Staying Roman: Conquest and Identity in Africa and the Mediterranean, 439-700 9780521196970, 2011047925

Citation preview

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 empire’s 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.

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:15:34, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:15:34, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

STAYING ROM AN Conquest and Identity in Africa and the Mediterranean, 439–700

JONATHAN CONANT

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:15:34, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, S˜ao 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, 439–700 / 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.

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:15:34, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

To Vanessa

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:15:34, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:16:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101

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 empire’s 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

2

Augustine, ‘Enarrationes’ in psalmos 58.1.21, ed. E. Dekkers and J. Fraipont, CCSL 38–40 (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

1

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 empire’s 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 533–4, 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

3

4

5

et le passage e´ conomique du Bas-empire au moyen aˆ ge’, in Morfologie sociali e culturali in Europa fra tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sull’alto Medioevo 45 (Spoleto, 1998), pp. 35–122. The classic studies are L. Schmidt, Geschichte der Wandalen (2nd. edn; Munich, 1942) and C. Courtois, Les Vandales et l’Afrique (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, L’Afrique byzantine: histoire de la domination byzantine en Afrique (533–709) (Paris, 1896). On the Moors, see Y. Mod´eran, ´ Les Maures et l’Afrique romaine (IVe–VIIe si`ecle), Biblioth`eque des Ecoles franc¸aises d’Ath`enes et de Rome 314 (Rome, 2003). See now W. Kaegi, Muslim Expansion and Byzantine Collapse in North Africa (Cambridge, 2010).

2

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

Introduction Roman’ by actively seeking to ensure the region’s 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 R´esistance africaine a` la romanisation, the Algerian scholar Marcel B´enabou 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, B´enabou’s 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. Frend’s The Donatist Church and Christian Courtois’s Les Vandales et l’Afrique.9 Both 6

7 8

9

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 300–900 (Cambridge, 2001); and C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford, 2005). See also, on a still wider canvas, B. Cunliffe, Europe between the Oceans, 9000 bc–ad 1000 (New Haven, Conn., 2008). M. B´enabou, La R´esistance africaine a` la romanisation (Paris, 1976). See, e.g., G. Woolf, Becoming Roman: The Origins of Provincial Civilization in Gaul (Cambridge, 1998), pp. 19–20 and G. Woolf, ‘Beyond Romans and Natives’, World Archaeology 28/3 (1997), pp. 340–1. W. H. C. Frend, The Donatist Church: A Movement of Protest in Roman North Africa (Oxford, 1952); for Courtois, see above, n. 3.

3

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 empire’s 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 metropolis’s 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, Rome’s 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, Vergil’s Aeneid and the Roman Self: Subject and Nation in Literary Discourse (Ann Arbor, Mich., 2005). Citizenship: see, e.g., P. D. A. Garnsey, ‘Rome’s 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. 93–151 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’, Antiquit´es africaines 23 (1987), p. 80; Garnsey, ‘Rome’s African Empire’, pp. 252–4.

4

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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. Mattingly’s 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, ‘Rome’s African Empire’, passim; the quotation is ibid., p. 235. Mattingly, ‘Libyans and the “Limes” ’, pp. 80–3. As Ramsay MacMullen has recently shown of Juba’s Mauretanian kingdom, a high degree of acculturation could accompany such reconciliation: R. MacMullen, Romanization in the Time of Augustus (New Haven, Conn., 2000), pp. 42–9. 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. 31–60. See also D. Cherry, Frontier and Society in Roman North Africa (Oxford, 1998). See, e.g., Wallace-Hadrill, Rome’s Cultural Revolution, esp. pp. 447–8 and for Africa, D. J. Mattingly and R. B. Hitchner, ‘Roman Africa: An Archaeological Review’, JRS 85 (1995), pp. 204–5. See also the similar debate surrounding the term ‘Hellenization’: e.g., G. W. Bowersock, Hellenism in Late Antiquity (Ann Arbor, Mich., 1990), pp. 6–7. Woolf, Becoming Roman, p. 7.

5

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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. 234–58; 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).

6

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 Cherry’s 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. 82–99. Augustine, De civitate Dei 14.1, ed. B. Dombart and A. Kalb, CCSL 47–8 (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 Cit´es de l’Afrique romaine au Bas-empire, 2 vols. (Paris, 1979–81); 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. 45–125.

7

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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, B´enabou 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 l’´education dans l’antiquit´e (7th edn; Paris, 1971), pp. 492–3; P. Rich´e, Education and Culture in the Barbarian West, from the Sixth through the Eighth Century, trans. J. J. Contreni (Columbia, SC, 1978), pp. 37–9. A. H. M. Jones, The Later Roman Empire, 284–602: A Social, Economic and Administrative Survey, 2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986), 1:474–5. 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, ‘Rome’s African Empire’, pp. 252–4. Mattingly and Hitchner, ‘Roman Africa’, p. 205. J. N. Adams, Bilingualism and the Latin Language (Cambridge, 2003), pp. 200–45; on Libyan, see ibid., pp. 245–7. See also below, Chapter 3.4. Mattingly and Hitchner, ‘Roman Africa’, p. 205.

8

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 empire’s 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 312–37) 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 56–7; A. Chastagnol, ‘Les Consulaires de Numidie’, in M´elanges d’arch´eologie, d’´epigraphie et d’histoire offerts a` J´erˆome Carcopino (Paris, 1966), pp. 215–28; and A. Chastagnol, ‘Les Gouverneurs de Byzac`ene et de Tripolitaine’, Antiquit´es africaines 1 (1967), pp. 119–34. Vicars: Arnheim, Senatorial Aristocracy, pp. 63–4. M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Sp¨atantike, Frankfurter althistorische Studien 7 (Kallm¨unz, 1973), pp. 29–30; Chastagnol, ‘Consulaires de Numidie’, p. 219.

9

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 Overbeck’s 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 L´egats du proconsul d’Afrique au Bas-empire’, Libyca 6 (1958), p. 12, repr. in ´ A. Chastagnol, L’Italie et l’Afrique au Bas-empire: Etudes administratives et prosopographiques, Scripta varia (Lille, 1987), pp. 67–82, here p. 72. Overbeck, Senatsadel, pp. 23–8; see also PLRE 1–2, fasti. Overbeck, Senatsadel, p. 33 rejects the argument of A. Chastagnol, La Pr´efecture urbaine a` Rome sous le Bas-empire, Publications de la Facult´e des lettres et sciences humaines d’Alger 34 (Paris, 1960), p. 431 that Chilo, Proconsul of Africa in 375, was himself African. Overbeck, Senatsadel, pp. 32–3. Chastagnol, ‘L´egats du proconsul’, p. 12; Overbeck, Senatsadel, pp. 29 and 31–2. Garnsey, ‘Rome’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 379–95), 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 Augustine’s 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 Augustine’s 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: ‘F¨ur 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 sp¨atantiken Gallien (T¨ubingen, 1948), pp. 5–42. Overbeck, Senatsadel, p. 34. On the Spanish senatorial elite, see also K. F. Stroheker, ‘Spanische Senatoren der sp¨atr¨omischen und westgotischen Zeit’, in Germanentum und Sp¨atantike (Zurich, 1965), pp. 54–87; and A. Chastagnol, ‘Les Espagnols dans l’aristocratie gouvernementale a` l’´epoque de Th´eodose’, in Les Empereurs romains d’Espagne (Paris, 1965), pp. 269–92.

11

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

Staying Roman bishops of small African towns.’45 For in the struggle to maintain the integrity of the empire’s 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 Italy’s 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 Hadrian’s 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 empire’s 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. 24–5; the quotation is from p. 25. See also Warmington, North African Provinces, pp. 106–8. Brown, Augustine, p. 25. Chastagnol, ‘Consulaires de Numidie’, p. 219; Overbeck, Senatsadel, pp. 34–5 and 41–2. On the social and economic power of the fourth-century western senatorial elite in general, see Arnheim, Senatorial Aristocracy, 143–68. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 province’s 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 K¨onigtum im Westgotenreich, Vortr¨age 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 364–425 (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 407–485 (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 554–800 (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 fr¨uhmittelalterlichen 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 Fr¨uhmittelalterforschung (Vienna, 1994), pp. 9–26. 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, 332–489 (Oxford, 1991); E. James, The Franks (Oxford, 1988); I. N. Wood, The Merovingian Kingdoms 450–751 (London, 1994); P. Amory, People and Identity in Ostrogothic Italy, 489–554, 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. 500–700, 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, Europe’s Barbarians, ad 200–600 (New York, 2009).

13

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 chr´etienne 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 Africa’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:17, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.001

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 east–west 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 [world’s] 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. 620–1 and C. Courtois, Les Vandales et l’Afrique (Paris, 1955), pp. 215–17, with the sceptical assessment of W. Goffart, Barbarians and Romans, ad 418–584: The Techniques of Accommodation (Princeton, NJ, 1980), pp. 231–4 and A. Merrills and R. Miles, The Vandals (Chichester, 2010), pp. 48–9.

19

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Guntheric’s 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 484–96) 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 425–55), 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

3

4 5

6

7

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. 3–10 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. 193–266. Geiseric’s 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’, Antiquit´e tardive 11 (2003), pp. 50–3. On the Laterculus regum Wandalorum, see R. Steinacher, ‘The So-Called Laterculus Regnum Vandalorum et Alanorum: A Sixth-Century African Addition to Prosper Tiro’s Chronicle?’, in A. H. Merrills (ed.), Vandals, Romans, and Berbers: New Perspectives on Late Antique North Africa (Aldershot, 2004), pp. 163–80. On regnal dating, see also below, Chapter 3.2.3. Clover, ‘Timekeeping and Dyarchy’, p. 46 argues that Geiseric’s 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. F´evrier, ‘Ostraka de la r´egion de Bir Trouch’, Bulletin d’arch´eologie alg´erienne 2 (1966–7), pp. 239–42 and 244–5, nos. 1–2 and 4; Clover, ‘Timekeeping and Dyarchy’, pp. 54–9. F. M. Clover, ‘Felix Karthago’, Dumbarton Oaks Papers 40 (1986), pp. 1–16. For a good recent synthesis on Vandal coinage, see C. Morrisson, ‘L’Atelier de Carthage et la diffusion de la monnaie frapp´ee dans l’Afrique vandale et byzantine (439–695)’, Antiquit´e tardive 11 (2003), pp. 66–74. 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. 48–135. 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. 53–4 and F. M. Clover, ‘Geiseric the Statesman: A Study of Vandal Foreign Policy’, Ph.D. diss., University of Chicago (1966), pp. 89–90.

21

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman

Map 1.2.

Late antique North Africa, 439–700

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 Geiseric’s 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 Valentinian’s laws for Africa in the 440s and 450s: Val. Nov. 13 (ad 445) and 34 (ad 451), pp. 95–7 and 140–1. ´ Y. Mod´eran, ‘L’Etablissement territorial des Vandales en Afrique’, Antiquit´e tardive 10 (2002), pp. 88–97.

22

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power Valentinian’s 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 408–50) 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 Aspar’s 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 Roman–Vandal 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 Valentinian’s 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. 11–18, esp. 12, ll. 24–9 and Merobaudes, Carmen 1.17–18, ed. Vollmer in MGH AA 14:3, with the commentary of Clover, Merobaudes, pp. 24 and 51–4, and Clover, ‘Geiseric’, pp. 96–7. Proc. BV 1.3.35–6, 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, 1883–5; repr. Hildesheim, 1963), 1:95 and 104; John Zonaras, Epitomae Historiarum 13.24, ed. L. Dindorf, 6 vols. (Leipzig, 1868–75), 3:245–6; for the date, see PLRE 2:164–9, 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. 13–14. 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 300–900 (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 aˆ ge’, in Morfologie sociali e culturali in Europa fra tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sull’alto Medioevo 45 (Spoleto, 1998), pp. 77–8. Proc. BV 1.4.13, 1:326 with Clover, Merobaudes, p. 53; see also Merobaudes, Carmen 1.7–8, p. 3 with the commentary by Clover, Merobaudes, pp. 20–1. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Eudocia’s sister Placidia – on the throne of the western empire.15 Priscus, a wellinformed contemporary, stated explicitly that it was Geiseric’s connection to Olybrius by marriage (¡ –x –pigam©av sugg”neia) that motivated the Vandal king’s efforts on the senator’s 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 (paèçhs©a) that was not available to other late Roman aristocrats.16 Geiseric also sought control of that portion of Valentinian’s 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 Geiseric’s death in 477.17 For, quite apart from the strategic importance of the lands Geiseric claimed, gaining recognition from the eastern imperial court of Eudocia’s marriage to Huneric seems to have been one of the driving forces of the Vandal king’s 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 king’s 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), 2:340–2; Proc. BV 1.6.6, 1:336. See also John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn, 1831), pp. 373–4, reporting Leo I’s fears that Geiseric and Olybrius might have ambitions for the senator’s 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. 340–1; cf. Proc. BV 1.6.6, 1:336 (kdov). Malalas, Chronographia, p. 373. Priscus, frags. 38.2 and 39.1, pp. 340–2; 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 d’Italia 117 (Rome, 1991), pp. 78–9. The alliance between the Vandals and the Visigoths is discussed by Clover, ‘Geiseric’, pp. 105–9. J. Evans Grubbs, Law and Family in Late Antiquity (Oxford, 1995), p. 143.

24

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 princess’s 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 Geiseric’s motivations, the fury which Theoderic felt at the treatment of his daughter doubtless accounts at least in part for the chill which settled upon Vandal–Visigothic 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 Huneric’s 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. 155–6 and CTh 3.16.2 (ad 421), pp. 157–8, discussed by Evans Grubbs, Law and Family, pp. 229–31 and 234–7. 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. 174–5 reads this as evidence of a temporary Vandal–Visigothic 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. 118–20. Jordanes, Getica 47.244, p. 100. On Euric’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman seems to have mattered to Geiseric that Eudocia marry Huneric specifically, the oldest of the Vandal king’s 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 Geiseric’s 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 king’s 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 Eudocia’s engagement and their eventual marriage, both of Geiseric’s 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 Eudocia’s son Hilderic.29 Shortly after his own accession to sole power in 428, though, Geiseric himself had been willing to execute his brother’s 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 son’s 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 city’s 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 2:572–3, s.n. ‘Hunericus’; ibid., pp. 502–3, s.n. ‘Genton 1’; ibid., p. 1073, s.n. ‘Theodericus 4’; though surely Proc. BV 1.8.6, 1:346, ‘–v aÉt¼n g‡r ¾ cr»nov ›fere t‡ prwte±a toÓ Gizer©cou g”nouv’, refers to Gunthamund – then Geiseric’s 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, 1:344–5; Courtois, Vandales, pp. 238–42; 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. 74–7. Sons of Theoderic: Vict. Vit. 2.12–14, pp. 28–9 (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. 525–6, s.n. ‘Gunthamundus’; ibid., pp. 1116–17, s.n. ‘Thrasamundus 1’. Vict. Vit. 2.14, p. 29.

26

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 emperor’s 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 A¨etius, 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 emperor’s lifetime.33 Moreover, Valentinian’s killing of A¨etius in 453 set in motion a series of events that led to the emperor’s 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 emperor’s two daughters was united with the new Caesar, Valentinian’s younger daughter, Placidia, had probably already wed Olybrius by the time of her father’s 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 7.441–9, ed. A. Loyen in Sidoine Apollinaire, 3 vols. (Paris, 1960–70), 1:71–2; Hydatius, Chronicon 160 (ad 455), p. 104; Priscus, frags. 30.1–3, pp. 330–2; 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 1.5.1–5, 1:331–2; Vict. Tonn. s.a. 455.15, pp. 7–8; Malalas, Chronographia, pp. 365–6; Theophanes, Chronographia AM 5947, 1:108–9. 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. 396–7, no. 17, ‘Eudocie’, and Clover, ‘Geiseric’, p. 186. Clover, ‘Geiseric’, pp. 129–36. On A¨etius, see PLRE 2:21–8, s.n. ‘A¨etius 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 f¨ur Alte Geschichte 27 (1978), pp. 174–82 argues that the two were only engaged in 455.

27

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 princess’s engagement to Huneric, Maximus risked Geiseric’s 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 450–7), 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 Valentinian’s 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 (¡ gunŸ toÓ patrik©ou ìOlubr©ou); 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. 54–5 seems to indicate that Olybrius and Placidia were married only after the princess’s return to Constantinople, which Evagrius places in the reign of Marcian (ad 450–7). 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 Placidia’s African sojourn. PLRE 2:407–8, s.n. ‘Eudocia 1’; contra S. I. Oost, ‘A¨etius and Majorian’, Classical Philology 59 (1964), pp. 27–8 and Clover, ‘Geiseric’, pp. 143–4. Malalas, Chronographia, p. 366 twice calls Eudocia a parq”nov (‘virgin’ or ‘young woman’) at the time of her marriage to Huneric; but on the relationship between parq”nov 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. 51–75 (thanks to Emmanuel Papoutsakis for this reference, and for a useful discussion of the significance of the word ‘parq”nov’ in Malalas’ Syriac milieu). 39 Vict. Tonn. s.a. 455.14, p. 7. Clover, ‘Geiseric’, p. 144 n. 2.

28

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power to the empire would follow from marriage into the Theodosian family. The usurper Maximus not only wed his son to the former emperor’s daughter; in a bid to legitimate his own seizure of power, the senator himself also forcibly married Valentinian’s 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 emperor’s 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, 425–76’, in A. Cameron, B. Ward-Perkins, and M. Whitby (eds.), Late Antiquity: Empire and Successors, ad 425–600, Cambridge Ancient History 14 (Cambridge, 2000), pp. 18–27; A. D. Lee, ‘The Eastern Empire: Theodosius to Anastasius’, ibid., pp. 42–52; and K. Holum, Theodosian Empresses: Women and Imperial Dominion in Late Antiquity, Transformation of the Classical Heritage 3 (Berkeley, Calif., 1982). See also PLRE 2:407–8, s.n. ‘Eudocia 1’; ibid., pp. 410–12, s.n. ‘Eudoxia 2’; ibid., pp. 714–15, s.n. ‘Marcianus’; ibid., p. 887, s.n. ‘Placidia 1’; ibid., pp. 929–30, s.n. ‘Pulcheria’; ibid., p. 1100, s.n. ‘Theodosius 6’; ibid., pp. 1138–9, s.n. ‘Valentinianus 4’; and ibid., pp. 1308–9, 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 1.4.36–9, 1:330–1; 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 34–5 (Bonn, 1838–9), 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. 11–12. The rumour is discussed and rejected by Clover, ‘Geiseric’, pp. 150–5. On the forced marriage of Eudoxia and Petronius Maximus, see PLRE 2:411, s.n. ‘Eudoxia 2’. Clover, ‘Geiseric’, pp. 54 and 87–100.

29

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman Table 1.1. Roman embassies to the Vandals, 455–84 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, Valentinian’s 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 Valentinian’s widow and daughters. However, neither Bleda’s 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 455–6) 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 1.4.13–14, 1:326. Clover, ‘Geiseric’, pp. 98–9, argues that Huneric had probably returned to Africa by the time Merobaudes delivered his panegyric in honour of A¨etius’ third consulate on 1 Jan. 446, and certainly by the time of the empress Gallia Placidia’s 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 1:271–3, s.n. ‘Deogratias 1’, p. 271; and Clover, ‘Geiseric’, pp. 134–6. 46 Merrills and Miles, Vandals, p. 118. Priscus, frag. 31.1, p. 334.

30

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 457–61) the situation degenerated into war. Outmanoeuvred by Geiseric, the western emperor was compelled to make peace on “shameful terms” (sunqkai a«scra©); 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 king’s attacks on Italy and Sicily, and in 461 or 462 the eastern emperor Leo I (ad 457–74) 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 Huneric’s marriage to Eudocia. The deal apparently also involved the release to the princess of some of her father’s eastern properties and probably also included a pledge on Leo’s 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 Eudocia’s 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 Geiseric’s candidate Olybrius. Leo sent Phylarchus back to Geiseric’s court, threatening to invade Africa if the Vandal king did not surrender his claims to Italy and Sicily. Still Geiseric refused to accept the emperor’s 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. 340–2. Priscus, frag. 36.2, p. 338; Clover, ‘Geiseric’, p. 179. On the conflict, see Merrills and Miles, Vandals, pp. 119–20. Priscus, frags. 38.1 and 39.1, pp. 340 and 342; Blockley, East Roman Foreign Policy, p. 72. Priscus, frags. 41.1–2, p. 346. Priscus, frag. 52, p. 360; Merrills and Miles, Vandals, p. 121.

31

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman 468 Leo and Geiseric negotiated another treaty.53 Finally, in 474, the new emperor in the East, Zeno (ad 474–91), 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 Geiseric’s 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 533–4, 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 Zeno’s first embassy to the court of Huneric (ad 477–84) was equally successful. In what was surely a calculated concession to Vandal dynastic claims, the eastern emperor Zeno sent Alexander, the guardian of Huneric’s 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 (st”rgoi t‡ ëRwma©wn), and that he was abandoning his father’s claims on Eudocia’s 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. 121–2. PLRE 2:796–8, s.n. ‘Olybrius 6’, pp. 797–8. 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 1.7.26–8, 1:344, but see also ibid., 1.16.13–14, 1:384, where Justinian is at pains to claim that he was not breaking the treaty. Blockley, East Roman Foreign Policy, pp. 79–80. 58 Malchus, frag. 17, p. 424. Malchus, frag. 5, p. 410, and Vict. Vit. 1.51, p. 22. Vict. Vit. 2.3–4, p. 25.

32

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power Presumably as a result of these actions, Huneric established himself as a friend of the emperor (f©lov 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 empire’s new, post-Theodosian rulers once again sought a military solution to the threat posed by Vandal raids, by the Vandal king’s 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 Geiseric’s status as ‘friend and ally’ (f©lov te kaª xÅmmacov) 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 Roman–Vandal 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 (f©lov 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 Huneric’s reign. Zeno’s envoy Reginus was present in Carthage in May of 483 when the Vandal king had an edict publicly read out insisting that the region’s Nicene bishops come to Carthage in the following year to debate the principles of Christian faith.65 A year later another of Zeno’s legates, Uranius, was sent to Huneric’s court in an attempt to persuade the king to stop the ensuing persecution of Africa’s Nicene Christians.66 And we hear nothing of Vandal–Roman relations under Huneric’s nephew and successor, Gunthamund (ad 484–96); though our evidence for contact 60 61

62 63 65

Malchus frag. 17, p. 424. ´ Mod´eran, ‘L’Etablissement territorial’, pp. 92–4; D. Braund, Rome and the Friendly King: The Character of the Client Kingship (London, 1984), pp. 23–37. In classical antiquity the title was granted most famously to Herod (see, e.g., Braund, Rome, pp. 24–5); 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. 233–42. ´ Mod´eran, ‘L’Etablissement territorial’, pp. 92–3; Heather, Fall of the Roman Empire, p. 292. 64 Malalas, Chronographia, pp. 373–4. Proc. BV 1.4.39, 1:331. 66 Vict. Vit. 3.32, p. 88. Vict. Vit. 2.38–9, pp. 38–9.

33

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman between the two powers in general diminishes after Huneric’s 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 496–523) was exceptionally well disposed towards the emperor Anastasius – or his particular friend (f©lov –v t‡ m†lista) – and that the same was true of Hilderic (ad 523–30) and Justinian.67 Envoys also passed between Justinian’s 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 Hilderic’s accession, in fact, we are told that imperial–Vandal 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 Justinian’s guest-friend (x”nov).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 Justinian’s uncle Justin to the imperial throne in 518 and Hilderic’s 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 Justinian’s 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 Vandal’s 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 Hilderic’s children ‘copious amounts of money’ (cržmata ¬kan†) as the great-grandchildren of Valentinian III.72 Indeed, once acknowledged, the Vandal royal family’s 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, 1895–8), 2:670–1. Proc. BV 1.9.5, 1:351–2. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Marcian’s choice of an Arian bishop to lead the mission to Geiseric in 455 or 456 was doubtless calculated to appeal to the Vandal king’s religious sensibilities. Alexander’s selection as an envoy to Huneric was, as we have seen, certainly a concession to Huneric’s 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 Huneric’s 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 Huneric’s part. Of course, the failure of Uranius’ embassy cannot be blamed on the envoy’s comparatively humble status; patrician status did not guarantee a mission’s 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 emperor’s 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, 411–533, Cambridge Studies in Medieval Life and Thought, 4th ser., 55 (Cambridge, 2003), pp. 231–8. For Tatianus’ career, see PLRE 2:1053–4, s.n. ‘Tatianus 1’. Malchus, frag. 5, p. 410. Vict. Vit. 3.32, p. 88. For his earlier career, see PLRE 2:1186–7, s.n. ‘Vranius 4’. Blockley, East Roman Foreign Policy, p. 153. On this point in general, see Gillett, Envoys, pp. 244–59.

35

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 king’s 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 1.25.5–6, 1:413. On Moorish–Vandal relations, see Y. Mod´eran, Les Maures et l’Afrique romaine (IVe –VIIe si`ecle), Biblioth`eque des e´ coles franc¸aises d’Ath`enes et de Rome 314 (Rome, 2003), esp. pp. 541–61. Geiseric: Vict. Vit. 1.35–8, pp. 16–17. Huneric: Vict. Vit. 2.26–37, pp. 33–8; see also ibid., 2.4, p. 25. Report and execution: Vict. Vit. 1.37, pp. 16–17.

36

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 530–3), 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 Vandal–Visigothic 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. 115–34 (quotation, ibid., p. 117); Warmington, North African Provinces, p. 72; Y. Mod´eran, ‘La Notitia provinciarum et civitatum Africae et l’histoire du royaume vandale’, Antiquit´e tardive 14 (2006), pp. 165–85. Proc. BV 2.4.26–8, 1:436. On the regularity of diplomatic contact, see Gillett, Envoys, pp. 74–5. Hydatius, Chronicon 186 (ad 458), p. 110 and ibid., p. 234 (ad 466–7), p. 118; Gillett, Envoys, pp. 68–9. 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 464–5), p. 116; Gillett, Envoys, p. 69. Jordanes, Getica 47.244, p. 100. Jordanes, Getica 36.184, pp. 78–9 (= Priscus, frag. 20.2, p. 306) and Priscus, frag. 20.1, pp. 304–6. Proc. BV 1.24.7–18, 1:411–12.

37

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Italy’s recognition of the legitimacy of Vandal claims to Sicily, and thus of Huneric and Eudocia’s marriage.91 Despite Huneric’s own renunciation of his claims to Eudocia’s 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 Theodoric’s 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 Ostrogothic–Vandal 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 Theodoric’s 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 37–38, ed. and trans. C. Rodr´ıguez Alonso, in Las Historias de los godos, v´andalos y suevos de Isidoro de Sevilla: Estudio, edici´on cr´ıtica y traducci´on, Fuentes y estudios de historia leonesa 13 (Le´on, 1975), pp. 232–4; Isidore, Historia Gothorum s. aera 545, ed. and ˚ J. Fridh, in trans. C. Rodr´ıguez Alonso, ibid., pp. 232–4; Cassiodorus, Variarum libri xii, ed. A. Magni Aurelii Cassiodori senatoris opera, Pars 1, CCSL 96 (Turnhout, 1973), 5.43–4, pp. 220–2; see also Proc. BG 1.12.46, 2:69; J. Moorhead, Theoderic in Italy (Oxford, 1992), p. 190. Proc. BV 2.4.34–41, 1:437–8. Vict. Vit. 1.14, p. 7. F. M. Clover, ‘A Game of Bluff: The Fate of Sicily after ad 476’, Historia: Zeitschrift f¨ur alte Geschichte 48 (1999), pp. 235–44 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. 63–5; and on the question of who initiated negotiations, H.-J. Diesner, Die

38

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Sicily’s 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 Theodoric’s envoy Senarius brought him not just to Constantinople and to Spain but also to Carthage.96 Agnellus, one of the higher officials of Theodoric’s court, was in Africa in the winter of 505–6, and then again probably in 507–8. These embassies took place within the context of worsening relations between Ravenna and Constantinople: by 507–8, 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; Theodoric’s 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 S¨achsischen Akademie der Wissenschaften zu Leipzig, Philologisch-historische Klasse 113/3 (Berlin, 1967), p. 4. Proc. BV 1.8.12–13, 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: Rome’s Economic Hinterland c.350–600 ad’, Papers of the British School at Rome 55 (1987), p. 181 and Heather, Goths, pp. 231–2 have read Thrasamund’s Africa as a client-kingdom of Theodoric’s 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 2:988–9, s.n. ‘Senarius’, and Gillett, Envoys, pp. 190–219. Cass. Var. 1.15.2, p. 25: ‘qui regnum petens alterius nostris est utilitatibus seruiturus’; Moorhead, Theoderic, pp. 182–3. For the dates and Agnellus’ possible offices, see PLRE 2:35–36, s.n.

39

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 king’s courts. Ostrogothic–Vandal relations reached their nadir in the reign of Thrasamund’s cousin and successor Hilderic. On Thrasamund’s 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 Hilderic’s 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 (di”f{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 Hilderic’s 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 king’s death that same year forestalled whatever attack he may have been planning, but the issue nevertheless prompted some diplomatic exchange between the court of Theodoric’s successor and the Vandal king Hilderic.105 We owe our knowledge of Ostrogothic–Vandal 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. 221–2. 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, 1:351–2; see also Moorhead, Theoderic, pp. 216–17 and 246–8. Cass. Var. 5.16–20, pp.195–9; Moorhead, Theoderic, pp. 246–8. See, however, the sceptical assessment of J. Pryor and E. M. Jeffreys, The Age of the Dromwn: The Byzantine Navy ca 500–1204, The Medieval Mediterranean 62 (Leiden, 2006), pp. 13–14. Cass. Var. 9.1, pp. 345–6.

40

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power From this perspective, Ostrogothic–Vandal 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 envoy’s 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 Theodoric’s 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 king’s 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 – Gelimer’s 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 2:988–9, s.n. ‘Senarius’. ¨ M. Sch¨onfeld, W¨orterbuch der altgermanischen Personen- und V¨olkernamen nach der Uberlieferung des klassischen Altertums bearbeitet (Heidelberg, 1911; repr. 1965), p. 202 and Gillett, Envoys, pp. 198–9. Cass. Var. 4.3, pp. 144–5 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 7:235–6, 240, and 258. 112 PLRE 2:35–6, s.n. ‘Agnellus’. Ennodius, Ep. 4.18, p. 143.

41

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Gelimer’s 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 Gelimer’s 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., 376–7, s.nn. ‘Fuscus 1–2’; 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 2:306–7, 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 2/1:879–80, 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. Sch¨onfeld, Altgermanischen Person- und V¨olkernamen, pp. 112–13. 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 ausw¨artigen diplomatischen Verkehr des r¨omischen Reiches im Zeitalter der Sp¨atantike’, Archiv f¨ur Urkundenforschung 12 (1931–2), p. 402 n. 4. C. Morrisson, C. Brenot, and J.-N. Barrandon, ‘L’Argent chez les Vandales: plats et monnaies’, in F. Baratte (ed.), Argenterie romaine et byzantine: actes de la table ronde, Paris 11–13 octobre 1983 (Paris, 1988), pp. 123–33.

42

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 another’s 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 couple’s son, King Hilderic (ad 523–30), 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 Valentiani 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 66–70.121 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 484–96) 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. 70–3; Berndt, Konflikt und Anpassung, pp. 205–6; M. E. Gil Egea, A´frica en tiempos de los v´andalos: continuidad y mutaciones de las estructuras sociopol´ıticas romanas, Memorias del Seminario de Historia Antigua 7 (Alcal´a de Henares, 1998), pp. 314–18; M. McCormick, Eternal Victory: Triumphal Rulership in Late Antiquity, Byzantium and the Early Medieval West (Cambridge, 1986), pp. 261–6; 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. 1–22; Schmidt, Wandalen, pp. 156–7; and Courtois, Vandales, pp. 242–8. Clover, ‘Carthage and the Vandals’, p. 9. Proc. BV 1.5.3–5 and 2.9.4–5, 1:331–2 and 1:456–7. On the Vandal sack of Rome, see Berndt, Konflikt und Anpassung, pp. 194–5. AL 371, ll. 12–14, 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. 50–9; 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. 85–121, esp. 109–10. Coinage: W. Hahn, Moneta Imperii Byzantini: Rekonstruktion des Pr¨ageaufbaues auf synoptisch-tabellarischer Grundlage, vol. 1, Von Anastasius I. bis Justinianus I. (491–565) einschliesslich der ostgotischen und vandalischen Pr¨agungen, Ver¨offentlichungen der Numismatischen

44

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 496–523) 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 Huneric’s reign; one, indeed, is dated according to the Vandal king’s 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

¨ Kommission 1/Osterreichische Akademie der Wissenschaften, Philosophisch-Historische Klasse, Denkschriften 109 (Vienna, 1973), pp. 94–5 and pl. 42.2–8, 10, 12 and 15–16. Inscription: CIL 8.10862. McCormick, Eternal Victory, pp. 263–6. Tablettes Albertini: actes priv´es de l’´epoque 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 3.3–14, pp. 39 and 72–8; fragment: ibid., 2.3–4, p. 25. Vict. Vit. 2.39, p. 39. P. Classen, Kaiserreskript und K¨onigsurkunde: Diplomatische Studien zum Problem der Kontinuit¨at 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., 3.3–14, pp. 72–8). Gil Egea, A´frica, pp. 318–20. Notitia, Byz. 107, p. 127; Courtois, Vandales, p. 243 n. 6.

45

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 201–5, pp. 150–3 (baths); AL 204, l. 1, p. 152 (fortifications); AL 204, ll. 5–6, 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’, Antiquit´e 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 g‡r ¾ Karchd»nov per©bolov oÌtw dŸ ˆphmelhm”nov ãste –sbat¼v –n cÛroiv pollo±v t boulom”n kaª eɔjodov –geg»nei’). A. Chastagnol and N. Duval, ‘Les Survivances du culte imp´erial dans l’Afrique du Nord a` l’´epoque vandale’, in M´elanges d’histoire ancienne offerts a` William Seston (Paris, 1974), pp. 87–118; but see F. M. Clover, ‘Emperor Worship in Vandal Africa’, in G. Wirth (ed.), RomanitasChristianitas: Untersuchungen zur Geschichte und Literatur der r¨omischen Kaiserzeit (Berlin, 1982), pp. 661–74 and F. M. Clover, ‘Le Culte des empereurs dans l’Afrique vandale’, BCTH ser. 2, 15–16 (1984), pp. 121–8, who suggests that the later Vandal kings allowed their Roman subjects to continue to venerate the emperor. Clover, ‘Carthage and the Vandals’, pp. 12–13; however, see Wickham, Framing, p. 637. Proc. BV 1.16.12, 1:384.

46

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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. Cyrila’s 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. 686–8, with the corrective offered by M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain, ad 300–750, BAR International Series 1135 (Oxford, 2003), p. 2 in the light of N. Duval, ´ Recherches arch´eologiques a` Ha¨ıdra, 2 vols., Collection de l’Ecole franc¸aise de Rome 18/1–2 (Rome, 1975–81), 1:87–8, no. 58 = AE (1968), 204, no. 638 (though note that ‘Vandalorum’ seems to be a later addition to this inscription: see Duval, Ha¨ıdra, 1:88 and 2:119 (La basilique ´ I dite de Mell´eus ou de Saint-Cyprien, Collection de l’Ecole franc¸aise de Rome 18/2 (Rome, 1981)).

47

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman wealth’ (ploÅt ˆkm†zwn), 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 senator’s estate in Byzacena, by contrast, seems to have been taken over by the Vandal king, for it proved recoverable when Gordian’s 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 1.5.11–15, 1:333–4; Vict. Vit. 1.13, p. 7; see also Val. Nov. 34.2–3 (ad 451), p. 141, which specifically mentions despoliations in Africa Proconsularis and Byzacena. On the geography, see J. Desanges, ‘Un T´emoignage peu connu de Procope sur la Numidie vandale et byzantine’ Byzantion 33 (1963), pp. 49–56. On imperial estates in Africa Proconsularis, see M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Sp¨atantike, Frankfurter althistorische Studien 7 (Kallm¨unz, 1973), pp. 40–1 and D. P. Kehoe, The Economics of Agriculture on Roman Imperial Estates in North Africa, Hyptomnemata: Untersuchungen zur Antike und zu ihrem Nachleben 89 (G¨ottingen, 1988). ´ On Vandal settlement, see esp. Mod´eran, ‘L’Etablissement 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 V¨olkerwanderungszeit 400–600. Berichte des Symposions der Kommission f¨ur Fr¨uhmittelalterforschung, 7. bis 9. Mai 1986, Stift Zwettl, Nieder¨osterreich, Ver¨offentlichungen der Kommission f¨ur Fr¨uhmittelalterforschung 11 (Vienna, 1988), pp. 21–72; J. Durliat, ‘Les Transferts fonciers apr`es la reconquˆete byzantine en Afrique et en Italie’, in E. Magnou-Nortier (ed.), Aux sources de la gestion publique, 2 vols. (Lille, 1993–5), 2:89–121 ; and A. Schwarcz, ‘The Settlement of the Vandals in North Africa’, in Merrills, Vandals, Romans, and Berbers, pp. 49–57 (who reject it). Vandal settlement is not discussed by Goffart, Barbarians and Romans. 142 V. Fulg. 1, p. 11. Leone, Changing Townscapes, pp. 145–8. Millenarius: Vict. Vit. 1.30–5, pp. 13–15. Fridamal: AL 299–300, pp. 246–7.

48

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power Though nobles like Fridamal and Victor’s millenarius owned rural or suburban estates, Vandals seem for the most part to have taken up residence in Africa’s 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

´ For the localization of Vandal settlement in general, see Mod´eran, ‘L’Etablissement territorial’, 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 7.21.89–91, ed. G. Lagarrigue, in Du Gouvernement de Dieu, vol. 2 of Œuvres, SC 220 (Paris, 1975), pp. 494–6; 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, 57–8 (Vienna, 1895–1923), here Ep. 185.1.1, CSEL 57:2; Olympiodorus, frag. 40, ed. Blockley, in Fragmentary Classicising Historians, 2:202–4; and pseudo-Augustine, Ep. 4, PL 33:1095. Clover, ‘Geiseric’, pp. 28–30 argues that the Goths in Boniface’s 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. 148–50 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. 232–3.

49

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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. 299–360; J. Kleemann, ‘Quelques r´eflexions sur l’interpr´etation ethnique des s´epultures habill´ees consid´er´ees comme vandales’, Antiquit´e tardive 10 (2002), pp. 123–9; P. von Rummel, ‘Habitus Vandalorum? Zur Frage nach einer gruppen-spezifischen Kleidung der Vandalen in Nordafrika’, Antiquit´e tardive 10 (2002), pp. 134–7; Christopher Eger, ‘Vandalisches Trachtzubeh¨or? Zu Herkunft, Verbreitung und Kontext ausgew¨ahlter 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; and S. Brather, ‘Kleidung, Grab und Identit¨at in Sp¨atantike und Fr¨uhmittelalter’, in Berndt and Steinacher, Das Reich der Vandalen, pp. 283–94. See also, in general, P. Amory, People and Identity in Ostrogothic Italy, 489–554, Cambridge Studies in Medieval Life and Thought, 4th ser. 33 (Cambridge, 1997), pp. 33–4 and 334–7.

50

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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. Az´ema, in Correspondance, 4 vols., SC 40, 98, 111, and 429 (Paris, 1964–98), 2:152–4 and Vict. Vit. 1.12 and 1.14, pp. 7–8. 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. 135–62 and Leone, Changing Townscapes, pp. 127–34. Note the apparent contraction of rural settlement in the inland regions of Dougga and Cillium-Thelepte: M. de Vos, Rus Africum: Terra, acqua, olio nell’Africa settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000), pp. 20–1, 72–5, and 81–2 and R. B. Hitchner, ‘The Kasserine Archeological Survey 1982–1986 (University of Virginia, USA – Institut national d’arch´eologie et d’art de Tunisie)’, Antiquit´es africaines 24 (1988), pp. 7–41, respectively. J. A. Greene, ‘Une Reconnaissance arch´eologique dans l’arri`ere-pays de la Carthage antique’, in A. Ennabli (ed.), Pour sauver Carthage: exploration et conservation de la cit´e punique, romaine et byzantine (Tunis, 1992), pp. 195–7. S. Dietz, ‘A Summary of the Field Project’, in S. Dietz, L. Ladjimi Seba¨ı, H. Ben Hassen, P. Ørsted, and J. Carlsen (eds.), Africa Proconsularis: Regional Studies of the Segermes Valley of Northern Tunisia, 3 vols. (Aarhus, 1995–2000), 2:781–2. ´ Leone, Changing Townscapes, pp. 128–9; M. Bonifay, Etudes sur la c´eramique romaine tardive d’Afrique, BAR International Series 1301 (Oxford, 2004), pp. 49 and 53–7; D. Barraud, M. Bonifay, F. Dridi, and J. F. Pichonneau, ‘L’Industrie c´eramique de l’antiquit´e tardive’, in H. Ben Hassen and L. Maurin (eds.), Oudhna (Uthina): La Red´ecouverte d’une ville antique de Tunisie (Bordeaux, 1998), pp. 139–67; and M. Mackensen, Die sp¨atantiken Sigillata- und Lampent¨opfereien von El Mahrine (Nordtunesien): Studien zur Nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts, M¨unchner Beitr¨age zur Vor- und Fr¨uhgeschichte 50, 2 vols. (Munich, 1993).

51

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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-Zˆateur 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; Victor’s 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 e´conomique: dix ans de recherche, Collection ´ de l’Ecole franc¸aise 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. 641–3; 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. Rich´e, Education and Culture in the Barbarian West, from the Sixth through the Eighth Century, trans. J. J. Contreni (Columbia, SC, 1978), pp. 37–9 and 64. K. V¨ossing, Schule und Bildung in Nordafrika der r¨omischen Kaiserzeit (Brussels, 1997), pp. 624–31 takes a darker view of the relations between Vandals and Romano-Africans. Koenig, ‘Wandalische Grabfunde’, pp. 308–9; Kleemann, ‘Quelques r´eflexions’, p. 126; Eger, ‘Vandalisches Trachtzubeh¨or?’, p. 189. Thus Rummel, ‘Habitus Vandalorum’, pp. 136–7; see also Rummel, Habitus barbarus: Kleidung und Repr¨asentation sp¨atantiker Eliten im 4. und 5. Jahrhundert, Erg¨anzungsb¨ande zum Reallexikon der Germanischen Altertumskunde 55 (Berlin, 2007), pp. 270–323 and Kleemann, ‘Quelques r´eflexions’, pp. 124–5. See, further, Amory, People and Identity, pp. 343–4. Claudius: V. Fulg. 5, pp. 29–31. Heldica and his family: Vict. Vit. 2.15, p. 29. Also two Vandal brothers: Vict. Vit. 3.38, p. 91.

52

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Becca’s ‘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. F¨orstemann, Altdeutsches Namenbuch, vol. 1, Personennamen (Munich, 1966), pp. 300–1, 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. 178–9. 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. 46–64, esp. 59 and pls. XVIII.40–XIX.43. Hunting themes: AL 299, 302, 329–30, and 355, pp. 246–7, 248–9, 264, and 278; see also AL 287 and 295, pp. 239 and 243. Beast hunts and amphitheatres: AL 341, 348–9, and 368, pp. 270, 273–5, and 285. Archaeology: Leone, Changing Townscapes, p. 140. Epigraphy: N. Duval and J. Mallon, ‘Les Inscriptions de la “chapelle vandale” a` Ha¨ıdra d’apr`es l’abb´e Delapard’, Bulletin de la Soci´et´e nationale des Antiquaires de France (1969), pp. 118–24 = AE (1973), 198, no. 622: Fecit Va/ricos ludos; the name seems to be the Punic ‘Baric’, but see Francovich Onesti, I Vandali, pp. 177–8, s.n. ‘Varica’.

53

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Gebamund’s 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 Virgil’s Aeneid and Georgics – to celebrate his marriage.172 The 164 165 166

167 168

169 170 171

172

CIL 8.25362; on Thrasamund’s baths, see above n. 133. On Gebamund, see Proc. BV 1.18.1 and 1.18.12–19, 1:388 and 1:390–1 and Vict. Tonn. s.a. 534.38, p. 38. AL 345 (De aquis calidis Cirnensibus), pp. 271–2. Y. Th´ebert, Thermes romains d’Afrique du Nord et leur contexte m´editerran´een: e´tudes d’histoire et d’arch´eologie, Biblioth`eque des e´ coles franc¸aises d’Ath`enes et de Rome 315 (Rome, 2003), pp. 418–21 and 482–3; Leone, Changing Townscapes, pp. 140–1 and 157–9; and Miles, ‘Anthologia Latina’, pp. 310–12. Proc. BV 1.17.10, 1:386–7. AL 327 and 364, pp. 262–3 and 283, respectively. The name Eugetius is Greek eÉgh{žv, ‘joyous’ or ‘cheerful’, not a corruption of Hoageis; on whom, see Courtois, Vandales, p. 399, no. 24, ‘Hoageis’, and ibid., p. 390, ‘Tableau g´en´ealogique des Hasdings’. AL 286, p. 238–9. AL 367 and 361, pp. 284 and 281, respectively. AL 72–4, pp. 76–8; pseudo-Dracontius, De origine rosarum, ed. E. Baehrens and F. Vollmer, in Poetae latini minores, 6 vols. (Leipzig, 1881–1911), 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), 1/1:79–82. This poem is not included in D. R. Shackleton Bailey’s edition of the Anthologia Latina, vol. 1/1 (Stuttgart, 1982).

54

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Rich´e 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 Gelimer’s 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; Rich´e, 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, Œuvres, 4 vols. (Paris, 1985–96), 3:134. AL 305, pp. 250–1. Luxorius also wrote about a female cithara-player named Gattula, whose body disgusted the poet: AL 356–7, pp. 279–80; Victor of Vita refers to Geiseric’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 city’s 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 s¯erik¯e [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. 8–9. 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. 241–4. 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. 101–32. 182 AL 371, l. 30, p. 288. AL 371, pp. 286–8, esp. ll. 9–18 and 28–36. 184 Discussed by Clover, ‘Felix Karthago’, pp. 1–10. Proc. BV 2.14.18–19, 1:484–5.

56

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 historian’s 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 malak©an), 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 Salvian’s censure, but the bishop reserved his strongest condemnation 185 186

187

188

Proc. BV 2.6.5–9, 1:443–4. 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 Antiquit´e tardive 8 (2000), a ´ special issue dedicated to the De aedificiis, esp. D. Feissel, ‘Les Edifices de Justinien au t´emoignage de Procope et de l’´epigraphie’, Antiquit´e tardive 8 (2000), pp. 81–104; and A. Kaldellis, Procopius of Caesarea: Tyranny, History, and Philosophy at the End of Antiquity (Philadelphia, Pa., 2004). Salvian, De gubernatione Dei 6.12.66–71, 7.13.54–7.22.100, and 8.2.9–8.5.25, pp. 404–8, 468– 502, and 517–27, esp. ibid., 7.13.56 and 7.22.94–100, pp. 470 and 498–502. On these points, see D. Lambert, ‘The Barbarians in Salvian’s De Gubernatione Dei’, in S. Mitchell and G. Greatrex (eds.), Ethnicity and Culture in Late Antiquity (London, 2000), pp. 103–15. 189 Proc. BP 1.17.37 and 2.8.6, 1:88 and 1:184–5. Malchus, frag. 17, p. 424.

57

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Salvian’s 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. Salvian’s denunciation of the circuses and theatres provides much of the matter for book 6 of his De gubernatione Dei. Salvian, De gubernatione Dei 6.8.39–45, pp. 388–92. 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 Salvian’s De Gubernatione Dei’, Augustinian Studies 30 (1999), pp. 115–30. Y. A. Dauge, Le Barbare: recherches sur la conception romaine de la barbarie et de la civilisation, Collection Latomus 176 (Brussels, 1981), pp. 433–4.

58

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Victor’s 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., 2.14.17–18, 1:484–5 (organized into cavalry units). Second wave: ibid., 2.19.3, 1:508. On this point, see Liebeschuetz, ‘Gens into Regnum’, pp. 71–2 and, more broadly, R. Steinacher, ‘Gruppen und Identit¨aten: Gedanken zur Bezeichnung “vandalisch”’, in Berndt and Steinacher, Das Reich der Vandalen, pp. 243–60. 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, 300–800, Transformation of the Roman World 2 (Leiden, 1998), pp. 17–69. 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. 134–5, 142–3, 181–2, 186, 249, and 258. Rarely can we be confident that they were themselves considered Vandals: Sidonius Apollinaris, Carmen 5.436–7, 1:44 (Geiseric’s sister); Vict. Vit. 2.14, p. 29 (Theoderic’s daughters); AL 340, pp. 269–70 (Damira). More usually they are the wives or mothers of Vandals: Vict. Vit. 1.35, p. 15 (millenarius’s widow); ibid., 2.12 (Theoderic’s wife), p. 28; ibid., 2.14 (Godagis’ wife; Gunderic’s widow), p. 29; ibid., 2.15 (wife of Heldica; Teucharia?), p. 29; ibid., 3.33, p. 88 (Dagila’s 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. 478–9 (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?’, 109–10. Vandals in military context: see, e.g., Possidius, Vita Augustini 28.4–5, pp. 148–50; 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 456–7), 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Geiseric’s 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 Carthage’s 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 2.348–50, 362–5, 368–70 and 379–80, 1:17–18 and ibid., 5.390–2 and 419–20, 1:43–4; 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, 1.24.1–4, and 2.3.8–14, 1:363–4, 1:410, and 1:429–30. Gunthimer: Vict. Tonn. s.a. 534.118, p. 38. Ammatas: Proc. BV 1.17.11 and 1.18.1–7, 1:387–9. Gebamund: ibid., 1.18.1 and 1.18.12–19, 1:388 and 1:390–1 and Vict. Tonn. s.a. 534.118, p. 38. Sidonius Apollinaris, Carmen 5.435–7, 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. 15–16. Cavalry: Sidonius Apollinaris, Carmen 5.397–424, 1:43–4; Coripp. Ioh. 3.236–55, pp. 56–7; 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, 450–900 (New York, 2003), pp. 180–8. Procopius estimated the Vandal force sent to recapture Sardinia in 533 at 5,000 (BV 1.11.23, 1:363–4) and the detachment under Gebamund’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Victor’s ‘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 Gelimer’s 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. 2.8–9, p. 27: ‘mares uel feminas in habitu barbaro’. Rummel, ‘Habitus Vandalorum’, pp. 133–4 is sceptical of Victor on this point. Rummel, ‘Habitus Vandalorum’, pp. 138–40; Rummel, Habitus barbarus, pp. 231–45; and Amory, People and Identity, pp. 341–3. 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. 109–10. Proc. BV 1.10.25–33, 1:359–60; cf. Heather, Goths, pp. 302–3.

61

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Staying Roman to wearing Victor’s 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 region’s 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 one’s identity was not invariably as easy as changing one’s 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 Jerome’s 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. 78–9; Berndt, Konfikt und Anpassung, pp. 223–4. 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. 39–74. Proc. BV 1.2.4, 1:311: ‘leuko© te g‡r Œpantev t‡ sÛmat† e«si kaª t‡v k»mav xanqoª, eÉmžkeiv te kaª ˆgaqoª t‡v Àyeiv’. 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), 1/1:230–6, 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. 3.55–60, pp. 99–102 with Clover, ‘Carthage and the Vandals’, pp. 15–17 for the archaeological evidence of a contemporary mass grave.

62

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

Vandal power grooming and gait may have helped cue individuals into one another’s 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 don’t 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 Cyrila’s 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. 51–61. ¨ die Sprache der Wandalen: Ein Beitrag zur germanischen Namen- und Dialektforschung, F. Wrede, Uber Quellen und Forschungen zur Sprach- und Kulturgeschichte der germanischen V¨olker 59 (Strasburg, 1886); Francovich Onesti, Vandali, pp. 133–202; 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. 51–71 and cf. Amory, People and Identity, pp. 102–8. Vict. Vit. 2.55, pp. 45–6: ‘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. 251–68. AL 206, p. 154.

63

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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, Africa’s 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 Africa’s 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 army’s invasion of Africa to the region’s 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 Riese’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Geiseric’s 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. Geiseric’s 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 Hilderic’s Africa, the poetry of Luxorius vividly illustrates that it mattered deeply, at least to 222

Proc. BV 1.16.13–14, 1:384.

65

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:32, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.002

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 Ruspe’s 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 region’s 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.

1

2 3

Vict. Vit. 1.15–16, 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. 34.2–3 (ad 451), p. 141, which refers to the Vandal expulsion of distinguished African property owners. Vict. Vit. 1.14, pp. 7–8; see also ibid., 1.12, p. 7, and Proc. BV 1.5.11, 1:333.

67

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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´ migr´es 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

5

6

Possidius, Vita Augustini 28.12–13, p. 154. On Geiseric’s later expulsion of some African bishops from their sees, see Prosper, Chronicon s.a. 437, p. 475, with PCBE 1:783–4, s.n. ‘Novatus’; ibid., pp. 890–6, 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, 1864–77), 2:221–5; PCBE 2/2:1561, s.n. ‘Optatus 2’; PCBE 1:801–2, s.n. ‘Optatus 4’; and J. Conant, ‘Europe and the African Cult of Saints, circa 350–900: An Essay in Mediterranean Communications’, Speculum 85 (2010), p. 43. Breviarium Hipponense, Brevis statutorum 27, ed. C. Munier, in Concilia Africae a.345–a.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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications undetermined 5%

laity 41%

Fig. 2.1.

churchmen 54%

Africans abroad, 439–533: 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, 439–533: 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 423–49 and 451–66) 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 457–61), 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. 29–32 and Epp. 34–6, 2:86–92 and 96–8; 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 100–2 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, 2:152–4 (Maria and Eudaemon). Theodoret, Ep. 70, 2:154, probably in the West (  ?0 ), though C. Courtois, Les Vandales et l’Afrique (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, 1960–70), 3:164 and 176. PLRE 2:374, s.n. ‘Domnulus 1’.

70

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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. 41–3 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. Alf¨oldy, Die r¨omischen Inschriften von Tarraco, Madrider Forschungen 10, 2 vols. (Berlin, 1975), 1:438–9, 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, 1991–3), 3:1111, no. 3180 = Handley, Dying on ˇ sel and J. Saˇ ˇ sel, Inscriptiones latinae quae in Iugoslavia inter Foreign Shores, no. 135 (Aquileia); A. Saˇ 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. 247–8, ed. C. Lambot, in S. Fulgentii episcopi Ruspensis opera, ed. J. Fraipont, 2 vols., CCSL 91–91A (Turnhout, 1968), 2:884 for Vandal-era African ship masters. On African economic connections abroad, see also below, section 2.

71

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Maria’s former handmaid, now fellow slave, continued to tend to her one-time mistress’s needs, washing her feet, looking after her bedding, and so on. The slave girl’s 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 city’s garrison pooled their resources to pay Maria’s 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 Maria’s 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 Maria’s 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 Theodoret’s 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. 37–8. Theodoret, Ep. 70, 2:152: ‘'@ A  *0  ; & B*0’. Ibid., 2:152–4. 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. 80–8. Courtois, Vandales, p. 280 n. 8 is sceptical about Maria’s Carthaginian origins. Courtois, Vandales, p. 280, notwithstanding the statements of Vict. Vit. 1.14, pp. 7–8 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 Theodoret’s own affairs).

72

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Maria’s owners, who from Theodoret’s 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 Theodoret’s 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 Theodoret’s mind this was something due to Maria primarily because of her noble birth, and only secondarily (if at all) because she was a Roman. Maria’s story unexpectedly illustrates another aspect of movement in the late antique Mediterranean: the importance of companions. Most of our fifty-four African e´ migr´es 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. 223–4. Brusin, Inscriptiones Aquileiae 3:1111, no. 3180 = Handley, Dying on Foreign Shores, no. 135; G. Rinaldi, ‘Osservazioni sull’epitaffio di Restuto’, in Aquileia e l’Africa, Antichit`a altoadriatiche 5 (Udine, 1974), pp. 181–9. Theodoret, Epp. 29, 31, and 35–6, 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. 3.29–30, pp. 86–7. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 402–5.35 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 35–6, 2:88, 90, and 98. Eudaemon: Theodoret, Ep. 70, 2:152. Gordian: V. Fulg. 1, p. 11. He was in Rome for Theoderic’s visit there in 500 (V. Fulg. 9, pp. 55–7). 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 1:507–13, s.n. ‘Fulgentius 1’. Silvagni, Monumenta epigraphica christiana 2/1, tab. 9.5 = Handley, Dying on Foreign Shores, no. 102 and Alf¨oldy, Die r¨omischen Inschriften von Tarraco i. 438–9, no. 996 = Handley, Dying on Foreign Shores, no. 68. Theodoret, Ep. 70, 2:152–4 with M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain, ad 300–750, BAR International Series 1135 (Oxford, 2003), pp. 98–100, 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 1:257–8, s.n. ‘Cyprianus 3–4’ and the sources cited there; see also Az´ema, Correspondence, 2:130 n. 1.

74

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 439–50; 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 province’s 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. 391–2; and Ferrandus, Breviatio canonum 121, ed. C. Munier in CCSL 149:297. Twenty: Canones in causa Apiarii 16, p. 105. See PCBE 1:947–9, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman 14

Number of groups

12 10 8 6 4 2 0

50

9– 43

Fig. 2.3.

–7

1 46

0

0

0

0

–6

1 45

–8

1 47

–9

1 48

00

–5 91

4 First visible

0

1 1– 50

0

–2

1 51

0

–3

1 52

33 1– 3 5

Africans abroad, 439–533: 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications

Map 2.1.

Africans abroad in the Mediterranean, 439–533

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 474–91) and the empress Ariadne. There they appear to have remained quite the cause c´el`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 Huneric’s 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, 1:345–6; 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. 192–9. Marcellinus comes, Chronicon s.a. 484.2, p. 93; Aeneas of Gaza, Theophrastus, PG 85, cols. 1000–1.

77

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman Table 2.1. Africans abroad, c.439–c.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

1

Africa – Cyrrhus – Ankyra – Constantinople

440/9

bishop (refugee)

Januarius Possessor – Justin

1 2

Byzacena – Constantinople Zabi – Constantinople

448 517

bishop bishop and deacon

Priscian

1

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

1

448?/451

bishop

Valerian 2

1

Africa – Constantinople? – Chalcedon Bassianensis – Constantinople? – Chalcedon

448?/451

bishop

1

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. 163–4. Vict. Tonn. s.aa. 479.50 and 566/7.173, pp. 16 and 54; Gregory I, Dialogi 3.32, ed. A. de Vog¨ue´ , in Diologues, 3 vols., SC 251, 260, 265 (Paris, 1978–80), 2:390–2. Perhaps the second-hand nature of his information helps to explain why Gregory misdates the persecution to the time of Justinian.

78

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 1:128–9, 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 1:1050–1, 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 1:129–130, s.n. ‘Aurelius 5’. Gesta actionis primae 330, in ACOec. 2/1/1:119–20 (Greek) and ACOec. 2/3/1:100 (Latin); PCBE 1:1136–8, s.nn. ‘Valerianus 1–2’. 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 1:129–30, 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 1:1136–8, s.nn. ‘Valerianus 1–2’. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 491–518).50 Prominent African landowners, presumably living in Constantinople, were said to have had Justinian’s 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, 1:92–4. Sacconius: Gelasius, Ep. 9, ed. Thiel in Epistolae Romanorum pontificum, pp. 339–41; see also Notitia, Proc. 7, p. 117. Hormisdas, Epp. 31, 115, and 124, pp. 805–6, 916–17, and 926–31. 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, 1855–9), 2:1. On the association of Caesarea with Caesarea in Mauretania, see Hertz, ibid., 2:p. vii; M. Schanz and C. Hosius, Geschichte der r¨omischen Literatur bis zum Gesetzgebungswerk des Kaisers Justinian, 4 vols. (Munich, 1907–20), 4/2:221–2; 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. W¨unsch (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 bc–ad 600 (Cambridge, 2007), pp. 259–70 and 516–76. John Lydus, De Magistratibus 3.73, pp. 165–7. On the date, see PLRE 2:881–2, 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. 65–7 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 2:1024–5, s.nn. ‘Speciosa’, and ‘Speciosus 1–2’, were all active in Italy, as was PLRE 3:1181, s.n. ‘Speciosus 2’. In addition to these individuals, sixth-century Italy boasted PCBE 2/2:2100–2, s.nn. ‘Speciosus 2–6’. With

80

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 East–West 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 2.1–2). These refugees included bishops and lay people, four men and one woman, all of whom seem to have followed different paths to northern Syria. Theodoret’s 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 419–58).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. 8–9, pp. 47–55. Vict. Vit. 1.29, p. 13. Fulgentius of Ruspe, De fide ad Petrum seu de regula fidei, ed. J. Fraipont, in CCSL 91A:711–60, esp. ibid., 1, p. 711. Theodoret, Ep. xxiii, 1:94 (Maximian), and Epp. 52–3, 2:128–30 (Cyprian). Az´ema is of the opinion that the last of these two letters were written together, and that, like the letters for Florentius (Theodoret, Ep. xxii, 1:92–4) and Maximian, they were written before the composition of those for Caelestiacus (Theodoret, Epp. 29–36, 2:86–100). As Az´ema dates the latter to ad 443–4, Theodoret, Epp. xxiii and 52–3 may be dated slightly earlier (Az´ema, Correspondence, 2:128–9, n. 2).

81

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

No.

Itinerary

Anonymous1

2+

Africa – East

468

merchants

Caelestiacus – wife – children – slaves

6

Carthage – Cyrrhus – Syria

c.440

aristocrat (refugee)

Cyprian

1

Africa (Thuburbo Maius?) – Ankyra – Cyrrhus – Edessa – Constantina in Osrhoene

435/57

bishop (refugee)

Maria – handmaiden

2

Africa – Cyrrhus – Aegae – West

443/8

slaves

Maximian

1

439/43

Quintianus

1

Carthage – Jerusalem – Cyrrhus – ? Urusi – Edessa in Macedonia

aristocrat (refugee) bishop (refugee)

1

Date

484/489?

Status

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

bishops twice as often as to lay people.58 Theodoret’s 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 A¨erius, 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. 31–2, 35–6, 52–3 and 70, 2:90–2, 96–8, 128–30. Laity (4 people): Theodoret, Ep. xxiii, 1:94, and Epp. 29–30 and 33–4, 2:86–8 and 94–6. Theodoret, Epp. 31–2, 35–6 and 52–3, 2:90–4, 96–100, and 128–30. Ankyra: Theodoret, Ep. xxii, 1:92 with n. 7. 61 Theodoret, Ep. 117, 3:72–4. Theodoret, Ep. xxiii, 1:94 and Ep. 30, 2:88–90.

82

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications Table 2.3. Africans abroad, c.439–c.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 before 492 465 487 465 487

diviner bishop monks (pilgrims) Pope bishop bishop bishop bishop

Victor

1

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

1 1 1 1 1

not their support had any connection to Theodoret’s 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 492–6), 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, 2:128–30. 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. 15–30. Cass. Var. 3.53, pp. 137–8. 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 2:78–80, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman Table 2.4. Africans abroad, c.439–c.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

1

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

1

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 Theoderic’s 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 438–50 (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 1.7.27–8, 2:37. Hilarius, Ep. 15.1, ed. Thiel in Epistolae Romanorum pontificum, p. 160. Felix II (III), Ep. 13.1–3, ed. Thiel in Epistolae Romanorum pontificum, pp. 259–61. V. Fulg. 9, pp. 55–7.

84

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 father’s 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. 140–1. Cass. Var. 12.9, pp. 473–4. The date of the letter (ad 533/7) probably places the exchange in Africa’s 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 1:947–9, 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 d’Italia 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 d’Abitina, la cit´e des c´el`ebres martyrs africains’, Comptes rendus des s´eances de l’Acad´emie des inscriptions et belles-lettres (1976), 255–66. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Africa’s sustained economic connections to the 75 76

77 78 79

Vict. Vit. 3.29, pp. 86–7. 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, 1937–42), 2:299–300; 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. 37–8. 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. 44–5 (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. 95–6. 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 1/2:527–8.

86

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 John’s 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. 209–10; Handley, Dying on Foreign Shores, p. 11. On Liwirit and Ampelius, see also Cass. Var. 5.39, pp. 212–15. 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 fun´eraires chr´etiennes de la basilique dite de Sainte-Monique a` Carthage, ´ Collection de l’Ecole franc¸aise de Rome 25 (Rome, 1975), p. 289, no. 200 and p. 345, no. 378 and L. Ennabli, Les Inscriptions fun´eraires chr´etiennes de Carthage II: la basilique de Mcidfa, Collection ´ de l’Ecole franc¸aise de Rome 62 (Rome, 1982), p. 277, no. 510 = Handley, Dying on Foreign Shores, nos. 361–3, all of whom are commemorated as peregrini. AL 288, 319, and 348–9, pp. 240, 258, and 273–5. 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. 1.19–21, pp. 9–10; Marcellinus comes, Chronicon s.a. 435.2, p. 79. See also PLRE 2:983–4, s.n. ‘Sebastianus 3’. Vict. Vit. 1.23, p. 11; Schmidt, Geschichte der Wandalen, pp. 93–4.

87

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 T´ebessa (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 Africa’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Maria’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Justinian’s 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 Africa’s 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 region’s 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 Africa’s 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 d’occidente e cristianit`a d’oriente (secoli VI–IX), 2 vols., Settimane di studio della Fondazione Centro italiano di studi sull’alto medioevo 51 (Spoleto, 2004), 2:1228–38. 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–), 3/2:613–97; P. Reynolds, Trade in the Western Mediterranean, ad 400–700: 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 c´eramique’, 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. 181–208; M. McCormick, Origins of the European Economy: Communications and Commerce, ad 300–900 (Cambridge, 2001), pp. 100–3; and C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford, 2005), pp. 708–20.

90

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Africa’s 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. Roug´e, SC 124 (Paris, 1966), p. 200. Basic studies include J. W. Hayes, Late Roman Pottery (London, 1972), pp. 13–299 with J. W. Hayes, A Supplement to Late Roman Pottery (London, 1980), pp. 484–523; 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), 33–56. 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 60–1, pp. 200–2 (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. 52–3 and 116–18; A. H. M. Jones, The Later Roman Empire, 284–602: A Social, Economic and Administrative Survey, 2 vols. (Oxford, 1964; repr. Baltimore, Md., 1986), 1:698 and 2:849–50; 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. 271–96. Expositio totius mundi 61, p. 200. Augustine, Confessions 6.9.14, ed. J. J. O’Donnell, 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. 1.24–5, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Foc¸a, 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., 1.22.17–18, 1:406–7 (Peloponnese); and Vict. Tonn. s.a. 455.15, pp. 7–8 (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 38.1–2, pp. 334 and 340 (Italy and Sicily); Sidonius Apollinaris, Carmen 5.388–90, 1:43 (Campania); and Proc. BV 1.5.23–5, 1:334–5 (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 275–425 (Cambridge, 2011). Hayes, Late Roman Pottery, p. 423; Panella, ‘Merci e scambi’, pp. 624–41; and Wickham, Framing, pp. 708–11. See also J. T. Pe˜na, ‘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 Colony’s Economy, Water Supply, a Public Bath, and the Mobilization of State Olive Oil, Journal of Roman Archaeology Supplementary Series 28 (Portsmouth, RI, 1998), pp. 117–238. 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 aˆ ge’, in Morfologie sociali e culturali in Europa fra tarda antichit`a e alto Medioevo, Settimane di studio del ´ Centro italiano di studi sull’alto Medioevo 45 (Spoleto, 1998), pp. 77–8; Bonifay, Etudes sur la c´eramique, p. 480. Panella, ‘Merci e scambi’, pp. 644–5 and Hayes, Late Roman Pottery, p. 423. Panella, ‘Merci e scambi’, pp. 639–40 and 644–5; Hayes, Late Roman Pottery, pp. 323–401 and 417–23; Wickham, Framing, pp. 714–15. Sodini, ‘Productions et e´ changes’, p. 188; Panella, ‘Merci e scambi’, p. 645; Hayes, Late Roman Pottery, pp. 417–18; C. Abadie-Reynal and J.-P. Sodini, La C´eramique pal´eochr´etienne de Thasos

92

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 city’s 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

´ ´ (Aliki, Delkos, fouilles anciennes), Ecole franc¸aise d’Ath`enes Etudes thasiennes 13 (Athens, 1992), pp. 79–83 and see also ibid., pp. 87–90; L. Anselmino and C. Pavolini, ‘Terra sigillata: lucerne’, Enciclopedia dell’arte antica, classica e orientale: Atlante delle forme ceramiche, vol. 1 (Rome, 1981), pp. 199–201. Sodini, ‘Productions et e´ changes’, pp. 188–90; Hayes, Late Roman Pottery, pp. 419–20. ´ Sodini, ‘Productions et e´ changes’, p. 190; Bonifay, Etudes sur la c´eramique, pp. 454–6; Hayes, Late Roman Pottery, pp. 420–1. ´ Bonifay, Etudes sur la c´eramique, pp. 454–6. ´ Panella, ‘Merci e scambi’, p. 645; Bonifay, Etudes sur la c´eramique, p. 456; Sodini, ‘Productions et e´ changes’, pp. 191–3. On olive oil in late antique Egypt, see Bagnall, Egypt in Late Antiquity, pp. 30–1, 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. 54–5 and 57; P. Pentz, From Roman Proconsularis to Islamic Ifr¯ıqiyah, 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. 266–72. 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. 85–124, esp. 121. Wickham, Framing, p. 711; Pentz, From Roman Proconsularis, pp. 151–4; Reynolds, Trade, p. 113; Panella, ‘Merci e scambi’, pp. 641–2 and 644. ´ J. Durliat, De la ville antique a` la ville byzantine: le probl`eme des subsistances, Collection de l’Ecole franc¸aise de Rome 136 (Rome, 1990), pp. 110–23 and fig. 1, p. 117.

93

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Vinalop´o 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 region’s 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. 114–15; 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. 247–59; 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), 2:420–7; Reynolds, Trade, pp. 113–14. Reynolds, Trade, pp. 114–15 and Hayes, Late Roman Pottery, p. 415. ´ Bonifay, Etudes sur la c´eramique, p. 457. Ibid., pp. 452 and 457; Sodini, ‘Productions et e´ changes’, p. 188; Hayes, Late Roman Pottery, pp. 402–4 and 415–16. M. Bonifay, ‘Observations sur les amphores tardives a` Marseille d’apr`es les fouilles de la Bourse (1980–1984)’, Revue arch´eologique de Narbonnaise 19 (1986), pp. 297–8; M. Bonifay and D. Pi´etri, ‘Amphores du Ve au VIIe s. a` Marseille: nouvelles donn´ees sur la typologie et le contenu’, ´ Journal of Roman Archaeology 8 (1995), p. 116; Bonifay, Etudes sur la c´eramique, p. 457.

94

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Africa’s 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. 135–6. ´ Bonifay, Etudes sur la c´eramique, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman Table 2.5. Travel from the African interior to the nearest port, c.439–c.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 31–4. 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 2.1–2.4. 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 c´eramique, pp. 454–6. Tablettes Albertini: actes priv´es de l’´epoque 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Januarilla’s 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. 213–27; D. J. Mattingly, ‘Olive Cultivation and the Albertini Tablets’, in A. Mastino (ed.), L’Africa romana: Atti del VI convegno di studio, Pubblicazioni del Dipartimento di storia dell’Universit`a di Sassari 14 (Sassari, 1989), pp. 403–15; 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. 124–42. 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), 1–6; and McCormick, Origins, pp. 75–7. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Huneric’s 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 city’s 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, Africa’s 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. 117–34; Concilium Carthaginense a. 525, pp. 255–82; Passio septem monachorum, ed. M. Petschenig in CSEL 7:108–14. 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 336–7, pp. 267–8; V. Fulg. 1 and 14, pp. 13 and 73; Proc. BV 1.5.15, 1.10.26–7, and 2.3.26, 1:333–4, 1:359, and 1:432; Liebeschuetz, ‘Gens into Regnum’, pp. 75–6; Wickham, Framing, pp. 89–92. ´ Keay, Late Roman Amphorae, 1:126–9; Bonifay, Etudes sur la c´eramique, p. 451. 130 Ibid., pp. 451–2. ´ Bonifay, Etudes sur la c´eramique, pp. 49 and 451.

98

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications the case, for example, in the Cillium-Thelepte area, in S´etif (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 Africa’s 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. A¨ıt Kaci, and N. Bounssair, ‘Prospections dans le Belezma: Rapport Pr´eliminaire’, in Actes du colloque international sur l’histoire de S´etif, Bulletin d’arch´eologie alg´erienne, Sup. 7 (Algiers, 1993), pp. 111–12; L. Neuru, ‘The Pottery of the Kasserine Survey’, Antiquit´es 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), 2:321–5; Hayes, ´ Late Roman Pottery, pp. 300–9; Pentz, From Roman Proconsularis, pp. 143–5; Bonifay, Etudes sur la c´eramique, pp. 53–7; 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. 128–9. Wickham, Framing, p. 721. B. Shaw, ‘Rural Markets in North Africa and the Political Economy of the Roman Empire’, Antiquit´es africaines 17 (1981), pp. 37–83; 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. 155–96. Wickham, Framing, pp. 698–9 and 722. Figs: Tablettes Albertini, acts 3, 4, 10–12, 16, 24, and 26, 1:218–27, 249–58, 268–9, 282–3, and 288; presumably also ibid., act 6, 1:233–6, a caprificus. Siteciae: ibid., acts 4 and 6, 1:223–7 and 233–6. Almonds: ibid., acts 4 and 19, 1:223–7 and 274 (the same cultivator). CIL 8.4508 (Zarai); R. Cagnat, ‘A New Roman Customs List’, JRS 4 (1914), pp. 142–6 (Lambaesis).

99

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Africa’s 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, 17–18, 20–1, and 26–8, pp. 73–7, 87–93, 99–107, and 123–35. 139 V. Fulg. 27, pp. 131–3. V. Fulg. 3 and 5–8, pp. 21–3 and 33–47. ´ Bonifay, Etudes sur la c´eramique, p. 451; Neuru, ‘Pottery of the Kasserine Survey’, p. 259; and see above, n. 128. M. L. Saccardy, conservateur des Eaux et Forˆets, chef du Service de la Conservation des Sols au G. G. de l’Alg´erie 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. 1–15 and D. D. Gilbertson, ‘Explanations: Environment as Agency’, in Barker, Gilbertson, Jones, and Mattingly, Farming the Desert, 1:291–317.

100

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications Table 2.6. The travels of Fulgentius of Ruspe, c.484–c.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. 5–6) Near Sicca – Medidi (cc. 7–8) Medidi – Carthage – Syracuse (c. 8) Syracuse – island off Sicily – Rome (c. 9) Rome – Sardinia – Medidi (c. 10) Medidi – estate of Silvestrio in Byzacena (cc. 10–11) Byzacena –monastery near Iunca (c. 12) Monastery near Iunca – monastery in Byzacena (c. 13) Monastery in Byzacena – Ruspe (c. 14) Ruspe – Carthage – Cagliari (cc. 17–18) 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. 26–7) 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

1

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 1994–1996): una sintesi preliminare’, in M. de Vos, Rus Africum: Terra, acqua, olio nell’Africa settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000), pp. 63 –5; Wickham, Framing, p. 721.

101

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman Exile within the territory of the Vandal kingdom also figures as a persistent theme in accounts of the Arian persecution of Africa’s 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 Gelimer’s 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 Mouza¨ıaville, 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. T´en`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. 95–6; Vict. Tonn. s.a. 479.50, p. 16; Laterculus regum Wandalorum 5 (Augiensis liber), p. 458. Vict. Vit. 2.26–37, pp. 33–8, 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. Bishop’s 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. 3.29–30, pp. 86–7; Proc. BV 1.10.25–6, 1.11.22–4, and 1.24.1, 1:359, 1:363–4, 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 Mod´eran, ‘Notitia provinciarum’, pp. 182–5. AE (1967), 212, no. 651 = Handley, Dying on Foreign Shores, no. 425. Vict. Vit. 2.10, p. 27.

102

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 bishop’s 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 Africa’s 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. 9–11, 13–15, 17–18, 22–3, 26, 28–30, 32–3, 35–40, 42, pp. 117–18, and ibid., Num. 68, p. 121; perhaps also ibid., Proc. 2–7, 24, 27, 41, and 44–6, pp. 117–19, and ibid., Byz. 20, 72, and 99, pp. 124, 126, and 127, mentioned simply as in exile. Sardinia: V. Fulg. 17–18, pp. 87–91, 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. 18–19, pp. 91–7.

103

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman rarely be attributed specifically to the Vandal period, they are nevertheless revealing of the general directions of Africa’s 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 498–514) 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 A¨ın Be¨ıda, 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. 327–41; 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. 31–54. Sigisteus: Epistola ad Parthemium, PLS 3:447–8 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. 41–2. Symmachus, Ep. 11, ed. Thiel in Epistolae Romanorum pontificum, pp. 708–9 = 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. 169–74. Liber Pontificalis 53.11, ed. Mommsen, p. 125; Stevens, ‘Circle of Bishop Fulgentius’, p. 338. See above, n. 49. CIL 8.2297 (A¨ın-Be¨ıda) = Cass. Var. 7.7.2, p. 268; Moorhead, Theoderic, p. 64 n. 133. See, however, Diesner, Auswirkungen der Religionspolitik, p. 5.

104

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Galla’s 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 Italy’s 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, 3:54–8; PLRE 2:491, s.n. ‘Galla 5’; Stevens, ‘Circle of Bishop Fulgentius’, p. 334. Fulg. Ep. 2.31, 1:208; see also Fulg. Epp. 3–4, 1:212–35. 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, 1:240–4; Gelasius, Ep. 41, p. 454; PLRE 2:1097–8, 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 f¨ur alte Geschichte 33 (1984), pp. 107–15. Fulg. Ep. 7, 1:244–54; 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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; Venantia’s 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, 1:235–40; Stevens, ‘Circle of Bishop Fulgentius’, p. 331. Romulus’s name suggests some intriguing possibilities: Moorhead, ‘Decii under Theoderic’, pp. 111–12, but see also PCBE 2/2:1916–19, s.nn. ‘Romulus 1–9’. Fulg. Epp. 6.1 and 7.1–2, 1:240 and 245. Stevens, ‘Circle of Bishop Fulgentius’, pp. 331–9, but for Theodore (and perhaps Venantia) see also Moorhead, ‘Decii under Theoderic’. V. Fulg. 25, p. 119. Fulg. Epp. 11–14 and 18, 1:357–444 and 2:619–24; Ferrandus, Ep. 7.2, PL 67:929C; Stevens, ‘Circle of Bishop Fulgentius’, pp. 336–7. 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:649–707; Fulgentius, Ad Monimum, ed. Fraipont in CCSL 91:1–64; and probably Fulg. Ep. 1, 1:189–97 also belong to this period. Return to Africa: Fulg. Epp. 8, 9, 10, 1:255–356; and Fulgentius, Contra Fabianum, ed. Fraipont, CCSL 91A:763–866. The Felix to whom Fulgentius wrote his Liber de Trinitate ad Felicem, ed. Fraipont, CCSL 91A:633–46 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. vi–vii.

106

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Christ’s 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 theologian’s early anti-Manichaean writings from his childhood friend Alypius, then 170

171

172 173 174

Fulg. Ep. 16, 2:551–62; R. H. Weaver, Divine Grace and Human Agency: A Study of the SemiPelagian Controversy, Patristic Monograph Series 15 (Macon, Ga., 1996), pp. 181–2 and see also ibid., pp. 165–80. Fulg. Epp. 15 and 17, 2:447–57 and 563–615, and Fulgentius, De veritate praedestinationis et gratiae, ed. J. Fraipont, CCSL 91A:458–548; V. Fulg. 25, pp. 119–20; Stevens, ‘Circle of Bishop Fulgentius’, p. 339; Weaver, Divine Grace, pp. 180–96. Hormisdas, Ep. 124.5, pp. 930–1, cited in Fulg. Ep. 15.18, 2:456. Fulg. Ep. 18, 2:619–24. Fulg. Ep. 5.12, 1:240; see also (in Africa) Fulg. Ep. 13.3, 1:386–7.

107

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Augustine’s 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 Augustine’s Literal Interpretation of Genesis and perhaps also his City of God in advance of compiling a florilegium of excerpts of the great theologian’s corpus around 500.178 Eugippius himself assumed that Augustine’s works were available above all in Rome, and indeed Jean-Paul Bouhot has argued the bishop’s library was transferred there from Hippo towards the beginning of Leo I’s pontificate (ad 440–61), 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 Augustine’s Confessions further survive in a sixth-century manuscript, as do the theologian’s 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 Capella’s 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. 3–4, ed. W. Hartel, CSEL 29 (1894), pp. 13–24; D. Trout, Paulinus of Nola: Life, Letters, and Poems, Transformation of the Classical Heritage 27 (Berkeley, Calif., 1999), pp. 202–6. CLA 11.1613; W. Green, ‘A Fourth Century Manuscript of Saint Augustine?’, Revue B´en´edictine 69 (1959), pp. 191–7. CLA 4.491; M. Gorman, ‘The Manuscript Traditions of St Augustine’s 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. 315–46; Graham, ‘Dissemination’, pp. 153–4. Eugippius, Excerpta ex operibus s. Augustini, ed. P. Kn¨oll, CSEL 9/1 (Vienna, 1885); M. Gorman, ‘Chapter Headings for St Augustine’s De Genesi ad litteram’, Revue des e´tudes augustiniennes 26 (1980), pp. 88–104 and M. Gorman, ‘A Survey of the Oldest Manuscripts of St Augustine’s De ciuitate dei’, JThS n.s. 33 (1982), pp. 398–410, repr. in M. Gorman, Manuscript Traditions as essay 2, pp. 44–60 and essay 6, pp. 178–90, respectively; Graham, ‘Dissemination’, p. 146. Eugippius, Epistula ad Probam virginem, ed. Kn¨oll, in Excerpta, pp. 1–2; J.-P. Bouhot, ‘La Transmission d’Hippon a` Rome des œuvres de saint Augustin’, in D. Nebbiai-Dalla Guarda and J.-F. Genest (eds.), Du copiste au collectionneur: m´elanges d’histoire des textes et des biblioth`eques en l’honneur d’Andr´e Vernet, Bibliologia Elementa ad Librorum Studia Pertinentia 18 (Turnhout, 1998), pp. 23–33. Gorman, ‘Manuscript Traditions’, 1:402. CLA 4.420a, 5.587, and 6.777; Gorman, ‘Manuscript Traditions’, 1:384.

108

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 bishop’s 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 advocate’s poems had spread from Africa to northern Italy and the Rhˆone 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 Le´on 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 Capella’s De Nuptiis Philologiae et Mercurii Book 1, University of California Publications in Classical Studies 32 (Berkeley, Calif., 1986), pp. 8–28; Graham, ‘Dissemination’, pp. 144–5. Pseudo-Vigilius of Thapsus, Contra Varimadum, ed. B. Schwank, in Florilegia Biblica Africana saec. V, CCSL 90 (Turnhout, 1961), pp. 1–134, with pp. vii–viii and xv; Caesarius of Arles, Breviarium adversus haereticos, ed. Morin, in Caesarii Arelatensis Opera omnia, 2:182–208; Graham, ‘Dissemination’, p. 188. Fulg. Ep. 5.12, 1:239; CLA 1.104a, see also ibid., 104b; Graham, ‘Dissemination’, pp. 190–2. 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, Œuvres, 1:100–1. ILCV 1645 (Le´on, ad 630) = Dracontius, Laudes Dei 1.611, ed. C. Moussy and C. Camus, in Œuvres, 1:183: ‘Omnibus hic mos est de flammis tollere flammas’; Moussy, ‘Introduction’ to Dracontius, Œuvres, 1:105–6; Graham, ‘Dissemination’, pp. 180–4. On other African texts in Spain, see ibid., pp. 177–80. Moussy, ‘Introduction’ to Dracontius, Œuvres, 1:102; Graham, ‘Dissemination’, pp. 165–7.

109

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Rhˆone 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 theologian’s body was moved twice, for before being brought to Pavia by the Lombard king Liutprand (ad 712–44) the bishop’s 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’, 160–7. See, e.g., T. Ruinart, Historia persecutionis Vandalicae (Paris, 1694), pp. 580–2; S. A. Morcelli, Africa christiana, 3 vols. (Brescia, 1816–17), 3:241; de Rossi, Roma sotterranea, 2:221–5; 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., 2:1093–103; 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 l’Ecole franc¸aise de Rome (Paris, 1982), 1:30 no. 13 (Thibiuca) and 2:729–30. Conant, ‘Europe and the African Cult of Saints’.

110

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Charlemagne’s 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. 189–98. However, see also Graham, ‘Dissemination’, pp. 79–84. Florus of Lyons, Carmina 13 (Rector magnificus) and 14 (Hac locuples), ed. E. D¨ummler, MGH Poet. 2 (Berlin, 1884), pp. 544–6; Florus, Martyrologium, in H. Quentin (ed.), Les Martyrologes historiques du moyen aˆ ge: e´tude 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. 507–14; and see also Annales Regni Francorum s.aa. 801–2, ed. G. H. Pertz and F. Kurze, MGH SRG 6 (Hanover, 1895; repr. 1950), pp. 116–17. On this translation, see McCormick, Origins, p. 891 no. 257 and ibid., pp. 890–1 nos. 254–6, pp. 890–1, but also C. Courtois, ‘Reliques carthaginoises et l´egende carolingienne’, Revue de l’histoire des religions 129 (1945), pp. 57–83. P. Chiesa, ‘Un testo agiografico africano ad Aquileia: gli Acta di Gallonio e dei martiri di Timida Regia’, Analecta Bollandiana 114 (1996), pp. 241–68 (Gallonius); B. de Gaiffier, ‘S. Marcel de Tanger ou de L´eon? e´ volution d’une l´egende’, Analecta Bollandiana 61 (1943), pp. 116–39 and (on the date of the manuscript) B. de Gaiffier, ‘Les Notices hispaniques dans le martyrologe d’Usuard’, Analecta Bollandiana 55 (1937), p. 271 n. 2 (Marcellus of Tingi); P. Chiesa, ‘Pellegrino martire in urbe Bolitana e Pellegrino di Ancona: un’altra agiografia africana ad Aquileia?’, Analecta Bollandiana 116 (1998), pp. 25–56 (Peregrinus); Peter the Subdeacon, Passio sanctae Restitutae, ed. E. D’Angelo in Pietro Suddiacono napoletano: L’opera agiografica (Florence, 2002), pp. 186–99 (Restituta); Usuard, Martyrologium, ed. J. Dubois, in Le Martyrologe d’Usuard, 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. 138–41 and Translatio duodecim martyrum, ed. Waitz in MGH SRL, pp. 574–6 (Twelve Brothers of Hadrumetum); J. de Guibert, ‘Saint Victor de C´esar´ee’, Analecta Bollandiana 24 (1905), 257–64 (Victor of Caesarea). For a seemingly earlier instance of this phenomenon, see H. Delehaye, ‘La Passion de S. F´elix de Thibiuca’, Analecta Bollandiana 39 (1921), 241–76. 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 hisp´anico de San Mill´an 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, 1889–93), 3:504; A. F´abrega Grau, Pasionario

111

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

hisp´anico (siglos VII–XI), 2 vols., Monumenta hispaniae sacra, Serie lit´urgica 6 (Madrid, 1953–5), 1:227–8; de Guibert, ‘Saint Victor de C´esar´ee’, pp. 257–60 and 262–4. H. Delehaye, Les Origines du culte des martyrs, Subsidia Hagiographica 20 (2nd edn; Brussels, 1933), pp. 50–3; McCulloh, ‘Cult of Relics’; Duval, Loca sanctorum, 2:549; Conant, ‘Europe and the African Cult of Saints’, pp. 42–3. Conant, ‘Europe and the African Cult of Saints’, pp. 4–34. Passio s. Polyeucti, ed. B. Aub´e, in Polyeucte dans l’histoire (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. 44–5 and Gregory, In gloria martyrum 57, pp. 527–8; 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. 287–9 and see above, previous n.

112

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications ecclesiastical calendar.201 By c.400 Cyprian’s 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 Perpetua’s visions, and at the same time the Greek version of her passio was known in the East.203 The later universality of Cyprian and Perpetua’s 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 d’Italia 81, 88, 90, and 91 (Rome, 1940–53), 2:18 (Perpetua) and ibid., s.d. xviii Kal. Oct., p. 26 (Cyprian). Prudentius, Peristefanon 11.237–8, ed. M. Cunningham, CCSL 126 (Turnhout, 1966), pp. 377–8 (dies sollemnis); see also ibid., 13, pp. 382–5 (passion) and ibid., 4.17–18, p. 286; Paulinus of Nola, Carmen 19.141–3, ed. W. Hartel, CSEL 30 (Vienna, 1894), p. 123. H. Schlunk, ‘Zu den fr¨uhchristlichen Sarkophagen aus der Bureba (Prov. Burgos)’, Madrider Mitteilungen 6 (1965), 139–66 at pp. 145–66; H. Schlunk and T. Hauschild, Die Denkm¨aler der fr¨uhchristlichen und westgotischen Zeit (Mainz am Rhein, 1978), pp. 141–3, and pl. 35; and Passio sanctarum Perpetuae et Felicitatis 4.3–7, ed. C. van Beek (Nijmegen, 1936), pp. 12–14 (Latin) and 13–15 (Greek). H. Delehaye, ‘Cyprien d’Antioche et Cyprien de Carthage’, Analecta Bollandiana 39 (1921), 314–32; 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., 1:137–40, nos. 452–61c. Zeno of Verona, Tractatus 1.39, ed. B. L¨ofstedt, CCSL 22 (Turnhout, 1971), pp. 107–10; Prudentius, Peristefanon, 4.45–8, p. 287; Paulinus of Nola, Carmen 19.144–8, p. 123.

113

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 saint’s 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 Llewellyn’s 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’, 36–45. 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), 105–24 with R. Bagnall, ‘Conversion and Onomastics: A Reply’, Zeitschrift f¨ur Papyrologie und Epigraphik 69 (1987), 243–50; L. Robert, Noms indig`enes dans l’Asie-Mineure gr´ecoromaine, Biblioth`eque arch´eologique et historique de l’Institut franc¸ais d’arch´eologie d’Istanbul 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 (498–506): A Preliminary Analysis’, Ancient Society 8 (1977), 245–75.

114

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Llewellyn’s 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 chr´etienne 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman Table 2.7. Comparison of nine names in PCBE 1–2 (Africa and Italy) before the Vandal capture of Carthage, c.300–439 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 chr´etienne du Bas-empire, vol. 1, Prosopographie de l’Afrique chr´etienne (303–533), ed. A. Mandouze (Paris, 1982); vol. 2, Prosopographie de l’Italie chr´etienne (313–604), ed. C. Pietri and L. Pietri (Rome, 1999–2000). 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 439–533), 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Flight and communications Table 2.8. Comparison of nine names in PLRE 1–2 before the Vandal capture of Carthage, c.260–c.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

1

1 1

2 1 4

1 3

1

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, 1971–92), vol. 1 (ad 260–395) and vol. 2 (ad 395–527).

Table 2.9. Comparison of nine names in PLRE 2–3 in the Vandal period, c.439–533 Name

Italy

Adeodatus Benenatus Cyprian Maximianus Victor

1 1 2 2 3

Constantinople

Thrace

Palestine

Egypt

‘East’

Total

1

1 1 3 3 6

1 1 1

1

Cresconius Donatus Peregrinus Victorinus: not attested Table based on the material in the Prosopography of the Later Roman Empire (Cambridge, 1971–92), vol. 2 (ad 395–527) and vol. 3 (ad 527–641).

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

pre-Vandal (300–439)

Vandal era (439–533)

Adeodatus Benenatus Cresconius Cyprian Donatus Maximianus Peregrinus Victor Victorinus

1 – 1 1 5 1 – 3 3

10 3 1 4 2 3 2 7 3

Table based on the material in Prosopographie chr´etienne du Bas-empire, vol. 2, Prosopographie de l’Italie chr´etienne (313–604), ed. C. Pietri and L. Pietri (Rome, 1999–2000).

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. 310–11.

118

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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. 311–12. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 440–61), 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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

4

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

Rome

5 6 7 8 9 10 11

391 394, 396, or 402 395 408 440–61 474 496

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

4

3

2

1

0 350–9 370–9 390–9 410–19 430–9 450–9 470–9 490–9 510–19 530–9 550–9 570–9 590–9

Date (earliest possible)

Fig. 2.4.

The names Adeodatus and Benenatus in dated Italian inscriptions, 350–599 (CIL and ICVR n.s.)

123

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

Staying Roman 6

Number

5 4 3 2 1 0 410–19 420–9 430–9 440–9 450–9 460–9 470–9 480–9 490–9 500–9 510–19 520–9 530–4

First visible

Fig. 2.5. The names Adeodatus and Benenatus, 410–534 in PCBE 2 (Italy) Figures are based on the materials in Prosopographie chr´etienne du Bas-empire (Paris, 1982–2000), vol. 2, Prosopographie de l’Italie chr´etienne (313–604). 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 5–6’. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 492–6) 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 2/1:19–24, s.nn. ‘Adeodatus 7–12’, 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 676–8); ibid., 81.17, p. 193 (Agatho, ad 678–81); and ibid., 83.4, p. 204 (Benedict II, ad 684–5). The Epitoma Cononiana, ed. Mommsen in MGH Gestorum Pontificum Romanorum 1:258 makes the same claim for Gelasius’ successor Symmachus (ad 498–514). 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 (384–99) Anastasius I (399–401) Innocent I (401–17) Zosimus (417–18) Boniface I (418–22) Celestine I (422–32) Xystus III (432–40) Leo I (440–61) Hilarus (461–8) Simplicius (468–83) Felix II (III) (483–92) Gelasius (492–6)

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 498–514); 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. 259–63. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Christ’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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, Africa’s 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 Africa’s 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. Thompson’s 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 Justinian’s ear, and to have convinced the emperor to reconquer their province for the 224 225

226

V. Fulg. 1, pp. 11–13. 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 400–700’, Reading Medieval Studies 6 (1980), 66–80. Gregory of Tours, Historiae 2.2–3, pp. 39–45 and Gregory, In gloria martyrum 57, pp. 527–8; Cyprian is the only other African saint included in this work, here ibid., 93, p. 550.

128

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Gordian’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:17:53, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.003

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 Victor’s thought world, the Vandals were a savage people hell-bent on the destruction of all things Roman. Victor’s own community – the Romano-African population of Nicene confession – was the New Israel; 1

Vict. Vit. 3.62, pp. 102–3: ‘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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Victor’s 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. Victor’s 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

4

Vict. Vit. 3.64–70, pp. 103–7. T. Howe, Vandalen, Barbaren, und Arianer bei Victor von Vita, Studien zur Alten Geschichte 7 (Frankfurt am Main, 2007), pp. 183–217 and 302–18; 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 57:484–96; Augustine, Sermones 344–45, PLS 3:417–840; Possidius, Vita Augustini 28.4–13, pp. 148–56; Capreolus of Carthage, Epistula ad synodum Ephesinum, in ACOec. 1/2:64–5; 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. 476–8. F. M. Clover, ‘Carthage in the Age of Augustine’, in

131

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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, Africa’s 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 forum’s loss of its civic function. This was not universal, for the forum did survive as a focus of urban activity at Sbe¨ıtla (class. Sufetula); but across Africa the old civic centre of the classical town was for the most part either neglected or

5

J. H. Humphrey (ed.), Excavations at Carthage 1976, Conducted by the University of Michigan, vol. 4 (Ann Arbor, Mich., 1978), p. 14; R. Gonz´alez Salinero, Poder y conflicto religioso en el norte de A´frica: Quodvultdeus de Cartago y los v´andalos, Graeco-Romanae Religionis Electa Collectio (GREC) 10 (Madrid, 2002), pp. 81–5; 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. 184–6. Vict. Vit. 1.3–12, pp. 3–7.

132

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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, Africa’s cities were clearly centres of economic production.11 The region’s 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 Africa’s 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

9

10 11

12 13

P. Pentz, From Roman Proconsularis to Islamic Ifr¯ıqiyah, GOTARC Serie B. Gothenberg Archaeological Theses 22 (Copenhagen, 2002), pp. 51–3; C. Wickham, Framing the Early Middle Ages: Europe and the Mediterranean, 400–800 (Oxford, 2005), pp. 636–7; A. Leone, Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest, Studi storici sulla Tarda Antichit`a 28 (Bari, 2007), pp. 135–7 and 159. Leone, Changing Townscapes, pp. 137–41 and 159; see also above, Chapter 1.3. S. Stevens, ‘S´epultures tardives intra-muros a` Carthage’, in P. Trousset (ed.), L’Afrique du Nord antique et m´edi´evale: monuments fun´eraires, institutions autochtones (Paris, 1995), pp. 207–17; A. Leone, ‘L’inumazione in “spazio urbano” a Cartagine tra V e VII secolo d. C’, Antiquit´e tardive 10 (2002), 233–48; Leone, Changing Townscapes, pp. 147–8 and 160. Pentz, From Africa Proconsularis, p. 44; Wickham, Framing, pp. 637–8; Leone, Changing Townscapes, pp. 148–54; 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 303–4, a church whose first phase was seemingly built in the Vandal era. See in general Pentz, From Africa Proconsularis, pp. 29–75 and Leone, Changing Townscapes, pp. 45–128. 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. 66–89; Pentz, From ´ Africa Proconsularis, pp. 48–9; M. Bonifay, Etudes sur la c´eramique romaine tardive d’Afrique, BAR International Series 1301 (Oxford, 2004), pp. 53–7. Leone, Changing Townscapes, pp. 140–1, but see also ibid., pp. 137–40; see also above, Chapter 1.3. Leone, Changing Townscapes, pp. 145–8 and 161–2.

133

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

Staying Roman the Vandal capital in glowing terms. Among the poet’s 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 Gelimer’s secretary Boniface was one of the king’s 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 Œuvres, 4 vols. (Paris, 1985–95), 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. 283–4, 342–3, 346–8, and 360, nos. 58 (Faustus), 59 (Felicianus), 124 (Pomerius), 126 (Priscian), and 138 (Speciosus); see also ibid., pp. 397–8 and 415–17, nos. 204 (Coronatus) and 235 (Luxorius). Probably also AL 373, pp. 289–90 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. 249–50 and 343–6, 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. 241–4, which places Fulgentius the Mythographer in the Byzantine period. Possibly also Kaster, Guardians of Language, pp. 385–6, no. 189 (Astyagius) and A. Lurius Geminius, a doctor and teacher of rhetoric from fourth- to sixth-century Mactar: Revue arch´eologique, 6th ser. 42 (1953), 180, no. 49. See also in general A. Merrills and R. Miles, The Vandals (Chichester, 2010), pp. 213–19. 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.199–224. Vict. Vit. 3.24 and 3.50, pp. 83 and 96; AL 297 and 304, pp. 244–5 and 249–50; and Merrills and Miles, Vandals, pp. 218–19. D. Shanzer, ‘Intentions and Audiences: History, Hagiography, Martyrdom, and Confession in Victor of Vita’s 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. 186–8 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. 611–12 s.v.

134

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 another’s 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 Medea’s 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. 196–7), Symphosius scholasticus (author of a book of riddles: AL 281, pp. 202–34), Tuccianus (AL 271–2, p. 198), and Vincentius (AL 273, pp. 198–9); note, however, that Modestinus (AL 267, p. 196) may be earlier: M. Schanz and C. Hosius, Geschichte der r¨omischen Literatur bis zum Gesetzgebungswerk des Kaisers Justinian, 4 vols. (Munich, 1907–20), 3:45–8, esp. 47 and ibid., 4/2:69–76. AL 282, pp. 235–6. 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. 159–60 and 162–3. D. Shanzer, A Philosophical and Literary Commentary on Martianus Capella’s De Nuptiis Philologiae et Mercurii, Book 1, University of California Publications in Classical Studies 32 (Berkeley, Calif., 1986), pp. 17–21. Aegritudo perdicae, ed. E. Baehrens and F. Vollmer, in Poetae latini minores, 6 vols. (Leipzig, 1881– 1911), 5:238–50, with ibid., p. viii.

135

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Africa’s 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. 177–86. Regianus: AL 264–6 (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 litt´erature et d’histoire anciennes 47 (1973), pp. 309–14. In addition to the poems cited above, see (on the hunt) AL 295 and 302, pp. 243 and 248–9. 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. 116–18. AL 288, 301, 319, 322–3, 331, 348–9, and 368, pp. 240, 248, 258, 260, 264–5, 273–5, and 285. Sexually active elders: AL 292, 296, 304, and 338–9, pp. 242, 244, 249–50, and 268–9; sodomites: AL 297, 316, and 331, pp. 244–5, 256, and 264–5; 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 356–7, pp. 250–1, 276–7, and 279–80; 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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. Geiseric’s 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 Vita’s 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 Geiseric’s reign, for example, Fulgentius of Ruspe’s 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 family’s landholdings 32 33 34

35 36

Vict. Vit. 1.48–50, 2.23, and 3.27, pp. 21–2, 32, and 85. Proc. BV 1.5.15, 1:333. See also above, Chapter 1.3. Tablettes Albertini: actes priv´es de l’´epoque vandale (fin du Ve si`ecle), ed. C. Courtois, L. Leschi, C. Perrat, and C. Saumagne, 2 vols. (Paris, 1952), acts 3, 4, 6–14, and 17–24, 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, 3:654–7, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

Staying Roman after his father’s 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. 10–11, pp. 59–61. V. Fulg. 1, p. 13. Tablettes Albertini, acts 4, 9–14, 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 Litt´eraires de l’Universit´e de Besanc¸on 447 (2nd edn; Paris, 1991), pp. 47–74 and D. P. Kehoe, The Economics of Agriculture on Roman Imperial Estates in North Africa, Hyptomnemata: Untersuchungen zur Antike und zu ihrem Nachleben 89 (G¨ottingen, 1988), pp. 28–70; but also C. R. Whittaker, ‘Land and Labour in North Africa’, Klio 60 (1978), 331–62, esp. 360–1 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. 213–16. Vict. Vit. 2.3–4, 2.39, and 3.3–14, pp. 25, 39, and 72–8; M. E. Gil Egea, A´frica en tiempos de los v´andalos: continuidad y mutaciones de las estructuras sociopol´ıticas romanas, Memorias del Seminario de Historia Antigua 7 (Alcal´a de Henares, 1998), pp. 318–20. Vict. Vit. 3.9 and 3.11–13, pp. 76–78; Quodvultdeus, Adversus quinque haereses 6.27, p. 284; V. Fulg. 7, pp. 43–5; Gil Egea, A´frica, pp. 283–5 and 287–9; and Gonz´alez Salinero, Poder y conflicto religioso, pp. 95–6. AL 290 and 335, pp. 241 and 266–7; Gil Egea, A´frica, pp. 289–91; Merrills and Miles, Vandals, pp. 218–19.

138

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Gelimer’s 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, Thrasamund’s 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, A´frica, pp. 287–8. AL 336–7, pp. 267–8. For the identification with Boniface, see Rosenblum, Luxorius, p. 219, seemingly accepted by PLRE 2:447, s.n. ‘Eutychus 2’. AL 248, ll. 19–23, 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 F´evrier, ‘Ostraka de Bir Trouch’, pp. 239–49; 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

Staying Roman and merchants gaining a freer hand in the running of the region’s 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–), 3/2:642–3 and Leone, Changing Townscapes, pp. 133–4. D. P. S. Peacock, F. Bejaoui, and N. Ben Lazreg, ‘Roman Amphora Production in the Sahel Region of Tunisia’, in Amphores romaines et histoire e´conomique: dix ans de recherche, Collection de ´ l’Ecole franc¸aise 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), 59–84. See further Leone, Changing Townscapes, pp. 133, 142, and 159–60. 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 dell’antichit`a 2 (1983), pp. 56–8 and Keay, Late Roman Amphorae, 2:420–3. M. Mackensen, Die sp¨atantiken Sigillata- und Lampent¨opfereien von El Mahrine (Nordtunesien): Studien zur Nordafrikanischen Feinkeramik des 4. bis 7. Jahrhunderts, M¨unchner Beitr¨age zur Vorund Fr¨uhgeschichte 50, 2 vols. (Munich, 1993), 1:487–91. Peacock, et al., ‘Roman Amphora Production’, pp. 183–8 and 199–200 and Peacock, et al., ‘Roman Pottery Production’, esp. p. 82. Peacock, et al., ‘Roman Pottery Production’, pp. 66–83. M. de Vos, Rus Africum: Terra, acqua, olio nell’Africa settentrionale: Scavo e ricognizione nei dintorni di Dougga (Alto Tell tunisino) (Trent, 2000), pp. 20–1.

140

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Ruspe’s sixth-century biographer tells us that the young man’s 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 Vandal–Roman 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 Hilderic’s Roman ancestors, the poet never mentions the king’s descent from the Hasding royal house. His king’s 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 Thrasamund’s 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 1982–1986’, pp. 7–41. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 imperial–Vandal treaties of 442 and 474, contracting and acknowledging Eudocia’s 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 336–7, pp. 267–8 mentioned above. Proc. BV 1.5.14, 1:333; Innes, ‘Land’, 66–74; and see also above Chapter 1.3–4.

142

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 484–96). 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. 1–3, pp. 13–23. Dracontius, Satisfactio, passim and Dracontius, Romulea 6, 4:1–5. F. M. Clover, ‘The Symbiosis of Romans and Vandals in Africa’, in E. Chrysos and A. Schwarcz (eds.), Das Reich und die Barbaren, ¨ Ver¨offentlichungen des Instituts f¨ur Osterreichische Geschichtsforschung 29 (Vienna, 1989), pp. 62–6. See also A. H. Merrills, ‘The Perils of Panegyric: The Lost Poem of Dracontius and its Consequences’, in Merrills, Vandals, Romans, and Berbers, pp. 145–62 M. Overbeck, Untersuchungen zum afrikanischen Senatsadel in der Sp¨atantike, Frankfurter althistorische Studien 7 (Kallm¨unz, 1973), pp. 23–40 and Warmington, North African Provinces, pp. 106–8.

143

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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 king’s 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 Huneric’s 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 Africa’s 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 l’Afrique (Paris, 1955), pp. 252–3; and Gil Egea, A´frica, pp. 276–9. 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. 475–6. 70 Vict. Vit. 2.8, p. 27. Vict. Vit. 1.47 and 3.34–7, pp. 20–1 and 89–91. 72 For these families, see Overbeck, Senatsadel, pp. 23–33. Wickham, Framing, p. 89.

144

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

The old ruling class Vandal king’s 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 Geiseric’s 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. 24–5, 30, 32–5, and 38–9; PLRE 1:745, s.nn. ‘Proculus 3–4’, and ibid., pp. 747–9, s.nn. ‘Proculus 11–12’, and see also ibid., p. 1147, stemma 30. Overbeck, Senatsadel, pp. 39–40; 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 1:271–2, s.nn. ‘Dracontius 3–4’.

145

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 333 337–8 340 351–2 321

Proconsul of Africa emissary of Geiseric

340 439/77 (after 457?)

magister privatae rei Africae Vicar of Africa advocatus

320–1 364–7 484/96

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. 36–7. He was probably not a career grammarian: Kaster, Guardians of Language, pp. 415–17, no. 235. Apart from the poet, the name is absent from PLRE 1–3 and PCBE 1. AL 201–5 and 248, pp. 150–3 and 186–8 (Felix); AL 214, 218 and 220, pp. 159–60 and 162–3 (Coronatus); CTh 6.21.1 (ad 425), p. 268; and Miles, ‘Anthologia Latina’, p. 309. However, Kaster, Guardians of Language, pp. 397–8, no. 204 is sceptical in the case of Coronatus.

146

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Thrasamund’s 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 Felix’s 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 Thrasamund’s 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 groom’s 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 Hilderic’s construction of a new audience hall at Anclae and, as discussed above (Chapter 1.3), lauded the king’s 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. 5–6, p. 287: ‘in quo concordant pietas, prudentia, mores, / virtus, forma, decus, animus cultusque virilis.’ AL 201–5, pp. 150–3. AL 205, p. 153; Schanz and Hosius, Geschichte der r¨omischen Literatur 4/2:71–2. Baehrens and Vollmer, Poetae latini minores, 5:237; Schanz and Hosius, Geschichte der r¨omischen Literatur 4/2:59 and 61. AL 382, p. 295. AL 336–7, pp. 267–8 (Eutychus); AL 293, p. 242 (eunuch); and AL 332, p. 265 (informer). AL 340, pp. 269–70 (Damira); AL 194, p. 145 (Anclae); and AL 206, p. 154 (praise). On the name Damira, which may be Germanic, see M. Sch¨onfeld, W¨orterbuch der altgermanischen Personen- und ¨ V¨olkernamen nach der Uberlieferung des klassischen Altertums bearbeitet (Heidelberg, 1911; repr. 1965), p. 70.

147

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 351–3) 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. 24–6. 88 Ibid., pp. 236–41, s.aa. 351–3. Ibid., pp. 226–7, s.a. 346. Ibid., pp. 332–5 and 342–5, s.aa. 399–400 and 404–5.

148

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

The old ruling class Mediterranean (ad 284–541) 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. One’s 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. 58–60. 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. 70–9 and A. Samuel, Greek and Roman Chronology: Calendars and Years in Classical Antiquity, Handbuch der Altertumswissenschaft 1/7 (Munich, 1972), here esp. pp. 245–8. See also H. Seyrig, ‘Antiquit´es syriennes’, Syria: revue d’art oriental et d’arch´eologie 39 (1962), pp. 42–4 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. 283–5 mandated the use of imperial regnal year, consular year, indiction, month, and day to date legal documents.

149

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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., 37a–h and probably ibid., 225 (Theoderic); and ibid., 38a–d (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 3:273–5, index 6E. ILCV 1721; M. Handley, Death, Society, and Culture: Inscriptions and Epitaphs from Gaul and Spain, ad 300–750, BAR International Series 1135 (Oxford, 2003), pp. 123–5. ILCV 1216 = CIL 12.2700 = E. F. Le Blant (ed.), Inscriptions chr´etiennes de la Gaule ant´erieures au VIIIe si`ecle, 2 vols. (Paris, 1856–65), 2:206–7, 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. 125–6; and see also ILCV 3:279–80, 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. 83–102.

150

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

The old ruling class Table 3.2. Consular and Imperial dating systems in late Roman Africa, c.407–54

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

Consular

Imperial

Total

1 1 – 1 2 – 1 6

5 – – – 1 1 – 7

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 d’Afrique (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 arch´eologique, 6th ser. 42 (1953), p. 178, no. 39 (ad 409); CIL 8.11127 (ad 427); Ennabli, Inscriptions chr´etiennes de Sainte-Monique, pp. 171–2, no. 46 (ad 439); AE (1967), 200 and 208–9, nos. 595 and 640 (both ad 452; note that the latter also uses an anno provinciae date); Revue arch´eologique, 5th ser. 20 (1924), p. 387, no. 58 (ad 454). Bagnall, et al., Consuls, pp. 58–60; see also Handley, Death, Society and Culture, p. 130. Clover, ‘Felix Karthago’, pp. 10–11.

151

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Sbe¨ıtla, 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 Geiseric’s 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 484–96), while the remaining epigraphic examples of the use of Vandal regnal dates were commissioned in or after the reign of Thrasamund (ad 496–523); 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 s´eances de l’Acad´emie des Inscriptions et Belles-Lettres (1914), p. 483; (3) F. B´ejaoui, ‘Les Vandales en Afrique: T´emoignages arch´eologiques. Les r´ecentes d´ecouvertes 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 arch´eologiques a` Ha¨ıdra, 2 vols., Collection de l’Ecole franc¸aise de Rome 18/1–2 (Rome, 1975–81), 1:281–3, no. 419; (6) B´ejaoui, ‘Vandales en Afrique’, p. 201; (7) B´ejaoui, ‘Vandales en Afrique’, pp. 200–1; (8) CIL 10516 = CIL 8.11528 = ILCV 388 = Courtois, Vandales, p. 379, no. 103 = Duval, Ha¨ıdra 1:273–7, 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. 384–8. Merlin and Monceaux, Comptes rendus de s´eances de l’Acad´emie 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 (1930–1), 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 arch´eologique, 6th ser. 48 (1956), p. 200, no. 125; (11) AE (1967), 200, no. 596.

152

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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

439–533

[]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. 503–26 July 508

Ge . . .

22 Feb. 510

Festa

17 Feb. 517

Fortunatiana

521

–

Geiseric (ad 439–77) or Thrasamund (ad 496–523) 2. Sbe¨ıtla, Byzacena natus anno XXVIII / regis Gesiric, pridie idus / septembres Gunthamund (ad 484–96) 3. El Ounaissia, Byzacena Thrasamund (ad 496–523) 4. T´ebessa

5.

Ha¨ıdra (class. Ammaedara)

6.

El Erg, Byzacena

7.

El Gousset, Byzacena

Hildiric (ad 523–30) 8. Ha¨ıdra 9.

Uppenna (mod. Hr. Chigarnia), Byzacena

Gelimer (ad 530–34) 10. Madauros

5

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

22 Jan. 531

Region is Africa Proconsularis, except where otherwise noted Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

...

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. A¨ın 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)

27 Feb. 463 or 520

Region is Africa Proconsularis, except where otherwise noted

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

The old ruling class at least, three ostraka from Bir Trouch, Algeria, use them interchangeably with Gunthamund’s 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 Justinian’s copper coins struck at Carthage bear an anno formula incorporating the emperor’s regnal year. Moreover, Justinian’s law of ad 537 mandating the use of consular dates also required the incorporation of the emperor’s 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 Justinian’s reign (Table 3.4, nos. 3–7). 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 F´evrier, ‘Ostraka de Bir Trouch’, pp. 239–42 and 244–5, nos. 1–2 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. 1–2; W. Hahn, Moneta Imperii Byzantini: Rekonstruktion des Pr¨ageaufbaues auf synoptisch-tabellarischer Grundlage, vol. 1, Von Anastasius I. bis Justinianus I. (491–565) einschliesslich der ostgotischen und vandalischen ¨ Pr¨agungen, Ver¨offentlichungen 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Gunthamund’s 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 Geiseric’s 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 Hilderic’s 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 Geiseric’s 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 Victor’s 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 Huneric’s regnal year, though again there was no love lost between this author and the Vandals.114 109

110

111 112 113 114

Bonnal and F´evrier, ‘Ostraka de Bir Trouch’, pp. 239–46, nos. 1–4 and probably no. 5; Tablettes Albertini, acts 1–4, 6–14, 16–24, 1:215, 217, 218, 223, 234, 238, 242, 246, 249, 253, 257, 259, 262, 269–71, 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 œuvre: e´tude critique (Algiers, 1954), p. 16. Passio septem monachorum 2, p. 108.

156

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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 Huneric’s council at Carthage (anno sexto regis Hunerici) does not correspond to the dates in Victor of Vita’s account, including those in Huneric’s law of 484 (quoted at Vict. Vit. 3.3–14, pp. 72–8), which place it in the king’s 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, ‘L’Odyss´ee du manuscrit de San Pietro qui renferme les œuvres 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. 233–46, but also E. A. Meyer, ‘Explaining the Epigraphic Habit in the Roman Empire: The Evidence of Epitaphs’, Journal of Roman Studies 80 (1990), pp. 74–96; Handley, Death, Society and Culture; N. Everett, Literacy in Lombard Italy, c. 568–774, Cambridge Studies in Medieval Life and Thought, 4th ser. 53 (Cambridge, 2003), pp. 235–76.

157

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 king’s regnal year. Setting aside for a moment the question of Vitalis presbyter’s epitaph, Gunthamund’s twelve-year reign produced at least one dated inscription; Thrasamund’s twenty-seven years, another two; Hilderic’s seven years, two; and Gelimer’s 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. Huneric’s reign, spanning the seven years from 477 to 484, has so far yielded no dated inscriptions – despite the fact that this same king’s 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 Huneric’s reign, and African elites had a brutal reminder of the transience of worldly things. The violence of Huneric’s persecution may have cut short a process of accommodation already under way in the later years of his father’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Christ’s 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. 88–90 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 l’uso del latino nel concilio cartaginese del 484’, in Oikoumene: Studi paleocristiani pubblicati in onore del Concilio Ecumenico Vaticano II (Catania, 1964), pp. 224–6; and Howe, Vandalen, pp. 156–82.

159

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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. 178–9. 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 Hen’s forthcoming study on the topic. Vict. Vit. 3.5, pp. 73–4. 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. 141–51 and D. H. Williams, Ambrose of Milan and the End of the Nicene-Arian Conflicts (Oxford, 1995; repr. 2002), pp. 11–37. On Vandal Arianism, however, see also Y. Mod´eran, ‘Une guerre de religion: les deux e´ glises d’Afrique a` l’´epoque vandale’, Antiquit´e tardive 11 (2003), 21–44, esp. 30–6. See in general R. P. C. Hanson, The Search for the Christian Doctrine of God: The Arian Controversy, 318–381 (Edinburgh, 1988). Socrates, Historia Ecclesiastica 6.8, PG 67:688–9; Sozomen, Historia Ecclesiastica 7.5–7 and 8.8, PG 67:1424–32 and 1536–7; 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, 489–554, Cambridge Studies in Medieval Life and Thought, 4th ser. 33 (Cambridge, 1997), pp. 246–7. 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. 486–7.

160

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Vita’s 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 historian’s 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. 5–11; Shanzer, ‘Intentions and Audiences’, pp. 272–8; Howe, Vandalen, pp. 61–119. Courtois, Victor de Vita, pp. 16–17 and Howe, Vandalen, pp. 38–60. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 kingdom’s desert fringes and Mediterranean islands. We have seen this already in the king’s 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. Huneric’s 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. 475–6 with Antoninus Honoratus, Epistola consolatoria ad Arcadium, PL 50:567–70; Vict. Vit. 3.41, p. 92 and perhaps ibid., 3.26–7, pp. 84–5; Passio septem monachorum 14, p. 113; Shanzer, ‘Intentions and Audiences’, pp. 282–4. Vict. Vit. 2.26–37, 3.15–20, 3.34–8, pp. 33–8, 78–81, and 89–91; 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 Carthage’s ordination in 480/1. Thrasamund: V. Fulg. 17–18, pp. 87–91 and Vict. Tonn. s.a. 497.78, p. 24. Geiseric: Vict. Vit. 1.51, p. 22. Exiles: Vict. Vit. 3.34 and 3.43–5, pp. 89 and 93–4. Recall: Vict. Tonn. s.a. 479.51–2, p. 16, followed by Isidore of Seville, Historia Vandalorum 80, ed. and trans. C. Rodr´ıguez Alonso,

162

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 king’s death and the accession of his successor Hilderic, though Fulgentius of Ruspe was briefly recalled to Carthage to answer Thrasamund’s 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 Mouza¨ıaville 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 Geiseric’s 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, v´andalos y suevos de Isidoro de Sevilla: Estudio, edici´on cr´ıtica y traducci´on, Fuentes y estudios de historia leonesa 13 (Le´on, 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. 20–1 and 25–7, pp. 99–107 and 121–7; Vict. Tonn. s.a. 523.106, p. 34; PCBE 1:507–13, s.n. ‘Fulgentius 1’, at pp. 510–11. Eugenius: Vict. Vit. 3.34 and 3.43–4, pp. 89 and 93–4 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 147–8. Vict. Vit. 1.22, pp. 10–11. Vict. Vit. 1.39–40, pp. 17–18. Under Huneric, exile was not always far from a bishop’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Ha¨ıdra (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 1.14–16, pp. 5–6 and 7–8; 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 l’Ecole franc¸aise de Rome (Paris, 1982), 2:675–7 and L. Ennabli, Inscriptions chr´etiennes de Sainte-Monique, pp. 12–16. 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. 42–3, 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 13:1219–30, 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, 1:7–10, no. 3 (Carthage); see also ibid., 2:691–2. The identification of the basilica of the Scillitans with that of the basilica Celerinae is accepted by L. Ennabli, Carthage: une m´etropole chr´etienne du IVe a` la fin du VIIe si`ecle (Paris, 1997), pp. 32–4. Proc. BV 1.21.17–25, 1:403–4, of one of the suburban churches of St Cyprian. Duval, Ha¨ıdra 1:87–8, no. 58, first announced in N. Duval, ‘Rapport pr´eliminaire sur les travaux effectu´es a` Ha¨ıdra en sept.-oct. 1967’, Africa 3–4 (1969–70), 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 1:271–3, s.n. ‘Deogratias 1’, at p. 271.

164

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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. 13–14 and 17, pp. 69–75 and 87–9. See also Isidore, Historia Vandalorum 81, pp. 302–4, 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. 6–7, pp. 35–45. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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. 1.35–8, pp. 15–17. Vict. Vit. 1.29, p. 13. For Vincent’s see, see Notitia, Proc. 41, p. 118. Vict. Vit. 1.43, pp. 18–19. Decree: Vict. Vit. 2.23, p. 32. Continued service: see below, this paragraph.

166

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 king’s demand, Victorianus was said to have been brutally punished and, Victor of Vita tells us, ‘accepted the crown of martyrdom’; but the ambiguity of Victor’s language makes it impossible to say with any real certainty whether the official was in fact put to death.157 Equally important to Geiseric’s 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 Geiseric’s reign, the procurator domus of prince Huneric’s 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 Ruspe’s 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. 475–6. Antoninus Honoratus, Epistola consolatoria, col. 567B–C: ‘Tecum est omnis chorus martyrum praecessorum tuorum; exspectant te martyres et defendunt, et coronam extendunt . . . Si mortuus fueris, certus esto, martyr eris.’ Vict. Vit. 1.19–21, pp. 9–10. See also Vict. Vit. 1.47, pp. 20–1, the case of Geiseric’s chief pantomime Mascula. Vict. Vit. 3.27, p. 85 and Shanzer, ‘Intentions and Audiences’, pp. 283–4, who nonetheless reads Victorianus as a martyr in the strong sense of the word. Vict. Vit. 1.43–6, pp. 19–20; the quotation is ibid., 1.44, p. 19. 160 Vict. Vit. 1.48–50, pp. 21–2. Vict. Vit. 2.10–11, pp. 27–8.

167

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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 Huneric’s 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 Africa’s Nicene population the laws through which the Roman emperors had sought to bring earlier heterodox communities to heel. Huneric’s 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 Huneric’s 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 1.8.9–10, 1:346–7. V. Fulg. 3, pp. 21–3. On the persecution of monks, see the Passio septem monachorum, pp. 108–14. Fulgentius of Ruspe, Abecedarium ll. 247–53, p. 884.

168

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 king’s cellarers, two doctors, two merchants, a bishop’s 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 Geiseric’s 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. 3.3–14, pp. 72–8; cf. esp. CTh 16.5.52 (ad 412), pp. 872–3, 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. 867–70, 873–4, 878–9, and 881–2, and Constitutio Sirmondiana 12 (ad 407), in CTh, pp. 916–17; Fournier, ‘Victor of Vita’, p. 75 n. 3. See also Vict. Tonn. s.a. 479.50, p. 16 (‘per totam Africam’) and Proc. BV 1.8.3–4, 1:345, who first mentions specifically religious persecution under Huneric. Proconsul of Carthage: Vict. Vit. 3.27, p. 85; secular aristocrats: ibid., 3.25–6, pp. 83–5; cellarer’s wife: ibid., 3.33, pp. 88–9; doctors: ibid., 3.24 and 3.50–1, pp. 83 and 96–7; merchants: ibid., 3.41, p. 92; bishop’s daughter: ibid., 3.24, p. 83; bishops: ibid., 3.34 and 3.43–6, pp. 89 and 93–5; Carthaginian clergy, esp. the deacon Muritta: ibid., 3.34–8, pp. 89–91; monks: ibid., 3.41, p. 92 and Passio septem monachorum, pp. 108–14. See also Vict. Vit. 3.49, p. 96, the forced rebaptism of an aristocratic boy. Thuburbo Maius: Vict. Vit. 3.25, pp. 83–4. Culusi: Vict. Vit. 3.26, pp. 84–5. Thambaia: Vict. Vit. 3.28, p. 86. For the locations, see Courtois, Victor de Vita, pp. 46–50, with map on p. 45. Passio septem monachorum 8, p. 110.

169

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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. Hilderic’s 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. Thrasamund’s 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 Vita’s 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 1:159–61, s.n. ‘Bonfatius 26’. Hays, ‘Romuleis Libicisque Litteris’, pp. 131–2 and Miles, ‘Anthologia Latina’, passim.

170

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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 Vita’s 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. 194–5 (for the date, see ibid., 628b, p. 196); Clover, ‘Carthage and the Vandals’, p. 4. Quodvultdeus, Liber promissionum et praedictorum Dei 4.8.15–16, ed. R. Braun in CCSL 60:200– 1; see also ibid., 5.7, pp. 194–5. On Quodvultdeus, see R. Gonz´alez Salinero, Poder y conflicto religioso en el norte de A´frica: Quodvultdeus de Cartago y los v´andalos, Graeco-Romanae Religionis Electa Collectio (GREC) 10 (Madrid, 2002). Quodvultdeus: De cataclysmo 3.12–24 and 5.3–9, ed. R. Braun in CCSL 60:411–12 and 414–15 (Pharaoh) and De tempore barbarico 1.3.18, ed. Braun in CCSL 60:427 (Nebuchadnezzar). Pseudo-Fulgentius, Sermones, PL 65:855–954, esp. Sermo 8, col. 868 (Nebuchadnezzar) and Sermones 13–15 and 78, cols. 874–5, 877–8, and 950 (Pharaoh). For the identification as African of these and all other anonymous sermons discussed here, see Dossey, ‘Christians and Romans’, pp. 366–78. 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. Barr´e, PLS 4, col. 1911.

171

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Huneric’s achievement in his reign of terror of 484. This was not just an abstract concern. Elpidoforus, one of Huneric’s 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 Mod´eran has convincingly argued that nearly 20 per cent of the African episcopate converted to the Arian confession in the persecution of 484.186 Thrasamund’s 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 535–6).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. 259–66. Quodvultdeus, Liber promissionum 4.5.7, p. 194. Vict. Vit. 3.34–7, pp. 89–91; 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 2.47–50, pp. 12, 30–1, and 42–4. 184 Fulgentius of Ruspe, Contra sermonem Fastidiosi 1.1, pp. 283–4. Vict. Vit. 3.39, pp. 91–2. Gregory of Tours, Historiae 2.3, p. 45. 187 AL 248, pp. 186–8. Mod´eran, ‘Notitia provinciarum’, pp. 172–81. Collectio Avellana 85–7, ed. O. G¨unther, in Epistulae imperatorum, pontificum, aliorum inde ab a. ccclxvii usque ad a. dliii datae Avellana quae dicitur collectio, CSEL 35/1 (Vienna, 1895), pp. 328–33; Dossey, ‘Last Days of Vandal Africa’, pp. 111–12.

172

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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. 1–145, here De sollemnitatibus, pp. 47–92; see also the Arian Contra Iudaeos and Contra paganos, ibid., pp. 93–117 and 118–40, 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. 147–53. 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. 2.38–3.15, pp. 38–79. Collatio beati Augustini cum Pascentio ariano, cols. 1156–62; Contra Felicianum Arianum de unitate Trinitatis, PL 62:333–52; Vigilius of Thapsa, Contra Arianos, Sabellianos, Photinianos dialogus, PL 62:179–238, 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 39:2198–200. Quodvultdeus, Contra Iudaeos, paganos et Arrianos, ed. R. Braun in CCSL 60:225–58 and Quodvultdeus, Adversus quinque haereses, pp. 261–301.

173

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 bishop’s 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 king’s 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 90:1–134 and pseudoVigilius, Solutiones obiectionum Arianorum, PL 62:469–72; pseudo-Augustine, Solutiones diuersarum quaestionum ab haereticis obiectarum, ed. Schwank, CCSL 90:141–223; also the anonymous Testimonia de Patre et Filio et Spiritu Sancto, ed. D. de Bruyne, CCSL 90:227–33 and Liber de Trinitate, ed. J. Fraipont, CCSL 90:239–60. Fulgentius of Ruspe, Abecedarium and Shanzer, ‘Intentions and Audiences’, pp. 285–6. Fulgentius, Ep. 8, 1:257–73. Fulgentius of Ruspe, Liber de Trinitate ad Felicem 1.1, p. 633. Fulgentius, Ep. 9, 1:277–80; Fastidiosus, Sermo, ed. J. Fraipont in CCSL 91:280–3; and Fulgentius of Ruspe, Contra sermonem Fastidiosi, pp. 283–308. Gennadius, De viris inlustribus, 74, 78–9 and 98, pp. 87–8 and 95–6; see also ibid., 97, p. 95 (Cerealis). Fulgentius of Ruspe, Dicta regis Trasamundi et contra ea responsionum liber unus, ed. J. Fraipont in CCSL 91:65–94 and Fulgentius, Ad Trasamundum regem libri iii, ed. Fraipont in CCSL 91:95–185.

174

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 579–c.600), had written against the Arians.206 So had Severus of M´alaga, 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 Hermenegild’s 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. 2.56–101, pp. 46–71, 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. 105–6); Lucifer of Cagliari (CPL no. 112–13); and Faustinus presbyter (CPL no. 120); also Marius Victorinus, an African (CPL nos. 95–6) and Pelagius, a Briton or perhaps an Irishman (CPL no. 748a). Gaul: Hilary of Poitiers (CPL nos. 434–5 and 462) and Phoebadius of Agen (CPL no. 473). Africa: Augustine (CPL no. 699–700 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. Codo˜ner Merino, in El ’De viris illustribus‘ de Isidoro de Sevilla: Estudio y edici´on cr´ıtica, Theses et Studia Philologica Salmanticensia 12 (Salamanca, 1964), pp. 149–50. Isidore, De viris inlustribus 30, ed. Codo˜ner 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: edici´on cr´ıtica y estudio hist´orico (Madrid, 1948), pp. 97–124. Gregory I, Dialogi 3.31, 2:384–90; 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. Codo˜ner Merino, pp. 149–50, and Greg. Ep. 9.229, 2:805–11. See further Gregory of Tours, Historiae 5.38, p. 244, who indicates that Hermenegild was converted by his Frankish wife Ingundis.

175

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

Staying Roman in a similar context. Bishop Avitus of Vienne (ad 490–518) 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 father’s lifetime.211 Earlier, another bishop in southern Gaul, Faustus of Riez (ad 461–c.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 502–43) 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 Ruspe’s biographer dismisses Thrasamund’s 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 Christian’s 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. 1–15. Gregory of Tours, Historiae 2.34, pp. 81–2. 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. 473–86. V. Fulg. 20, p. 99. 216 Ibid., 2.14.2–6, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

The old ruling class a new temporal king, Quodvultdeus’ words made an unmistakable and powerful statement that the Nicene Christian’s true loyalties were not to this world, but to God’s 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 Habetdeum’s emperor to be God; but it is likely that there is an element of double entendre in Victor of Vita’s 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 Huneric’s 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 Huneric’s 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. Theoderic’s wife and the couple’s 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 Huneric’s brothers, was also exiled together with 218 219 220 221 222

Vict. Vit. 3.46, pp. 94–5: ‘lectitanda imperatori meo transmisi.’ Vict. Vit. 2.12–13, p. 28; Merrills and Miles, Vandals, p. 75. Vict. Vit. 2.12, p. 28; see also ibid., 3.19–20, pp. 80–1. Vict. Vit. 2.13, p. 28; Proc. BV 1.7.29, 1:344–5; Jordanes, Getica 33.169, p. 72. Merrills and Miles, Vandals, pp. 75–6.

177

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

Staying Roman his wife.223 The king furthermore ordered the execution of the retired official Heldica – Geiseric’s 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 Victor’s 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 Hilderic’s succession were banished to Corsica and condemned to hard labour, on the grounds that they did not want the king’s 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. 5:34–7).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 Huneric’s house and the succession of his son. From the king’s 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. Victor’s 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. 2.12–14, pp. 28–9. For the difficulties to Hilderic’s immediate succession posed by Theoderic and Genton’s male heirs, see Courtois, Vandales, pp. 399–404 and ibid., p. 390, ‘Tableau g´en´ealogique’. 225 Vict. Vit. 2.13 and 2.16, pp. 28 and 30. Vict. Vit. 2.15–16, pp. 29–30. Vict. Vit. 1.44 and 2.13, pp. 19 and 28. Vict. Vit. 3.17–20, pp. 79–81; Mod´eran, ‘Notitia provinciarum’, p. 181. See, however, Merrills, ‘Perils of Panegyric’, pp. 154–5 and Merrills and Miles, Vandals, p. 75; see also Howe, Vandalen, pp. 276–8. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 bishop’s 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 Huneric’s 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. 866–7; 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. 952–3: ‘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. 1.41–2, p. 18. Dossey, ‘Last Days of Vandal Africa’, pp. 109–10. Vict. Vit. 3.48, pp. 95–6; see also ibid., 3.46–7, pp. 94–5. 238 V. Fulg. 6–7, pp. 35–43. Vict. Vit. 3.47, p. 95.

179

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Africa’s 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 Victor’s 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, Huneric’s 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, Victor’s 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 king’s 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 Vita’s 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. 156–82 and 302–18. Vict. Vit. 3.38, p. 91; Howe, Vandalen, pp. 168–76.

180

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Huneric’s 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 father’s 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 king’s 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 Huneric’s 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 Huneric’s 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. 2.8–9, p. 27; see also ibid., 2.11, p. 28. Vict. Vit. 2.12, p. 28: ‘sese iamdudum omnibus lenem ostenderat.’ For important limitations on Victor’s definition of omnes, see also Vict. Vit. 2.1–2, 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 1.5.11–14, 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. Mod´eran, ‘L’Etablissement territorial des Vandales en Afrique’, Antiquit´e tardive 10 (2002), pp. 107–10. 248 Vict. Vit. 3.3–14, pp. 72–8. Vict. Vit. 2.40, pp. 39–40.

181

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 region’s 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. 872–3. 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. 268–76, who sees this ecumenism as new to Huneric’s 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 87:47–92; see also above, n. 189) and Fastidiosus, Sermo, ed. J. Fraipont in CCSL 91:280–3. Vict. Vit. 3.1, p. 72 and Shanzer, ‘Intentions and Audiences’, p. 285. Fastidiosus, Sermo cc. 3 and 5, pp. 281–2; see also Fulg. Ep. 9.4, 1:279. My thanks to Robin Whelan for kindly alerting me to these passages.

182

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 king’s understanding of his people’s providential history and his sacral conception of his own rulership. Huneric’s 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 Geiseric’s 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 God’s 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, A´frica, p. 317. 257 Jordanes, Getica 33.169, p. 72. Dracontius, Satisfactio 41 and 107–8, 2:178 and 181. Proc. BV 1.5.24–5, 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, 1:115–16; Proc. BV 1.3.14– 36 and 1.6.1–26, 1:320–4 and 1:335–40; Jordanes, Getica 33.167 and 33.172, pp. 72 and 74, and see also ibid., 36.184 and 47.244, pp. 78–9 and 100; Jordanes, Romana 330, p. 42; John Malalas, Chronographia, ed. L. Dindorf, Corpus Scriptorum Historiae Byzantinae 15 (Bonn, 1831), pp. 372–3; John of Antioch, frag. 196, ed. K. M¨uller, in Fragmenta historicorum Graecorum, 5 vols. (Paris, 1841–70), 4:613; and Theophanes, Chronographia AM 5931, 1:93–5 (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. 41–55 and 111–24.

183

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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, Huneric’s 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 Africa’s 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. 2.23–4, pp. 32–3. 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 7.21.89–7.22.100, pp. 494–524 (morality), and in general Gil Egea, A´frica, pp. 315–18; see also Lambert, ‘Barbarians in Salvian’s De Gubernatione Dei’, pp. 109–12. Vict. Vit. 2.1–2, p. 24; Liber genealogus 616 and 618, pp. 194–5; 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 87:93–117 and 118–41, respectively (and see above, n. 189).

184

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Geiseric’s 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. 88–90. 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 1.5.24–5, 1:335. Isidore, Historia Vandalorum 74, p. 294.

185

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 work’s modern editor for the manuscript’s 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 Augustine’s 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. 104–5. ´ V. V¨aa¨ n¨anen, Etude sur le texte et la langue des Tablettes Albertini, Suomalaisen Tiedeakatemian toimituksia, ser. B, 141/2 (Helsinki, 1965), pp. 48–9, s.nn. gemio and maforsenum; J. N. Adams, Bilingualism and the Latin Language (Cambridge, 2003), pp. 454–5; and J. N. Adams, The Regional Diversification of Latin 200 bc–ad 600 (Cambridge, 2007), pp. 556–8 and 561. On the distinctive features of the Latin of the Albertini Tablets in general, see ibid., pp. 549–62. Countryside: Augustine, Epp. 66.2 and 108.5.14, CSEL 34:236 and 628; ibid., 209.2–3, 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. 161–2; see also Augustine, De haeresibus 87, ed. R. Vander Plaetse and C. Beukers, CCSL 46 (Turnhout, 1969), pp. 339–40. Hippo: Augustine, Sermones 113.2.2 and 167.3.4, PL 38:648 and 910. The classic study is W. Green, ‘Augustine’s 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. 179–90; see also C. Krahmalkov,

187

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 14–15, 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), 126–34; D. J. Mattingly, Tripolitania (London, 1995), pp. 162–6; C. Lepelley, ‘T´emoignages de saint Augustin sur l’ampleur et les limites de l’usage de la langue punique dans l’Afrique de son temps’, in C. Briand-Ponsart (ed.), Identit´es et culture dans l’Alg´erie antique, Publications de l’Universit´e de Rouen 377 (Rouen, 2005), pp. 127–53; and in general Bowersock, Brown, and Grabar, Late Antiquity, pp. 656–7, s.v. ‘Punic’; but also P. Brown, ‘Christianity and Local Culture in Late Roman Africa’, JRS 58 (1968), 85–95 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 (T¨ubingen, 2005), pp. 105– 6. The most important collections include R. G. Goodchild, ‘La necropoli Romano-Libica di Bir ed-Dr´eder’, Quaderni di Archeologia della Libia 3 (1954), 91–107; F. Vattioni, ‘Glosse Puniche’, Augustinianum 16 (1976), 536–55; and the material cited in Reynolds and WardPerkins, Inscriptions of Roman Tripolitania, pp. 10–13. See also R. Marichal, Les Ostraca de Bu Njem, Suppl´ements 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 18.22–3, ed. L. Ian and C. Mayhoff, 6 vols. (Stuttgart, 1892–1909; repr. 1967–70), 3:147–8. Devotional literature: Augustine, Enarratio in Psalmum 118.32.8, 40:1776 with Lepelley, ‘T´emoignages’, p. 132. Augustine, Epp. 16.2 and 17.2, CSEL 34:37–8 and 41–2; 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, 1955–91), 2/1:109. Augustine, Ep. 17.2, CSEL 34:41–2. Thanks to Brian Stock for this reference.

188

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Augustine’s 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 Justinian’s 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, 1120–1, 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, 1955–65), 1:148; Epitome de Caesaribus 20.8, ed. F. Pichlmayr in Sexti Aurelii Victoris Liber de Caesaribus (Leipzig, 1966), pp. 155–6; and Statius, Silvae 4.5.45–6, ed. A. Marastoni (Leipzig, 1961), p. 92. R. Bartoccini, ‘Scavi e rinvenimenti in Tripolitania negli anni 1926–1927’, Africa Italiana 2 (1928), 187–200 with Jongeling and Kerr, Late Punic Epigraphy, pp. 71–74. 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, 1855–80), 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 3.146–7, 3.322–3, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 514–23) 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 Vita’s 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 Paul’s 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 empire’s 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 (Paul’s 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 39:1683–4: ‘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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 individual’s 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 Victor’s 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 Victor’s 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 Victor’s 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 Victor’s 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. 226–8 and 234–7. Augustine, De civitate Dei, 14.1, 48:414. On this point, see also Howe, Vandalen, pp. 328–33.

192

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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. Victor’s 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 Vita’s 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. 17–22. Shanzer, ‘Intentions and Audiences’, pp. 279–90 and Howe, Vandalen, pp. 334–56; see also Costanza, ‘Uuandali-Arriani’, p. 237.

193

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Vandal–African 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 Victor’s 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 Victor’s 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 Huneric’s reign, the Arianizing push to which that king had been so committed stalled and died under his son Hilderic. It is not unreasonable 194

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 Thrasamund’s 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 Thrasamund’s reign in which the manuscript was produced. Fulgentius and his circle, then, would seem to have accepted Thrasamund’s 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 king’s conversion to the Nicene faith. To Fulgentius, Thrasamund’s 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 God’s eternal law or against it. The answer to that question, of course, depended on one’s 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 Victor’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:09, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.004

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 533–4, nearly a century of maturing Vandal–African 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 province’s 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. Justinian’s 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: R¯um (Romans or Byzantines), Af¯ariq or Africans, and Berbers or Moors. Ibn c Abd al-Hakam, Fut¯uh Ifr¯ıqiya w’al-Andalus, ed. and trans. A. Gateau, in Conquˆete de l’Afrique du nord et de l’Espagne (2nd edn; Algiers, 1947).

196

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans the Vandal yoke; but it also points out that, in facilitating the territory’s ‘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 province’s 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: Africa’s 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 region’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 strat¯egos Liby¯es.5 The difficulties of administering such a far-flung province also ensured that even in Justinian’s 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

6

CJ 1.27.1 (ad 534), pp. 77–9. The most recent full-length synthesis of the history of Byzantine Africa is still C. Diehl, L’Afrique byzantine: histoire de la domination byzantine en Afrique (533–709) (Paris, 1896), here pp. 97–137 and 483–502 (civil and military organization of the province); on Byzantine Africa in general, see also Averil Cameron, ‘Gelimer’s 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. 171–90. CJ 1.27.2 (ad 534), pp. 79–81. 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 Hi´erarchie militaire en Afrique byzantine’, Antiquit´e tardive 10 (2002), 169–75. 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. 29–62.

198

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Africa’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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, Cameron’s 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 elite’s 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 565–78) 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 region’s 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. Mod´eran, Les Maures et l’Afrique romaine (IVe –VIIe si`ecle), Biblioth`eque des e´ coles franc¸aises d’Ath`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. 11–12 = Just. Nov. 149.1, p. 724. A. H. M. Jones, The Later Roman Empire, 284–602: 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. 800–1 and Jones, Later Roman Empire, 1:389. T. S. Brown, Gentlemen and Officers: Imperial Administration and Aristocratic Power in Byzantine Italy ad 554–800 (London, 1984), pp. 64–5.

200

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans Italians: Fidelis (ad 537–8) was a native of Milan; Reparatus (ad 538–9), 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 610–41), at least fourteen seem themselves to have been Egyptians.13 After Justinian’s 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 2.12.27–8, 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 2:469–70, s.n. ‘Fidelis’ and ibid., pp. 939–40, 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 3:96–8, s.n. ‘Apion 3’; ibid., pp. 118–19, s.n. ‘Aristomachus 2’; ibid., pp. 262–3, s.n. ‘Callinicus 4’; ibid., p. 423, s.n. ‘Dorotheus 7’; ibid., pp. 437–8, s.n. ‘Elias 3’; ibid., pp. 582–3, 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. 1121–2, s.n. ‘Senuthius 1’; and ibid., pp. 1372–3, s.n. ‘Victor 4’. See also ibid., pp. 408–9, s.n. ‘Domentianus’; ibid., pp. 733–4, 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, 300–700 (Cambridge, 2007), pp. 245–9; 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 l’Egypte byzantine, Biblioth`eque de l’Ecole des hautes e´ tudes, Sciences historiques et philologiques 201 (Paris, 1912), here pp. 72–9; on the governors’ regional origins, see ibid., p. 83. Service in native province: PLRE 3:664, s.n. ‘Ioannes 59’, and ibid., pp. 1372–3, s.n. ‘Victor 4’; see also PLRE 3:105, s.n. ‘Archelaus 2’, and ibid., pp. 733–4, s.n. ‘Iulianus 12’. Non-Egyptians: see PLRE 2:677–81, s.n. ‘Liberius 3’, and PLRE 3:928–30, s.n. ‘Narses 2’; ibid., pp. 940–3, s.n. ‘Nicetas 7’; ibid., pp. 1085–6, s.n. ‘Rhodon’; and ibid., p. 1418, s.n. ‘Zeno 1’. See also perhaps PLRE 3:750–4, s.n. ‘Iustinus 4’.

201

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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

533–4

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) 534–6; (2) 539–44 536–9 544–5 545 545–6 546–551/2 563–564/5 570 571 (1) 577–85 (2) 591–8 602?–10 645–7

‘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 2–3 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 Tr¯ogilos (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. Jireˇcek identified this Germania with what is now the territory of western Bulgaria: ‘Arch¨aologische Fragmente aus Bulgarien’, Archaeologisch-epigraphische Mittheilungen aus Oesterreich-Ungarn 10 (1886), pp. 71–2. The term Germania typically referred to the Rhineland provinces: see, e.g., OCD, pp. 633–4, s.v. ‘Germania’. PLRE 3:644, s.n. ‘Ioannes qui et Troglita 36’. Tr¯ogilos was also the name of a district in Sicily.

202

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 empire’s militarily active frontiers, and in 600 we hear of an African holding the province’s 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 2.3.12–13, 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 1.10.13–16, 1:47–8 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. 109–10, 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 3:546–7, s.nn. ‘Gregoria 1–3’, and ibid., pp. 549–54, s.nn. ‘Gregorius 4, 7, 10–13, 17–18 and 20’. Greek seals: ibid., pp. 552–5, s.nn. ‘Gregorius 8–9 and 21–6’. Italy: ibid., pp. 547–53, s.nn. ‘Gregorius 1, 5–6, and 15’. Gaul: ibid., pp. 548–9, s.n. ‘Gregorius 3’. The Vita S. Gregorii Agrigentini, PG 98:549–716, 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 2.11.2–38, 1:198–203 and John of Ephesus, Historiae ecclesiasticae pars tertia 6.6, trans. E. W. Brooks, CSCO (Scr. Syr.) ser. 3, 3 (Louvain, 1936), pp. 221–2 with ODB 1:127, s.n. ‘Apameia’.

203

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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

Dates

Regional origins

Archelaus Solomon

534 (1) 534–6 (2) 539–44 536–9 544–5 545–c.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.640 641–2 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 2–3 fastes.

pontiff refers to the prefect as Augustine of Hippo’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Justinian’s 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, 1:529–30. 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. 66–7 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 3:107–9, 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 15–16 (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 2:133–4, s.n. ‘Archelaus 5’. The fact is strongly suggestive of the great expectations Justinian entertained for a reconquered Africa. PLRE 3:142–4, s.n. ‘Athanasius 1’.

205

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman tells us that, with the exception of Solomon and the Hun A¨ıgan, 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 Za¨ıdus and Sarapis were among the non-Thracians as well: Za¨ıdus is perhaps the Arabic name Sac¯ıd 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 3:1414–15, s.n. ‘Za¨ıdus’ and ibid., p. 1114, s.n. ‘Sarapis’; I. Shahˆıd, Byzantium and the Arabs in the Sixth Century, 2 vols. (Washington, DC, 1995–2009), 1/1:180 is more cautious, but see also ibid., pp. 181–2. Proc. BV 1.11.14, 1:362. Proc. BV 1.11.11 and 2.4.29, 1:362 and 1:436 (Pharas); Proc. BG 4.19.6–7, 2:585–6 (Sinnion). Proc. BV 2.10.3, 1:459. Proc. BV 1.11.5, 1:361; PLRE 3:1167–77, s.n. ‘Solomon 1’, at p. 1169. Proc. BV 2.15.51, 1:495; PLRE 3:814, s.n. ‘Marcellus 2’. Proc. BV 2.5.2–4 and 2.15.50, 1:439 and 1:495; PLRE 3:371–2, s.n. ‘Cyrillus 2’.

206

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 A¨ıgan (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. 102–4. Proc. BV 2.13.2, 1:475 (Althias), ibid., 2.14.40, 1:488 (Valerian), and ibid., 2.15.50–1, 1:495 (Marcellus, Cyril, Barbatus, Sarapis, and Terentius). Jordanes, Romana 369, p. 48; but see also Proc. BV 2.15.50–1, 1:495. Proc. BV 2.14.37–40, 1:487–8. Proc. BV 2.5.5, 1:439; on his career in general, see PLRE 3:644–9, 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 2.10.3–11, 1:459–60. Proc. BV 2.5.6, 1:439–40, and PLRE 3:635, s.n. ‘Ioannes 12’.

207

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman Moors of that province began under the chieftain Antalas, and a certain Peter (another Thracian who had been one of Solomon’s 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 empire’s 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 Troglita’s 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 Troglita’s army in the 540s. The Gregory who in the winter of 546 or 547 fought under John’s 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., 2.28.24–33, 1:548–50, where he assists Artabanes in dispatching Guntharis’ bodyguards. Himerius: ibid., 2.23.3–17, 1:525–7; Coripp. Ioh. 4.8–65, pp. 66–8. Contrast the purge of Lycians from the eastern imperial administration in 392–5 discussed by C. Kelly, Ruling the Later Roman Empire, Revealing Antiquity 15 (Cambridge, Mass., 2004), pp. 48–9. Proc. BV 2.23.6–9 and 2.23.17, 1:526–7. 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 2:872–3, s.n. ‘Pharesmanes 3’). Coripp. Ioh. 5.248–51, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Troglita’s 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 Ners¯eh 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 Troglita’s 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 3:547–8, s.n. ‘Gregorius 2’. Artabanes’ Armenian cousin: Proc. BV 2.27.10–19, 2.28.7–10, and 2.28.14–16, 1:541–3 and 1:545–8. The military commander under John Troglita: Coripp. Ioh. 4.487–8, 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. Garso¨ıan, ‘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. 61–6. Proc. BP 1.19.37, 1:106; PLRE 3:928–30, 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 hi´erarchie militaire’, pp. 174–5. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Theodore’s 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 Troglita’s wars against the Moors. One was a certain John, John Troglita’s 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 Troglita’s 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 John’s commanders of the field army, see PLRE 3:547–8, 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 3.18–19, pp. 146–50. 65 Proc. BV 2.18.3, 1:505–6. 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 2.14.39–41, 1:488) but need not have been there continuously from 533: see PLRE 3:1060–6, s.n. ‘Procopius 2’, at p. 1061. Proc. BV 2.5.10, 2.21.3, and 2.21.13–15, 1:440 and 1:518–19; native of Tripolitania: ibid., 1.10.22, 1:359. John: Coripp. Ioh. 7.242–61, pp. 153–4. Liberatus Caecilides: Coripp. Ioh. 3.47–51, pp. 48–9. 70 Coripp. Ioh. 3.41–51, pp. 48–9. Maspero, Organisation militaire, p. 95.

210

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 2:1042–3, s.n. ‘John’ and ODB 3:1604, 1636, 1953, and 2039, s.nn. ‘Paul’, ‘Peter’, ‘Stephen’, and ‘Theodore’. On the last two, see PLRE 2–3 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 650–950’, in Ahrweiler and Laiou, Internal Diaspora, pp. 20–1. 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), 210–11, no. 703. On these names, see PLRE 2–3 s.nn. PCBE 1 lists seven Theodores (pp. 1107–9), seven Leontii and one Leontia (pp. 632–4), nine Peters (pp. 870–4), and seventeen Pauls (pp. 839–48) for the pre-Byzantine period. On names in Byzantine Italy, see Brown, Gentlemen and Officers, p. 68.

211

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Ha¨ıdra 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 641–68) 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, 1:251–2. Martyr: Duval, Loca sanctorum, 1:121–3 no. 56 (Ha¨ıdra) and 2:665–6; see also C. Courtois, ‘Sur un carreau de terre cuite repr´esentant saint Pantal´eon’, Karthago 3 (1951–2), pp. 209–13. The quotation is from Y. Duval, Loca sanctorum ´ franc¸aise Africae: le culte des martyrs en Afrique du IVe au VIIe si`ecle, 2 vols., Collection de l’Ecole 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 3:723–6, s.nn. ‘Isidorus 1–13’. C. Morrisson and W. Seibt, ‘Sceaux de commerciaires byzantins du VIIe si`ecle trouv´es a` Carthage’, Revue numismatique, 6th ser. 24 (1982), p. 237, nos. 19–20; Pringle, Defence, p. 334, no. 48 = ILCV 234a. On the popularity of the name in Italy: PLRE 3:854–5, s.n. ‘Mauricius 2’, and ibid., pp. 861–2, s.nn. ‘Mauricius 8–9’ (3 of 10 examples), with PBE 1, s.nn. ‘Maurikios 1–4 and 7–8’ (6 of 8 seventh-century examples); see also ibid., s.n. ‘Maurikios 6’ (Sicily).

212

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 individual’s 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, 230–31, and 233–4, nos. 5–6, 10–11, and 16. Fifth and early sixth centuries: PLRE 2:994– 5, s.nn. ‘Sergius 1–9’. Later sixth and seventh century, eastern connotations: PLRE 3:1123–35, s.nn. ‘Sergius 1–4, 6–12, 15, and 38–43’; see also Greek seals and weights: ibid., pp. 1131–7, s.nn. ‘Sergius 18–23, 25, 28–37, and 45–54’. Italy: ibid., pp. 1128, 1130–1, and 1137, s.nn. ‘Sergius 5, 16 and 55’; see also Latin seals: ibid., pp. 1131–2, 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 3:236–7, s.n. ‘Boethius 1’ (prefect). On the name Boethius, see PLRE 2–3, 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 1:712–14, s.nn. ‘Pompeianus 1–11’; also the proconsul of Africa ad 400–1: PLRE 2:897–8, 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 2:1067–8. However, see also A. D. E. Cameron, ‘The House of Anastasius’, Greek, Roman, and Byzantine Studies 19 (1978), pp. 259–76, esp. 259–63 and stemma p. 274 on the name Pompeius in the Anastasian house. Greg. Ep. 1.74, 1:82–3. 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 1–5’; see also Kajanto, Onomastic Studies, pp. 65–7. 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, 489–554, Cambridge Studies in

213

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 province’s 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, Justinian’s 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. 87–91, 97–102, and 263–72; for Vandal Africa, see above, Chapter 1.3. Proc. BV 1.10.14, 1:357–8.

214

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans Table 4.3. Constantinople to Africa: late ancient itineraries Date Itinerary

Nautical miles Comments

533 Constantinople–Heraclea 1,400 (class. Perinthus, mod. Marmaraere˘glisi)–Abydus (mod. C ¸ anakkale)–Sigeum (mod. Yenis¸ehir)–Malea–Caenopolis (class. Taenarum, mod. Cape Matapan)–Methone–Zacynthus– Sicily near Mt. Etna–Caucana (near Santa Croce Camerina, Sicily)–Gaulus (mod. Gozo)– Malta–Caput Vada 534 Lesbos–Peleponnese–Africa (near 1,200 Numidia and Mauretania?) 546 Constantinople–Abydus–Sigeum– 1,300 Aegean–Adriatic–Sicily– Caucana–Caput Vada

Belisarius/Procopius: Delays at Heraclea (5 days), Abydus (4 days), Methone (unspecified length); strong winds Abydus–Sigeum and Malta– Caput Vada; gentle winds Sigeum–Malea and Zacynthus–Sicily (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., 1.12.1–1.14.17, 1:365–76. 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. 123–61. 90 Proc. BV 1.14.3 and 1.14.8–10, 1:373–4. McCormick, Origins, pp. 482 and 491. Coripp. Ioh. 1.159–371, pp. 9–18; 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, 649–1571 (Cambridge, 1988), pp. 12–101;

215

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 commander’s 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. 133–43; and M. McCormick, Origins of the European Economy: Communications and Commerce, ad 300–900 (Cambridge, 2001), pp. 92–8. M. McCormick, ‘Bateaux de vie, bateaux de mort: maladie, commerce, transports annonaires et le passage e´ conomique du Bas-empire au moyen aˆ ge’, in Morfologie sociali e culturali in Europa fra tarda antichit`a e alto Medioevo, Settimane di studio del Centro italiano di studi sull’alto Medioevo 45 (Spoleto, 1998), pp. 77–8; see also Pryor, Geography, pp. 94–5. Speed of travel: McCormick, ‘Bateaux de vie’, p. 102. Comparable voyages: McCormick, Origins, pp. 481–500. 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 44–5; 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 1:549–51; 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. 4.222–35, pp. 74–5; Agathias, Historiae 1.prooem.25, p. 8. Proc. BV 2.18.1–18, 1:505–8. Proc. BV 2.22.1–6, 1:522–3. Eventually, John son of Sisiniolus moved against the Moors: ibid., 2.23.1–32, 1:525–9.

216

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 empire’s 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. 495–6; Theophanes, Chronographia AM 6055, 1:238–9. On his office, see PLRE 3:670, s.n. ‘Ioannes qui et Rogathinus 75’.

217

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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

Dates

Previous career

Belisarius

533–4

Solomon

(1) 534–6; (2) 539–44

Germanus

536–9

Sergius Areobindus Artabanes

544–5 545 545–6

John Troglita

546–551/2

Marcian Theoctistus

563–564/5 570

Amabilis Gennadius Heraclius

571 (1) 577–85; (2) 591–8 602?-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

637 645–7

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 Justinian’s 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 empire’s eastern borders, rather than in Thrace or in Italy. Early in Justinian’s reign, from 527 to 529, 98 99

See also Jones, Later Roman Empire, 1:386. Proc. BG 3.40.5–6, 2:476; PLRE 2:505–7, s.n. ‘Germanus 4’. On Thrace in the sixth century, ¨ see P. Soustal, Thrakien (Thrak¯e, Rodop¯e und Haimimontos), Osterreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, Denkschriften 221/Tabula Imperii Byzantini 6 (Vienna, 1991), pp. 69–73.

218

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 John’s 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 Justinian’s 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 empire’s southern Mediterranean provinces.106

100 101 102

103 104

105

106

PLRE 3:181–224, s.n. ‘Fl. Belisarius 1’, at pp. 182–7. Battle of Nisibis: Coripp. Ioh. 1.56–67, pp. 5–6; Proc. BP 2.18.16–23, 1:226–7. Service in Mesopotamia: PLRE 3:645–6, s.n. ‘Ioannes 36’. Zachariah of Mitylene, Chronicle 9.2, p. 223 (accompanied Belisarius to Mesopotamia); Proc. BV 1.11.5–6, 1:361 (commander of foederati on Vandal campaign). See also PLRE 3:1167–77, s.n. ‘Solomon 1’, at pp. 1168–9. Kaegi, Heraclius, pp. 21–5; see further PLRE 3:584–6, s.n. ‘Heraclius 3’, and the sources cited there. Killed two Byzantine governors: Proc. BV 2.27.17, 1:542 with Proc. BP 2.3.6–7 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 3:125–30, s.n. ‘Artabanes 2’, at p. 125. Guntharis’ coup: Proc. BV 2.25.1–2.26.33, 1:532–540; 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 2.27.9–10 and 2.28.1–41, 1:541 and 1:545–51; 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. 4.232–42, 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. 64–5.

219

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 senator’s 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 Justinian’s niece, Praejecta (see above, section 2.1). To be sure, Areobindus was appointed in the midst of Antalas’ revolt, when many of the empire’s 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, 1:529–30 (‘! 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 3:821–3, ‘Marcianus 7’, whose previous career is not clear. Proc. BV 2.22.1–2.24.4, 1:522–30. See also Jones, Later Roman Empire, 1:386.

220

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans Table 4.5. Duces of Egypt, c.538–641: previous and subsequent careers Name

Dates

Previous career

Alexandria Hephaestus Liberius 3

?545/6 538/9–c.542

see below, Thebaid PPO of Italy, c.493–500 PPO of Gaul, 510–34 (both under Ostrogoths)

Nicetas 7

?610–17

Led military expedition to Libya and Egypt, 609 praefectus Augustalis, 538 ?MVM in Egypt, 640–1

Rhodon 538–9 Theodorus 166 641–2 Arcadia Domentianus Marcianus 6

640–1 549/50

Thebaid Apion 3

548–50

Military commander in Italy and Spain (despite frets about his lack of military experience) Patrician and comes excubitorum 612–13 (–?)

? 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), 566–8

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, 551–2 (cont.)

221

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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, 538–40 ?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 Justinian’s magistri militum were relatives of the emperor. Germanus was Justinian’s 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 Justinian’s favour.113 Once in Africa, however, even Belisarius’ loyalty was open to question. Procopius claims that the general’s 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 Justinian’s 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 Marcian’s 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 1.12.20–1, 1:58. PLRE 3:553, s.n. ‘Gregorius 16’. Proc. BP 1.24.40–54, 1:130–3; Malalas, Chronographia, p. 476; Chronicon Paschale s.a. 532, 1:621; Theophanes, Chronographia AM 6024, 1:181–6, esp. p. 185; Cedrenus, Compendium Historiarum 1:647; and Nicephorus Callistus, Historia Ecclesiastica 17.10, PG 147:244C-D.

222

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 general’s 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 emperor’s 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 Solomon’s nephew. Procopius indicates that the young man’s appointment was a product of Justinian’s 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 one’s 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 2.8.1–8, 1:452–3. 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 1.11.5–6, 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, 1:517–20 and 1:522–3, 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 hi´erarchie militaire’, p. 171. Respect for Solomon: Proc. BV 2.22.11, 1:523–4.

223

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 region’s 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 John’s service on the Persian frontier, Procopius and Corippus give us very different appraisals of the commander’s 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 John’s 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 John’s subsequent tenure as supreme commander of the Byzantine forces in Africa is cursory; the historian’s final assessment of conditions in Africa under John is positive but far from exuberant; and the general’s 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 578–82), 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 2.21.13–16 and 2.21.19, 1:519–20. 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 3:1124–8, s.n. ‘Sergius 4’, at p. 1125. Coripp. Ioh. 3.318, p. 59; Proc. BV 2.17.6 and 2.17.13–17, 1:501–3, and see also a little below, ibid., 2.17.19, 1:503 where John’s troops are conspicuously not mentioned. Proc. BV 2.28.45–52, 1:551–2; see also Proc. BG 4.17.20–2, 2:579. John of Biclar, Chronicon, s.a. 577.47, ed. Collins, p. 69; CIL 8.2245 + 17671 = ILCV 795 = J. ´ Durliat, Les D´edicaces d’ouvrages de d´efense dans l’Afrique byzantine, Collection de l’Ecole franc¸aise de Rome 49 (Rome, 1981), no. 28, pp. 67–71 = Pringle, Defence, p. 329, no. 33; Durliat,

224

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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

D´edicaces, no. 29, pp. 71–7 = Pringle, Defence, p. 328, no. 31; and Pringle, Defence, p. 334, no. 49. See also CIL 8.12035 = ILCV 793 = Durliat, D´edicaces, no. 30, pp. 77–8 = 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. 28–37; V. Laurent, ‘Une Effigie in´edite de Saint Augustin sur la sceau du duc byzantin de Numidie Pierre’, Cahiers de Byrsa 2 (1952), pp. 87–95; N. Duval, ‘Nouvelles recherches d’arch´eologie et d’´epigraphie chr´etiennes a` Sufetula ´ (Byzac`ene)’, M´elanges d’arch´eologie et d’histoire de l’Ecole franc¸aise de Rome 68 (1956), pp. 284–6, no. 9; V. Laurent, Les Sceaux byzantins du M´edaillier Vatican, Medagliere della Biblioteca vaticana 1 (Vatican City, 1962), pp. 85–7, no. 92; Y. Duval, ‘Le Patrice Pierre, exarque d’Afrique?’, Antiquit´es africaines 5 (1971), pp. 209–14; Pringle, Defence, p. 336, no. 54 (see also ibid, pp. 37–8, no. 55); and in general PLRE 3:1013, s.n. ‘Petrus 70’. On Peter, see also Maximus Confessor, Epistolae, PG 91:363–649, here Epp. 13–14, cols. 509–44; Diffloratio ex epistola s. Maximi scripta ad Petrum illustrem, in Maximus, Opuscula Theologica et Polemica, PG 91:141–6; and Maximus, Brevis enarratio christiani paschatis, PG 19:1217–80. John of Biclar, Chronicon, s.a. 569.11, ed. Collins, p. 62; A. Merlin, in BCTH (1925), pp. ccliv–cclv.

225

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Justinian’s 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 565–78) and Tiberius II (ad 578–82).130 As with the region’s 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., 2.8.17–19, 1:186–7 (abandoned Antioch); ibid., 2.16.17–19, 1:223–4 (persuaded to participate in the invasion of Persia in 541); ibid., 2.19.33–4, 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 3:1226–7, s.n. ‘Theoctistus 2’. PLRE 3:1317–19, s.n. ‘Thomas 15’. PLRE 3 lists 36 Thomases for the period ad 527–641. Durliat, D´edicaces, nos. 25 and 27–8, pp. 59–62 and 64–71 = Pringle, Defence, pp. 327 and 329, nos. 29 and 32–3 (respectively).

226

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 (536–9) dux Byzacenae (545–6) dux Numidiae (545) magister militum Byzacenae (sixth–seventh centuries) magister militum Byzacenae (sixth–seventh centuries) magister militum (Carthage, sixth–seventh 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, 1972–85), 1/3:1643, no. 2885 and P. Monceaux, ‘Enquˆete sur l’´epigraphie chr´etienne d’Afrique’, Revue arch´eologique, 4th ser. 2 (1903), 75, no. 15. See also ODB 3:1935–6, s.v. ‘Spatharios’; ODB 2:1154, s.v. ‘Koubikoularios’; and M. McCormick, ‘Emperor and Court’, in A. Cameron, B. Ward-Perkins, and

227

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman the office of comes excubitorum, or commander of a select corps of imperial guards.133 The senator Domnicus accompanied Justinian’s 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 Egypt’s 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 uncle’s 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 Solomon’s 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 425–600, Cambridge Ancient History 14 (Cambridge, 2000), pp. 135–63, esp. 151–2. Cf. Brown, Gentlemen and Officers, pp. 65–6. Proc. BV 2.12.17 and 2.14.35, 1:472 and 1:487 with PLRE 3:1248, s.n. ‘Theodorus 9’, and ODB 1:646–7, s.v. ‘Domestikos ton Exkoubiton’. Proc. BV 2.16.2, 1:497; thus PLRE 3:415–16, 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 3:96–8, s.n. ‘Apion 3’, who was consul and comes domesticorum before becoming dux Thabaidis; see also ibid., pp. 1256–7, 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 2.22.12–20, 1:520 and 1:524–5. Cyrus and Sergius: BV 2.21.1, 1:517–18 with PLRE 3:374 and 1125, s.nn. ‘Cyrus 3’ and ‘Sergius 4’.

228

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans Table 4.7. Early commanders in Africa: previous careers Name

Position

Previous career

A¨ıgan

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 533–4 served at the battle of Dara (530); commander of foederati 533–4 served at the battle of Dara (530); commander of foederati 533–4 served in Persian wars under Sittas; commander of foederati 533–4 served at the battle of Dara (530); commander of allied Heruls 533–4 member of Belisarius’ household; cavalry commander, bandifer 533–4

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 Kˆısˆıl, 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 A¨ıgan, who served in Byzacena after the reconquest.142 A¨ıgan, moreover, had been a member 139

140 141 142

John of Nikiu, Chronicle, 109.24, p. 176 with PLRE 3:762, s.n. ‘Kˆısˆıl’. 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 exarch’s jurisdiction: John of Nikiu, Chronicle, 107.4 and 12–13, pp. 167–8 (the quotation is from 107.4, p. 167) with PLRE 3:779–80, s.n. ‘Leontius 28’. Greg. Ep. 9.9, 2:570. Solomon: PLRE 3:1167–77, s.n. ‘Solomon 1’, at pp. 1168–9. Marcellus: Proc. BP 1.13.21, 1:62. A¨ıgan: Proc. BP 1.13.20, 1.14.39, and 1.14.44, 1:62 and 1:71–2; Cyril: Proc. BP 1.13.21, 1:62; Pharas: Proc. BP 1.13.19–27, 1.14.32–3, and 1.14.39, 1:62–3 and 1:70–1.

229

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Solomon’s 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 Justinian’s 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 Justinian’s 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 545–6, 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 Justinian’s 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 1.15.3–17, 1:74–7; Malalas, Chronographia, pp. 469 and 472–3; and PLRE 3:420–1, s.n. ‘Dorotheus 2’. Martin: Proc. BP 1.21.27–1.22.2, 1:114–15. On his career in general, see PLRE 3:1160–3, s.n. ‘Sittas 1’. Proc. BV 2.27.4–6 and 2.27.31, 1:541 and 1:544 (Byzacena); Proc. BG 2.5.1, 2:170 (Italy). PLRE 3:818–19, 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 2.23.3–17, 1:525–7 (Himerius); see also ibid., 2.21.26–8, 1:521–2 (death of Solomon).

230

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 general’s 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. 44–5, 48–9, and 173–4. Proc. BG 3.10.1–3, 2:336–7; see also ibid., 3.12.4 and 3.39.16–17, 2:347 and 2:473–4; 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. 1.15–20, 3:7–9.

231

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 A¨ıgan 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 Troglita’s 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. 61–2 and 69. 154 Proc. BV 2.21.15, 1:519. Proc. BV 2.10.3–11, 1:459–60; 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-Yahy¯a al-Bal¯adhur¯ı, Kit¯ab fut¯uh al-buld¯an, 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), Kit¯ab al-c Unw¯an, ed. and trans. A. Vasiliev as Kitab al-c Unvan: histoire universelle e´crite par Agapius (Mahboub) de Menbidj, PO 5, 7, 8, 11 (Paris, 1910–15), 8:479; Michael the Syrian, Chronicle 11.10, ed. and trans. J. B. Chabot in Chronique de Michel le Syrien, Patriarche Jacobite d’Antioche (1166–1199), 4 vols. (Paris, 1899–1910), 2:440; Gregorius Ab¯u 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. 203–4, 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 2.24.9–14, 1:531–2; Marcellinus comes, Chronicon addit. s.a. 545.2, p. 107; Coripp. Ioh. 4.103–200, pp. 70–3. Proc. BV 2.24.15, 1:532. First five: Proc. BV 2.15.50–9, 1:495–6. 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. 1.390–403, pp. 18–19.

232

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 province’s 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 Gelimer’s 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 Justinian’s 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, 1:381–2 and Kelly, Ruling the Later Roman Empire, pp. 37–43, 89–93, and 194–5. Proc. BV 2.8.1–8, 1:452–453. Proc. BV 1.12.1–2, 1:365 (departure in Mar. 533) and ibid., 1.15.31, 1:380 (arrival in Africa three months later); see also ibid., 1.21.17–25, 1:402–4 (entry into Carthage in mid September 533). Recall shortly after late March 534: ibid., 2.7.1–17 and 2.8.1–8, 1:448–51 and 1:452–3. Proc. BV 2.22.2, 1:522 (misrule); ibid., 2.24.4–6, 1:530 (shared command of Africa with Areobindus); ibid., 2.24.7–8, 1:530–1 (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 3.31.2–4, 2:431. Proc. BV 2.16.1, 1:497 (appointment); ibid., 2.19.1, 1:508 (recall ad 539/540); Proc. BP 2.6.9–10, 1:174–5 (sent to Antioch). Proc. BV 2.28.45, 1:551; Proc. BG 4.17.20–2 and 4.24.33–7, 2:579 and 2:622–3; Marcellinus comes, Chronicon addit. s.a. 547.6, p. 108; Jordanes, Romana 385, pp. 51–2; and Coripp. Ioh., esp. ibid., 1.48–53, p. 5. Proc. BV 2.14.37–41 and 2.21.26–8, 1:487–8 and 1:521–2.

233

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman with Solomon’s administration of Africa and were it not for his untimely death the general’s 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 582–602). 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, 1.72–3, 4.7, 6.62, 7.3, and 9.9, 1:70–1, 1:80–2, 1:223, 1:436–7, 1:445–6, 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 3:1164–6, s.n. ‘Smaragdus 2’ (two terms, each perhaps 5 years); ibid., pp. 1092–3, s.n. ‘Romanus 7’ (died in office after perhaps 7 years); ibid., pp. 264–5, s.n. ‘Callinicus 10’ (perhaps 6 years); ibid., pp. 435–6, 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 5–6 years and mode of 5 years. PLRE 3:142–4, s.n. ‘Athanasius 1’.

234

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 George’s 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 641–68). 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 3:521–2, s.n. ‘Georgius 50’. PLRE 3:1317–19, s.n. ‘Thomas 15’. Proc. BV 2.16.2, 1:497 (appointment) and ibid., 2.19.1, 1:508 (recall). PLRE 2:469–70, s.n. ‘Fidelis’ (died in office after perhaps slightly over 1 year); ibid., pp. 939–40, s.n. ‘Reparatus 1’ (died in office after roughly 1 year); PLRE 3:142–4, 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. 515–16, 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 91:364–92. Maximus Confessor, Ep. 12, PG 91:460–509; on this incident see also ibid., 18, PG 91:584–9.

235

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman and commerciarius’).182 In the opinion of C´ecile 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 Martina’s fall from power in December 641 or January 642. George’s 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 Solomon’s 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 Solomon’s command.187 Again, though, Justinian’s 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. 232–3, nos. 13–14. 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 2.27.5–6 and 2.27.31, 1:541 and 1:544; Coripp. Ioh. 4.532–40 and 5.447–9, pp. 87 and 109. For the identification, see PLRE 3:818–19, s.n. ‘Marcentius’. Proc. BV 2.8.24, 2.15.49, and 2.17.5–6, 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., 2.14.32–41, 1:487–8 (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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

New Rome, New Romans Only among regimental commanders and Justinian’s 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 533–4 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 empire’s 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., 2.18.3–4, 1:505–6 (informed by Asclepiades about the plot of Maximinus). Service against Vandals: Coripp. Ioh. 7.385, p. 158. Last appearance: ibid., 7.374–498, pp. 158–62; and in general PLRE 3:790–1, s.n. ‘Liberatus’. Proc. BV 2.21.15, 1:519 (death) and for his career see above, n. 67. Egypt: Maspero, Organisation militaire, pp. 88–91; 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. Beˇsevliev, 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 1.24.18–20, 2:120; Marcellinus comes, Chronicon addit. s.a. 537.2, p. 105. Cyprian: Proc. BG 1.23.19–20, 2:116 (Rome); ibid., 2.23.2, 2.24.18, and 2.27.26, 2:251, 2:259–60, and 2:273–4 (Fiesole); ibid., 3.5.4, 2:318 (Florence); ibid., 3.6.8, 3.12.18–20, 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 3:368–70, 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 John’s exploits in Mesopotamia, Procopius typically in a much less flattering light: Proc. BP 2.18.16–23, 1:229–31; cf. Coripp. Ioh. 1.56–98, pp. 5–7. John in Mesopotamia in 541: Proc. BP 2.14.12, 1:215. Stationed in Africa until 540/1: Coripp. Ioh.

237

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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)

538–40/549

2–11 years

535–46/550 593–5 537–48

11–15 years 2 + years 11 years

Constantinus 3 ?MVM vacans Cyprianus ?commander of foederati MVM vacans Francio 1 MVM

535–7/538 537–40

3 years 8 years (total)

killed by his troops executed murdered

Fulcarius Herodianus 1

553 535–40

approx. 20 years >1 year 5 years

killed in action defected to Goths

542–5 537/8–540 535–7/546 537–49 535–44 537–40 598–9 591–2 538–40/543

3 years 2 years 2–10 years 12 years 9 + years 3 years 11/2 + years 1 + year 2–6 years

535–7

2 + years

535–9/543 538–9 552–3 553–66

4–8 years >1 year 1 year 13 years

535–7 544–5

4–10 years (total)

magister militum magister militum patrician MVM magister militum infantry commander

536–40 547–54/555 559 547–50 591–2 535–7

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

540–5 c.568–c.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.

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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.

3.29–34, 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Prefect’s 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 Justinian’s 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 533–4, 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 Solomon’s general overhaul of the army in Africa in 539–40, 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. 4.82–5, 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. 51–2. (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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Justinian’s 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 region’s 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 Solomon’s 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. 79–80; 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. 79–85, but also, e.g., Jones, Later Roman Empire, 1:614–19 ¨ and J. Haldon, Recruitment and Conscription in the Byzantine Army c.550–950, Osterreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, Sitzungsberichte 357 (Vienna, 1979), pp. 20–8. 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. 260–1.

241

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 taxe¯ot¯es (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 wife’s 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 2.14.8–10, 1:483–4; W. Kaegi, Byzantine Military Unrest, 471–843: An Interpretation (Amsterdam, 1981), pp. 47–50. On soldiers’ marriages in general, see Lee, War in Late Antiquity, pp. 149–53. 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, Perª tän –n Sin ‰g©wn pat”rwn 40, ed. F. Nau, ‘Le Texte grec des r´ecits du moine Anastase sur les saints p`eres du Sina¨ı’, Oriens Christianus 2 (1902), p. 83. Coripp. Ioh. 1.197–202, p. 11. Proc. BG 3.31.13, 2:432; see also Proc. BV 2.28.43–4, 1:551 and Proc. BG 3.31.2–16, 2:431–3. Guntharis himself had hoped to regularize his rule by marrying Praejecta: Proc. BV 2.27.20–2, 1:543 and Jordanes, Romana 384, p. 51.

242

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 exarch’s wife remained in the imperial capital, as did the fianc´ee 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 general’s 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 rebel’s 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. M¨uller, in Fragmenta historicorum Graecorum, 5:37; Histoire nestorienne in´edite (Chronique de S´eert) 2.82, ed. A. Scher, PO 4, 5, 7, and 13 (Paris, 1908–18), 13:526–7; Agapius, Kit¯ab al-c Unw¯an, 8:449; Cedrenus, Compendium Historiarum 1:711–12; 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. 25–37. John of Antioch, frag. 218f, ed. M¨uller, 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. 1.15–20, 3:7–9. Coripp. Ioh. 1.197–207, p. 11. Ibn c Abd al-Hakam, Fut¯uh Ifr¯ıqiya w’al-Andalus, p. 46. As told in the Arabic sources for the conquest of Africa the presence of Gregory’s 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, 1975–86), 2:490–5; 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. 29–40. For a fuller discussion of the early Arabic historical

243

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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. 1.375–404, pp. 18–19. 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. 22–3.

244

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Greek–Latin 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 city’s rectangular harbour, Mandrakion, is a Greek word meaning ‘little square’, while the name for the palace prison, Ank¯on, is Greek for ‘the corner’. Procopius records a children’s 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, D´edicaces, no. 2, pp. 9–11 = Pringle, Defence, pp. 322–3, no. 15 (Bordj Hellal); Durliat, D´edicaces, no. 6, pp. 18–21 = Pringle, Defence, p. 319, no. 5 (Madaura, mod. M’daourouch); and perhaps Durliat, D´edicaces, B, p. 90 = Pringle, Defence, p. 320, no. 8 (Churisa, mod. La Kessera). Greek inscriptions: Durliat, D´edicaces, no. 10, pp. 25–6 = Pringle, Defence, p. 321, no. 12 (Thagura, mod. Taoura); Durliat, D´edicaces, no. 14, pp. 35–7 = Pringle, Defence, pp. 320–1, no. 9 (Sufes, mod. Sbiba); and perhaps Durliat, D´edicaces, no. 34, pp. 85–6 = Pringle, Defence, p. 322, no. 13 (Thagura), which may be a copy of the first inscription from Thagura. Durliat, D´edicaces, no. 26, pp. 62–4 = Pringle, Defence, p. 328, no. 30 (Sidi Gherib), a bilingual inscription, dates to the joint reign of Justin II and Tiberius II (ad 574–8). Christides, Byzantine Libya, p. 12; see also Durliat, D´edicaces, 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. 10–11.

245

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman Greek–Latin inscription dedicating a fortification and dating to the joint reign of Justin II and Tiberius II (ad 574–8) 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 capital’s two languages: Latin. Junillus had never learned Greek, he couldn’t 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, D´edicaces, no. 26, pp. 62–4 = Pringle, Defence, p. 328, no. 30 (Sidi Gherib). See also in general J. Desanges, ‘Quelques consid´erations sur l’usage du grec dans les ports de l’Afrique romaine’, Graeco-Arabica 6 (1995), pp. 27–36. Dossey, ‘Last Days of Vandal Africa’, pp. 113–16. Eastern connotations: see PLRE 1:559, s.n. ‘Marianus 2’; PLRE 2:772, s.n. ‘Marianus 2’; and PLRE 3:829–30, s.nn. ‘Marianus 1–2 and 4–5’. 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. 20.17–20, 3:127–8; see, however, M. Maas, Exegesis and Empire in the Early Byzantine Mediterranean, Studien und Texte zu Antike und Christentum 17 (T¨ubingen, 2003), pp. 1–115, esp. 12–13, and below, Chapter 6, n. 149.

246

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Sbe¨ıtla, 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 Sbe¨ıtla, 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 Bora¨ıdes.230 Perhaps the most fully documented epigraphic corpus from Byzantine Africa, however, comes from Ha¨ıdra in Byzacena, near the modern border between Tunisia and Algeria. The vast majority of the names recorded in the churches of Byzantine Ha¨ıdra speak of a singularly 227

228

229 230

Italy: Brown, Gentlemen and Officers, pp. 68–9. 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. 260–1. ´ N. Duval, ‘Inscriptions byzantines de Sbe¨ıtla (Tunisie) III’, M´elanges de l’Ecole franc¸aise de Rome (Antiquit´e) 83 (1971), pp. 431–3 = AE (1971), 174, no. 495 = Pringle, Defence, pp. 335–6, 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 Sbe¨ıtla (Tunisie) III’, pp. 428–31 = AE (1971), 173–4, 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; Beˇsevliev, Personennamen, p. 79; and PLRE 3:245–6, s.n. ‘Boraides’ (a cousin of Justinian); but see also K. Jongeling, North-African Names from Latin Sources (Leiden, 1994), p. 28.

247

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman Christian, Latin, African identity. The city’s 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 city’s 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 Ha¨ıdra, they seem not to have greatly affected the existing structures of power. Certainly, the city’s 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

´ N. Duval, Recherches arch´eologiques a` Ha¨ıdra, 2 vols., Collection de l’Ecole franc¸aise de Rome 18/1–2 (Rome, 1975–81), 1:413–16. The subsequent publication of numerous supplements to this catalogue in Duval, Recherches arch´eologiques a` Ha¨ıdra II, pp. 215–24; F. Baratte, F. B´ejaoui, ´ and Z. Ben Abdallah (eds.), Recherches arch´eologiques a` Ha¨ıdra. Miscellanea 2, Collection de l’Ecole franc¸aise de Rome 17/2 (Rome, 1999), pp. 143–7, 217–25, and 233; and F. Baratte, F. B´ejaoui, ´ and Z. Ben Abdallah (eds.), Recherches arch´eologiques a` Ha¨ıdra III, Collection de l’Ecole franc¸aise de Rome 18/3 (Rome, 2009), pp. 131–55 has not substantially changed the situation. PCBE 1:333, s.n. ‘Dynamius’. The name Dynamius in Byzantine Ha¨ıdra: Duval, Ha¨ıdra 1:48–9, 57–8, 96–7, 109–10, and 277–8, nos. 23, 31, 66, 77, and 414; ibid., pp. 287–9, no. 424, which also mentions the name Dynamius, appears to have been Vandal-era. The name Innocent: Duval, Ha¨ıdra 1:103–4, 112–13, 259–60, 284, and 291–2, 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, Ha¨ıdra 1:401–3. The name Theodore (Duval, Ha¨ıdra 1:50–1 and 64–5, nos. 24 and 38) was popular throughout the late Roman world; for Evodia (Duval, Ha¨ıdra 1:129–30, no. 97: Ivuza) and Evodius, see Notitia, Byz. 6, p. 124 and PCBE 1:366–74, s.nn. ‘Evodius 1–2’; for Diotimus (Duval, Ha¨ıdra 1:36–7, 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, Ha¨ıdra 1:122, no. 88), see PCBE 1:534, s.nn. ‘Gennadius 1–2’. Less well attested in Africa are Theophilus: Duval, Ha¨ıdra 1:214–15, no. 203 (tentative: ‘Teauf?il?[i]’); see also PLRE 3:1307–9, s.nn. ‘Theophilus 1–5’. Zosimus: Duval, Ha¨ıdra 1:129, no. 96 (tentative: ‘Zosiu’); see also PLRE 3:1421, s.n. ‘Zosimus’. Antiochia: Duval, Ha¨ıdra 1:232–3, no. 304; see also PLRE 3:90–1, s.nn. ‘Antiochus 1–6’, most of whom are attested only in the sigillographic record. See also Duval, Ha¨ıdra 1:77–8 and 125, nos. 50–1 (Poemenius) and 91 (Polibius), neither of whose names appears in PLRE 3. Duval, Ha¨ıdra 1:193–5 and 256–8, nos. 200A (Marcellus) and 402 (Silvanianus). Two earlier inscriptions, presumably Vandal-era, also record one Albucius inlustris (ibid., pp. 53–4, no. 26) and Astius Vindicianus vir clarissimus et flamen perpetuus (ibid., pp. 254–5, no. 401).

248

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 Ha¨ıdra is an optio with the Latin name Maurianus.235 A similar pattern emerges in what is now eastern Algeria, where the citizens of A¨ın el-Ksar constructed a castrum in their town at some point in the reign of Tiberius II Constantine (ad 578–82). 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 garrison’s 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 A¨ın 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 A¨ın 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 region’s defence. Three brothers named Maximianus, Stephen, and Mellosus – presumably local landowners – built 234

235 236

Duval, Ha¨ıdra 1:155–8, nos. 121–2; see also ibid., pp. 140–1, no. 109, perhaps the tomb of his wife. On the title, see the discussion ibid., pp. 448–9. For the local currency of the name, see Astius Mustelus, flamen perpetuus christianus: ibid., pp. 273–7, no. 413. Duval, Ha¨ıdra 1:64, no. 37, with the discussion ibid., p. 449. Durliat, D´edicaces, no. 29, pp. 71–7 = Pringle, Defence, p. 328, no. 31, and see the discussions at Durliat, D´edicaces, pp. 75–7 (who suggests that Phocas was the maˆıtre-d’œuvre) and Pringle, Defence, pp. 74–7. On the name Guntharith, see M. Sch¨onfeld, W¨orterbuch der altgermanischen ¨ Personen- und V¨olkernamen 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. xxiv–xxv. 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. 395–409.

249

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

Staying Roman a tower there at some point in the late sixth or early seventh century, in the reign of Maurice (ad 582–602) 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 Sbe¨ıtla, bore a Greek name.240 The primicerii John and Donatus, on the other hand, both of whom were buried in T´ebessa 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, D´edicaces, no. 30, pp. 77–8 = Pringle, Defence, p. 330, no. 36 (Ksar Lemsa). On local management of public works, including fortifications, see CJ 1.4.26 (ad 530), pp. 42–4. On defensive works in Africa in general, see Durliat, D´edicaces, pp. 93–114, esp. 112–13. Durliat, D´edicaces, no. 28, pp. 67–71 = Pringle, Defence, p. 329, no. 33, with ibid., p. 218. Durliat, D´edicaces, no. 32, pp. 80–3 = Pringle, Defence, p. 331, no. 39. Cosmas: Duval, ‘Inscriptions byzantines de Sbe¨ıtla (Tunisie) III’, pp. 439–41 = 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 empire’s 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. Augustine’s 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 emperor’s family, his former bodyguards, commanders of palace troops, high-ranking senators, and others who had gained the autarch’s 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. 55–120.

251

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.005

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 533–4. In the poet’s 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 poet’s 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.006

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 Moorish–imperial 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

2

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. 22–3; 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’, Rome’s Fall and After (London, 1989), pp. 1–32, esp. 22–5.

253

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.006

Staying Roman forceful diplomacy with the politics of conciliation’ that seems to have characterized the Roman empire’s original expansion into the Maghribi hinterland.3 In the wake of his army’s 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 empire’s 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 empire’s 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 army’s 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. Mod´eran, Les Maures et l’Afrique romaine (IVe –VIIe si`ecle), Biblioth`eque des e´ coles franc¸aises d’Ath`enes et de Rome 314 (Rome, 2003), pp. 585–93. 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

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.006

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 – Justinian’s 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 Moorish–imperial relations in the Byzantine period is perhaps a corollary of the emperor’s 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 province’s 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, 2.4.10–12, 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, 1:432–3 (beauty of these slaves); ibid., 2.14.8, 1:483 (enslaved women married to their captors). Proc. BV 2.4.10–12, 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., 2.14.17–18, 1:484–5 (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. 81–4. Proc. BV 1.25.3–8, 1:412–13. See, however, Vict. Vit. 1.25, 1.35–8, 2.4, 2.28, 2.32–3, 2.36, and 3.68, pp. 12, 16–17, 25, 34, 36–7, 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 18.5–9, 3:42 and 112–13; M. Cesa, ‘La politica di Giustiniano verso l’occidente nel giudizio di Procopio’, Athenaeum, Studi Periodici di Letteratura e Storia

255

Downloaded from https://www.cambridge.org/core. Staatsbibliothek zu Berlin - Preussischer Kulturbesitz, on 18 Aug 2020 at 21:18:44, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139048101.006

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

dell’Antichit`a n.s. 59 (1981), pp. 389–409; and A. Kaldellis, Procopius of Caesarea: Tyranny, History, and Philosophy at the End of Antiquity (Philadelphia, Pa., 2004), pp. 17–24. 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. 167–80. See also, however, Kaldellis, Procopius of Caesarea, p. 92. Proc. BP 1.19.13, 1:102 (Arabia); ibid., 1.19.29, 104–5 (Upper Nile); Proc. BG 2.15.16–19, 2:217 (Scandinavia: Scrithiphini); Proc. BP 1.15.21, 1:77 and much more expansively Proc. Aed. 3.6.2–5, 4:95–6 (Caucasus: Tzani), and Proc. BP 2.28.27, 1:286: ‘!/ A 0!