios Christides has recently observed, though, the Arabic sources
indicate that the Afariqa, the Romano-African population, consistently
and wholeheartedly sided with the Byzantines in these struggles.258
Finally, in 697, Muslim forces captured Carthage. Even so, though,
the imperial administration was not willing to surrender Africa without
a fight. That same year, the emperor Leontius (ad 6958) mounted a
massive counter-attack under the patricius John that recovered Carthage
and reoccupied both the metropolis and the surrounding region. The
Byzantine troops wintered in Africa, but the following year the caliph
c
Abd al-Malik sent a still larger force to recover the province for Islam.
The Byzantines were once again expelled, and withdrew to Crete, perhaps to await reinforcements. However, Johns army rebelled, elected the
sub-commander (droungarios) Tiberius III Apsimar emperor, and sailed
on the imperial capital. The military coup succeeded, and as with
Majorian nearly two and a half centuries earlier Leontius paid the price
for his military failure. The emperor was deposed, mutilated, and confined to a monastery.259 The withdrawal of the provincial administration
253
254
255
256

257
258
259

Nicephorus, Breviarium 8, p. 48.

Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, pp. 5660; Christides, Byzantine Libya, p. 43;
and Kaegi, Muslim Expansion, pp. 1112. See also al-Baladhur, Kitab futuh al-buldan, pp. 2278.
Theophanes, Chronographia AM 61601, 1:3512; see also ibid., AM 6155, 1:348, which almost
certainly must have happened after the assassination of Constans II, and below, next n.
Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, pp. 608; Ahmad ibn-Yahya al-Baladhur, Kitab
futuh al-buldan, pp. 2278; Christides, Byzantine Libya, pp. 434; and Kaegi, Muslim Expansion,
pp. 1315 and 2206.
Christides, Byzantine Libya, pp. 456 and Kaegi, Muslim Expansion, pp. 22646.
Christides, Byzantine Libya, pp. 634.
Nicephorus, Breviarium 41, pp. 98100; Theophanes, Chronographia AM 6190, 1:3701; Ibn
c Abd al-Hakam, Fut
uh Ifrqiya wal-Andalus, p. 76; Christides, Byzantine Libya, pp. 469; Kaegi,
Muslim Expansion, pp. 24765.

358

The dilemma of dissent

(and mint) to Sardinia kept Byzantine Africa on the books for a little
while longer, but, by the opening years of the eighth century, direct
imperial control of the African mainland had ended for good.260
Persians and Muslims, Maximus Confessor and his followers, Jews, the
exarch Gregory: these were Romes perceived enemies, without and
within. But from an African perspective, things looked very different
indeed. Maximus Confessor, at least, was a paragon of Chalcedonian
orthodoxy who received the enthusiastic support of the African church.
Gregorys rebellion is more of an open question. The sources give us no
real sense as to how far Africans supported their rebel exarch. If Gregory
did pose as the defender of Chalcedon, it is not unreasonable to expect
that the African episcopate may have regarded his insurrection with at
least cautious optimism. But it is also important to remember that as late
as the year before Gregorys rebellion that same episcopate had not yet
abandoned all hope of convincing Constans II and his patriarch to abandon Monotheletism. Just as the empire did not abandon Africa, so too
Africa remained engaged with the empire; and at the Sixth Ecumenical
Council at Constantinople (ad 6801) the African bishops finally saw
the triumph of their vision of the universal church, and the condemnation once and for all of the Monothelete heresy. And though there
was no love lost between Maximus Confessor (and, we can only imagine, the rest of the African Christian ecclesiastical hierarchy) on the one
hand and Africas Jews on the other, it was the Persian and Muslim wars
that inspired the most apocalyptic visions on the part of seventh-century
Byzantines living in Africa. The Roman or Byzantine empire was seen
as the fourth kingdom in the vision of Daniel. Barbarian victories heralded not just the diminution of the empire but the end of time.261 To
Maximus Confessor, they represented the coming of Antichrist.262
The Byzantine elite in North Africa remained integrated into the larger
imperial world right up until the very end. Ideologically, the Byzantine
reconquest was an unparalleled success certainly, within the empire but
(insofar as it provoked any comment there at all) also beyond the imperial
frontiers. Within Africa itself, perhaps the most vociferous proponent of
the restored empire was the poet Fl. Cresconius Corippus. His Iohannis
is a testament to the almost principled lack of differentiation in the
260
261

262

McCormick, Origins, p. 356.

Doctrina Jacobi 3.8, 3.10, 3.12, 4.5, and 5.5, pp. 16573, 1813, and 1913; see also ibid., 5.1,
pp. 1835 and the commentary at pp. 2656. The association of the Roman empire with the
fourth kingdom had already been made by Augustines day: De civitate Dei 20.23, 48:742.
Maximus Confessor, Ep. 14, PG 91:540B.

359

Staying Roman
poets mind between Romano-Africans and Byzantines. As we have seen
(above, Chapter 4.2.1), Corippus consistently refers to the Byzantine
magister militum Africae Artabanes as the Armenian (Armenius); but, for
the most part, the meaningful distinction in Corippus thought world was
between Romans and non-Romans; between the Mediterranean culture
shared by the Romano-Africans and Byzantines on the one hand and,
on the other, the strange, frightening, and somewhat repulsive world of
the Moors (see above, Chapter 5). Nor was Corippus alone among the
Romano-African elite in his enthusiasm for the empire: this he shared
with Junillus Africanus, Fulgentius the Mythographer, and seemingly a
whole host of local urban notables.
Perhaps more telling, though, is the fact that both in their words and
in their actions African bishops, priests, and deacons sought with unrelenting determination to ensure the empire-wide acceptance of Chalcedonian orthodoxy throughout the sixth and seventh centuries. Indeed,
African churchmen seem fundamentally to have accepted the empire
as such: they frequently appealed to the Pope and even sometimes to
the emperor to adjudicate disputes; Justinian was able quickly to bring
the African church to heel; and in any case African theologians had
expended considerable time and effort in an attempt to convince the
emperor that his ecclesiastical course was misguided. Taken together, the
evidence would seem to suggest that African resistance to the empire was
not systemic but rather was directed at the specific policies of particular
emperors. Even when defensive of their own conciliar autonomy, African
ecclesiastics understood and accepted the fact that they were themselves
implicated in the larger imperial and papal structures of power that for
them still bound the Mediterranean together.
As for disloyalty to the empire or at least to different emperors this
certainly existed as well in the sixth and seventh centuries. Even after the
endemic military revolts of the first decade of Byzantine rule had come
to an end, the seventh century saw the rebellions of the exarchs Heraclius
and Gregory, the recall (in what seems to have been disgrace) of the Praetorian Prefect George, and the trial of Maximus Confessor for treason.
For the most part, when not simply the product of imperial paranoia, the
untrustworthiness of high officials can probably be attributed to the lusts
and anxious desires of ambition. Maximus Confessor, however, represents
something of a different story. The account of his trial that survives was
written by one of his sympathizers, and therefore it is hardly surprising
that the accusations of treason do not really seem to stick. Even so, the
primary issue at stake in Maximus trial seems to have been his refusal to
concede to the emperor the right to define orthodoxy. In this, Maximus
situation was strikingly parallel to that of the empires Jewish subjects.
360

The dilemma of dissent

To judge from the Doctrina Jacobi, North African Jews (and, indeed, Jews
throughout the empire) saw themselves as no less loyal to the emperor
for having refused to convert to Christianity. The emperor and his representatives, however, appear not to have seen the recalcitrance either of
Jacob and his co-religionists or of Maximus and his companions in so
non-threatening a light. In the face of pressure, competing definitions
of Romanness were fast collapsing in on themselves. Under the Vandal
regime, loyalty to the emperor and adherence to the imperially sponsored
interpretation of Christianity provided related but distinct definitions of
what it meant to be Roman in Africa both of which, of course, were
inimical to the Vandal kings. Under the restored imperial regime, loyalty
to the emperor continued to be an important element in the definition
of Romanness; but, increasingly over the sixth and seventh centuries,
loyalty to the emperor and adherence to his interpretation of the faith
were coming to be synonymous.

361

Chapter 7

AFTER MATH

Your nobility sent us a letter within the year asking that we ordain the
priest Servandus as bishop according to the Christian system. We have
undertaken to do this, since your request seemed fair and honest.1

Thus wrote Pope Gregory VII in 1076 in one of the hundreds of letters
preserved in his papal register. This letter, however, is perhaps most
remarkable not for its author but for its recipient: one Anazir (al-Nasir),
a Muslim ruler whose kingdom was centred on the Atlas Mountains of
what is now eastern Algeria. In Gregorys mind, the territory was still
called Mauretania Sitifensis, while the bishopric in question was the port
city of Hippo Regius (mod. Annaba). Nor does it seem that the see had
simply been reconstituted in Gregorys own day. Insofar as it is possible to
tell, a Christian community had continued to exist at Hippo for centuries
after the Islamic occupation.
The cultural reorientation of North Africa in the wake of the Islamic
invasion had corresponded only roughly with the Byzantines loss of
political control over the region. It was, moreover, a gradual process.
Indeed, our sources for early Islamic North Africa, such as they are,
suggest that the transition to Arab rule may not have been so radically
different from the Vandal and Byzantine conquests that had come before
it. As elsewhere in the early Islamic Mediterranean, a new coinage initially imitated Byzantine models, though potentially offensive images
such as the cross were removed.2 As in the eastern Mediterranean, Africa
1

Gregory VII, Epistula 3.21, ed. E. Caspar, 2 vols., MGH Epist. Select. 2 (Berlin, 19203),
1:287: Nobilitas tua hoc in anno litteras suas nobis misit, quatenus Servandum presbyterum
episcopum secundum christianam constitutionem ordinaremus. Quod, quia petitio tua iusta et
optima videbatur, facere studuimus. To be read in conjunction with Gregory VII, Epp. 3.1920,
1:2857.
On early Islamic coinage in the Umayyad Maghrib, see J. Walker, A Catalogue of the Muhammadan
Coins in the British Museum, vol. 2, A Catalogue of the Arab-Byzantine and Post-Reform Umaiyad
Coins (London, 1956), pp. 5473, 76, and 789; M. Bates, History, Geography and Numismatics
in the First Century of Islamic Coinage, Revue suisse de numismatique 65 (1986), pp. 23162;
M. Bates, Roman and Early Muslim Coinage in North Africa, in North Africa from Antiquity

362

Aftermath
was eventually reorganized into a new province with the same name:
Ifrqiya. As with previous conquests, the initial Arab presence seems to
have been fairly small, and more or less restricted to the cantonment of
Kairouan and its immediate environs. Only slowly were the two great
languages of elite Roman cultural identity, Latin and Greek, replaced
by Arabic. For hundreds of years after the conquest, Afariqa continued
to speak a dialect of Latin that al-Idrs (ad 1100c.1165) called al-latn
al-afrq or African Latin. This dialect was spoken in Gafsa into the
twelfth century at least.3 As an epigraphic language it is attested both in
Kairouan and in the region of Tripoli down to the eleventh century.4
The emergence of a new identity that looked to the Islamic East rather
than to Italy or Constantinople was just as gradual in North Africa as it
was in the rest of the Islamic world.
Perhaps more significantly, Christianity in Africa remained strong on
the eve of the Arab conquest and for centuries into the Islamic period.
Richard W. Bulliet, in what remains, to the best of my knowledge,
the only study that attempts to assess the rate of conversion to Islam
in the medieval period, concludes that in Tunisia this process probably
took place at much the same rate as in the rest of the former Byzantine
world.5 Muslims, he suggests, did not become a majority in Tunisia until

3
4

to Islam: Papers of a Conference Held at Bristol, October 1994, ed. M. Horton and T. Wiedemann,
Centre for Mediterranean Studies (University of Bristol), Occasional Paper 13 (Bristol, 1995),
pp. 1215; A. Balaguer, Early Islamic Transitional Gold Issues of North Africa and Spain in the
American Numismatic Society, American Numismatic Society Museum Notes 24 (1979), pp. 225
41; E. Leuthold, Jr., Due rare testimonianze della prima monetazione musulmane a Cartagine,
Rivista italiana di numismatica e scienze affini 69 (1967), pp. 939. My thanks to Michael L. Bates
for these references.
Al-Idrs, Nuzhat al-mushtaq f ikhtiraq al-afaq, ed. and trans. R. Dozy and M. J. de Goeje (Leiden,
1866), pp. 1045 (p. 122 in the French translation), of Gafsa.
C. Saumagne and L. Poinssot, in BCTH (19289), pp. 3701; W. Seston, Sur les derniers temps

du christianisme en Afrique, Melanges dArcheologie et dHistoire de lEcole

francaise de Rome (1936),
pp. 10124; C. Courtois, Gregoire VII et lAfrique du Nord: Remarques sur les communautes
chretiennes dAfrique au XIe si`ecle, Revue historique (1945), pp. 97122 and 193226; J. B.
Ward-Perkins and R. G. Goodchild, The Christian Antiquities of Tripolitania, Archaeologia:
Miscellaneous Tracts Relating to Antiquity 95 (1953), pp. 212; AE (1965), 478, no. 147 (Kairouan);
A. Mahjoubi, Nouveau temoignage e pigraphique sur la communaute chretienne de Kairouan
au XIe si`ecle, Africa 1 (1966), pp. 8796; A. di Vita, La diffusione del Cristianesimo nellinterno
della Tripolitania attraverso i monumenti e sue sopravvivenze nella Tripolitania araba, Quaderni di
Archeologia della Libia 5 (1967), pp. 1349; G. Gualandi, La presenza cristiana nellIfriqiya. Larea
cimiteriale dEn-Ngila (Tripoli), Felix Ravenna 1056 (1973), pp. 25779; R. Bartoccini and
D. Mazzoleni, Le inscrizioni del cimitero di En Ngila, Rivista di archeologia cristiana 53 (1977),
pp. 15798. See also J.-M. Lass`ere, Diffusion et persistance des traditions latines dans le Maghreb
medieval, in La latinite, hier, aujourdhui, demain (Avignon, 1981), pp. 27785, esp. 2812.
R. W. Bulliet, Conversion to Islam in the Medieval Period: An Essay in Quantitative History (Cambridge,
Mass., 1979), esp. pp. 769 and 92103. Bulliet made a preliminary sketch for his approach nearly a
decade earlier: A Quantitative Approach to Medieval Muslim Biographical Dictionaries, Journal
of the Economic and Social History of the Orient 13 (1970), pp. 195211. The work has been extensively

363

Staying Roman
quite late in the ninth century, and did not become a vast majority until
some time in the tenth. Bulliet bases this argument on the incidence
in Ifriqiya of individuals bearing the five distinctively Muslim names of
Muhammad, Ahmad, c Al, al-Hassan, and al-Hussain.6 There are clear
methodological objections to the kind of statistical modelling Bulliet
employs, especially in the light of what he would be the first to concede
is a very small data set indeed,7 but his conclusions remain challenging
nevertheless.
Moreover, Bulliets arguments seem to receive some support from
other evidence that also indicates the survival of a Christian population in
North Africa down into the twelfth century and beyond. This Christian
community becomes difficult to see in the years following the Arab conquest but not impossible.8 Elizabeth Savage has gathered evidence for
over thirty Christian communities in North Africa ranging in date from
the eighth to the fourteenth centuries, and possibly later. These included
the bishoprics of Carthage, Gumi, Sabratha, Hadrumetum (Sousse), and
Tripoli, the monastic community at Monastir, and Christian populations
in Kairouan, Gab`es, and Fez.9 A prosperous Christian community also
survived in ninth-century Tiaret (Tahart), the capital of Rustamid Algeria, where Christians were involved in trade and numbered among the

7
8

reviewed and variously received by the scholarly community: see the reviews of U. Abd-Allah
in Journal of Near Eastern Studies 44 (1985), pp. 23941; J. Calmard in Studia Iranica 12 (1983),
pp. 1334; R. Cleveland and K. B. Leyton-Brown in International History Review 4 (1982), pp.
45961; T. Glick in Journal of Historical Geography 7 (1981), pp. 2245; H. Kennedy in International
Journal of Middle East Studies 13 (1981), pp. 2502; I. Lapidus in American Historical Review 86
(1981), pp. 1878; D. Little in The Middle East Journal (1980), p. 369; H. T. Norris in Bulletin of
the School of Oriental and African Studies, University of London 44 (1981), pp. 1623; J. Voll in Journal
of Interdisciplinary History 11 (1981), pp. 5223; and J. Waltz in Speculum 56 (1981), pp. 3602. N.
Levtzion (ed.), Conversion to Islam (New York, 1979) contains no essays on conversion in early
medieval North Africa. Many thanks to Deborah G. Tor and David Cook for their own insights
into the debate surrounding Bulliets thesis.
Bulliet, Conversion to Islam, pp. 6472, with the (much richer) comparative evidence of Iran. For
the incidence of the five Muslim names in Tunisia prior to AH 300 (ad 913), Bulliet draws upon
the biographies of noted Maghribi members of the Maliki law school compiled by Ibn Farhun,
Kitab al-dbaj al-mudhahhab f mac rifa ac yan c ulama al-madhhab (Cairo, 1932); thereafter, on the
biographical dictionary of Abu al-Falah c Abd al-Hayy Ibn al-c Imad, Shadharat al-dhahab f akhbar
man dhahab, 8 vols. (Cairo, 19312).
Bulliet, Conversion to Islam, pp. 64 and 76.
On the survival of a Christian community in Islamic North Africa, see also M. Handley, Disputing
the End of African Christianity, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New
Perspectives on Late Antique North Africa (Aldershot, 2004), pp. 291310.
E. Savage, A Gateway to Hell, a Gateway to Paradise: The North African Response to the Arab Conquest,
Studies in Late Antiquity and Early Islam 7 (Princeton, NJ, 1997), pp. 10710. See also R. M.
Speight, Temoignage des sources musulmanes sur la presence chretienne au Maghreb de 26/747
a` 184/800, Institut des belles lettres arabes 129 (1972), pp. 7396 and Speight, The Place of the
Christians in Ninth Century North Africa, According to Muslim Sources, Islamochristiana 4
(1978), pp. 4765.

364

Aftermath
citys notables.10 The Sicilian saint Elias the Younger was twice captured
by Christian African slave merchants when he was a boy, the second
occasion leading to perhaps forty years in captivity as the slave of a
Christian African tanner.11 More spectacular are the handful of Latin
tombstones that attest a continued Christian population in Kairouan and
at En Ngila, near Tripoli, as late as the eleventh century.12 But the Christians of Islamic North Africa are also mentioned at least sporadically in
the works of medieval Europeans, for the Christian communities of the
Mediterranean world remained in contact for centuries after the Islamic
conquest.
Connections were certainly maintained with the East. Some time
between 758 and 767, a ship arrived in Rome bearing a representative of
the Patriarch of Alexandria who had travelled to Italy by way of Africa.13
Slightly earlier, the Palestinian monk George had found that the African
Church was being severely beaten by the attacks of tyrants when he
travelled there from St Sabas monastery in Jerusalem on the orders of
his abbot. George had been sent to Africa on account of the monks
stipend (ob stipendium monachorum), a richly suggestive phrase that raises
more questions than it answers. It is clear that St Sabas monastery enjoyed
or hoped for some kind of financial support from the Christian community of Ifriqiya, but the precise nature of the subsidy is no longer
clear to us. Given the dire situation in Africa, George continued on to
Spain.14 It is difficult to believe that he found conditions there to be
much more congenial. The monk found himself caught up in the Cordoban martyrs movement, and in July 852 he was executed along with
a number of Spanish Christians by the local Muslim authorities.15 The
Life of Elias the Younger gives further evidence of connections between
Africa and the Holy Land in the ninth century, for having purchased
his freedom and outworn his welcome in North Africa Elias travelled
to the East on a pilgrimage.16 Travelling in the opposite direction, a
group of Frankish penitents returned to Rome from Jerusalem by way of
Egypt and Africa. In Carthage they visited the sepulchre of St Cyprian,
10
11

12
13
14
15
16

Savage, Gateway to Hell, p. 101.

Vita s. Eliae iunioris 417, ed. G. Rossi Taibbi, in Vita di SantElia il Giovane, Istituto Siciliano di
Studi Bizantini e Neoellenici Testi e Monumenti 7 (Palermo, 1962), pp. 626; M. McCormick,
Origins of the European Economy: Communications and Commerce, ad 300900 (Cambridge, 2001),
pp. 2467.
See above, n. 4.
Paul I to Pippin (= Codex Carolinus 40), ed. Gundlach in MGH Epist. 3:553.
Eulogius of Toledo, Memorialae sanctorum 2.10.23, PL 115, cols. 7867: vapulare asperius Dei
Ecclesiam incursatione tyrannorum.
Eulogius of Toledo, Memorialae sanctorum 2.10, cols. 77792.
Vita s. Eliae iunioris 14 and 17, pp. 202 and 246.

365

Staying Roman
where many great works and many miracles are frequently revealed by
the Lord.17 Links between Africa and the East are further hinted at in a
collection of Latin liturgical manuscripts a Psalter, an epistolary, and an
antiphonary that are now housed at St Catherines monastery in Sinai.
All three manuscripts appear to have been produced in a single, eastern
scriptorium, perhaps in the tenth century, and their script shows signs
of contemporary Greek, Arabic, and Syriac influence; yet the epistolary
used the African text of the Old Latin Bible and a liturgical calendar included with the Psalter is indicative of devotion to a distinctively
African constellation of saints. These manuscripts would seem, then, to
have been produced for a community of African Christians resident in
the East, perhaps even in the Holy Land itself.18
But it was the old connection with Italy and Sicily and now, through
Italy, the transalpine world of the Frankish empire that seems to have
remained the strongest over the course of the Middle Ages. An epitaph
of Pope Hadrian I (ad 77295) found in a ninth-century codex from
Regensburg indicates that, nearly a century after the fall of Carthage
to the Arabs, Africa, held captive for so many years, rejoices to have
merited bishops by your prayers.19 Einhard tells us that Charlemagne
entered into diplomatic relations with the rulers of Syria, Egypt, and
Africa, so as to care for the poverty-stricken Christians of Jerusalem,
Alexandria, and Carthage, respectively.20 Notker the Stammerer remembered Charlemagnes gifts of wealth, grain, wine, and oil to the Africans
specifically. Indeed, African poverty was (enigmatically) proverbial by
Notkers day.21 As we have seen, in 801, Charlemagne was also said to
17

18

19
20
21

Gesta sanctorum Rotonensium 3.8, ed. C. Brett, in The Monks of Redon (Woodbridge, 1989), pp.
2079: direxerunt gressum ad Africam uisitare sepulchrum sancti Cypriani archiepiscopi et
martyris Christi, qui secundo miliario ab urbe Carthaginensi requiescit iuxta mare, ubi multae
uirtutes et multa miracula a Domino saepius ostenduntur.
E. A. Lowe, An Unknown Latin Psalter on Mount Sinai, Scriptorium 9 (1955), pp. 17799; E.
A. Lowe, Two New Latin Liturgical Fragments on Mount Sinai, Revue Benedictine 74 (1964),
pp. 25283; E. A. Lowe, Two Other Unknown Latin Liturgical Fragments on Mount Sinai,
Scriptorium 19 (1965), pp. 329; J. Gribomont, Le Mysterieux Calendrier latin du Sina: e dition
et commentaire, Analecta Bollandiana 75 (1957), pp. 10534; and B. Fischer, Zur Liturgie der
lateinischen Handschriften vom Sinai, Revue Benedictine 74 (1964), pp. 28497.
Silverius, Epitaphium Adriani I Papae, ed. E. Dummler, MGH Poet. 1 (Berlin, 1881), p. 114:
Africa laetatur, multos captiva per annos, / Pontifices precibus promeruisse tuis.
Einhard, Vita Karoli magni 27, ed. G. H. Pertz and G. Waitz, MGH SRG (5th edn; Hanover,
1905), pp. 278.
Notker Balbulus, Gesta Karoli magni imperatoris 1.26 and 2.9, ed. H. F. Haefele, MGH SRG n.s. 12
(Berlin, 1962), pp. 37 and 623; see also ibid., 2.14, p. 77, which implies that African merchants
were a regular sight in ninth-century Gallia Narbonensis. On proverbs, see H. Walther, Lateinische Sprichworter und Sentenzen des Mittelalters in alphabetischen Anordnung, Carmina Medii Aevi
Posterioris Latina 2, 9 vols. (Gottingen, 196386); A. Otto, Die Sprichworter und sprichwortlichen
Redensarten der Romer (Leipzig, 1890; repr. Hildesheim, 1988); and S. Singer, et al., Thesaurus

366

Aftermath
have secured the translation of Cyprians relics, as well as those of St
Pantaleon and the Scillitan martyrs, from Africa to the Frankish empire
(see above, Chapter 2.4.2).22 Perhaps significantly, in the eighth and ninth
centuries, two different military commanders from Byzantine Sicily were
said to have fled the displeasure of Constantinople and sought refuge in
Africa, where each also had himself proclaimed emperor.23 In the late
ninth century, two Sicilian boys named Antoninus and Peter were captured in Arab raids and carried off to Africa, where they were raised as
Muslims and eventually appointed to official posts at the Aghlabid court.
However, they secretly converted to Christianity, were executed for their
beliefs, and came to be venerated as martyrs in the Byzantine church.24
Earlier, in 813, disaster befell a Muslim fleet sent from North Africa to
raid Sardinia and 100 ships were said to have been sucked into some
sort of a whirlpool or watery chasm (vorago). News of the catastrophe
reached the governor of Byzantine Sicily, Gregory, by way of a letter
written by an African Christian to a highly placed friend in the imperial
administration of the province either Gregory himself or his legate
Theopistus notarius. Pope Leo IIIs own legate passed the news on to
Rome, whence it travelled to the aging Charlemagne.25 By the end of
the ninth century, the African church was split by a schism once again,
and, during the pontificate of Formosus (ad 8916), legates from Africa
travelled to Rome in an attempt to resolve the dispute.26
Rome remained in contact with the African church over the course
of the following centuries, and, indeed, the Pope came to invest the
metropolitans of the province. In the tenth century, the clergy and people
of Carthage were said to have elected a bishop named James (or Jacob)
and to have sent him to Rome to be consecrated by Pope Benedict VII
(ad 97584). Benedict examined his orthodoxy and having found his

22

23
24
25
26

proverbiorum medii aevi: Lexikon der Sprichworter des romanischgermanischen Mittelalters, 13 vols.
(Berlin, 19956).
On these translations, see, in addition to the material cited above, Chapter 2.4.2 n. 195,
P. Franchi de Cavalieri, Le Reliquie dei martiri Scillitani, Romische Quartalschrift fur Christliche
Altertumskunde und fur Kirchengeschichte 17 (1903), pp. 20921; H. Quentin, Les Martyrologes
historiques du moyen a ge: etude sur la formation du Martyrologe romain (2nd edn; Paris, 1908),
pp. 50714; and F. Ruggiero, Atti dei martiri Scilitani: Introduzione, testo, traduzione, testimonianze
e commento, Atti della Accademia Nazionale dei Lincei, Memorie, 9th ser., 1/2 (Rome, 1991),
p. 53.
Theophanes, Chronographia AM 6274, 1:4556 and Theophanes Continuatus, Chronographia 2.27,
ed. I. Bekker (Bonn, 1838), pp. 813.
Synaxarium Constantinopolitanae s.d. Sept. 23, cols. 724.
Leo III, Epistula 7, ed. K. Hampe, MGH Epist. 5 (Berlin, 1899), pp. 979.
Flodoard of Rheims, Historia Remensis ecclesiae 4.2, ed. M. Stratmann, MGH Scriptores 36
(Hanover, 1998), p. 372.

367

Staying Roman
faith satisfactory ordained him and sent him back to Africa.27 Papal
concerns about African orthodoxy were not new in the tenth century:
from the late fifth century down to the mid eleventh, the anxieties of
the bishops of Rome in this regard were sufficiently strong for them to
warn against even accepting Africans into ecclesiastical orders.28 By the
mid eleventh century the Christian community in Africa was certainly
in desperate straits, though not for theological reasons. Towards the end
of his pontificate, Leo IX (ad 104954) wrote to the African bishops
Thomas, Peter, and John. In his letter, Leo comments on the fact that
205 bishops had once attended a council in Carthage, and that now
there were scarcely five bishops in all of Africa.29 If the Pope was not
exaggerating, then we know the names of almost the entire African
episcopate in 1053: Leos letters to Thomas, Peter, and John concerned
their unruly fellow bishop, Gummitanus. Unfortunately, for the most part
we do not know where their sees were located. Twenty years later, there
was a Christian community not only at Carthage (where in 1053 Thomas
was probably metropolitan) but at Hippo Regius as well, and it seems
reasonable to suppose that one of the three other bishops mentioned in
Leos correspondence had his see there in the mid eleventh century. Even
at this late date the African church seems to have preserved its peculiar
tradition of recognizing provincial primacy according to seniority. With
only five bishops this would almost seem a hollow honour, but the
number of ecclesiastical provinces in Africa had diminished in the
1070s, Hippo Regius was in the province of Mauretania Sitifensis, not
Numidia as in late antiquity and the honour was clearly not hollow in
the eyes of the African bishops themselves. Indeed, Leo was at pains to
stress that the bishop of Carthage presided over the entire African church,
second only to the Pope, from whom the Carthaginian metropolitan
alone received the pallium.30
27
28

29

30

Leo abbas et legatus, Epistola ad Hugonem et Rotbertum reges, ed. G. H. Pertz, MGH Scriptores 3
(Hanover, 1839), p. 689.
See Gelasius Is formula on accepting Africans into ecclesiastical orders: Constituta quae episcopi
in sua ordinatione accipiunt, PL 59, col. 137d; reiterated in Greg. Ep. 2.31, 1:118; Gregory II to
Boniface (= Boniface, Ep. 18), ed. E. Dummler in MGH Epist. 3:267; Liber diurnus Romanorum
pontificum 6, ed. T. E. von Sickel (Vienna, 1889), p. 6; Nicholas II, Ep. 25, PL 143, col. 1347a;
and see the discussion of Markus, Donatism: The Last Phase, pp. 1245 and R. A. Markus,
Country Bishops in Byzantine Africa, in D. Baker (ed.), The Church in Town and Countryside,
Studies in Church History 16 (Oxford, 1979), pp. 56.
Leo IX, Epistulae 834, PL 143, cols. 72831. Another eleventh-century source, Ralph Glaber,
is too loose with his use of the term Africa to tell whether the Christian individuals and
communities he places there existed in North Africa itself, in Muslim Spain, or, indeed, even
in Italy: Historiarum libri quinque, ed. and trans. J. France (Oxford, 1989), pp. 1252, esp. ibid.,
4.7.22 and 5.1.13, pp. 206 and 234; but see also ibid., p. 74 n. 2.
Leo IX, Ep. 84, cols. 72931.

368

Aftermath
Twenty years later, Gregory VII (ad 107385) was in contact with
another Bishop of Carthage, Cyriacus, as well as with the clergy and
people of Hippo.31 Gregorys letters reveal a deepening crisis within the
African church, which was both divided against itself and experiencing
persecution from the Muslim majority.32 By 1076, there were only two
bishops left in Africa. The church was no longer able to sustain itself,
for canon law required that a new bishop be ordained by three sitting
colleagues.33 As we have seen, Gregory intervened, consecrating a new
bishop for Hippo Regius.34 But the writing was already on the wall. The
church in Africa was dying.
The timing of these appearances of the African church in our sources
for medieval Europe may be significant. There are two clear peaks to these
references, one in the ninth century, the other in the eleventh. If Bulliet
is correct in his projection of the curve of conversion, the first of these
peaks corresponds to the initial large-scale acceptance of Islam in North
Africa and presumably the earliest awareness of a serious Islamic threat
to African Christianity. The Life of Elias the Younger, probably written
in the 930s, makes the point fairly explicit: from a Byzantine point of
view, already in the ninth century North African Christianity was no
longer firmly established (82).35 A similar attitude is visible in
the earlier poetry of the Frankish deacon Florus of Lyons celebrating the
translation of Cyprians relics to Charlemagnes empire.36 The second
peak corresponds to the laggard phase of conversion: that period in
which the vast majority of the North African population had already
converted to Islam, and the Christian community had become a rapidly
dwindling minority. In this interpretation, medieval European interest in
contemporary Africa peaked at crisis moments in the life of the African
church. This may in part help explain Notkers puzzling statements about
African poverty; and, as we have seen, this interpretation seems to be
borne out in the correspondence of Pope Gregory VII.
How long the African church survived after Gregorys pontificate is
anybodys guess. Virginie Prevost has recently argued that the last North
African Christian communities, concentrated in the Jard and Nafzawa
oases of southern Tunisia, are unlikely to have survived past the mid
thirteenth century.37 Other scholars have argued that a native Christian
31
32
34
35
36
37

Gregory VII, Epp. 1.23 (ad 1073) and 3.1921 (ad 1076), 1:3940 and 2858.
33 Gregory VII, Ep. 3.19, 1:285.
Gregory VII, Ep. 1.23, 1:3940.
Gregory VII, Epp. 3.201, 1:2868.
Vita S. Eliae iunioris 4, p. 6. See also Speight, Place of the Christians, pp. 635.
Florus, Carmen 13 (Rector magnificus), pp. 5445.
V. Prevost, Les Derni`eres Communautes chretiennes autochtones dAfrique du Nord, Revue de
lhistoire des religions 224 (2007), pp. 46183.

369

Staying Roman
population may have survived in Africa into the fourteenth century.38
Certainly, as late as the 1350s, Boccaccio still thought of Gafsa as a
city with a significant community of Christians. In one of the more
memorable stories from the Decameron, the young, innocent Muslim girl
Alibech was inspired to travel to the wilderness of the Thebaid where
the unscrupulous monk Rustico taught her to put the devil back into
Hell after hearing the Christians of Gafsa praise the desert ascetics. But
Boccaccios story was set once upon a time (gi`a), in the sexual fantasy
world of a fourteenth-century Florentine, and was never intended to
reflect the authors contemporary reality.39
Ultimately, it is true, the North African church was unable to maintain
itself under Arab rule.40 But Christianity survived in Carthage for at least
400 years after the Islamic conquest, and in some of the more remote
corners of North Africa perhaps longer. The question of why Christianity
died out there some time between the twelfth and nineteenth centuries
surely has its answer in the history of the Islamic period not in late
antiquity.
38

39

40

Savage, Gateway to Hell, pp. 1079 and T. Lewicki, The Ibadites in North Africa and the
Sudan to the Fourteenth Century, Journal of Western History/Cahiers dhistoire mondiale 13 (1971),
pp. 83130.
Giovanni Boccaccio, Decameron 3.10, ed. V. Branca (Turin, 1980), pp. 44350; see also, however,
A. Delattre, Le Culte de la Sainte Vierge en Afrique dapr`es les monuments archeologiques (Paris, 1907),
pp. 15563, a fourteenth-century medallion of the Virgin, designated Mother of God (\[2]
J[]#), with the invocation in Arabic Mary, protect your servant (
). On the
late medieval cult of Mary and contemporary Christian-Muslim relations, see A. Remensnyder,
Christian Captives, Muslim Maidens, and Mary, Speculum 82 (2007), pp. 64277.
See also Cameron, Byzantine Africa, pp. 4950.

370

CONCLUSIONS

Empires are defined by their peripheries. Control of the African periphery was particularly important to late Roman emperors, with the result
that even after the Vandal conquest ended effective imperial control of the
region, a succession of emperors nevertheless sought to sustain Roman
hegemony there. The techniques that they employed in pursuit of this
endeavour changed over time: the diplomacy and political marriage of
the fifth century gave way in the sixth to Justinians forcible reintegration
of Africa into the political and military structures of the empire. Thereafter the reconquered province was fortified with defensive works that
projected imperial indomitability, and was administered by close, highly
trusted, and experienced associates of the emperor. All the way down to
the final Islamic conquest, though, the imperial administration struggled
to maintain Africas integration into the larger Roman world.
This integration was not purely political. To be sure, the sixth- and
seventh-century emperors extracted revenue from the province in the
form of taxation; but throughout the period from the fifth to the early
eighth centuries African goods also continued to reach markets across the
Mediterranean, even if in progressively dwindling quantities. The flow
of goods accompanied a flow of individuals, including not only merchants, but also bishops, aristocrats, soldiers, office-holders, diplomats,
teachers, poets, monks, and deacons. In the course of their journeys to
attend church councils, to flee warfare or persecution, to discharge official duties, to seek the aid of miracle-working saints, to further personal
ambitions, or simply to find a better life these travellers crossed back
and forth between Africa on the one hand and Rome, Constantinople,
Spain, Gaul, Sardinia, Sicily, southern Italy, the Balkans, Asia Minor,
the Levant, and Egypt on the other. Regardless of the directions of
their movement, they brought with them devotion to the holy men
and women of their native provinces and injected new elements into
established name-stocks; some also bore letters or books for emperors,
intellectuals, administrators, and military officers interested in developments elsewhere in the Mediterranean. By the admittedly diminishing
371

Staying Roman
standards of late antiquity, Africa remained well connected to the outside
world.
One important consequence of the continued interconnectedness of
Africa and the Mediterranean was the need of all of those who struggled for political domination in the region to legitimate their rule to
multiple audiences at the same time: the imperial administration, other
powers in the Mediterranean, and African populations which expected
authority to be expressed and exercised in Roman terms. The case was
perhaps the easiest for Justinian to make. He was, after all, the Roman
emperor, and Africa had once been a Roman province. The right that
Justinian asserted to intervene in African affairs was disputed at the time
by the Vandal king Gelimer and may not have been universally accepted
even within the empire; but success reshapes opinions. After the fact,
Justinians detractors continued to question his motives, doubt the wisdom of his actions, and reject his methods; but they nevertheless seem
to have accepted his reconquest of Africa as a fundamentally legitimate
restoration of Roman power. Similarly, the Vandal kings and the Moorish leaders of the African interior emphasized the continuities between
their rule and what came before. As with the Byzantine reconquest, the
establishment of autonomous Vandal and Moorish kingdoms in Africa
in fact represented a profound discontinuity with the imperial past; but
through the maintenance of Roman offices and institutions, Africas new
rulers sought to underscore the ongoing stability of the political and
legal conventions that had structured the lives of Africans in the late
Roman period. But Romano-Africans were not the sole audience of
these Moorish and Vandalic efforts at political legitimation. Centuries of
unquestioned Roman hegemony in the West had left its mark on nonRoman populations as well. Vandals and Afri barbari, too, expected to be
addressed in the Roman vocabulary of power. The new rulers appeals
to what we perceive as Roman political and cultural forms sought to
impress these kings regal dignity and grandeur on their entire populations, Roman and barbarian alike. Much the same was true on the
diplomatic level of exchanges with neighbouring and transmarine barbarian kingdoms. The Roman language of power was the only such
language spoken in the late antique Mediterranean.
Internally these efforts at legitimation appear to have been remarkably
successful. It was at most a matter of decades before Romano-Africans
began dating inscriptions and manuscripts according to the Vandal regnal
year, and, indeed, within the first generation of the conquest at least
some members of the old ruling class were already working in the Vandal
administration. The evidence is thinner in the Moorish kingdoms, but
the same basic trends seem to have held there as well. Romano-Africans
372

Conclusions
were also more or less immediately recruited into the Byzantine administration and army one or two to very prominent positions at the imperial
court though a string of eastern emperors simultaneously elicited loud
choruses of protest from the African episcopate for what were perceived
as presumptuous and heretical ecclesiastical policies. Popes, too, met with
censure from the peppery African bishops, both for ecclesiastical prevarication and for overstepping their authority by meddling in African affairs.
This was, however, a far cry from systemic resistance to, or rejection of,
either the empire or the church as such.
Indisputably, though, there were some Romano-Africans who rejected
the claims of Vandal kings and Byzantine emperors to political legitimacy.
Victor of Vita and Dracontius are the most visible cases in the Vandal
century; in his older, more embittered phase Facundus of Hermiane
is the best-known example under the Byzantines. But the same was true
in territories long established as part of the empire, where Procopius,
for example, was to accuse Justinian of being a demon incarnate.1 Disaffection is a common thread running through most political systems
perhaps through all of them. As a rule, however, Africans seem to have
internalized the legitimacy of the Vandal and Byzantine political structures, and to have expressed whatever disaffection they felt accordingly.
For Vandal and Moorish rulers, at least, the greater challenge lay in
reassuring the centres of imperial power of their willingness to work
with the empire while at the same time maintaining practical autonomy.
In their different ways, both Vandals and Moors sought to do precisely
that; but in neither case did their assurances equate to an unchallenged
acceptance of imperial terms and conditions. Indeed, in its initial phases,
working out the shape that such political accommodation would take
was frequently both violent and rhetorically polemical. Thus, the mid
fifth century saw frequent Vandal raids on imperial territory which sporadically erupted into full-scale war with the empire, a pattern that was
to repeat itself in Moorishimperial relations in the mid sixth century.
Moreover, just as Africas new rulers sought to legitimate their power
through an appeal to continuity of Roman political form, authors writing
outside the territory of these new kingdoms deployed venerable GrecoRoman stereotypes of the barbarian against them. Regional differences
in material culture, religious belief, and social practice characteristic of
the Mediterranean in all periods, regardless of its degree of interconnectedness had long been understood in ethnographic terms in the ancient
world, and in the fifth, sixth, and seventh centuries such differences could
be politicized in an effort to delegitimize competing claims to authority
1

Proc. Anecd. 12.1427, 18.1, 18.367, and 30.34, 3:7982, 11112, 118, and 186.

373

Staying Roman
in times of conflict. In other words, in the changed and charged
circumstances of the fifth and later centuries, whether one was Roman
or not was very much a question of perspective.
Indeed, the fifth century saw a fracturing of Roman identity. The
political fragmentation of the Mediterranean world combined with its
sustained economic, social, and cultural interconnectedness to force
Africans who were no longer subject to imperial authority but who
continued to identify themselves with the empire and its history, creed,
and institutions to reconsider the terms in which their Romanness could
be defined. Politics, high culture, and religion became the axes of Roman
identity in fifth-century North Africa, and remained so through to at
least the beginnings of the eighth century.
To Procopius and other authors writing within imperial territory,
Romans were first and foremost subjects of the emperor. Yet in the Vandal period it was precisely this political definition of collective identity
that had so manifestly failed Africans who wanted to stay Roman. In
any case, even among authors like Procopius other issues were also at
stake. Loyalty to the emperor was one, though of course among the
emperors subjects there were rebellious Romans and, say, loyal Armenians. Perceived cultural difference was also a factor, one which is strikingly
illustrated by Procopius and Corippus accounts of the Moors, including
those Moors who accepted the authority of and remained loyal to the
emperor. From a twenty-first century perspective, though, it is hard to
tell how firmly grounded perceptions of this sort were in lived experience. Indeed, in the Vandal kingdom it would seem that over time
distinctions along cultural lines became increasingly difficult to draw. At
least the Roman aristocratic lifestyle probably did not successfully define
Romanness there for very long. By most accounts the Vandals took very
quickly to the good life, dressing in gold and silks, adorning their suburban villas with splendid works of art, educating themselves or their
children alongside Romano-African youths, and even patronizing Latin
poetry. Elite Romano-Africans meanwhile stitched together a new reality for themselves pieced from inherited fragments of antiquity, took an
interest in the remote or mythological past, and lamented the jabbering
toasts and calls for food and drink that passed for la grande vie among
Germanic barbarians. Yet by the late fifth century many members of the
old ruling class perhaps even most seem to have been willing to reach
a measure of accommodation with the Vandals. Wealth, power, a sincere
enthusiasm for classical learning, and the polish of aristocratic culture
were the entree into the world of the Romano-African nobility. These
the Vandal elite had acquired with some zeal. The aristocratic gripe
was as much against the lower orders rustics and provincials, slaves,
374

Conclusions
labourers, and, indeed, the vast majority of humanity as it was against
the barbarians. Then, too, the accession of Hilderic, the son of the Vandal
prince Huneric and the imperial princess Eudocia, probably helped to
resolve still further many Romano-African ambiguities about the Vandal
ruling class: when Luxorius celebrated Hilderic as the heir of a twin
crown, it was the kings Roman not his Vandal ancestors who drew
the poets praise.
The Christian religion also provided its adherents with a further definition of Romanness. At least, as they were separated from the empire
politically under the Vandal regime, African churchmen of Nicene confession sought to impress on their audiences both at home and abroad
that what united them to other Romans throughout the Mediterranean
was their adherence to Catholic Christianity. Their position developed
naturally from Augustines thought; for, even as the great theologian
conceded that Romanness had become the predominant political identity throughout his world, he argued forcefully that it had simultaneously
ceased to matter. What distinguished human societies in Augustines mind
was the belief or lack thereof in the truth of Christian revelation.2 The
fifth- and sixth-century emperors in Ravenna and Constantinople seem
by and large to have accepted this argument. They saw themselves as protectors of the Nicene community of the Vandal kingdom: Valentinian III
and Zeno both intervened diplomatically in African affairs to secure
the ordination of a new metropolitan bishop for Carthage, and Justinian
reconquered the region in part to conserve and reinforce its orthodoxy.
Thereafter debates about the nature of the faith were not cast in terms
of Roman and non-Roman, but rather in terms of the degree to
which confessional refinements either preserved or did violence to the
decisions of Nicaea and Chalcedon. In the world of late antique religious polemic, Vandals and their faith could plausibly be discredited as
barbarian; Byzantine emperors and their theological innovations no
matter how heretical they might seem could not. Yet we should be
wary of accepting the Nicene characterization of Arianism as a distinctively barbarian confession. The few hints that survive suggest that by the
late fifth century, if not before, Arians in the Vandal kingdom saw their
faith as a universal one too, and thus sought both the conversion of the
entire local populace (at all levels of society) and the protection of Arian
communities abroad.
Theology was naturally a matter of some importance to Christian
clergy, but it is difficult to tell how far such questions mattered to
2

Augustine, Enarrationes in Psalmos 58.1.21, 39:744, and Augustine, De civitate Dei 14.1, 48:414,
cited above, Introduction and Chapter 3.4, respectively.

375

Staying Roman
laymen at this time. Junillus Africanus, a lawyer and advocate of imperial
theocracy, seems to have been deeply interested in them. Yet Corippus
never mentions the Three Chapters controversy, though he was writing
at almost exactly the moment that Facundus went into hiding and that
Victor of Tonnena underwent imprisonment and exile. The same was
true of Fulgentius the Mythographer, who wrote somewhat later. The
Vandal-era secular poets Luxorius, Florentius, and Felix likewise never
mention Arianism or the persecution of Nicene Christians. Perhaps it
would have been impolitic for any of these men to have done so, as
most of them (in both the Vandal and the Byzantine periods) enjoyed
or sought to gain a position at the relevant court. Fulgentius of Ruspes
biographer tells us that in the late 490s the saint, then a young monk,
decided to forego a trip to Egypt because the patriarchates of the East
were divided from his own in the West by the Acacian schism; but such
considerations seem not to have bothered the grammaticus Priscian, or
for that matter the subdeacon Reparatus and his fellow confessors from
Tipasa, all of whom travelled to Constantinople in the throes of the same
schism. On the other hand, Fulgentius correspondence concerning theological questions indicates a clear interest in such matters on the part of
the educated laity, as does pseudo-Origens Commentary on Job from the
Arian side. Victor of Vitas history claims that a handful of prominent
laymen were willing to undergo exile and even execution rather than
apostatize the Nicene confession. As A. H. M. Jones observed long ago,
religion probably enjoyed the place in late antique and early medieval
society that political ideology enjoyed in the modern world during the
Cold War.3 Those who chose not to pursue it as a vocation will have
found theology important to greater and lesser degrees, but few will have
remained untouched by its influence.
To some extent, however, this division between the ecclesiastical and
secular elites is an artificial one. Both were drawn from the same social
world. They came from the same families, and indeed in some instances
they were doubtless the same individuals. This is demonstrably the case
only with Fulgentius of Ruspe, who began his career as a tax-collector
and ended it as a bishop, but there must have been others like him.
Thrasamunds poet Felix sought to become one: he petitioned the Vandal
kings primiscrinarius for an ecclesiastical office. In any case, secular and
ecclesiastical writers seem to have shared a common vision of the world
and the peoples encroaching on the empire from beyond its frontiers.
Victor of Vita loathed Vandals because they were Arians, but he also hated
3

A. H. M. Jones, Were the Ancient Heresies National or Social Movements in Disguise?, JThS
n.s. 10 (1959), p. 295.

376

Conclusions
and feared them because they were non-Romans; and he had no qualms
giving voice to his vitriolic anti-barbarian prejudices in some of the
most passionate passages to emerge from his poison pen. Ecclesiastical
writers approached Moors with no more sympathy or understanding
than their secular contemporaries. The same was true of Muslims. As
those who were perceived as non-Romans encroached upon and were
felt to threaten the late Roman world, ecclesiastical writers did not
draw quite the same conclusions about Romanness as had Augustine.
Romanness did matter after all. Only now it was to be defined in terms of
confession.
Thus at the end of antiquity Roman identity had come to be defined
along multiple axes, the most prominent of which included politics,
high culture, and religion. These definitions could overlap and inform
one another, but they were not always mutually reinforcing. Sixth- and
seventh-century emperors demanded religious conformity as a sign of
political self-identification with the empire and its interests; but Maximus Confessor, Africas Jews, and (I have argued) the Nicene ecclesiastical hierarchy rejected the idea that the first was a necessary consequence
of the second. Similarly, in the Vandal period Fulgentius of Ruspe strenuously opposed Thrasamunds religious policy and expressed surprise to
the king himself at how culturally like a Roman the ruler was in his interest in intellectual debates; but even when in exile in Sardinia Fulgentius
accepted the fundamental political legitimacy of Vandal rule. In the 480s,
other North Africans similarly found that it was possible to go Vandal
politically (in the sense of taking up an office at the Vandal court) and yet
stay Roman religiously, by clinging to their Nicene faith. Indeed, this
was in part precisely what Hunerics persecution was designed to prevent.
Vandals could, and did, go Roman culturally, and yet remain Vandals
politically and religiously. Moors like Cusina might be loyal subjects of
the emperor and even adherents of the Christian faith, but they were still
regarded culturally as Moors and so on. The circumstances of the fifth,
sixth, and seventh centuries allowed for virtually endless permutations.
Ultimately, such permutations made Roman identity a very versatile
concept in late antiquity. It could be used with considerable flexibility
to create a vocabulary of commonality or to reinforce perceived differences across a remarkable range of situations. Yet over the long run this
versatility was to prove a weakness as well as a strength. As non-Romans
came to control the provinces of the western empire, Roman identity was
increasingly decoupled from the Roman state. An indisputable indication
of the long-term success that the empire enjoyed in shaping attitudes,
outlooks, and beliefs in the greater Mediterranean world, this decoupling ironically eased the empires disintegration. The imperial ideal, the
377

Staying Roman
tradition of public authority, and a predictable structure of social relationships regulated by the law came to characterize not only the truncated
Roman state but also its successor kingdoms in the barbarian West. Even
so, in late antiquity the idea of Romanness was not just a question of
nostalgia or of attachment to a vanished past. It had a valuable function,
in that it helped for a time to bind together a new world in which all
the trends were towards fissiparousness rather than unity. In the end, the
political unity of the Mediterranean under Rome was, as we well know,
only a moment in the history of the great inland sea. As that moment
began to fade away, those who dwelled around that sea Roman and
barbarian alike cast about for strategies to ensure some kind of continued Mediterranean unity. Diplomatic, military, and theological answers
all ultimately failed in this task, but the drive to remain connected
and thus to stay Roman remained strong among inhabitants of the
medieval Mediterranean into the eighth century and beyond.

378

B I BL I OG R A P HY

PRIM ARY SOURCES

Acta Conciliorum Oecuminicorum, ed. E. Schwartz and J. Straub, 4 vols. (Berlin, 1914
83).
Acta Conciliorum Oecuminicorum series Secunda, ed. R. Riedinger (Berlin, 1984).
Ado of Vienne, Chronicon, ed. G. H. Pertz, MGH Scriptores 2 (Hanover, 1829),
pp. 31523.
Martyrologium, in H. Quentin (ed.), Les Martyrologes historiques du moyen a ge: etude
sur la formation du Martyrologe romain (2nd edn; Paris, 1908), pp. 465681.
Aegritudo perdicae, ed. E. Baehrens and F. Vollmer, in Poetae latini minores, 6 vols.
(Leipzig, 1914), 5:23850.
Aeneas of Gaza, Theophrastus, PG 85:8721004.
Agapius of Menbij (Mabbug), Kitab al-Unvan: histoire universelle ecrite par Agapius
(Mahboub) de Menbidj, ed. A. Vasiliev, PO 5, 7, 8, 11 (Paris, 191015).
Agathias, Historiae, ed. R. Keydell, in Agathiae Myrinaei historiarum libri quinque
(Berlin, 1967).
Al-Idrs, Nuzhat al-mushtaq f ikhtiraq al-afaq, ed. and trans. R. Dozy and M. J. de
Goeje (Leiden, 1866).
Ambrose of Milan, De bono mortis, ed. C. Schenkl, in Sancti Ambrosii Opera, CSEL
32/1 (Leipzig, 1897), pp. 70353.
Ammianus Marcellinus, Res gestae, ed. W. Seyfarth, in Ammiani Marcellini Rerum
gestarum libri qui supersunt, 2 vols. (Leipzig, 1978).
Anastasios the Sinaite, Per tn n Sin gwn patrwn, ed. F. Nau, Le Texte grec
des recits du moine Anastase sur les saints p`eres du Sina, Oriens Christianus, 2
(1902), pp. 5989.
Annales Regni Francorum, ed. G. H. Pertz and F. Kurze, MGH SRG 6 (Hanover,
1895; repr. 1950).
Anthologia Latina, ed. D. R. Shackleton Bailey, vol. 1, Carmina in codicibus scripta,
fasc. 1: Libri Salmasiani aliorumque carmina (Stuttgart, 1982).
Anthologia Latina, ed. A. Riese, F. Buecheler, and E. Lommatzsch, 2 vols. (Leipzig,
18941926).
Antoninus Honoratus, Epistola consolatoria ad Arcadium, PL 50:56770.
Apringius of Beja, Tractatus in Apocalypsin, ed. M. Ferotin, in Apringius de Beja: son
commentaire de lApocalypse, Biblioth`eque patrologique 1 (Paris, 1900).
Apuleius, Apologia, ed. R. Helm, in Apulei Platonici Madaurensis Opera quae supersunt,
3 vols. (Leipzig, 195591).

379

Bibliography
Arnobius Junior, Commentarii in Psalmos, ed. K.-D. Daur, CCSL 25 (Turnhout,
1990).
Auctores Antiquissimi in Monumenta Germaniae Historica (Hanover and Berlin,
1819).
Augustine, Confessions, ed. J. J. ODonnell, 3 vols. (Oxford, 1992).
Contra Cresconium grammaticum et Donatistam libri iiii, ed. M. Petschenig, CSEL 52
(Vienna, 1909), pp. 323582.
Contra Iulianum libri vi, PL 44:641874.
Contra secundam Juliani responsionem, PL 45:1049608.
De civitate Dei, ed. B. Dombart and A. Kalb, CCSL 478 (Turnhout, 1955).
De gratia Christi et de peccato originali, ed. C. F. Urba and J. Zycha, CSEL 42 (Vienna,
1902), pp. 123206.
De haeresibus, ed. R. Vander Plaetse and C. Beukers, CCSL 46 (Turnhout, 1969),
pp. 283345.
De nuptiis et concupiscentia, ed. C. F. Urba and J. Zycha, CSEL 42 (Vienna, 1902),
pp. 207319.
De unitate ecclesiae [= Epistula ad catholicos de secta Donatistarum], ed. M. Petschenig,
CSEL 52 (Vienna, 1909), pp. 229322.
De utilitate credendi, ed. J. Zycha, CSEL 25 (Vienna, 1891), pp. 148.
Enarrationes in psalmos, ed. E. Dekkers and J. Fraipont, CCSL 3840 (Turnhout,
1956).
Epistolae ad Romanos inchoata expositio, ed. J. Divjak, CSEL 84 (Vienna, 1971).
Epistulae, ed. A. Goldbacher, CSEL 34, 44, 578 (Vienna, 18951923).
Epistulae nuper in lucem prolatae, ed. J. Divjak, CSEL 88 (Vienna, 1981).
In epistolam Joannis ad Parthos, PL 25:19772062.
In Iohannis evangelium tractatus, ed. R. Willems, CCSL 36 (Turnhout, 1954).
Sermo Morin Guelferbytanus, 17, PLS 2:5825.
Sermones, PL 389; PLS 3:417840.
[pseudo-]Augustine, Collatio beati Augustini cum Pascentio ariano, PL 33:115662.
Contra Felicianum Arianum de unitate Trinitatis, PL 62:33352.
De natali domini (Homilia Vindobonense 5a), ed. H. Barre, PLS 4:191112.
De navitate domini (sermo Mai 117), PLS 2:12213.
De occisione infantium 2 (sermo Mai 151), PLS 2:12479.
Epistola 4, PL 33:1095.
In natali domini V (Caillau-Saint-Yves 1.11), PLS 2:92931.
In natali domini VIII (Caillau-Saint-Yves 1.14), PLS 2:9356.
In natali martyris Vincentii II (Caillau-Saint-Yves 1.48), PLS 2:10057.
Quaestiones Veteris et Novi Testamenti, PL 35: 2207386.
Sermo 195, PL 39:210710.
Sermo 245, PL 39:21968.
Sermo 246, PL 39:2198200.
Sermo 381 (De Natali apostolorum Petri et Pauli), PL 39:16834.
Sermo de natiuitate domini, ed. H. Barre, PLS 4:19936.
Solutiones diversarum quaestionum ab haereticis obiuctarum, ed. B. Schwank, CCSL 90,
pp. 14122.
Avitus of Vienne, Dialogi cum Gundobado rege uel librorum contra Arrianos reliquiae, ed.
R. Peiper, MGH AA 6/2 (Berlin, 1883), pp. 115.
Epistulae ad diversos, ed. R. Peiper, MGH AA 6/2 (Berlin, 1883), pp. 35103.

380

Bibliography
Baladhur, Ahmad ibn-Yahya al-, Kitab futuh al-buldan, ed. M. J. de Goeje, in Liber
expugnationis regionum (2nd edn; Leiden, 1968).
Bar Hebraeus, Gregorius Abu-l-Faraj (Abulpharagius), Chronography, ed. and trans.
E. A. W. Budge, 2 vols. (London, 1932).
Bede, De temporum ratione, ed. C. Jones, CCSL 123B (Turnhout, 1977).
Historia ecclesiastica gentis Anglorum, ed. and trans. B. Colgrave and R. A. B. Mynors,
in Bedes Ecclesiastical History of the English People (Oxford, 1969).
Boccaccio, Giovanni, Decameron, ed. V. Branca (Turin, 1980).
Boethius, De differentiis topicis, PL 64:1173216.
Boniface, Epistolae, ed. E. Dummler, MGH Epist. 3 (Berlin, 1892), pp. 215433.
Breviarium Hipponense, ed. C. Munier, in Concilia Africae a.345a.525, CCSL 149
(Turnhout, 1974), pp. 2253.
Breviarium Syriacum, ed. and (Greek) trans. L. Duchesne, AASS Nov. 2/1 (Brussels,
1894), pp. liilxiv.
Caesarius of Arles, Breviarium adversus haereticos, ed. G. Morin, in S. Caesarii opera
omnia, 2 vols. (Maredsous, 1942), 2:182208.
Expositio de Apocalypsi sancti Iohannis, ed. G. Morin, in S. Caesarii opera omnia,
2 vols. (Maredsous, 1942), 2:21077.
Candidus, fragments, ed. R. C. Blockley, in The Fragmentary Classicising Historians
of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and Malchus, 2 vols.
(Liverpool, 1983), 2:46470.
Canones in causa Apiarii, ed. C. Munier, in Concilia Africae a.345a.525, CCSL 149
(Turnhout, 1974), pp. 98148.
Capreolus of Carthage, Epistula ad synodum Ephesinum, ed. E. Scwartz, in ACOec.
1/2, pp. 645.
Cassiodorus, Chronicon, ed. T. Mommsen, MGH AA 11/Chron. Min. 2 (Berlin,
1894), pp. 12061.
De orthographia, ed. H. Keil, in Grammatici latini, 8 vols. (Leipzig, 185580), 7:143
210.
Expositio psalmorum, ed. M. Adriaen, 2 vols., CCSL 978 (Turnhout, 1958).
Institutiones, ed. R. A. B. Mynors (Oxford, 1937).
Variarum libri xii, ed. A. J. Fridh, CCSL 96/1 (Turnhout, 1973).
[pseudo-]Cassiodorus, De artibus ac disciplines liberalium litterarum, PL 70:1149220.
Cassius Felix, De medicina, ed. V. Rose (Leipzig, 1879).
Cedrenus, George, Compendium Historiarum, ed. I. Bekker, 2 vols., Corpus Scriptorum Historiae Byzantinae 345 (Bonn, 18389).
Cerealis of Castellum Ripensis, Libellus contra Maximinum Arianum, PL 58:757
68.
Chronica Alexandrina, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin, 1892),
pp. 89140.
Chronica Gallica a. dxi, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin, 1892),
pp. 63244 and 64766.
Chronicon ad annum Domini 1234 pertinens, trans. J.-B. Chabot, CSCO (Scr. Syr.) ser.
3, 14 (Paris, 1920), pp. 17266.
Chronicon Paschale, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 1516
(Bonn, 1832).
Claudian, Carmina, ed. J. B. Hall, in Claudiani Carmina (Leipzig, 1985), pp. 1288.
In Gildonem, ed. J. B. Hall, in Claudianus Carmina, pp. 10727.

381

Bibliography
Codex Carolinus, ed. W. Gundlach, MGH Epist. 3 (Berlin, 1892), pp. 469
657.
Codex Justinianus, ed. P. Kruger, in Corpus Juris Civilis, vol. 2 (Berlin, 1892).
Codex Theodosianus, ed. T. Mommsen and P. Meyer, 2 vols. (Berlin, 1905).
Collatio cum Maximino Arianorum episcopo, PL 42:70942.
Collectio Arriana Veronensis, ed. R. Gryson, in Scripta Arriana Latina, CCSL 87 (Turnhout, 1982), pp. 1145.
Collectio Avellana, ed. O. Guenther, in Epistulae imperatorum, pontificum, aliorum inde
ab a. ccclxvii usque ad a. dliii datae Avellana quae dicitur collectio, 2 vols., CSEL 35
(Vienna, 18958).
Concilium Carthaginense a.525, ed. C. Munier, in Concilia Africae a.345a.525, CCSL
149 (Turnhout, 1974), pp. 25482.
Continuatio Hauniensis Prosperi, ed. T. Mommsen, MGH AA 9/Chron. Min. 1
(Berlin, 1961), pp. 298339.
Contra Felicianum Arianum de unitate Trinitatis, PL 62:33352.
Contra Judaeos, ed. R. Gryson, in Scripta Arriana Latina, CCSL 87 (Turnhout, 1982),
pp. 93117.
Contra paganos, ed. R. Gryson, in Scripta Arriana Latina, CCSL 87 (Turnhout, 1982),
pp. 11841.
Corippus, Flavius Cresconius, In laudem Iustini Augusti minoris, ed. S. Ant`es, in Eloge
de lempereur Justin II (Paris, 1981).
Iohannis seu de bellis Libycis Libri VIII, ed. J. Diggle and F. R. D. Goodyear (Cambridge, 1970).
Corpus Juris Civilis, ed. P. Krueger, T. Mommsen, R. Scholl, and W. Kroll, 3 vols.
(Berlin, 18925; repr. 1954).
Cyprian, Firminus, and Viventius, Messianus presbyter and Stephen the Deacon, Vita
S. Caesarii episcopi Arelatensis, ed. G. Morin, in Sancti Caesarii episcopi Arelatensis
Opera omnia, 2 vols. (Maredsous, 193742), 2:296345.
Cyril of Scythopolis, Vita Sabae, ed. E. Schwartz, in Kyrillos von Skythopolis, Texte
und Untersuchungen zur Geschichte der altchristlichen Literatur 49/2 (Leipzig,
1939).
De ratione paschae, ed. B. Krusch, in Studien zur christlich-mittelalterlichen Chronologie,
2 vols. (Leipzig, 1880), 1:27997.
Depositio martyrum, ed. R. Valentini and G. Zucchetti, in Codice Topografico della Citt`a
di Roma, 4 vols., Fonti per la Storia dItalia 81, 88, 90, and 91 (Rome, 194053),
2:1728.
Digesta, ed. T. Mommsen, in Corpus Juris Civilis, vol. 1 (Berlin, 1895).
Dio Cassius, Historiae Romanae, ed. U. P. Boissevain, in Cassii Dionis Cocceiani Historiarum Romanarum quae supersunt, 5 vols. (Berlin, 18951931).
Doctrina Jacobi nuper baptisati, ed. and trans. G. Dagron and V. Deroche, in Juifs
et Chretiens dans lOrient du VIIe si`ecle, Travaux et Memoires 11 (1991),
pp. 17273.
Dracontius, Blossius Aemilius, Laudes Dei, ed. C. Moussy and C. Camus, uvres, 4
vols. (Paris, 198596), 1:151233 and 2:1654.
Romulea, ed. J. Bouquet and E. Wolff, uvres, 4 vols. (Paris, 198596), 3:13460
and 4:176.
Satisfactio ad Gunthamundum regem Wandalorum, ed. C. Moussy, uvres, 4 vols.
(Paris, 198596), 2:17691.

382

Bibliography
[pseudo-]Dracontius, De origine rosarum, ed. E. Baehrens and F. Vollmer, in Poetae
latini minores, 6 vols. (Leipzig, 18811911), 5:237.
Edictum Diocletiani et collegarum de pretiis rerum venalium, ed. T. Mommsen, CIL
3:190953.
Einhard, Vita Karoli magni, ed. G. H. Pertz and G. Waitz, MGH SRG (5th edn;
Hanover, 1905).
Ennodius, Epistulae, ed. F. Vogel, MGH AA 7 (Berlin, 1885).
Panegyricus dictus clementissimo regi Theoderico, ed. F. Vogel, MGH AA 7 (Berlin,
1885), pp. 20314.
Epistolae aevi Merowingici collectae, ed. W. Gundlach, MGH Epist. 3 (Berlin, 1892),
pp. 43468.
Epistolae Austrasicae, ed. W. Gundlach, MGH Epist. 3 (Berlin, 1892), pp. 11053.
Epitoma Cononiana, ed. T. Mommsen, MGH Gestorum Pontificum Romanorum 1
(Berlin, 1898), pp. 22763.
Epitome de Caesaribus, ed. F. Pichlmayr, in Sexti Aurelii Victoris Liber de Caesaribus
(Leipzig, 1966).
Eugippius, Epistula ad Probam virginem, ed. P. Knoll, CSEL 9/1 (Vienna, 1885),
pp. 12.
Excerpta ex operibus s. Augustini, ed. P. Knoll, CSEL 9/1 (Vienna, 1885).
Eulogius of Toledo, Memorialis sanctorum, PL 115:731818.
Eutychius of Alexandria (Sacd ibn Batrk), Annales ecclesiasticae, ed. and trans.
M. Breydy, in Das Annalenwerk des Eutychios von Alexandrien. Ausgewahlte
Geschichten und Legenden kompiliert von Said ibn Batrq um 935 ad, CSCO (Scr.
Arab.) 445 (Louvain, 1985).
Evagrius Scholasticus, Historia Ecclesiastica, ed. J. Bidez and L. Parmentier, in The
Ecclesiastical History of Evagrius with the Scholia (London, 1898).
Excerpta Valesiana, ed. J. Moreau and V. Velkov (2nd edn; Leipzig, 1968).
Expositio totius mundi et gentium, ed. J. Rouge, SC 124 (Paris, 1966).
Facundus of Hermiane, Contra Mocianum scholasticum, Opera omnia, ed. J.-M.
Clement and R. Vander Plaetse, 2 vols., CCSL 90A (Turnhout, 1974),
pp. 399416.
Epistula fidei catholicae in defensione trium capitulorum, Opera omnia, ed. J.-M.
Clement and R. Vander Plaetse, 2 vols., CCSL 90A (Turnhout, 1974), pp. 419
34.
Pro defensione trium capitulorum, Opera omnia, ed. J.-M. Clement and R. Vander
Plaetse, 2 vols., CCSL 90A (Turnhout, 1974), pp. 3398.
Fasti Vindobonenses Priores, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin,
1892), pp. 274336.
Fastidiosus, Sermo, ed. J. Fraipont, CCSL 91 (Turnhout, 1968), pp. 2803.
Felix II (III), Epistulae, ed. A. Thiel, in Epistolae Romanorum pontificum genuinae
(Braunsberg, 1868; repr. Hildesheim, 1974), pp. 22184.
Ferrandus of Carthage, Breviatio canonum, ed. C. Munier, in Concilia Africae a.345
a.525, CCSL 149 (Turnhout, 1974), pp. 284311.
Epistulae, PL 67:887950.
Vita S. Fulgentii episcopi Ruspensis, ed. G.-G. Lapeyre, in Vie de Saint Fulgence de
Ruspe (Paris, 1929).
Flodoard of Rheims, Historiae Remensis ecclesiae, ed. M. Stratmann, MGH Scriptores
36 (Hanover, 1998).

383

Bibliography
Florus of Lyons, Carmina, ed. E. Dummler, MGH Poet. 2 (Berlin, 1884), pp. 507
66.
Martyrologium, in H. Quentin (ed.), Les Martyrologes historiques du moyen a ge:
etude sur la formation du Martyrologe romain (2nd edn; Paris, 1908), pp. 222
408.
[pseudo-]Fredegarius scholasticus, Chronicorum libri iv cum continuationibus, ed. B.
Krusch, MGH SRM 2 (Hanover, 1888), pp. 18193.
Fulgentius Mythographer, De aetatibus mundi et hominis, ed. R. Helm, Opera (Leipzig,
1898; repr. 1970), pp. 12779.
Expositio sermonum antiquorum, ed. R. Helm, Opera (Leipzig, 1898; repr. 1970),
pp 10926.
Expositio Virgilianae continentiae, ed. R. Helm, Opera (Leipzig, 1898; repr. 1970),
pp. 81107.
Mitologiarum libri tres, ed. R. Helm, Opera (Leipzig, 1898; repr. 1970), pp. 380.
Fulgentius of Ruspe, Abecedarium, ed. C. Lambot, S. Fulgentii episcopi Ruspensis opera,
2 vols., CCSL 9191A (Turnhout, 1968), 2:87585.
Ad Euthymium de remissione peccatorum libri II, ed. J. Fraipont, S. Fulgentii episcopi
Ruspensis opera, 2 vols., CCSL 9191A (Turnhout, 1968), 2:649707.
Ad Monimum, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera, 2 vols., CCSL
9191A (Turnhout, 1968), 1:164.
Ad Trasamundum regem libri iii, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera,
2 vols., CCSL 9191A (Turnhout, 1968), 1:95185.
Contra Fabianum fragmenta, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera,
2 vols., CCSL 9191A (Turnhout, 1968), 2:763866.
Contra sermonem Fastidiosi Ariani, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera,
2 vols., CCSL 9191A (Turnhout, 1968), 1:283308.
De fide ad Petrum seu de regula fidei, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis
opera, 2 vols., CCSL 9191A (Turnhout, 1968), 2:71160.
De veritate praedestinationis et gratiae, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis
opera, 2 vols., CCSL 9191A (Turnhout, 1968), 2:458548.
Dicta regis Trasamundi et contra ea responsionum liber unus, ed. J. Fraipont, S. Fulgentii
episcopi Ruspensis opera, 2 vols., CCSL 9191A (Turnhout, 1968), 1:6594.
Epistulae, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera, 2 vols.,
CCSL 9191A (Turnhout, 1968), 1:189280, 31112, 359444, 2:44757, 551
624.
Liber de Trinitate ad Felicem, ed. J. Fraipont, S. Fulgentii episcopi Ruspensis opera,
2 vols., CCSL 9191A (Turnhout, 1968), 2:63346.
[pseudo-]Fulgentius, Sermo de Symbolo, ed. G. Morin, PLS 3:13706.
Sermones, PL 65:855954.
Gelasius I, Constituta quae episcopi in sua ordinatione accipiunt, PL 59:1378.
Epistolae, ed. A. Thiel, in Epistolae Romanorum pontificum genuinae (Braunsberg,
1868; repr. Hildesheim, 1974), pp. 287613.
Gennadius, De viris inlustribus, ed. E. C. Richardson, in Hieronymus liber De viris
inlustribus; Gennadius liber De viris inlustribus, Texte und Untersuchungen zur
Geschichte der altchristlichen Literatur 14/1 (Leipzig, 1896), pp. 5797.
George of Cyprus, Descriptio orbis Romani, ed. H. Gelzer (Leipzig, 1890).
George the Monk, Chronicon breve, PG 110.
Gesta apud Zenophilum, ed. C. Ziwsa, CSEL 26 (Vienna, 1893), pp. 18597.

384

Bibliography
Gesta conlationis Carthaginiensis, anno 411, ed. S. Lancel, CCSL 149A (Turnhout,
1974).
Gesta episcoporum Neapolitanorum, ed. G. Waitz, MGH SRL (Hanover, 1878),
pp. 40236.
Gesta sanctorum Rotonensium, ed. C. Brett, in The Monks of Redon (Woodbridge,
1989), pp. 107219.
Glaber, Ralph, Historiarum libri quinque, ed. and trans. J. France (Oxford, 1989).
Gregory I, Dialogi, ed. A. de Vogue , in Diologues, 3 vols., SC 251, 260, 265 (Paris,
197880).
Registrum Epistularum, ed. D. Norberg, CCSL 140140A (Turnhout, 1982).
Gregory VII, Epistulae, ed. E. Caspar, 2 vols., MGH Epist. Select. 2 (Berlin, 19203).
Gregory of Tours, In gloria martyrum, ed. B. Krusch, MGH SRM 1/2 (Hanover,
1885), pp. 484561.
Liber vitae patrum, ed. B. Krusch, MGH SRM 1/2 (Hanover, 1885), pp. 661744.
Libri Historiarum X, ed. B. Krusch and W. Levison, MGH SRM 1/1 (2nd edn;
Hanover, 1951).
Gulosus, Probus, et al., Epistula ad Paulum episcopum Constantinopolitanum,
ed. R. Riedinger, ACOec.2 , 1:8095.
Hilarius, Epistulae, ed. A. Thiel, in Epistulae Romanorum pontificum genuinae, vol. 1
(Braunsberg, 1867; repr. Hildesheim, 1974), pp. 12670.
Histoire nestorienne inedite (Chronique de Seert), ed. and trans. A. Scher (Paris, 1908
18).
Honorius I, Epistolae, PL 80:46984.
Hormisdas, Epistulae, ed. A. Thiel, in Epistulae Romanorum pontificum genuinae,
vol. 1 (Braunsberg, 1867; repr. Hildesheim, 1974), pp. 741990.
Hydatius, Chronicon, ed. and trans. R. W. Burgess, in The Chronicle of Hydatius and
the Consularia Constantinopolitana: Two Contemporary Accounts of the Final Years of
the Roman Empire (Oxford, 1993), pp. 70122.
Ibn c Abd al-Hakam, Futuh Ifrqiya wal-Andalus, ed. and trans. A. Gateau, in Conquete
de lAfrique du nord et de lEspagne (2nd edn; Algiers, 1947).
Ildefonsus of Toledo, De viris illustribus, ed. C. Codoner Merino, CCSL 114A
(Turnhout, 2007).
Isidore of Seville, De origine Gothorum, ed. and trans. C. Rodrguez Alonso, in Las
Historias de los godos, vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica y
traduccion, Fuentes y estudios de historia leonesa 13 (Leon, 1975), pp. 172280.
De viris inlustribus, ed. C. Codoner Merino, in El De viris illustribus de Isidoro de
Sevilla: Estudio y edicion crtica, Theses et Studia Philologica Salmanticensia 12
(Salamanca, 1964).
Etymologiae, ed. W. M. Lindsay, in Isidori Hispalensis episcopi Etymologiarum sive
originum libri XX, 2 vols. (Oxford, 1911).
Historia Gothorum, ed. and trans. C. Rodrguez Alonso, Las Historias de los godos,
vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica y traduccion, Fuentes y
estudios de historia leonesa 13 (Leon, 1975), pp. 172274.
Historia Vandalorum, ed. and trans. C. Rodrguez Alonso, Las Historias de los godos,
vandalos y suevos de Isidoro de Sevilla: Estudio, edicion crtica y traduccion, Fuentes y
estudios de historia leonesa 13 (Leon, 1975), pp. 288321.
Jerome, Adversus Jovinianum, PL 23:221352.
Commentarius in Ionam, ed. Y. Duval, SC 323 (Paris, 1985), pp. 160316.

385

Bibliography
John Lydus, De Magistratibus Populi Romani, ed. R. Wunsch (Leipzig, 1903).
John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn, 1831).
John Moschus, Pratum spirituale, PG 87/3:28523112.
John of Antioch, fragments, ed. K. Muller, in Fragmenta Historicorum Graecorum,
5 vols. (Paris, 184170), 4:535622 and 5:2739.
John of Biclar, Chronicon, ed. R. Collins, CCSL 173A (Turnhout, 2001), pp. 5783.
John of Ephesus, Historiae ecclesiasticae pars tertia, ed. and trans. E. W. Brooks, CSCO
(Scr. Syr.) ser. 3, 3 (Louvain, 1936).
John of Nikiu, The Chronicle of John (c.690 ad), Coptic bishop of Nikiu, trans. R. H.
Charles (Amsterdam, 1982).
Jordanes, Getica, ed. F. Giunta and A. Grillone, in Iordanis De origine actibusque
Getarum, Fonti per la storia dItalia 117 (Rome, 1991).
Romana, ed. T. Mommsen, MGH AA 5 (Berlin, 1882), pp. 152.
Julius Caesar, Bellum Gallicum, ed. W. Hering, in Commentarium rerum gestarum,
vol. 1 (Leipzig, 1987).
Junillus Africanus, Instituta regularia diuinae legis, ed. H. Kihn, in Theodor von Mopsuestia und Junilius Africanus als Exegeten: Nebst einer kritischen Textausgabe von des
letzteren Instituta regularia divinae legis (Freiburg im Breisgau, 1880), pp. 465528
= M. Maas, Exegesis and Empire in the Early Byzantine Mediterranean, Studien
und Texte zu Antike und Christentum 17 (Tubingen, 2003), pp. 118234.
Justinian, Novellae, ed. R. Scholl and W. Kroll, in Corpus Juris Civilis, vol. 3 (Berlin,
1895; repr. 1954).
Kalendarium carthaginiense, PL 13:121930.
Laterculus regum Wandalorum et Alanorum, ed. T. Mommsen, MGH AA 13/Chron.
Min. 3 (Berlin, 1898), pp. 45660.
Leo I, Epistolae, PL 54:5931218.
Leo III, Epistolae, ed. K. Hampe, MGH Epist. 5 (Berlin, 1899), pp. 85104.
Leo IX, Epistolae, PL 143:591794.
Leo abbas et legatus, Epistola ad Hugonem et Robertum reges, ed. G. H. Pertz, MGH
Scriptores 3 (Hanover, 1839), pp. 68690.
Liber de Trinitate, ed. J. Fraipont, in Florilegia Biblica Africana saec. V, CCSL 90
(Turnhout, 1961), pp. 23960.
Liber diurnus Romanorum pontificum, ed. T. E. von Sickel (Vienna, 1889).
Liber genealogus, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin, 1892),
pp. 15496.
Liber generationis, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin, 1892),
pp. 89140.
Liber Pontificalis, ed. T. Mommsen, MGH Gestorum Pontificum Romanorum 1
(Berlin, 1898).
Liberatus, Breuarium causae Nestorianorum et Eutychianorum, ed. E. Schwartz, in
ACOec. 2/5 (Berlin, 1936), pp. 98141.
Licinianus of Cartagena, Epistulae, ed. J. Madoz, in Liciniano de Cartagena y sus Cartas:
edicion crtica y estudio historico (Madrid, 1948).
Malchus of Philadelphia, fragments, ed. R. C. Blockley, in The Fragmentary Classicising Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and
Malchus, 2 vols. (Liverpool, 1983), 2:40262.

386

Bibliography
Marcellinus comes, Additamentum ad a.DXLVIII, ed. T. Mommsen, MGH AA
11/Chron. Min. 2 (Berlin, 1894), pp. 1048.
Chronicon ad a.DXVIII continuatum ad a.DXXXIV, ed. T. Mommsen, MGH AA
11/Chron. Min. 2 (Berlin, 1894), pp. 60104.
Marius of Avenches, Chronica, ed. T. Mommsen, MGH AA 11/Chron. Min. 2
(Berlin, 1894), pp. 22539.
Martianus Capella, De nuptiis Philologiae et Mercurii, ed. J. Willis (Leipzig, 1983).
Martin I, Epistolae, PL 87:119204.
Martyrologium Hieronymianum, ed. H. Quentin and H. Delehaye, AASS Nov. Propylaeum (Brussels, 1931).
Maximus Confessor, Disputatio cum Pyrrho, PG 91:288353.
Epistolae, PG 91:363649.
Opuscula Theologica et Polemica, PG 91:9285.
Merobaudes, Flavius, Carmina, ed. F. Vollmer, MGH AA 14 (Berlin, 1905), pp. 36.
Panegyricus II, ed. F. Vollmer, MGH AA 14 (Berlin, 1905), pp. 1118.
Michael the Syrian, Chronicle, ed. and trans. J. B. Chabot, in Chronique de Michel le
Syrien, Patriarche Jacobite dAntioche (11661199), 4 vols. (Paris, 18991910).
Miracles of St Artemios: A Collection of Miracle Stories by an Anonymous Author of
Seventh-Century Byzantium, ed. and trans. V. S. Crisafulli and J. W. Nesbitt, The
Medieval Mediterranean 13 (Leiden, 1997).
Nicephorus Callistus, Historia Ecclesiastica, PG 1457.
Nicephorus of Constantinople, Breviarium, ed. and trans. C. Mango, in Nikephoros
Patriarch of Constantinople: Short History, Corpus Fontium Historiae Byzantinae
13 (Washington, DC, 1990).
Nicholas II, Epistolae et diplomata, PL 143:130162.
Notitia provinciarum et civitatum Africae, ed. M. Petschenig, CSEL 7 (Vienna, 1881),
pp. 11534.
Notker Balbulus, Gesta Karoli magni imperatoris, ed. H. Haefele, MGH SRG n.s. 12
(Berlin, 1962).
Olympiodorus, fragments, ed. R. C. Blockley, in The Fragmentary Classicising Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and Malchus,
2 vols. (Liverpool, 1983), 2:152220.
Orosius, Historiae adversum paganos, ed. C. Zangemeister, CSEL 5 (Vienna, 1882).
Parthemii presbyteri rescriptum ad Sigisteum, PLS 3:448.
Passio s. Polyeucti, ed. B. Aube, in Polyeucte dans lhistoire (Paris, 1882), pp. 73104.
Passio s. Victoris martyris Caesareae, ed. Bollandists, in Catalogus codicum hagiographicorum
latinorum antiquiorum saeculo XVI qui asservantur in Bibliotheca Nationali Parisiensi,
3 vols., Subsidia hagiographica 2 (Brussels, 188993), 3:5046.
Passio sanctarum Perpetuae et Felicitatis, ed. C. van Beek (Nijmegen, 1936).
Passio septem monachorum, ed. M. Petschenig, CSEL 7 (Vienna, 1881), pp. 10814.
Paul the Deacon, Historia Langobardorum, ed. L. Capo, in Paolo Diacono, Storia dei
Longobardi (Milan, 1992).
Paulinus, Epigramma, ed. C. Schenkl, CSEL 16/1 (Vienna, 1888), pp. 5038.
Paulinus of Nola, Carmina. ed. W. Hartel, CSEL 30 (Vienna, 1894).
Epistulae, ed. W. Hartel, CSEL 29 (Vienna, 1894).
Pelagius, In defensione trium capitulorum, ed. R. Devreesse, Studi e testi 57 (Vatican
City, 1932).

387

Bibliography
Peter the Subdeacon, Passio sanctae Restitutae, ed. E. DAngelo, in Pietro suddiacono
napoletano: Lopera agiografica (Florence, 2002), pp. 18399.
Pliny, Naturalis historia, ed. L. Ian and C. Mayhoff, 6 vols. (Stuttgart, 18921909;
repr. 196770).
Pompeius, Commentum artis Donati, ed. H. Keil, in Grammatici Latini, 8 vols. (Leipzig,
1867), 5:81312.
Pontianus, Epistola Pontiani episcopi ad Justinianum imperatorem de tribus capitulis, PL
67:9958.
Possidius of Calama, Vita Augustini, ed. M. Pellegrino, in Vita di S. Agostino, Verba
seniorum 4 (Alba, 1955).
Primasius of Hadrumetum, Commentarius in Apocalypsin, ed. A. W. Adams, CCSL
92 (Turnhout, 1985).
Priscian, Institutionum grammaticarum libri xviii, ed. M. Hertz, in Grammatici Latini,
ed. H. Keil, 8 vols. (Leipzig, 18559), vols. 23.
Priscus, fragments, ed. and trans. R.C. Blockley, in The Fragmentary Classicising
Historians of the Later Roman Empire: Eunapius, Olympiodoros, Priscus and Malchus,
2 vols. (Liverpool, 1983), 2:222400.
Procopius of Caesarea, Opera omnia, ed. J. Haury, 4 vols. (Leipzig, 190513; repr.
19624).
Prosper, Epitoma de chronicon, ed. T. Mommsen, MGH AA 9/Chron. Min. 1 (Berlin,
1892), pp. 385485.
Prudentius, Peristefanon, ed. M. Cunningham, CCSL 126 (Turnhout, 1966),
pp. 251389.
Quodvultdeus of Carthage, Adversus quinque haereses, in Opera Quodvultdeo
Carthaginiensi episcopo tributa, ed. R. Braun, CCSL 60 (Turnhout, 1976),
pp. 259301.
Contra Iudaeos, paganos et Arrianos, in Opera Quodvultdeo Carthaginiensi episcopo
tributa, ed. R. Braun, CCSL 60 (Turnhout, 1976), pp. 22558.
De accedentibus ad gratiam 2, in Opera Quodvultdeo Carthaginiensi episcopo tributa, ed.
R. Braun, CCSL 60 (Turnhout, 1976), pp. 45970.
De cataclysmo, in Opera Quodvultdeo Carthaginiensi episcopo tributa, ed. R. Braun,
CCSL 60 (Turnhout, 1976), pp. 40720.
De quattuor virtutibus caritatis, in Opera Quodvultdeo Carthaginiensi episcopo tributa,
ed. R. Braun, CCSL 60 (Turnhout, 1976), pp. 36578.
De tempore barbarico 1, in Opera Quodvultdeo Carthaginiensi episcopo tributa,
ed. R. Braun, CCSL 60 (Turnhout, 1976), pp. 42137.
De tempore barbarico 2, in Opera Quodvultdeo Carthaginiensi episcopo tributa,
ed. R. Braun, CCSL 60 (Turnhout, 1976), pp. 47186.
Liber promissionum et praedictorum Dei, in Opera Quodvultdeo Carthaginiensi episcopo
tributa, ed. R. Braun, CCSL 60 (Turnhout, 1976), pp. 1223.
Registri Ecclesiae Carthaginensis Excerpta, ed. C. Munier, in Concilia Africae a.345a.525,
CCSL 149 (Turnhout, 1974), pp. 173247.
Relatio factae motionis inter domnum Maximum monachum et socium eius coram principibus
in secretario, ed. P. Allen and B. Neil, in Scripta saeculi VII vitam Maximi Confessoris
illustrantia, Corpus Christianorum Series Graeca 39 (Turnhout, 1999), pp. 151.
Sallust (C. Sallustus Crispus), Iugurtha, ed. L. D. Reynolds, in C. Sallusi Crispi
Catalina, Iugurtha, Historiarum Fragmenta Selecta (Oxford, 1991), pp. 54149.

388

Bibliography
Salvian of Marseilles, De gubernatione Dei, ed. G. Lagarrigue, in Du Gouvernement de
Dieu, vol. 2 of uvres, SC 220 (Paris, 1975).
Scriptores Historiae Augustae, ed. E. Hohl, 2 vols. (Leipzig, 195565).
Sebeos, The Armenian History Attributed to Sebeos, trans. R. Thomson (Liverpool,
1999).
Senarii v.i. comitis patrimonii et patricii epitaphium, ed. T. Mommsen, MGH AA 12
(Berlin, 1894), pp. 499.
Servius, Commentarius in Vergilii Aeneidos, ed. G. Thilo, in Servii grammatici qui feruntur
in Vergilii carmina commentarii, ed. G. Thilo and H. Hager, 3 vols. (Leipzig, 1881
1902).
Sidonius Apollinaris, Carmina, ed. A. Loyen, in Sidoine Appollinaire, 3 vols. (Paris,
1960), vol. 1.
Epistulae, ed. A. Loyen, in Sidoine Appollinaire, 3 vols. (Paris, 196070), vols. 23.
Sigisteus comes, Epistola ad Parthemium, PLS 3:4478.
Silverius, Epitaphium Adriani I Papae, ed. E. Dummler, MGH Poet. 1 (Berlin, 1881),
p. 114.
Socrates, Historia Ecclesiastica, PG 67:29841.
Solinus, Collectanea rerum memorabilium, ed. T. Mommsen (Berlin, 1895; repr. 1958).
Sophronius of Jerusalem, Miracula sanctorum Cyri et Ioannis, ed. N. Fernandez Marcos,
in Los Thaumata de Sofronio: Contribucion al estudio de la Incubatio cristiana (Madrid,
1975).
Sozomen, Historia Ecclesiastica, PG 67:8441629.
Statius, Silvae, ed. A. Marastoni (Leipzig, 1961).
Symmachus, Epistulae, ed. A. Thiel, in Epistolae Romanorum pontificum genuinae,
vol. 1 (Braunsberg, 1868; repr. Hildesheim, 1974), pp. 641734.
Symmachus, Q. Aurelius, Epistulae, ed. O. Seeck, MGH AA 6/1 (Berlin, 1883),
pp. 1278.
Synaxarium ecclesiae Constantinopolitanae e codice Sirmondiano nunc Berolinensi adiectis
synaxariis selectis, ed. H. Delehaye, AASS Nov. Propylaeum (Brussels, 1902;
repr. 1985).
Tablettes Albertini: actes prives de lepoque vandale (fin du Ve si`ecle), ed. C. Courtois,
L. Leschi, C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952).
Tacitus, Annales, ed. H. Heubner (Stuttgart, 1983).
Testimonia de Patre et Filio et Spiritu Sancto, ed. D. de Bruyne, in Florilegia biblica
Africana saec. V, CCSL 90 (Turnhout, 1961), pp. 22733.
Theodoret of Cyrrhus, Epistolae, ed. and trans. Y. Azema, in Correspondance, 4 vols.,
SC 40, 98, 111, and 429 (Paris, 196498).
Theodosius, De situ terrae sanctae, ed. P. Geyer, CCSL 175 (Turnhout, 1965),
pp. 11525.
Theophanes, Chronographia, ed. C. de Boor, 2 vols. (Leipzig, 18835; repr.
Hildesheim, 1963).
Theophanes Continuatus, Chronographia, ed. I. Bekker (Bonn, 1838).
Theophylactus Simocatta, Historiae, ed. C. de Boor and P. Wirth (Stuttgart, 1972).
Translatio duodecim martyrum, ed. G. Waitz, MGH SRL (Hanover, 1878), pp. 5746.
Usuard, Martyrologium, ed. J. Dubois, in Le Martyrologe dUsuard, Subsidia Hagiographica 40 (Brussels, 1965).
Valentinian III, Novellae, ed. T. Mommsen and P. Meyer, in CTh 2:69154.

389

Bibliography
Vegetius Renatus, Flavius, Digestorum artis mulomedicinae libri, ed. E. Lommatzsch
(Leipzig, 1903).
Victor of Carthage, Epistula ad Theodorum papam, ed. R. Riedinger, ACOec.2 , 1:99
103.
Victor of Vita, Historia persecutionis Africanae provinciae, ed. M. Petschenig, CSEL 7
(Vienna, 1881).
Victor Tonnennensis, Chronicon, ed. C. Cardelle de Hartmann, CCSL 173A (Turnhout, 2001).
Vigilius of Thapsa, Contra Arianos, Sabellianos, Photinianos dialogus, PL 62:179238.
[pseudo-]Vigilius of Thapsa, Contra Varimadum, ed. B. Schwank, in Florilegia Biblica
Africana saec. V, CCSL 90 (Turnhout, 1961), pp. 1134.
Solutiones obiectionum Arianorum, PL 62:46972.
Virgil (P. Vergilius Maro), Aeneidos, ed. R. A. B. Mynors, in P. Vergili Maronis Opera
(Oxford, 1969), pp. 103422.
Vita S. Eliae iunioris, ed. G. Rossi Taibbi, in Vita di SantElia il Giovane, Istituto
Siciliano di Studi Bizantini e Neoellenici Testi e Monumenti 7 (Palermo,
1962).
Vitas Sanctorum Patrum Emeretensium, ed. and trans. J. Garvin, Studies in Medieval
and Renaissance Latin Language and Literature 19 (Washington, DC, 1946).
Zachariah of Mitylene, The Syriac Chronicle Known as that of Zachariah of Mitylene,
trans. F. J. Hamilton and E. W. Brooks (London, 1899).
Zeno of Verona, Tractatus, ed. B. Lofstedt, CCSL 22 (Turnhout, 1971).
Zonaras, John, Epitomae Historiarum, ed. L. Dindorf, 6 vols. (Leipzig, 186875).

EPIGRAPHIC SOURCES

Alfoldy, G., Die romischen Inschriften von Tarraco, Madrider Forschungen 10, 2 vols.
(Berlin, 1975).
LAnnee epigraphique, Revue des publications epigraphiques relatives a lantiquite romaine
(1962).
Brusin, G., Inscriptiones Aquileiae, Pubblicazioni della Deputazione di storia patria
per il Friuli 20, 3 vols. (Udine, 19913).
Cagnat, R. and A. Merlin, Inscriptions latines dAfrique (Tripolitaine, Tunisie, Maroc)
(Paris, 1923).
Corpus inscriptionum Graecarum, vol. 4, ed. E. Curtius and A. Kirchhoff (Berlin, 1877;
repr. Hildesheim, 1977).
Corpus inscriptionum Latinarum (Berlin, 1862).
Diehl, E., Inscriptiones Latinae Christianae Veteres, 3 vols. (Berlin, 192531).
Ennabli, L., Inscriptions de Bir el Knissia, in S. Stevens (ed.), Bir el Knissia at
Carthage: A Re-Discovered Cemetery Church. Report No. 1, Journal of Roman
Archaeology Supplementary Series 7 (Ann Arbor, Mich., 1993), pp. 25788.
Les Inscriptions funeraires chretiennes de Carthage II: la basilique de Mcidfa, Collection

de lEcole
francaise de Rome 62 (Rome, 1982).
Les Inscriptions funeraires chretiennes de Carthage III: Carthage intra et extra muros,

Collection de lEcole
francaise de Rome 151 (Rome, 1991).
Les Inscriptions funeraires chretiennes de la basilique dite de Sainte-Monique a` Carthage,

Collection de lEcole
francaise de Rome 25 (Rome, 1975).

390

Bibliography
Euzennat, M. and J. Marion (eds.), Inscriptions antiques du Maroc, vol. 2, Inscriptions
latines (Paris, 1982).
Gsell, S. and H.-G. Pflaum, Inscriptions latines de lAlgerie, 2 vols. (Paris, 192276).
Inscriptiones Christianae urbis Romae septimo saeculo antiquiores, nova serie (Rome,
1922).
Le Blant, E. F. (ed.), Inscriptions chretiennes de la Gaule anterieures au VIIIe si`ecle, 2 vols.
(Paris, 185665).
Marcillet-Jaubert, J., Les Inscriptions dAltava, Publications des Annales de la Faculte
des Lettres n.s. 65 (Aix-en-Provence, 1968).
Merlin, A., Inscriptions latines de la Tunisie (Paris, 1944).
Noy, D., Jewish Inscriptions of Western Europe, vol. 1, Italy (Excluding the City of Rome),
Spain and Gaul (Cambridge, 1993).
Revue archeologique (18811961).
Reynolds, J. M. and J. B. Ward-Perkins, Inscriptions of Roman Tripolitania (Rome,
1952).
sel, A. and J. Sa
sel, Inscriptiones latinae quae in Iugoslavia inter annos MCMII et
Sa
MCMXL repertae et editae sunt, Dissertationes Musei Nationalis Labacensis 25
(Ljubljana, 1986).
Silvagni, A. (ed.), Monumenta epigraphica christiana saeculo XIII antiquiora quae in Italiae
finibus adhuc existant iussu Pii XII pontificis maximi, 4 vols. (Vatican City, 1943).
Supplementum Epigraphicum Graecum (1923 ).
Vives, J., Inscripciones cristianas de la Espana romana y visigoda, Monumenta Hispaniae
sacra, serie patrstica 2 (2nd edn; Barcelona, 1969).
Wessel, C., A. Ferrua, and C. Carletti (eds.), Inscriptiones Graecae Christianae veteres
Occidentis, Inscriptiones Christianae Italiae Subsidia 1 (Bari, 1989).

SECONDARY LITERATURE

Abadie-Reynal, C. and J.-P. Sodini., La Ceramique paleochretienne de Thasos (Aliki,

Delkos, fouilles anciennes), Ecole

francaise dAth`enes Etudes
thasiennes 13
(Athens, 1992).
Abramowski, L., Die Zitate in der Schrift In Defensione Trium Capitulorum der
romischen Diakons Pelagius, Vigiliae Christianae 10 (1956), pp. 16093.
Adams, J. N., Bilingualism and the Latin Language (Cambridge, 2003).
The Regional Diversification of Latin 200 bcad 600 (Cambridge, 2007).
Albertini, E., Ostrakon byzantin de Negrine (Numidie), Cinquantenaire de la Faculte
des Lettres dAlger (18811931) (Algiers, 1932), pp. 5362.
Amory, P., People and Identity in Ostrogothic Italy, 489554, Cambridge Studies in
Medieval Life and Thought, 4th ser. 33 (Cambridge, 1997).
Anselmino, L. and C. Pavolini, Terra sigillata: lucerne, Enciclopedia dellarte antica,
classica e orientale: Atlante delle forme ceramiche, vol. 1 (Rome, 1981), pp. 184207.
Arnheim, M. T. W., The Senatorial Aristocracy in the Later Roman Empire (Oxford,
1972).
Arthur, P., Naples: From Roman Town to City-State, Archaeological Monographs of
the British School at Rome 12 (London, 2002).
Naples: Notes on the Economy of a Dark Age City, in C. Malone and
S. Stoddart (eds.), Papers in Italian Archaeology IV, The Cambridge Conference,

391

Bibliography
vol. 4, Classical and Medieval Archaeology, BAR International Series 246 (Oxford,
1985), pp. 24759.
Azema, Correspondence [= Theodoret of Cyrrhus, Epistolae (q.v.)].
Bagnall, R., The Camel, the Wagon and the Donkey in Later Roman Egypt,
Bulletin of the American Society of Papyrologists 22 (1985), pp. 16.
Conversion and Onomastics: A Reply, Zeitschrift fur Papyrologie und Epigraphik
69 (1987), pp. 24350.
Egypt in Late Antiquity (Princeton, NJ, 1993).
Religious Conversion and Onomastic Change in Early Byzantine Egypt, Bulletin
of the American Society of Papyrologists 19 (1982), pp. 10524.
Bagnall, R., A. Cameron, S. Schwartz, and K. Worp, Consuls of the Later Roman
Empire, Philological Monographs of the American Philological Association 36
(Atlanta, Ga., 1987).
Bagnall, R. and K. Worp, The Chronological Systems of Byzantine Egypt (2nd edn;
Leiden, 2004).
Regnal Formulas in Byzantine Egypt, Bulletin of the American Society of Papyrologists suppl. 2 (Missoula, Mont., 1979).
Balaguer, A., Early Islamic Transitional Gold Issues of North Africa and Spain in
the American Numismatic Society, American Numismatic Society Museum Notes
24 (1979), pp. 22541.
Bales, K., The Social Psychology of Modern Slavery, Scientific American (Apr. 2002),
pp. 808.
Baratte, F., F. Bejaoui, and Z. Ben Abdallah (eds.), Recherches archeologiques a` Hadra

III, Collection de lEcole

francaise de Rome 18/3 (Rome, 2009).

Recherches archeologiques a` Hadra, Miscellanea 2, Collection de lEcole

francaise de
Rome 17/2 (Rome, 1999).
Barker, G. W. W., D. D. Gilbertson, C. O. Hunt, and D. J. Mattingly, Romano
Libyan Agriculture: Integrated Models, in G. W. W. Barker, D. D. Gilbertson,
B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The UNESCO Libyan
Valleys Archaeological Survey, 2 vols. (Paris and London, 1996), 1:26590.
Barnes, T. D., Athanasius and Constantius: Theology and Politics in the Constantinian
Empire (Cambridge, Mass., 1993).
Barnish, S. J. B., Pigs, Plebians and Potentes: Romes Economic Hinterland c.350
600 ad, Papers of the British School at Rome 55 (1987), pp. 15786.
Barraud, D., M. Bonifay, F. Dridi, and J. F. Pichonneau, LIndustrie ceramique de
lantiquite tardive, in H. Ben Hassen and L. Maurin (eds.), Oudhna (Uthina):
La Redecouverte dune ville antique de Tunisie (Bordeaux, 1998), pp. 13967.
Barre, H. Le Culte marial en Afrique apr`es saint Augustin, Revue des etudes augustiniennes 13 (1967), pp. 285317.
Bartoccini, C. R., Scavi e rinvenimenti in Tripolitania negli anni 19261927, Africa
Italiana 2 (1928), pp. 187200.
Bartoccini, R. and D. Mazzoleni, Le inscrizioni del cimitero di En Ngila, Rivista
di archeologia cristiana 53 (1977), pp. 15798.
Bates, Michael, History, Geography, and Numismatics in the First Century of
Islamic Coinage, Revue suisse de numismatique 65 (1986), pp. 23162.
Roman and Early Muslim Coinage in North Africa, in North Africa from Antiquity
to Islam: Papers of a Conference Held at Bristol, October 1994, ed. M. Horton

392

Bibliography
and T. Wiedemann, Centre for Mediterranean Studies (University of Bristol),
Occasional Paper 13 (Bristol, 1995), pp. 1215.
Bejaoui, F., Les Vandales en Afrique: Temoignages archeologiques. Les recentes
decouvertes en Tunisie, in G. Berndt and R. Steinacher (eds.), Das Reich

der Vandalen und seine (Vor-) Geschichten, Osterreichische

Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften 366/Forschungen
zur Geschichte des Mittelalters 13 (Vienna, 2008), pp. 197212.
Nouvelles donnees archeologiques a` Sbetla, Africa 14 (1996), pp. 3763.
Benabou, M., La Resistance africaine a` la romanisation (Paris, 1976).
Berndt, G., Konflikt und Anpassung: Studien zu Migration und Ethnogenese der Vandalen,
Historische Studien 489 (Husum, 2007).
Berndt, G. and R. Steinacher (eds.), Das Reich der Vandalen und seine (Vor-)

Geschichten, Osterreichische
Akademie der Wissenschaften, Philosophischhistorische Klasse, Denkschriften 366/Forschungen zur Geschichte des Mittelalters 13 (Vienna, 2008).
Beschaouch, A., Communication sur la localisation dAbitina, la cite des cel`ebres
martyrs africains, Comptes rendus des seances de lAcademie des inscriptions et belleslettres (1976), pp. 25566.
Besevliev, V., Untersuchungen u ber die Personennamen bei den Thrakern (Amsterdam,
1970).
Bickerman, E. J., Chronology of the Ancient World (2nd edn; Ithaca, NY, 1980).
Bischoff, B. and M. Lapidge (eds.), Biblical Commentaries from the Canterbury School of
Theodore and Hadrian, Cambridge Studies in Anglo-Saxon England 10 (Cambridge, 1994).
Blockley, R. C., East Roman Foreign Policy: Formation and Conduct from Diocletian to
Anastasius, ARCA Classical and Medieval Texts, Papers and Monographs 30
(Leeds, 1992).

Bonifay, M., Etudes

sur la ceramique romaine tardive dAfrique, BAR International Series
1301 (Oxford, 2004).
Observations sur les amphores tardives a` Marseille dapr`es les fouilles de la Bourse
(19801984), Revue archeologique de Narbonnaise 19 (1986), pp. 269305.
Bonifay, M. and D. Pietri, Amphores du Ve au VIIe s. a` Marseille: nouvelles donnees
sur la typologie et le contenu, Journal of Roman Archaeology 8 (1995), pp. 94120.
Bonnal, J.-P. and P.-A. Fevrier, Ostraka de la region de Bir Trouch, Bulletin
darcheologie algerienne 2 (19667), pp. 23949.
Bouhot, J.-P., La Transmission dHippon a` Rome des uvres de saint Augustin,
in D. Nebbiai-Dalla Guarda and J.-F. Genest (eds.), Du copiste au collectionneur: melanges dhistoire des textes et des biblioth`eques en lhonneur dAndre Vernet,
Bibliologia Elementa ad Librorum Studia Pertinentia 18 (Turnhout, 1998),
pp. 2333.
Bovill, E. W., The Golden Trade of the Moors: West African Kingdoms in the Fourteenth
Century (London, 1958).
Bowersock, G. W., Hellenism in Late Antiquity (Ann Arbor, Mich., 1990).
Bowersock, G. W., P. Brown, and O. Grabar (eds.), Late Antiquity: A Guide to the
Postclassical World (Cambridge, Mass., 1999).
Boyle, L., The Basilicanus of Hilary Revisited, in E. Condello and G. De
Gregorio (eds.), Scribi e colofoni: le sottoscrizioni di copisti dalle origini allavvento

393

Bibliography
della stampa, Biblioteca del Centro per il collegamento degli studi medievali e
umanistici in Umbria 14 (Spoleto, 1995), pp. 93105.
Brather, S., Kleidung, Grab und Identitat in Spatantike und Fruhmittelalter,
in G. Berndt and R. Steinacher (eds.), Das Reich der Vandalen und seine

(Vor-)Geschichten,
Osterreichische
Akademie
der
Wissenschaften,
Philosophisch-historische Klasse, Denkschriften 366/Forschungen zur
Geschichte des Mittelalters 13 (Vienna, 2008), pp. 28394.
Braund, D., Rome and the Friendly King: The Character of the Client Kingship (London,
1984).
Brett, M., The Arab Conquest and the Rise of Islam in North Africa, in
J. D. Fage and R. Oliver (eds.), The Cambridge History of Africa, 8 vols.
(Cambridge, 197586), 2:490555.
Brett, M. and E. Fentress, The Berbers (Oxford, 1996).
Brock, S., An Early Syriac Life of Maximus the Confessor, Analecta Bollandiana 91
(1973), pp. 299346.
Brogan, O. and D. J. Smith, Ghirza: A Libyan Settlement in the Roman Period (Tripoli,
1984).
Broise, H. and Y. Thebert, Recherches archeologiques franco-tunisiennes a` Bulla Regia II/1:

les thermes Memmiens, Collection de lEcole

francaise de Rome 28/2 (Rome,
1993).
Brown, P., Augustine of Hippo: A Biography (Berkeley, Calif., 1969).
Christianity and Local Culture in Late Roman Africa, JRS 58 (1968), pp. 85
95.
Brown, T. S., Gentlemen and Officers: Imperial Administration and Aristocratic Power in
Byzantine Italy ad 554800 (London, 1984).
Bulliet, R. W., The Camel and the Wheel (Cambridge, Mass., 1975).
Conversion to Islam in the Medieval Period: An Essay in Quantitative History (Cambridge, Mass., 1979).
A Quantitative Approach to Medieval Muslim Biographical Dictionaries, Journal
of the Economic and Social History of the Orient 13 (1970), pp. 195211.
Cagnat, R., A New Roman Customs List, JRS 4 (1914), pp. 1436.
Cameron, A. D. E., The House of Anastasius, Greek, Roman, and Byzantine Studies
19 (1978), pp. 25976.
Cameron, Averil, Byzantine Africa The Literary Evidence, in J. H. Humphrey
(ed.), Excavations at Carthage 1978, Conducted by the University of Michigan, vol. 7
(Ann Arbor, Mich., 1982), pp. 2962.
The Byzantine Reconquest of N. Africa and the Impact of Greek Culture,
Graeco-Arabica 5 (1993), pp. 15365.
Changing Cultures in Early Byzantium (Aldershot, 1996).
Corippus Iohannis: Epic of Byzantine Africa, in F. Cairns, F. Williams, and
S. Cairns (eds.), Papers of the Liverpool Latin Seminar, Fourth Volume, 1983 (Liverpool, 1984), pp. 16780.
Gelimers Laughter: The Case of Byzantine Africa, in F. M. Clover and
R. S. Humphreys (eds.), Tradition and Innovation in Late Antiquity (Madison,
Wis., 1989), pp. 17190.
Procopius and the Sixth Century, Transformation of the Classical Heritage 10 (Berkeley, Calif., 1985).

394

Bibliography
Camps, G., De Masuna a` Koceila: les destinees de la Mauretanie aux VIe au VIIe
si`ecles, Bulletin archeologique du Comite des Travaux Historiques et Scientifiques n.s.
19B (1985), pp. 30725.
Rex gentium Maurorum et Romanorum: recherches sur les royaumes de Mauretanie
des VIe et VIIe si`ecles, Antiquites africaines 20 (1984), pp. 183218.
Carcopino, J., Un empereur maure inconnu dapr`es une inscription latine
recemment decouverte dans lAur`es, Revue des etudes anciennes 46 (1944),
pp. 94120.
Encore Masties, lempereur maure inconnu, Revue africaine 100 (1956), pp. 339
48.
Le Maroc antique (8th edn; Paris, 1943).
Caron, B. and C. Lavoie, Les Recherches canadiennes dans le quartier de la
Rotonde de lOdeon a` Carthage: un ensemble paleochretien des IVe Ve
si`ecles ou une phase doccupation et de construction du VIIIe si`ecle?, Antiquite
tardive 10 (2002), pp. 24961.
Caspar, E., Geschichte des Papsttums von den Anfangen bis zur Hohe der Weltherrschaft,
2 vols. (Tubingen, 19303).
Cesa, M., La politica di Giustiniano verso loccidente nel giudizio di Procopio,
Athenaeum, Studi Periodici di Letteratura e Storia dellAntichit`a n.s. 59 (1981),
pp. 389409.
Chadwick, H., John Moschus and his Friend Sophronius the Sophist, JThS n.s. 25
(1974), pp. 4174.
Chastagnol, A., Les Consulaires de Numidie, in Melanges darcheologie, depigraphie
et dhistoire offerts a` Jerome Carcopino (Paris, 1966), pp. 21528.
Les Espagnols dans laristocratie gouvernementale a` lepoque de Theodose, in
Les Empereurs romains dEspagne (Paris, 1965), pp. 26992.
Les Gouverneurs de Byzac`ene et de Tripolitaine, Antiquites africaines 1 (1967),
pp. 11934.

LItalie et lAfrique au Bas-empire: Etudes

administratives et prosopographiques, Scripta
varia (Lille, 1987).
Les Legats du proconsul dAfrique au Bas-empire, Libyca 6 (1958), pp. 719.
La Prefecture urbaine a` Rome sous le Bas-empire, Publications de la Faculte des lettres
et sciences humaines dAlger 34 (Paris, 1960).
Chastagnol, A. and N. Duval, Les Survivances du culte imperial dans lAfrique du
Nord a` lepoque vandale, in Melanges dhistoire ancienne offerts a` William Seston
(Paris, 1974), pp. 87118.
Chazelle, C. and C. Cubitt (eds.), The Crisis of the Oikoumene: The Three Chapters and the Failed Quest for Unity in the Sixth-century Mediterranean (Turnhout,
2007).
Cherry, D., Frontier and Society in Roman North Africa (Oxford, 1998).
Chiesa, P., Pellegrino martire in urbe Bolitana e Pellegrino di Ancona: unaltra
agiografia africana ad Aquileia?, Analecta Bollandiana 116 (1998), pp. 25
56.
Un testo agiografico africano ad Aquileia: gli Acta di Gallonio e dei martiri di
Timida Regia, Analecta Bollandiana 114 (1996), pp. 24168.
Christides, V., Byzantine Libya and the March of the Arabs towards the West of North
Africa, BAR International Series 851 (Oxford, 2000).

395

Bibliography
Chrysos, E., Zur Datierung und Tendenz der Werke des Facundus von Hermiane,
Kleronomia 1 (1969), pp. 31124.
Ciotola, A., Il materiale ceramico rinvenuto nella ricognizione intorno a Dougga
(campagne 19941996): una sintesi preliminare, in M. de Vos, Rus Africum:
Terra, acqua, olio nellAfrica settentrionale: Scavo e ricognizione nei dintorni di Dougga
(Alto Tell tunisino) (Trent, 2000), pp. 5877.
Classen, P., Kaiserreskript und Konigsurkunde: Diplomatische Studien zum Problem der
Kontinuitat zwischen Altertum und Mittelalter, Vyzantina Keimena kai Meletai 15
(Thessaloniki, 1977).
Claude, D., Adel, Kirche und Konigtum im Westgotenreich, Vortrage und Forschungen
8 (Sigmaringen, 1971).
Clifford, J., The Predicament of Culture: Twentieth-Century Ethnography, Literature, and
Art (Cambridge, Mass., 1988).
Clover, F. M., Carthage and the Vandals, in J. H. Humphrey (ed.), Excavations at
Carthage 1978, Conducted by the University of Michigan, vol. 7 (Ann Arbor, Mich.,
1982), pp. 122.
Carthage in the Age of Augustine, in J. H. Humphrey (ed.), Excavations at
Carthage 1976, Conducted by the University of Michigan, vol. 4 (Ann Arbor, Mich.,
1978), pp. 114.
Le Culte des empereurs dans lAfrique vandale, BCTH ser. 2, 1516 (1984),
pp. 1218.
Emperor Worship in Vandal Africa, in G. Wirth (ed.), Romanitas-Christianitas:
Untersuchungen zur Geschichte und Literatur der romischen Kaiserzeit (Berlin, 1982),
pp. 66174.
The Family and Early Career of Anicius Olybrius, Historia: Zeitschrift fur Alte
Geschichte 27 (1978), pp. 16996.
Felix Karthago, Dumbarton Oaks Papers 40 (1986), pp. 116.
Flavius Merobaudes: A Translation and Historical Commentary, Transactions of the
American Philosophical Society n.s. 61/1 (Philadelphia, Pa., 1971).
A Game of Bluff: The Fate of Sicily after ad 476, Historia: Zeitschrift fur alte
Geschichte 48 (1999), pp. 23544.
Geiseric the Statesman: A Study of Vandal Foreign Policy, Ph.D. dissertation,
University of Chicago (1966).
The Symbiosis of Romans and Vandals in Africa, in E. Chrysos and
A. Schwarcz (eds.), Das Reich und die Barbaren, Veroffentlichungen des Instituts

fur Osterreichische
Geschichtsforschung 29 (Vienna, 1989), pp. 5773.
Timekeeping and Dyarchy in Vandal Africa, Antiquite tardive 11 (2003), pp. 45
63.
Conant, J., Europe and the African Cult of Saints, circa 350900: An Essay in
Mediterranean Communications, Speculum 85 (2010), pp. 146.
Literacy and Private Documentation in Vandal North Africa: The Case of the
Albertini Tablets, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New
Perspectives on Late Antique North Africa (Aldershot, 2004), pp. 199224.
Costanza, S., Uuandali-Arriani e Romani-Catholici nella Historia persecutionis Africanae provinciae di Vittore di Vita. Una controversia per luso del latino
nel concilio cartaginese del 484, in Oikoumene: Studi paleocristiani pubblicati in
onore del Concilio Ecumenico Vaticano II (Catania, 1964), pp. 22341.

396

Bibliography
Courtois, C., Gregoire VII et lAfrique du Nord: Remarques sur les communautes
chretiennes dAfrique au XIe si`ecle, Revue historique (1945), pp. 96122 and
193226.
Reliques carthaginoises et legende carolingienne, Revue de lhistoire des religions
129 (1945), pp. 5783.
Sur un carreau de terre cuite representant saint Pantaleon, Karthago 3 (19512),
pp. 20913.
Les Vandales et lAfrique (Paris, 1955).
Victor de Vita et son uvre: etude critique (Algiers, 1954).
Cross, F. L. and E. A. Livingstone (eds.), The Oxford Dictionary of the Christian Church
(3rd edn; Oxford, 1997).
Cunliffe, B., Europe between the Oceans, 9000 bcad 1000 (New Haven, Conn.,
2008).
Curta, F., (ed.), Borders, Barriers, and Ethnogenesis: Frontiers in Late Antiquity and the
Middle Ages (Turnhout, 2005).
The Making of the Slavs: History and Archaeology of the Lower Danube Region, ca. 500
700, Cambridge Studies in Medieval Life and Thought, 4th ser. 52 (Cambridge,
2001).
Darmon, J.-P., Les Mosaques de la synagogue de Hammam Lif: un reexamen du
dossier, in R. Ling (ed.), Fifth International Colloquium on Ancient Mosaics, Held at
Bath, England, on September 512, 1987, 2 vols., Journal of Roman Archaeology
Supplementary Series 9/Colloque international pour letude de la mosaque
antique 5 (Ann Arbor, Mich., 1995), 2:729.
Dauge, Y. A., Le Barbare: recherches sur la conception romaine de la barbarie et de la
civilisation, Collection Latomus 176 (Brussels, 1981).
Davis, D., Resurrecting the Granary of Rome: Environmental History and French Colonial
Expansion in North Africa (Athens, Ohio, 2007).
de Gaiffier, B., S. Marcel de Tanger ou de Leon? e volution dune legende, Analecta
Bollandiana 61 (1943), pp. 11639.
Les Notices hispaniques dans le martyrologe dUsuard, Analecta Bollandiana 55
(1937), pp. 26883.
de Guibert, J., Saint Victor de Cesaree, Analecta Bollandiana 24 (1905), pp. 25764.
de Ligt, L., Fairs and Markets in the Roman Empire: Economic and Social Aspects of
Periodic Trade in a Pre-Industrial Society, Dutch Monograms on Ancient History
and Archaeology 11 (Amsterdam, 1993), pp. 15596.
de Rossi, G. B., La Roma sotterranea cristiana, 3 vols. (Rome, 1864).
de Vos, M., Rus Africum. Terra, acqua, olio nellAfrica settentrionale: Scavo e ricognizione
nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000).
Dekkers, E., Clavis Patrum Latinorum qua in Corpus Christianorum edendum optimas
quasque scriptorum recensiones a Tertulliano ad Bedam, CCSL (3rd edn; Steenbrugge,
1995).
Delattre, A., Le Culte de la Sainte Vierge en Afrique dapr`es les monuments archeologiques
(Paris, 1907).
Delattre, A. and P. Monceaux, Sceaux byzantins decouverts a` Carthage, Bulletin de
la Societe nationale des Antiquaires de France (1912), pp. 3313.
Delehaye, H., Cyprien dAntioche et Cyprien de Carthage, Analecta Bollandiana
39 (1921), pp. 31432.

397

Bibliography
Les Origines du culte des martyrs, Subsidia Hagiographica 20 (2nd edn; Brussels,
1933).
La Passion de S. Felix de Thibiuca, Analecta Bollandiana 39 (1921), pp. 24176.
Dench, E., RomulusAsylum: Roman Identities from the Age of Alexander to the Age of
Hadrian (Oxford, 2005).
Desanges, J., Catalogue des tribus africaines de lantiquite classique a` louest du Nil (Dakar,
1962).
The Iconography of the Black in Ancient North Africa, in J. Vercoutter (ed.),
The Image of the Black in Western Art, vol. 1, From the Pharaohs to the Fall of the
Roman Empire (Cambridge, Mass., 1976), pp. 24668.
Quelques considerations sur lusage du grec dans les ports de lAfrique romaine,
Graeco-Arabica 6 (1995), pp. 2736.
Un Temoignage peu connu de Procope sur la Numidie vandale et byzantine,
Byzantion 33 (1963), pp. 4169.

Detschew, D., Die thrakischen Sprachreste, Osterreichische

Akademie der Wissenschaften, Philosophisch-historische Klasse, Schriften der Balkankommission,
Linguistische Abteilung 14 (Vienna, 1957).
Devreesse, R., La Fin inedite dune lettre de Saint Maxime: un bapteme force de
Juifs et de Samaritains a` Carthage en 632, Revue des sciences religieuses 17 (1937),
pp. 2535.
di Vita, A., La diffusione del Cristianesimo nellinterno della Tripolitania attraverso i
monumenti e sue sopravvivenze nella Tripolitania araba, Quaderni di Archeologia
della Libia 5 (1967), pp. 12142.
Diehl, C., LAfrique byzantine: histoire de la domination byzantine en Afrique (533709)
(Paris, 1896).
Diesner, H.-J., Die Auswirkungen der Religionspolitik Thrasamunds und Hilderichs auf
Ostgoten und Byzantiner, Sitzungsberichte der Sachsischen Akademie der Wissenschaften zu Leipzig, Philologisch-historische Klasse 113/3 (Berlin, 1967).
Dietz, S., A Summary of the Field Project, in S. Dietz, L. Ladjimi Seba, H. Ben
Hassen, P. rsted, and J. Carlsen (eds.), Africa Proconsularis: Regional Studies of
the Segermes Valley of Northern Tunisia, 3 vols. (Aarhus, 19952000), 2:77199.

Disdier, M.-T., Le Temoignage spiritual de Thalassius le Lybien, Etudes

byzantines
2 (1944), pp. 79118.
Donner, F., The Early Islamic Conquests (Princeton, NJ, 1981).
Dore, J., The UNESCO Libyan Valleys Archaeological Survey Pottery, in G. W.
W. Barker, D. D. Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the
Desert: The UNESCO Libyan Valleys Archaeological Survey, 2 vols. (Paris and
London, 1996), 2:31789.
Dossey, L., Christians and Romans: Aspiration, Assimilation, and Conflict in the
North African Countryside, Ph.D. dissertation, Harvard University (1998).
The Last Days of Vandal Africa: An Arian Commentary on Job and its Historical
Context, JThS n.s. 54 (2003), pp. 60138.
Peasant and Empire in Christian North Africa, Transformation of the Classical Heritage 47 (Berkeley, Calif., 2010).
Downs, J., The Christian Tomb Mosaics from Tabarka: Status and Identity in
a North African Roman Town, Ph.D. dissertation, University of Michigan
(2007).

398

Bibliography
Drinkwater, J. and H. Elton (eds.), Fifth-Century Gaul: A Crisis of Identity? (Cambridge, 1992).
du Coudray La Blanch`ere, R. and P. Gauckler, Catalogue du Musee Alaoui, Catalogue
des Musees et Collections Archeologiques de lAlgerie et de la Tunisie 7 (Paris,
1897).
Dunbabin, K., The Mosaics of Roman North Africa: Studies in Iconography and Patronage
(Oxford, 1978).
Durliat, J., Les Dedicaces douvrages de defense dans lAfrique byzantine, Collection de

lEcole
francaise de Rome 49 (Rome, 1981).
Le Salaire de la paix sociale dans les royaumes barbares (Ve VIe si`ecles), in
H. Wolfram and A. Schwarcz (eds.), Anerkennung und Integration: Zu den
wirtschaftlichen Grundlagen der Volkerwanderungszeit 400600. Berichte des Symposions der Kommission fur Fruhmittelalterforschung, 7. bis 9. Mai 1986, Stift Zwettl,
Niederosterreich, Veroffentlichungen der Kommission fur Fruhmittelalterforschung 11 (Vienna, 1988), pp. 2172.
Les Transferts fonciers apr`es la reconquete byzantine en Afrique et en Italie, in
E. Magnou-Nortier (ed.), Aux sources de la gestion publique, 2 vols. (Lille, 1993
5), 2:89121.

De la ville antique a` la ville byzantine: le probl`eme des subsistances, Collection de lEcole

francaise de Rome 136 (Rome, 1990).
Duval, N., Les Byzantines a` Rusguniae, Bulletin archeologique du Comite des Travaux
archeologiques et historiques n.s. 19B (1983), pp. 34160.

Inscriptions byzantines de Sbetla (Tunisie) III, Melanges de lEcole

francaise de
Rome (Antiquite) 83 (1971), pp. 42343.
Nouvelles recherches darcheologie et depigraphie chretiennes a` Sufetula

(Byzac`ene), Melanges darcheologie et dhistoire de lEcole

francaise de Rome 68
(1956), pp. 27698.
Rapport preliminaire sur les travaux effectues a` Hadra en sept.oct. 1967, Africa
34 (196970), pp. 193224.

Recherches archeologiques a` Hadra, 2 vols., Collection de lEcole

francaise de Rome
18/12 (Rome, 197581).
Duval, N. and J. Mallon, Les Inscriptions de la chapelle vandale a` Hadra dapr`es
labbe Delapard, Bulletin de la Societe nationale des Antiquaires de France (1969),
pp. 99124.
Duval, Y., Loca sanctorum Africae: le culte des martyrs en Afrique du IVe au VIIe si`ecle,

2 vols., Collection de lEcole

francaise de Rome (Paris, 1982).
La Mauretanie Sitifienne a` lepoque byzantine, Latomus 29 (1970), pp. 15761.
Le Patrice Pierre, exarque dAfrique?, Antiquites africaines 5 (1971), pp. 20914.
Eger, C., Vandalisches Trachtzubehor? Zu Herkunft, Verbreitung und Kontext
ausgewahlter Fibeltypen in Nordafrika, in G. Berndt and R. Steinacher

(eds.), Das Reich der Vandalen und seine (Vor-) Geschichten, Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften
366/Forschungen zur Geschichte des Mittelalters 13 (Vienna, 2008), pp. 183
96.
Elton, H., Defining Romans, Barbarians, and the Roman Frontier, in R. Mathisen
and H. Sivan (eds.), Shifting Frontiers in Late Antiquity (Aldershot, 1996),
pp. 12635.

399

Bibliography
Ennabli, L., Carthage: une metropole chretienne du IVe a` la fin du VIIe si`ecle (Paris, 1997).
Evans Grubbs, J., Law and Family in Late Antiquity (Oxford, 1995).
Everett, N., Literacy in Lombard Italy, c.568774, Cambridge Studies in Medieval Life
and Thought, 4th ser. 53 (Cambridge, 2003).
Fabrega Grau, A., Pasionario hispanico (siglos VIIXI), 2 vols., Monumenta hispaniae
sacra, Serie liturgica 6 (Madrid, 19535).
Fanning, S., Emperors and Empires in Fifth-Century Gaul, in J. Drinkwater and
H. Elton (eds.), Fifth Century Gaul: A Crisis of Identity? (Cambridge, 1992).
Fasola, U. M., Le catacombe di S. Gennaro a Capodimonte (Rome, 1975).
Feissel, D., Un Acte de la prefecture dAfrique sur leglise de Byzac`ene au debut du
r`egne de Justin II, Antiquite tardive 11 (2003), pp. 97112.

Les Edifices
de Justinien au temoignage de Procope et de lepigraphie, Antiquite
tardive 8 (2000), pp. 81104.
Fentress, E., A. At Kaci, and N. Bounssair, Prospections dans le Belezma: Rapport Preliminaire, in Actes du colloque international sur lhistoire de Setif, Bulletin
darcheologie algerienne, Sup. 7 (Algiers, 1993), pp. 10727.
Fentress, E., A. Drine, and R. Holod, An Island through Time: Jerba Studies, vol. 1,
The Punic and Roman Periods, Journal of Roman Archaeology Supplementary
Series 71 (Portsmouth, RI, 2009).
Fevrier, P.-A., Differences et conflits: Maures et Barbares, in Approches du Maghreb
romain: pouvoirs, differences et conflits, 2 vols. (Aix-en-Provence, 198990), 2:139
57.
Masuna et Masties, Antiquites africaines 24 (1988), pp. 13347.
Fischer, B., Zur Liturgie der lateinischen Handschriften vom Sinai, Revue
Benedictine 74 (1964), pp. 28497
Fontaine, J., Isidore de Seville et la culture classique dans lEspagne wisigothique, 3 vols.
(2nd edn; Paris, 1983).
Forstemann, E., Altdeutsches Namenbuch, vol. 1, Personennamen (Munich, 1966).
Fournier, E., Victor of Vita and the Vandal Persecution: Interpreting Exile in Late
Antiquity, Ph.D. dissertation, University of California, Santa Barbara (2008).
Franchi de Cavalieri, P., Le Reliquie dei martiri Scillitani, Romische Quartalschrift
fur Christliche Altertumskunde und fur Kirchengeschichte 17 (1903), pp. 20921.
Francovich Onesti, N., I Vandali: Lingua e storia, Lingue e letterature Carocci 14
(Rome, 2002).
Frend, W. H. C., The Christian Period in Mediterranean Africa, c.ad 200 to 700,
in J. D. Fage and R. Oliver (eds.), The Cambridge History of Africa, 8 vols.
(Cambridge, 197586), 2:41089.
Donatist and Catholic: The Organization of Christian Communities in the North
African Countryside, in Cristianizzazione ed organizzazione ecclesiastica delle campagne nellalto medioevo: espansione e resistenze, 2 vols., Settimane di studio del
Centro italiano di studi sullalto medioevo 28 (Spoleto, 1982), 2:60134.
The Donatist Church: A Movement of Protest in Roman North Africa (Oxford, 1952).
From Donatist Opposition to Byzantine Loyalism: The Cult of Martyrs in North
Africa 350650, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New
Perspectives on Late Antique North Africa (Aldershot, 2004), pp. 25969.
Fulford, M. G., Carthage: Overseas Trade and the Political Economy, c.ad 400700,
Reading Medieval Studies 6 (1980), pp. 6680.

400

Bibliography
Garnsey, P. D. A., Romes African Empire under the Principate, in P. D. A. Garnsey
and C. R. Whittaker (eds.), Imperialism in the Ancient World (Cambridge, 1978),
pp. 22354.
Garsoan, N., The Problem of Armenian Integration into the Byzantine Empire, in
H. Ahrweiler and A. Laiou (eds.), Studies on the Internal Diaspora of the Byzantine
Empire (Washington, DC, 1998), pp. 53124.
Gauckler, P., L. Poinssot, A. Merlin, L. Drappier, and L. Hautcoeur, Catalogue du
Musee Alaoui, Supplement, Catalogue des Musees et Collections Archeologiques
de lAlgerie et de la Tunisie 15/1 (Paris, 1910).
Geary, P. J., Before France and Germany: The Creation and Transformation of the Merovingian World (New York, 1988).
Geertz, C., The Interpretation of Cultures (New York, 1973).
Gil Egea, M. E., Africa en tiempos de los vandalos: continuidad y mutaciones de las estructuras
sociopolticas romanas, Memorias del Seminario de Historia Antigua 7 (Alcala de
Henares, 1998).
Gilbertson, D. D., Explanations: Environment as Agency, in G. W. W. Barker,
D. D. Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The
UNESCO Libyan Valleys Archaeological Survey, 2 vols. (Paris and London, 1996),
1:291317.
Gilbertson D. D. and C. O. Hunt, RomanoLibyan Agriculture: Walls and Floodwater Farming, in G. W. W. Barker, D. D. Gilbertson, B. Jones, and D. J.
Mattingly (eds.), Farming the Desert: The UNESCO Libyan Valleys Archaeological
Survey, 2 vols. (Paris and London, 1996), 1:191225.
Gillett, A., Envoys and Political Communication in the Late Antique West, 411533,
Cambridge Studies in Medieval Life and Thought, 4th ser., 55 (Cambridge,
2003).
Was Ethnicity Politicized in the Earliest Medieval Kingdoms?, in A. Gillett (ed.),
On Barbarian Identity: Critical Approaches to Ethnicity in the Early Middle Ages,
Studies in the Early Middle Ages 4 (Turnhout, 2002), pp. 85121.
Gilliam, J. F., Romanization of the Greek East: The Role of the Army, Bulletin of
the American Society of Papyrologists 2 (1965), pp. 6573.
Giroire, C. and D. Roger, Roman Art from the Louvre (New York, 2007).
Goffart, W., Barbarian Tides: The Migration Age and the Later Roman Empire (Philadelphia, Pa., 2006).
Barbarians and Romans, ad 418584: The Techniques of Accommodation (Princeton,
NJ, 1980).
Rome, Constantinople, and the Barbarians, Romes Fall and After (London, 1989),
pp. 132.
Goitein, S. D., A Mediterranean Society: The Jewish Communities of the Arab World as
Portrayed in the Documents of the Cairo Geniza, 6 vols. (Berkeley, Calif., 196793).
Gonzalez Salinero, R., Poder y conflicto religioso en el norte de Africa: Quodvultdeus de
Cartago y los vandalos, Graeco-Romanae Religionis Electa Collectio (GREC)
10 (Madrid, 2002).
Goodchild, R. G., La necropoli Romano-Libica di Bir ed-Dreder, Quaderni di
Archeologia della Libia 3 (1954), pp. 91107.
Gorman, M., Chapter Headings for St Augustines De Genesi ad litteram, Revue des
etudes augustiniennes 26 (1980), pp. 88104.

401

Bibliography
The Manuscript Traditions of St Augustines Major Works, in V. Grossi (ed.), Atti
del Congresso internazionale su S. Agostino nel XVI Centenario della Conversione,
3 vols., Studia Ephemeridis Augustinianum 24 (Rome, 1987), 1:381412.
The Manuscript Traditions of the Works of Augustine, Millennio Medievale 27 (Florence, 2001).
A Survey of the Oldest Manuscripts of St Augustines De ciuitate dei, JThS n.s.
33 (1982), pp. 398410.
Graham, S. R., The Dissemination of North African Christian and Intellectual
Culture in Late Antiquity, Ph.D. dissertation, University of California, Los
Angeles (2005).
Greatrex, G., Roman Identity in the Sixth Century, in S. Mitchell and G. Greatrex
(eds.), Ethnicity and Culture in Late Antiquity (London, 2000), pp. 26792.
Green, W., Augustines Use of Punic, in W. Fischel (ed.), Semitic and Oriental Studies:
A Volume Presented to William Popper, Professor of Semitic Languages, Emeritus, on
the Occasion of his Seventy-Fifth Birthday, October 29, 1949, University of California
Publications in Semitic Philology 11 (Berkeley, Calif., 1951), pp. 17990.
A Fourth Century Manuscript of Saint Augustine?, Revue Benedictine 69 (1959),
pp. 1917.
Greene, J. A, Une Reconnaissance archeologique dans larri`ere-pays de la Carthage
antique, in A. Ennabli (ed.), Pour sauver Carthage: exploration et conservation de la
cite punique, romaine et byzantine (Tunis, 1992), pp. 1957.
Gribomont, J., Le Mysterieux Calendrier latin du Sina: e dition et commentaire,
Analecta Bollandiana 75 (1957), pp. 10534.
Grierson, P., Mathasuntha or Mastinas: A Reattribution, Numismatic Chronicle
(1959), pp. 11930.
The Tablettes Albertini and the Value of the Solidus in the Fifth and Sixth Centuries
ad JRS 49 (1959), pp. 7380.
Gualandi, G., La presenza cristiana nellIfriqiya. Larea cimiteriale dEn-Ngila
(Tripoli), Felix Ravenna 1056 (1973), pp. 25779.
Guery, R., Survivance de la vie sedentaire pendant les invasions arabes en Tunisie
centrale: lexemple de Rougga, Bulletin archeologique du Comite des Travaux
archeologiques et historiques n.s. 19B (1983), pp. 399410.
Guery, R., C. Morrisson, and H. Slim, Recherches archeologiques franco-tunisiennes a`

Rouga III: le tresor de monnaies dor byzantine, Collection de lEcole

francaise de
Rome 60 (Rome, 1982).
Hahn, W., Moneta Imperii Byzantini: Rekonstruktion des Prageaufbaues auf synoptischtabellarischer Grundlage, vol. 1, Veroffentlichungen der Numismatischen Kom
mission 1/Osterreichische
Akademie der Wissenschaften, Philosophischhistorische Klasse, Denkschriften 109 (Vienna, 1973).
Haldon, J., Byzantium in the Seventh Century: The Transformation of a Culture
(Cambridge, 1990).
The Idea of the Town in the Byzantine Empire, in G. P. Brogiolo and B. WardPerkins (eds.), The Idea and Ideal of the Town between Late Antiquity and the Early
Middle Ages (Leiden, 1999), pp. 123.

Recruitment and Conscription in the Byzantine Army c.550950, Osterreichische

Akademie der Wissenschaften, Philosophisch-historische Klasse, Sitzungsberichte 357 (Vienna, 1979).
Warfare, State and Society in the Byzantine World, 5651204 (London, 1999).

402

Bibliography
Halkin, F. (ed.), Bibliotheca hagiographica graeca, 3 vols., Subsidia Hagiographica 8a
(3rd edn; Brussels, 1957; repr. Wetteren, 1986).
Halsall, G., Gender and the End of Empire, Journal of Medieval and Early Modern
Studies 34 (2004), pp. 1739.
Warfare and Society in the Barbarian West, 450900 (New York, 2003).
Handley, M., Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain,
ad 300750, BAR International Series 1135 (Oxford, 2003).
Disputing the End of African Christianity, in A. H. Merrills (ed.), Vandals,
Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot,
2004), pp. 291310.
Dying on Foreign Shores: Travel and Mobility in the Late Antique West, Journal of
Roman Archaeology Supplementary Series 86 (Portsmouth, RI, 2011).
Inscribing Time and Identity in the Kingdom of Burgundy, in S. Mitchell
and G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London, 2000),
pp. 83102.
Hanson, R. P. C., The Search for the Christian Doctrine of God: The Arian Controversy,
318381 (Edinburgh, 1988).
Harper, K., The New Coordinates of the Slave Trade in Late Antiquity: Africa,
the Mediterranean, and the Near East, Paper delivered to the 124th Annual
Meeting of the American Historical Association, 9 Jan. 2010 (San Diego,
Calif.).
Slavery in the Late Roman World, ad 275425 (Cambridge, 2011).
Harries, J., Sidonius Apollinaris and the Fall of Rome, ad 407485 (Oxford, 1994).
Harris, M., Theories of Culture in Postmodern Times (Walnut Creek, Calif., 1999).
Haussleiter, J., Leben und Werke Bischofs Primasius von Hadrumetum: Eine Untersuchung
(Erlangen, 1887).
Hayes, J. W., Excavations at Sarachane in Istanbul, vol. 2 (Princeton, NJ, 1992).
Late Roman Pottery (London, 1972).
Pottery Report 1976, in J. H. Humphrey (ed.), Excavations at Carthage 1976,
Conducted by the University of Michigan, vol. 4 (Ann Arbor, Mich., 1978),
pp. 2398.
Probl`emes de la ceramique des VIIe` me IXe` me si`ecles a Salamine et a Chypre, in
Salamine de Chypre: histoire et archeologie: etat des recherches: Lyon, 1317 mars 1978
(Paris, 1980), pp. 3758.
A Supplement to Late Roman Pottery (London, 1980).
Hays, G., The Date and Identity of the Mythographer Fulgentius, Journal of Medieval
Latin 13 (2003), pp. 163252.
Romuleis Libicisque Litteris: Fulgentius and the Vandal Renaissance , in
A. H. Merrills (ed.), Vandals, Romans, and Berbers: New Perspectives on Late
Antique North Africa (Aldershot, 2004), pp. 10132.
Haywood, R. M., Roman Africa, in T. Frank (ed.), An Economic Survey of Ancient
Rome, vol. 4 (Baltimore, Md., 1938), pp. 1119.
Heather, P., The Barbarian in Late Antiquity: Image, Reality, and Transformation, in R. Miles (ed.), Constructing Identities in Late Antiquity (London, 1999),
pp. 23458.
The Fall of the Roman Empire: A New History of Rome and the Barbarians (Oxford,
2006).
The Goths (Oxford, 1996).

403

Bibliography
Goths and Romans, 332489 (Oxford, 1991).
(ed.), The Visigoths from the Migration Period to the Seventh Century: An Ethnographic Perspective, Studies in Historical Archaeoethnology 4 (Woodbridge,
1999).
The Western Empire, 425476, in A. Cameron, B. Ward-Perkins, and M. Whitby
(eds.), Late Antiquity: Empire and Successors, ad 425600, Cambridge Ancient
History 14 (Cambridge, 2000), pp. 132.
Helm, R., Untersuchungen u ber den auswartigen diplomatischen Verkehr des
romischen Reiches im Zeitalter der Spatantike, Archiv fur Urkundenforschung
12 (19312), pp. 375436.
Hitchner, R. B., Historical Text and Archaeological Context in Roman North
Africa: The Albertini Tablets and the Kasserine Survey, in D. Small (ed.),
Methods in the Mediterranean: Historical and Archaeological Views on Texts and
Archaeology, Mnemosyne suppl. 135 (Leiden, 1995), pp. 12442.
The Kasserine Archeological Survey 19821986 (University of Virginia, USA
Institut national darcheologie et dart de Tunisie), Antiquites africaines 24
(1988), pp. 741.
The Kasserine Archaeological Survey 1987, Antiquites africaines 26 (1990),
pp. 23159.
Holum, K., Theodosian Empresses: Women and Imperial Dominion in Late Antiquity,
Transformation of the Classical Heritage 3 (Berkeley, Calif., 1982).
Horden, P. and N. Purcell., The Corrupting Sea: A Study of Mediterranean History
(Oxford, 2000).
Howe, T., Vandalen, Barbaren, und Arianer bei Victor von Vita, Studien zur Alten
Geschichte 7 (Frankfurt am Main, 2007).
Hubert, J., J. Porcher, and W. F. Volbach, The Carolingian Renaissance (New York,
1970).
Humbert, M., Enfants a` louer ou a` vendre: Augustin et lautorite parentale (Epist.
10* et 24*), in Les lettres de saint Augustin decouvertes par Johannes Divjak.
Communications presentees au colloque des 20 et 21 Septembre 1982 (Paris, 1983),
pp. 189204.
Iancu, C., and J.-M. Lass`ere (eds.), Juifs et judasme en Afrique du Nord dans lantiquite
et le haut moyen-age (Montpellier, 1985).
Icard, F., Sceaux et medailles de plomb trouves a` Carthage, BCTH (1917),
pp. 118.
Sceaux et plombs marques decouverts a` Carthage, BCTH (1927), pp. 47987.
Innes, M., Land, Freedom and the Making of the Medieval West, Transactions of the
Royal Historical Society 16 (2006), pp. 3974.
Isaac, B., The Meaning of the Terms Limes and Limitanei, JRS 78 (1988),
pp. 12547.
Isola, A. (trans.), Vita di San Fulgenzio, Collana di testi patristici 65 (Rome, 1987).
James, E., Europes Barbarians, ad 200600 (New York, 2009).
The Franks (Oxford, 1988).
Jeffreys, E., B. Croke, and R. Scott (eds.), Studies in John Malalas, Byzantina Australiensia 6 (Sydney, 1990).
Jirecek, C., Archaologische Fragmente aus Bulgarien, Archaeologisch-epigraphische
Mittheilungen aus Oesterreich-Ungarn 10 (1886), pp. 43104 and 129209.

404

Bibliography
Jones, A. H. M., The Later Roman Empire, 284602: A Social, Economic and Administrative Survey, 2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986).
Were the Ancient Heresies National or Social Movements in Disguise?, JThS
n.s. 10 (1959), pp. 28098.
Jongeling, K., North-African Names from Latin Sources (Leiden, 1994).
Jongeling, K. and R. Kerr (eds.), Late Punic Epigraphy: An Introduction to the Study of
Neo-Punic and Latino-Punic Inscriptions (Tubingen, 2005).
Justi, F., Iranisches Namenbuch (Marburg, 1895; repr. Hildesheim, 1963).
Kaegi, W., Arianism and the Byzantine Army in Africa 533546, Traditio 21 (1965),
pp. 2353.
Byzantium and the Early Islamic Conquests (Cambridge, 1992).
Byzantine Military Unrest, 471843: An Interpretation (Amsterdam, 1981).
Heraclius, Emperor of Byzantium (Cambridge, 2003).
Muslim Expansion and Byzantine Collapse in North Africa (Cambridge, 2010).
Society and Institutions in Byzantine Africa, in P. Corrias and S. Cosentino (eds.),
Ai confini dellimpero. Storia, arte e archeologia della Sardegna bizantina (Cagliari,
2002), pp. 1528.
Kajanto, I., Onomastic Studies in the Early Christian Inscriptions of Rome and Carthage,
Acta Instituti Romani Finlandiae 2/1 (Helsinki, 1963).
Peculiarities of Latin Nomenclature in North Africa, Philologus 108 (1964),
pp. 31012.
Kaldellis, A., Procopius of Caesarea: Tyranny, History, and Philosophy at the End of
Antiquity (Philadelphia, Pa., 2004).
Kamesar, A., The Virgin of Isaiah 7:14: The Philological Argument from the Second
to the Fifth Century, JThS n.s. 41 (1990), pp. 5175.
Kapitan, G., The Church Wreck off Marzamemi, Archaeology 22 (1969), pp. 12233.
Schiffsfrachten antiker Baugesteine und Architekturteile vor den Kusten Ostsiziliens, Klio 39 (1961), pp. 276318.
Kaster, R., Guardians of Language: The Grammarian and Society in Late Antiquity,
Transformation of the Classical Heritage 11 (Berkeley, Calif., 1988).
Keay, S. J., Late Roman Amphorae in the Western Mediterranean. A Typology and Economic
Study: The Catalan Evidence, 2 vols., BAR International Series 196 (Oxford,
1984).
Keenan, J, Evidence for the Byzantine Army in the Syene Papyri, Bulletin of the
American Society of Papyrologists 27 (1990), pp. 13950.
Kehoe, D. P., The Economics of Agriculture on Roman Imperial Estates in North Africa,
Hyptomnemata: Untersuchungen zur Antike und zu ihrem Nachleben 89
(Gottingen, 1988).
Keil, H., De grammaticis quibusdam latinis infimae aetatis commentatio (Erlangen, 1868).
Kelly, C., Ruling the Later Roman Empire, Revealing Antiquity 15 (Cambridge, Mass.,
2004).
Kennell, S. A. H., Magnus Felix Ennodius: A Gentleman of the Church (Ann Arbor,
Mich., 2000).
Kery, L., Canonical Collections of the Early Middle Ages (c.4001140): A Bibliographical
Guide to the Manuscripts and Literature (Washington, DC, 1999).
Kingsley, S., The Economic Impact of the Palestinian Wine Trade in Late Antiquity, in S. Kingsley and M. Decker (eds.), Economy and Exchange in the East

405

Bibliography
Mediterranean during Late Antiquity, Proceedings of a Conference at Somerville College,
Oxford, 29th May, 1999 (Oxford, 2001), pp. 4468.
Kleemann, J., Quelques reflexions sur linterpretation ethnique des sepultures
habillees considerees comme vandales, Antiquite tardive 10 (2002), pp. 123
9.
Klingshirn, W. E., Caesarius of Arles: The Making of a Christian Community in Late
Antique Gaul, Cambridge Studies in Medieval Life and Thought, 4th ser. 22
(Cambridge, 1994).
Koenig, G., Wandalische Grabfunde des 5. und 6. Jhs, Madrider Mitteilungen 22
(1981), pp. 299360.
Kolendo, J., Le Colonat en Afrique sous le Haut-empire, Annales Litteraires de
lUniversite de Besancon 447 (2nd edn; Paris, 1991).
Krahmalkov, C., A Phoenician-Punic Grammar, Handbook of Oriental Studies/Handbuch der Orientalistik, Section One, The Near and Middle East 54
(Leiden, 2001).
Kulikowski, M., Late Roman Spain and its Cities (Baltimore, Md., 2004).
Lackner, W., Westliche Heilige des 5. und 6. Jahrhunderts im Synaxarium Ecclesiae Constantinopolitanae, Jahrbuch der osterreichischen Byzantinistik 19 (1970),
pp. 185202.
Ladner, G., On Roman Attitudes toward Barbarians in Late Antiquity, Viator 7
(1976), pp. 126.
Lambert, D., The Barbarians in Salvians De Gubernatione Dei, in S. Mitchell and
G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London, 2000),
pp. 10315.
The Uses of Decay: History in Salvians De Gubernatione Dei, Augustinian Studies
30 (1999), pp. 11530.
Lancel, S., Tipasitana II: une inscription martyrologique de Tipasa, Bulletin
archeologique algerienne 2 (19667), pp. 2519.
Langlois, P., Peut-on dater Reposianus (Anth. Lat. 253, Riese)?, Revue de philologie,
de litterature et dhistoire anciennes 47 (1973), pp. 30914.
Lanzoni, F., Le diocesi dItalia dalle origini al principio del secolo VII (an. 604), Studio
critico, 2 vols. (2nd edn; Faenza, 1927).
Laporte, J.-P., Zabi, Friki: notes sur la Mauretanie et la Numidie de Justinien,
Antiquite tardive 10 (2003), pp. 15167.
Lass`ere, J.-M., Diffusion et persistance des traditions latines dans le Maghreb
medieval, in La latinite, hier, aujourdhui, demain (Avignon, 1981), pp. 27785.
Laurence, R. and J. Berry (eds.), Cultural Identity in the Roman Empire (London,
1998).
Laurent, V., Une Effigie inedite de Saint Augustin sur la sceau du duc byzantin de
Numidie Pierre, Cahiers de Byrsa 2 (1952), p. 8.
Les Sceaux byzantins du Medaillier Vatican, Medagliere della Biblioteca vaticana 1
(Vatican City, 1962).
Lee, A. D., The Eastern Empire: Theodosius to Anastasius, in A. Cameron,
B. Ward-Perkins, and M. Whitby (eds.), Late Antiquity: Empire and Successors,
ad 425600, Cambridge Ancient History 14 (Cambridge, 2000), pp. 3362.
War in Late Antiquity: A Social History (Oxford, 2007).

406

Bibliography
Leone, A., Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest,
Studi storici sulla Tarda Antichit`a 28 (Bari, 2007).
Linumazione in spazio urbano a Cartagine tra V e VII secolo d. C, Antiquite
tardive 10 (2002), pp. 23348.
Leone, A. and D. J. Mattingly, Vandal, Byzantine and Arab Rural Landscapes in
North Africa, in N. Christie (ed.), Landscapes of Change: Rural Evolutions in
Late Antiquity and the Early Middle Ages (Aldershot, 2004), pp. 13562.
Lepelley, C., Les Cites de lAfrique romaine au Bas-empire, 2 vols. (Paris, 197981).
Temoignages de saint Augustin sur lampleur et les limites de lusage de la langue
punique dans lAfrique de son temps, in C. Briand-Ponsart (ed.), Identites et
culture dans lAlgerie antique, Publications de lUniversite de Rouen 377 (Rouen,
2005), pp. 12754.
Leuthold, E., Jr., Due rare testimonianze della prima monetazione musulmane a
Cartagine, Rivista italiana di numismatica e scienze affini 69 (1967), pp. 939.
Leveau, P., Caesarea de Mauretanie: une ville romaine et ses campagnes, Collection de

lEcole
francaise de Rome 70 (Rome, 1984).
Recherches historiques sur une region montagneuse de Mauretanie Cesarienne:

des Tigava Castra a` la mer, Melanges de lEcole

francaise de Rome, Antiquite 89
(1977), pp. 257311.
Levtzion, N. (ed.), Conversion to Islam (New York, 1979).
Lewicki, T., The Ibadites in North Africa and the Sudan to the Fourteenth Century,
Journal of Western History/Cahiers dhistoire mondiale 13 (1971), pp. 83130.
Leyser, C., A Wall Protecting the City: Conflict and Authority in the Life of
Fulgentius of Ruspe, in A. Camplani and G. Filoramo (eds.), Foundations of
Power and Conflicts of Authority in Late-Antique Monasticism, Proceedings of the
International Seminar, Turin, December 24, 2004, Orientalia Lovaniensia Analecta
157 (Louvain, 2007), pp. 17592.
Liebeschuetz, J. H. W. G., Gens into Regnum: The Vandals, in H.-W. Goetz, J.
Jarnut, and W. Pohl (eds.), Regna et Gentes: The Relationship between Late Antique
and Early Medieval Peoples and Kingdoms in the Transformation of the Roman World,
Transformation of the Roman World 13 (Leiden, 2003), pp. 5583.
Llewellyn, P. A. B., The Roman Church during the Laurentian Schism: Priests and
Senators, Church History 45 (1976), pp. 41727.
The Roman Clergy during the Laurentian Schism (488506): A Preliminary
Analysis, Ancient Society 8 (1977), pp. 24575.
Lowe, E. A., Codices Latini Antiquiores: A Palaeographical Guide to Latin Manuscripts
Prior to the Ninth Century, 11 vols. (Oxford, 193471).
Two New Latin Liturgical Fragments on Mount Sinai, Revue Benedictine 74
(1964), pp. 25283.
Two Other Unknown Latin Liturgical Fragments on Mount Sinai, Scriptorium
19 (1965), pp. 329.
An Unknown Latin Psalter on Mount Sinai, Scriptorium 9 (1955), pp. 17799.
Lund, J., A Synagogue at Carthage? Menorah-Lamps from the Danish Excavations,
Journal of Roman Archaeology 8 (1995), pp. 24562.
Maas, M., Exegesis and Empire in the Early Byzantine Mediterranean, Studien und Texte
zu Antike und Christentum 17 (Tubingen, 2003).

407

Bibliography
MacCormack, S., Art and Ceremony in Late Antiquity, Transformation of the Classical
Heritage 1 (Berkeley, Calif., 1981).
McCormick, M., Bateaux de vie, bateaux de mort: maladie, commerce, transports
annonaires et le passage e conomique du Bas-empire au moyen a ge, in Morfologie
sociali e culturali in Europa fra tarda antichit`a e alto medioevo, Settimane di studio
del Centro italiano di studi sullalto medioevo 45 (Spoleto, 1998), pp. 35122.
Clovis at Tours: Byzantine Public Ritual and the Origins of Medieval Ruler
Symbolism, in E. Chrysos and A. Schwarcz (eds.), Das Reich und die Barbaren,

Veroffentlichungen des Instituts fur Osterreichische

Geschichtsforschung 29
(Vienna, 1989), pp. 15580.
Emperor and Court, in A. Cameron, B. Ward-Perkins, and M. Whitby (eds.),
Late Antiquity: Empire and Successors, ad 425600, Cambridge Ancient History
14 (Cambridge, 2000), pp. 13563.
Eternal Victory: Triumphal Rulership in Late Antiquity, Byzantium and the Early
Medieval West (Cambridge, 1986).
The Imperial Edge: Italo-Byzantine Identity, Movement and Integration,
ad 650950, in H. Ahrweiler and A. Laiou (eds.), Studies on the Internal Diaspora
of the Byzantine Empire (Washington, DC, 1998), pp. 1752.
Origins of the European Economy: Communications and Commerce, ad 300900 (Cambridge, 2001).
McCulloh, J., The Cult of Relics in the Letters and Dialoguesof Pope Gregory
the Great: A Lexicographical Study, Traditio 32 (1976), pp. 14584.
Macias i Sole, J. M. and J. A. Remol`a Vallverdu, Tarraco visigoda: caracterizacion
del material ceramico del siglo VII d. C, in J. M. Gurt and N. Tena (eds.),
V Reunio dArqueologia Cristiana Hisp`anica [= V Reunion de Arqueologa Cristiana
Hispanica], Monografies de la Seccio Hist`orico-Arqueol`ogica 7 (Barcelona,
2000), pp. 48597.
Mackensen, M., Die spatantiken Sigillata- und Lampentopfereien von El Mahrine (Nordtunesien): Studien zur Nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts,
Munchner Beitrage zur Vor- und Fruhgeschichte 50, 2 vols. (Munich, 1993).
Mackensen, M. and G. Schneider, Production Centres of African Red Slip Ware
(3rd7th c.), in Northern and Central Tunisia: Archaeological Provenance and
Reference Groups based on Chemical Analysis, Journal of Roman Archaeology
15 (2002), pp. 12158.
McLynn, N., From Palladius to Maximinus: Passing the Arian Torch, Journal of
Early Christian Studies 4 (1996), pp. 47793.
MacMullen, R., The Epigraphic Habit in the Roman Empire, American Journal of
Philology 103 (1982), pp. 23346.
Romanization in the Time of Augustus (New Haven, Conn., 2000).
Mahjoubi, A., Nouveau temoignage e pigraphique sur la communaute chretienne
de Kairouan au XIe si`ecle, Africa 1 (1966), pp. 8796.
Recherches dhistoire et darcheologie a` Henchir el-Faouar (Tunisie): la cite des Belalitani
Maiores, Publications de lUniversite de Tunis, Faculte des lettres et sciences
humaines, ser. 1, Archeologie histoire 12 (Tunis, 1978).

Maier, J. L., LEpiscopat

de lAfrique romaine, vandale et byzantine, Bibliotheca Helvetica
Romana 11 (Rome, 1973).

408

Bibliography
Mallardo, D., Il calendario marmoreo di Napoli, Biblioteca Ephemerides Liturgicae 18
(Rome, 1947).
Manitius, M., Geschichte der lateinischen Literatur des Mittelalters, 3 vols., Handbuch
der Altertumswissenschaft, 9. Abt., 2 (Munich, 1911).
Marichal, R., Les Ostraca de Bu Njem, Supplements de Libya Antiqua 7 (Tripoli,
1992).
Markey, T., Germanic in the Mediterranean: Lombards, Vandals, and Visigoths,
in F. M. Clover and R. S. Humphreys (eds.), Tradition and Innovation in Late
Antiquity (Madison, Wis., 1989), pp. 5171.
Markus, R. A., Christianity and Dissent in Roman North Africa: Changing Perspectives in Recent Work, in D. Baker (ed.), Schism, Heresy
and Religious Protest, Studies in Church History 9 (Cambridge, 1972),
pp. 2136.
Country Bishops in Byzantine Africa, in D. Baker (ed.), The Church in Town and
Countryside, Studies in Church History 16 (Oxford, 1979), pp. 115.
Donatism: The Last Phase, in C. W. Dugmore and C. Duggan (eds.), Papers Read
at the First Winter and Summer Meetings of the Ecclesiastical History Society, Studies
in Church History 1 (Leiden, 1964), pp. 11826.
Reflections on Religious Dissent in North Africa in the Byzantine Period, in
G. J. Cuming (ed.), Papers Read at the Third Winter and Summer Meetings of
the Ecclesiastical History Society, Studies in Church History 3 (Leiden, 1966),
pp. 1409.
The Problem of Donatism in the Sixth Century, in Gregorio Magno e il suo

tempo: XIX Incontro di studiosi dellantichit`a cristiana in collaborazione con lEcole

francaise de Rome,Roma, 912 maggio 1990, Studia ephemeridis Augustinianum
33, 2 vols. (Rome, 1991), 1:15966.
Marrou, H.-I., Histoire de leducation dans lantiquite (7th edn; Paris, 1971).

Maspero, J., Organisation militaire de lEgypte

byzantine, Biblioth`eque de lEcole
des
hautes e tudes, Sciences historiques et philologiques 201 (Paris, 1912).
Mathisen, R. W., Barbarian Bishops and the Churches in Barbaricis Gentibus
during Late Antiquity, Speculum 72 (1997), pp. 66497.
The Ecclesiastical Aristocracy of Fifth-Century Gaul: A Regional Analysis of
Family Structure, Ph.D. dissertation, University of Wisconsin (1979).
Roman Aristocrats in Barbarian Gaul: Strategies for Survival in an Age of Transition
(Austin, Tex., 1993).
Matthews, J., Western Aristocracies and Imperial Court, ad 364425 (Oxford, 1990).
Mattingly, D. J., Explanations: People as Agency, in G. W. W. Barker, D. D. Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The UNESCO
Libyan Valleys Archaeological Survey, 2 vols. (Paris and London, 1996), 1:31942.
The Laguatan: A Libyan Tribal Confederation in the Late Roman Empire, Libyan
Studies 14 (1983), pp. 96108.
Libyans and the Limes: Culture and Society in Roman Tripolitania, Antiquites
africaines 23 (1987), pp. 7194.
Oil for Export? A Comparison of Libyan, Spanish and Tunisian Olive Oil
Production in the Roman Empire, Journal of Roman Archaeology 1 (1988),
pp. 3356.

409

Bibliography
Olive Cultivation and the Albertini Tablets, in A. Mastino (ed.), LAfrica
romana: Atti del VI convegno di studio, Pubblicazioni del Dipartimento di storia
dellUniversit`a di Sassari 14 (Sassari, 1989), pp. 40315.
RomanoLibyan Settlement: Typology and Chronology, in G. W. W. Barker,
D. D. Gilbertson, B. Jones, and D. J. Mattingly (eds.), Farming the Desert: The
UNESCO Libyan Valleys Archaeological Survey, 2 vols. (Paris and London, 1996),
1:11158.
Tripolitania (London, 1995).
War and Peace in Roman North Africa: Observations and Models of State-Tribe
Interaction, in R. B. Ferguson and N. Whitehead (eds.), War in the Tribal Zone:
Expanding States and Indigenous Warfare (Santa Fe, NM, 1992), pp. 3160.
Mattingly, D. J. and R. B. Hitchner, Roman Africa: An Archaeological Review,
JRS 85 (1995), pp. 165213.
Mattingly, D. J., D. Stone, L. Stirling, and N. Ben Lazreg, Leptiminus (Tunisia):
A Producer City?, in D. J. Mattingly and J. Salmon (eds.), Economies beyond
Agriculture in the Classical World (London, 2000), pp. 6689.
Melluso, M., Alcune testimonianze in tema di mercati di schiavi nel tardo antico,
in M. Garrido-Hory (ed.), Routes et marches desclaves, 26e colloque du GIREA
(Paris, 2002), pp. 34570.
Merlin, A. and R. Lantier, Catalogue du Musee Alaoui, 2e Supplement, Catalogue des
musees et collections archeologiques de lAlgerie et de la Tunisie 15/2 (Paris,
1922).
Merrills, A. H., The Perils of Panegyric: The Lost Poem of Dracontius and its
Consequences, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New
Perspectives on Late Antique North Africa (Aldershot, 2004), pp. 14562.
(ed.), Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa
(Aldershot, 2004).
Merrills, A. H. and R. Miles, The Vandals (Chichester, 2010).
Meyer, E. A., Explaining the Epigraphic Habit in the Roman Empire: The Evidence
of Epitaphs, Journal of Roman Studies 80 (1990), pp. 7496.
Miles, R., The Anthologia Latina and the Creation of a Secular Space in Vandal
Carthage, Antiquite tardive 13 (2005), pp. 30520.
Millar, F., Local Cultures in the Roman Empire: Libyan, Punic and Latin in Roman
Africa, JRS 58 (1968), pp. 12634.
Mitchell, S. and G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London,
2000).
Moderan, Y., LAfrique reconquise et les Trois Chapitres, in C. Chazelle and C.
Cubitt (eds.), The Crisis of the Oikoumene: The Three Chapters and the Failed
Quest for Unity in the Sixth-century Mediterranean (Turnhout, 2007), pp. 39
82.

LEtablissement
territorial des Vandales en Afrique, Antiquite tardive 10 (2002),
pp. 87122.
Les Fronti`eres mouvantes du royaume vandale, in C. Lepelley and X. Dupuis
(eds.), Fronti`eres et limites geographiques de lAfrique du Nord antique (Paris, 1999),
pp. 24163.
Une Guerre de religion: les deux e glises dAfrique a` lepoque vandale, Antiquite
tardive 11 (2003), pp. 2144.

410

Bibliography
Kusayla, lAfrique et les Arabes, in C. Briand-Ponsart (ed.), Identites et culture
dans lAlgerie antique, Publications de lUniversite de Rouen 377 (Mont-SaintAignan, 2005), pp. 42357.
Les Maures de lAfrique romaine dans lantiquite tardive, Revue des etudes latines
82 (2004), pp. 24969.
Les Maures et lAfrique romaine (IVe VIIe si`ecle), Biblioth`eque des e coles francaises
dAth`enes et de Rome 314 (Rome, 2003).
La Notitia provinciarum et civitatum Africae et lhistoire du royaume vandale, Antiquite tardive 14 (2006), pp. 16585.
Monceaux, P., Enquete sur lepigraphie chretienne dAfrique, Revue archeologique,
4th ser. 2 (1903), pp. 5990.
Moorhead, J., The Decii under Theoderic, Historia: Zeitschrift fur alte Geschichte 33
(1984), pp. 10715.
Theoderic in Italy (Oxford, 1992).
Moralee, J., Maximinus Thrax and the Politics of Race in Late Antiquity, Greece
& Rome 55 (2008), pp. 5582.
Morcelli, S. A., Africa christiana, 3 vols. (Brescia, 181617).
Morizot, P., LAur`es et lolivier, Antiquites africaines 29 (1993), pp. 177240.

Byzantines,
Masties fut-il jamais imperator?, in XXe Congr`es international des Etudes
Coll`ege de France Sorbonne, 1925 aout 2001, Pre-Actes, vol. 2, Tables rondes (Paris,
2001), pp. 2623.
Pour une nouvelle lecture de lelogium de Masties, Antiquites africaines 25 (1989),
pp. 26384.
Morrisson, C., LAtelier de Carthage et la diffusion de la monnaie frappee dans
lAfrique vandale et byzantine (439695), Antiquite tardive 11 (2003), pp. 65
84.
Catalogue des monnaies byzantines de la Biblioth`eque Nationale, 2 vols. (Paris, 1970).
Un Tresor de solidi de Constantin IV de Carthage, Revue numismatique 6th ser.
22 (1980), pp. 15560.
Morrisson, C. and W. Seibt, Sceaux de commerciaires byzantins du VIIe si`ecle
trouves a` Carthage, Revue numismatique, 6th ser. 24 (1982), pp. 22240.
Morrisson, C., C. Brenot, and J.-N. Barrandon, LArgent chez les Vandales: plats
et monnaies, in F. Baratte (ed.), Argenterie romaine et byzantine: actes de la table
ronde, Paris 1113 octobre 1983 (Paris, 1988), pp. 12333.
Moussy, C., Introduction, in Dracontius, uvres, 4 vols. (Paris, 198596), 1:7136.
Muhlberger, S., The Fifth-century Chroniclers: Prosper, Hydatius, and the Gallic Chronicler
of 452, ARCA Classical and Medieval Texts, Papers and Monographs 27 (Leeds,
1990).
Murialdo, G., Le Anfore da trasporto, in T. Mannoni and G. Murialdo (eds.),
S. Antonino: un insediamento fortificato nella Liguria bizantina, Collezione di monografie preistoriche ed archeologiche 12, 2 vols. (Bordighera, 2001), pp. 25596.
Musurillo, H., The Acts of the Christian Martyrs (Oxford, 1972).
Nesbitt, J. and N. Oikonomides (eds.), Catalogue of Byzantine Seals at Dumbarton
Oaks and in the Fogg Museum of Art, vol. 1, Italy, North of the Balkans, North of
the Black Sea (Washington, DC, 1991).
Neuru, L., The Pottery of the Kasserine Survey, Antiquites africaines 26 (1990),
pp. 2559.

411

Bibliography
Northcote, J. S. and W. R. Brownlow, Rome souterraine, trans. P. Allard (2nd edn;
Paris, 1874).
Roma sotterranea (2nd edn; London, 1879).
Noth, A. and L. Conrad, The Early Arabic Historical Tradition: A Source-Critical Study,
trans. M. Bonner, Studies in Late Antiquity and Early Islam 3 (2nd edn; Princeton, NJ, 1994).
Noy, D., Immigrants in Late Imperial Rome, in S. Mitchell and G. Greatrex (eds.),
Ethnicity and Culture in Late Antiquity (London, 2000), pp. 1530.
ODonnell, J. J., Cassiodorus (Berkeley, Calif., 1979).
Oost, S. I., Aetius and Majorian, Classical Philology 59 (1964), pp. 239.
Otto, A., Die Sprichworter und sprichwortlichen Redensarten der Romer (Leipzig, 1890;
repr. Hildesheim, 1988).
Overbeck, M., Untersuchungen zum afrikanischen Senatsadel in der Spatantike, Frankfurter althistorische Studien 7 (Kallmunz, 1973).
The Oxford Classical Dictionary, ed. S. Hornblower and A. Spawforth (3rd edn;
Oxford, 1996).
The Oxford Dictionary of Byzantium, ed. A. Kazhdan, 3 vols. (Oxford, 1991).
Pallas-Brown, R., East Roman Perceptions of the Avars in the Mid- and Late-sixth
Century, in S. Mitchell and G. Greatrex (eds.), Ethnicity and Culture in Late
Antiquity (London, 2000), pp. 30929.
Palme, B., The Imperial Presence: Government and Army, in R. Bagnall (ed.),
Egypt in the Byzantine World, 300700 (Cambridge, 2007), pp. 24470.
Panella, C., Le anfore di Cartagine: nuovi elementi per la ricostruzione dei flussi
commerciali del Mediterraneo in et`a imperiale romana, Opus: Rivista internazionale per la storia economica e sociale dellantichit`a 2 (1983), pp. 5374.
Le anfore tardoantiche: centri di produzione e mercati preferenziali, in A. Giardina (ed.), Societ`a Romana e imperio tardoantico, III, Le merci. Gli insediamenti
(Rome, 1986), pp. 25172.
Merci e scambi nel Mediterraneo tardoantico, in A. Momigliano and A. Schiavone (eds.), Storia di Roma, 4 vols. (Turin, 1988 ), 3/2: 61397.
Panella, C. and L. Sagu`, Consumo e produzione a Roma tra tardoantico e altomedioevo: le merci, i contesti, in Roma nellalto medioevo, 2 vols., Settimane di
studio del Centro italiano di studi sullalto medioevo 48 (Spoleto, 2001), 2:757
820.
Parker, A. J., Ancient Shipwrecks of the Mediterranean and the Roman Provinces, BAR
International Series 580 (Oxford, 1992).
Peacock, D. P. S., F. Bejaoui, and N. Ben Lazreg, Roman Amphora Production
in the Sahel Region of Tunisia, in Amphores romaines et histoire economique: dix

ans de recherche, Collection de lEcole

francaise de Rome 114 (Rome, 1989),
pp. 179222.
Roman Pottery Production in Central Tunisia, Journal of Roman Archaeology 3
(1990), pp. 5984.
Pena, J. T., The Mobilization of State Olive Oil in Roman Africa: The Evidence
of Late 4th c. Ostraca from Carthage, in J. H. Humphrey (ed.), Carthage Papers:
The Early Colonys Economy, Water Supply, a Public Bath, and the Mobilization
of State Olive Oil, Journal of Roman Archaeology Supplementary Series 28
(Portsmouth, RI, 1998), pp. 117238.

412

Bibliography
Pentz, P., From Roman Proconsularis to Islamic Ifrqiyah, GOTARC Serie B. Gothenberg
Archaeological Theses 22 (Copenhagen, 2002).
Percival, J., Culturae Mancianae: Field Patterns in the Albertini Tablets, in B. Levick
(ed.), The Ancient Historian and his Materials: Essays in Honour of C. E. Stevens
on his Seventieth Birthday (Farnborough, 1975), pp. 21327.
Pewesin, W., Imperium, Ecclesia universalis, Rom: Der Kampf der afrikanischen Kirche
um die Mitte des 6. Jahrhunderts, Geistige Grundlagen romischer Kirchenpolitik,
Forschungen zur Kirchen- und Geistesgeschichte 11 (Stuttgart, 1937).
Pirenne, H., Mahomet et Charlemagne (7th edn; Paris, 1937).
Pohl, W., Telling the Difference: Signs of Ethnic Identity, in W. Pohl and
H. Reimitz (eds.), Strategies of Distinction: The Construction of Ethnic Communities, 300800, Transformation of the Roman World 2 (Leiden, 1998), pp.
1769.
Tradition, Ethnogenese und literarische Gestaltung: Eine Zwischenbilanz, in K.

Brunner and B. Merta (eds.), Ethnogenese und Uberlieferung:

Angewandte Methoden
der Fruhmittelalterforschung (Vienna, 1994), pp. 926.
Prevost, V., Les Derni`eres Communautes chretiennes autochtones dAfrique du
Nord, Revue de lhistoire des religions 224 (2007), pp. 46183.
Pringle, D., The Defence of Byzantine Africa from Justinian to the Arab Conquest: An
Account of the Military History and Archaeology of the African Provinces in the Sixth
and Seventh Centuries, BAR International Series 99 (Oxford, 2001).
Prosopographie chretienne du Bas-empire, vol. 1, Prosopographie de lAfrique chretienne
(303533), ed. A. Mandouze (Paris, 1982); vol. 2, Prosopographie de lItalie
chretienne (313604), ed. C. Pietri and L. Pietri (Rome, 19992000).
Prosopography of the Byzantine Empire, ed. J. R. Martindale (Ashgate, 2001) (CDRom).
The Prosopography of the Later Roman Empire, ed. A. H. M. Jones, J. R. Martindale,
and J. Morris, 3 vols. (Cambridge, 197192).
Pryor, J., Geography, Technology, and War: Studies in the Maritime History of the Mediterranean, 6491571 (Cambridge, 1988).
Pryor, J. and E. M. Jeffreys, The Age of the Dromwn: The Byzantine Navy ca 5001204,
The Medieval Mediterranean 62 (Leiden, 2006).
Quentin, H., Les Martyrologes historiques du moyen a ge: etude sur la formation du Martyrologe romain (2nd edn; Paris, 1908).
Rance, P., Battle, in P. Sabin, H. Van Wees, and M. Whitby (eds.), The Cambridge
History of Greek and Roman Warfare, 2 vols. (Cambridge, 2007), 2:34278.
Campidoctores Vicarii vel Tribuni: The Senior Regimental Officers of the Late
Roman Army and the Rise of the Campidoctor, in A. Lewin and P. Pellegrini
(eds.), The Late Roman Army in the Near East from Diocletian to the Arab Conquest,
BAR International Series 1717 (Oxford, 2007), pp. 395410.
Rapp, C., Hagiography and Monastic Literature between Greek East and Latin West
in Late Antiquity, in Cristianit`a doccidente e cristianit`a doriente (secoli VIIX),
2 vols., Settimane di studio del Centro italiano di studi sullalto medioevo 51
(Spoleto, 2004), 2:122180.
Raven, S., Rome in Africa (3rd edn; London, 1993).
Rebuffat, R., Recherches sur le bassin du Sebou (Maroc), Comptes rendus de seances
de lAcademie des Inscriptions et Belles-Lettres (1986), pp. 63361.

413

Bibliography
Remensnyder, A., Christian Captives, Muslim Maidens, and Mary, Speculum 82
(2007), pp. 64277.
Reynolds, P., Trade in the Western Mediterranean, ad 400700: The Ceramic Evidence,
BAR International Series 604 (Oxford, 1995).
Riche, P., Education and Culture in the Barbarian West, from the Sixth through the Eighth
Century, trans. J. J. Contreni (Columbia, SC, 1978).
Riedinger, R., Die Lateranakten von 649: Ein Werk der Byzantiner um Maximos
Homologetes, Byzantina 13/1 (1985), pp. 51734.
Riley, J. A., The Pottery from the Cisterns 1977.1, 1977.2 and 1977.3, in J. H.
Humphrey (ed.), Excavations at Carthage 1977, Conducted by the University of
Michigan, vol. 6 (Ann Arbor, Mich., 1981), pp. 85124.
Rinaldi, G., Osservazioni sullepitaffio di Restuto, in Aquileia e lAfrica, Antichit`a
alto adriatiche 5 (Udine, 1974), pp. 1819.
Robert, L., Noms indig`enes dans lAsie-Mineure greco-romaine, Biblioth`eque
archeologique et historique de lInstitut francais darcheologie dIstanbul 13
(Paris, 1963).
Rodolfi, A., Procopius and the Vandals: How the Byzantine Propaganda Constructs and Changes African Identity, in G. Berndt and R. Steinacher

(eds.), Das Reich der Vandalen und seine (Vor-) Geschichten, Osterreichische
Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften
366/Forschungen zur Geschichte des Mittelalters 13 (Vienna, 2008),
pp. 23342.
Rosenblum, M., Luxorius: A Latin Poet among the Vandals (New York, 1961).
Rouge, J., Escroquerie et brigandage en Afrique romaine au temps de saint Augustin
(Epist. 8* et 10*), in Les lettres de saint Augustin decouvertes par Johannes Divjak:
communications presentees au colloque des 20 et 21 Septembre 1982 (Paris, 1983),
pp. 17788.
Rowland, R., Jr., The Sojourn of the Body of St Augustine in Sardinia, in
J. Schnaubelt and F. Van Fleteren (eds.), Augustine in Iconography: History and
Legend (New York, 1999), pp. 18998.
Ruggiero, F., Atti dei martiri Scilitani: Introduzione, testo, traduzione, testimonianze e
commento, Atti della Accademia Nazionale dei Lincei, Memorie, 9th ser., 1/2
(Rome, 1991).
Ruinart, T., Historia persecutionis Vandalicae (Paris, 1694).
Rummel, P. von., Habitus barbarus: Kleidung und Reprasentation spatantiker Eliten im
4. und 5. Jahrhundert, Erganzungsbande zum Reallexikon der Germanischen
Altertumskunde 55 (Berlin, 2007).
Habitus Vandalorum? Zur Frage nach einer gruppen-spezifischen Kleidung der
Vandalen in Nordafrika, Antiquite tardive 10 (2002), pp. 13141.
Rushworth, A., From Arzuges to Rustamids: State Formation and Regional Identity in the Pre-Saharan Zone, in A. H. Merrills (ed.), Vandals, Romans,
and Berbers: New Perspectives on Late Antique North Africa (Aldershot, 2004),
pp. 7798.
From Periphery to Core in Late Antique Mauretania, in G. Fincham, G. Harrison, R. Holland, and L. Revell (eds.), TRAC 99, Proceedings of the Ninth
Annual Theoretical Roman Archaeology Conference, Durham 1999 (Oxford, 2000),
pp. 90103.

414

Bibliography
Sallares, R., Ecology, in W. Scheidel, I. Morris, and R. Saller (eds.), Cambridge
Economic History of the Greco-Roman World (Cambridge, 2007), pp. 1537.
Samuel, A., Greek and Roman Chronology: Calendars and Years in Classical Antiquity,
Handbuch der Altertumswissenschaft 1/7 (Munich, 1972).
Sarris, P., Economy and Society in the Age of Justinian (Cambridge, 2006).
Savage, E., A Gateway to Hell, a Gateway to Paradise: The North African Response to
the Arab Conquest, Studies in Late Antiquity and Early Islam 7 (Princeton, NJ,
1997).
Schanz, M. and C. Hosius, Geschichte der romischen Literatur bis zum Gesetzgebungswerk
des Kaisers Justinian, 4 vols. (Munich, 190720).
Scharf, U., Straenkleidung der romischen Frau, Geschichte und ihre Hilfswissenschaften
ser. 3, 585 (Frankfurt am Main, 1994).
Schlunk, H., Zu den fruhchristlichen Sarkophagen aus der Bureba (Prov. Burgos),
Madrider Mitteilungen 6 (1965), pp. 13966.
Schlunk, H. and T. Hauschild, Die Denkmaler der fruhchristlichen und westgotischen Zeit
(Mainz am Rhein, 1978).
Schmidt, L., Geschichte der Wandalen (2nd. edn; Munich, 1942).
Schmidt, L. Zur Frage nach der Volkszahl der Wandalen, Byzantinische Zeitschrift
15 (1906), pp. 6201.
Schonfeld, M., Worterbuch der altgermanischen Personen- und Volkernamen nach

der Uberlieferung
des klassischen Altertums bearbeitet (Heidelberg, 1911; repr.
1965).
Schopflin, G., Nations, Identity, Power (New York, 2000).
Schramm, P. E., Herrschaftszeichen und Staatssymbolik: Beitrage zu ihrer Geschichte vom
dritten bis zum sechzehnten Jahrhundert, Schriften der Monumenta Germaniae
historica 13, 3 vols. (Stuttgart, 19546).
Schwarcz, A., The Settlement of the Vandals in North Africa, in A. H. Merrills
(ed.), Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa
(Aldershot, 2004), pp. 4957.
Schwartz, J., The (Primarily) Mammalian Fauna, in H. R. Hurst and S. P. Roskams,
Excavations at Carthage: The British Mission, 2 vols. (Sheffield 1984), 1/1:229
56.
Sears, G., Late Roman African Urbanism: Continuity and Transformation in the City,
BAR International Series 1693 (Oxford, 2007).
Seston, W., Sur les derniers temps du christianisme en Afrique, Melanges darcheologie

et dhistoire de lEcole
francaise de Rome (1936), pp. 10124.
Seyrig, H., Antiquites syriennes, Syria: revue dart oriental et darcheologie 39 (1962),
pp. 424.
Shahd, I., Byzantium and the Arabs in the Sixth Century, 2 vols. (Washington, DC,
19952009).
Shanzer, D., Intentions and Audiences: History, Hagiography, Martyrdom, and
Confession in Victor of Vitas Historia Persecutionis, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot,
2004), pp. 27190.
A Philosophical and Literary Commentary on Martianus Capellas De Nuptiis Philologiae et Mercurii, Book 1, University of California Publications in Classical
Studies 32 (Berkeley, Calif., 1986).

415

Bibliography
Shaw, B., Autonomy and Tribute: Mountain and Plain in Mauretania Tingitana, in
P. Baudel (ed.), Desert et montagne: hommage a` Jean Dresch, Revue de lOccident
Musulman et de la Mediterranee 412 (Aix-en-Provence, 1986), pp. 6689.
Eaters of Flesh, Drinkers of Milk: The Ancient Mediterranean Ideology of the
Pastoral Nomad, Ancient Society 13/14 (1982/3), pp. 531.
Fear and Loathing: The Nomad Menace and Roman Africa, in C. M. Wells
(ed.), Roman Africa/LAfrique romaine: The 1980 Governor-General Vanier Lectures,
Revue de lUniversite dOttawa 52 (Ottawa, 1982), pp. 2546.
Rural Markets in North Africa and the Political Economy of the Roman Empire,
Antiquites africaines 17 (1981), pp. 3783.
Sherwin-White, A. N., The Roman Citizenship (2nd edn; Oxford, 1973).
Sherwood, P., An Annotated Date-List of the Works of Maximus Confessor, Studia
Anselmiana 30 (Rome, 1952).
Singer, S., W. Ziltener, C. Hostettler, and Kuratorium Singer, Thesaurus proverbiorum
medii aevi: Lexikon der Sprichworter des romanischgermanischen Mittelalters, 13 vols.
(Berlin, 19956).
Sivan, H., Ausonius of Bordeaux: Genesis of a Gallic Aristocracy (New York, 1993).
Why Not Marry a Barbarian? Marital Frontiers in Late Antiquity (The Example
of CTh 3.14.1), in R. Mathisen and H. Sivan (eds.), Shifting Frontiers in Late
Antiquity (Aldershot, 1996), pp. 13645.
Sjostrom, I., Tripolitania in Transition: Late Roman to Early Islamic Settlement (Aldershot,
1993).
Smalley, B., Sallust in the Middle Ages, in R. R. Bolgar (ed.), Classical Influences
on European Culture ad 5001500: Proceedings of an International Conference Held
at Kings College, Cambridge, April 1969 (Cambridge, 1971), pp. 16575.
Snowden, F., Jr., Before Color Prejudice: The Ancient View of Blacks (Cambridge, Mass.,
1983).
Sodini, J.-P., Productions et e changes dans le monde protobyzantin (IVe VIIe s.): le
cas de la ceramique, in K. Belke, F. Hild, J. Koder, and P. Soustal (eds.), Byzanz

als Raum: Zu Methoden und Inhalten der historischen Geographie des Ostlichen
Mit
telmeerraumes, Osterreichische Akademie der Wissenschaften, Philosophischhistorische Klasse, Denkschriften 283/Tabula Imperii Byzantini 7 (Vienna,
2000), pp. 181208.
Sotinel, C., Emperors and Popes in the Sixth Century: The Western View, in
M. Maas (ed.),The Cambridge Companion to the Age of Justinian (Cambridge,
2005), pp. 26790.

Soustal, P., Thrakien (Thrake, Rodope und Haimimontos), Osterreichische

Akademie der
Wissenschaften, Philosophisch-historische Klasse, Denkschriften 221/Tabula
Imperii Byzantini 6 (Vienna, 1991).
Speight, R. M., The Place of the Christians in Ninth Century North Africa,
According to Muslim Sources, Islamochristiana 4 (1978), pp. 4765.
Temoignage des sources musulmanes sur la presence chretienne au Maghreb de
26/647 a` 184/800, Institut des belles lettres arabes 129 (1972), pp. 7396.
Stein, E., Histoire du Bas-empire, ed. J.-R. Palanque, 2 vols. (Paris, 194959).
Steinacher, R., Gruppen und Identitaten: Gedanken zur Bezeichnung
vandalisch, in G. Berndt and R. Steinacher (eds.), Das Reich der Vandalen

Akademie der Wissenschaften,

und seine (Vor-) Geschichten, Osterreichische

416

Bibliography
Philosophisch-historische Klasse, Denkschriften 366/Forschungen zur
Geschichte des Mittelalters 13 (Vienna, 2008), pp. 24360.
The So-Called Laterculus Regnum Vandalorum et Alanorum: A Sixth-Century
African Addition to Prosper Tiros Chronicle?, in A. H. Merrills (ed.), Vandals,
Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot,
2004), pp. 16380.
Stern, K., Inscribing Devotion and Death: Archaeological Evidence for Jewish Populations
of North Africa, Religions in the Graeco-Roman World 161 (Leiden, 2008).
Stevens, S. (ed.), Bir el Knissia at Carthage: A Re-Discovered Cemetery Church, Report
No. 1, Journal of Roman Archaeology Supplementary Series 7 (Ann Arbor,
Mich., 1993).
The Circle of Bishop Fulgentius, Traditio 38 (1982), pp. 32741.
Sepultures tardives intra muros a` Carthage, in P. Trousset (ed.), LAfrique du Nord
antique et medievale: monuments funeraires, institutions autochtones (Paris, 1995),
pp. 20717.
Stout, A., Jewelry as a Symbol of Status in the Roman Empire, in J. L. Sebesta
and L. Bonfante (eds.), The World of Roman Costume (Madison, Wis., 1994),
pp. 77100.
Stroheker, K. F., Der senatorische Adel im spatantiken Gallien (Tubingen, 1948).
Spanische Senatoren der spatromischen und westgotischen Zeit, in Germanentum
und Spatantike (Zurich, 1965), pp. 5487.
Syed, Y., Vergils Aeneid and the Roman Self: Subject and Nation in Literary Discourse
(Ann Arbor, Mich., 2005).
Syme, R., Tacitus, 2 vols. (Oxford, 1958).
Szidat, J., Zum Sklavenhandel in der Spatantike (Aug. epist. 10*), Historia: Zeitschrift
fur alte Geschichte 34 (1985), pp. 36071.
Terry, J., Christian Tomb Mosaics of Late Roman, Vandalic and Byzantine Byzacena, Ph.D. dissertation, University of Missouri, Columbia (1998).
Thebert, Y., Thermes romains dAfrique du Nord et leur contexte mediterraneen: etudes
dhistoire et darcheologie, Biblioth`eque des e coles francaises dAth`enes et de
Rome 315 (Rome, 2003).
Thompson, E. A., The End of Roman Spain, Nottingham Medieval Studies 20 (1976),
328 (Part I); 21 (1977), 331 (Part II); 22 (1978), 322 (Part III); 23 (1979),
121 (Part IV).
Thompson, L., Blacks and Romans (Norman, Okla., 1989).
Tiefenbach, H., Das wandalische Domine miserere, Historische Sprachforschung (Historical Linguistics) 104 (1991), pp. 25168.
Troncarelli, F., LOdissea di unodissea: note sullIlario basilicano (Arch. S. Pietro D
182), Scriptorium 45 (1991), pp. 321.
Trout, D., Paulinus of Nola: Life, Letters, and Poems (Berkeley, Calif., 1999).
Uhalde, K., Expectations of Justice in the Age of Augustine (Philadelphia, Pa., 2007).
Ulin, R. C., Understanding Cultures: Perspectives in Anthropology and Social Theory (2nd
edn; Malden, Mass., 2001).

Vaa nanen, V., Etude

sur le texte et la langue des Tablettes Albertini, Suomalaisen
Tiedeakatemian toimituksia, ser. B, 141/2 (Helsinki, 1965).
van der Veen, M., A. Grant, and G. Barker, RomanoLibyan Agriculture: Crops and
Animals, in G. W. W. Barker, D. D. Gilbertson, B. Jones, and D. J. Mattingly

417

Bibliography
(eds.), Farming the Desert: The UNESCO Libyan Valleys Archaeological Survey,
2 vols. (Paris and London, 1996), 1:22763.
van Parys, M., Un Matre spiritual oublie: Thalassios de Libye, Irenikon 52 (1979),
pp. 21440.
Vattioni, F., Glosse Puniche, Augustinianum 16 (1976), pp. 50555.
Vitelli, G., Islamic Carthage: The Archaeological, Historical and Ceramic Evidence, Dossiers
CEDAC 2 (Carthage, 1981).
Vives, J., El supuesto Pasionario hispanico de San Millan de la Cogolla, Hispania
Sacra 12 (1959), pp. 44553.
von den Brincken, A.-D., Studien zur lateinischen Weltchronistik bis in das Zeitalter
Ottos von Freising (Dusseldorf, 1957).
Vossing, K., Schule und Bildung in Nordafrika der romischen Kaiserzeit (Brussels, 1997).
Walker, J., A Catalogue of the Muhammadan Coins in the British Museum, 2 vols.
(London, 194156).
Wallace-Hadrill, A., Romes Cultural Revolution (Cambridge, 2008).
Walther, H., Lateinische Sprichworter und Sentenzen des Mittelalters in alphabetischen
Anordnung, Carmina Medii Aevi Posterioris Latina 2, 9 vols. (Gottingen, 1963
86).
Wanscher, O., Sella Curulis: The Folding Stool, an Ancient Symbol of Dignity (Copenhagen, 1980).
Ward-Perkins, B., The Fall of Rome and the End of Civilization (Oxford, 2005).
Ward-Perkins, J. B. and R. G. Goodchild, The Christian Antiquities of Tripolitania,
Archeologia: Miscellaneous Tracts Relating to Antiquity 95 (1953), pp. 184.
Warmington, B. H., The North African Provinces from Diocletian to the Vandal Conquest
(Cambridge, 1954).
Warwick, H. H., A Vergil Concordance (Minneapolis, Minn., 1975).
Weaver, R. H., Divine Grace and Human Agency: A Study of the Semi-Pelagian Controversy, Patristic Monograph Series 15 (Macon, Ga., 1996).
Wenskus, R., Stammesbildung und Verfassung: Das Werden der fruhmittelalterlichen Gentes
(Cologne, 1961).
Wessel, H., Das Recht der Tablettes Albertini (Berlin, 2003).
Whitehouse, D., L. Costantini, F. Guidobaldi, S. Passi, P. Pensabene, S. Pratt,
R. Reece, and D. Reese, The Schola Praeconum II, Papers of the British
School at Rome 53 (1985), pp. 163210.
Whittaker, C. R., Frontiers of the Roman Empire: A Social and Economic Study (Baltimore, Md., 1994).
Land and Labour in North Africa, Klio 60 (1978), pp. 33162.
Wickham, C., Framing the Early Middle Ages: Europe and the Mediterranean, 400800
(Oxford, 2005).
Williams, D. H., Ambrose of Milan and the End of the Arian-Nicene Conflicts (Oxford,
1995; repr. 2002).
Wilmart, A., LOdyssee du manuscrit de San Pietro qui renferme les uvres de
saint Hilaire, in L. W. Jones (ed.), Classical and Mediaeval Studies in Honor of
Edward Kennard Rand (New York, 1938), pp. 293305.
Wilson, A., Timgad and Textile Production, in D. J. Mattingly and J. Salmon (eds.),
Economies beyond Agriculture in the Classical World, Leicester-Nottingham Studies
in Ancient Society 9 (London, 2001), pp. 27196.

418

Bibliography
Wolfram, H., History of the Goths, trans. T. J. Dunlap (Berkeley, Calif., 1988).
The Roman Empire and its Germanic Peoples, trans. T. J. Dunlap (Berkeley, Calif.,
1997).
Wood, I. N. (ed.), Franks and Alamanni in the Merovingian Period: An Ethnographic
Perspective, Studies in Historical Archaeoethnology 3 (Woodbridge, 1998).
The Merovingian Kingdoms, 450751 (London, 1994).
Woolf, G., Becoming Roman: The Origins of Provincial Civilization in Gaul (Cambridge,
1998).
Beyond Romans and Natives, World Archaeology 28/3 (1997), pp. 33950.

Wrede, F., Uber

die Sprache der Wandalen: Ein Beitrag zur germanischen Namen- und
Dialektforschung, Quellen und Forschungen zur Sprach- und Kulturgeschichte
der germanischen Volker 59 (Strasburg, 1886).
Wroth, W., Catalogue of the Coins of the Vandals, Ostrogoths and Lombards and of the
Empires of Thessalonica, Nicaea and Trebizond in the British Museum (Oxford,
1911).
Zacos, G. and A. Veglery, Byzantine Lead Seals, 2 vols. (Basle, 197285).
Zechiel-Eckes, K., Die Concordia canonum des Cresconius: Studien und Edition, 2 vols.,
Freiburger Beitrage zur mittelalterlichen Geschichte 5 (Frankfurt am Main,
1992).
Zuckerman, C., La Haute Hierarchie militaire en Afrique byzantine, Antiquite
tardive 10 (2002), pp. 16975.

419

INDEX

c Abd

Allah ibn Saad, 335

Achilles, 55
Adeodatus (name), 11527, 283
administration, Byzantine
Africans in, 2034, 211, 237
financial, 199, 2356
high officials
court connections, 2045, 2203, 226
regional origins, 2024
promotion within, 2234
senior officers
families of, 2424
military background, 21820
prior service in Africa, 2246
term of appointment, 2335
ties to predecessors, 223
size, 240
structure of, 1978
subordinate officers
court connections, 2278
recommendation of, 229, 230
regional origins, 2056, 20711
term of appointment, 2369
ties to superiors, 22831
administration, Roman, 912
administration, Vandal, 46, 1434
Aegean, 92, 215, 337
Aetius, 27
Afariqa, see Romano-Africans
Africa
Byzantine
military crises in, 21617, 3012
reconquest of, 2545
defined, 16
identification of Vandals with, 56
Islamic (Ifriqiya)
Byzantine-Islamic transition, 3623
Christianity in, 36370
communications with Italy, 3669
conquest, 111, 232, 280, 305, 335, 353,
3589
local pride in, 556

Roman
distinctive culture of, 89
Romanness of, 34, 78
social and political ties to Italy, 912
Vandal conquest of, 1312
Africa Proconsularis
Arian persecution in, 144, 1645, 166, 181
Byzantine administration of, 197, 207
Byzantine fortification of, 299
dating systems, 1501
economy, 51, 91, 934, 334, 336
location of, 7
provincial church council of, 322, 348,
357
Roman administration of, 910
Vandal control of, 22
Vandal settlement in, 4750
African red slip ware
as economic evidence, 91, 333
distribution in Africa, 989, 1001, 287,
288, 334
exports, 924, 95, 336, 337
production, 51, 91, 98, 140, 336
Agapetus (Pope), 172, 347
Agnellus (Ostrogothic ambassador), 39, 41
Agan (Byzantine officer), 207, 22930, 232,
300
Albertini Tablets
economy and society in, 978, 99, 141,
2813
estate management and, 139
language of, 187
literacy in, 134
property rights in, 138
regnal dating in, 45, 156
written on non-local wood, 100
Alexander (ambassador), 32
Alexandria, 74, 93, 337, 366
Altava, 2767, 278, 28990, 294, 302
Amalafrida (Ostrogothic princess), 3840,
315
Amalasuntha (Ostrogothic queen), 315

420

Index
ambassadors
Avar, 260
East Roman
to Moors, 210
to Vandals, 30, 312, 33, 34, 356
Frankish, 3489
Moorish, 279, 294
multiple, 412
Ostrogothic, 39, 41
status of, 35
Vandal, 32, 35, 37, 412, 314
Visigothic, 25
West Roman, 301
amphitheatres, 53, 57, 133
amphorae
African exports, 935, 336, 337
as economic evidence, 91, 3334
distribution in Africa, 98, 100, 334
imports into Africa, 93, 337
production, 51, 140
Anastasius I (emperor), 34, 80, 213
anno, anno Karthaginis, see dating, year of
Carthage
Antalas
Berber name, 294
blackness of, 271
military tactics of, 266
provincial origins of, 276
relations with Byzantines, 298, 300, 302, 339
territory of, 268, 279
Antichrist, 80, 171, 359
Antioch, 57, 337
apocalypticism, 171, 272, 359
apostasy, see conversion
Arab invasion, see Africa, Islamic
Arabs
in Byzantine army, 206
medieval historians, 305, 335, 358, see also
Ibn c Abd al-Hakam
Procopius on, 256, 257
service in wars against, 225
Arcadius of Caesarea, St, 113
Archelaus (Praetorian Prefect), 205, 206
architectural fittings, 348
Areobindus, 204, 219, 220, 223, 242
Arianism
anti-Arian literature, 109, 1736
Arian-Nicene debate, 1735
barbarian identity and, 1923
catholicity of, 1812, 184
in Byzantine army, 321
Nicene conversion to, see conversion
patriarchate of Carthage, 467
political loyalty and, 1767, 1789
Roman identity and, 160, 194

suppression of, 3201

tenets of, 15960
Vandal conversion to, 15960, 185
Vandal identity and, 62, 159, 164, 1823
Vandal kingship and, 161, 1834
aristocratic lifestyle, 535, 136
Armenia, 219, 230, 242, 257
Armenians, 203, 2089, 219, 245, 350
army
Byzantine
Africans in, 210, 237, 2401, 339
Arians in, 321
Belisarius command corps, 2057
death in service, 2313
eastern reinforcements of, 240
field promotion by, 2234
Greek in, 245
ideological role of, 254
Latin in, 244
limitanei, 198, 2401, 249
local recruitment into, 210, 2401
local society and, 2412, 24750
looting and, 31011
military tactics, 2667, 297
military unrest and, 216, 3012, 321, 358
officers as local notables, 333
provisioning of, 311
raids of, 299
Thrace as recruiting ground for, 231
Roman, 49, 68
Vandal, 47, 5960, 168, 25960
Artabanes (magister militum)
Armenian, 203, 208
family of, 2089, 232, 242, 244
military career of, 219
recall of, 233
Artemios, St, 341
Arzugitanus poeta, 281
astrologers, 86
Athanagild (Visigothic prince), 3489
Athanasius (Praetorian Prefect), 205, 234
Attila, 37
Augustine of Hippo
circulation of writings, 1078
early career of, 1112
ethnic distinctions in, 192
family of, 295
illicit slaving in, 299
Moors in, 267, 274, 283
Punic identity in, 1879
Romanness in, 18990, 192, 375
translation of relics, 11011
Aur`es Mountains
Byzantine fortifications in, 250
Christianity in, 268, 269, 288

421

Index
Aur`es Mountains (cont.)
economic and cultural ties, 100, 288
estate management in, 341
Islamic invasion of, 280
Moorish power in, 278, 287, 302
Vandals flee to, 56
Avars, 260, 264, 351
Avitus (emperor), 30
Balearic Islands, 94, 288, 323
Balkans, 91, 160, 202, 2056, 2078, 2301
banishment, 67, 73, 86, 162
barbarian, Greco-Roman ideology of, 58,
1301, 193, 25661
barley, 263, 264, 289, see also grain
Basiliscus, 310
bath complexes
abandonment of, 54
construction of, 45, 534, 147
maintenance of, 54, 133
repurposing of, 133
use of, 56, 354
Belezma, 99, 277, 288
Belisarius
African policy of, 31011, 313
Balkan origins of, 202
career of, 219, 222
family of, 242, 243
key associates of, 223, 22930
loyalty of, 2223
Moors and, 279
recall of, 223, 233, 297
Benenatus (name), 11527, 283, 293
Bir el Knissia (basilica), 336
bishops, Arian, see also Bleda; Cyrila; Jucundus
as ambassadors to Vandals, 30, 35
epigraphy and, 164
judicial function of, 138, 165
Nicene apostates among, 172
Nicene disputation with, 36, 173
persecution and, 179
preach in Latin, 182
bishops, Nicene
Arian patriarchate and, 467
as local notables, 333, 33940
as travellers
epigraphic evidence, 701
limitations on overseas travel, 68
numbers, 68
Byzantine religious policy and, 3201, 357,
359, see also Maximus Confessor;
Three Chapters Controversy
care for refugees, 812
conversion to Arianism among, 172
cult of relics and, 340

end of persecution and, 170

flee Vandal invasion, 68
Gregory I and, see Gregory I
in Islamic Africa, 362, 364, 3689
in western Africa, 2913
naming patterns among, 2923
of Carthage, see Carthage
ordination of
Gelasius and, 1256
Islamic-era, 362, 366, 3679
Vandal-era, 30, 32, 1645, 166, 170, 184
political loyalties of, 1769
summoned to Constantinople, 3212, 323
targets of Arian persecution, 1626, 170
blackness, see skin colour
Bleda (ambassador), 30
Boccaccio, 370
Boethius (philosopher), 105, 108, 213
Boethius (primate of Byzacena), 322
Boniface (comes Africae), 68
Boniface (secretary of Gelimer), 134, 139
Bordj Djedid, mosaics, 61
buildings, see also amphitheatres; bath
complexes; churches; circuses;
fortifications; houses; theatres
Justinianic ideology and, 333
Vandal kingship and, 456, 147
Byzacena
Byzantine
Arab attacks reach, 358
as province, 197
council of bishops, 321, 322, 3278
fortification of, 287, 299
magistri militum in, 213, 227
military administration of, 198, 207
military unrest in, 302
Moorish attacks in, 216, 298, 299, 300,
352
Solomons campaign in, 300
subordinate officers in, 207, 208, 22930
dating systems, 1501
economy, 91, 94, 98, 99, 100, 140
exile to, 163, 167
foreign monks in, 878
location of, 7
Moors of, 255, 268, 278, 279, 297, 299
Roman administration of, 9
Vandal
control of, 22
persecution in, 169
royal estates in, 478
Vandal settlement and, 49
Caelestiacus (refugee), 70, 73, 74, 89
Caesarea in Mauretania, 88, 207, 2878

422

Index
Caesarius of Arles, 86, 109, 176, 345
camels, 97, 167, 266
Cartagena, 336
Cartennae, 102, 163
Carthage
African exports via, 93
bishops of (Arian), see Cyrila, Jucundus
bishops of (Nicene), see also Deogratias;
Eugenius; Quodvultdeus; etc.
exile of, 67, 73, 86, 102
Islamic-era, 364, 3689
Justinian and, 3212
ordination of, 30, 170, 184, 322, 3678
Papacy and, 3256, 348, 357, 3679
patriarch of Constantinople and, 357
vacant see, 1645
Byzantine
capital of Africa, 197
commanders in, 207
communications with Constantinople,
3456
Jewish community, 354
market in, 311
military unrest in, 3012
plague in, 242, 325, 351
resistance to empire in, 313
Vandal refugees in, 2545
clergy of (Nicene), see also Ferrandus;
Liberatus
Arian persecution and, 169
exile of, 67, 73, 85, 102, 162
Islamic-era, 367
recall from exile, 32, 1623
reclamation of St Cyprians basilica by, 164
cult of the Virgin in, 342
cultural life in, 134, 136, 336
Greek spoken in, 245
imports into, 93, 337
Islamic
Charlemagnes charity to, 366
conquest of, 358
continuity of occupation, 331
patriarchate of (Arian), see Arianism
public processions in, 59, 269, 272
rural economy of, 51, 334
Salvian of Marseilles and, 57
silver-working in, 91
spread of saints cults from, 112
ties to African interior, 95, 96, 98
travel time to Constantinople, 21416
Vandal
Arian persecution in, 1801
capital of kingdom, 2
celebration of, 56
coinage and, 56

communications hub, 88
confiscations in, 1634
exile near, 167
Roman aristocrats banished from, 67
seat of proconsul, 46
Vandal diet and, 62
Vandal kingship and, 203
Vandal settlement and, 49
Vandal-era poetry and, 56
year of, see dating
Cassian of Tingi, St, 113
Cassiodorus
African authors and, 338, 344
monastic library of (Vivarium), 109, 3445
Punic language and, 189
Variae of, 401, 85, 104
Cato (poet), 147
Caucasus, 2567, 278, see also Armenia
ceramics, see African red slip ware; amphorae;
lamps
Chalcedon
Chalcedonian orthodoxy
African defense of, 157, 323, 357
eastern threats to, 7980
Gregory (rebel) as defender of, 355
Maximus Confessor and, 356, see also
Maximus Confessor
Three Chapters and, 316, 31720, see also
Three Chapters
Council of, Africans at, 79
exile to, 322
chariot racing, 87, 136, see also circuses
Charlemagne, 111, 277, 3667
Childebert II (Frankish king), 3489
Christianity
Arian, see Arianism
in African borderlands, see Aur`es Mountains;
Mauretania; Moors; Numidia;
Tripolitania
in Islamic Africa, see Africa, Islamic
Nicene
debates within, see Chalcedon; Three
Chapters Controversy; Maximus
Confessor; etc.
Romanness and, see Romanness
churches
building of, 456, 2689
confiscation of, 1634, 3267
dedication of, 342
economic production and, 333
in Moorish hinterland, 268
remodelling of, 336
urban landscape and, 133, 332
Vandals take refuge in, 254
Cillium, see Thelepte

423

Index
circuses, 58, 133
cities
as index of Romanization, 7
changes in urban fabric, 1323, 3313
continuity of settlement, 331
cultural life and, 578, 1334, 2902
economic activity and, 51, 989, 133, 288,
333
focus of Vandal settlement, 4950, 258
moralists and, 578
municipal institutions and, 1323, 333
clergy
Arian, 48, 164, 165, 178, 179, 320
Nicene, see also Carthage, clergy (Nicene) of
acceptance of Africans into, 368
appeals to Papacy among, 3269
as local notables, 333
conversion to Arianism among, 84, 172
dedication of altars and, 340
exile of, 102
Islamic-era, 362, 367, 369
naming patterns among, 292
persecution of 484 and, 169
Roman, expansion of under Gelasius, 125
use of Vandal dates among, 152, 1568
clothing, see also jewellery
African export, 91
Albertini Tablets and, 2812
as ethnic signifier, 7
as index of Romanization, 7
barbarian dress, 602
imperial purple, 44
imports into Africa, 348
Moorish
battle-gear, 2656
cap, 279
cloak, 263, 265, 279
shoes, 334
silk, 56
Symmachus charity and, 104
trade within Africa, 99
coinage
as index of Romanization, 7
Byzantine
dating on, 155
Islamic invasion and, 335, 352
Islamic, 362
Vandal
Carthage on, 21, 56
dating on, 155
Justin I on, 313
Victory on, 45
Columbanus, 10910
Columbus (bishop), 325, 326, 328
commerciarii, 199, 213, 2356

Constans II, 235, 337, 356, 3578, 359

Constantine I, 910, 11, 145, 184
Constantinople
African exports to, 95, 3367
African grain and, 92, 336
African saints cults in, 112
Arianism in, 160, 184
as imperial residence, Constans II and, 3578
Byzantine officers wives kept in, 2423
captured by Heraclius, 355
circulation of manuscripts in, 108, 320,
3434
communications with Carthage, 3456
coronation ceremonial in, 302
exports to Africa, 337
First Council of, 160
importance as travel destination, 8990
Latin in, 8990
Second Council of, 322
Sixth Council of, 359
travel time to Carthage, 21416
Vandal envoys to, 32, 314
conversion
to Arianism
demanded by Huneric, 1812
ethnic significance of, 182
Nicene, 84, 1712
of Geiseric, 185
of Vandals, 15960
political significance of, 177
to Islam, 3634, 369
to Nicene Christianity
among Vandals, 180
evangelization of Moors and, 268
of Arian kings, 1756
of Arians, 320, 3478
of Jews, forced, 354
of Vandal kings, hoped for, 1746
Corbie, 108
Corippus
African perspective of, 199, 256
Africans in Byzantine army in, 210, 237
Armenians in, 2089
barbarians in, 25960
Byzantine officials in
Athanasius (prefect), 234
John Troglita, 219, 224
Thomas (prefect), 226
ConstantinopleAfrica route in, 215
education/intellectual outlook of, 259, 340
in Constantinople, 260, 338
Moors in
Antalas, 276
Berber names of, 2934
blackness of, 269, 271

424

Index
Byzantine reconquest and, 252
Cusina, 3034
fear of, 252
Latin-speakers among, 294
long hair of, 264
military dress of, 2656
military tactics of, 266
non-Romanness of, 2723
paganism of, 267
repulsiveness of, 272
rusticity and poverty of, 303
Roman identity in, 35960
Three Chapters and, 376
Corsica, 102, 103, 162, 178, 206, 347
court, imperial
African refugees at, 80, 85
Corippus at, 260, 338
episcopal appeals to, 3279, 346
Hilderic at, 34
Huneric at, 23, 2930
ideology of reconquest, 1967
Junillus at, 338
militarism at, see militarism
opposition to reconquest at, 310
court, Vandal
Arianism at, see persecution
barbarian dress and, 61
Carthage as focus of, 49
draw of, 98
political purges at, 1778
word Roman at, 186
Crete, 216, 358
cursus publicus, 46, 245
Cusina, 268, 27980, 294, 298, 3034
Cyprian, St
Byzantine reconquest and, 309, 342
churches of, 1634
cult of, 11213, 3656
Punic identity and, 1889
translation of, 111, 367
Cyprus, 92
Cyrila (Arian patriarch), 47, 63
Damira (Vandal princess), 59 n. 197, 147
Datianus (bishop), 321, 327, 346
dating
consular, 14850, 151, 284
Egyptian calendar, 87, 34950
episcopal, 339
from creation, 149
imperial, 1501, 155, 280, 284, 338,
340
indiction, 149, 338
local era, 149, 150, 151, 288, 290
Olympiads, 149

royal, 150, 1512, 1558

year of Carthage, 201, 1526, 157
decadence, 568
Deogratias of Carthage, 30, 162, 1645
diet, 62, 256, 2634
dignities, Roman, 35, 46, 1456, 248, 2856
Djebel Mrata, see Albertini Tablets
Djedar, 268, 277, 278
doctors, 169, 198
Doctrina Jacobi nuper baptizati, 354, 361
Dominicus of Carthage, 3256, 348
Domnulus (Quaestor of the Sacred Palace), 70,
89
Donatists, Donatism
Gregory I and, 324, 325, 326, 327, 328
Justinian and, 321
Vandals and, 156, 171, 180, 184
Dracontius
barbarians in, 59
circulation of writings, 109
education of, 141
Gunthamund and, 143, 183
legal career of, 137
literary circle of, 1356
name and family of, 145
Roman identity in, 190
Thrasamund and, 147
Vandal kingship and, 142, 147, 183
dress, see clothing
Droctulf (Byzantine officer), 20910, 229
dux
appointment of, 2289
Armenians as, 209
continuity of tenure, 236
in Egypt, 220, 228
Moorish rulers as, 278, 294
of Byzacena, 198, 207, 230, 236
of Mauretania, 198
of Mesopotamia, 21819, 237
of Numidia, 198, 206, 223, 229, 230,
275
of Sardinia, 198
of Tripolitania, 198, 209, 229, 279, see also
John Troglita; Sergius
economy
African production, 512, 13941, 2879
African prosperity, 3336
exchange within Africa, 98101
export commodities, see exports
long-distance exchange, 904, 2879,
3367
urban, 1323
Ecthesis, 356
education, 55, 134, 141, 174, 3401

425

Index
Egypt, see also Alexandria
African exports to, 93, 337
African troops posted in, 241
Arian monastery in, 184
Byzantine administration of
civilian nature of, 220
deployments, 237
imperial intimates in, 228
Libya Pentapolis, 228
locals in, 201, 210, 241, 247
non-locals in, 209
size of, 240
Byzantine duces of, see dux
ceramics production in, 92
Charlemagnes charity to, 366
commanders from, in Africa, 206
exile to, 323
Islamic raids from, 358
Israelites bondage in, 171
Persian and Islamic conquests of, 347, 357
saints cults from, in Africa, 343
Three Chapters and, 324
El Mahrine, 96, 98, 101, 140
Elias the Younger, 365, 369
elite, definition of, 15
Ennodius, 41, 86, 108, 109
epigraphy
Arian bishops and, 164
as index of Romanization, 7
bath complexes and, 534
Byzantine presence in Africa and, 24750
Christians of Ifriqiya and, 363, 365
circulation of books and letters and, 104, 109
cult of relics and, see relics
cult of saints and, see saints, cult of
dating systems and, 14859
exile and, 102
fortifications and, see fortifications
Greek-language inscriptions, 87, 88, 189,
245, 34950
Greek-Latin bilingual, 2456, 349
metrical, 341
Moorish Christianity and, 2689
Moorish seizure of interior and, 276,
278
Punic Christianity and, 189
Punic-language inscriptions, 1878, 189
Roman identity and, 2846, 288, 28992,
2945, 338
settlement of easterners and, 87, 88, 209,
34950
spread of African names and, 11822,
1267
transhumance and, 283
travel as revealed in, 702, 73, 74, 288

Vandal royal title and, 42, 44

Vandal settlement and, 49
estates, see also economy; property
as cities, 283
imperial, 47, 85, 168, 249, 285
management of, 139, 141, 341
markets on, 99
of Africans, in Italy, 85
of Byzantine elite, 242
of Italo-Roman elite, 12
of Romano-Africans, 48, 56, 97, 1378, 299
of Vandal elite, 47, 48, 2412
of Vandal royal family, 478, 54, 311
suppression of heresy on, 168
ethnicity
Arianism and, 1823
clothing and, 49, 52, 602, 2646
in early medieval West, 13
in Vandal kingdom, 49, 523, 5864
marks of distinction, 7
Moors and, 26175
names and, 49, 523, 2945
Romanness and, 190
royal titulature and, 2767
Eudaemon (father of Maria), 72, 74
Eudocia
brought to Africa, 267
death in Jerusalem, 32
Hilderic as son of, 434
inheritance due to, 24, 301, 32
marriage to Huneric
engagement, 223
importance of, 246
recognition of, 24, 30, 31, 38
threatened, 278
Eudoxia (empress), 267, 29, 31
Eugenius of Carthage
banishment to Albi, 86
death, 165
exile within Africa, 102, 163
ordination of, 165, 184
Vandal persecution and, 180
venerated as saint, 112, 128
writings of, 104, 1745
Eugippius (abbot), 105, 1078, 109
eunuchs, 147, 227, 228, 346
Euric (Visigothic king), 25, 37, 176
Eutyches, Eutychianism, 79, 318, 319, 320,
323
exarch, see also Gennadius; Gregory; Heraclius
Armenians as, 203
families of, 2423
military backgrounds of, 219
prior service in Africa, 2245
responsibilities of, 198

426

Index
revolts of, 355
subordinate appointments and, 2289
term of appointment, 234
exile, see also banishment
biblical references and, 163, 171
communications during, 104, 1057
Nicene bishops diet in, 264
of heretics, in Roman law, 168
of Vandal warriors, 255
recall from, 32, 1623, 170, 323
under Byzantines, 3223, 350, 357
under Vandals, 67, 87, 1778, see also
persecution
Vandal dating and, 157
exports (commodities), 91, 99, 334
Facundus of Hermiane
circulation of writings, 320, 321,
3434
disillusionment of, 329
parallels with Maximus Confessor, 356
Three Chapters and, 31920, 321, 3224
Faustinus of Tebessa (bishop), 33940
Felix (poet), 139, 147, 172, 376
Ferrandus, 106, 31819, 326, 329, 344,
345
figs, 91, 99, 141, 289
Firmus (primate of Numidia), 322
flamines perpetui, 46, 137, 156
Florentius (poet), 56, 1334, 147, 376
foederati
Byzantine commanders of, 2067, 2323
ideology of barbarian and, 261
Moorish, seizure of frontier zone and,
277
Vandals as, 21
fortifications
Byzantine
cities and, 332
epigraphy and, 2456, 24950, 332
frontiers and, 2067, 279, 287, 299
locals and, 24950
private, 299
late Roman, 283
Moorish, 288, 289
Vandal poetry and, 456, 56
Franks, see also Charlemagne; Childebert II;
Ingundis
dating systems of, 150
Procopius on, 257, 279
Fridamal, 48
friendship, diplomatic, 334
frontier zone
careers of Byzantine officials and, 21819
cultural identity in, 2804

origins of Byzantine officials and, 2024,

2079
Romans and local powers in, 5
Fulgentius of Ruspe
age of, on travels, 74
anti-Arian writings, 174
beating of, near Sicca, 165, 179
circulation of books and letters and, 1057,
109, 174, 344
communications during exile, 1057
correspondants of, 1057
early career of, 141, 143
evades ordination, 165
exile to Sardinia, 100, 103
family of, 70, 85, 1378, 145, 246
founder of monastery, 138
knowledge of Greek, 246
letter-collection of, 104
monastic conversion of, 168
recall from exile, 163
Redemptus (monk) and, 74
regnal dating and, 157
Romanness in, 1901
Thrasamund and, 157, 174, 176, 178, 195
travels of, 81, 84, 89, 95, 100
Fulgentius the Mythographer, 556, 1867,
299, 339, 340, 376
Fundus Tuletianos, see Albertini Tablets
Gafsa, 40, 97, 98, 363, 370
Galicia, 37
gardens, 48, 54, 567
Garmules (Moorish king), 232
garum, 99, 100, 289
Gaul
African saints cults in, 112, 114
anti-Arian literature in, 1756
circulation of manuscripts in, 109, 345
exports from Africa to, 94, 336
Gallo-Romans in administration of, 11
illict slave trade in, 300
imports into Africa from, 94
Vandal invasion of, 2
Gebamund (Vandal prince), 534
Geiseric
anti-Arian treatise sent to, 175
as Antichrist, 171
consolidation of power, 67
conversion to Arianism, 185
external relations
with empire, 224, 2932, 33, 356
with Huns, 37
with Moors, 36
with Odoacer, 38
with Visigoths, 245, 37

427

Index
Geiseric (cont.)
property settlement of, 478, 137
religious policies of, see persecution
Roman advisers of, 87, 144, 167
royal ideology of, 201, 183
sack of Rome (455), 268, 29
Vandal succession and, 256, 177
Gelasius (Pope), 75, 7980, 83, 89, 1256
Gelimer
as refugee among Moors, 37, 55
capture of, 254
end of the Vandal kingdom and, 308, 313,
315
Justinian and, 31415
officials of, 61, 134, 139
regnal dating, 152
relations with Ostrogoths, 42, 315
relations with Visigoths, 378, 42, 315
succession of, 312, 314, 315
titulature, 42, 44
Geminius Catullinus, Flavius, 137, 141, 282,
283
Gennadius (exarch)
Gregory I and, 229, 324, 328, 347
kills Garmules, 232
term of appointment, 2245, 234
George (Palestinian monk), 365
George (Praetorian Prefect), 2356
George of Cyprus, 331, 332
Germanus (magister militum), 205, 218, 222,
228, 233, 237
Ghirza, 267, 268, 283
Gisaleic, 378, 39
Godas, 61, 313
Gordian (Carthaginian senator), 48, 70, 74, 85
Goths, see also Godas; Ostrogoths; Visigoths
apocalypticism and, 272
Arianism of, 185
entertainments of, 634, 142
in late Roman army, 49, 68
Vandals as, 62, 179
Gotthaeus (ambassador), 37, 412
grain, see also barley; wheat
African exports
to Constantinople, 92, 336
to East, 31415
to empire, 23, 91, 357
to Italy, 1, 12, 23, 92
imports into Africa, 86, 366
Moorish consumption of, 2634
grammarians (grammatici), see also Corippus;
Pomerius; Priscian
Byzantine, 198
late Roman, 188, 2734
Vandal-era, 55, 134, 135, 141, 146

Greek (language), 63, 1278, 171, 2446, 341,

363, see also epigraphy, Greek-language
inscriptions
Gregory (rebel)
death, 232, 355
family of, 203, 222, 243
Maximus Confessor and, 355, 356
previous career, 225
rebellion of, 355
Gregory I (Pope)
African church and, 3249, 330, 348
confessors of Tipasa and, 78, 309
cult of relics and, 326, 342
Gennadius and, 229, 324, 328, 347
Innocent and, 2034
perceptions of Moors, 272
Three Chapters and, 324
Gregory of Tours, 86, 128, 172
Gregory VII (Pope), 362, 369
Gunthamund
Dracontius and, 143, 183
external relations of, 33, 38
recalls exiled bishops, 162
regnal dating, 21, 1525, 156
titulature, 44
Guntharis (rebel), 216, 219, 2234, 230, 302
Hadrian (abbot), 341, 352
Hadrian I (Pope), 366
Hadrumetum
communications hub, 88
foreigners in, 87, 178, 34950
Islamic period, 364
market, 311
Moors and, 271, 339, 352
notables, 98, 339, 352
renamed Unuricopolis, 45
urban continuity in, 331
Hadra, 164, 212, 2479
Hasdings, see Vandals
Heraclius (emperor), 235, 243, 301, 354, 355,
357
Heraclius (exarch), 203, 219, 234, 243,
355
Heruls, 206, 257, 259
Hilderic
Arian persecution ended by, 170
as scion of Theodosian house, 434
children of, 34
coup against, 65, 314
in Constantinople, 34
relations with empire, 34, 31314
relations with Ostrogoths, 40
succession of, 177, 178
supporters of in Constantinople, 314

428

Index
Hippo Regius
Arianism in, 160
barbarian jewellery at, 50
Byzantine army in, 247
eastern migrants to, 349
Islamic-era bishopric, 362, 369
Punic spoken in, 187
refugee bishops at, 68
Roman army in, 49, 68
Vandal refugees in, 255
Vandal siege of, 68
Vandal treasure captured at, 38
Hoageis (Vandal prince), 54, 147
Hoamer (Vandal prince), 55
Hormisdas (Pope), 80, 104, 107, 191
houses, private, 48, 54, 133, 299
Huneric
anti-Arian treatise sent to, 175
convenes Council of 484, 46
death, 156
eastern Arian church and, 184
hostage in Ravenna, 2930
legislation of, 45, 138, 1689
marriage to Eudocia
effected by Geiseric, 27
engagement, 223
ideological significance of, 434
recognition of, 24, 31, 38
threatened, 278
Vandal succession and, 256
marriage to Visigothic princess, 245
panegyric to, 147
political purge under, 1778
princely household of, 144
relations with empire, 323, 35
relations with Moors, 367, 287
religious policies of, see persecution
renames Hadrumetum, 45
royal ideology of, 45, 183
titulature, 44
Vandal succession and, 1778
Huns, 37, 206, 259, 270, see also Agan; Attila
hunting, 53, 56, 61, 87, 136
Hydatius, 27, 185

African saints cults in, 11213

Africans in administration of, 11
Africans in church of?, 1246
anti-Arian literature in, 175
Arianism in, 160
Byzantine administration of
Armenians in, 209
assimilation of troops, 241
exarchs, 219, 234
locals in, 2001, 247
Praetorian Prefects, 235
reposting from Africa, 233, 237
senior officers, 237
subordinate appointments, 228
Christians of Ifriqiya and, 3667
circulation of manuscripts in, 10710,
3445
consular dating in, 14850
East Roman attacks on, 39
eastern names in, 211, 213
Italo-Roman elite and Africa, 1056
regnal dating in, 150
Romanization of, 5
saints cults of, among Africans, 104,
342
Three Chapters and, 324
Vandal attacks on, 30, 31
Vandal claims to, 31
Vandal slaving in, 91

Ibas of Edessa, see Three Chapters Controversy

Ibn c Abd al-Hakam, 243, 280, 335
Ingundis (Frankish princess), 348
Innocent (Praetorian Prefect), 203
Isidore of Seville, 109, 175, 185, 269, 309,
345
Islamic conquest, see Africa, Islamic
Italy, see also Rome; Naples; Ravenna
African exports to, 23, 934, 3367
African names in, 11527

Jerba, 99, 334

Jerusalem
Charlemagnes charity to, 366
treasures from Temple of, 44
jewellery
as booty, 44
barbarian, 4950, 52
brooches, 279
earrings, 264, 2812
gold, 56, 279
Jews, 173, 180, 321, 348, 354,
361
John Malalas, 24, 33, 308, 314, 315
John Moschus, 351, 352
John the Lydian, 80, 308
John Troglita
Balkan origins of, 202
Corippus and, 338
dux in Africa, 207
dux Mesopotamiae, 219
family of, 232, 242, 243
Procopius and, 224
service in Africa, 224
term of appointment, 233
Jordanes, 25, 37, 183, 308

429

Index
Jucundus (patriarch of Carthage), 165,
178
Junillus, 106, 246, 338, 3401, 376
Justin I, 34, 31314
Justin II, 200, 260, 323, 327, 338
Justinian
African church and, 3203
African legislation of, 1968, 2401, 254,
307, 3201
African refugees and, 80
Antalas and, 300
children of Hilderic and, 34
cult of saints and, 342
ecclesiastical architecture and, 333
Egypt (administration) and, 201
fortification of Africa and, 279, 299
Gelimer and, 31415
Hilderic and, 34
Italy (administration) and, 200
legitimation of reconquest, 645, 1967,
30712
militarism of, see militarism
Moors in ideology of, 2535
rebels against Vandals appeal to, 313
reorganization of Africa, 1978, 2401
suppression of Arianism, 3201
Three Chapters and, 316, 321
titulature, 61
Kahina, 305
Kairouan, 358, 363, 364, 365
Kusayla, 280, 294, 305
Laetus (bishop), 309
Laguatan, 267, 277, 279, 298, 299, 301
Lambaesis, 99, 283
lamps, 92, 94, 333
Lares, see Sicca
Late Roman C, see Phocaean red slip ware
Lateran Synod (649), 357
Latin
African v. Italian, 80
Arian bishops preach in, 182
as index of Romanization, 7
brief on eastern theology in, 318
Byzantine elite and, 244
eastern liturgical manuscripts in, 366
education in, 341
epigraphic language, 363, see also epigraphy
in Constantinople, 80, 8990
language of diplomacy, 41, 294
poetry in, see poetry
Punic identity and, 189
spoken language, 63, 263, 363
Latin Anthology, 54, 63, 271, 281

Laurence, St, 286, 342

law
canon, 68, 326, 327, 369
form v. social practice, 64
Roman
captive citizens in, 300
codification of, 8
ideology of reconquest in, 1967, 307
imperial legislation for Africa, see Justin II;
Justinian; Valentinian III
in Vandal kingdom, 138, 282
of marriage, 24
Punic language in, 189
soldiers in, 311
suppression of heresy in, 168
Romanness and, 4, 7
Vandal, 45, 1389, 1689, 1812, 183
lawyers, 134, 138, see also Dracontius; Junillus
Lazi, 208, 256, 278
Leo I (emperor), 312, 33
Leo IX (Pope), 368
Leontius (emperor), 358
Leptis Magna, 71, 342
letters, see also Cassidorus; Fulgentius of Ruspe;
Gregory I; Gregory VII; Leo IX;
Theodoret of Cyrrhus
as rhetorical device, 298, 31112, 314
circulation of
AfricaConstantinople, 222, 235, 319,
320, 3456, 357
AfricaRome, 104, 31819, 3478, 357
AfricaSicily, 367
AugustineJerome, 74
ConstantinopleRome, 7980
SardiniaConstantinople, 1067
within Africa, 104, 167, 3467
Liberatus (deacon), 318, 323
Liberatus Caecilides, 21011, 237, 242
Libyan, 1867, 196
Lilybaeum, 39
limes, see frontier zone
literacy, rural, 134
literary circles, 1356
Lixus, 289
Luxorius
Arianism and, 376
blackness in, 2701
elite lifestyle in, 535, 136
Hilderic and, 434
injustice in, 139
literary circle of, 135
Punic identity and, 1867, 189
status and family of, 1456
Vandal regime in, 1478
Vandalic language in, 63

430

Index
Madauros, 102, 157, 163
magister militum Africae, 198, see also
administration, Byzantine; Artabanes;
Belisarius; Germanus; Sergius;
Solomon; etc.
Majorian (emperor), 31
Malchus, 35, 57, 58
Mallorca, see Balearic Islands
Mancian tenure, 138
Manicheans, 180, 184
manuscripts
African community in East and, 366
circulation of, 10710, 3435
Marcellinus comes, 77, 3078, 312
Marcian (emperor), 28, 30, 35
Maria (enslaved aristocrat), 723, 74
markets, 99, 311
Marseilles, 94, 121, 336, 348
Martianus Capella, 1089, 135
Martina (empress), 2356
martyrs/martyr cults, see persecution, Arian;
relics; saints, cult of
Massa Candida, 113
Masties, 269, 278, 280, 287, 293
Masuna, 276, 278, 280, 293
Mauretania
Arian persecution in, 163
Byzantine, 197, 198, 207, 287
Christianity in, 2689, 286, 2913, 362, 368
communications with Rome, 286
dating systems in, 151
economy, 91, 2878, 289
exile in, 102
location of, 9
Moorish seizure of, 2767
Moors of, 252, 255, 278, 279
refugees in, 56, 85, 285, 301
retroceded to empire, 22
Roman identity in, 274, 2846, 28992
RomanBerber intermarriage in, 295
spread of saints cults from, 112, 113
Vandal control of, 21, 36, 2867
Maurice (emperor), 234, 328, 349, 350
Maurice (general), 232, 2434, 247
Maximus Confessor, 235, 345, 347, 351, 3557,
359
medicine, 54, 109, 134, see also doctors
merchants, 71, 89, 169, 348, 349, 354, 365
militarism, 2534, 261
millenarii, 48, 60, 64
monasteries, see monks
monks, see also Fulgentius of Ruspe; Maximus
Confessor
Arian persecution and, 102, 162, 16970
Arian, in Egypt, 184

conversion to Arianism among, 172

correspondants of Fulgentius of Ruspe,
1057
Decameron and, 370
discipline of, 325
Dythelete, 345
eastern, in Africa, 878, 178, 351, 365
Islamic-era, 364
of Gafsa, 98, 170
overseas travel, 68, 352
Scythian, 1067
Three Chapters and, 323
travel within Africa, 98
travellers from Africa as, 85, 341, 352
Monoenergism, 353
Monophysites/Monophysitism, 235, 316, 351,
353, 356, see also Eutychianism
Monotheletism, 345, 351, 353, 356, 357, 359
Moors
blackness of, 26971
Byzantine subsidies to, 298
Christianity among, 2679
demonic nature of, 2712
evangelization of, 166, 267, 268
legendary history of, 2612
paganism among, 2678
raids of, 297300
receive Amalafrida, 40
relations with empire, 255, 27880, 297302
relations with Vandals, 367, 56, 102, 166
Roman perceptions of, 26173
symbols of power, 36, 2789, 283
take over African interior, 2758
term Afri barbari and, 274
term Mauri and, 2735
term Moors and, 16
titulature, 276, 278, 279
Vandal-era authors and, 255
mosaics, 21, 53, 54, see also Bordj Djedid,
mosaics; Tabarka
Mouzaaville, 102, 163, 276
names, see also Adeodatus; Benenatus;
Quodvultdeus; Sergius
African
distinctiveness of, 11516, 11819
Italy and, 1267
movement of, 11427
Arabic, 206
Berber, 290, 2934
Byzantine, in Africa, 21114
ethnic identity and, 523, 2945
Germanic, 49, 248, 249
Greco-Roman, 2923, 295
Greek, 248, 249, 250

431

Index
names (cont.)
Islamic, 364
Latin, 24850, 283, 290, 294
mixed, 53, 2945
Punic, 119, 290, 292, 294
Roman
as index of Romanization, 7
changes in patterns of, 6
Naples, 85, 93, 105, 108, 112, 336
Nestorianism, 316, 31819, 320
nomads, 258, 259, 263, see also pastoralists
Notitia provinciarum et civitatum Africae, 36, 157,
292
Numidia
Arian persecution in, 163, 167
Byzantine
as province, 197
fortification of, 206, 287, 299
military administration, 198
military careers and, 2256
military unrest in, 207, 232
officers in, 347
recruitment in, 241
subordinate officers in, 207, 229
Christianity in, 2689, 286, 2913
church of, and Gregory I, 3247, 3289
communications with Rome, 288
dating systems, 150, 284
illicit slave trade in, 299
location of, 7
Moorish attacks in, 299
Moors of, 252, 255, 278, 279, 297
provincial church council of, 322
Punic language in, 187
refugees settled in, 85, 285
retroceded to empire, 22
Roman administration of, 9
Roman identity in, 2846
Vandal control of, 212, 36, 2867
Vandal royal estates in, 47
warlords in, 275, 277
nuns, 169, 172, 235, 351
nuts, 99, 289
Odoacer, 38
olives
cultivation of, 97, 99, 141, 287,
335
groves of, used for ambushes, 266
oil
amphorae and, 91
assessment of, 341
bottling of, 51, 140
exports from Africa, 92, 93, 334, 335
imports into Africa, 366

production of, 91, 99, 133, 289, 333

skins and, 334
trade within Africa, 98, 100
Olybrius, Anicius, 24, 278, 31, 32, 105
Ostrogoths, see also Amalafrida; Amalasuntha;
Theodoric
Italo-Roman elite under, 105, 2001
Procopius attitude toward, 258
regnal dating, 150
relations with Vandals, 34, 3842, 315
relations with Visigoths, 25
Oudhna, 96, 98
pagans, 173, 180, 188, 321, see also Moors,
paganism among
panegyric, 434, 56, 1468, 278, 338
Pantaleon, St, 212, 343, 367
Papacy, see also letters; Rome; travellers;
individual Popes
appeals to, 320, 3267, 3289
Islamic Africa and, 362, 3679
Parthemius presbyter, 55, 104
Passio septem monachorum, 156
pastoralists, 99, 277, 283, 289
Paul (Numidian bishop), 3289
Paul II (patriarch of Constantinople), 346, 357
Perpetua, St, 11213, 163
persecution, see also Donatists; Jews;
Manicheans
Arian
attrition of Nicene sees as, 1645, 166
Byzantine reconquest and, 307, 309, 314
confiscation of property as, 1634, 168
end of, 170, 313
enticements as, 1678, 172
exile as, 1023, 1623, 165, 166, 168
goals of, 162, 170, 184, 185
Gregory of Tours and, 128
intensification under Huneric, 76, 1689
interdict of Nicene liturgy as, 181
local violence and, 179
martyrdom (execution) and, 162, 165,
166, 167, 170, 179
refugees from, see refugees
restriction of office-holding as, 1667, 168
Roman models for, 1689
Romano-African response to, 1712, 193
secular poets and, 376
social degradation as, 167
social relations and, 1656, 1801
translation of relics and, 11014
imperial
of Arians, 3201
of hard-line Chalcedonians, 3213, see also
Maximus Confessor

432

Index
Persians, Persian Empire
apocalypticism and, 359
besiege Constantinople, 351
conquests in East, 353, 357
embassies to, 35
imperial frontier with, 56, 2023, 233,
255
Maximus Confessor as son of, 351
Procopius attitude toward, 258
refugees from, 351
service in wars against, 21819, 2256, 229,
230
Peter (magister militum), 347, 356
Peter, St, 326, 342
Petronius Maximus, 278, 29
Phocaean red slip ware, 92
Phocas (emperor), 229, 2423, 301, 355
Phuscias (ambassador), 37, 41
Phylarchus (ambassador), 31
Placidia, 24, 278, 31, 32
poetry
Arian persecution and, 376
Byzantine-era, see Corippus; Fulgentius the
Mythographer
Carthage in, 21, 56, 1334
exchange of complements in, 55
Frankish, African Christianity in, 369
inscriptions and, 288, 341
panegyric, see panegyric
provincial, 281
social advancement and, 146, see also
Corippus
St Cyprian in, 113
terms of Vandal-Roman accommodation in,
1412, 190, 193
theology and, 170
Vandal sensitivity to, 55
Vandal-era, see Dracontius; Latin Anthology;
Luxorius; etc.
Vandals and law in, 139
victory-ideology in, 45
Virgil, see Virgil
weddings and, 545
Pomerius (grammarian), 75, 86, 89
Pontianus of Thaenae (bishop), 318
Possessor of Zabi (bishop), 73, 80, 104, 107
Praejecta (niece of Justinian), 204, 233, 242, 243
Praetorian Prefects (Byzantine), see also
Archelaus; Athanasius; George;
Innocent; etc.
death in service, 232
officium of, 1978
prior experience, 205, 226
recall of, 2356
term of appointment, 225, 2345

Primasius of Hadrumetum, 322, 323, 338,

3445, 346
Primosus of Carthage (bishop), 322
Priscian (grammarian), 75, 80, 89, 376
proconsul
dating and, 151
palace of, 44
Roman, 145
Vandal, 46, 144, see also Victorianus
Procopius
as source for Africa, 57, 2556
audience of, 312
barbarians in
diet, 256
faithlessness, 257
human sacrifice, 267
major v. minor barbarians, 2589
scarcity, poverty, 256
simplicity, 259
violence, 2567
Byzantine officers in
Areobindus, 220
Artabanes, 242
Belisarius, 222
John Troglita, 219, 224
regional origins of, 202, 2056
Sergius, 223, 224, 228
Solomon, 223
Byzantine reconquest in, 64, 254, 308, 309,
31011
captive-taking in, 300
confessors of Tipasa and, 778
ConstantinopleCarthage route in, 21415
flight from Africa, 301
gardens in, 54
Greek childrens chant in, 245
intellectual concerns of
diplomacy, 34
idea of city, 332
tyranny, 311
urban lifestyle, 57
Junillus and, 246
Justinian in
building programme, 333
demonic nature, 373
legitimation of reconquest, 310, 31112
St Prima and, 342
suppression of Arianism, 321
Moors in
battle-gear, 265
blackness of, 269, 270
clothing, 2645
diet, 2634
military tactics, 2667
military uprisings, 2978

433

Index
Procopius (cont.)
not properly human, 263, 273
Phoenician people, 2612
Phoenician tongue, 2623
poverty, hardiness, 263, 303
symbols of power, 2789
VandalMoorish relations, 36
Palestinian origins of, 210
Romanness in, 190, 196
Vandals in
Amalafrida affair, 40
Arian persecution, 168
decadence, 568
distinctiveness, 59, 62
exporpriations, 137
Gelimer, 38, 314
hunting, 53
kingship, 183
relations with empire, 33, 34
revolts against, 313
Roman support for, 313
Proculus (aristocrat), 1445
property
Byzantine troops and, 2412
expropriation of, 478, 1634, 1678, 3201
of heretics, in Roman law, 168
Romano-Africans and, 1378
Vandals and, 478, 241
Prosper of Aquitaine, 166
Pudentius, 210, 211, 232, 237
Pulcheria, Aelia, 28
Punic (language and identity), 1869, 262, 270
Pyrrhus (patriarch of Constantinople), 351, 355
Quodvultdeus (name), 1267, 283, 293
Quodvultdeus of Carthage
age of, 75
Arian persecution in, 171, 172, 173, 1767
banishment of, 67, 73, 85
Moors in, 272
Romanness in, 191
Ravenna, 85, 89, 110, 122
rebaptism, 84, 177, 179, 307
refugees
from Byzantines, 2545
from Muslims, 125, 352
from Persians, 351
from Vandals, 6886, 8890, 114, 285
royal, see Gelimer; Gisaleic
spread of names and, 126
spread of relics and, 11014
Reginus (ambassador), 33
Reginus comes (correspondent of Fulgentius),
107, 329

relics, see also Augustine of Hippo; St Cyprian;

saints, cult of
African, 326, 367
attitudes toward, 114
contact relics, 104, 112
deposition of, 340, 342, 343
eastern, 286, 341, 343
Roman, 286, 342
resistance
to Byzantine empire, 31630
to Byzantine reconquest, 2545, 313
to Papacy, 32430
to Roman empire, 34
Ricimer, 31
Romanization (concept), 5
Romanness
Christianity and, 1903, 295, 354, 361
communications and, 3, 1214
fracturing of, 131
loyalty to empire and, 3534, 361
modern perspectives on, 39
Moors as rhetorical inversion of, 26173,
274, 3023
Procopius and, 196
Salvian of Marseilles and, 58
secular African authors and, 186, 18990
slavery and, 73
Theodoret of Cyrrhus and, 723
Vandals as assailants of, 1301
Romano-Africans
Afariqa, 358, 363
appeal of Vandal identity to, 62, 142
definition of, 1516
identification with Byzantine empire,
33840, 358, see also Corippus
in Moorish kingdoms, 277, 28496
new elite among, 1446
service in Vandal administration, 1434
Rome
African exports to, 92, 93, 336
African names in, 11926
African refugees in, 114
African saints cults in, 11213
Africans in church of?, 1246
Constans II and, 357
consular dating in, 14850
cultural melting pot, 5
gens Aradia and, 145
manuscripts in, 108, 344
Numidia and, 288, see also Gregory I
officials from, 910
population decline in, 93
Romanness and, 190
saints cults in Africa, 286, 342
Vandal sack of (455), 268, 29, 44, 261

434

Index
Romulus, 190
Rufinus (Byzantine officer), 206, 207, 230,
232, 300
rural society, 259, 2989, 313, see also Albertini
Tablets; economy
sailors, 71
saints, cult of, 11014, 291, 309, 3413, 366,
367, see also relics; Virgin; individual
saints
Sallust, 304
Salvian of Marseilles, 49, 578, 184, 267
San Antonino di Perti, 336
Sardinia
Augustines relics and, 111
Byzantine
administration of, 197, 198, 359
flight from Africa to, 352
reconquest of, 206
Muslim attacks on, 367
Vandal
administration of, 61
communications hub, 1023
exile to, 103, 162, 165, 168
exiles recalled from, 163, 170
rebellion of, 313, 315
Symmachus and, 104
Saturus (procurator domus), 144, 167
Sbetla, 132, 152, 247, 250, 331, 341
schism, 81, 114, 367, 376
Sebastian comes, 87, 144, 167
Segermes, 51, 99, 334
Senarius (Ostrogothic ambassador), 39, 41
Septem, 198, 207, 240, 287, 342, 350
Sergius (Byzantine official)
appointment of, 223
as dux Tripolitaniae, 224, 228, 298, 299
ineptitude of, 216
Mesopotamian origins of, 202
recall of, 233
Sergius (name), 21213
Setif, 99, 285, 286, 287
Severus (ambassador), 32, 356
sexuality, 57, 136, 138
Sicca, 100, 102, 179, 287, 331
Sicily, see also Syracuse
cession to Odoacer, 38
exile to, 168
Muslim attacks on, 358, 367
Vandal attacks on, 31, 38
Vandal claims to, 31, 38, 39
Vandal slaving in, 91
Sidi Marzouk Tounsi, 93, 96, 140
sigillata, see African red slip ware
Sigisteus comes, 55, 104

skin colour, 62, 26971

slaves, see also Elias the Younger; Maria
accompany refugees, 73
African trade in, 91, 99
Albertini Tablets and, 98, 270
Arian persecution and, 1656, 167
armed skirmish over, 300
as Vandal officials, 61
counted among Vandals, 19
diet, 264
eastern markets for, 72, 91, 256
Exarch Gregorys daughter as, 243
illicit trade in, 299300
imperial subjects as, 279
in Byzantine army, 313
Mauretania as source of, 91
Maximus Confessor as son of, 351
Moors as, 269, 271, 299
redemption of, 32, 36, 72, 73
Roman captives as, 26, 36, 51, 91, 299,
335
Romanness and, 73
Romano-Africans as, 67, 130
skin colour and, 26970, 271
Vandal women and children as, 241, 255
Solomon
appointment of, 223, 233
Belisarius and, 223
death, 233, 300
eunuch, 228
flight from Africa, 233, 301
Mesopotamian origins of, 2023
military career of, 206, 219
Moors and, 233, 298, 299, 300
overhaul of army, 240
Stotzas revolt and, 233, 242
terms of appointment, 233
uncle of Sergius, 223
Vandal policy of, 2412
Sophronius of Jerusalem, 351, 352
sortes Vandalorum, 181
Spain
African exports to, 94, 3367
African saints cults in, 11214
anti-Arian literature in, 175
circulation of manuscripts in, 108, 109,
345
dating systems, 150
diversion of wheat from, 86
saints cults from, in Africa, 3423
Vandal invasion of, 2
Vandal settlement in, 21
Vandal slaving in, 91
Vandal treasury to be sent to, 38
St Catherines, Sinai, 366

435

Index
Stotzas
called tyrannus, 216
death of, 232
followers of, 302
revolt of, 207, 232, 242, 321, 339
seeks refuge with Moors, 301
succession
imperial, 289, 302
Vandal law of, 26, 65, 177, 178
Sueves, 25, 37
Symmachus (Pope), 104, 11415, 126
Syracuse, 357
Syria, 91, 337, 366
Syriac (language), 263
Syrians, in Byzantine administration, 203, 208
Tabarka
Byzantine post and, 245
mosaics, 265, 2812
Vandal settlement, 48
Tarragona, 74, 336
Tatianus (ambassador), 31, 35
taxation
Byzantine, 198, 199, 241, 327, 336
late Roman, 99, 284
Vandal, 98
Temple of Jupiter Optimus Maximus, 26, 44
Thalassius (theologian), 351
theatres, 56, 578, 133
Thelepte, 99, 100, 141, 331
Theoderic (Vandal prince), 26, 55, 144, 1778
Theoderic (Visigothic king), 245
Theodore of Mopsuestia, see Three Chapters
Controversy
Theodore, St, 343
Theodoret of Cyrrhus
African refugees in, 70, 723, 813
Three Chapters and, 83, 316, 318
Theodoric (Ostrogothic king), 34, 3840, 41,
86, 150, 258
Theodosian house
extinction of, 289, 32
Petronius Maximus and, 279
Vandals and, 224, 256, 29
Theodosius II, 23, 28, 29
Thrace, 184, 218, 231
Thracians, 202, 2058, 2301, 247
Thrasamund
building programme, 45
Fulgentius of Ruspe and, 157, 163, 174, 176,
178, 191
Gisaleic and, 38
panegyric to, 56, 147
regnal dating, 152, 155, 157
relations with empire, 34

relations with Ostrogoths, 3840

religious policy, see persecution
victory ideology and, 45
Three Chapters Controversy, 83, 31624, 329,
3434, 346, 376
Tiberius II, 249
Tiberius III Apsimar, 358
Tipasa
confessors of
Byzantine reconquest and, 307, 309
flight to Spain, 73, 86
in Constantinople, 73, 768, 129,
376
venerated as saints, 112
port city, 88, 287
Trastevere, 114
travel
along coast, 98, 100, 102
combined land and sea, 100
Constantinople-Africa, routes, 21415,
216
interior to coast, 956
overland, 93, 97, 214
planned travel, 74, 81
travel times, 21415, 216
within interior, 979, 1002, 289
travellers, from Africa
refugees, see refugees
to Alexandria, 323
to Anatolia, 81
to Balearic Islands, 288, 323
to Carthage, 956, 98, 288
to Constantinople
flight from Moorish raids, 352
Hilderic, 34
imperial patronage and, 338
on ecclesiastical business, 346
supporters of Hilderic, 314
testicular ailment and, 341
Three Chapters and, 3212, 323
Vandal-era, 7680
via Rome, 329
via Syria, 82
to Gaul, 86, 352
to Italy, 845, 341
to Jerusalem, 32, 81, 352, 365
to Macedonia, 81
to Rome
appeals to Papal court, 3267, 3289
flight from Arabs, 352
flight from Vandals, 68
late Roman, 11
on ecclesiastical business, 348
Vandal-era, 834
to Sardinia, 1023, 168, 352

436

Index
to Sicily, 81, 85, 168, 301
to Spain, 73, 86, 352, 365
to Syracuse, 71
to Syria, 72, 74, 79, 813
uncompleted travel, 74, 327
Vandal-era
ages, 34, 745
chronology, 75
companions, 734
enslaved, 72
gender, 712
motivations for travel, 889
religious v. lay, 689, 71
social status, 6871
travel horizons, 8890
via Syracuse, 81, 327
travellers, to Africa
enslaved, 365
exiles, 350
Frankish, 3489, 3656
from Asia Minor, 349
from Constantinople, 349, 351
from East, 878
from Egypt, 87, 270, 34950,
351
from Jerusalem, 365
from Rome, 286, 355
from Sicily, 365, 367
from Spain, 867
from Syria, 348, 3501
monks, 878, 365
via Constantinople, 348
travellers, via Africa
EgyptRome, 365
GaulConstantinople, 3489
JerusalemSpain, 365
SpainConstantinople, 348
Tripolitania
Byzantine, 197, 198, 207, 212,
279
Christianity in, 268, 3645
economy and society, 99, 2889
Islamic-era, 3645
late Roman, 5
Latin in, 363
Moorish, 277, 279, 2889
paganism in, 267
poetry in, 281
Punic language in, 188, 189
Vandal, 22, 313
Typos, 356
Ucutamani, king of, 269, 278
Uranius (ambassador), 33, 35
urban life, see cities

Valentinian III
African legislation of, 85, 275, 2856
death, 279, 2867
Hilderic as grandson of, 34, 434, 147
relations with Vandals, 213, 27, 2930, 164,
375
Vandalic (language), 634, 182
Vandals
aristocratic lifestyle and, 525, 567
conquest of Africa, 1312
conversion to Nicene Christianity among,
180
deportation of, 242, 255
deportees return to Africa, 215
foreign policy
diplomatic relations, 2143
Theodosian house and, 224, 29, 301,
32, 345
identity
Africa and, 56
distinguishability, 623
military service and, 60
social status and, 601
kings
Carthage and, 201, 56
conversion hoped for, 1746
ecclesiastical responsibilities of, 1834
imperial pretentions of, 437
sacral ideology of, 183
Theodosian house and, 234, 256, 32,
434
titulature, 42, 44, 277
numbers, 19
property ownership among, 478, 241
providential history of, 183
raids of, 30, 36, 50
relations with Romano-Africans, 31213
ruling class of, 612
sack of Rome (455), 269
settlement in Africa of, 4950
succession, law of, see succession
treaties with empire
treaty of 442, 212
treaty of 474, 32
women in Byzantine Africa, 2412, 255
Verecundus of Iunca, 322
Victor of Tonnena, 28, 78, 309, 312, 323,
344
Victor of Vita
Arian persecution in
as Babylonian Captivity, 1301
attrition of Nicene sees, 166
exile, 162
intensification under Huneric, 169, 1801
local violence and, 179

437

Index
Victor of Vita (cont.)
martyrdom (execution), 165, 167
passports and, 179
scalpings, 59
social degradation, 167
background of, 134, 161
barbarian dress in, 602
confessors of Tipasa and, 77
Hunerics political purge in, 177
ideology of barbarian in, 1301
Moors in, 36, 264, 268
Nicene loyalties in, 177, 178
regnal dating in, 156
Romanness in, 1913
Romano-African office-holders in, 137,
144
Roman-Vandal accommodation and, 180
Vandal conquest in, 67, 132
Vandal injustice in, 139
Vandal millenarius in, 48
Vandal succession in, 1778
Vandalic language and, 63
Victorianus (proconsul), 98, 144, 167
Vigilius (Pope), 320, 324
Vigilius of Thapsa, 109, 173
villas, see estates; houses
Vincent, St, 332, 342
Virgil, 54, 135, 190, 341
Virgin, cult of, 307, 342

Visigoths, see also Athanagild; Euric; Gisaleic;

Spain; Theoderic
Arianism and, 15960, 175, 176
regnal dating, 150
relations with Vandals, 245, 378, 412,
315
Vivarium, see Cassiodorus, monastic library of
Volubilis, 289, 294
warlords, 2757
water-diviner (aquilex), 83, 89
wheat, 86, 256, 264, 289, see also grain
whiteness, see skin colour
wine
barbarians and, 256, 263
exports from Africa, 1
imports into Africa, 93, 366
trade within Africa, 99, 100
Zachariah of Mitylene, 263, 308,
309
Zarai, 99, 283
Zeno (emperor)
confessors of Tipasa and, 77
embassies to Vandals, 323, 35
Nicene clergy and, 32, 33, 163, 165, 184,
375
Zeno of Verona, 113
Zeugi Carthago, see Africa Proconsularis

438