Roman Edessa: Politics and Culture on the Eastern Fringes of the Roman Empire (Routledge Classical Monographs)

ROMAN EDESSA ROMAN EDESSA Politics and culture on the eastern fringes of the Roman Empire, 114–242 ce Steven K. Ross...

77 downloads 1226 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

ROMAN EDESSA

ROMAN EDESSA Politics and culture on the eastern fringes of the Roman Empire, 114–242 ce

Steven K. Ross

London and New York

First published 2001 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 2001 Steven K. Ross The right of Steven K. Ross to be identiﬁed as the author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988 All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Ross, Steven K., 1955– Roman Edessa : politics and culture on the eastern fringes of the Roman Empire, 114–242 C.E. / Steven K. Ross. p. cm. Includes bibliographical references and index. 1. S¸anlıurfa (Turkey) – History. 2. Romans – Turkey – S¸anlıurfa. I. Title. DS51.S22 R67 2000 939′.3 – dc21 99–046969 ISBN 0-203-99197-4 Master e-book ISBN

ISBN 0–415–18787–7

FOR YOLANDA

CONTENTS

List of ﬁgures Preface Acknowledgements Abbreviations

viii ix xi xii

Introduction

1

1

The earliest Edessa

5

2

The coming of Rome

29

3

From kingdom to province

46

4

A king in Rome’s service

69

5

A ‘Golden Age’?: the culture of pre-Christian Edessa

83

6

Early Christianity and Edessan culture

117

7

Conclusion

139

Appendix: Numismatic notes Notes Bibliography Index

145 163 185 197

vii

FIGURES

Int.1 1.1 5.1 5.2 5.3 5.4 5.5 5.6 5.7 App.1a App.1b App.2 App.3 App.4a App.4b

The Shelmath inscription Map of the area around Edessa The Funerary Couch mosaic The Tripod mosaic The Phoenix mosaic The Orpheus mosaic Town plan of Urfa The Aptuh.a mosaic The Family Portrait mosaic Abgar X, the ‘Presentation’ coinage: obv. (Emperor Gordian III) Abgar X, the ‘Presentation’ coinage: rev. Abgar X and Gordian III (rev.) The ‘Adventus’ of Abgar X at Edessa (rev.) Coin of Colonia Edessa under Gordian III: obv. Colonia Edessa under Gordian III: rev., showing city’s Tyche

viii

2 6 93 95 97 98 103 112 114 146 147 148 149 159 160

PREFACE

This work arose from a dissertation submitted in 1997 at the University of California at Berkeley. Its original kernel, however, was a paper written for a seminar on Documents from Roman Syria under the direction of Professor Glen Bowersock, of the Institute for Advanced Studies at Princeton. The focus of that paper was on the Syriac parchments from Edessa found along with the Euphrates Papyri cache, and on historical conclusions that could be reached from those documents in combination with numismatic data. Although the ﬁnal project has advanced far beyond those concerns, they remain at the center of this work, as will be seen. Although the intent has been to deal, as much as possible, with the evidence for early (pre-Christian) Edessa, the city’s later importance as a religious and literary center demands some treatment of Edessan Christianity and culture, the arrival of the new faith and the evidence offered by early Syriac literature. It is in these areas that I feel my credentials to be most lacking, and it is only with trepidation that I offer any opinion at all on controversial topics. Although their names are mentioned in the Acknowledgements, I would like to express here again my deepest personal gratitude to the giants of scholarship in this area, Professors Sebastian Brock, H. J. W. Drijvers, and J. B. Segal. Each of them has been most generous with suggestions and assistance, and their kindness has rescued me countless times from the brink of error. A number of crucial suggestions were provided by the publisher’s anonymous referee, with the result that this work approximates much more closely to a full and balanced treatment. I am fully responsible for any errors and weaknesses that remain. Throughout the extended period of work on this subject I have been sagely counseled, and gently prodded, by Prof. Erich S. Gruen, the mentor of mentors. Glen Bowersock, the original inspiration for ix

ROMAN EDESSA

my work on Edessa, has proven to be an endless source of cheer and encouragement. Finally, I owe an undying debt of gratitude to my long-suffering wife and family. Steven K. Ross Louisiana State University Baton Rouge

x

ACKNOWLEDGEMENTS

The following individuals and institutions have been instrumental in bringing this work to completion, and sincere thanks are offered to all: The American Research Institute in Turkey, The Turkish Ministry of Antiquities, The Archaeological Museum at Urfa, The American Numismatic Society, Prof. Glen Bowersock, Prof. Sebastian Brock, Prof. H. J. W. Drijvers, Prof. Bill Metcalf, Curator of Roman Coins at the American Numismatic Society, Prof. J. B. Segal, and Prof. Javier Teixidor. Grateful permission is also extended to the American Numismatic Society for rights to reproduce photos of Edessan coins studied there during my term as Fellow in Roman Coinage Studies (Figs. App.1–4), and to Prof. Segal for reproductions of, and reproductive rights to the Shelmath inscription (Fig. Int.1), hand-drawn illustrations of Edessan mosaics by Mrs Seton Lloyd (Figs. 5.1–4, 5.7), the photograph of the Mosaic of Aptuh.a (Fig. 5.6), and the Plan of Urfa (Fig. 5.5). Thanks are also due to Prof. Segal for permission to quote the lengthy extract from the Chronicle of Edessa in his translation (Chapter 5). Finally, I acknowledge my debt of gratitude to Richard Stoneman, Coco Stevenson and the staff at Routledge, for their patience during a long and trying period of waiting. Sincere thanks also go to any and all whose assistance I may have failed to acknowledge. S. R.

xi

ABBREVIATIONS

AE BLC

l’Année Epigraphique H. J. W. Drijvers, The Book of the Laws of Countries: Dialogue on Fate of Bardais.an of Edessa, ed. J. H. Hospers and Th. C. Vriezen (Assen: Van Gorcum, 1965): Text (even-numbered pages) with English translation (odd-numbered pages). BMC Arabia, etc. G. F. Hill, Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia (London: British Museum, 1922). BMC Roman Empire Coins of the Roman Empire in the British Museum (London: British Museum, 1923–62). Vol. I: Augustus to Vitellius; Vol. II: Vespasian to Domitian; Vol. III: Nerva to Hadrian; Vol. IV: Antoninus Pius to Commodus; Vol. V: Pertinax to Elagabalus; Vol. VI: Severus Alexander to Balbinus and Pupienus. Bull. Epigr. Bulletin Epigraphique. CAH Cambridge Ancient History (Cambridge: Cambridge University Press 1970– ). Chron. Min. Chronica Minora (CSCO vols. 1–6; Scr. Syr. ser. 3, vol. 4). Cited by page numbers of text/ translation. Chron. Zuq. Chronicle of Zuqnin (‘Chronicle of PseudoDionysius of Tell-Mah.re¯’). CSCO 91, 121, 104; Scr. Syr. ser. 3, vols. 1–2. Cited by page numbers of text/translation. CIG Corpus Inscriptionum Graecarum. CIL Corpus Inscriptionum Latinarum.

xii

A B B R E V I AT I O N S

CRAI

Comptes Rendues de l’Académie des Inscriptions et Belles-Lettres. CSCO Corpus Scriptorum Christianorum Orientalum (Louvain); Scr. Syr. = Scriptores Syri. Dessau Hermann Dessau, Inscriptiones Latinae Selectae. FHG Karl Müller, Fragmenta Historicorum Graecorum (1841–70). Geog. Graec. Min. Karl Müller (ed.), Geographi Graeci Minores (Paris: Firmin-Didot, 1855–82). HA Historia Augusta. IG Inscriptiones Graecae. IGLS L. Jalabert, R. Mouterde et al., Inscriptions greques et latines de la Syrie (1929– ). IGR Inscriptiones Graecae ad Res Romanae Pertinendae. Migne P. G. J.-P. Migne, Patrologiae Cursus Completus, Series Graeca (Paris: Migne, 1857–87). Mommsen, R. Gesch. Theodor Mommsen, Römische Geschichte (Berlin: Weidmann, 1854–94). NHC Nag Hammadi Corpus, cited by codex and tractate number. PIR2 Prosopographia Imperii Romani 2. RE A. F. von Pauly and Georg Wissowa, Realencyclopädie der classischen Altertumswissenschaft (Stuttgart: J. B. Metzler, 1893–1963). RIC Harold Mattingly et al., Roman Imperial Coinage (London: Spink and Son, 1923–94). SEG Supplementum Epigraphicum Graecum (1923– ). YCS Yale Classical Studies. ZPE Zeitschrift für Papyrologie und Epigraphik.

xiii

INTRODUCTION

Visitors to the modern-day economic center and provincial capital of Urfa (S¸anlıurfa) in southeastern Turkey will, almost inevitably, ﬁnd their attention drawn to the most prominent physical reminders of that city’s pre-Byzantine history: the two columns on the rock outcropping known as the city’s ‘Citadel,’ at the southern end of town. Standing sentinel over the modern-day market district and the mosques which are the focus of devotion among the region’s faithful, the two columns are together known as the ‘Throne of Nimrud’ in accordance with local legends relating them to the town’s reputed Patriarchal history. The columns’ bossed drums and Corinthianinﬂuenced capitals clearly reﬂect the inﬂuence of Greco-Roman culture in this community, once known by its Greek name, Edessa – the capital of the small northern Mesopotamian district known as Osrhoene. It is only by a closer approach, and by a careful examination of one of the two, that the observer discovers another telling feature: an inscription in early Syriac characters carved into its face (Fig. Int. 1): I Aphtuh.a, nu[hadra?], son of Bars[—] made this column and the statue which is on it to Shelmath the Queen, daughter of Ma nu the pa[s.]griba, wife of the [king?] my lady [——].1 With this honorary inscription, one of very few surviving from the independent monarchy, a notable pays his respects to the daughter of Ma nu the pas.griba, or heir apparent to the Edessan throne. The date is probably in the ﬁrst third of the third century ce, and Ma nu is likely to have been second-to-last in the royal line, father of Abgar X, the last king of Edessa (239–42 ce). By the time this last of the native monarchs took the throne, the 1

ROMAN EDESSA

Figure Int.1 Early Syriac inscription on a Roman-era column on Urfa’s Citadel hill. It honors ‘Shelmath, daughter of Ma nu the pas.griba’, a member of the Edessan royal family. It is one of very few surviving monuments from the period of Edessa’s native monarchy that can be seen by visitors to Urfa today. Photo courtesy J. B. Segal.

absorption of Edessa and its territory into the Roman Empire was nearly complete, and in fact this king’s reign itself was anomalous. The column of Shelmath, however, is interesting for another reason: the mixture of cultural forces it represents. The language of the inscription is Syriac, a form of Aramaic that evolved in northern Mesopotamia and eastern Syria in the Hellenistic and Parthian periods, but which would produce an impressive corpus of literature only after the annexation by Rome, at the hands of the Christian Fathers. The names of the individuals named in the inscription – Shelmath herself, Ma nu, and Aphtuh.a – show strong afﬁnities with the Arab and other Semitic peoples that surrounded Edessa, which was the center of Syriac language and culture. At the same time, embedded in the inscription is the term pas.griba, a royal title which is not Semitic but of Persian/Parthian derivation. This represents a Parthian strain that also crops up in other contexts when considering Edessan culture. Finally there is the archaeological locus of the inscription: a column in a Classical order showing a marked Greek or Roman inﬂuence. Whether this architectural style is only an expression of the spread of Hellenism around the Near and Middle 2

INTRODUCTION

East, and its survival after the fall of the Seleucid Kingdom, or whether, on the other hand, it is also part of a revival or injection of ‘Classical’ culture under the more recent inﬂuence of Rome, may never be known. In any case, the column and its mate spring out of the cultural context of the late monarchy, and in some degree, they express the layering or mixture of cultural and political inﬂuences to which the kingdom was subject during the more than three centuries of its existence. This book traces the history of Edessa’s encounter with Rome and seeks to clarify the circumstances under which it was eventually absorbed, as well as the reasons for that absorption, and its consequences. Although much of this tale has been told before, it is now possible to bring to bear new evidence and to interpret some of the old evidence in a different light. Moreover, by focusing on events in the ﬁrst two-and-a-half centuries of the Common Era – before there is any convincing evidence of ‘ofﬁcial’ Christianity at Edessa – it is possible to evaluate the events of this period in their own right, uninﬂuenced by the city’s later importance in the history of the Christian Empire. This is, thus, the beginning of an attempt to write the needed ‘history of profane Edessa,’ in the words of the historian Peter Brown.2 That history cannot, however, be written without some attempt to address the other questions posed by the Citadel column and the Shelmath inscription: those of Edessa’s cultural antecedents and inﬂuences, and the signiﬁcance of the mixture that we see in the column, in the inscription and in other expressions of Edessan culture. The answers to some of these questions may be, at the moment, beyond our reach; but this is our second aim: to come as close to those answers as possible, using the evidence of language, art, religion and iconography. The two projects – political history and cultural interpretation – are fairly clearly delimited in the pages that follow, but they illuminate each other at various points, and are intended to form the two parts of a coherent whole. Before taking on that bipartite task it is necessary to deal with some preliminaries, beginning with the historical geography of Edessa: this includes such questions as the potential inﬂuence of long-range trade routes on the city’s prosperity; the site’s strategic signiﬁcance; and the extent of Osrhoene, Edessa’s kingdom and later a Roman province. The Hellenistic origins of Edessa, the arrival of the Abgarid kings during the second century bce, and Parthian overlordship may also be considered preliminaries to the main project, as are Edessa’s early contacts with Rome. After considering all 3

ROMAN EDESSA

these questions, we embark on the main journey, taking our starting point from the ﬁrst full-scale Roman invasion of Mesopotamia, under the Emperor Trajan in 114 ce. The political history of Rome and the Edessan kingdom carries us down to the reign of the last native king, up to Edessa’s ﬁnal incorporation into the empire c. 242 ce. The reader will note that a disproportionate amount of attention is paid to the events of the last few years before that incorporation. This is because it is to these years that the most recently discovered documentary evidence pertains, as does the material that forms the main subject of our numismatic study. At some time in the years leading up to the kingdom’s ﬁnal incorporation, and possibly during the reign of the last Abgar, the inscription of Shelmath was carved into the face of the Citadel column. In our ﬁnal chapters we will turn to the questions posed by that monument, those of Edessa’s cultural identity – by looking at literature, religion, language, art and archaeology. If at the end of this inquiry many more questions than answers still remain, this should indicate not the sterility of the topic, but its potential ability to lead us to a new and deeper understanding of the fascinating place that is the Roman Near East. To ﬁnd the ﬁnal answers to all these questions may be the task of hands more skilled than ours; here, we can but point the way.

4

1 THE EARLIEST EDESSA

Blessed with generous springs that made it a welcome stopping point for travelers in the parched steppeland of middle Mesopotamia, Edessa must certainly have attracted visitors and inhabitants from a very early date. Despite this likelihood, however, Edessa does not enter the historical record until the Hellenistic period. The city is known in Syriac and other Semitic languages as Orhai or Urhai, but it is hard to ﬁx on the pre-Hellenistic traveler’s landscape, by either name. Cuneiform records supply a large number of toponyms in this part of Mesopotamia, but Orhai has never been securely identiﬁed. Nevertheless, given its desirability as a waystation, it seems highly probable that Edessa is mentioned under another name. The best candidate is DM , mentioned in several old-Assyrian and Babylonian itineraries as a staging point near H . arran/Carrhae (Harrak 1992). The Syriac–Arabic lexicon of Bar-Bahloul, although late in date (tenth century), offers some conﬁrmation; it identiﬁes Adme (alternatively Admi or Admum) as ‘the name of a city, Al-Rahha’— which in its turn is the Arabic form of Orhai.1 Although this name is never used for Edessa in extant Syriac literature, the location of Adme/Admi/Admum in cuneiform sources – three stations away from H . arran – helps to consolidate the identiﬁcation (Goetze 1953: 51–2; Harrak 1992: 213). There can be little doubt that the ‘exceptionally well-endowed and water-rich oasis of Urfa’ was frequently visited by commercial travelers of the Ancient Near East (Astour 1989: 687). The reason for its relative lack of prominence in the pre-Hellenistic period is a matter of conjecture. The picture, as far as we can tell, remained essentially unchanged under the Persian Empire, which dominated the Middle East from 550–330 bce. Orhai and the nearby H . arran may have been visited frequently by travelers along the Royal Road from Sardis in Asia Minor to Susa in the Persian heartland under Achaemenid rule, but 5

ROMAN EDESSA

Herodotus’s description of that route makes it seem as if that road (or at least the most frequently traveled version of it) passed through Mesopotamia further north (Hdt. 5.52; cf. Oates 1968: 10). The future Edessa remained a relatively minor waystation remote from the centers of power, despite the potential strategic and economic advantages of its situation (Fig. 1.1). The arrival of the Greco-Macedonian armies of Alexander the

Figure 1.1 Map of Upper Mesopotamia showing the location of Edessa and nearby communities. Edessa’s well-watered oasis made it a frequent stopping point for travelers across the parched expanse to the east of the Euphrates, from a time long before the founding of the Hellenistic city.

6

THE EARLIEST EDESSA

Great and his generals in the fourth century bce changed the face of the eastern Mediterranean world thoroughly and permanently. Alexander swept across the Persian Empire in a wave of conquest and destruction beginning in 334 bce, putting an end to Achaemenid rule. By the time of his death in Babylon, 323 bce, the lands formerly ruled by the Persians, from Egypt and Asia Minor as far east as Bactria and the borders of India, were potentially part of a uniﬁed Macedonian-ruled empire that would also have included the Aegean islands, Macedonia and the Greek mainland. By failing to arrange for a stable succession, however, Alexander practically guaranteed that this potential unity would collapse under the pressure of intense rivalries among his successors – Ptolemy, Antigonus, Lysimachus, Antipater, Cassander and Seleucus I Nicator (‘the Conqueror’). Seleucus’s recapture of Babylon in 312/311 bce—later marked as the starting-point of a new dating system, the Seleucid Era – was the beginning of momentous changes for Edessa and the entire Middle East. One of the hallmarks of the Hellenistic period is the settlement, on broad areas of the newly conquered land, of Greeks and Macedonians – including not only the veterans of the wars of conquest and succession, but former residents of the European homeland whose need for land and new sources of wealth could now be more easily satisﬁed. The inﬂux of settlers spread Greek culture across a vast expanse, bringing a new language, new artistic forms and new modes of religious and political expression, at least to the urban centers in the new territories. Alexander himself is renowned as the founder of numerous colonies, but the city-founding activities of Seleucus and his immediate successors also became legendary, especially in the Syrian portion of their kingdom. The founding of such Syrian colonies as Antioch, Seleucia-in-Pieria, Laodicea and Apamea was among the most successful of all Hellenistic urban endeavors. Antioch itself survived for centuries as an economic and cultural metropolis, long past the end of the Hellenistic period and even beyond Classical Antiquity. Seleucid colonization and cultivation of urban centers, however, was not limited to the Levant. By concentrating on Antioch and its environs, the historian may receive the impression that the Seleucid kings, too, focused their attentions on this, as the area of their kingdom that was in closest contact with the Greek world. A case can be made, however, that the rich region of Mesopotamia, with Babylon at its focus, was just as central to the kings’ thinking. It was certainly more central, geographically speaking, to the whole kingdom. 7

ROMAN EDESSA

In this context, the situation of the future Edessa could hardly have been more advantageous. Unlike the Syro-Palestinian littoral, which was always a bone of contention between the Seleucids and the powerful Ptolemaic kingdom based in Egypt, the northern Mesopotamian region was ﬁrmly in the Seleucid realm. It would seem that one of the early Seleucid kings – according to tradition, Seleucus I himself – soon recognized the economic and strategic advantages of the site, close to the rich Syrian territories but in touch also with the kingdom’s Mesopotamian center of gravity, and capitalized on them by implanting there a new urban settlement.2 The new polis received, like many other Seleucid colonies, the name of a city in the Macedonian homeland (Appian 203; Stephanus Byzantinus s.v. Εδεσσα). According to Pliny the Elder, the city at some point bore the Seleucid royal name of Antioch; but this is not likely, since a dynastic name like this for a royal foundation was a mark of honor, and would not have been dropped in favor of the name Edessa.3 According to one tradition, Seleucus I is supposed to have referred to Edessa as Αντι χεια Μιξοβα´ρβαρο: ‘half-barbarian Antioch’ (Malalas 17.418). The source for this information – the Late Roman Antiochene orator Malalas – is not the most reliable, but in this instance, he may be reporting a truthful tradition. If so, it does not necessarily indicate that the city ever was ofﬁcially named Antioch. The king may have been joking. In any case, it was the name Edessa that stuck – except among speakers and writers of Syriac. These habitually used the name Orhai, which over the course of the centuries evolved into the present-day Urfa. If the purported quotation from Seleucus I is in fact genuine, it suggests that at Edessa, the importation of Greek ways of life and thought, and perhaps of Greek settlers themselves, was less thoroughgoing than elsewhere in Seleucid realms. There is very little information on which to base a judgment of this matter, but the persistence of the Semitic name – which may well have been the name in use before the Seleucid colonization – is suggestive. Even in the most thoroughly Hellenized places, such as Antioch, a large segment of society – the rural populace and the urban underclass – never took on more than a thin veneer, if that, of Hellenism. As it had under Persian rule, Aramaic remained in widespread use. Edessa, in the centuries after the breakup of the Seleucid kingdom, seems to have maintained rather a tenuous connection with Hellenism, and this is more easily understandable if the city was only slightly Hellenized to begin with. By the end of the second century bce, the breakup of the Seleucid 8

THE EARLIEST EDESSA

kingdom, under the pressures of internal rivalries for power and external competition, was practically complete. To the west, the squabbling Seleucids had to deal with the threat of ascendant Roman power as well as their traditional rivals, the Ptolemies; in the east, they faced the Arsacids (Parthians), coming from the northerly reaches of the old Persian Empire, who laid claim to the mantle of the Achaemenid monarchy and had begun to take over sections of the Seleucid kingdom as early as 247 bce. When Rome arrived in the Middle East in the ﬁrst century, it found that the territory beyond the Euphrates was under Parthian hegemony, and the Roman–Parthian contretemps set the tone for the military and diplomatic activity of the next three centuries. For the Greek cities of Mesopotamia, including Edessa, Parthian rule meant dealing with a different overlord, but not necessarily a different form of overlordship. The Parthian administrative system was not greatly different from the ‘satrapal’ organization that the Seleucids had adopted from their Persian predecessors. Signiﬁcantly, it seems that in some cities the most prominent element of society continued to be the descendants of the original Greco-Macedonian settlers (Bauer 1933; Welles 1951).4 At Edessa, events took a different course. Simultaneously with the arrival of the Parthians, a new local regime took power: a series of kings or ‘toparchs’ bearing Semitic names that show clear afﬁnities to Arabic, which historians call the Abgarid dynasty after the name that recurs most often, Abgar.5 How or why this came about is quite uncertain; presumably, however, the Abgarids were in some way allied to the Parthian invaders or simply took advantage of the unsettled conditions to impose their rule. The date of this event can be approximated: one source, the Chronicle of Zuqnin, states that in the ‘Year of Abraham’ 2233 the monarchy came to an end, after lasting for 352 years, which would place the installation of the monarchy in 135 bce. According to the preserved fragments of the Chronicle of Edessa, however, the kingdom was established in (Seleucid) Year 180, 132/131 bce (Chron. Ed. 2).6 Although many aspects of Edessan culture under the late monarchy betray Parthian inﬂuence, there is no telling whether, and how much, Edessa’s overall ethnic make-up changed at the time of the installation of the Abgarids. If Hellenistic Edessa was indeed ‘halfbarbarian’ already, there may have been little change in the general population below the level of the ruling elite. The ﬁrst decades of Abgarid rule in Edessa lie in near-total obscurity, relieved only by the names of the ﬁrst kings as given by 9

ROMAN EDESSA

the Zuqnin chronicler: a list which is of dubious reliability.7 After the organization of the Province of Syria by Pompey in 64 bce, however, Edessa began to be noticed by Rome as an important principality in the region just beyond the frontier, and its kings to be considered as potential friends and helpers. When Pompey’s general Afranius traversed northern Mesopotamia he required the assistance of the local population (Dio 37.5.5, speciﬁcally mentioning the people of Carrhae but not Edessa). A friendly relationship seems to have been achieved between Edessa and Rome, for when Pompey’s ally in the ‘First Triumvirate,’ Marcus Licinius Crassus, arrived on the scene ten years later to seek his military fortune against the Parthians, he relied on the assistance of Abgar of Edessa to guide him on unfamiliar ground (Plutarch Crassus 21–2; Dio 40.20 ff.). The expedition ended in disaster for the triumvir, allegedly because of the treachery of Abgar; yet it is entirely uncertain whether he was led astray by this betrayal or by his own ambition and poor generalship. The incident does, however, conﬁrm an early willingness on Edessa’s part to form a friendly relationship with Rome, even while it remained in what was supposedly the Parthian sphere of inﬂuence. Up until the end of the ﬁrst century ce, however, there was little reason to expect any further extension of the limits of the empire to the east. The boundary with Parthia had been more or less ﬁxed along the lines of the upper Euphrates since pre-Augustan times, and despite the activities of adventurers such as Crassus during the Late Republic, the Augustan diplomacy that recovered Crassus’s standards was seen, at least in imperial propaganda, as establishing a stable Roman–Parthian relationship (and one that was favorable to Rome).8 A key element of this arrangement was the Armenian buffer kingdom, and disputes over Armenia brought Roman legions to the east repeatedly. A second example of the Roman willingness to seek the cause of failures in this region in local treachery is Tacitus’s description of the behavior of another Abgar during a Roman attempt to install Meherdates, a candidate for the Armenian throne, in 49 ce. According to Tacitus (Ann. 12.12 and 14), Abgar went to receive Meherdates and detained him treacherously, eventually abandoning him in Adiabene and leaving him to be defeated by his rival Gotarzes. Again, regardless of the truth of this speciﬁc charge, the background to the alleged betrayal indicates that Rome saw in Edessa at least a potentially useful ally. In each case, the accusation of treachery against the Edessan potentates may be grounded more in an attitude of distrust toward ‘Orientals’ than in the facts, but a ﬁnal conclusion on the issue is 10

THE EARLIEST EDESSA

impossible. It is worth observing that Edessa remained under Parthian hegemony, and would have been unwise to seek too close a relationship with Rome. Rome did not offer a client relationship such as that with Armenia, nor did Edessa seek one. In the circumstances, it was able to play off against each other two empires, to both of which it was but a remote and minor principality. However critical as an arena of operations, Mesopotamia was secondary to Armenia to the north and to the Province of Syria as factors in Rome’s eastern policy. These priorities were apparently conﬁrmed when Nero’s general Corbulo established, and later abandoned, a foothold across the Euphrates in 62–4 ce in pursuit of a settlement with Parthia over Armenia (Tac. Ann. 15.9–12; Dio 62.20–3). In this situation, Edessa found its own interests best served by maintaining a ﬂexibility in its relations with the superpowers. It was not until Rome displayed a new, expansionist attitude in the second century that Edessa’s own approach changed, and then only under duress. It is at this point that we will take up the narrative again, in the next chapter.

Language and culture The issue of Edessan culture is viewed here, in the ﬁrst place, through the perhaps distorting prism of Hellenism. As our very ﬁrst example, the Citadel column, makes clear (Introduction), it would be impossible to do justice to this complicated issue by considering it only in terms of Greek culture versus some indistinct and undifferentiated substratum, whether that substratum be conceived as ‘Syriac’ culture, a Parthian/Semitic mixture, or simply ‘Oriental’ culture in general – a concept that, in any case, is now understood as misleading or useless. Yet a barometer of some sort is needed, and the fact that Edessa (as far as we can tell) was ﬁrst urbanized under the Greco-Macedonian conquerors of the Near East makes Hellenism a logical one. Founded as a Greek city near the center of the Seleucid Kingdom, it underwent a number of political transformations, until its absorption into another ‘Western’ empire. The question here is to what extent those transformations affected Edessan culture, and whether an underlying strain of Hellenism survived the period when the region was controlled by Syriac-speaking monarchs and the Parthian kings. Taking the admittedly meager evidence as a whole, it would seem that the answer to the second question is yes – but it is not as simple a matter as it might seem. The fact that Edessa, after the arrival of 11

ROMAN EDESSA

Christianity, was the scene of heated philosophical/theological controversies, proves the inﬂuence at that time of various schools of thought inspired by Greek philosophy. The later existence at Edessa of a Christian literary ‘school’ where the Greek authors of Classical antiquity were read, translated and transmitted to posterity (often by Arabic-language intermediaries) similarly indicates a familiarity with Hellenism; but in neither case, without any direct evidence, do we have proof of a continuity of tradition in the ﬁrst through the third centuries. Nor does the importance of Syriac as a vehicle for much of the Christian literature throughout the Greek East prove anything in this connection. To conclude that the people of Edessa had any knowledge of Greek or any sort of consciousness of themselves as ‘Hellenes’ before the third century ce, we will have to examine various forms of evidence that are, in the main, indirect and subject to varying interpretations.9 It is, in fact, the language itself that provides us with our greatest problem. If Edessa was founded as a Greek community, the language of at least its leading inhabitants must have been Greek – as was the case at Antioch and other Seleucid colonies. At Edessa, however, Greek was replaced by Syriac, and so completely that there is no direct evidence of any knowledge of Greek by the people or the rulers of Edessa before the third century ce. This is in contrast to the situation at Palmyra in Syria, also under the rule of an Aramaicspeaking elite, where bilingual (Palmyrene-Greek) inscriptions are common; and at Dura-Europos, the military outpost on the middle Euphrates, where inscriptions and grafﬁti attest the continuous use of Greek through the period of Parthian rule. The record at both these sites is far more complete than that at Edessa, and it is possible that the picture may change as further discoveries are made. Those that have been made to date simply reinforce the impression that, under the monarchy, Syriac replaced Greek completely. This is an argument e silentio, and under other circumstances, it might seem to strain positivism beyond the limits of credulity. Greek language and culture, presumably, prevailed among the Greco-Macedonian settlers of Edessa; they were known, at least to the literary class, in Roman Edessa as early as the fourth century ce. Again, this does not prove a continuity of usage. Events in the third and fourth centuries ce, during the ﬁrst century of Edessa’s existence as a Roman city, make it conceivable that Greek began to be commonly used there only after a period in which it was all but completely suppressed by the inﬂuence of Syriac. First, the end of the Abgarid monarchy made Edessa a municipality of the Roman 12

THE EARLIEST EDESSA

Empire, which in its eastern provinces used Greek as a language of administration. Second, the Christianization of the empire after Constantine elevated the status of Edessa, which was already largely Christian, and gave new standing to its intellectual community. A third event, possibly of greater signiﬁcance, also took place toward the beginning of this period: it was the activity of the Edessan philosopher Bardais.an, active in the late second and early third centuries ce. What is known of this thinker’s work betrays a familiarity with Greek philosophy and perhaps with the Greek language as well, and Bardais.an became the founder of a long-lived ‘school’ or Christian sect that transmitted and at times transformed his teachings. Conceivably, it was not until some or all of these events had taken place that the knowledge of Greek spread at Edessa, as only one aspect of a general cultural movement that may have had its origins in the reign of Bardais.an’s patron Abgar VIII, one of the last of the kings. Language is only one aspect of the problem, and I will argue that Hellenism of some sort survived at Edessa despite a possible suppression of Greek. Hellenism was, however, subject to a number of other inﬂuences, and these, too, will need to be examined. Despite the priority given in this account to Greek artistic, linguistic and other inﬂuences, it is extremely difﬁcult to identify any of the various strains of identity as Edessa’s primary cultural afﬁliation at the end of the second and the beginning of the third century ce, based on the evidence at our disposal. The historian seeking clues to the importance of any ancient city will often look to its geographic location, and it is to this subject that we now turn, before returning in the following chapters to the historical narrative and the question of culture. The intent is to give a better deﬁnition to the terms ‘Edessa’ and ‘Osrhoene’ and to understand the commercial and strategic advantages of the site. These geographical questions present us with a bewildering variety of sources, beginning with the physical evidence of the terrain itself.

Historical geography Edessa as a waystation In some respects the site of ancient Edessa is not at ﬁrst promising. Although it lies in the midst of a fertile territory where the north Mesopotamian plain meets the ﬁrst hills leading up to the Armenian highlands, the city is not within easy reach of either the Tigris or the Euphrates. Even after the Hellenistic city’s appearance in the 13

ROMAN EDESSA

historical record, it is still not known as a prominent waystation on antiquity’s commercial thoroughfares, in any period before the Late Roman. Nor does the site in itself exhibit any striking strategic advantages, other than its steady water supply. It is perhaps partly due to these relative disadvantages that it remained for long uncertain whether or not Edessa was mentioned in ancient cuneiform texts at all, despite the prominence therein of its neighbors, H . arran (Carrhae) to the southeast and Nisibis to the east. Nevertheless, during the Hellenistic and Roman period Edessa became the site of several major battles and a key player in the East– West conﬂict, and also achieved a good degree of prosperity. Finally it grew to be one of the greatest cities of northern Mesopotamia in the Late Antique and Byzantine periods. This was surely a product of speciﬁc historical events and processes, rather than (in a deterministic sense) the result of topographical advantages. This does not, of course, mean that geography is irrelevant to the history of Edessa. To mention just one factor, the city lies within the 200 mm isohyet, the line that marks the limit of the area considered capable of sustaining rainfed agriculture (Oates 1968: 2–3, ﬁg. 1; Dillemann 1962: 64). Because of this, Edessa was (and modern Urfa is) the center of a cereal-producing district. Past the city (in preByzantine times, through it) ran the river Skirtos or Dais.an, the modern Kara Koyun, on its way to join with the Balikh, a tributary of the Euphrates. This stream was known chieﬂy for the extreme ﬂuctuations in volume it underwent throughout the course of the year. A steady source of water, however, also existed in the perennial springs that fed the pools at the southern end of town, which still, as in antiquity, teem with shoals of sacred ﬁsh (Itinerarium Egeriae [ed. Arce, 1980]: 19.7). It is hard to imagine that an oasis such as this could have failed to serve as at least a minor stopping-point on overland routes through the region. By the mid-fourth century, at the latest, it was more than that. In September of 354 ce a Persian general, ﬁnding Mesopotamia too well guarded, tried the unusual step of invading the Roman Empire via Edessa’s province, Osrhoene – a move by which, if successful, he would have caused widespread devastation, ‘fulminis modo’ (Amm. Marc. 14.3.2). What made the maneuver particularly dangerous was the fact that this was the time of the annual fair at Batnae southwest of Edessa, coinciding with a religious festival, to which traders and people of all stations came, ‘ad commercanda quae Indi mittunt et Seres, aliaque plurima vehi terra marique consueta’ (Amm. Marc. 14.3.4). 14

THE EARLIEST EDESSA

Batnae was the best place to hold this trade fair because, as Ammianus notes, it lay within easy reach of the Euphrates. In this period, however, Edessa was the administrative center of this district, and certainly beneﬁted – probably more than the little municipality of Batnae itself – from the trade in the goods which India and Seres (China) ‘sent.’ In another account of Late Roman date, the cities of Mesopotamia – that is, the Roman province of that name – and of Osrhoene are described as ‘multas et varias,’ and Edessa, along with Nisibis to the east, as containing men who were both generally excellent and adept at business: ‘Sunt ergo Nisibis et Edessa, quae in omnibus viros habent optimos et in negotio valde acutos . . .’ (Expositio totius mundi et gentium, 22 [Geog. Graec. Min. 2.517, 22]). This account (the ‘Liber Junioris Philosophi’) describes the inhabitants of both cities as ‘et divites et omnibus bonis ornati’ due to their merchants’ industry in obtaining goods coming from Persia (and, presumably, beyond) in exchange for the products of the Roman Empire – except for iron and bronze, ‘quia non licet hostibus dare aeramen aut ferrum’ (ibid.). There is good reason to believe that the author of this account may have confused Edessa with another city, Amida to the northeast.10 Despite this confusion, however, he gives us good evidence for Edessa’s prosperity. In this account, which is presented in the form of an itinerary from East to West, Nisibis appears as the ﬁrst city after entering Roman territory (‘nostra terra’). Hence the text must have been written before the treaty of 363 which followed the death of Julian and ceded that outpost to the Persians (Amm. Marc. 25.7.9–11). Another thing is clear: despite some limitations on items of military value, the hostility between the two empires did not prevent their peoples from engaging in mutually beneﬁcial commerce. After 363, with the two empires at peace though in a state of cold war, trade surely continued to exchange hands. With Nisibis in Persian hands, Amida and Edessa were now the ﬁrst cities in Roman territory, Amida on the route toward the heart of Armenia, and Edessa on the more difﬁcult, but more direct route to the Euphrates and the Province of Syria beyond. Therefore Edessa probably grew even more prosperous during the last third of the fourth century. These sources, though of relatively late date, prove that Edessa was in a position to prosper on the basis of long-distance trade. It remains unclear, however, to what degree it took advantage of that position before the fourth century.11 As we have seen, there is little certitude as to the site’s identity in Ancient Near Eastern cuneiform 15

ROMAN EDESSA

records. If the proposed identiﬁcation with DM is correct, it does not seem to have been a major stopping point. Earlier accounts of the route from West to East, such as that of Isidore of Charax near the beginning of the Christian era, stress the route down the Euphrates to Seleucia–Ctesiphon and thence eastwards (Geog. Graec. Min. 1.244–54). In the itinerary related by Isidore, the traveler crosses the Euphrates at Zeugma and heads overland to the Balikh, then travels down that river’s valley to rejoin the Euphrates. Edessa does not play a major role, but a reference to its oasis and Hellenistic fortiﬁcations may lie in the mention of Μανοοορρα Αυρθ [or Αυρρ], χρωµα κα κρνη, situated to the ‘right’ of another stronghold, Κορα!α "ν Βατα´νη (ibid. 246). The position of this site in the itinerary, after passing through Anthemusia (the territory of Batnae) and before reaching the Balikh, ﬁts Edessa very approximately. This led Gutschmid to emend the text, reading Μα´ννου Ορροα!ου Αυρ for Μανοοορρα Αυρθ and taking it as a reference to ‘Mannos of Orraeos’ or Ma nu III Saphlul, whose reign over Edessa he dated between the years 23 and 4 bce. The traveler heading east from Batnae, however, would have to turn left – not right – to reach Edessa. Μανοοορρα Αυρθ is probably elsewhere, and may even be a conﬂation of two different toponyms (Dillemann 1962: 178–80). Edessa, then, is not mentioned by this important early source on the Eastern trade routes. Strabo’s Geography also describes a route across the desert after the crossing at Zeugma, mentioning explicitly that the reason for leaving the river was to avoid the heavy exactions of the communities along the valley (Strabo 16.1.27). In the route described by Strabo, however, the merchant heads straight for the Khabur from Anthemusia, and thence across uninhabited areas to ‘Scenae’ on the borders of Babylonia, with the aid of camel-driving Scenitae or nomads (ibid.). This route apparently involved no stops at Edessa or any other community between Anthemusia and Babylonia. Although Strabo mentions Edessa, he confuses it with Hierapolis/ Bambyce on the Syrian side of the river. Whether or not the identiﬁcation of Edessa with a watering-place in Isidore’s account is correct, and keeping in mind that information on these points is very sparse, Edessa does not seem to have derived much proﬁt from the trade route at this early date. For a source indicating that trade in silk or other items of long-distance commerce had a signiﬁcant role in the city, we must look to the ﬁfthcentury Syriac Doctrine of Addai. This document mentions the role played in the evangelization of Edessa by ‘the Jews . . . who traded 16

THE EARLIEST EDESSA

in silk’ (trans. Howard, 1981: 69; Syriac 68). In the same narrative, one of the ﬁrst Edessan disciples is Aggai, ‘who made regal silks and tiaras’ (ibid. 71; Syriac 70). The events in this tale, however, if they have any historical validity at all, cannot be earlier than about 200 ce.12 The appearance of silk as a commodity and a luxury item may possibly indicate that during the centuries since Isidore’s account was written, a growing portion of the trafﬁc to the East had begun to use the overland route via Edessa to Nisibis. From here it either passed down the Khabur to rejoin the Euphrates route, or continued eastwards to the Tigris and its crossings. Again, we are very poorly informed; yet the role played by Edessa and its rulers in the events of the second century and later suggest that it had managed to reach a new level of prosperity. This prosperity and the greater importance of the city, both as a trading station and militarily, may be reﬂected in the Antonine Itinerary, which probably reﬂects the topography of the empire at the time of the Emperor Caracalla (Itinerarium Antonini: 184–92). Edessa (‘Edissa’) is the only major stopping-point east of the Euphrates in the description of the eastern frontier road network. Although Carrhae appears as well, it is decidedly secondary in rank, being mentioned only once, as the endpoint of one route (to Hierapolis – ibid. 192). By contrast, six roads to Edessa are listed, from as far away as Nicopolis in the heart of Syria. Caracalla himself visited the region in 217, during the course of his campaign against the Parthians, and was assassinated while traveling from Edessa to Carrhae to visit the ancient temple of the moon-god Sin (Dio 78.5.4–5; Hdn. 4.13.5; HA Caracalla 6.6, 7.1–2). To judge from the Itinerary, this was nothing but a side-trip to a city that had ﬁnally been eclipsed in a long-running rivalry with Edessa. The Peutinger Tablet, one of the most famous of all Roman maps, represents the empire at a much later date.13 In the Mesopotamian portion, however, it places the end of the area of Roman control (‘Areae ﬁnes Romanorum, Fines Exercitus Syriaticae et commertium Barbarorum’) west of the Euphrates, and it preserves at least one city – Hatra – that was long abandoned by the time the map was apparently drawn up in the fourth century (Tabula Peutingeriana: p. X). Hence it seems likely that in this area the Tablet’s information derives from an itinerary similar to that of Isidore of Charax describing the Parthian road network, or from a combination of such sources (Dillemann 1962: 134–6). On this map Edessa appears to be a crossroads of similar status to Rhesaina, Nisibis and Singhara – each of them being represented by a graphic depiction of its city walls. 17

ROMAN EDESSA

The main route from Zeugma to Rhesaina and eastward passes south of Edessa, through Carrhae; but another route seems to connect Edessa to Rhesaina (here labeled Fons Scaborae) via Carrhae.14 It is difﬁcult to assess the signiﬁcance of this, since the map is far from exact topographically, and some places appear more than once on different routes (e.g. Charra/Charris, Singhara/Sirgora). It would seem, however, to conﬁrm that Edessa was growing in importance during the ﬁrst two centuries ce. Strategic importance If it is correct that Edessa gained in prosperity and importance during the Common era – whether this be due to an increase in commercial trafﬁc or to any other factor – this makes all the more compelling the question of why it was so badly overshadowed by H . arran/Carrhae before this, to the point that it cannot be securely identiﬁed in any pre-Hellenistic record. Here again, topography may provide a clue. To the visitor walking southward down one of Urfa’s main streets, the ‘Citadel’ of Edessa presents an imposing prospect. The nearly sheer north face of the fortress, with its remnants of Crusader fortiﬁcations, seems unassailable. Yet after one mounts the Citadel and surveys the lie of the land, the true picture becomes plain: this is not a defensible position. Far from being an isolated eminence, such as the Athenian Acropolis, Edessa’s Citadel is more of a rock outcropping joined on its southern and western sides by hills that slope more gently down to the level of the plain. A narrow moat has been cut into the rock on these sides, but this does not offer anything approaching the defensive advantage of the northern cliff face. Moreover, by contrast with the well-watered oasis at the foot of the hill, the fortress is without an internal water supply. What little archaeological excavation has been done has not yet produced any evidence of cisterns or similar systems for enabling the fortress to withstand long-term siege, although it would presumably have been useful as a base for a last-stand defense. By contrast with Urfa’s topography, H . arran and other ancient cities of the Mesopotamian plain were usually situated on naturally defensible eminences or mounds produced by centuries of habitation – the classic Middle Eastern tell – and encircled by walls, although water supply arrangements varied from site to site. Edessa did not develop in this manner in the early period, even though the site certainly held the potential for such urban growth. The reasons are 18

THE EARLIEST EDESSA

beyond retrieval at present, but may have to do with the domination of the area by H . arran, other political rivalries, or with a ﬂuctuation in patterns of trade and travel. Whatever the reason, it would seem that the turning point in Edessa’s urban development came with the walling of the lower city and the inclusion within its defensible boundaries of the springs, at the time of the city’s Seleucid colonization. The Citadel was now truly an inner keep, at the southwest corner of a fortiﬁcation system which would remain basically unchanged into the Byzantine period.15 The importance of the springs as an internal water source is apparent in the story told to Egeria, the Western pilgrim who visited Edessa in the course of her travels in the late fourth century (Gingras 1970: 79): when they [the Persian invaders, at some unspeciﬁed date] saw that they would be unable to penetrate into the city, they decided to kill those within by cutting off the water supply. . . . On that day and at the very hour when the Persians diverted the water supply, these springs, which you see here, burst forth immediately, by the command of God. From that day up to the present time these springs have continued to be here through the grace of God. However, the very water which the Persians had diverted dried up in that hour so completely that those who were besieging the city did not have one day’s supply of water. Although this story obviously cannot be true – a conclusion supported not only by common sense but by sources of the period of the monarchy that mention the very same springs – it points up the defensive advantages of Edessa’s situation.16 By Egeria’s time a promise of invulnerability for the city had been incorporated into the myth of Jesus’s letter to King Abgar, possibly because of the city’s success in withstanding repeated attacks.17 It is true that, at least as early as the beginning of the fourth century, Edessa’s growth and the construction of facilities such as the city baths called for aqueducts to supplement the water supply.18 However, the springs were themselves ample to support a sizable population in time of need. Supported by such resources, before Rome arrived on the Near Eastern scene, the Seleucid colony of Edessa had already become an autonomous kingdom, or ‘toparchy,’ under the rulers of the Abgarid 19

ROMAN EDESSA

line. In this guise it took part in the give-and-take between Rome and Parthia for the next 200 years. Yet throughout this period the entire northern Mesopotamian region stayed more or less within the Parthian sphere, with the de facto border between the two empires at the Euphrates (Strabo 11.9.2; Dio 40.14; Pliny HN 5.88; Hdn. 4.10). This did not, of course, prevent all cross-border communications, as we have seen in the case of Pompey’s general Afranius and the supposed ‘betrayal’ of Crassus by a King Abgar. The position of Osrhoene on one of the northward routes out of Mesopotamia meant that Edessa also played a part in the history of Armenia and in the conﬂicts over it, and that Edessa’s fate was sometimes tied to that of its northern neighbor. Control of the region surrounding Edessa within the bend of the Euphrates was an important factor in this long-running conﬂict; when the Roman general Corbulo negotiated a Parthian withdrawal from Armenia in 62 ce, the price was an evacuation of positions east of the Euphrates to leave the border between the two empires, ut olim, at the river (Tac. Ann. 15.17). The relative positions of Osrhoene and Armenia ﬁgured, as well, in the earlier episode of Abgar’s supposed betrayal of the Armenian pretender Meherdates. It seems that Rome expected its favorite candidate for the Armenian throne to proceed apace, directly north from Edessa through the Armenian Gates on the Euphrates – even though this was a more difﬁcult and dangerous route than the one through Adiabene chosen by Abgar. The slowness of the Adiabene route meant that Meherdates arrived on the Armenian scene too late, and Abgar was blamed for sabotaging the project (Tac. Ann. 12.12, 12.14).19 When Rome gained a longer-lived foothold east of the Euphrates, ﬁrst under the Emperor Lucius Verus (161–9) and again under Septimius Severus (193–211), this region increased tremendously in strategic importance. When Severus moved the arena of conﬂict eastward with the creation of two new provinces, his professed aim was to provide a ‘bulwark’ (πρ βολο) for Syria (Dio 75.3.2–3). It may have been just as important, however, to cut off Armenia from Parthia, and control of Osrhoene was clearly a necessity for this. Regardless of the reasons for this move, the conﬂict remained focused on northern Mesopotamia until the Islamic Conquest more than 400 years later. Edessa did not, perhaps, rival Nisibis in its importance for the control of north Mesopotamia. It was impossible, however, for either of the parties to the conﬂict to ignore Edessa, well situated and well watered as it was. Edessa’s troublemaking potential was demon20

THE EARLIEST EDESSA

strated in 115, when it and other nearby cities rose in rebellion after initially submitting to the invading armies of the Emperor Trajan. The revolt cut off the emperor – who had advanced as far as Ctesiphon – from his home base and placed him in extreme danger; Trajan was forced to send his general Lusius Quietus to reduce the town (Dio 68.29.4). It was in Nisibis, not in Edessa, that Roman troops were apparently left, perhaps to guard against Parthian advances, after the invasion of Mesopotamia by Lucius Verus in the 160s (Dio 75.1.2). But Edessa’s position controlling the western portion of northern Mesopotamia meant that these troops again had no direct access to Syria, the nearest Roman province – a circumstance of which Edessa and its allies tried to take advantage by attacking Nisibis in 194 (Dio 75.1.2–3). When the army of Severus, in response, invaded Mesopotamia and marched to Nisibis, the soldiers suffered extreme hardship due to lack of water, underlining the importance of the Edessan oasis (Dio 75.2.2). Edessa would continue to grow in importance as an advance base of operations against Persia – a role in which we frequently ﬁnd it during the fourth century (Amm. Marc. 18.7.7, 20.11.4, 21.7.7, 21.13.1). Perhaps even more importantly, however, it was a stronghold which the Persians themselves could not afford to leave at their rear when they attempted to strike at the rich province of Syria. This was apparent at least as early as 260, the date of Shapur I’s victory over and capture of the Emperor Valerian. The climactic battle of this campaign, according to the inscription of Shapur at Naksh-i-Rustam, took place under the walls of the besieged Edessa (Greek text, Maricq 1965: 52–5, ll. 19–20, 24). According to Byzantine sources, even after this victory and the ensuing devastation of the region, Shapur was obliged to pay off the garrison at Edessa to ensure his safe return home (Petrus Patricius, frag. 11 [FHG 4, p. 187]). This means either that he had captured the city but neglected to arrange sufﬁciently for its secure occupation, or that he had not bothered to see the siege through to completion after his victory over Valerian. In either case, the threat posed to his safe withdrawal illustrates Edessa’s strategic importance.20 One hundred years later, Persian successes at Singhara and elsewhere were squandered, said the turncoat Antoninus, when the victors failed to follow them up by capturing Edessa: iam inde quadragesimi anni memoriam replicabat, post bellorum assiduos casus, et maxime apud Hileiam et Singaram, 21

ROMAN EDESSA

ubi acerrima illa nocturna concertatione pugnatum est . . . Persas nondum Edessam nec pontes Euphratis tetigisse. . . . (Amm. Marc. 18.5.7 [348 ce]). Thus, while control of Edessa was critical to the Roman hold on the region, it was no less a key to any westward successes by Rome’s Persian rivals. That Rome was able to pursue its ambitions east of the Euphrates for so long is in large part due to the continued loyalty of the leaders and people of Edessa – a loyalty which may, ironically, have been sealed by the emergence of the powerful Sassanid regime in 226. The Asian empire’s Zoroastrian rulers could be even less tolerant of Christianity than the emperors, and even at this early date Edessa had a sizable Christian community. Allegiance to Rome may have seemed prudent to its people and, before the end of the monarchy around 240, to its rulers – even at times when Rome’s rivals had the upper hand in this eastern corner of the empire. ‘Osrhoene’: kingdom and province One gauge of Edessa’s importance is the extent of the territory it controlled. In general this can be equated with Osrhoene or Orrhoene (from Orhai, the Syriac name of Edessa), which Procopius deﬁnes as ‘ Εδεσσα . . . ξ'ν τοι αµφ ατ)ν χωρ!οι’ (Procop. Wars 1.17.24).21 This author’s account of Justinian’s building work in Osrhoene mentions by name only Carrhae, Callinicum and Batnae in addition to Edessa (Aed. 2.7.1, 2.7.18). From this, and from the locations listed in the Notitia Dignitatum, it appears that the Late Roman province extended from the Euphrates eastwards as far as, or almost as far as, Rhesaina/Theodosiopolis at the headwaters of the Khabur, and was bordered on the east by the Khabur until its junction with the Euphrates at Circesium (Not. Dig. Or. 35).22 At an earlier date, when Edessa was ruled by the toparchs of the Abgarid line – with military resources that could not compare with those of the Later Roman Empire – the area under its control may well have been smaller. Nor do the Diocletianic provinces of Mesopotamia and Osrhoene necessarily coincide with the similarly named ones organized by Septimius Severus when he annexed the region. The elder Pliny, in his description of Mesopotamia, introduces Osrhoene, or ‘Osrhoenian Arabia,’ as the area lying on the left bank of the Euphrates after the river ﬂows through Armenia (HN 5.20.2: ‘Arabiam inde laeva Orrhoeon dictam regionem, trischoena mensura, dextraque Commagenem [sc. Euphrates] disterminat’). 22

THE EARLIEST EDESSA

This district, he says, contained both Edessa and Carrhae – ‘Carrhas Crassi clade nobiles’ (HN 5.21.1). Pliny’s value for determining the precise limits of Osrhoene, however, is diminished by his statement that to the east, the ‘Arabes Orroei’ extended as far as the Tigris, which in turn separated them from Adiabene (HN 6.31). This statement makes the Osrhoenians inhabit all of upper Mesopotamia, without any mention of the intervening territories of Rhesaina, Nisibis and Singhara. Since Pliny elsewhere distinguishes between city-dwelling or sedentarizing ‘Arabes’ such as the Orroei or Praetavi and the nomadic Scenitae – ‘ipsi vagi sed a tabernaculis cognominati’ (HN 6.32) – it seems unlikely that he pictured the people of Osrhoene as wandering throughout the steppe to the east, although some relationship of political or military hegemony by Edessa in part of this area is not out of the question. Why Pliny should have enlarged the territory of Osrhoene so greatly in his account remains a question to be answered, but it seems possible that he simply confused its inhabitants with other Arabs of the region. Both Nisibis and Batnae seem to have remained independent of Edessa at least until the invasion of Trajan in 114 ce, when the emperor conquered them both and received the title Parthicus (Dio 68.23). Singhara, too, was independent, being ruled by a monarch named Mannos (Ma nu – Dio 68.21.22). Although Pliny’s account shows that Edessa and Carrhae were considered to be closely related to each other, this does not justify the inclusion of Carrhae as one of the oppida of Osrhoene. The two Seleucid colonies remained independent of each other and developed along different lines socially and politically. Edessa’s population was quite mixed at an early stage, and its original constitution gave way to monarchic rule by the Abgars in the second century, while the inhabitants of Carrhae (or the leading members of the community) were apparently of Macedonian descent and sympathetic to Rome at the time of Pompey’s invasion, 65 bce (Dio 37.5.5: Καρραιοι, Μακεδ νων τε αποικοι *ντε κα "νταυθα´ που ο+κουντε). Therefore, if the ‘Arabes Orroei’ were as widespread as Pliny claims, and if their territory was identical with ‘Osrhoene,’ various communities remained independent of Edessa despite being surrounded by its territory. Surely it is better to assume that Pliny has indeed confused the various peoples populating northern Mesopotamia (who may, admittedly, have been closely related), and to give to Osrhoene its natural meaning, that is, the χ,ρα or immediate area of Orhai/Edessa.23 This, however, is more or less a negative 23

ROMAN EDESSA

conclusion. A small amount of evidence exists enabling us to delineate the area of Edessan control more clearly. The Chronicle of Edessa, one of the very few Syriac historical records surviving from the time of the Edessan monarchy, reports that as one of the measures taken to help Edessa recover from the disastrous ﬂood of 201, the magnanimous King Abgar VIII remitted taxes on the ﬂooded ‘villages and ﬁelds’ for a period of ﬁve years (CSCO Syr. ser. 3, vol. 4, p. 4; Syriac, p. 3). This information, while suggestive, is not very useful since it does not name any of the communities concerned. From epigraphy, however, it appears that on the west, Edessan authority or inﬂuence extended to the Euphrates from an early date. The earliest known Syriac inscription, dating to 6 ce, comes from Birecik (Birtha/Makedonopolis) on the Euphrates. It is the dedication of a tomb by ZRBYN, who identiﬁes himself as the ruler (shallita) of Birtha and tutor of WYDNT, son of Ma nu bar Ma nu and possibly a member of the Edessan royal family (Drijvers 1972: no. 1; Maricq 1965: 127–39; Segal 1970: 23, n. 3). Some decades later, but still in the ﬁrst century, a certain Ma nu erected a tomb tower at Serrin on the eastern bank of the Euphrates, placing upon it an inscription, again in Syriac, that bore the date 385 of the Seleucid era (73 ce – Drijvers 1972: no. 2; Maricq 1965: 134–6; Segal 1970: 23, n. 4). Neither of these inscriptions contains an explicit reference to Edessa or its authority. Their use of the Edessan script and dialect, however, is interesting, and the fact that Serrin lies well south of Batnae – at nearly the same latitude as Hierapolis in Syria – perhaps indicates a broadening of Edessan inﬂuence along the river. When Abgar of Edessa (Abgar V?) became involved in the Roman attempt to install Meherdates on the Armenian throne in 49 ce, the king met Meherdates at the river crossing, which may help to conﬁrm this impression of an extension of Edessan power as far as the river and some way along its banks (Tac. Ann. 12.12, 12.14). To the north, Osrhoene extended as far as the mountains of Armenia and the ‘Armenian Gates,’ where those mountains come closest to the Euphrates. In the south, however, Edessan authority was limited by the territories of Carrhae and of Batnae/Anthemusia (the latter was ruled by a phylarch named Sporaces at the time of Trajan’s invasion – Dio 68.21). To the east, the natural limit of its χ,ρα was the Tektek Dagh, a spur extending southward from the Tur Abdin mountain range some 30 kilometers away. Yet it is on this mountain that important evidence is found indicating that, at least in the second century ce, Edessa had extended its authority 24

THE EARLIEST EDESSA

east of the mountains. Sumatar Harabesi on the Tektek Dagh is the site of a spring around which shepherds of the area gather to water their ﬂocks, and where a small village has developed in recent years (Drijvers 1980: 122).24 It is also the site of a group of ancient monuments that are important for the history and religion of Edessa. Among these is a series of rock-cut inscriptions, two of which are inscribed by an ofﬁcial, the shallita de- Arab (‘ruler of Arab’ or Arabarchos), in honor of ‘the King.’ Some of the inscriptions are dated to 165, a time just before the replacement of Edessa’s proParthian King Wael by a monarch more favorable to Roman interests (Segal 1970: 56–7; Drijvers 1972: nos 23, 24). A parallel has been drawn between the ofﬁcial mentioned here and Mettolbaesos, the Αραβα´ρχο named in a parchment of 121 found at Dura-Europos, who in addition to holding that title was the strategos and tax-collector of ‘Mesopotamia and Parapotamia’ (Welles, Fink and Gilliam 1959: no. 20, l. 5). It would seem that this Arabarchos was responsible for protecting and administering the territories inhabited by the sedentary or seminomadic peoples known as Araboi.25 The ofﬁcials mentioned in the Sumatar inscriptions may or may not have had a similar role, but we do have some indication of the area of their authority. East of the Tektek was a region, extending at least as far as Tella northwest of Resaina (later renamed Constantia – modern Virans¸ehir), and known as Arab or Arob; it was here, seemingly, that the shallita de- Arab held sway. Based on the parallel with the Αραβα´ρχο of Dura Parchment 20 – who bears Parthian court titles – and the fact that, until 165, this entire region was under Parthian hegemony, one might expect the ofﬁcials in the Sumatar inscriptions to be Parthian ones as well. The ‘king’ to whom the inscriptions refer is not named, and could thus be the Parthian ‘king of kings’ (cf. Dura Parchment 20, l. 1: Βασιλεοντο βασιλ-ω βασιλ-ων Αρσα´κου εεργ-του, κτλ.). Yet the fact that the inscriptions are in Syriac argues for an association with the Edessan king. An undated inscription in a nearby cave refers to an Aurelius H . afsai ( WRYLWS H . FSY), identiﬁed as a freedman of ‘Antoninus’ ([ N]TWNYNS) and as ‘lord and benefactor’ of the dedicator, who is another ‘Ruler of Arab’ (Drijvers 1972: no. 7; cf. idem 1980: 130–1). H . afsai is a name that will recur later in Edessan history, and its occurrence here strengthens the connection with Edessa. The fact that this individual bears a Roman nomen is in itself signiﬁcant, and as we shall see, the inscriptions of Sumatar Harabesi play an important role in puzzling out the political history of Edessa 25

ROMAN EDESSA

in the second century. For now, however, we are concerned only with the geographical question. Aurelius H . afsai strengthens the connection between Edessa on the one hand and Sumatar Harabesi and the Arab on the other.26 The precise extent of this district is a vexed question; H. Pognon, who ﬁrst published the Sumatar inscriptions, saw it as extending all the way to the Tigris (Pognon 1907: 34–5; already contested by Theodor Nöldeke [Nöldeke 1908: 153]). The assumption that the Arabarchoi named here exerted their authority so far east runs into the same objections as does Pliny’s imprecise statement about the extent of the ‘Arabes Orroei’.27 Questions may remain, but as far as Edessan territory is concerned, the most conservative approach is to conclude that Arab designates a fairly restricted area around Tella and Rhesaina (Dillemann 1962: 75–6). Hence, by around the middle of the second century ce, Edessan inﬂuence extended as far west as the Euphrates and as far east as Tella/Virans¸ehir, and for some distance beyond in the illdeﬁned steppe between Tella and Nisibis. This conclusion has the advantage of avoiding giving the kings of Edessa a domain far larger than the Late Roman province of Osrhoene, which in itself would be improbable. Rhesaina and the line of the lower Khabur probably represented the maximum eastward extent of Osrhoenian authority under normal circumstances. Drijvers’s statement that the shallita de- Arab ‘held sway over the desert area east of Edessa as far as the Tigris’ depends upon an extensive interpretation of the word Arab; it is hardly justiﬁed by his reference to the Thesaurus Syriacus’ deﬁnition of RB as ‘regio Mesopotamiae in vicinia Edessae’ (cf. Drijvers 1980: 130, with n. 21). This picture of the eastward extent of Edessan authority appears to ﬁt the circumstances of 194, when Edessa joined forces with Adiabene in an attack on Nisibis (Dio 75.1.2–3). Plainly the fortress on the ‘Mygdonius’ (the central branch of the upper Khabur) was a tempting target for a pincer movement, and despite the presence there of Roman troops, there was little or no impediment to the Edessan forces attacking from the west. King Abgar was to ﬁnd his domains substantially reduced by Septimius Severus’s settlement of the region after his Parthian victories. The coins of the Roman municipalities established at surrounding cities, and archaeological evidence of the new western border of Abgar’s kingdom, testify to the restriction of Edessan territory in the ensuing period.28 Edessa itself was soon to become one of the cities of the Roman province of Osrhoene; yet it still, seemingly, had 26

THE EARLIEST EDESSA

some sort of authority over at least one outlying district in or near Batnae. It is in this connection that the recently discovered documents of the reign of Gordian III begin to provide us with new information (Feissel and Gascou 1989; Teixidor 1990). These two Syriac documents, known as A and B, were drawn up under Edessan legal authority, but not in Edessa itself. Rather, Document A was transacted in 240 – under King Abgar X – in a place called ‘New Haikla-Karka da-S.ida,’ while Document B was drawn up in 242 in ‘Marcopolis Thera’ (Teixidor 1990: 147, 154). Both documents concern the legal affairs of one Worod, and one might assume that they were both transacted in the same place, referred to ﬁrst by its Syriac, and then by its Greek name.29 Haikla da-S.ida may be the same as Charax Sidou, which Isidore of Charax gives as a name for Anthemusia, one of the ﬁrst stations on the eastward journey after the Euphrates (Geog. Graec. Min. 245: Χα´ραξ Σ!δου, 2π3 δ4 5 Ελλνων Ανθεµουσια` π λι). Anthemusia was long ago identiﬁed as Batnae, the modern Eski Serug southwest of Urfa (Regling 1901–2: 450; Dillemann 1962: 178–9). But since Marcopolis and Batnae are listed as separate bishoprics in the Notitia Antiochena, it is not possible simply to state that these are all names for the same place, namely Batnae. Marcopolis and Batnae may, however, have lain close to each other within Anthemusia, the ‘basin of Serug’ (Dillemann 1962: 102). Ecclesiastical history can provide some help in delimiting Osrhoene. Eusebius of Caesarea, in a passage introduced by the succession of Victor to the episcopate of Rome in the tenth year of Commodus (189 ce), describes the participation of the bishops of ‘Osrhoene and the cities there’ (τ)ν Οσροην)ν κα τα` "κεισε π λει) in a controversy over the proper date of the Easter celebration (Eccl. Hist. 5.23.4). By the 190s, therefore, Osrhoene was conceived of – at least in episcopal administration – as encompassing more than just Edessa and its immediate territory.30 To sum up, Edessa’s position at a crossing-point between the legendary East–West ‘Silk Road’ and routes connecting Armenia with southern Mesopotamia gave it great potential importance, enhanced by its ample sources of groundwater. This economic potential began to be realized probably by the beginning of the third century ce, possibly much earlier. It was certainly recognized in the fourth century, when Ammianus describes the fair at Batnae dealing in silks and other Eastern trade goods. Edessa’s position also gave it great strategic importance in the back-and-forth conﬂict between East and West; neither Rome nor its Parthian and Persian 27

ROMAN EDESSA

adversaries could afford to ignore such a well founded and well watered stronghold in the course of their mutual raiding. It was more than an isolated fort, however. Our wide-ranging discussion of the extent of Edessa’s inﬂuence has shown that already under the monarchy, almost certainly by the beginning of the Common Era (the Birecik and Serrin inscriptions), Edessan inﬂuence extended to the banks of the Euphrates in the west, and a good way southwards along the river. To the east, the city’s own territory was limited by the Tektek Dagh range, but its rulers sought, at times successfully, to extend their sway as far beyond this as possible in the ill-deﬁned Arab region. Our only explicit evidence of this expansionist drive comes from the last third of the second century ce, ﬁrst in the form of the Sumatar inscriptions and then in Abgar VIII’s opportunistic moves against Nisibis. It may well be that it was not until then that Edessa was prosperous and powerful enough to ﬂex its muscles in this way; if so, this might reﬂect the fact that the lucrative Eastern trade began to pass more frequently through Osrhoene at some point in the second century. There is no way to be certain of the extent of the Kingdom of Osrhoene before it became a Roman province. The presence of H . arran and Batnae to the south could have set limits on expansion in this direction, and if so, the kings’ authority should be seen as reaching no farther than 25 miles or so to the south. The Serrin inscription, however, suggests a broader area of inﬂuence – extending well past Batnae. It is possible, indeed, that at times under the kings, ‘Osrhoene’ can be envisaged as occupying the entire region limited on the west and south by the Euphrates and on the east by the Khabur, as it did in Late Antiquity. The documents from the period of the monarchy – the ﬂood narrative in the Chronicle of Edessa and ‘Document A’ of the reign of Abgar X – give valuable additional data. They show, in the one case, a king ruling a sizable, prosperous town and in control of the revenues of outlying villages and regions, and in the other case, a king still – on the verge of the monarchy’s extinction – holding sway over at least one community, ‘Haikla da-S.ida,’ outside the immediate area of Edessa. When the Roman emperors won the tug-of-war over Edessa and its territories, they obtained a sizable and wealthy addition to the empire, with a vibrant and multifaceted cultural identity.

28

2 THE COMING OF ROME

Rome’s preoccupation with the security of its Armenian client kingdom and the integrity of the Euphrates frontier may have given the rulers of Edessa hope that they could continue to conciliate both of the major powers between which they found themselves. If so, the situation changed dramatically in the second century ce, with the beginning of serious and sustained Roman involvement beyond the Euphrates under the Emperor Trajan (97–117). This emperor had already shown his willingness to push beyond earlier limits, both in his Dacian wars and in the Middle East, with the acquisition of the Nabataean Kingdom and its transformation into the Province of Arabia.1 In Mesopotamia, he departed from earlier practice by attempting to retain conquered territory in the form of new provinces, although his successor Hadrian (117–138) thought better of the attempt. Fifty years after Trajan, the pattern of aggressive campaigning in the north and expansion in the east reappeared under the co-emperors Marcus Aurelius and Lucius Verus (161–169). The effects of Verus’s eastern campaign, however, were longer-lasting, both in Edessa and elsewhere in northern Mesopotamia. By the end of the century between Trajan and Septimius Severus, the king of Edessa was squarely within Roman clientela, and the groundwork was laid for the even ﬁrmer incorporation of his realm into the empire.

Abgar and Trajan Events at Edessa in the second century are reﬂected, if darkly, by the eighth-century Syriac Chronicle of Zuqnin.2 The chronicle dates important events in world history by several different schemes including the Seleucid era and the reigns of Roman emperors, but the main era employed is the ‘Year of Abraham.’ The section dealing 29

ROMAN EDESSA

with the period from creation up to 313 ce is based on Eusebius and a variety of Greek and Syriac sources, some of them apocryphal but others certainly reliable. For our purposes the most useful feature of the chronicle is that it preserves what appears to be nearly a full listing of the kings of Edessa and the lengths of their reigns. Although at least some of the earliest of these are almost certainly legendary, the list is much more accurate for the kings of Trajan’s day and later. A recent study showed that the dates in the chronicle can be reconciled with other historical records by adding 26 years to the dates given for the reigns of individual kings (Brock 1992b: 10–11). The chronicle’s account of events surrounding the arrival of Rome is a case in point. A crucial event in this section of the Zuqnin chronicle is the death of a King Abgar of Edessa, which it places in the Year 2106 (Chron. Zuq. 119/89). After this, according to the chronicler, ‘because of the rivalries for power,’ the people of Edessa were unable to agree on a leader for two years (ibid.). This interregnum was followed by the reign of ‘Ilour [? or Yalour] Pharnataspat.’ – or Ilour/Yalour and Pharnataspat. – lasting three years and ten months. The chronicle subsequently has Pharnataspat. ruling alone for ten more months, until the long reign of Ma nu the son of Izates (16 years, 8 months) begins (Chron. Zuq. 121–2/91). Synchronisms elsewhere in the text, and an explicit statement by the chronicler, show that the Year of Abraham is in fact the Seleucid year plus 1706.3 The date of the death of Abgar, therefore, is given here as Seleucid Year 401, or 89/90 ce. This is unlikely to be the correct date. The events which, according to the chronicle, followed the king’s death are much more believable if one assumes that they happened, not in 89/90, but 26 years later – during the ﬁrst major Roman invasion of Mesopotamia, under Trajan. A brief look at that invasion will provide justiﬁcation for the adjustment to the Zuqnin chronicle’s dating. More importantly, however, it will provide the background to the important events at Edessa during these years, and some indication of the city’s role in Roman imperial policy.

Trajan’s war The motives behind Trajan’s ‘Parthian War,’ and the exact sequence of events, have deﬁed historians’ inquiries despite renewed efforts in this century (Guey 1937; Lepper 1948; Lightfoot 1990). The only ancient authority to comment on the subject, the historian Cassius Dio, states that Trajan mounted his eastern campaign ‘out of a desire 30

THE COMING OF ROME

for glory’ (δ ξη "πιθυµ!6 – 48.7.1), although the publicly stated reason was the Armenian security crisis. The two forms of motivation are not mutually exclusive – a leader’s desire for glory can coexist with more ‘legitimate’ reasons for going to war; moreover, plans, policies and objectives may well change as events warrant. According to the narrative as it can be reconstructed from Dio and preserved fragments of Arrian (supplemented by Malalas and others), Trajan departed from Rome toward the end of 113 and wintered in Antioch. He campaigned ﬁrst in Armenia, where the Arsacid King Parthamasires had been behaving undiplomatically and defying Rome. Upon arrival there, the Roman legions swept across the country with little or no opposition (αµαχε!) and reached Elegeia, where Parthamasires was forced to submit and Rome took direct control of his territory (Dio 68.19.20; Arrian [Roos] 38–40). These activities occupied only a part of the campaigning season of 114, with the legions accomplishing their aims more easily than expected. As a result, Rome found itself making dispositions on the eastern borders of Armenia, beyond Lake Van, where the potentates of the region submitted to Roman diplomacy and had their thrones conﬁrmed or ‘assigned,’ in the expressive language of a coin issue: regna adsignata (BMC Roman Empire III 115, 120, 122; Dio 68.20). Thus, by the end of 114, Rome found open to it an eastern salient reaching far into the Parthian empire. Whether this was one of the original war aims, or whether it came about through ‘absentminded’ imperialism, cannot be decided. Certainly, however, it was an unstable situation. The Armenian salient was either a ﬂank exposed to Parthian attack or a base for an aggressive push southwards, depending on one’s point of view. It is here, perhaps, that we must allow for possible changing objectives. It is certainly possible that Trajan had, from the beginning, envisaged a sweeping effort to advance the eastern frontier by the incorporation of new Mesopotamian territory; on the other hand, this would have marked a departure from earlier Roman policy in a way that the settlement of the Armenian situation (a primary Roman aim since the Julio-Claudian era) did not. It is possible, too, that the advance into Mesopotamia was ﬁnancially motivated, so that the revenues from the newly incorporated area could help pay for the costs of holding Armenia. In any case, the ensuing push southwards into Mesopotamia, and on to Edessa via Nisibis, was a move that Trajan had not originally planned, or at least had not discussed with the Senate (Dio 68.21.1). Two things, however, made this an opportune time for a move in this direction: the perceived 31

ROMAN EDESSA

weakness of the Parthian ruler and the lack of resistance to Trajan’s Armenian activities, at the time of an apparent succession conﬂict within the Parthian kingdom (Dio 68.26). Faced with this opportunity, Trajan apparently conceived the further ambition of pushing on as far as Ctesiphon. Even without such an ambition, however, the situation demanded action, in order to remove the threat posed by the new salient. In the year that followed, Trajan’s conquests proceeded so far and so rapidly that the Senate was bewildered by the rapid-ﬁre succession of dispatches reporting the conquest of far-ﬂung peoples and territories of whom it had never heard – and which were surely not among the war’s initial public objectives (Dio 68.29.3). In moving so far and so fast, was Trajan (as Dio implies) subordinating rational strategy to the desire for personal glory? The emperor’s early victories, including the capture of Nisibis and Batnae, had earned him the victory-title of Parthicus. But he is reported to have expressed more pride in the title of Optimus Princeps, also ofﬁcially conﬁrmed around this time, ‘inasmuch as it referred rather to his character than to his arms’ (Dio 68.23). Such a preference does not mean that Trajan put no stock at all in notable military accomplishments, and the collection of titles he amassed speaks for itself in that regard: by his death he was styled imp xiii, avg ger(manicvs) dac(icvs) par(thicvs). His pride in being known as ‘the Best’ itself speaks to his concern for his public image, and Dio, who relays the story without comment, did not let it affect his judgment of the emperor’s desire for glory. The historian’s view of Trajan’s motivations should, however, be considered in light of the events of Dio’s own era, when the Emperors Septimius Severus (193–211), and later, his son Caracalla (211–17), mounted their own Parthian offensives. In a famous passage, Dio makes apparent his feeling that the Severan adventures had been more trouble than they were worth, and that, rather than increasing the safety of the empire, they imperiled it by adding to its foreign entanglements (Dio 75.3.2–3). He is hardly more ﬂattering when speaking of Caracalla’s activities. Although Trajan’s eastward push resulted in no permanent addition of territory to the empire, for Dio it may have set the pattern for the later campaigns. Both Trajan and Severus pushed on beyond northern Mesopotamia and both unsuccessfully besieged the fortress of Hatra, making the parallel seem even closer. Dio plainly saw Trajan’s campaign in a similar light to that of Severus, for he used nearly identical language to describe both emperors’ motives.4 32

THE COMING OF ROME

As we have seen, there were probably sound strategic reasons for the southwards movement of 115 ce. It is after this point, however, that it begins to look very much as if the voice of sound judgment was drowned out by that of ambition. After his northern successes, it would seem, Trajan felt impelled, rather than attempting to stabilize the situation, to capitalize on his victory over the Parthians. Leaving Edessa and northern Mesopotamia in his rear, he pushed on as far as Ctesiphon, which he sacked – conﬁrming for himself the title of Parthicus – and placed a client, Parthamaspates, on the Parthian throne (Dio 68.28.1; 68.30.3). He then took something of a tourist trip down the Tigris as far as the head of the Persian Gulf – where he engaged in wistful speculation, comparing his case to that of Alexander (Dio 68.28.3, 29.1). Inspired by the emperor’s absence on this voyage and very possibly by Parthian diplomacy, the peoples of northern Mesopotamia rose in a rebellion that was suppressed by Trajan’s general Lusius Quietus, with much destruction and loss of life at Edessa among other places (Dio 68.30.2). This preserved the region in Roman hands as the emperor headed home toward Rome. He never reached the capital, however – dying along the way in 117. Trajan’s heavy commitment of resources to the east had added several new provinces to the empire; they are listed by Byzantine chronographers as Armenia, Mesopotamia, Assyria and Arabia (though the acquisition of Arabia certainly occurred before the ‘Parthian War,’ and there is some question as to the genuineness of the Province of Assyria – Festus 15.30; Jer. Chron. 2.163– 5).5 Yet one of the ﬁrst actions of Trajan’s successor, the governor of Syria Publius Aelius Hadrianus (Hadrian, 117–38), was to abandon all the conquests beyond the Euphrates (Dio 68.33). Thus it was easy for the later historian to conclude that all Trajan’s accomplishments in the Parthian War had been vanities, no more substantial than the arch in his honor that stood for a brief time outside the gates of Dura-Europos (Baur [ed.] 1933: 56–65). For Edessa, however, the events of this period were far from inconsequential. Regnum adsignatum? The diplomacy that preceded the emperor’s arrival in the region of Edessa should be understood with the example in mind of the northern rulers who had submitted to Trajan without resistance, and whose acceptance into Roman alliance was commemorated on the regna adsignata coinage. Before arriving on the scene, Trajan 33

ROMAN EDESSA

made overtures to the principalities of the area including Anthemusia, Nisibis and Osrhoene, as well as ‘Arabia’ (apparently in this case to be identiﬁed with the territory of Singhara – Dio 68.21). Unlike their northern counterparts, however, the leaders of these communities declined to reciprocate by appearing in Trajan’s presence to have their positions ratiﬁed. The failure of one of them, Ma nu (Μα´ννο) of Singhara, to appear is cited as the occasion for Quietus to occupy that city and the neighboring territory without a battle (Dio 68.21.22). Abgar of Edessa was, at ﬁrst, no more eager than his neighbors to take up the double-edged offer of Roman friendship. Dio explicitly states that the Edessan ruler, aware of the sensitivity of the situation, sent gifts and a friendly message (δωρα δ4 δ) κα λ γου φιλ!ου) to Trajan at Antioch, the emperor’s ﬁrst stopping-point on his eastern campaign, but that he did not wish to antagonize the Parthians and therefore tried to play both sides ("πηµφοτ-ριζε – Dio 68.18). It was not until Roman forces were already in control of Nisibis, Singhara and Batnae that Abgar, perhaps realizing the hopelessness of his position, went over wholeheartedly to the Roman side and met the emperor on the road to his city (Dio 68.21). His apologies accepted, Abgar hosted the emperor at a banquet of friendship, during which the entertainment was provided by the king’s beautiful son performing a ‘barbaric’ dance. Despite his failure immediately to embrace the Roman standard, Abgar became a friend of Trajan and kept his throne until the alliance was overtaken by the events of the coming year. During this period, the Edessan king occupied a position analogous to that of the other rulers whose thrones had been ‘assigned’ during the previous year, with the important difference that his little kingdom was now totally surrounded by Romancontrolled territory. Although Edessa is explicitly attested as having taken part in the uprising that swept the region in Trajan’s absence during 116, there is no indication from the Western sources of Abgar’s role in these events. On the other hand, it should now be possible to reconcile the information about the rulers of Edessa as given by the Chronicle of Zuqnin with the outline of events as just given. The year in which Abgar died was not 89/90, as the chronicle would have it (Chron. Zuq. 119/89), but it may very well have been 26 years later, in 115/ 116 – for the events of 116 provide a very convincing context for the death of a King Abgar. Given the king’s ambivalence about falling in with Rome in the ﬁrst place, he may have been one of the leaders of the rebellion. In that case, he almost certainly either died in battle 34

THE COMING OF ROME

against the Roman forces or was dethroned and executed when Lusius Quietus suppressed the uprising. However, it may be thought unlikely that the anti-Roman party would have trusted him as a leader, if Dio’s report about his behavior during the visit of Trajan is true. It is more likely that Abgar died at the hands of the rebels, while defending his newfound alliance. As the aftermath showed, the Roman presence in the region was still strong enough to deal a decisive blow, and Abgar was most probably well aware of this fact. All we know for certain from the chronicle is that he died in the midst of some sort of power struggle. For the two years following the king’s death, Edessa was ‘without a ruler,’ according to the same document. If we have correctly located these events in time, this refers to nothing less than a period of Roman military occupation as the forces of Quietus worked to incorporate Abgar’s realm safely into Trajan’s new province of Mesopotamia, the experiment with client kingship having failed. One ﬁnal piece of evidence secures the link between these events and the activities of Trajan: it is the reign of Ilour/Yalour and Pharnataspat. that comes after the interregnum. Pharnataspat. is apparently an Aramaization of a Parthian name, and it coincides closely enough (as was seen already by Gutschmid) with Παρθαµασπα´τη or Παρθεµασπα´τη, who was the Roman candidate for the Parthian throne as named by Roman authors (Dio 68.33; Malalas 1.352, 357). It was not long after Trajan’s death and Hadrian’s renunciation of his eastern conquests that the Parthian kingdom rejected Parthamaspates, yet he apparently still held sway in this part of Mesopotamia. If the identiﬁcation of Parthamaspates with the individual in the Zuqnin Chronicle’s king-list is correct, the years of his reign at Edessa might indicate that despite Hadrian’s withdrawal, the kingdom maintained a close relationship with Rome. This is not entirely clear, however; as a member of Parthia’s ruling nobility he may have held a position at Edessa more like that of a Parthian satrap than that of a Roman client. After four years and eight months of this ambiguous situation, the chronicle reports the return to the throne of a local ruler in the person of ‘Ma nu the son of Izates’ (Chron. Zuq. 121–2/91). The patronymic of this individual suggests a possible link with the rulers of the neighboring kingdom of Adiabene. The chronicle at this point gives no indication that this accession represented any estrangement from the eastern empire. The kingdom’s recent experiences, however, meant that a pro-Roman faction had been either created or strengthened there – a situation that was to 35

ROMAN EDESSA

lead to a rift when Rome returned to the area ﬁfty years after Trajan’s expedition.

A ‘friend of Rome’ It is during this period that numismatic evidence begins to be of some use as a supplement to the historical record. Among the ﬁrst coins of Edessa are the silver issues of a King Ma nu who styles himself Φιλορ,µαιο, a ‘friend of the Romans’ (BMC Arabia, etc. 92–3 nos 5–9; rev. legend ΜΑΝΝΟΣ ΦΙΛΟΡΩΜΑ[ΙΟΣ]). These coins, all with Greek legends, bear obverse portraits of Marcus Aurelius and Lucius Verus, placing them in the joint reign of the two emperors (161–9). These are not, however, the very ﬁrst Edessan coins. They are preceded by bronze pieces which, on one side, bear the Syriac legend W L MLK – King Wael – and on the other either an image of a temple or the bust of a Parthian king (BMC Arabia, etc. 91–2 nos. 1–3). These, probably the oldest coins of Edessa, were therefore produced during a period of Parthian hegemony; the portrait bust has been identiﬁed as that of Vologaeses III (148–92).6 Again, the Chronicle of Zuqnin serves to place the Edessan rulers depicted on these coins within the succession list. It informs us that after the reigns of Pharnataspat. and of Ma nu the son of Izates, the latter’s son – also named Ma nu – began a 24-year reign (Chron. Zuq. 123/92). This event is placed in the Year of Abraham 2130, therefore some 24 years after the death of Abgar in 116. Ma nu is replaced by Wael, the son of Sahru, for a two-year reign beginning in 2154 of the Abrahamic era – this is the only Wael mentioned by the chronicle, and certainly the one who issued the coins in honor of the Parthian king with Syriac inscriptions W L MLK (Chron. Zuq. 125/ 94). The chronicle does not, however, say that Ma nu had died; rather, he ‘went over to Roman territory’ (BT RWMY – Chron. Zuq. 123/92). After a two-year interval he was back, returning from among the Romans to unseat Wael (Chron. Zuq. 125/94).7 All of this evidence combines to paint a compelling picture of conﬂict at Edessa between pro-Roman and pro-Parthian factions. Indeed the outline of events, though sketchy, makes it look as if Ma nu, by making overtures to the Romans, may himself have brought upon Edessa a Parthian intervention and the installation of Wael. Ma nu returned to the throne, according to the chronicle, 50 years after the death of Abgar, in Year of Abraham 2156: a date which can be corrected to 165/6. His return, therefore, was undoubtedly the result of the siege and capture of Edessa by Lucius Verus’s armies, which was seem36

THE COMING OF ROME

ingly brought about with the help of a faction within the city (Lucian Hist. Conscr. 22; Procopius Persian Wars 2.12; HA Verus 7).

The campaign of Lucius Verus It is understandable that the Parthian ruler may have wanted to make sure that his own man was on the throne of Edessa in 163. For at that time a renewed confrontation with Rome over Armenia was well under way.8 Early in the reign of Marcus Aurelius and Lucius Verus, Vologaeses had entered Armenia and installed his own choice, Pacorus, as its king; an attempt by Severianus, the governor of Cappadocia, to put the situation right ended abruptly in Severianus’s death and the destruction of his army at Elegia (Lucian Alexander 27; Hist. Conscr. 21, 24, 25). With the Armenian alliance thus in abeyance and the eastern frontier under Parthian threat, trouble began brewing within the borders of the empire itself, in the province of Syria. It was decided that the situation was serious enough to warrant the presence on the scene of an emperor, and Verus was the one chosen to head the armies (Dio 71.1.3; HA Marcus 8.9, Verus 5.8). Before his arrival in the East the new governor of Cappadocia, Statius Priscus, arrived, and took the Parthian campaign in hand (Lucian Hist. Conscr. 20; HA Verus 7.1; Dio 71.3.1). His capture of the Armenian capital of Artaxata in 163 turned the situation in the Romans’ favor, but Vologaeses was unwilling to listen to Roman peace overtures. As the emperor under whose auspices the Armenian successes were achieved, Lucius Verus was entitled to, and took, the victorytitle Armeniacus even though he had done no actual ﬁghting. This pattern of events was to continue, whereby Verus remained in the rear – usually at Antioch or the nearby pleasure-resort of Daphne – while his generals pursued the war. This is not to say that his presence did not contribute to Roman successes; he apparently took a hand in planning the campaign, in whipping the Syrian troops into shape and he certainly served to inspire them to greater efforts (Dio 71.2.2; Fronto Ad Verum Imp. 2.1). Yet later historians, and indeed his contemporaries, took critical note of his absence from the front. The Historia Augusta accuses the emperor of ‘lingering amid the debaucheries of Antioch and Daphne’ and taking credit for the work of his legates (HA Marcus 8.12). The same passage, however, shows how Verus’s reputation suffered unfairly by being linked with the name of his co-emperor. Marcus is given credit for keeping watch over all the affairs of the state and with planning and executing the 37

ROMAN EDESSA

conduct of the war himself, after which, plainly, Verus could not be admitted too much of a role if the reputation of Marcus was not to suffer. After the successes of Statius Priscus in Armenia and the installation of a pro-Roman king, Rome again took the offensive. As in the case of the Trajanic invasion, the reasons for the continued campaigning are unclear. It may be, however, that Parthian intransigence convinced the Romans that Armenia must not again be left vulnerable to attack from that quarter. Taking over command from Priscus, the energetic and popular Avidius Cassius swept down through Mesopotamia, took Nisibis and, like Trajan, went as far as Ctesiphon, which he sacked along with the nearby Greek city of Seleucia (Dio 71.2.3–3.1; Lucian Hist. Conscr. 15, 19). Verus again took great pride in his deputy’s accomplishments, which ended the Parthian threat for the foreseeable future and left Mesopotamian territory for the ﬁrst time permanently in Roman hands. Where Trajan had attempted and failed to create new provinces in the territories he had conquered beyond the Euphrates, Verus’s only addition to the territory of the empire was the extension of the Province of Syria to the south along the river, where it now included the fortress at Dura-Europos.9 Yet at the same time, Nisibis now contained a Roman garrison, and the Syrian legions apparently occupied advance positions in Osrhoene: a state of de facto annexation that was to lead to the next major Roman intervention in the region, under Septimius Severus. Osrhoene itself was occupied during the course of the invasion by a Roman force under the command of M. Claudius Fronto, whose career inscription lists him as leg. Augg. pr. pr. exercitus legionarii | et auxilior. per Orientem in Armeniam | et Osrhoenam et Anthemusiam ducto | rum, leg. Augg. legioni primae Minervi | ae in exspeditionem Parthicam deducen | dae. (CIL 6.1377 = Dessau 1098, ll. 14–18; cf. Lucian, Hist. Conscr. 21) It was surely Fronto’s armies that besieged Edessa, possibly being admitted to the city by partisans who also helped dispatch the Parthian occupying force (Lucian Hist. Conscr. 22; Procopius Persian Wars 2.12; HA Verus 7). The reign of Wael ended, and Ma nu returned to the throne, at the hands of the Roman legions. With such forceful backing Ma nu was able to proclaim the alliance with Rome publicly on his coins, and to rule for another 12 years until his 38

THE COMING OF ROME

son, Abgar VIII, came to the throne – according to our chronicler in Year of Abraham 2169 (Seleucid 464 – Chron. Zuq. 126/94). Applying the 26-year correction to this date we arrive at Seleucid 490, or 178/9 ce, for the inauguration of this Abgar, whose ﬁrst coins bear obverse portraits of Marcus Aurelius’s son and successor, Commodus (177–92 – BMC Arabia, etc. 93–4, nos 10–13). The Chronicle of Zuqnin gives Abgar a reign of 35 years – years that were to see some of the most momentous events in Edessa’s history, and that would end with the kingdom in an even closer alliance with Rome after a ﬁnal attempt to assert its independence.

The record of Sumatar As we have seen, the series of votive inscriptions on the hillside at Sumatar Harabesi southeast of Edessa bears witness to the attachment of Sumatar, and the Arab region east of it, to Edessa, at least at one point in the kingdom’s history. Most of these inscriptions are undated. Those that are dated, however, all come from the year 476 (Seleucid Era), which is 164/5 ce – shortly before the arrival of Rome’s armies to recapture Edessa for Ma nu Philorhomaios (Drijvers 1972: nos. 3–24). This in itself is intriguing, but more interesting still is the content of some of the inscriptions. In addition to the ofﬁce of shallita de- Arab or ‘ruler of the Arab region’ that has been discussed earlier, some individuals mentioned in these inscriptions bear the title of NWHDR or military commander, and others that of BWDR, an ofﬁce apparently with religious functions devoted to the deity worshipped at the site, who may have been the moon-god Sin. The majority of the inscriptions are simple ex votos, in which the dedicant asks to be remembered or seeks a blessing for himself or others, for instance: May Absamya son of Adona be remembered, the military commander (NWHDR ); may he be remembered before Ma¯rila¯ha¯; may BBS be remembered and Tiridates sons of ( A)bsamya. (Segal 1954: 21; Drijvers 1972: no. 18; 1980: 124) An exception to the rule is a lengthy inscription that deals with the succession to the ofﬁce of BWDR. In this inscription Tiridates the son of Adona, probably the brother of Absamya, is enjoined to care for the safety of a baetyl and a religious stool erected by his brothers, apparently upon Tiridates’s installation as ruler and BWDR 39

ROMAN EDESSA

(Segal 1954: 26–8; 1970: 57; Drijvers 1972: no. 24). The inscription enjoins Tiridates to hand over his religious functions to ‘him whom my ruler feeds’ – that is, his intended successor in ofﬁce – and invokes unspeciﬁed consequences if he should fail in this duty: ‘. . . if he withholds the stool, the baetyl will be ruined. He, the god, lives’. – ll. 8–9. This and a companion inscription, in which Tiridates commemorates his own erection of a baetyl and dedicates it ‘for the life of my king and his sons and for the life of Adona,’ are among the texts dated to 165 at the site. The inscription signed by Tiridates’s brothers is the only one that envisages the interruption of the cult practiced at Sumatar. It is most natural to try to connect this situation with the political and military events of 165, and to use this connection to illuminate the site’s whole history. This is what H. J. W. Drijvers has done, in the course of a study of preChristian Edessan religion. In the theory of Drijvers, the ofﬁcials named BWDR of Ma¯rila¯ha¯ are dedicated to Sin, who is mentioned in some inscriptions at the site, and for whom, Drijvers postulates, the pro-Parthian King Wael son of Sahru felt a ‘particular reverence’ (Drijvers 1980: 137). The ofﬁce of ‘ruler of Arab’ represents Wael’s attempt, with Parthian backing, to break away from Rome and to exert his inﬂuence over the region to the east. One of the holders of this ofﬁce was named Wael son of Wael, and this, says Drijvers, may be a relative of the king, perhaps one of his sons (ibid. 142). The inscriptions of Tiridates are dated to Shebat 476 or approximately February 165, which is before the Roman armies would have arrived. Therefore the king honored in Tiridates’s text can only be Wael himself; Tiridates was a member of the proParthian party, and the concern with the continuity of the cult that is expressed in one of these inscriptions reﬂects anxiety over the events to come. This reconstruction, although developed in the course of an investigation devoted more to religious than to political history, has something to be said for it. In fact the activity of the rulers of Arab at Sumatar is not attested before the time of Wael, and there is no telling for certain how long it survived his reign. It is plausible that, as hypothesized by Drijvers, Tiridates did have to resign as BWDR and ruler of Arab upon the arrival of the Romans in 165. From this point, however, events may have taken a different course from that envisaged by Drijvers. The inscription naming Wael son of Wael as ruler of Arab is one of a small group on the walls of a cave below the summit where the 40

THE COMING OF ROME

long inscriptions of Tiridates and his brothers are found. It is accompanied by an inscription naming another Wael, perhaps son of the ﬁrst one, who is called ‘ruler of SˇWDR’ (Drijvers 1972: nos. 3, 4, and 5; 1980: 131). Accompanying these inscriptions in the cave, which was perhaps the location of investiture or other ceremonies for the rulers of Arab and the holders of the religious ofﬁce of BWDR, are several more, inscribed seemingly in chronological order (Drijvers 1980: 130, 138–9). After Wael come inscriptions dedicated to Tiridates son of Adona, whose ofﬁce is unspeciﬁed here but whom the summit inscription identiﬁes as a ruler of Arab; another text honoring an Abgar, ruler of Arab; and ﬁnally one in honor of a certain Aurelius H . afsai ‘freedman of Antoninus Caesar,’ dedicated by Bar Nahar, another ruler of Arab (Drijvers 1972: nos. 12, 9, 7). The inscription honoring Abgar was dedicated by Ma nu son of Moqimu. Drijvers, observing that these names as well as those of the two Waels display some connection with the names of the kings of Edessa, suggests that the ofﬁces of ruler of Arab, of NWHDR and of BWDR of the god may have been exercised by relatives or allies of the Edessan king during the period when the king was exerting control over Sumatar and its region. Accordingly, he concludes that Tiridates son of Adona – whose inscription in honor of ‘my lord the king and his sons’ makes the royal connection explicit – had to leave the ofﬁce of BWDR in 165, soon after the text dealing with this possibility was inscribed; and that at this point the ofﬁce passed to the family of Wael’s rival Ma nu in the person of an Abgar, possibly the son of Ma nu who, 12 years later, took the throne as Abgar VIII (Drijvers 1980: 132). Drijvers further posits that Bar Nahar took over Abgar’s function when the latter became king, and that he was the last to hold this ofﬁce, losing it when the realm of Edessa was curtailed during the reign of Septimius Severus (Drijvers 1980: 134). The chronology is partly supported by the inscription in honor of Aurelius H . afsai, the freedman of an Emperor Antoninus. For Drijvers, this individual owed his name and his status to Marcus Aurelius Antoninus, so that this inscription comes in the period not long after the armies of Lucius Verus arrived in Osrhoene (Drijvers 1980: 131). If this interpretation of the Sumatar evidence is correct, the year 165 saw a smooth transition of power between the party of Wael (as represented by Tiridates) and that of Ma nu, whose son Abgar took control at Sumatar as soon as his family was restored to the Edessan throne. This, however, would seem to contradict the inscription of Tiridates’s brothers, predicting serious consequences, including the 41

ROMAN EDESSA

destruction of the baetyl or cult image, if Tiridates were not to see to the proper transfer of his religious functions. Given what we know of the antagonism between the pro-Parthian and the pro-Roman factions at Edessa, it is hard to conceive that the latter faction had a share in the Sumatar cult and the rule of Arab during Wael’s reign, or that Abgar son of Ma nu was being ‘reared’ as successor during that period. Therefore, if Abgar ruler of Arab really was Ma nu’s son or a member of his family, we might imagine some sort of break before his installation, and the continued dominance of the Arab by Edessa under the pro-Roman faction. This ‘break’ need not have been of the cataclysmic nature apparently foreseen by the inscription of Tiridates’s brothers; perhaps, indeed, one group of Edessan nobles simply gave way to a rival group. If the maintenance of the Sumatar cult without disruption was considered to be of vital importance for the religious well-being of the city, such a smooth transition is conceivable. A slightly different arrangement of events, however, may better ﬁt the circumstances surrounding the Roman invasion. It is, in fact, not quite clear whether the inscription of Abgar or that of Bar Nahar comes ﬁrst. If it was Bar Nahar (whose inscription honors the imperial freedman) who held the ofﬁce of ruler of Arab immediately after Tiridates, his inscription may represent a period when Rome was exerting closer control over the Sumatar region before allowing it to revert to a member of the old Edessan ruling class. There is, however, little to decide between these two possible reconstructions. Whichever alternative is adopted, the interpretation of these texts revolves around the concept that the civic ofﬁces of ruler of Arab and NWHDR and the religious function of BWDR formed a complex. This complex somehow embodied the civic and military authority of the individual delegated to hold them, along with a cultic relationship with the peoples of the area – who seem to have shared the devotion to Sin for which the nearby city of H . arran is famous. According to this line of thinking, Edessan patronage of the Sumatar cult would have been an expression of the kings’ intention to assert control over the eastward region. There is one point in the period after 165 when that intention forcefully reappeared, and this supplies yet a third possible reconstruction of the sequence of Sumatar inscriptions. In 194, Edessa cooperated with the Adiabenians in an attack on the stronghold of Nisibis, taking advantage of Rome’s preoccupation with the internal struggles surrounding the accession of Septimius Severus (Dio 75.1.2–3). Since, as we have allowed, it 42

THE COMING OF ROME

may have been Bar Nahar who held the ofﬁce of ruler of Arab immediately after the invasion of 165, the renewed territorial assertiveness of Edessa under King Abgar VIII would seem to be an appropriate time for the appearance in this ofﬁce of an Abgar. We know that during the reign of Wael son of Sahru, a Wael son of Wael – perhaps the king’s own son – held the ofﬁce; perhaps the appearance of Abgar is a resurrection of the same custom. For it is known that Abgar VIII had a son, Abgar Severus, who ruled ‘with his son’ for one year and seven months after the 35-year reign of his father (Chron. Zuq. 128/96). We have it on Cassius Dio’s authority that at the time of the joint Edessan/Adiabenian attack, the countryside between Edessa and Nisibis contained Roman forts and garrisons (Dio 75.1.2–3). It would thus seem appropriate for the Arab region to have been under the authority of an individual loyal to Rome in this period, and Bar Nahar’s dedication to Aurelius H . afsai would seem to make him ﬁt the bill. This is not to say that Roman and Edessan inﬂuence in the region were mutually exclusive; in fact, both Bar Nahar and his patron were, very possibly, members of Edessan families. In a Syriac document on parchment dating from 243 – discovered at Dura-Europos but drawn up in Edessa – an individual named Aurelius H . afsai signs as a witness (Drijvers 1972: 54, ll. 20 and verso l. 2; Bellinger and Welles 1935: 95; Goldstein 1966). In addition, the History of Michael the Syrian names Aurelianus son of H . afsai as the ﬁrst governor of Edessa after the end of the monarchy, during the 240s (Mich. Syr. [ed. Chabot] 1.120). If our hypothesis concerning the date of Bar Nahar is correct, this Aurelianus could be the son of Bar Nahar’s patron Aurelius H . afsai. This suggested rearrangement of the rulers of Arab does not detract from Drijvers’s main thesis, that the Sumatar cult and its associated ofﬁces were an expression of Edessan authority in the area. It does, however, place less stress on the cult of Sin as a private enthusiasm of King Wael and his circle, and more on its political role. By extending the sequence of rulers of Arab to take account of the well-attested Edessan activity in this region at the time of Septimius Severus, we have been able to read into these inscriptions a nearly complete record of events from 165 onwards. Yet the facts that only a few of the texts are dated, and that the chronological order of the inscriptions in the cave is only hypothetical, may undermine conﬁdence in this effort.10

43

ROMAN EDESSA

Conclusion Much of what has been said in the last section is necessarily hypothetical. Nevertheless, the Sumatar evidence – in combination with data from coins, the Classical historians and the Syriac Zuqnin chronography – provides a sense of the upheavals that the secondcentury Roman invasions brought to Edessa. The Parthian War of Trajan, although ostensibly launched in response to problems in Armenia, was also the beginning of serious Roman concern with territories further south in Mesopotamia. At Edessa, the result was the growth of a factionalism which, if not new, received added energy from the geopolitical implications of the kingdom’s choice of sides. Direct evidence of Parthian involvement in these events is lacking, but the general uprising that arose when Trajan annexed Osrhoene and Mesopotamia and proceeded on his southern pursuits is likely to have had some encouragement from Ctesiphon. Whether or not this was the case, Edessa learned in the episode that the consequences of faction were deadly serious. The Chronicle of Zuqnin tells us that King Abgar VII lost his life at this juncture, and that the ensuing period was one of ‘rivalry for power.’ While not certain, it is a reasonable conjecture that the king’s death had something to do with the factional violence or with Lusius Quietus’s forceful suppression of the rebellion. No longer could Edessa temporize between the Parthian and Roman realms; expressions of friendship such as those that had at ﬁrst left the kingdom in an ambiguous position (and roused early Roman suspicions) were insufﬁcient to satisfy the invader. If Parthia did not take as active a role in ensuring the loyalties of this region at the time of Trajan, it corrected this oversight when tensions over the Armenian question ﬂared again. The Zuqnin Chronicle and the coins reﬂect the installation of the pro-Parthian Wael in 163, suggesting that Vologaeses considered Edessa critical to his further plans to secure his northern ﬂank. Wael, however, was replaced two years later by Ma nu, the ‘Friend of the Romans’ on his coins, who had spent the interim among the Romans. The Sumatar record is evidence that, even in a period when the great powers were making their presence and their interest in the region increasingly felt, Edessa could hope to extend its regional hegemony under the wing of one or the other of them. Like the coins and the Edessan king-list, however, the inscriptions also reﬂect factional tensions. Chronological uncertainties prevent them from being an adequate record, but even if all our hypotheses concerning 44

THE COMING OF ROME

the sequence of control over this region should prove incorrect, there remains the one remarkable inscription of the brothers of Tiridates. Dated in the months just before the rearrival of Ma nu and the Roman armies, the inscription seems to foresee a period of trouble, the possible interruption of the cult and destruction of the holy objects. Admittedly, this sort of anxiety could result from an awareness that war was imminent, without any implications for the question of factionalism. But the fact that Tiridates is followed in ofﬁce by an Abgar – whose name is associated with the pro-Roman faction of Ma nu – and by the protégé of an imperial freedman, puts a different color on things, regardless of the exact chronology. The advent of Rome and of serious attention to the region by both great powers meant that, like it or not, whatever political or military moves Edessa should make would have consequences beyond its immediate neighborhood. This was a lesson that would be driven home when Ma nu’s son, Abgar ‘the Great,’ tried to take advantage of a seeming opportunity for aggrandizement – nearly the last gasp of the Edessan spirit of independence and self-assertion.

45

3 FROM KINGDOM TO PROVINCE

Severus and Abgar: rebellion and submission Abgar VIII Bar Ma nu, the king who owed his throne to the Roman troops who reinstated his father, can with reason be called a ‘client’ of Roman power from the beginning of his reign. This is clearly enough shown by the king’s coins, the earliest of which bear portraits of the Emperor Commodus (177–192: BMC Arabia, etc. 93–4, nos. 10–13). The precise status of Edessa and Osrhoene after the campaigns of Lucius Verus, however, is not totally clear. Although Roman troops were apparently present in some force in the region, and in Nisibis to the east, Abgar’s subsequent behavior shows that he was neither very closely watched, nor very ﬁrmly attached to Rome, during the ﬁrst part of his reign.1 Edessan policy as we have reconstructed it had always been to seek to preserve the kingdom’s independence and to expand its inﬂuence as far as was safely possible in the shadow of Roman and Parthian power, and Abgar’s behavior at the end of the second century ﬁts this pattern. In attempting to take advantage of Rome’s internal troubles, however, this king accomplished only the reduction of his kingdom – and brought his own dynasty to the verge of extinction. This picture of Abgar’s policy and motives is open to question, since it needs to be pieced together from the scraps of evidence that can be assembled. On the other hand, Rome’s own actions in this period, if better documented, are not necessarily better understood. It was ‘out of a desire for glory’ ("πιθυµ!6 δ ξη), wrote Cassius Dio, that the Emperor Septimius Severus (Dio’s contemporary) ‘made a campaign against the barbarians – against the Osrhoeni, the Adiabeni, and the Arabians’ (Dio 75.1.1) early in his reign. Whatever this phrase about glory means – and we have seen that Dio uses the same language to characterize the behavior of Trajan – it is 46

FROM KINGDOM TO PROVINCE

unlikely to represent a mere egotistical impulse on Severus’s part. Historians in the modern era have tended to stress the geopolitical element in Rome’s eastern frontier activities.2 Severus mounted this campaign, however, only shortly after he had won his ﬁrst successes against his rivals – and certainly before eliminating the last of them. It can also now be demonstrated, as we shall see, that the formal annexation of Osrhoene and its organization as a Roman province came early in the reign. In assessing the reasons for these actions, we may do well to remember that during these years Septimius Severus was very heavily occupied with civil conﬂicts. It might also be observed, in reference to Dio’s comment about the emperor’s ‘glory,’ that glory could prove a useful commodity to a pretender attempting to shore up his power base. It was the war against Pescennius Niger, who governed Syria at the time of Pertinax’s death, that ﬁrst brought the new emperor to the Orient, and resulted in his ﬁrst campaign across the Euphrates. This crucial episode is mentioned only brieﬂy in the excerpts of Dio’s history, and Herodian’s account is seriously defective. Yet a careful examination of these two sources, along with documentary evidence, provides a reasonably clear idea of events. In Dio the extract in question follows immediately on the account of the capture of Byzantium. In modern editions that capture comes after the ﬁnal defeat of Niger’s forces at Issus in 194, Niger’s death and Severus’s punishment of some of his rival’s supporters.3 The Osrhoeni and the Adiabeni had revolted and laid siege to Nisibis, and had been defeated by Severus; but now, after Niger’s death, they sent an embassy to him, not, indeed, to ask his pardon, as if they had committed any wrong, but to demand reciprocal favors, pretending that they had acted as they had on his behalf; for they claimed it was for his sake that they had destroyed the soldiers who favored Niger’s cause. They also sent him some gifts and promised to restore the captives and whatever spoils there were still left; yet they were unwilling either to abandon the forts that they had captured or to receive garrisons, but actually demanded the removal from their country of such garrisons as still remained. It was this that led to the present war.4 This passage is difﬁcult for several reasons, not least because it mentions a defeat of the Osrhoenian and allied forces by Severus before his campaign against them. If we follow the sequence of the 47

ROMAN EDESSA

extract strictly, this defeat also took place before the death of Niger. These statements may imply that Osrhoene and Adiabene provided forces to Niger, and that these were among the armies defeated during the succession conﬂict (cf. J. Sturm, RE 17, col. 736 s.v. Nisibis). Favoring this interpretation is the fact that among the targets of Severus’s campaign were ‘Arabians’ (Dio 75.1.1), although these were not charged with participating in the attack on Nisibis. The Arab ruler of Hatra, Barsemias, did support Niger (Hdn. 3.1.2), but Severus’s attempt to get revenge on Hatra forms a separate and notorious episode in Dio (75.10–11). The terms ‘Arab’ and ‘Arabia’ as used by Classical writers are confusing; in many places they seem to refer to Mesopotamia and its inhabitants generally. It is noteworthy, indeed, that Bardais.an, the Edessan contemporary of Dio, is presented as speaking of the Roman conquest of ‘Arabia’ in a way that makes it look as if he is talking not about the Roman province of Arabia, but of Mesopotamia.5 Among the targets of Severus’s ﬁrst eastern campaign, then, may have been other residents of ‘Arabia’/ northern Mesopotamia, what eventually became Severus’s Province of Mesopotamia. As we shall see, the subsequent course of events makes it appear unlikely that Abgar VIII of Edessa declared for Niger so openly. In addition to the implication contained in the phrase ττηθ-ντε 2π3 Σεουρου, the only other piece of evidence possibly suggesting that Edessa supported Niger against Severus is its cooperation with Adiabene in the siege of Nisibis. Scholarship on this point is by no means unanimous, but the interpretation of this action as pro-Niger is the prevailing viewpoint among those who see events through the lens of Roman history.6 It is true that Dio describes the kingdoms as being in a state of rebellion (αποστα´ντε), but this does not necessarily indicate a rebellion against Severus and for Niger. It is, in fact, more likely that Osrhoene and Adiabene were cooperating in an attempt to evict Roman forces from Mesopotamia altogether, and that this is what is implied by the word αποστα´ντε.7 The implications of the attack on Nisibis are hard to grasp, because its status at the end of the second century is almost completely unknown. The city had come under siege during the Parthian campaign of Lucius Verus in the 160s, at which time a plague broke out within the walls (Lucian, Hist. Conscr. 15); but the result of the siege is unknown. Presumably, however, it ended in success for the Romans, since there was a garrison there to be besieged almost 30 years later. Moreover, Dio’s mention of fortresses 48

FROM KINGDOM TO PROVINCE

(τε!χη) and prisoners (α+χµα´λωτοι) taken during the siege implies a Roman presence of some substance in this region of Mesopotamia east of Osrhoene. Even less is known about the position of Nisibis and its garrison in the conﬂict between Niger and Severus. Although the scornful tone of Dio’s report on the embassy to Severus indicates that the ambassadors’ claim of action in Severus’s interest was accorded little credit, Dio does not seem to doubt that some, at least, of the soldiers destroyed by the besiegers favored the side of Niger. For this reason, although some scholars have concluded from the fact of the attack on Nisibis that the city favored Severus, others have seen in it evidence of loyalty to Niger on the part of the troops stationed there (most recently Millar 1993: 113). In either case, to conclude from the attack on Nisibis anything about the allegiance to either side of Osrhoene and Adiabene risks circularity. One of the surest indicators of a community’s position before the fall of Niger may be the treatment it received from Severus after it. The emperor wasted little time in dealing out rewards and punishments as soon as he had done away with his eastern rival. Dio and Herodian provide explicit evidence of the treatment of both individuals and communities, including the demotion in rank of Greek cities and their attribution to the territory of their closest (and therefore most hated) neighbors (Hdn. 3.6.9; Dio 75.8–9). In SyroMesopotamia the most notorious case is that of Antioch, which had served as Niger’s base of operations and his mint, and which was demoted to the status of a ‘village’ (κ,µη) in the territory of its rival Laodicea.8 In the case of Nisibis, Severus placed the city under the command of an equestrian ofﬁcer and rewarded it with unspeciﬁed honors (αξ!ωµα): the strongest indication that he saw in it no evidence of disloyalty (Dio 75.3.2). Until recently, it was considered certain by many that Abgar of Edessa was severely punished, giving evidence that he had, indeed, been disloyal to Severus (for the classic statement of this position see Sturm, RE 17, col. 737). The punishment was the transformation of his entire kingdom, including the capital, into the Roman province of Osrhoene, whose ﬁrst procuratorial governor, C. Julius Pacatianus, has long been known on the basis of the inscription giving his cursus honorum.9 The sequence of positions Pacatianus held implied, though it did not prove, that the province was created relatively early in Severus’s reign, thus putting all of Osrhoene, like the area around Nisibis, under ﬁrm military control.10 Abgar thus paid for his disloyalty, it was thought, with his crown. The Edessan mint, however, produced a sizable series of coins bearing the 49

ROMAN EDESSA

portraits of the king and Septimius Severus on opposite sides, proving that Abgar adopted Severus’s name, and seeming to show a relationship similar to that which he had earlier had with Commodus (Babelon 1893: 251–7; BMC Arabia, etc. 94–6). On these coins Abgar styles himself Lucius Aelius Septimius Abgar or Lucius Aelius Aurelius Septimius Abgar. Therefore, those who propounded the theory of the king’s dethronement were forced to hypothesize that the Edessan king later managed to ingratiate himself with Severus and regain the throne (Hasebroek 1921: 76). From another inscription mentioning Pacatianus and published in 1983, however, it is now clear that the establishment of the province did not mean the immediate dethronement of King Abgar: ex auctoritate Imp. Caes. L. Septimi Severi Pii Pertinacis Aug. Arab. Adiab. pontif. max., trib. pot. III, imp. VII, cos. II, p. p., C. Iul. Pacatianus proc. Aug. inter provinciam Osrhoenam et regnum Abgari ﬁnes posuit. (Wagner 1983: 113; cf. Wagner 1985: 64) This inscription was found near the village of Kızılburç some 25 km northwest of Urfa. Since it records the setting of boundaries between the new province and the kingdom of Abgar, it is now plain that the former did not replace the latter, but coexisted with it. Moreover, the inscription’s mention of the third year of Severus’s tribunician power (the ﬁrst was 193) puts the inscription in 195, the very year in which the emperor mounted his ﬁrst campaign east of the Euphrates. Severus did not put off dealing with Abgar until his return to the area in 197 – but neither did he punish him severely with dethronement. Just how severe the punishment in fact was is difﬁcult to ascertain. Certainly Abgar lost a substantial amount of his kingdom. Although the boundary inscription of Pacatianus was not found in situ, it is unlikely to have migrated far from its original position. Moreover, the approximate location of the border between the province and the kingdom is conﬁrmed by a milestone of 197, putting the distance from the Euphrates to the border at 48 millia passuum (Wagner 1983: 115). If the limit of the province was this close to 50

FROM KINGDOM TO PROVINCE

Edessa on the west, the loss of territory would seem to amount to approximately the western half of the kingdom. To the south, Abgar’s remaining territory was probably similarly limited by the district of Carrhae, which became a Roman colonia at some point in Severus’s reign. If it was equally restricted on all sides (though there is no ﬁrm proof of this), the remnant amounted to ‘nur noch das Stadtgebiet von Edessa’ (Wagner 1985: 64, with map p. 42). To the east, however, we have no evidence for the extent of territory controlled by the Roman forces at Nisibis.11 The remains of Roman military installations uncovered during survey work in 1979 and 1980 are all to the west and northwest of the city, as well (Wagner 1983: 107–10). It is only on the west, therefore, that we have ﬁrm evidence of the boundary between the province and the kingdom. Of course, it is unlikely that Severus left Abgar with any substantial military resources with which again to threaten the Roman presence. However, two additional observations suggest that, however limited his territory, the Edessan monarch retained more than a token kingship. The ﬁrst is the fact that Edessa retained minting rights. This contrasts with the treatment of communities that supported Niger, such as Antioch. The second is Abgar’s later visit to Rome, apparently at Severus’s personal invitation, in the midst of great pomp and ceremony (Dio 79.16.2; Procop. Bell. Pers. 2.12). The combined evidence indicates that Severus may have pondered a role for Abgar as a client king in the eastern frontier system, analogous to that of the former Armenian King Tiridates. This analogy is suggested by the fact that Cassius Dio compares Abgar’s ostentatious reception at Rome to that of Tiridates when he visited during the reign of Nero. The analogy may stretch the point somewhat – Armenia was always more important to Roman interests than Osrhoene – but in any case, Abgar’s survival shows one thing: that Severus did not see the attack on Nisibis as a serious threat to himself. Even so, what led Severus to allow a local ruler who had taken direct military action against Roman interests to retain his throne? The answer must be that, during the ﬁrst few years of his reign, Severus was still preoccupied with eliminating his rivals for the throne and their supporters. Abgar may have survived because he was able to represent himself as an asset to the emperor in this ongoing campaign. Loyalty to Severus, at this juncture, was more important than loyalty to Rome – and the evidence seems to compel the conclusion that, in spite of Dio’s phrase ττηθ-ντε 2π3 Σεουρου (75.1.2), the Osrhoenian forces never put up any serious 51

ROMAN EDESSA

resistance to Severus once he arrived in the East.12 Although Severus took the titles Parthicus, Adiabenicus, and Arabicus (later dropping the ‘Parthicus’ for fear of offending the Parthians), there is no evidence of Osrhoenicus or anything similar to it as an imperial title (HA Severus 9.9–11; Aur. Victor De Caes. 20.14–17; Eutropius, Breviarum 8.18.4).13 Severus did receive three acclamations on the conclusion of his ﬁrst Mesopotamian campaign; but these must be for his Arab, Adiabenian, and Parthian victories – leaving no room for an Osrhoenian one.14 A passage in the contemporary history of Herodian, although difﬁcult to pin down chronologically, gives a good insight into the dynamics of Abgar VIII’s relationship with Severus: Abgar, the king of Osrhoene, also took refuge with the emperor and handed over his children as hostages to guarantee his good faith. He brought a large number of archers to be auxiliary forces for Severus.15 Herodian puts this episode, along with the submission of the king of Armenia, in the context of Severus’s second Parthian expedition, in 197. Some who have studied the issue believe Abgar’s submission to the emperor (practically an act of deditio) ﬁts better with the events of 195, when the king might naturally have been desperate to demonstrate that he was no threat to the new emperor.16 If this is right it would help explain Severus’s forgiving attitude toward Abgar. Even if Herodian is correct in placing this episode among the events of the second Parthian campaign, however, the fact that Abgar had a force of archers to offer could be slight evidence that Severus saw in him a potentially useful ally, even at the time of his earlier campaign, and chose for that reason to spare him.17 If we are right, Severus received the submission of Osrhoene without much of a ﬁght. Nonetheless, the territory did not escape unharmed. After reporting on the failure of the rebels’ embassy to Severus, Dio reports: Afterwards [after crossing the Euphrates and a parched countryside] Severus reached Nisibis, and tarrying there himself, sent Lateranus, Candidus, and Laetus in various directions among the barbarians named; and these generals upon reaching their goals proceeded to lay waste the barbarians’ land and to capture their cities.18 52

FROM KINGDOM TO PROVINCE

This passage is found in the text of Xiphilinus’s abridgement of Dio at 303, 21–304. Immediately preceding it is the statement that Severus mounted his campaign ‘out of a desire for glory’ against the Osrhoeni, the Adiabeni, and the Arabs (Xiph. 299, 25–303, 21). There can be no doubt, therefore, that the phrase το' προειρηµ-νου βαρβα´ρου is intended to signify these three peoples. Dio’s account of the march to Nisibis highlights the extreme thirst which the army suffered, describing an incident in which Severus was forced to drain a cup of foul water to assure his weary soldiers that it was drinkable (Dio 75.2.2). The army clearly took a lengthy and roundabout route, apparently skirting the southern edge of the territory controlled by Abgar. Therefore they were unable to enjoy the beneﬁts of the Edessan oasis until after reaching Nisibis and turning back to approach it from the east.19 To sum up: Severus’s treatment of Abgar and Osrhoene, which was relatively lenient but whose degree of harshness cannot be precisely gauged, implies that he saw in the kingdom’s behavior no evidence of serious disloyalty to his person. This applies speciﬁcally to the attack on Nisibis, but may also indicate that Abgar did not in any other way contribute to Niger’s bid for empire. What then of Dio’s claim that Abgar’s forces and those of his Adiabenian and Arab allies had been defeated by Severus? This must refer to the outcome of the attack on Nisibis. As Dio describes it, the attack resulted in a stalemate. Although they had achieved some initial victories resulting in the capture of fortresses and prisoners, the attackers failed to achieve their primary objective, the important stronghold of Nisibis. As we also saw, Severus rewarded Nisibis for its loyalty to him, which must mean that its garrison declared in his favor (possibly only after receiving the report of his victory at Issus). When the native rulers’ siege of Nisibis failed, therefore, they could be described as having been ‘defeated by’ Severus, that is, by forces loyal to him though not under his command. The fact that Severus nevertheless treated Abgar leniently, in addition to suggesting that he saw the ruler as potentially useful, also reﬂects the ambiguity of the whole situation. Severus chose to accept the declaration of loyalty on the part of Nisibis, but also that of Abgar who had attacked the city. Very possibly he had the sense that things could have gone the other way – the eastern outpost might as easily have chosen the side of his rival, whose Syrian power base was so close.

53

ROMAN EDESSA

Securing Osrhoene and Mesopotamia Julius Pacatianus’s early organizing activity in the new province bespeaks Severus’s concern that the security threat it presented be quickly brought under control. The same concern had been evident in the new emperor’s treatment of Syria. Himself a former governor of that province, he realized, soon after Niger’s death, that the position concentrated too much power in one man’s hands, and divided Syria immediately into the provinces of Syria Phoenice and Syria Coele.20 Pacatianus held command of one of Severus’s three newly formed Parthian legions, possibly the First – either concurrently with, or directly after, his procuratorship of Osrhoene. He was soon called away, however, probably for service in the ultimate battle with Clodius Albinus.21 It was one of his successors as procurator of Osrhoene, Aelius Ianuarius, who was responsible for the completion of the road from Zeugma to the borders of Abgar’s kingdom, a job completed by 205.22 In the meantime, the paciﬁcation of the countryside proceeded, and a vexillation of Legio IV Scythica constructed the substantial fortress at Eski Hissar, northwest of Edessa, by 197 – in time for Severus’s second Parthian campaign (Wagner 1983: 112–13 [TRIB POT V]). The composition of Osrhoene’s garrison at this time is an open question. Since the location of Legio I Parthica is still unknown, and Pacatianus is listed as having held the command of one of the Parthian legions, it would seem possible that this legion was in the new province (E. Ritterling, RE 12 cols. 1308– 9 s.v. ‘Legio’; Millar 1993: 126). There is no positive evidence for this, however. What is certain is that the build-up of Roman forces in the new province was heavy. By the time of composition of the late Roman Notitia Dignitatum (c. 395), the dux of Osrhoene had under his command nine units of equites including the Equites sagittarii indigenae primi Osrhoeni at ‘Rasin’; the Legio IV Parthica at Circesium; six alae and two cohorts (Not. Dig. Or. [Seeck] 35). Eleven castella are named in Osrhoene. Severus’s annexation, therefore, began a process which was to see the center of gravity shift permanently to the east. The inscription recording the completion of the road from the Euphrates to the border of Abgar’s kingdom in 205 says that Severus and his sons built the road a novo, which may mean either that it was then constructed for the ﬁrst time or that it was repaired (AE 1984, no. 18). If this was a repair project (a road may have been built at the time of L. Verus’s invasion and capture of Edessa), it adds one element to the emerging picture of a program of renewal 54

FROM KINGDOM TO PROVINCE

on the eastern frontier following the Severan annexations. The clues are sparse, but include a group of inscriptions recording the restoration of a bridge at Cendere, in the frontier zone west of the Euphrates. The bridge carried the military road north from Commagene and Syria to Cappadocia, and was restored under Severus, Caracalla and Geta – probably around 199 (CIL 3, 6711; IGLS 41; Mitford 1980: 1206–7; Mitford 1972: inscription no. 83). According to these inscriptions, the joint emperors ‘restored the bridge of the River Chabina and renewed the roadway’ (ll. 16–20). This work ﬁts into the context of military operations surrounding Severus’s second Parthian war. Another text in the area just west of Osrhoene, however, suggests that such a program of renewal, or at least the rhetoric surrounding it, had begun even earlier, after the completion of the ﬁrst campaign. It records in Greek the restoration of ‘the primeval altar’ of Zeus at the imposing site of Derik Kale by ‘Candidus.’23 The dedicator is otherwise unidentiﬁed but may be Ti. Claudius Candidus, one of the three generals to whom Severus assigned the reduction of the hostile peoples after his arrival at Nisibis (Dio 75.2.3). Hence the editor’s estimate of 195 as the date for this inscription. It was at this same time, as we have seen, that C. Iulius Pacatianus was organizing Provincia Osrhoenae, and a vexillation of Legio IV Scythica was engaged in the construction of the fortress at Eski Hissar. Osrhoene may have been literally on the frontier – the last province of Rome, bordering the Parthian Empire – until the establishment of the province of Mesopotamia. This critical strategic situation makes it seem likely that the newly formed First Parthian Legion was located in Osrhoene during Pacatianus’s procuratorship, before Severus’s return to the East to pursue his Parthian war (Kennedy 1987: 61–2). Some few years later, however, both the First and the Third Parthian Legions are found in Mesopotamia itself, while the Second was based in Italy (Dio 55.24.4; Ritterling, RE 12 cols. 1308–9). Pacatianus’s job as prefect of the First may have been simply to conduct it to its Mesopotamian base – perhaps Singhara – before taking up his western command in the Alps.24 From Rhesaina, coins which apparently bear an image of the legionary standard of the Third Parthian Legion suggest that this city was its base.25 A relatively small number of inscriptions supplements this evidence, leaving a clear impression that Mesopotamia was a province by the end of the second Parthian campaign, in 199.26 For the status of cities located within the new province during this period the best evidence comes from their coins. In the next 55

ROMAN EDESSA

few decades cities such as Nisibis, Rhesaina and Singhara bear names such as Septimia Colonia on their coinage, indicating that municipal or colonial status was bestowed on them in the reign of Severus.27 The most likely time for this to have happened is during the ﬁrst few years of the reign, when Severus was directly involved in the region.

Honors for the king The rapid incorporation of these easternmost of Rome’s subject cities into the provincial system is surely a result of Severus’s need to ensure their security and loyalty during a risky period, as he left them at his back and marched ﬁrst on Albinus, then on Ctesiphon (Dio 75.4–8, 9–10). All the more remarkable, then, was the reign of Abgar at Edessa. As we have seen, Abgar saved his neck and what remained of his kingdom by a swift submission to Severus once the emperor arrived on the scene. Beyond that, however, he seems to have established an unusual relationship with the Roman emperor. Cassius Dio’s mention of an Abgar’s visit to Rome during the reign of Severus (Dio 80.16.2) can only refer to Abgar VIII, and makes clear that this visit made quite an impression.28 A story current in a later era, though largely legendary, helps discredit the idea that the splendid procession of Abgar mentioned by Dio happened in Mesopotamia, not Rome (Magie 1950: 1542). In the context of a Persian attack on Edessa in 540, Procopius (Persian Wars 2.12) explains the city’s importance by reference to its early history, including the story of the supposed correspondence between an Abgar and Jesus Christ. In the same passage it is reported that Abgar visited Rome, proving to be so wise and pleasant that the emperor did not want to let him return to Edessa, until he made a clever demonstration in the arena. The story ends with the king returning to Edessa with the gift of a hippodrome from Augustus, and his epigram describing the gift as bringing ‘pain without loss and pleasure without gain.’29 Previous students have connected this story with the visit of Abgar VIII to Rome, sometimes concluding that the visit was ascribed to Jesus’s contemporary, Abgar V, under the inﬂuence of Armenian Christian legends (Duval 1892: 214–15; the hippodrome: 100). Clearly there can be no certainty in the case of an account written so long after the event, and one which contains so many legendary elements. The story may have been invented to give a context to the epigram. Yet in combination with the mention in Dio, it helps 56

FROM KINGDOM TO PROVINCE

support the supposition that a reigning Abgar did visit Rome, and if so, it must have been Abgar VIII. It remains to note that the Dio passage is meant, not to draw a parallel between the political position of Abgar and that of Tiridates, but only between the pomp (and perhaps the exotica) that surrounded their visits and that of the procession honoring Aurelius Zoticus. It was under Abgar VIII, who bears the epithet ‘Great’ in Syriac literature (but not on his coins – cf. Babelon 1893: 256), that preChristian Edessan culture reached its zenith, as is demonstrated by the remains of the literature of the period, artistic monuments (mosaics) and texts, including the Chronicle of Edessa’s account of the ‘Great King’s’ magnanimous response to the destruction caused by the ﬂood of 201. At Edessa, and surely in the surrounding region, the outpouring of coins produced under Abgar VIII and bearing the images of emperor and king on opposite sides publicized a close association between the two rulers (BMC Arabia, etc. 94–6, nos. 14–35; Babelon 1893: 250–5). For all these reasons – the story of the friendship between king and emperor, the ﬂowering of Edessan culture during the Severan period, and the impression given by the coinage – one concludes that Septimius Severus was more than grudgingly forgiving of Abgar. This relationship between the two men may have cushioned Edessa from the effects of Roman rule during this critical period of its development, and it may to some degree account for the leading cultural role the city would take in the Late Roman and Byzantine periods. In any event, Edessa was safe from more direct Roman control as long as both Severus and Abgar ruled.

Without a king: Colonia Edessa As with the Severan coloniae of Mesopotamia, the coins of Edessa bear witness to its transformation into a Roman colonia in the ﬁrst half of the third century (BMC Arabia, etc. 99–112 [rev. legends ΜΗΤ. ΚΟΛ. Ε∆ΕΣΣΗΝΩΝ, etc., beginning with Elagabalus]; Babelon 1893: 268–86). Here, however, the municipal title contains no reference to Septimius himself. Edessa had remained a monarchy, and the Abgarid family retained the throne, until the emperor’s son and successor Caracalla removed it in an act of deception related by Dio: [Antoninus (Caracalla)] tricked the king of the Osrhoeni, Abgarus, inducing him to visit as a friend, and then arresting 57

ROMAN EDESSA

and imprisoning him; and so, Osrhoene being thus left without a king, he subdued it.30 This episode, then, marks the interruption of royal rule and the establishment of Edessa’s self-governing municipal government – though not necessarily the point at which the city became a colonia. Dio’s language would seem to imply that any remaining area beyond the city’s own territory was absorbed into the existing province of Osrhoene. The date of this change can be fairly precisely ﬁxed with the use of the Syriac documents from the reign of Gordian III, drawn up under the municipal government of the city. These documents are dated in multiple forms: by the year of the emperor’s reign, the consular year, the Seleucid era and the year of the city’s ‘freedom,’ as well as the years in ofﬁce of local magistrates (Teixidor 1990; Bellinger and Welles 1935; Goldstein 1966). ‘Document B’ from the Beth Phouraia archive is dated to Elul (approximately September) of the ﬁfth year of Gordian III and the consulship of Vetius Atticus and Lepidus Praetextatus, which was 242; while the parchment P. Dura 28 – an Edessan slave-sale contract that was discovered in the excavations at Dura-Europos – comes from Iyyar (roughly May) of Gordian’s sixth year and the consulship of Annius Arrianus and Cervonius Papus, 243. In other words, Document B was drawn up near the end of the Seleucid year 553, or 241/2, and P. Dura 28 in the middle of 554 (242/3).31 The dates are also expressed as Year 30 and Year 31, respectively, of the ‘freedom’ of Edessa Antoniana (Antoniana Edessa), Colonia Metropolis Aurelia Alexandria. Counting backwards, this gives 212/13, Seleucid year 524, as the ﬁrst year of the city’s ‘freedom.’32 The ‘freedom’ which is celebrated by this era is likely to have been the city’s liberation from royal rule, as is suggested by the passage which, in modern editions, immediately precedes Caracalla’s ouster of the Edessan king: Abgarus, king of the Osrhoeni, when he had once got control of the kindred tribes, visited upon their leaders all the worst forms of cruelty. Nominally he was compelling them to change to Roman customs, but in fact he was indulging his authority over them to the full.33 It is possible that the king lost his throne because of this arbitrary behavior, and this has been the interpretation adopted by many.34 It is too little remarked, however, that Dio states no reason for 58

FROM KINGDOM TO PROVINCE

Caracalla’s action. It is only the editor’s judgment that has placed the characterization of the king’s rule in close association with the report of his removal, which comes from an entirely separate manuscript context. The point to be observed here, however, is that there is some reason to believe that the people of Edessa (or at least their ‘leaders’ – the heads of local clans) may have seen the king’s removal as a liberating event. In the ﬁrst published treatment of the parchment found at Dura, the editors, without argument, took the date of this event as being the foundation date for Colonia Edessa (Bellinger and Welles 1935: 143). This required them to explain the appearance of the name ‘Alexandria’ in the colony’s full name as referring to Caracalla’s supposed obsession with Alexander the Great, on the assumption that Edessa took on, in 212/13, the full titulature that is attested by the later documents (Bellinger and Welles: 143 n. 4; Hdn. 4.8.1). It is very likely, however, that Edessa achieved colonial and metropolitan status not in one great step, but by stages. In that case, the city’s full name as attested on the documents of the mid-third century – Edessa Antoniana Colonia Metropolis Aurelia Alexandria – is well explained by the hypothesis that it became ‘Antoniana’ (more correctly, Antoniniana) upon its conversion to a free municipium in 212/13 – taking on the name of Antoninus/Caracalla – and ‘Aurelia’ during the short reign of Elagabalus, under whom it attained colonial rank.35 It would then have become ‘Alexandria’ between 222 and 235, being raised to metropolitan rank by the Emperor Severus Alexander. Such a progression would follow the pattern of gradual elevation in civic status that is seen elsewhere in the empire, and that was an important mechanism in Rome’s absorption of alien territories and peoples. From the earliest days, the enfranchisement of a community as a municipium with civic ofﬁces, social and religious privileges and perhaps the ius Italica was a mark of distinction. A further step beyond this was the granting of colonial status, the original signiﬁcance of which was that the community’s inhabitants were actually entitled to Roman citizenship. Such a colonia could be literally a ‘colony’ in the sense that Roman veterans were settled there as a reward for their service; but under the Principate, especially from the second century, such rights were granted to an increasing number of local communities simply as a mark of honor. The ﬁnal step beyond this, in what eventually comes to seem like an inﬂationary process, is the status of metropolis or ‘mother city,’ equivalent to a city’s being the capital of a district or the leading 59

ROMAN EDESSA

member of a regional κοιν or association of cities. When Edessa proclaimed itself as Metropolis, it was thus claiming equivalence with Antioch, the metropolis of Syria and one of the greatest cities of the empire. The signiﬁcance of the change from royal to municipal government, of course, is more than merely chronological. Before dealing with it, however, it is well to attempt to identify the king whom Caracalla removed. At 78.12.12 Dio gives his name only as ‘Agbaros’ ( Αγβαρο), muddling the spelling as he often does.36 He gives no indication that the throne of Edessa had changed hands since his last mention of Abgar VIII. From other sources, however, it seems that the king deposed by Caracalla was not the ‘Great King’ who had surrendered to Rome but his son, Abgar Severus. The Chronicle of Zuqnin states that, beginning in ‘the year 2203,’ a king named Abgar Severus reigned at Edessa, ‘with his son,’ for one year and seven months. The last dynastic entry before his is the 35-year reign of Abgar the Great, so this is almost certainly the son of that king. The late seventh-century Syriac historian Jacob of Edessa also mentions Abgar Severus, and speaks of the end of his reign as the point at which Edessa came permanently under Roman control.37 Here the chronology is confused; Jacob places the end of Abgar Severus’s reign in the ﬁfth year of the Emperor Philip I, by which time the monarchy had already been defunct for at least six years. That Abgar VIII might have given his son and heir the surname Severus in honor of his Roman benefactor is understandable, and makes even more sense if it is true that the king had another son named Antoninus (after Caracalla) who resided at Rome (IG 14.1315). The reign of Abgar Severus is independently attested by coins of Caracalla bearing his name, and by the legend of the Apostle Addai, supposed bringer of Christianity to Osrhoene, who is supposed to have been persecuted by a cruel king of this name.38 In this tale, Abgar’s son Severus is a cruel character who is held responsible for the execution of the apostle (cf. Gutschmid 1887: 15–16; Duval 1892: 222). Although the Doctrine of Addai is of questionable reliability, there is no apparent reason for its author to have invented a son of Abgar VIII named Severus. This effort to accumulate evidence for the reign of Abgar Severus (also known as Severus Abgar, by the testimony of his coins) would hardly be necessary, were it not for the fact that the Zuqnin document’s authority on this point has been variously contested. The coins of Edessa, so often used to clarify points that remain obscure in the sources, work to the opposite effect in this case, for the coin 60

FROM KINGDOM TO PROVINCE

portraits of Abgar Severus seem to show that he was a young man at the time of his rule – too young to have a son as a co-ruler, according to some interpreters. Hence Gutschmid rearranged the text of the chronicle to have it say that ‘Abgar [the Great] reigned at Edessa with his son Severus,’ thus doing away with ‘Abgar Severus’ entirely as a separate ruler (Gutschmid 1887: 43; cf. Duval 1892: 221–3). Others have wished to interpret ‘Abgar Severus’ in the Chronicle of Zuqnin as another way of denominating Abgar VIII.39 Under this interpretation his year and seven months’ rule ‘with his son’ would have been a joint rule with Ma nu, whom the chronicle names as the next king of Edessa but who never held effective power after Caracalla’s creation of the municipality of Edessa.40 Much easier than any of this is to allow Abgar Severus his reign and his son. Abgar VIII, the ‘Great King’ who ruled for 35 years, was surely old enough at the end of his reign to have not only a grown son, but a grandson – Ma nu – who was able to rule jointly with the son and to step into the role of ‘Crown Prince’ after his father’s removal, less than two years after Abgar VIII’s death. The reason for the youthful coin portrait of ‘Severus Abgar’ may remain a mystery, but it is better to let that mystery stand than to contradict the written sources unnecessarily or to rearrange them arbitrarily, as if the coins were a photographic record of the contemporary ruler’s age and appearance. It is also unlikely that ‘Abgar Severus’ is actually Abgar VIII, for the chronicler would not be expected to have added a new entry to record the year and seven months of joint rule, if this period was in fact part of the whole reign of the king who had been on the throne for more than 30 years. The king who reigned during this period and who was removed by Caracalla, in all probability, was Abgar VIII’s son Abgar Severus. The son who ruled with him was probably the Ma nu who ‘ruled’ for another 26 years as pas.griba or heir-apparent after Caracalla’s move. This resolution of the question of Abgar Severus’s identity leaves us in a position to take up the more interesting matter of his removal, and to attempt to understand his supposed attempt to impose ‘Roman customs’ (τα` των 5 Ρωµα!ων ;θη – Dio 78.12.1a). What Dio means by this phrase is not readily apparent. It is unlikely, however, that a sudden attempt to impose Latinitas or other stereotypical elements of ‘Romanization’ is intended; there is no evidence for the use of Latin and other such customs at any time among the Edessan population. Alternatively, Abgar Severus may have tried to stamp out customs, such as self-emasculation, that seemed barbaric to the Roman rulers. Yet we have it on the evidence 61

ROMAN EDESSA

of the Book of the Laws of Countries from the school of Bardais.an that Abgar VIII himself banned this practice, with no apparent damage to his popularity either with Rome or with his own people (cf. Kennedy 1987: 58–9). The text of Dio 78.12.1a itself may offer a hitherto underutilized clue to the signiﬁcance of Abgar Severus’s action. Dio writes that the king displayed his cruelty, not (as Hill would have it) ‘as soon as he had established his power over his people,’ but after he had extended his authority over related tribespeople: "πειδ) α<παξ "ν κρα´τει των =µοφλων "γ-νετο (cf. BMC Arabia, etc. ci–cii). Since his reign lasted less than two years, this must have occurred not too long after he came to power. It is possible, indeed, that it occurred simultaneously with his coronation; but whether it did or not, Dio relates the trespasses of Abgar Severus not to his assumption of the Edessan throne but to this extension of his power over the =µοφλοι. Moreover, Dio says that the king’s cruelty affected, not the general population of his own community or related peoples, but their leaders – το' προ-χοντα ατων. Any attempt to guess what those ‘cruel’ actions may have been needs to take into consideration the historical context of the period following Severus’s annexations, as well as one other factor: whatever those actions were, Abgar Severus apparently thought that the rationale – the λ γο – of imposing ‘Roman customs’ would make them desirable or acceptable to Rome (this is separate from the question of whether or not the actions were Caracalla’s pretext for his removal; the point is that he felt the need to justify them in terms of Roman practices). Pulling together all these strands of conjecture, the following picture emerges. In the 17 or 18 years between the creation of the province of Osrhoene and the removal of Abgar Severus, the Roman forces were surely concerned, not only with solidifying the occupation, but with making it pay for itself through control and taxation of the local population and economy. The same management approach was evident in other areas of Roman control, including the neighboring provinces of Arabia and Syria, and was a decisive ‘Romanizing’ inﬂuence (Millar 1993: 52; cf. 97). It was always through the cities, with their command of the agricultural resources of the hinterland, that these Roman aims were best served. A region containing a large nomadic or semi-nomadic population, however, could present real problems of security and of ﬁscal control. Mesopotamia and Osrhoene were like Syria in that they did contain a number of ‘Greek’ cities, but at the same time populations of nomads – ‘Saracens’ or ‘Scenite Arabs’ – inhabited the area. The 62

FROM KINGDOM TO PROVINCE

people of Edessa itself under the Abgarid dynasty, speaking a Semitic dialect akin to Arabic, were closely enough related to these other groups that Western writers sometimes called them ‘Arabs’ (Pliny, HN 5.20.2, 6.9.1 [the ‘Arabes Orrhoei’]; Tac. Ann. 11.10, 12.12, 14). Dio’s statement about Abgar Severus’s extension of power over the =µοφλοι could very well mean that, as the protégé of the Roman emperor, the king of Edessa gained some kind of explicit or implicit authority over these related tribes and their leaders. This relationship is very likely to have been established during the reign of Abgar VIII; if so, it apparently succeeded in large part because of that king’s authority gained through long tenure of the throne. It was when his less experienced son took over the position that problems arose. If all this is true, in what way was Abgar Severus attempting to impose ‘Roman customs’? This phrase could mean that he presented his extension of power over related tribes and their leaders to Rome as a step in forming them into communities for easier handling by the military and tax-gathering authorities. The leaders of the related tribes might understandably have been offended by such an attempt, especially if Dio is correct and Abgar Severus was using the process as an excuse to seize more power for himself. It is not hard to imagine those leaders’ spirited resistance leading to Abgar Severus’s extreme measures (των δεινοτα´των).41 An interesting chronological coincidence helps ﬂesh out this picture. From the evidence of the colonial-era documents, Caracalla deposed Abgar in 212/13. It was in this same year that Caracalla enacted what may be his best-known legal measure, the ‘Constitutio Antoniniana’ conferring Roman citizenship on every free inhabitant of the empire (Ulpian, Dig. 1.5.17). Cassius Dio, in the passage reporting on this measure, explicitly connects it with the emperor’s desire to enlarge the empire’s tax base (Dio 77.9.5). If there is any truth to this, it could support the notion that in the period leading up to this enactment, the Edessan king was trying to meet the emperor’s demands by exerting greater control over the related tribespeople in his region. It was only when this attempt faltered – perhaps backﬁring by arousing the resistance of tribal leaders – that Caracalla took the next step, guaranteeing the efﬁcient incorporation of Edessa and its territory into the imperial system even as he rationalized the status of free inhabitants throughout the empire. It is interesting that Jacob of Edessa, in his own passage recording the end of the monarchy, also connects that event with the imposition of taxes (Chron. Min. 282/211). 63

ROMAN EDESSA

This reconstruction seems to make the most consistent use of the bits of evidence available. In particular, to interpret Dio’s phrase τα` των 5 Ρωµα!ων ;θη as a reference to an attempt to ﬁt the tribes into the Roman provincial and taxation system seems to make more sense than to see it as signalling the abolition of outlandish religious customs. The former is something the occupying authorities may well have wished to see happen, for practical reasons, while the latter was probably less important to them. The Romans seem to have been willing to shrug off practices such as circumcision, and even self-emasculation, as long as the ‘Orientals’ who practiced them stayed off the throne of Rome (Lucian, Dea Syria 51; cf. Herodian 5, 6, 3–5, HA Elagabalus 7.1–5). If Abgar Severus did exceed the mandate he was given for the administration of the area, it was a serious mistake. It is not necessary to believe that Caracalla punished the Edessan monarch simply for mistreating his own people, although it is conceivable that he did so. With equal probability, the emperor perceived such a move to restore Abgarid power as a threat – either because it raised the prospect of a newly powerful ruler to challenge Rome, or because it could exacerbate resentment of Rome and unsettle the region. Whether or not these concerns motivated his deposition of the king, Caracalla was amply motivated to seek the incorporation of Edessa and its territory into the provincial system. In 212/13 the emperor was already planning his Parthian campaign, and needed to ensure the security of all cities in the Mesopotamian region. It was in similar circumstances that Severus had established Provincia Osrhoena, as well as the province of Mesopotamia with its coloniae. At that time, Abgar VIII went to great lengths to establish a relationship of trust with the emperor. When the two rulers’ sons came to power within a few months of each other, however, the basis for the special relationship was completely undone. In any case, whatever the ‘Roman customs’ were that Abgar Severus had been trying to spread, Caracalla eventually decided that the best and most efﬁcient means of doing this would be the transformation of his realm into a municipium – the traditional means of ‘Romanizing’ outlying parts of the empire.

A city of the empire As with so much about early Edessa, evidence about the evolution of society and politics in the years after the king’s removal is very sparse. About the only thing that can be done is to make cautious 64

FROM KINGDOM TO PROVINCE

conjectures based on what is known of military and political events in this period. With that in mind and following the model of betterattested cities in the Eastern Empire, we might hypothesize that one of the forces shaping Edessan society – in addition to a generally greater openness to the rest of the empire – was the social incentive to advance through public service, that is, by serving as a magistrate in a municipium organized along the traditional lines of the Greek polis. Another force shaping Edessan society during this time would have been the presence and activity of Roman soldiers from elsewhere in the empire, such as the legionaries from IV Scythica at Eski Hissar. As we have seen, however, Osrhoene itself had contributed to Roman manpower at the time of Abgar VIII’s submission to Severus, and the experience of the Osrhoenian auxiliaries probably also affected Edessan society and attitudes to Rome. Finally, it is possible that Caracalla’s citizenship decree had something of an equalizing effect, in that the free people of Edessa and Osrhoene were all brought into the empire with the theoretically equivalent status of citizen. Edessa also fell, more than many other cities, under the shadow of the military and political disturbances that threatened the Roman Empire during the ‘Crisis of the Third Century.’ In the years after Edessa gained its ‘freedom’ from royal rule, changes were taking place in the Parthian kingdom that would engulf the entire Near East in a new series of wars. In 224 the Parthian ruler Artabanus was defeated by a Persian rebel named Ardashir, who claimed descent from a soldier by the name of Sasan.42 This was the turning-point in the downfall of the Parthian Arsacids, and led to the reassertion of control over the entire eastern realm by Ardashir’s family, the Sassanids, who appealed directly to the Persian Empire’s Achaemenid heritage. The ambition of Ardashir, and of his even more energetic son Shapur (Sapor), to resurrect past Persian glories meant the end of what had been something of a stalemate all along the border with the Roman Empire. The Sassanid threat led eventually to one of the most remarkable episodes in our entire story: the brief suspension of colonial government and the resurrection of the monarchy. There is no reason to assume that, in the ﬁrst years of the municipium’s life, this possibility was foreseen by anyone (although Ma nu, the son of Abgar VIII, apparently lived on and held the strictly honorary title of pas.griba). For most of the ﬁrst 26 years, the newly Roman Edessa lived under the shadow of storm clouds in the east. An ominous passage near the end of Xiphilinus’s abridgement of 65

ROMAN EDESSA

Dio echoes the momentous events of the early to mid-third century. After telling of concerns over the renewed activity of Iran, the historian reports a number of mutinies, and points to the danger inherent in the shaky loyalties of the region’s garrisons (Dio 80.3). Against this background of instability, certain strands of evidence combine to suggest that the period up to the reign of Gordian III was a crucial one for Edessa’s development within the Roman Empire. Ironically, it seems to have been an episode of mutiny that led to the steps by which Osrhoene was, during this period, more ﬁrmly established as a ‘Roman’ city. Partly as a result of Severus’s marriage to the Emesene aristocrat Julia Domna, Syria remained a central concern of Roman policy during the generation after his death. Yet Caracalla’s abortive Parthian campaign launched in 214, and his assassination on the very outskirts of Edessa, were the last events east of the Euphrates to hold any particular signiﬁcance for the empire as a whole, until the resurgence of Iran under Ardashir in 230 (Dio 79.1.1–2; 5.4–5). After Macrinus’s defeat by the Parthians and subsequent abandonment of his predecessor’s aggressive policy, it would seem that a more or less stable situation prevailed, although a purported speech to the Senate of Severus Alexander shows that some in Rome may have blamed Elagabalus for neglecting the frontier region.43 There may have been some substance to these charges of neglect, to judge by Ardashir’s initial success in his attempt to recover the area, and the confused stories of rebellion and discontent in and around Edessa at the time of the war. As described by Herodian, this war involved a three-pronged offensive on the eastern front led by the young Emperor Severus Alexander, after diplomatic exchanges in response to the reassertion of Persian territorial rights failed to bring satisfaction (Hdn. 6.2.1– 2; 6.2.5; 6.4.5; 6.5.1–2). Despite Herodian’s dismissive tone, it is apparent that this campaign succeeded at least in putting a temporary check to Persian ambitions, as the historian is forced to acknowledge in describing the damage suffered by both sides.44 Yet the display of Persian military power must have made a lasting impression on the people and the armies of the Orient, and Dio is not the only writer to talk of unrest in these quarters. Just as he was to cross into Mesopotamia, the emperor was faced with a rebellion among soldiers of the Egyptian and Syrian legions, and was forced to reassign them to different sectors of the frontier (Hdn. 6.4.7). The Byzantine historians Zosimus and Syncellus, possibly both deriving from Dexippus, talk of military rebellions at the time of the eastern 66

FROM KINGDOM TO PROVINCE

campaign, in which usurpers named ‘Uranius’ and ‘Antoninus’ were proclaimed.45 According to Syncellus, Edessa itself was the center of this unrest. Evidence is lacking for the garrison of Edessa at this time; but if the accounts of Herodian and the later historians can be linked, they may provide a clue that now, as under Septimius Severus, it consisted at least partly of vexillations from the Syrian legions. These accounts speak of attempts to replace the emperor with candidates of the rebels’ own choosing, not to join forces with the Sassanid. Yet the passage of Dio cited earlier makes plain that the people and the legions of the areas under Persian attack harbored, or were suspected of harboring, sympathies with the attacker. It is hard not to read this as applying to Edessa and the surrounding area, though the only details in Dio’s account refer not to Osrhoene but to Mesopotamia (Dio 80.4.1–2: the death of the commander Flavius Heraclio). The city which had so recently been ruled by Abgar, ‘king of the Persians,’ was apparently still unsure of its loyalties – this although Edessa seems to have achieved the status of colonia already under Elagabalus, and although the abolition of the monarchy had created ‘freedom’ and opened the way for the development of municipal institutions. In terms of social and political development, however, an even more important factor than the change in Edessa’s juridical status may have been the earlier incorporation of Osrhoenians into the Roman army, beginning with the force of archers proffered by Abgar VIII on his ﬁrst submission to Severus (Hdn. 3.9.2). Inscriptional and literary evidence suggests that these sagittarii Osroeni, ﬁghting alongside other provincial auxiliaries, became an important element in the Roman strike force in the decades following Severus.46 When Severus Alexander came to the East, 35 years after Abgar’s submission to Severus, some of these auxiliaries were surely of retirement age. The later-attested titulature of Colonia Edessa makes it plausible that Alexander granted the city metropolitan status at this time, simultaneously with the settlement there of a group of retired Osrhoenian auxiliaries. These veterans, after a generation spent in service to Rome, might well have been a more reliable replacement for the rebellious troops earlier stationed at Edessa. They would also have formed an important core around which a Romanizing aristocracy could form. The ﬁrst promotion of Edessa to colonia under Elagabalus may also have marked a settlement of veterans or similar occasion; but Severus Alexander was apparently able to command Edessan loyalties in a way that his 67

ROMAN EDESSA

cousin could not. Osrhoenian troops are speciﬁcally named as leaders in the revolt against Maximinus in Germany that followed his seizure of power, and Edessa’s mint produced no coins in the name of Alexander’s successor.47 By this time, 20 years after the end of the monarchy, Edessa had apparently found an emperor on whom to focus its loyalties, an important step toward becoming a truly Roman city. The legal documents of the reign of Gordian III give us our only direct look at the inner workings of the municipality. Although these documents postdate the reign of Abgar X that interrupted the colonial regime, they must surely reﬂect in all essential respects the city’s organization during the 26 years that preceded that reign. It is fascinating to ﬁnd, therefore, that the documents are dated not only by regnal and consular years, but by the strategia of the ruling magistrates, and even by the year in ofﬁce of the city’s public priest (who by one reading is identiﬁed by the Syriac transliteration of the Greek >ερε). The magistrates and priests go on record with the preﬁxed name to which Caracalla’s citizenship decree entitled all of them, Marcus Aurelius. One of them is even recorded as being a member of Rome’s equestrian order.48 Nothing, it would seem, could be more Greek – or perhaps one should say, more provincially Roman. With a mixed heritage that reached back as far as the Hellenistic era, Edessa must have been fertile ground for the seeds of Romanization. One ﬁnal note is of interest. As we have seen, the Syriac historian Jacob of Edessa reports that the departure from the throne of Abgar Severus was followed by the imposition of taxes and the appointment of a governor or YGMWN (γεµ,ν) ‘in place of a king’ (Chron. Min. 282/211). Jacob gives the governor’s name as Aurelianus the son of ‘H . absay’ ( WRYLY NWS BR H . BSY), which by an easy emendation of the text could be read as H . afsai. The eponymous STRTGWS (στρατηγ ) of Edessa at the time of the inscription of the slave-sale document P. Dura 28 was a son of H . afsai, going by the name of Marcus Aurelius Abgar (P. Dura 28, ll. 3–4). The name also occurs in the context of the Sumatar Harabesi inscriptions, as we saw in Chapter 2. It would seem, therefore (unless the historian is mistaken) that the ﬁrst Roman governor of Edessa was a member of one of the city’s own leading families, perhaps a relative of the H . afsai named as a ‘freedman of Antoninus Caesar’: a pioneer in the Romanization of Edessa.

68

4 A KING IN ROME’S SERVICE

By the middle of the third decade of the third century ce, Edessa’s Abgarid monarchy was little more than a memory, the remnant of which lived on in the person of the pas.griba Ma nu. The city was integrated into the empire; it had achieved the gloriﬁed status of a metropolis; its people, now Roman citizens, attached ‘Aurelius’ or ‘Aurelia’ to their Semitic, Iranian or Greek names, signalling their self-identiﬁcation with the new order. Edessa retained its remarkable mix of Semitic, Iranian and Greek identities, and indeed was about to enter into the phase of Christian literary activity that represented the greatest ﬂowering of Syriac culture. Its transformation into Roman Edessa, however, was nearly complete. Before the ﬁnal extinction of the Abgarids, the royal name returned to the scene in the person of Abgar X or Aelius Septimius Abgar, the contemporary of the Emperor Gordian III (238–44), as has long been known on the basis of coins bearing the names and images of both rulers.1 Although the circumstances are still very unclear, this local ruler apparently held his throne brieﬂy as a result of Roman attempts to stave off catastrophe in the confrontation with resurgent Sassanid Iran. There is much about the reign of this last of the Abgars that is still a mystery, and will undoubtedly remain so, despite the publication within the last few years of a genuine Edessan document that has answered a number of questions. In this chapter we shall look at that document and its companions from the period shortly after the end of Abgar’s reign. We also consider the broader implications of his coinage, reserving detailed numismatic analysis for an appendix. The two forms of information together are sufﬁcient to allow us to form a conclusion about the position and attitude of this local dynast in relation to Rome. Abgar X took the throne not as a rebel or as a potential ally of the Sassanids, but explicitly in the Roman interest 69

ROMAN EDESSA

and under the auspices of his emperor. Although this might have been implied, in any case, by his coinage in Gordian’s name, the example of his ancestor Abgar VIII – who started as the client of Commodus but nevertheless took up arms against Roman interests – might lead us to expect similar vacillation on the later Abgar’s part. On the contrary, it can now be shown that Abgar not only acted on Rome’s behalf, but actually received his throne – and some sort of ofﬁcial position in the Roman imperial structure – from the hand of Gordian himself. This unexpected piece of information results from the documents of Abgar’s reign studied in conjunction with other, little-noticed sources concerning that of Gordian.

Gordian III: a young emperor at war2 Marcus Antonius Gordianus was, it seems, only 13 when raised to the throne as the result of a bloody sequence of events which saw the deaths of his grandfather and father (or uncle – Gordian I and II) in suicide and battle respectively, and the short-lived reign of the senatorial candidates Pupienus and Balbinus.3 The exact date of Gordian III’s accession is unknown, and a recent study of this period concludes that it may have been at any time between May and August 238 (Bland 1991: 45; cf. Loriot 1974: 297–312; Sartre 1984: 49–61). The best that we can say is probably that it happened around the middle of that year. After this, the next point that we can grasp with any ﬁrmness is the young emperor’s marriage to Furia Sabinia Tranquillina, which the Historia Augusta places in the consulship of Pompeianus and Gordian, or 241 (HA Gord. 23.6). The imperial groom, then, was only 16, or perhaps 17, years old. By this marriage he was allied with Tranquillina’s father Timesitheus (called Misitheus by the Historia Augusta), who became Praetorian Prefect with a leading military and, we may assume, political, role.4 Soon after his marriage Gordian formally opened the doors of the Temple of Janus, signifying a state of war with Persia. The planned campaign, however, was delayed by military operations under Timesitheus in the Balkans, and the emperor and his army did not arrive in the East until at least the end of the summer of 242.5 This Eastern expedition, which was to see the deaths of both Timesitheus and the emperor, is well documented in comparison with the other events of the reign. In addition to the Western sources, which are generally either late or notoriously unreliable for this period (especially the Historia Augusta, nevertheless our best 70

A K I N G I N R O M E ’ S S E RV I C E

source for some purposes), since 1937 historians have beneﬁted from the great inscription of Shapur I at Naqsh-e Rustam, the so-called Res Gestae divi Saporis or ‘Shapur I at the Kaabah of Zoroaster’ (SKZ).6 This serves as an independent check on the historiographers, though not one without its own biases and tendencies. From the Roman and Byzantine historians supplemented by Shapur’s inscription, we learn that Gordian’s expedition was at ﬁrst marked by success. Setting out from Antioch in the spring of 243, the Roman army crossed into Mesopotamia and soon liberated both Carrhae and Nisibis, which had been under Persian occupation for at least four years, and perhaps twice as long.7 Our sources are unclear on whether Edessa had also been occupied, but if it was, it now returned to Roman control along with the other cities. Gordian then successfully confronted the Persian forces in a major battle at Rhesaina east of Edessa (Amm. Marc. 5.17). As far as the occupation of the Roman province of Mesopotamia was concerned, this battle was decisive: Shapur was forced to retreat southwards. When Rome attempted to follow up on this success, however, its leaders met with difﬁculty. To begin with, Timesitheus was killed, to be replaced as prefect by the future Emperor Marcus Junius Philippus (Philip I, ‘the Arab,’ 244–9: HA Gord. 28). The biographer’s charge that Philip hastened the prefect’s death by interfering in his medical treatment is likely mere fancy; but the majority of the Western sources also accuse Philip of treacherously bringing about the disaster that then befell Gordian. In emulation of the feats of his predecessors, Gordian pushed on through Mesopotamia, determined to deal a decisive stroke by assaulting the ancient Parthian capital of Ctesiphon, and perhaps – like Trajan some 130 years earlier – to march as far as the shores of the Persian Gulf.8 For the result we may turn to Shapur’s inscription: When at ﬁrst we had become established in the empire, Gordian Caesar raised in all of the Roman Empire a force from the Goth and German realms and marched on Asuristan against the Empire of Iran and against us. On the border of Asuristan at Misikhe, a great ‘frontal’ battle occurred. Gordian Caesar was killed and the Roman force was destroyed. And the Romans made Philip Caesar. Then Philip Caesar came to us for terms, and to ransom their lives, gave us 500,000 denars, and became tributary to us. And for this reason we have renamed Misikhe Peroz-Shapur.9

71

ROMAN EDESSA

Notably, Shapur lays no claim to responsibility for Gordian’s death, in contrast to his account of the capture of the Emperor Valerian later in the same inscription.10 Gordian therefore seems not to have died at Persian hands, though what did happen is unclear. The most widely believed story in antiquity was that Philip instigated mutiny among the Roman troops by withholding supplies, so that he could beneﬁt from the resulting assassination plot.11 Another strand of tradition, preserved in Byzantine histories, claims that Gordian died as a result of a fall from his horse in the ultimate battle with Shapur, and this may be correct.12 In any case, Philip became the next emperor and made peace with Shapur. Meanwhile, the Romans erected a monument to Gordian at Zaitha near Circesium, where Ammianus (23.5.7) saw it some 120 years later.

Abgar and Gordian: the ‘Presentation’ coinage Despite some gaps and uncertainties, this outline of events is fairly clear. It provides a fairly ﬁrm framework in which to try to locate the reign of Abgar X, and a background for the coins minted at Edessa in his and Gordian’s names (BMC Arabia, etc. 113–14). These came in three denominations and a variety of types, but the most interesting are the large bronzes that show Gordian seated on a stool on a low platform, facing right and holding a lance or arrow, while a standing Abgar presents him with a Victory ﬁgurine (see Appendix, Figs. App. 1–2). The scene is rich in symbolism, derived from both the Roman iconographical lexicon and Eastern religious and political art. The message of the coins at its most basic, however, is the close association between the Roman and the Edessan rulers, identiﬁed in the legends as ΑΥΤΟΚ ΓΟΡ∆ΙΑΝΟΣ and ΒΑΣΙΛΕΥΣ ΑΒΓΑΡΟΣ. In this state of the question, it was inevitable that scholars should wish to see the campaign of 242–3 as the occasion of Abgar X’s investiture, and of this ‘Presentation’ coin type – and nearly all of them have done so.13 The circumstances ﬁt so well that, without any countervailing evidence, the inference is almost unavoidable: what better occasion for such a reception scene than an imperial visit to Edessa, and what better interpretation of the scene’s victory symbolism than as a salute to the emperor’s initial success against the Persian invaders? Hence the campaign of Gordian served many as the anchor-point for dating the reign of Abgar. It is only recently that additional evidence pertaining to this question has come to light, although the earlier work of A. R. Bellinger had already pointed the way. 72

A K I N G I N R O M E ’ S S E RV I C E

By combining the evidence of P. Dura 28, the Edessan legal document discovered at Dura-Europos in the 1930s, with that of the two parchment documents from a district under Edessan control whose discovery was announced in 1989, it is now possible to trace fairly accurately Edessa’s political history during this period. Abgar X, it can now be shown, took the throne probably in 239, or possibly late 238, much earlier than formerly supposed. Moreover, the Dura discovery and the second parchment from the new archive prove that, far from beginning with the emperor’s arrival in 242, Abgar’s reign was probably already over by that time. For by 1 September 242, Year 30 of the city’s ‘freedom,’ his kingdom was styled Colonia Edessa once more.14

Three parchments It is thus necessary to seek a new explanation for Abgar’s reign and his Presentation coinage. We may start by considering the character of his rule, insofar as it can be deduced from the ﬁrst parchment, Document A from the Euphrates Papyri archive (December 240). In the month of December of the year 552, the third year of the Emperor Caesar Marcus Antonius Gordianus, Fortunate and Victorious, in the second year of King Aelius Septimius Abgar, son of Ma nu the pas.griba, son of King Abgar, who was honored with the consulship at Orhai, the baris city which is the grandmother of all the cities of Mesopotamia. This document was written in Haikla¯-Karka¯ of S.ida¯ the New, of King Abgar, on the 28th day.15 This prescript to a contract drawn up during the reign of Abgar X differs in several ways from those of Document B and Dura Parchment 28, legal documents contracted under the colonia, which parallel each other closely.16 Where both of those documents begin with the regnal year of the Emperor Gordian and follow this with the consular dating, then the year of the Seleucid era, the scribe working under Abgar X starts with the Seleucid year. Then comes Gordian’s regnal year, and ﬁnally the year of the reign of Abgar himself, supplanting the consular dates of the other two documents. Where the documents of the colonia seem to include the Seleucid year almost as an afterthought, and explicitly call this ‘the former reckoning,’ that of Abgar’s administration holds ﬁrmly to the old method of counting the years. Also missing in 73

ROMAN EDESSA

Document A, of course, are the magistrates’ years which are found in the other two. Potentially more signiﬁcant than the dating systems employed, however, are differences of language. The epithets given to the emperor in Abgar’s document, ‘Fortunate and Victorious,’ are pure Syriac: GDY WZKY . The imperial epithets in Document B and P. Dura 28, by contrast, are transliterations of the Greek: WSBWS SBST.WS (= εσεβ) σεβαστ ) and WSBWS WT.WKS SBST.WS (= εσεβ) ετυχ) σεβαστ ) respectively. All three documents use the transliteration WT.QRT.WR (= ατοκρα´τωρ) for the emperor. Whereas both of the colonial documents call the city Edessa – transliterating the Greek name into Syriac characters – under Abgar, the Syriac name of Orhai has returned. The use of the Iranian term pas.griba to designate Abgar’s father adds to the ‘Oriental,’ or at least non-Greco-Roman, atmosphere of Document A. Moreover, it would seem that ‘Haikla¯-Karka¯ of S.ida¯ the New,’ which translates as ‘New Hunting Town Palace,’ is the same place as the Marcopolis Thera of Document B, with the name expressed in Syriac (Teixidor 1990: 155–6; Brock 1991: 262 n. 20, 264). In this case, as indeed with the rest of the differences here noted, it is impossible to say which rendering ought to take priority. In other words, we cannot know whether the usage that we ﬁnd in Document B and P. Dura 28 reﬂects the way things were done during the ﬁrst 26 years of the colonia’s existence, or whether these dating practices, and the new name of Abgar’s ‘Hunting Town,’ were introduced only after the re-establishment of the colonia. Considering, however, that the monarchy had been in abeyance for a quarter-century, it seems justiﬁable to talk as if the practices of the royal administration supplanted those which we see in the other two documents, even though this reverses their true chronology. Immediately following the words ‘at Orhai’ (B RHY) in Document A come the letters B RS, which Teixidor takes as baris, meaning castle (Teixidor 1990: 155; cf. Will 1987). However, the letters R and D are the easiest of all Syriac characters to confuse, differing only in the placement of a diacritical mark (where such marks are used; none appear in the texts under consideration). Hence the suggestion that B RS could instead be B DS – ‘at Edess(a).’17 The use of both the Greek and the Syriac name together in this way would be unusual. There may be something to commend this reading, however, for if one follows Brock’s suggestion it enables one to take RBT , ‘great’ with MDYNT , ‘city,’ rather than with M – thus giving a better rendering of the idea of a ‘metropolis.’ This seems in 74

A K I N G I N R O M E ’ S S E RV I C E

keeping with the apparent tendency to translate rather than transliterate Greek terms. The effect is to imply that Abgar’s city is still ‘Edessa Colonia Metropolis,’ even though now it is again under the leadership of its royal family. An even more enigmatic element is the reference to a ‘consulship’ (HPT.Y , a transliteration of the Greek 2πατε!α). We will, however, defer discussion of this mystery until we have completed our analysis of all the remaining evidence. For the moment we may observe that, despite the loss of direct Roman control in this region around 235–8, and despite the temptation that Abgar X must have felt to assert his full independence – perhaps to ally himself to the rising Sassanid star – it seems as if he was fully in the Roman camp. Considering the shortage of information about these events, there has until now been little that could be said about the circumstances of Abgar’s rise to power. One line of speculation has been that he came to the fore in the struggle against the Persian invaders, defending Edessa against their attack, or perhaps even evicting an occupying force – in either case explaining why Gordian III did not need to recapture Edessa along with Carrhae, Nisibis and Rhesaina. In any event, past scholars have interpreted Abgar’s reign in the light of the generally disturbed conditions in the region at this time, with more or less vague references to Abgar’s taking power ‘by right or by force,’ to a ‘royauté de circonstance’ or to the emperor ‘forced to tolerate a fait accompli.’18

239 ce: two rulers meet Documents A and B from the Mesopotamia archive allow us to assert that Abgar had a relatively short tenure on the throne that started early in the reign of Gordian III, and ended before that emperor took the ﬁeld against Shapur in 242. To be precise, the king had a reign of at least a year, and perhaps as much as three years, nine months – from December 238 to September 242, just before the date of the ﬁrst of the ‘colonial’ documents. Another piece of evidence has been more or less overlooked by earlier scholars, although it would seem to have at least some small value for the chronology of Abgar X. Like its Mesopotamian sister cities, Edessa minted coins with obverse types and legends honoring Gordian’s wife, Tranquillina.19 Gordian married Tranquillina in 241, the year before his arrival in the East. This marriage was politically important, as the expression of Gordian’s alliance with Timesitheus – the strategist behind Rome’s resurgence until he died during the Persian 75

ROMAN EDESSA

campaign. By means of the marriage Timesitheus became, if not the heir apparent to the throne, then at least a bearer of the imperial mantle, as the potential grandfather of a future emperor. The alliance with the young Gordianus also strengthened his position in dealing with the Praetorian Guard, who had, like the Roman populus, rioted to show their disapproval of rival candidates until the young scion of the Gordian family was installed (HA Gord. 22.5–6; Max. et Balb. 14.6–8). The important marriage was celebrated on both Roman and provincial coinage, and Edessa was no exception. First-denomination coins from this mint which bear obverse portraits of Tranquillina share a common reverse type with those of Gordian: the veiled and turreted bust of the city-goddess facing left with a small statue on a pedestal before her. Edessan coins of Tranquillina are relatively rare. But the fact that they exist at all highlights the fact that no coin of Tranquillina and Abgar X has been found. This in itself does not mean that none were minted, but the argument from silence is strengthened by the abundance of the coins of Abgar as compared with the colonial issues. The greater comparative volume of Abgar’s surviving coinage suggests that if he had minted any coins with Tranquillina on the obverse, at least a few of them ought to have been discovered. Abgar must have departed the throne, or at least stopped minting, by the middle of 242, when the news of Gordian’s marriage can be expected to have long since reached the East. To state more or less the same thing rather more persuasively (since the king could have had his own reasons to mint for Gordian and not for his wife), the appearance of Tranquillina among the colonial coins under Gordian argues strongly for placing this series after, not before, those of Abgar.20 It is, then, clear for a number of reasons that the coins of Abgar X cannot be as late as the autumn of 242, nor can the meeting of the two rulers portrayed in the Presentation scene on them. If Abgar had indeed seized power in the unsettled circumstances after 235, one might explain his Presentation coins honoring Gordian by supposing that, upon hearing of Gordian’s planned expedition and expected arrival in the East, he issued a prodigious series of coins with ‘anticipatory’ types, justifying his seizure of power in Edessa (or at least making it seem less suspicious) by displaying his loyalty to the emperor. Another explanation might be that the meeting of the rulers portrayed on the coins is merely ‘symbolic,’ having no reference to a real or anticipated meeting.21 It would be preferable, however, to relate the scene to a real meeting between the two, if the 76

A K I N G I N R O M E ’ S S E RV I C E

proper circumstances could be found – in other words, if it can be shown that there is a good chance the king and the emperor did meet, at some point earlier than the normally assumed date. For this purpose, the theory that Gordian III visited Antioch early in his reign seems to hold promise, and may itself derive some support from a re-examination in the light of the Edessan coins. The idea of such a visit was raised tentatively by Xavier Loriot on the basis of a ‘lost’ rescript of Gordian III: a legal text that was not included in the Codex Justinianus but transmitted independently, and whose authenticity, therefore, was (and remains) subject to question (Loriot 1971; cf. Volterra 1971). The rescript reads: Imp. GORDIANVS A. rationalibus Manifestum est nuptiis contra mandata contractis, dotem, quae data illo tempore, cum traducta est, fuerat, iuxta sententiam Divi Severi ﬁeri caducam, nec si consensu postea coepisse videatur matrimonium, in praeteritum commisso vitio potuit mederi. Dat. Kal. april, Antiochiae, Gordiano A. et Aviola coss. The consulship of Gordian Augustus and Acilius Aviola was 239, so from this text it would appear that Gordian was in residence at Antioch on 1 April during the ﬁrst year of his reign. A number of problems surround this text, which apparently was preserved only as a marginal addition in one copy of the Lex Romana Wisigothica. It is possible that Antioch is the location, not of its origin, but of the petitioner to whom it was addressed. Uncertainty also attaches to the date; other cases of probable misdatings on Gordian’s constitutions are known. Nevertheless, the most recent scholar to study this issue concludes, with some hesitation, that ‘on balance it seems that Gordian did visit Antioch in the spring of 239’ (Bland 1991: 490). That statement was made in the context of a study of the mint of Antioch, with the immediate purpose of explaining reports of gold aurei and Adventus types from early in Gordian’s reign at the Antioch mint. The production of gold is usually associated with the imperial presence, and Adventus types are normally interpreted as celebrating an emperor’s arrival in a city. Unfortunately the existence of the Antioch aurei is hard to conﬁrm, since there are only two specimens, both found only in old catalogues, and none has recently been reported.22 The Adventus type on silver coins of Antioch is better documented; it is found on radiates of Gordian’s second tribunician year.23 In the next year, TR P III, it was the Roman mint 77

ROMAN EDESSA

that issued Adventus types for Gordian, this time showing him on horseback with a spear (RIC 4.3, 24 nos. 80, 81; 47 no. 295). As Bland suggests, ‘These could well be associated with a triumphal entry into Rome at the end of an expedition’ (Bland 1991: 490). Since there is no record of any other campaign by the young emperor this early in his reign, the hypothesized expedition to Syria is a likely candidate. All the evidence cited so far, however, remains inconclusive. Bland supplements it with one additional numismatic clue: a tetradrachm from the mint of Alexandria dated to Year 2 of Gordian (Bland 1991: 489). He takes up the argument of Markus Weder, to the effect that this coin – the reverse of which shows the emperor riding down an eastern barbarian (identiﬁable by his ‘Phrygian cap’) – shows that Gordian was on campaign in the East during his second tribunician year, 238–9.24

Gordian’s early movements and Abgar’s role This is plausible, and leaves us with the question of the circumstances surrounding such a campaign. According to the Historia Augusta (Max. et Balb. 13.5), before their assassination Gordian’s predecessors Balbinus and Pupienus were planning two campaigns: one against the ‘Germans’ in the north and one against the ‘Parthians’ in the east. Roman Mesopotamia, after a period of relative calm following the eastern campaign of Severus Alexander, had apparently come under increasing Persian pressure. Edessa’s eastern neighbor, Nisibis, and its southern one, Carrhae, had been occupied since sometime in the reign of Maximinus, if not before, and Hatra – the Arab kingdom on the empire’s eastern fringes that had been, in some sense, allied with Rome since around 235 – was under increasing pressure that was to lead to its conquest and abandonment in 241.25 In the midst of this threatening atmosphere comes a grafﬁto scratched on the wall of a house at Dura-Europos, and dated to 239: κατ-βη "φ µων Π-ρση – ‘The Persian descended upon us’ (SEG 7.743b; Bauer 1933: 112–14). It is, therefore, quite believable that the new regime of Gordian III should have felt the need to ‘show the ﬂag’ in the Eastern provinces. The emperor did not take the ﬁeld against Persia – at least there was no formal declaration of war – but Gordian’s presence in the East might have been intended to serve as a rallying-point, and to let the beleaguered residents of the Eastern Empire know that they had not been forgotten.26 If anything beyond this was planned, the attempt seems to have been cut short, perhaps by even more 78

A K I N G I N R O M E ’ S S E RV I C E

pressing events in the Western Empire.27 The coins apparently celebrating Gordian’s adventus at Rome in 240 make it seem as if he spent a year, at most – probably much less – in the East. For all these reasons – the impossibility of imagining a meeting between Abgar and Gordian in the autumn of 242; the king’s allegiance to the Roman cause; the numismatic clues and the legal text suggesting that Gordian was in the East early in his reign – by far the best explanation for the meeting of the two rulers shown on the coin is the hypothesis that they did meet early in the reign. The range of possible dates for the beginning of Abgar’s reign, between December 238 and December 239, makes it ﬁt comfortably with an Eastern visit by Gordian in the spring of 239 – well enough that this observation should itself be added to the dossier of evidence concerning Gordian’s movements (such a meeting probably would have taken place at Antioch rather than Edessa). It was just at this time, apparently, that a special relationship between Abgar and Rome was established. Exactly what that relationship was, it is extremely difﬁcult to say. Although we may conjecture that Abgar played a military role of some kind, there is no evidence to support this beyond the fact that on his coins he wears a sword. The reference in the Syriac Document A to a ‘consulship’ held by Abgar may, however, hint that the king was given a role in the region’s defense, along with a position in the Roman command structure. The nature of this consulship is unknown, though it is on the face of it plausible that he received the ornamenta consularia, marks of distinction that gave the bearer a ceremonial rank equivalent to that of a consul but that involved no tenureship of ofﬁce. In support of this idea the case of Septimius Odenathus of Palmyra – called λαµπρ τατο 2πατικ (clarissimus consularis) in a number of inscriptions from the 250s – has been cited, on the assumption that Odenathus and Abgar held analogous positions.28 It is, however, quite difﬁcult to translate the phrase concerned with any certainty.29 Only a very small number of examples of allied rulers’ being honored with the ornamenta can be cited.30 Moreover, when such an honor is given it is never termed a 2πατε!α; the term used in Greek texts, rather, is τιµα` 2πατικα´ (Dio 60.8.2 [cf. 78.13.1]; Syncellus 717.20). A clue to the signiﬁcance of Abgar’s Roman title may be provided by two Greek texts on papyrus from the same Euphrates Papyri archive that supplied the Syriac texts, Documents A and B (Feissel and Gascou 1989: 557–8, nos. 1, 3–4). Both of these texts mention equestrian ofﬁcials in administrative positions in the mid-third 79

ROMAN EDESSA

century, and in each case the ofﬁcial is described as δι-πων τ)ν 2πατε!αν, ‘in charge of the consulship.’ In P. Euphr. 3–4 (two copies of the same text, datable to sometime after 252), the ofﬁcial so described is Pomponius Laetianus, also known as a procuratorial governor of Syria from a slightly earlier document at Dura-Europos.31 It appears as if he still held the same position at the time the document was written, although if the document is substantially later than 252 he may have been reassigned. In any case, the petitioners of P. Euphr. 3–4 hope to appeal to him as a higher authority (µε!ζων δικαστ), having failed to obtain satisfaction in earlier petitions to the consular governor of Syria (Feissel and Gascou 1989: 554, n. 76). One of the uncertainties surrounding this text, however, is whether Laetianus is actually viewed as an authority above the consular governor, or whether he has administratively replaced the governor and the petitioners are now placing their hopes in a personal appearance before a judge who is ‘greater’ than the Prefect Julius Proculus to whom Document 3–4 is addressed.32 In the case of the document P. Euphr. 1, of August 245, the ofﬁcial holding the 2πατε!α is certainly a higher ofﬁcial. He is Julius Priscus, brother of the Emperor Philip I and – as the editors observe – one of the most powerful men of his day (Feissel and Gascou 1989: 552–4; Pﬂaum 1960: 2.833–6; PIR2 J 488). In addition to being charged with the 2πατε!α, he is also identiﬁed as the prefect of Mesopotamia, a position which he apparently held from the beginning of his brother’s reign in 244. The petitioners of Document 1, however, address him while he is at Antioch, where he has been performing his administrative duties for at least eight months (ll. 2, 7). Priscus’s ‘consulship,’ then, extended over at least the province of Coele Syria, in addition to Mesopotamia. Comparing it with the 2πατε!α of Laetianus, the editors initially proposed that in both cases, such a non-senatorial governorship was a form of maius imperium which may even have extended over the entire Roman Orient (Feissel and Gascou 1989: 553–4). In the full publication of this document, however, they have reconsidered this proposition and the entire notion of the 2πατε!α. Tracing the evolution of this term they show that 2πατε!α, normally the equivalent of the Latin consulatus, is here used instead for the function of the consularis – the consular governorship (Feissel and Gascou 1995: 81–3, n. 69). We therefore have a parallel for the 2πατε!α of Abgar that involves neither a real consulship nor a grant of consular ornamenta. Like Julius Priscus only a few years later, the king held a position called a ‘consulship’ although he was not a member of the Senate; 80

A K I N G I N R O M E ’ S S E RV I C E

also like him, he administered that position in a region that was on the front line of the confrontation with Sassanid Persia.33 As we have noted, Edessa’s neighbor cities of Nisibis and Carrhae were under Persian occupation, and evidence from Dura bespeaks the instability of the Mesopotamian region at the time of the beginning of Abgar’s reign. All this, along with the clues suggesting that Gordian III had to cut short a visit to the Orient under pressing circumstances, make it appear as if Abgar’s assumption of the 2πατε!α was an administrative makeshift designed to place the critical region of Osrhoene under a strong hand and ensure its loyalty until the emperor could return and set matters right by a decisive victory. Whether or not Abgar himself had any military role is uncertain, but it is possible to interpret his gesture in the Presentation coinage as declaring his intention to do battle on the empire’s behalf. Considering the shaky strategic situation, his ‘consulship’ may have involved some sort of imperium to enable him to hold the line in case of emergency. In this sense, there is some justiﬁcation for seeing his position in 239–41 as preﬁguring the roles of later Near Eastern potentates – Odenathus ﬁrst of all – in the empire’s defense.34

The character of Abgar’s rule The totality of the evidence, such as it is, for Abgar’s role and his relationship with Rome leaves the forceful impression that his reign in no way represented an attempt to regain independence for Edessa. Therefore the contrast that at ﬁrst strikes the eye between the Syriac titles in Document A and the transliterated Greek titulature used for the emperor in the other two Syriac documents probably does not, as might otherwise be thought, signify a deliberate break. The same could be said for the use of ‘Orhai’ in place of ‘Edessa.’ At the same time, the language of the king’s document is certainly intended to remind the reader of Abgar’s royal heritage, with its explicit mention of his descent from Abgar VIII, the Great King. We have no legal documents from the pre-Roman monarchy to compare, but it is plausible that the language of Document A reﬂects the usage of that period. The titulature used for the emperor, and the overall sense of an attempt to avoid transliterated Greek terms, may result from an attempt to combine the forms which had come into use since the creation of the Edessan municipium with the remembered usage of the monarchy. Our ﬁndings concerning the historical background also enable us to read the rhetorical program of the Presentation coinage in its full 81

ROMAN EDESSA

context. The large-denomination bronze coins with the ceremonial scene joining Gordian and Abgar are the most common ones of this size, but there are two other reverse types: one showing the two rulers standing and facing one another and a third bearing the image of Abgar on horseback with his arm raised (Appendix, Figs. App. 3– 4). There is no reason to believe the coins were issued in any particular order, and work on the dies has so far failed to produce such a sequencing; but the whole group makes most sense if interpreted as showing, ﬁrst, the encounter of the two standing rulers, with the emperor in his military paludamentum and cuirass; next, Abgar’s pledge of loyalty to Gordian as the latter (now seated and in civilian toga) prepares to return to the West and leave the war in Abgar’s hands; and ﬁnally, the king’s triumphal return to Edessa, ratiﬁed in kingship by the ruler of the Roman world. By appealing to the emperor’s backing, the series of images was designed to inspire loyalty and conﬁdence in the people of Edessa, who now had the dual status of subjects of the king and citizens of the empire. Royal rhetoric may indeed have tried to preserve the connection with the greatness of the kingdom’s past. Yet, by his coins’ use of this political imagery, by the fusing of monarchical with Roman forms and usage in the legal document of his era, and above all, by his service in behalf of the empire, Abgar X showed how much had changed since the era of Abgar the Great.

82

5 A ‘GOLDEN AGE’? THE CULTURE OF PRE-CHRISTIAN EDESSA

In the course of its history up to the middle of the third century ce, Edessa underwent at least eight changes of government or political allegiance. The establishment of Seleucid rule in the region was followed by the Parthian domination and establishment of the monarchy; the Trajanic invasion and subsequent rebellion; the invasion of Lucius Verus; the rebellion of Abgar VIII and his submission to Septimius; and ﬁnally, the city’s incorporation into the Roman frontier system as a municipality under Caracalla in 212/13. Despite various lacunae, this overall outline of events is fairly clear. Far less clear is the development of Edessan society and culture, which probably both inﬂuenced, and were inﬂuenced by, political and military developments. Rather than being attempts to achieve pure independence, Edessa’s resistance to Roman rule during the second century might be thought to represent a basic afﬁnity with the ‘Orient’ – with the Parthian kingdom to which it had been allied for more than 200 years. Such an afﬁnity is certainly conceivable, given that Edessa during the Parthian period was under the control of a series of kings with Arabic names speaking a Semitic language, rather than a Greek-speaking aristocracy as at Dura-Europos and Seleucia on the Eulaeus. There is no doubt that the Parthian connection exerted a good deal of inﬂuence on Edessan culture. Such inﬂuence is plain in the adoption of Parthian loan-words such as pas.griba in Edessan texts and inscriptions, in the use of the Parthian-style high tiara by Edessa’s kings, and in the clothing and headwear of individuals pictured in mosaic portraits. Parthian names such as Tiridates and Meherdates appear in the rock-cut inscriptions at Sumatar Harabesi and in the Doctrine of Addai.1 Yet, while Abgarid Edessa was far from being in exactly the same Kulturgebiet as Antioch, it still retained a 83

ROMAN EDESSA

link with the Hellenistic culture of the Syrian cities, separated though they were by the political boundary between Parthia and the Roman Empire.2 In an article surveying the archaeological evidence for ‘Classical’ – that is, Greco-Roman – culture in northern Mesopotamia, Marlia Mango stated that the region had absorbed that culture in three ‘waves,’ coinciding with the Seleuid, the Parthian, and the Roman periods of control (M. M. Mango 1982: 117–19). She found substantial archaeological and art-historical evidence that the Classical traditions thus absorbed were continuously preserved up until the time of the Muslim Conquest and beyond. Almost all the evidence examined, however, dates to the period after the arrival of the third ‘wave’ of Classicism with the Roman creation of the provinces of Mesopotamia and Osrhoene, and most of it postdates the end of the Edessan monarchy. Given the state of the archaeological record, this could hardly be avoided. Yet it means that we still lack a broad-based study of the culture under the kings of the Abgarid line. In this chapter I collect various forms of information, chieﬂy documentary and archaeological, to sketch the picture of early Edessan culture in terms of religion, art and archaeology. The impressive surviving corpus of literature, almost all of it produced in the service of the Church, that expressed and interpreted Classical culture in Syriac will also bear on the question, and deserves a separate chapter for its consideration. Both in the case of the physical evidence and in that of the literature, however, the record of the early years is nearly blank. The literature does not start until the work of Bardais.an of Edessa around the beginning of the third century ce. As for the physical record, clues including the documentary remnants of Edessan (pre-Christian) religion, art, architecture and the city’s plan can help us draw the broad outlines of early Edessan society and its culture; but there is, in effect, nothing yet discovered to give direct evidence of the situation before the last third of the second century ce. This fact actually helps to support the idea that the concluding half-century of the Edessan monarchy was a period of prosperity and heightened cultural activity, recalled with admiration during the Christian period. It is perhaps only a slight exaggeration to speak of the last 65 years or so of the monarchy – the period of Abgar VIII and his successors – as Edessa’s Golden Age.

84

A ‘GOLDEN AGE’?

The gods of Edessa The ﬁrst thing that must be done in the attempt to outline Edessan religion is to identify as many as possible of the gods worshipped there. In fact it is not possible to go far beyond this given the state of our evidence, although the literary sources and the archaeological remains do provide some information on questions of ritual practice, personal beliefs and expectations, and the like. Our survey of Edessan religion will concentrate on identifying the gods and drawing connections with the pantheons of other cultures, in the hope that this will help to pin down Edessa’s cultural afﬁnities. In the process we will gather as much information about practices and beliefs as possible. The two earliest texts on the subject, which can give us a general picture of the situation to begin with, are the Bardesanite Book of the Laws of Countries and the Syriac text supposedly preserving an oration by ‘Meliton the Philosopher’ (Syriac PLSWP ), published together in the nineteenth century but possibly of very different dates. Although both are very brief mentions of Edessan religion, they deserve to be given due consideration as the only seemingly contemporary literary descriptions. We shall begin with the Oration of Meliton, supposedly delivered before ‘Antoninus Caesar.’3 Though the date as well as the original language of this work are uncertain (if the oration was indeed delivered before an emperor, the original must have been in Greek), ‘Meliton’ seems well informed about religious practices in Syria and Mesopotamia, for which he uses Euhemeristic explanations as part of an argument against idolworship. Concerning the practices of the region around Edessa, he says, ‘The people of Mesopotamia also worshipped Kutbai, a Hebrew [or ‘Arab’] woman who saved Bakru, the patrician [? Syriac BY ] of Edessa from his enemies.’4 The identity of ‘Kutbai’ is unclear, but since the name contains the root KTB (to write), it has been conjectured that this deity was in some way related to the art of writing.5 It is certain that a deity with a very similar name was venerated as part of the Nabataean pantheon (Strugnell 1959: 29–36; Teixidor 1961: 22–6). Kutbai at Edessa may have been a manifestation of Nebo, who was the god of literature and the arts, among other things. The Book of the Laws of Countries adds to this the statement, ‘In Syria and Edessa ( WRHY) there was the custom of self-emasculation in honor of Tar atha (Atargatis),’ the deity worshipped at a venerable shrine at Hierapolis/Mabbug and known to the Greeks and Romans 85

ROMAN EDESSA

as the ‘Syrian Goddess’ (BLC 58–9). This passage, its context and implications for Edessan culture and the beginnings of Christianity will be more fully discussed later; but for the moment we may take it and ‘Meliton’s’ testimony as preliminary indications of the nature of pre-Christian practice. They seem to show that Edessan religion in this period was similar to that of the surrounding Semitic peoples. The passage from the Book of the Laws is in concord with a number of other sources indicating that Atargatis was one of the most important deities worshipped at Edessa. Except for these and the essay De Syria Dea of the second-century satirist Lucian, which clariﬁes the nature of the Atargatis cult, the only sources that have anything substantial to contribute to the question of the Edessan pantheon are a late Syriac tract known as the Doctrine of Addai, the ‘Homily on the Fall of the Idols’ by Jacob of Serug, and a single oration by the Emperor Julian.6 None of these three sources dates to the period with which we are concerned; Julian, the earliest, dates to the second half of the fourth century, and Addai and Jacob of Serug to the ﬁfth. Each of the three, moreover, has a polemical axe to grind: Addai on behalf of one community of Edessan Christians against another; Jacob for Christianity against paganism; and Julian, of course, in favor of the traditional religions of the Roman Empire.7 This should inspire due caution; yet while we may recognize that each author used the fact of pre-Christian Edessan paganism for his own purposes, there is no easily discernible reason for any of them to have intentionally misidentiﬁed the gods of Edessa. Both Addai (a locally-produced text) and Jacob of Serug (= Batnae, a short distance to the southwest of Edessa) presumably drew on local traditions for their knowledge of the subject. A passage in the former work also suggests that its author had some ﬁrst-hand knowledge of pagan practices. After the mass conversion of the Edessans, the formerly pagan priests among them were said to rush to pull down their altars, ‘except for the great high place which was in the midst of the city’ (Doct. Addai [Howard] 68–9). This may mean that the altar was still standing at the time of composition, and that the author of Addai had seen it. The Chronicle of Joshua the Stylite also proves that at least some vestiges of pagan practice survived to the late ﬁfth century.

86

A ‘GOLDEN AGE’?

Bel, Nebo and the Sun In the Doctrine of Addai, the disciple supposedly sent by Jesus to evangelize Edessa delivers a sermon to the entire community intended to deter it from the worship of idols (Doct. Addai [Howard] 38–69). A passage in this speech, if it contains genuine information, illuminates the nature of Edessan religion: Who is this [man-]made idol Nebo which you worship, and Bel, which you honor? Behold there are those among you who worship Bath Nical, like the inhabitants of H . arran your neighbors, and Tar atha [Atargatis], like the inhabitants of Mabbug, and the Eagle, like the Arabs, and the sun and the moon, like the rest of the inhabitants of H . arran, who are like you. Do not be led captive by dazzling lights or the brilliance of a star because everyone who worships created things is cursed before God. (ibid. 48–9) Among the gods named here, Bel and Nebo are known from the Babylonian pantheon and from Palmyra, where, as is well known, a major sanctuary was dedicated to Bel (Seyrig 1971: 85–114; Seyrig, Amy and Will 1975). ‘Bath Nical’ (the daughter of Nical/Nikkal), the sun, and the moon are deities whose worship was shared with the people of H . arran, according to the speech. All three ﬁt well with knowledge from other sources of paganism at H . arran (Drijvers 1980: 143, n. 50). Tar atha is Atargatis, while the ‘eagle’ is identiﬁed as the god of ‘the Arabians.’ The last sentence in the quoted passage seems to describe an astral cult of the type known at H . arran and elsewhere in the region, though which star is meant is not explicit. The overall argument against the worship of ‘creatures’ is unremarkable, being a standard line used in anti-polytheistic polemic. What is worthy of note here is the number of connecting lines drawn between Edessan beliefs and those of surrounding Mesopotamian and Aramaic communities. None of the deities mentioned is a Greek or Roman one, although to be sure the Greeks were skilled at seeing their gods behind the mask of foreign divinities everywhere. The ‘Homily on the Fall of the Idols’ by Jacob of Serug, celebrating the supposed collapse of pagan religion after the birth of Christ, contains the following lines:

87

ROMAN EDESSA

He [i.e. Satan] put Apollo as idol in Antioch and others with him, In Edessa he set Nebo and Bel together with many others. . . . (Martin 1875: 108; trans. Drijvers 1980: 38). This text agrees with Addai in locating the worship of Nebo and Bel at Edessa; from the fact that they are the only ones named, it appears that these two Babylonian deities were the two chief members in Edessa’s pantheon. When speaking of the Greek gods, Jacob uses the appropriate Greek names, transliterated into Syriac characters: he talks of the worship of Artemis ( RT.MYS) at Ephesus, of Zeus (ZWS) at Rome, of Apollo ( PLW) and other divinities at Antioch, and of Aphrodite ( PRWDYT. ) at Baalbek (Discours 45–65). Therefore there is no need to suppose that the bishop is Aramaizing the names of the gods worshipped at Edessa. Jacob is in fact quite careful in identifying the national gods of many nations, and his testimony concerning the cults of Babylonian gods at Edessa ought to be believed. The notice concerning Edessan religion in Julian’s oration in honor of Helios is problematic. The emperor identiﬁes Edessa as >ερ3ν "ξ α+ωνο 5 Ηλ!ου χωρ!ον, ‘a place forever sacred to the Sun,’ and says that its inhabitants associate with that deity ‘Monimos and Azizos’ (Julian, 150 C–D). These two, he says, are actually Hermes and Ares, and are worshipped as the πα´ρεδροι – assessors or assistants – of the Sun. Numerous scholars, observing that the Sun as supreme deity is more appropriate to Emesa than to Edessa, have assumed a textual error and corrected accordingly: thus Spanheim and, following him, many others including Mouterde and du Mesnil du Buisson (Mouterde 1959: 76–8; du Mesnil du Buisson 1962: 333). Most recently, J. B. Segal adopts this interpretation in his history of Edessa; on the other hand, Drijvers concludes that it is ‘entirely unnecessary to alter Julian’s text’ (Segal 1970: 106 n. 1; Drijvers 1980: 159). It may be impossible conclusively to solve this question, but given the independent witness of the Doctrine of Addai for a Sun cult at Edessa, and the unanimous manuscript witness for Edessa in the text of Julian, the reading should be preserved. In that case, the oration serves as another example of the closeness of Edessan religion to that of the surrounding peoples. For Azizos and Monimos are the Arab gods Aziz and Mun im, the manifestations of the planet Venus as Morning and Evening star (Dussaud 1903: I, 128–42; Dussaud 1955: 142; Henninger 1976: 129–68; Drijvers 88

A ‘GOLDEN AGE’?

1980: 149–53). It is not necessary to pursue Julian’s belief that the gods represent the Greek deities Hermes and Ares, which he says he derives from Iamblichos. More interesting is the association with Arab religion, and the chance of identifying, by means of this reference, the brilliant star mentioned in the speech of Addai. It appears to be Venus. Atargatis and Hadad Among all these deities, the goddess Tar atha/Atargatis has the strongest support from the earlier literary sources and other evidence. In addition to the statement from the Book of the Laws of Countries, the cult’s importance for the people of Edessa and the surrounding region is probably also reﬂected in the statement by Lucian of Samosata, in what is the best source for the study of this cult, that the goddess had devotees ‘from beyond the Euphrates’ (π-ρηθεν του Εφρτεω) – that is, from Mesopotamia – among many others (Lucian Syr. Dea 13). Lucian’s account has the advantages of being nearly contemporary with the period of our concern and of being written by an eyewitness and apparent devotee of the cult, although some aspects of his description are marred by an overzealous interpretatio Graeca. A notable feature of the cult was its emphasis on the spring at Hierapolis and its life-giving waters, celebrated in a festival which involved transporting sea-water to the temple in sealed containers (ibid. 13, 48). The Hierapolis temple complex included pools containing swarms of sacred ﬁsh, which are also found at Atargatis temples elsewhere in the empire (ibid. 45).8 The goddess herself was accompanied by her consort, the sky-god Hadad (whom Lucian calls ‘Zeus,’ while admitting that his local name was different), and by a third deity, ‘Apollo’ or Nebo (ibid. 31, 35). Atargatis was borne on the backs of lions, and Hadad on those of bulls (ibid. 15, 31). The goddess’s cult statue is described as bearing a ‘tower’ (πργο) on its head, among other attributes (ibid. 15, 32).9 Between Hadad and Atargatis in the sanctuary was an image (ξ ανον) unlike any of the others, with ‘no shape of its own’ or even any name of its own; Lucian says the people of the region called it simply the σηµιον, the sign or symbol (ibid. 33). Many of the features of the Hierapolis cult are paralleled at Edessa. Immediately to the north of the town’s Citadel rise the springs which feed a complex of freshwater pools teeming with the carp which were described by the Christian pilgrim Egeria, and 89

ROMAN EDESSA

which still can be seen today. Iconographic evidence includes a stone relief in the Urfa museum depicting a female ﬁgure with a ﬁsh’s tail, which has been interpreted as having some relationship with the cult of Atargatis and its holy ﬁsh (Drijvers 1980: 84, Pl. XXI). This cannot be taken as established, for although Lucian does mention similar images of ‘Derketo’/Atargatis in the cults of ‘Phoenicia,’ he emphasizes that at Hierapolis the goddess was entirely human in form (Dea Syr. 14).10 On another relief in the same museum, a female ﬁgure and a male one are shown ﬂanking a large, roughly conical object with a pointed top and arms extending on either side (Drijvers 1980: 80–3). This object is very similar in its overall shape to one pictured within a temple on the coins of Hierapolis, which is almost certainly the semeion mentioned by Lucian – the aniconic cult object of the temple – or the shrine containing the semeion itself (Oden 1977: 132–3, 160; ﬁgs. 1, 2). The Urfa relief would seem to portray, in an expanded form freed from the restrictions on numismatic art, the scene within the naos of the Hierapolis temple. The only difference is that the lions and bulls are missing on the relief, although they are present on the coins. The third-century coinage of the Edessan colonia depicts a goddess wearing a mural crown, who may well be conceived as the ‘tower’bearing Atargatis in her role as Tyche or patroness of the city.11 If so, it is interesting that, by comparison with the scene on the Hierapolis coins, this image has quite a ‘Greek’ feel about it, and that some or all of the coins use the established imagery for a city Tyche based on an Antiochene model. It is not proven that the goddess on these coins is Atargatis, but her popularity in Edessa makes it seem very possible. The ﬁgure accompanying Atargatis on the semeion relief raises the question of the goddess’s consort. As we have seen, Nebo is mentioned as one of the main gods of Edessa in the Doctrine of Addai and Jacob of Serug’s homily. Neither in late sources nor in contemporary texts, however, is there any mention of Hadad, which might be expected if Edessan religion was closely related to that practiced in the Syrian cult center. It is not easy to explain his absence, but from the late sources quoted above it seems that Bel, whose worship the Abgarids may have brought with them when they took over Edessa, was the primary male deity. It is possible that his importance was too great to allow him to be supplanted by Atargatis’s consort Hadad, the ‘Syrian Zeus.’ It is also possible that the Edessan worshippers of Atargatis conceived of her consort as identical to Bel. Be that as it may, it seems that Bel, Tar atha, and Nebo were the three main gods of Edessa. 90

A ‘GOLDEN AGE’?

Sin, Ba alshamin, Nah.ay Another source of direct evidence for the Edessan pantheon is the collection of inscriptions and rock reliefs at the hillside site of Sumatar Harabesi, which we have considered in prior chapters as evidence of Edessan political and military involvement in this region to the southeast. The dedicatory texts and other inscriptions at this site express their authors’ devotion to MRLH , a divine name that can be vocalized as either Ma¯rila¯ha¯ or Marilahe¯, either ‘the lord God’ or ‘Master of the gods.’ If the former is correct, the reference may be to the ‘Lord of the Heavens’ Ba alshamin, whose worship at Edessa is otherwise attested only by a name in a mosaic inscription (Segal 1970: 55; Segal 1963a: 213). The second reading, however, has been used as part of an argument that the paramount deity worshipped at this site was Sin, the moon-god whose cult was centered at nearby H . arran (Drijvers 1980: 122–45; Green 1992: 65–8). This interpretation is probably correct, since Sin (SYN LH ) is mentioned by name in another text at the site (Drijvers 1972: no. 14, l. 3).12 It is possible that Sin was assimilated to Ba alshamin, at least in the context of the Sumatar cult. Regardless of the identity of the god, however, the evidence ties the cult, once again, to the Semitic environment, with the erection of a baetyl (NS.BT ) for the god mentioned in two of the inscriptions (Drijvers 1972: no. 23 l. 3, no. 24 ll. 4, 9). The walls of a cave below the summit of the Sumatar peak contain several reliefs and identifying inscriptions, which we have studied for their bearing on the events of the second century. In addition to these, to both sides of one of these reliefs are depictions of ‘a horned pillar of oval shape,’ which is very likely the cult object itself (Pognon 1907: 34–50; Drijvers 1980: 130–1; Segal 1970: 58). Some of these individuals, holding the positions of ‘ruler of Arab’ and BWDR of the god, bore names that seem to relate to the royal dynasty of Edessa at a critical time in its relations with Rome. It may be that this cult site was frequented only or mainly by the ofﬁcials whose names and titles appear in the inscriptions, and by their entourages. If so, this collection of inscriptions may be good evidence for the character of the cult of the royal family, but unfortunately it is restricted in two ways. First, the only dated Sumatar inscriptions come from 165, and we are unable to say how long before or after that time Edessan control of the area, and the cult of Ma¯rila¯ha¯ there, lasted. It is, second, unclear how inﬂuential this cult was among the general population of Edessa, Osrhoene or the Arab region.13 91

ROMAN EDESSA

A connection has been proposed between the cult represented by the Sumatar inscriptions, on the one hand, and one of the enigmatic features of Edessan numismatic art, on the other. Among the earliest coins of the Edessan monarchy are small bronze pieces with obverse portraits of King Wael Bar Sahru (163–5) and reverses depicting a pedimental temple containing a cubic form that has been interpreted as a cult object or baetyl (BMC Arabia, etc. Pl. 13, no. 7). While the king is clearly identiﬁed by the Syriac obverse legend, the reverse legend has puzzled scholars. The traces were long thought to correspond most closely to the letters LH LWL, which would mean ‘the god Elul.’ While Elul is well enough known as the Aramaic name for the month roughly corresponding to September, the consecration of this or any month as an object of worship is completely unknown (BMC Arabia, etc. 91, no. 2; Naster 1968: 5–13; cf. Drijvers 1972: no. M2). This has led to repeated attempts to read the letters in other ways, and a consensus is now forming around another reading: the god in question is very likely to be, not ‘Elul,’ but Nah.ay (Drijvers and Healey 1999: 226–7; Segal 1970: 58 n. 5). This deity is already known as an element of some Edessan theophoric names, and as the patron god of the ofﬁcial whose tomb tower was erected at Serrin in 73 (Drijvers 1980: 155–6; 1972: nos. 2–3). The later coins of the colonial era portray a similar temple, engraved in miniature, so frequently that it has been taken as an identifying mark for coinage struck in the city’s mint.14 If this is the same temple as the one on Wael’s coin, Nah.ay probably represents an important ﬁgure in the Edessan pantheon, and one to which the people of Edessa remained devoted long after the short reign of Wael. Because of the date, Drijvers has proposed linking the coin with the contemporary inscriptions at Sumatar, and taking it as honoring the MRLH of that site, either Sin or Ba alshamin (Drijvers 1980: 137). In some Arabic inscriptions, however, a deity with a name similar to Nah.ay appears to be associated with the sun. The coin of Wael’s reign and the continued use of the temple imagery on later coinage may, therefore, conﬁrm the impression that the solar god was an important one at Edessa. Personal names from Edessan documents offer a limited amount of help in outlining the religious situation. Bel appears occasionally as an element of theophoric names, and Nebo less often.15 On one mosaic tomb inscription, the ‘Funerary Couch Mosaic’ (Fig. 5.1), the dedicator of the tomb identiﬁes himself as the son of ‘Bar-Ba sˇmen,’ a name which may contain the divine name Ba alshamin (Drijvers 1972: no. 51, ll. 2, 11, 20; Segal 1963a: 205, 214). The most 92

Figure 5.1 The ‘Funerary Couch’ Mosaic: probably 277/8 ce. In this tomb mosaic, the deceased is shown relaxing at a meal surrounded by his family, a theme very common in the funerary art of Palmyra, the ‘caravan city’ in the Syrian desert. Hand-colored reproduction by Mrs Seton Lloyd; courtesy of J. B. Segal.

ROMAN EDESSA

common divinity to ﬁgure in theophoric names among the ‘OldSyriac’ inscriptions, however, is Tar atha/Atargatis (Syriac T – Drijvers 1972: no. 25 ( BD T ); nos 34, 38, 44, 54 (BR T ); P (MTR T ); no. 48 (SLM T ). Together, these examples conﬁrm the picture that we haveˇ already sketched of the city’s main divinities. This outline of Edessan religion reveals several points of contact with Palmyra, where the deities Bel, Nebo and Atargatis were also revered, with the ﬁrst-named occupying the most impressive temple in the city. The mosaic art and sculptures found in tombs around Edessa conﬁrm this similarity between Edessan and Palmyrene forms of devotion. In one piece of artwork, the ‘Tripod Mosaic’ (Fig. 5.2), the family’s paterfamilias and dedicator of the tomb reaches out a hand holding what looks like a leaf over a tripod, in a ritual gesture very similar to that depicted at Palmyra and in wall-paintings in the Temple of the Palmyrene Gods (Temple of Bel) and elsewhere at Dura-Europos (Segal 1970: Pl. 3; Cumont 1926: 29–168; Breasted 1922: 187–206; Perkins 1973: 114–26). Even more strikingly, several mosaics as well as rock carvings in Edessan tombs depict the deceased at a so-called ‘funerary banquet,’ reclining on a couch with cup in hand, sometimes surrounded by his loved ones – a motif well known from Palmyrene funerary art.16 In an essay on the religious practices of the population of DuraEuropos in the Roman period, C. B. Welles wrote, ‘The natives joined the Macedonians in the worship of Zeus, Apollo, and Artemis, who were in their eyes Ba alshamin, Hadad, and Atargatis – to mention only three possibilities. . . . With the passage of time, the Macedonians and their descendants of mixed blood must have lost all feeling of distinction’ (Welles 1970: 54). Like Edessa, Dura-Europos owed its foundation as a military colony to Seleucus Nicator, and so might be expected to suggest parallels for social and religious development. However, this proposed picture of the evolution of popular religion at Dura cannot hold true for Edessa. All in all, the evidence for Edessan pre-Christian religion indicates that both the city’s common people and its rulers worshipped the ancestral Aramaean and Babylonian deities, with minimal inﬂuence from those of Greece and Rome. The iconography of Atargatis, as described by Lucian, was indeed syncretistic in the number and variety of her attributes. If the goddess depicted on the coins of the Edessan colonia is Atargatis, her Greek appearance could have some signiﬁcance for the cultural afﬁnities of her Edessan devotees, possibly implying that they adopted Greek 94

Figure 5.2 The ‘Tripod’ Mosaic (third century ce). The central ﬁgure performs a religious ceremony reminiscent of another theme in Palmyrene art; the surrounding members of his family hold objects denoting their authority or function in the household. Hand-colored reproduction by Mrs Seton Lloyd; courtesy of J. B. Segal.

ROMAN EDESSA

forms of worship to some extent. On the other hand, the image of the goddess here might be only a result of the conventions of numismatic art. Given this uncertainty and the fact that there is no certainty as to the identity of the goddess depicted, these coins cannot prove anything one way or the other about the afﬁnities of Edessan religion. There seems to be no indication in any of our sources – including the author of the Book of the Laws of Countries and Jacob of Serug, both of whom were certainly familiar with the primary Greco-Roman deities – that the people of Edessa went so far as to subordinate the gods they worshipped to those of Greece. In that sense, there is no element of interpretatio Graeca in Edessan religion. Belief or non-belief in an afterlife, and attitudes toward death in general, are another aspect of religious expression. Funerary art, of course, is relevant to such beliefs, and in this connection it is interesting that two mosaic tomb ﬂoors at Edessa contain Greek mythological themes. They are the beautifully executed ‘Phoenix’ and ‘Orpheus’ mosaics, their subjects with names in . both picturing . transliteration (PNKS, RPWS) amid pastoral scenes and ﬁgured borders which show marked Greek inﬂuence (Figs. 5.3 and 5.4).17 Although it has been asserted that the Orpheus Mosaic ‘points to the existence at Edessa of the cult whose followers observed forms of ceremonial and moral self-abnegation,’ i.e. Orphism, the presence of this artistic motif does not justify such a conclusion, nor does the Phoenix Mosaic necessarily indicate, as Segal also asserts, a belief in the resurrection of the body (Segal 1970: 55–6). It is better to take both images as the expressions of more or less stereotypical sentiments occasioned by the thought of death, drawing on an iconographical stock that was common throughout the Hellenistic and Roman world to console the living, rather than to express a belief in the eventual return of the departed. Each mosaic contains a Syriac text in which the founder of the tomb describes it as a ‘House of Eternity’ (BYT LM ) which he has built for himself and his family ‘for the days of Eternity’ (LYWMT LM – Drijvers 1972: nos. 49, 50). The best evidence for Edessan attitudes to death is found in such expressions as these, and in this text from the Tripod Mosaic: Whoever despises the expectations of [his] last [days] and mourns [his] ﬁrst [days], – he shall have a goodly latter end. (Drijvers 1972: no. 48) 96

Figure 5.3 The Phoenix Mosaic (235/6 ce). The mythical bird is identiﬁed by its Greek name (transliterated into Syriac characters to its left). This, along with the high quality of the workmanship, bespeaks the westward-looking attitude of Edessa’s Hellenizing upper class, as well as their wealth. Courtesy of J. B. Segal.

Figure 5.4 The Orpheus Mosaic (227/8 ce). Here, as in the Phoenix Mosaic, the artist draws on Greek mythological themes to console the bereaved with thoughts of life after death. The accompanying Syriac inscription refers to the tomb as a ‘House of Eternity.’ Courtesy of J. B. Segal.

A ‘GOLDEN AGE’?

Segal interprets this as meaning that ‘only the man who scorns long life and repents of his youthful errors may deem himself worthy of life after death.’ The words about the ‘end’ in the last line, however, may refer to the end of life rather than to life after death; in other words, the inscription would seem to advocate a philosophical attitude toward death, perhaps with no expectation of an afterlife implied. In this case, the exhortation to ‘despise the expectations of [one’s] last [days]’ would advocate an attitude of stoic resignation, a renunciation of the hope of living beyond one’s allotted time. Therefore we cannot conclude from these mosaics, or from the number of funerary inscriptions referring to ‘Houses of Eternity,’ that Edessan religious beliefs, on the whole, included any deﬁnite belief in an afterlife or in resurrection. A ‘House of Eternity’ could certainly mean the ‘house’ in which one expected one’s body to remain for all time.18 The appearance of these Greek mythological motifs, labeled with their Greek names, might support our point about a lack of identiﬁcation between the native deities and the Greco-Roman ones. Again, while the creators of these images and their employers were presumably familiar with the myths they represented, there is no sign here of a melding or syncretism of these themes with ideas of Eastern origin. The assertion that Edessan religion lacked any element of interpretatio Graeca makes it necessary to deal with the case of Lucian, who as we have seen claimed to be a dedicant of the Hierapolis cult, and yet consistently uses only Greek names for the gods worshipped there. If his is an example of the mind-set of a typical worshipper, his essay might cause one to object that Lucian, and perhaps other worshippers of the Syrian goddess, actually thought of the goddess as ‘Hera’ and her consort as ‘Zeus.’ There are two ways to answer this objection. First, since the De Syria Dea was written in Greek, presumably Lucian used the names that would be more familiar to his Greek-reading audience. In the second place, the use of Greek names ﬁts well with what seems to have been the author’s parodic intent. The language of the De Syria Dea is not Lucian’s normal Atticizing Greek, but a conscious imitation of the Ionic dialect of Herodotus, the use of which Lucian elsewhere condemns (Oden 1977: 11–24; Allison 1886: 203–17). This fact has at times been used to attack the Lucianic authorship of the essay; more recently, however, scholars have seen in it an element of satire that, along with other clues, makes the work ﬁt in well with the corpus of Lucian. If the work does parody Herodotus, the fact that its author uses only the names of 99

ROMAN EDESSA

Greek gods – even though he surely knew the true names of the Hierapolis divinities – is easily explained as an aspect of the parody.

Survivals The religious life of pre-Christian Edessa had deep roots, and the old beliefs and practices did not simply disappear with the arrival of Christianity. As we have seen, the Great Altar in the center of town remained as a monument to the old forms of belief. Moreover, a rescript of Theodosius dated to 382 presupposes not only the preservation at Edessa of pagan monuments, but the continued existence of temples and the preservation therein of simulacra, images of the pagan gods (Cod. Theod. 16.10.8: giving permission for the people of Edessa to assemble in a temple, on the condition that rites of sacriﬁce not be performed). The Chronicle of Joshua the Stylite describes an annual springtime festival of lights, in which lighted lamps were brought to line the banks of the river as it passed through the city, and the population delighted in performances of music, dancing and the singing of ‘heathen tales’ (Josh. Styl. 27, 30, 33). The chronicler records these events as evidence of the depravity that he thought was responsible for the many plagues and invasions suffered by the region, and notes with relief the fact that when the time came around for the celebration of the festival in the year 499, a plague of locusts prevented the solemnities from occurring (ibid. 33). The festival described by Joshua is reminiscent in some ways of the ‘Descent to the Lake’ festival at Hierapolis described by Lucian (Syr. Dea 47). At the same time, it parallels the Maioumas water festival celebrated at Daphne and elsewhere in the Eastern Empire, which was harshly condemned as a lascivious pagan rite by Christian moralists including John Chrysostom (In Matt. 7.6: Migne P. G. 57 79–81). The role of lighted lanterns in the festival described by Joshua, however, suggests another connection: with the cult of the brilliant star against which the ﬁctional Apostle Addai inveighed. The ‘dazzling lights’ in Addai’s speech could well be the lanterns of the religious procession in Joshua’s chronicle – another indication that the author of the Doctrine of Addai was well acquainted with elements of the older religion which long survived the ofﬁcial Christianization of Edessa and the empire. Our conclusions concerning those older traditions can be summarized by saying that the rulers and people of Edessa revered a pantheon of deities very similar to those worshipped elsewhere in Mesopotamia and Syria, and connecting lines can also be drawn to 100

A ‘GOLDEN AGE’?

Arab and Nabataean religion. The important Babylonian gods Bel and Nebo occupied prime places in the Edessan pantheon, as in that of Palmyra; nearly equal, if not completely equal, in importance was the ‘Syrian goddess’ Tar atha or Atargatis. There is some evidence, although less than we might expect, of the worship of the ‘lord of the heavens’ Ba alshamin, and it remains a question, as well, whether and under what form Atargatis’s consort Hadad occupied an important position there. As for the cults of heavenly bodies, the moon-god Sin was less important than at nearby H . arran, although some of the evidence from Sumatar points to Sin’s prominence at least at one point during the second century ce. Three textual sources (the oration of Julian, the speech in the Doctrine of Addai and the homily of Jacob of Serug) testify to the worship of the sun and the morning and evening stars. In addition to these afﬁnities with the Semitic environment, we have also seen that at least some elements in Edessan society were familiar with Greek beliefs and mythology. This familiarity, however, did not (as far as we know) have any inﬂuence on actual religious beliefs and practices. What little evidence we do have for practices – the worship of aniconic cult objects like the baetyl seen on the coin of Wael bar Sahru; the practice of self-emasculation in honor of Atargatis – seems decidedly un-Greek. We might conjecture that if Zeus, Aphrodite, or the ‘ancestral gods’ Apollo and Artemis were worshipped at Edessa as at Dura-Europos, it was they that were subsumed under their Semitic counterparts, rather than subsuming them.

The face of the city Another way to gauge the degree of Hellenization or Romanization of a city in the ancient Mediterranean world is to assess its outward conformity with what are normally considered the typical characteristics of a city in each society: the presence of an agora or a forum; public structures such as temples, theaters, gymnasia, administrative buildings, baths and the like; a ‘rational’ if not an orthogonal street plan, and an encircling defensive wall, often including a defensible acropolis within the bounds of the city. To the student looking for such clues, the record of Edessa’s urban development is a metaphor for the condition of all the historical sources. Although there is a relative abundance of information concerning the city and its buildings in Late Antiquity, it is very hard to tell how much of it is relevant to the earlier period, and the continued occupation of the 101

ROMAN EDESSA

site makes archaeological investigation almost impossible. Though some features such as the basic layout of the city walls almost undoubtedly derive from the original Hellenistic ground plan, there is very little evidence, written or physical, for the city’s development between the second century bce and the fourth century ce. In what follows, an attempt will be made to sift that evidence from the relics both of the earliest and of the latest stages, on the assumption that only what can be deﬁnitely identiﬁed as coming from the period of the monarchy gives a true picture of the cultural life of that period. The two columns with Corinthian capitals that still stand on the summit of Edessa’s Citadel are the most prominent relic of the monarchy. They undoubtedly represent the remnants of some kind of monumental architecture on the Citadel, but whether they belong to a palace, temple, or other public building cannot be ascertained in the absence of excavation. The inscription of Queen Shelmath, however, is ample evidence for dating the columns to the monarchy, probably to the third century ce, and thus for the importance of this defensive bastion in that period – even though its fortiﬁcations were eventually deemed inadequate and needed considerable strengthening in the sixth century (Procop. Aed. 2.7.13–16). The nearly sheer northern face of the Citadel mount bears the remnants of fortiﬁcation systems that are still inadequately explored, but which probably contain contributions from the many periods, Hellenistic to Crusader, during which this served as an important fortress. Immediately to the north arises the system of freshwater springs feeding the ﬁsh pools that formed part of the palace complex of the Abgarid kings, and may have been a center also for the worship of Atargatis. Today they water the complex of buildings associated with the great mosque in the shadow of the Citadel’s north face. The shape of the ancient city is also recalled by another waterway: the Kara Koyun, the ancient Skirtos or Dais.an. This stream bed, ﬂowing from west to east and nearly dry in the summer months, runs through the center of the modern town. But the river course actually follows the line of the ancient enclosure wall’s northern side, and is the work of Justinian’s engineers, who deﬂected the bulk of the stream from its original destructive path further south, through what was then the center of the city (Procop. Aed. 2.7.4–10). Beyond these observations, it is hazardous to use the plan of modern Urfa as a guide to the layout of Edessa. Even J. B. Segal, whose plan of the city contains a great deal of informed speculation as to the location of monuments, can offer no more than conjectures in most cases (Fig. 5.5: Segal 1970: 262–3). The street plan of modern 102

Figure 5.5 Plan of Urfa, showing the location of some ancient monuments. The River Dais.an (present-day Kara Koyun) originally ﬂowed through the city by means of sluice gates; its course was redirected to skirt the walls during the Byzantine period. The only pre-Byzantine remains still visible above ground are on the Citadel, in the far southwest corner of the city. Plan from Edessa: ‘the Blessed City’ by J. B. Segal; used by permission of the author.

ROMAN EDESSA

Urfa certainly displays little of the rectilinearity one might ask of a Hellenistic polis; nevertheless, the town’s main north–south thoroughfare intersects a street joining two gates at the western and eastern sides of the city in a central square, now occupied by the Ulu Camı mosque. Segal places here the Church of St Stephen, ‘formerly a synagogue?’, but, given its central location, we might guess that here was found the pagan ‘great high place [altar] in the midst of the city’ of which the Doctrine of Addai speaks (Howard 1981: 68–9). Of Edessa’s encircling defensive walls, substantial sections remain, mostly on the western and southeastern sides. These contain gates which must correspond to some of those mentioned by the Late Antique sources. Like those sources, however, these remains bear witness to what the city became in the ﬁfth century and later, rather than to its condition under the kings. Nevertheless, it can hardly be doubted that the walls and gates stand very nearly in the same places their predecessors did. The attempt to place the features of the ancient city upon the map of modern-day Urfa is frustrated by its continued habitation and ongoing development; yet modern activities can occasionally reveal hidden features. In this way the remains of the hypocaust of what was presumably a municipal bath complex were revealed during construction work on the city’s water system in the summer of 1959. According to the report of J. B. Segal, who saw the hypocaust ‘brieﬂy’ before it was covered up again, these baths were located ‘not far north of the ﬁsh pools’ in the city’s southern quarter (Segal 1970: 32 n. 1).

A picture of the king’s city The difﬁculties in unearthing the archaeological record again highlight the importance of the literary sources, which, although mostly later than our period, are not universally unreliable. One invaluable source is the opening section of the Chronicle of Edessa from the city archives. This passage describing a major ﬂood in the city is the longest preserved item in the chronicle, and opens a fascinating window on some features of the city as it appeared in the latter decades of the monarchy. We shall quote it here in extenso, in the translation of J. B. Segal (Segal 1970: 24–5; Chron. Min. 1–3/3–4): In the year 513 in the reign of [Septimius] Severus and the reign of King Abgar, son of King Ma nu [201 ce], in the month of the latter Teshrin [November], the spring of water 104

A ‘GOLDEN AGE’?

that comes forth from the great palace of King Abgar the Great became abundant; and it rose abundantly as had been its wont previously, and it became full and overﬂowed on all sides. The royal courtyards and porticoes and rooms began to be ﬁlled with water. When our lord King Abgar saw this he went up to a safe place on the hill, above his palace where the workmen of the royal works reside and dwell. While then the experts [‘wise men’] were considering what to do about the excess of waters which had been added, there took place a great and abundant downpour of rain during the night. The [river] Dais.an came before the usual time and month and foreign waters came, and they found the sluices closed with large plated iron [bars] and with reinforced iron bolts. Since no ingress was found for the waters, a great lake formed outside the city walls and the waters began to descend between the battlements of the walls into the city. King Abgar standing on the great tower called ‘[the tower] of the Persians,’ saw the waters by [the light of] burning torches and ordered that the gates and the eight sluices of the eastern wall19 of the city should be removed from [the place] where the river came out. But at that very moment the waters broke down the western wall of the city and entered into the city. They destroyed the great and beautiful palace of our lord king and removed everything that was found in their path – the charming and beautiful buildings of the city, everything that was near the river to the south and north. They caused damage, moreover, to the nave [Syriac haikla] of the church of the Christians. In this incident there died more than two thousand persons; while many of them were asleep at night, the waters entered upon them suddenly and they were drowned. When the city was full of the sound of wailing and when King Abgar had seen this damage that had taken place, he ordered that all the craftsmen of the city should take away their booths from beside the river, and that no one should build a booth for himself beside the river; through the expert [skill] of the surveyors and knowledgeable men, the booths were placed as far as the breadth of the river [allowed] and they added to its former measurements. For even though the waters were great and abundant, the actual breadth of the river was small; it received the waters of twenty-ﬁve streams in their conﬂuence from all sides. King 105

ROMAN EDESSA

Abgar ordered that all those who resided in the portico and carried out their occupation opposite the river should not pass the night in their booths from the former Teshrin to Nisan, but that all the winter time ﬁve of the geziraye who guard the city should pass the night on the wall above the place where the waters enter the city. When at night they observed and heard the sound of foreign waters beginning to enter the city and . . . Whoever heard [this] sound and was negligent and did not go out [and shout], ‘Behold the waters’ would be punished for contempt because he had despised the order of the king. This order was instituted from the time when the event happened in this wise until eternity. But our lord King Abgar ordered a building to be built as his royal dwelling, a winter house [in] Beth Tabara – and there he used to dwell all the winter time; in the summer he would go down to the new palace that had been built for him by the source of the spring [of water]. His nobles also built for themselves buildings as dwelling places in the neighborhood in which the king was, in the High Street called Beth Sah.raye. In order that the former tranquillity of the city should be established, King Abgar ordered that unpaid taxes from those who were inside the city and from those who dwelt in the villages and on farms should be remitted, and that taxes should be suspended from them for ﬁve years until the city had grown rich in its population and adorned with its buildings. Maryahb son of Shemesh and Qayoma son of Magrat.at. – these scribes of Edessa wrote down this event at the order of King Abgar, and Bardin and Bulid who were in charge of the archives of Edessa received it and placed it inside them as sharrire of the city. The clear and concise style, along with the absence of tendentious passages, in this narrative – the most extensive of the preserved sections of the Chronicle of Edessa – mark it as a genuine extract from the Edessan city archives.20 The sequence of events that it narrates is quite clear: after early and unexpected winter rains, the Dais.an, overcoming the normal ﬂood control measures, burst into the city from the west and backed up against the closed sluice-gates at the eastern exit, which had failed to be opened in time. The narrator is in error in saying that the king ordered the western gates to be opened ﬁrst, since to do so before opening the eastern gates would 106

A ‘GOLDEN AGE’?

have led to the ﬂooding of the city even if the waters did not surmount the battlements. The devastation was the result of a lack of vigilance, which the king attempted to correct by the posting of geziraye (watchmen) at the river-gate. Nevertheless, the city was to suffer again and again from the river’s effects; even the hydraulic works of Justinian in 525 (Procop. Aed. 2.7.4–10) could not prevent serious reoccurrences in the succeeding centuries. The narrative is of most use, however, as a record of the city as it appeared during the reign of Abgar the Great. In the hours preceding the great ﬂood, the copious springs of the Edessan oasis mounted to an unusual, though not unprecedented, rate of ﬂow, ﬂooding the buildings of the palace complex. This conﬁrms the hypothesis that the king’s dwelling at that time – the royal ‘courtyards and porticoes and rooms’ – centered around the pools just north of the Citadel. The Citadel itself – the ‘hill above his palace’ – was, at this time, neither a royal abode nor, apparently, a place of any particular distinction. Rather, it was the district containing the homes of ‘the workmen of the royal works.’ It was here that the king repaired to consult with his engineers on measures to take in the light of his palace’s inundation, which was inconvenient but not a matter of urgency – unlike the events that ensued. King Abgar viewed those events from the ‘Tower of the Persians,’ the name of which suggests that it was a defensive bastion erected during the era of Parthian domination. Its location is not speciﬁed, but such a structure would very likely have been a part of the Citadel’s fortiﬁcation system, possibly on the higher ground at the western end of the acropolis; this location also makes sense given the king’s earlier movements. From here the king saw the light of the torches gleaming on the surface of the water rising outside the western sluice-gates, less than 500 meters to the north; his attempt to take precautionary measures, however, came too late. The ﬂood destroyed the king’s own dwelling, his ‘great and beautiful palace,’ so the river’s point of entry into the city must have been toward the southern end of town near the springs. It continued its destructive path through the city, sweeping away structures on both its north and its south banks and killing more than two thousand. We may guess that this ﬁgure includes both members of the nobility whose houses, built in proximity to the palace, were among the ‘charming and beautiful buildings’ there, and members of the artisanate whose custom it apparently was to sleep in their booths by the riverside. In the case of the craftsmen, the king attempted to prevent future tragedies by outlawing this custom 107

ROMAN EDESSA

henceforth and relocating all booths as far away as possible from the riverbank, even though the preventive widening of the river bed may have made this difﬁcult. In the case of the nobility, however, their relocation was voluntary; all that was necessary was for the king to move his own residence. The location of ‘Beth Tabara,’ his new winter palace, is unknown, as is ‘Beth Sah.raye’; it has been reasonably conjectured, however, that the king took up residence on the high ground of the Citadel, with the nobility nearby. At the same time, the lower palace among the pools was rebuilt, a welcome refuge in the hot summer months. King Abgar thus showed himself to be deserving of his epithet ‘Great,’ being as concerned with his subjects’ welfare as with his own comfort. His efforts at recovery went beyond construction and engineering, for he granted both forgiveness of back taxes and a ﬁveyear tax suspension, until the city should regain some of its former prosperity. The fact that the king was able to be so magnanimous suggests that Abgar ruled during a period of security and increased prosperity, whatever the reason. The ﬂood narrative includes a brief but signiﬁcant statement to the effect that one of the buildings damaged was ‘the church of the Christians.’ Although accepted by most scholars, this has been challenged by some as an interpolation, the apparent aim of which would have been to provide evidence for the early Christianization of Edessa.21 The challenge is partly justiﬁed by the fact that the Christian building would appear to have been singled out for mention; the only other building speciﬁcally named as having suffered in the ﬂood is the royal palace. This brief passage, however, does not serve the purposes that a fabrication might have been intended to accomplish. As we shall see (Chapter 6), the later controversies among different factions of Edessan Christians produced works of ﬁction that gave one group or the other a claim to priority. The Doctrine of Addai was intended not primarily to prove that Christianity came to Edessa at an early date, but to buttress one particular group’s claim to special authority. A generalized reference to ‘the Church of the Christians’ does not serve such a purpose. If it could be established, however, that one particular church near the river was associated with one of the factions, for instance the Marcionites, the case would be different. In that case the insertion of a reference to the early existence of ‘the Church of the Christians’ at that location might well have served the purposes of the followers of that group.22 There is good cause to believe, on the basis of independent data, that Christianity of one form or another had indeed been established 108

A ‘GOLDEN AGE’?

at Edessa before the end of the second century. The ﬂood narrative’s mention of the church is consistent with the existence of such a group, but does not imply that it had any ofﬁcial or preferred status.23 We will provisionally accept the evidence that in the days of Abgar VIII a building existed near the river that was recognized by some as the Christian church of Edessa, though the precise nature of the Christianity practiced there is unknown, and although there may have been more than one such congregation.

Late records The evidence of the ﬂood narrative for Abgar’s ‘summer palace’ at the pools below the Citadel receives conﬁrmation from the account of the Spanish pilgrim Egeria. Because this traveler visited Edessa toward the end of the fourth century ce, her description can, in general, only be taken to apply to the features of the Late Antique city. At one point, however, she speciﬁcally describes the ‘the palace of King Abgar,’ which contained a statue of the king and whence arose the springs that fed the pools containing the holy ﬁsh.24 Egeria believed this palace to have been the home of Abgar Ukkama, who corresponded with Jesus according to the legend. If that king had a palace on this site, however, it was destroyed or seriously damaged by the ﬂood of 201, if not earlier. The palace which Egeria saw, which also contained a beautiful statue of the king’s son ‘Magnus,’ was almost certainly the one built by Abgar VIII, who had a son named Ma nu. The descriptions of the pools and the sacred ﬁsh they contained support the idea that in or adjacent to the palace-pool complex was the temple of Atargatis, whose worship is attested at Edessa and whose cult at Hierapolis displayed just these features. The inviolability of the ﬁsh in earlier tradition was not carried forth into the Christian period, for Egeria (Itinerarium Egeriae [Arce 1980] 19.6–7) speaks of the ﬁshes’ incomparable taste as well as their size and beauty. Older than the account of Egeria but still belonging to the postmonarchy period is the seemingly genuine narrative of the martyrdom of Shmona and Guria. This, one of the earliest genuine documents of ‘Roman’ Edessa, narrates the reluctant persecution of the Edessan ‘confessors’ by Mysianus, the Roman governor (γεµ,ν – Syriac YGMWN ) of Edessa under Diocletian. In it we learn that the administration of the city was carried out at an ofﬁcial building called the Basilica (BSLYQ , Shmona and Guria [Burkitt 1913] 41), which was located near the ‘Winter Baths’ (BL NY STWYT ).25 109

ROMAN EDESSA

These structures were apparently located in the western section of the city, since after ﬁnally ordering the martyrs’ execution the governor has them hastily removed from town by the Western Gate (ibid. 56). In them we see, a century after the events in the ﬂood narrative, evidence of the transformation of Abgar’s capital into a Roman provincial city like any other. The process presumably began not long after the death of the great king, upon his son’s removal, the foundation of the municipium of Edessa and the appointment of the ﬁrst γεµ,ν in 212/13 (Jacob of Edessa [CSCO Scr. Syr. 3, vol. 4] 282). As with almost all else, the source of the greatest abundance of information for the face of the city is the literature of the Christian period. We learn, for instance, that outside Edessa on its northwest side stood a hippodrome (Procop. Persian War 2.12; Aed. 2.7.9–10), and that on its eastern side near the outlet of the river was a theater (Josh. Styl. 27). Joshua the Stylite says that the lamps of the pagan festival were hung ‘in the stoas and the Antiphoros’ (B STW WB NTYPWRWS) among other places (ibid.); the latter structure, also mentioned by Procopius (= καλοµενο Αντ!φορο, Aed. 2.7.15), may have been so named because of its location across from the city’s forum.26 Joshua also mentions both the summer and the winter baths, the inns, colonnades and the ξενοδοχειον (Josh. Styl. 42–3). If it could be shown that some or all of these features existed at an early date, they would be important evidence for the inﬂuence of Greek or Roman cultural values at that time. Duval, on the assumption that Abgar VIII was the ﬁrst Christian king, dates to his reign the construction of the hippodrome which Procopius tells us was the gift to the city of Augustus (Duval 1892: 100; Procop. Aed. 2.7.15). Since both the story of the hippodrome and this king’s Christianity are questionable, there is no need to keep this assumption; the supposed hippodrome may – like most such structures throughout the Roman Orient – have been of a much later date.27 In conclusion, until some of the urban features mentioned in the late sources can be ﬁrmly located and dated by excavation, they cannot be used as evidence for deciding the cultural question. Some features such as the overall city plan may be considered survivals from the Hellenistic period, and we may assume that the city as ﬁrst laid out contained features such as public buildings, open spaces and temples, and that some of these probably remained in use during the monarchy, as did the Hellenistic-era walls. The description of Abgar’s palace complex in the Chronicle of Edessa, with ‘royal courtyards and porticoes ( STW or stoai) and rooms’ centered around the 110

A ‘GOLDEN AGE’?

springs of the Edessan oasis, is not speciﬁc enough to allow any conclusions as to the style of architecture employed, but it would not be surprising if this Romanophile king, the client of Septimius Severus, had a palace built in a ‘Classical,’ Western style. Beyond these minimal conclusions, we may seek some guidance from the few remnants of Edessan art that have been preserved and studied.

The art of the kingdom As we have seen, the Phoenix and Orpheus mosaics offer a picture of what Mango calls the ‘Edessan school’ of mosaics (Mango 1982: 118). They are, however, among the most Classicizing of all the Edessan mosaics. The pavement containing a supposed portrait of Abgar VIII from a tomb in the S¸ehitlik Mahallesi area, which is the most recent one discovered and published, provides a good illustration of what Mango calls a ‘nonclassical element in Edessene art’ (Drijvers 1982: passim). This scene decorated the ﬂoor of a tomb prepared for the family of one Barsimya, an Edessan noble. It contains portraits of ﬁve individuals in two registers, an arrangement similar to that of the previously discovered mosaics of Belai Bar Gusai and that of Aptuh.a (Fig. 5.6; cf. Segal 1972: Pll. 16b, 17a). All ﬁve are identiﬁed by their names in Syriac: ‘Barsimya son of Asˇdu,’ ‘Abgar son of Ma nu,’ ‘Asˇdu son of Aqrab,’ ‘H . anan son of Asˇdu,’ and ‘ ZL mother of Barsimya.’ Although there is no date, the form of the Syriac script and the resemblance to similar mosaic scenes which bear dates in the 220s and 230s suggest a date in the ﬁrst half of the third century. The mosaic includes a central inscription, ‘For the life of my lord Abgar Bar Ma nu,’ under the central portrait in the upper register, which makes plausible the connection with Abgar VIII, whose father was named Ma nu (Drijvers 1982: 182–3). This identiﬁcation is certainly not beyond question. If the Abgar Bar Ma nu portrayed in this mosaic were a king, one would expect to see him wearing the high tiara, unmistakable symbol of kingship in Edessa as throughout the Parthian Near East (see the treatment of this issue in the Appendix). Instead, the Abgar in this portrait wears a soft tiara, similar to those of the other males portrayed. Abgar’s tiara, however, is depicted in a lighter color, perhaps indicating that it was made of precious materials. Abgar’s beard contains prominent stripes of white, which would be consistent with an image of Abgar VIII toward the end of his long reign. There was, however, another King Abgar, son of Ma nu, in this period – Abgar X, grandson of 111

ROMAN EDESSA

Figure 5.6 The Mosaic of Aptuh.a (undated). Like a number of other Edessan mosaics, this one consists of busts of the deceased and his family in two (sometimes more) registers. This mosaic is preserved at the Istanbul Museum, having escaped the destruction that has greeted almost all other Edessan mosaics (including that dedicated by a Barsimya to ‘Abgar bar Ma nu’, not illustrated here). Courtesy of J. B. Segal.

the great king and last of his line. Abgar X had only a short reign, but as a member of the royal family and son of Ma nu the Crown Prince, he was probably an important enough individual, even before his coronation, to have served as the lord and benefactor of a noble such as Barsimya. For that reason, we may guess that this mosaic portrays the future Abgar X, in his soft tiara, before receiving the crown at the hands of Emperor Gordian III. This would place the scene somewhere in the 230s, which ﬁts very well with what is known of the dates of all other similar mosaics.28 The portraits, of four bearded men and a woman with a tall headdress, are presented in a frontal style that has in the past been taken as characteristic of ‘Parthian art,’ but which is in reality a feature of the art of this region of the Near East under both Roman and Parthian rule.29 The men in the upper register – the patron Abgar, Barsimya and Barsimya’s father – all wear ﬂowing robes along with their soft tiaras. The mantle over Abgar’s shoulders is ﬁgured in bold, brightly colored patterns, and his right hand crosses his chest and holds a staff or scepter, presumably as a symbol of royalty, while 112

A ‘GOLDEN AGE’?

Barsimya holds a round object in a similar position: possibly the royal seal or another indicator of his close relationship with the royal ﬁgure (Drijvers 1982: 182). Under the central portrait of Abgar is the foundation inscription: ‘I Barsimya son of Asˇdu made for myself this house of eternity, for myself and for my children and for my brother, for the life of Abgar, my lord and benefactor.’ Although Abgar was not to be interred in this tomb (nor were Barsimya’s parents, pictured on the right side of the scene), the structure was ‘for [his] life’ ( L H . YY), that is, in his honor. Examples of other dedications by one person ‘for the life of’ another person can be found at Sumatar Harabesi and elsewhere in Semitic epigraphy, as at Hatra and Palmyra.30 It is rare that such a formula occurs in a tomb inscription such as this; its occurrence here suggests that Barsimya considered his relationship with Abgar to be of paramount importance. All around the portraits in the scene runs a decorative border with geometric patterns that recall the Antiochene ﬂoor mosaics of Severan and later date.31 Like the Phoenix and Orpheus mosaics, which have similar borders, these motifs highlight the availability of artistic models from the ‘Greek’ milieu of Roman Syria to artists in the Mesopotamian principality. On the other hand, the costumes and hieratic poses of the busts in the Abgar mosaic, as well as of the fulllength ﬁgures in the Family Portrait (Fig. 5.7), Funerary Couch (Fig. 5.1) and Tripod (Fig. 5.2) mosaics, tie them in to the Mesopotamian cultural context. The language of all the inscribed texts in the mosaics is Syriac, in the particular ‘proto-Estrangelo’ script of the third century ce. The ‘funerary banquet’ poses of the families in the mosaics have their closest parallels in the funerary art of Palmyra. In general, despite the presence of some ‘Greek’ artistic motifs, the third-century Edessan mosaics reveal, not Hellenism per se, but a culture that was under a great variety of inﬂuences. They were produced for patrons who were clearly in the upper ranks of society, and they served as expressions of those patrons’ pride in their social position. The rich, almost ostentatious, imagery drawn from the Greek tradition can be seen to support the assertion of a high social position by patrons who could afford to hire artists trained in the Greek mosaic tradition, and who may have been imported from the Roman province to the west. Seen in this light, the third-century date of the mosaics (which is ﬁrm in some cases and probable in all of them) takes on a certain signiﬁcance.32 As a body, they ﬁt in with the picture of a cultural renaissance accompanying Edessa’s increased economic and strategic 113

ROMAN EDESSA

Figure 5.7 The ‘Family Portrait’ Mosaic, found in a tomb at Urfa and probably from the mid-third century ce. The costumes and poses of the individuals in this and other Edessan mosaics tie them to the Near Eastern cultural context, whereas the workmanship and the geometric border motif display clear inﬂuence from neighboring Roman Syria. Hand-colored reproduction by Mrs Seton Lloyd; courtesy of J. B. Segal.

importance in and after the reign of Abgar VIII. Just as the king himself cultivated the association with Rome and patronized Bardais.an the Syriac interpreter of Greek philosophy, the nobles of his court – of whom the funerary art is virtually the only record – incorporated elements of Western artistic traditions in their memorials. Here again, however, the Edessan tradition is eclectic. Although it begins to adopt a Greek ‘accent,’ Edessa still expresses itself in Syriac. A comparison between the earliest native Edessan coins and those of the colonia, toward the end of autonomous Edessan coinage, shows some of the signs of a shift toward Hellenism. The ﬁrst coins, those of the pro-Roman King Ma nu and the Parthian client Wael Bar Sahru, bear legends in Syriac, and obverse portraits resembling those of the king with diademed tiara on Parthian coins.33 The temple of ‘the God Nah.ay (?)’ on the reverse of one coin is vaguely Greek in style, but the cubical object of worship portrayed inside it conveys a distinctly non-Greek feeling. The bust of the Parthian king himself (Vologaeses III?) which appears back-to-back with that of the Edessan ruler on some coins is another non-Greek element (BMC 114

A ‘GOLDEN AGE’?

Arabia, etc. 91, no. 1). By contrast, when the ‘free’ municipality begins to mint, beginning in 212/13, the reverse type is based on a thoroughly Greek model, the Tyche of Antioch. Even though this goddess may have been conceived as the ‘Syrian Goddess,’ Edessa’s patroness Tar atha/Atargatis, the use of this model for her portrait testiﬁes to the power now exerted by Greco-Roman artistic models, at least in the somewhat restricted genre of numismatic art. As one would expect of such a process, the shift was not a smooth progression. It is upon the reinstallation of the pro-Roman Ma nu after the invasion of Lucius Verus that Edessan coins begin to be inscribed in Greek, which remains their language until the end.34 On some of them Ma nu takes the appellation ΦΙΛΟΡΩΜΑΙΟΣ, ‘Friend of the Romans’ (BMC Arabia, etc. 92–3, nos. 5–9). The switch to the use of Greek on the coins thus parallels, and may be caused by, political changes. The presence of the Roman emperor’s portrait on all Edessan coins after Ma nu, which of course also reﬂects politics, brings the appearance of these coins more into line with that of other provincial cities’ coinage. The portraits of the Edessan kings, however, including that of the strongly pro-Roman Abgar X, always retained the strong resemblance to those of the kings of Parthia. In general, the art of the second and third centuries shows both Western and Eastern inﬂuences, with no single element dominating. It is impossible to say for certain whether this period was more ‘Hellenized’ than the period before it, since there is even less evidence for the ﬁrst century than for later. Edessa was part of the whole cultural sphere of northern Mesopotamia which was subject to successive waves of inﬂuence under the three different imperial powers that ruled there between the fourth century bce and the end of the Roman domination, so that Greek traditions always held some sway. The ﬁnal wave, characterized by the Classical motifs on Edessan mosaic pavements and the Corinthian capitals of the Citadel columns, arrived in the wake of the Roman alliance. Even in this period, Hellenism in Edessan art has a distinctly Semitic/Parthian ﬂavor, so that it may not be too much of a stretch to imagine that, before the arrival of Rome, that ﬂavor was much stronger. In art as in other respects, it is certainly not justiﬁed to say that Edessa ‘was as Hellenized as the rest of Syria and the Roman Empire in general.’35

Conclusion Although they lived inside the walls, and walked the stoai, of Hellenistic Edessa, the people of Parthian Edessa lived in cultural 115

ROMAN EDESSA

communion – as represented most of all by the evidence for their religious beliefs – with the peoples of the desert and elsewhere in the Semitic Middle East. Like Romanization, however, Hellenization is a social as well as a cultural issue. The Edessan elite, like the aristocracy of many another provincial city under the Roman Empire, found it in its interest, after the consolidation of Roman rule, to adopt Roman names and titles, and participate in a newly reorganized local government along Roman lines. The wave of what might be called ‘Westernization’ had begun with Abgar VIII, Rome’s client. These nobles depended on the king for their standing, as is shown by their erection of new homes near his hilltop palace after the ﬂood of 201 ce, and by the king’s importance in the mosaic of Barsimya. The adoption of Greek motifs and imagery by the mosaic artists the aristocracy employed, as well as the transliterated names of Orpheus and the Phoenix, bespeak an awareness of and admiration for Greek culture, and its use as a symbol of high social standing. The fact that the names were transliterated rather than written in Greek underlines the message that is conveyed, however faintly, by Edessan religion and the other physical remains: that while an appreciation for the Hellenic culture that dominated the Roman East was growing stronger by the end of the second century, in language and in outlook Edessa remained a Mesopotamian city for many years after ofﬁcially becoming a ‘Roman’ one. The changing strategic situation and resulting economic changes can also be seen as being among the causes, if not the primary causes, of cultural changes. The erasure of the political boundary along the Euphrates by Septimius Severus meant that Edessa, which must always have been to some extent dominated by Antioch, came even more strongly under the inﬂuence of the Greek-speaking metropolis of Syria. Trafﬁc between the two cities must have increased greatly, especially if the eastern trade was beginning to follow the northern route through Osrhoene more frequently. Above all, direct Roman military involvement in the region as witnessed by Severus’s creation of the Parthian legions, and most probably the subsidizing of the Edessan king, brought the prosperity that can so often lead to a boom in cultural expression. It may be in this sense, ﬁnally, that it becomes appropriate to think of Edessa after about 200 as indeed an Eastern version of Athens, which long before proved the importance of economic resources in sustaining cultural activity.

116

6 EARLY CHRISTIANITY AND EDESSAN CULTURE

In the year 2046, Abgar, the king of Edessa, addressed a letter to Jesus Christ in the country of Jerusalem. (Chron. Zuq. 4/5) In these words the Chronicle of Zuqnin tersely reﬂects the legend that gave Edessa empire-wide fame before the end of the fourth century. According to the legend a King Abgar (supposedly Abgar V Ukkama, Jesus’s contemporary) wrote to Jesus in Jerusalem asking to be healed and inviting Jesus to visit Edessa. The text of that letter, and Jesus’s reply, exist in many versions in almost all the languages of the Roman Empire, in keeping with the belief that the texts themselves had a sanctifying and protective power. The importance of this legend for the reputation of the city illustrates the central fact of the post-monarchical period: regardless of preChristian Edessa’s primary cultural orientation – whether it was to the East or to the West – the crucial factor in its later identity was its prominence as a center of Mesopotamian Christianity – the ‘First Christian Kingdom’ or the ‘Blessed City’ – and it was this factor that preserved the name and status of Edessa through the Byzantine Empire and beyond. The origins of Edessan Christianity have been hotly disputed, one reason being that there is no contemporary evidence. The only solid terminus that can be assigned is the Chronicle of Edessa’s mention of the Church of Christians, existing in Edessa at the time of the great ﬂood of 201. Another indicator is the Abercius Inscription, a text discovered in Phrygia and dating to before 215 ce (Grégoire 1955– 7; Bundy 1989, 1992). In this inscription Abercius talks of having traveled as far east as Nisibis, and encountered fellow Christians everywhere along the way.1 Therefore Christianity had arrived in the region, and almost certainly in Edessa itself, by the second half of 117

ROMAN EDESSA

the second century. There is no need to assume that it had any ‘ofﬁcial’ standing at this time, or any connection with the royal family. The story of Abgar’s conversion, and all of the written accounts that purport to tell the story of the city’s evangelization, tell us more about the theological disputes of the third and fourth centuries, when they were written, than about ﬁrst- and secondcentury events. For this reason, in looking at Edessan Christianity as an indicator of the social and cultural milieu of the late monarchy, we shall for the most part sidestep the entire problem of origins. Toward the end of this chapter, however, we shall look at some of the proposed motives for the invention and development of the legend of Abgar, Addai and Jesus. Edessan Christianity was always marked by controversy; our sources record more than one instance of energetic and at times violent confrontation between the adherents of rival theological schools. This disputatiousness, which reﬂected the controversies over orthodoxy and heresy in the Roman Empire as a whole, continued into the later period. We must assume that a large amount of theological literature produced in this period has simply disappeared, since it was deemed offensive to orthodoxy. The few heterodox materials that survive, however, are highly suggestive; they can be supplemented with the aid of orthodox polemics. The period with which we are concerned, roughly the ﬁrst two to three centuries of the Common Era, saw a number of attempts to encompass Christianity within the context of existing (pre-Christian) cultural and belief systems, and at Edessa the results of some of these attempts had a lasting impact. In particular, the complex of philosophical and religious ideas collectively known as Gnosticism had a strong appeal in Edessa and Syro-Mesopotamia generally, and underwent some of its most interesting later mutations there. Some early Syriac literature, including what remains of the teachings of Bardais.an (154–222), can be used to assess the extent of Gnostic inﬂuence at Edessa. Bardais.an himself was apparently an adept of a number of different cults or belief systems, and has been taken as a key ﬁgure in the development and transmission of the esoteric Hermetic corpus. From the point of view of Hellenism, the arrival and success of these various belief systems at Edessa meant that the concepts they incorporated, many of which stemmed from Greek philosophy, were current there as well. It is tempting to take these intellectual currents as evidence of a sort of predisposition in favor of Greek culture at Edessa. However, as we have seen, the availability of Greek 118

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

mythological and artistic motifs to the artists who produced the mosaics of the late monarchy is not enough to allow us to call those artists’ patrons, or Edessa itself, thoroughly Hellenized. Similarly, the popularity of the ‘Greek’ ideas that came to be associated with Christianity cannot, in all strictness, justify any conclusions on the ‘Greekness’ of the intellectual climate before, or outside of, the theological disputes of the later second century.

Bardais.an and the wisdom of the Greeks Three apparently early Syriac texts – the Book of the Laws of Countries of the school of Bardais.an, the Oration of Meliton ‘the Philosopher’ (Syriac PLSWP ) to ‘Antoninus Caesar,’ and the ‘Letter of Mara Bar Serapion to his Son’ – were published together in 1855 (Cureton 1855).2 The ﬁrst of these, and the only one to have been re-edited since Cureton, is also the only one the existence of which at an early date is independently attested (it was known to Eusebius of Caesarea, who refers to it [Hist. Eccl. 4.30] as the ‘Dialogue on Fate’).3 It is thus the most genuine and compelling evidence of intellectual trends in our time period, and it has been called the ‘clearest example’ of a ‘fusion of cultures and traditions in this area’ (Millar 1987: 160).4 It is, in fact, the central text for the understanding of Edessan culture and Hellenism in our period. The other best source for Bardesanite philosophy happens to be among the most hostile: it is the anti-heretical work of the fourthcentury poet, hymnographer and theologian Ephrem Syrus. Ephrem was neither a contemporary nor an admirer of Bardais.an – he wrote most of his polemical works after 363 – but his hostility to the Edessan’s theology cannot disguise his genuine appreciation of his work. Ephrem preserves many fragments of it intact, in his hymns Contra Haereses and his Prose Refutations.5 Nearly equal in value to Ephrem is Eusebius of Caesarea (c. 260–340), who read Syriac and saw some of Bardais.an’s work in the original, and relates some information in his Historia Ecclesiastica and Praeparatio Evangelica. In contrast to Ephrem’s attitude to Bardais.an’s work, that of Eusebius is still fairly complimentary, showing that in his time the boundaries of Christian orthodoxy were still somewhat ﬂexible.6 The Book of the Laws comes from the pen, not of Bardais.an himself, but of one of his followers (probably after the teacher’s death), and hence it cannot be taken as a deﬁnitive statement of his philosophy. Indeed it contains very little that even the strictest advocates of orthodoxy would ﬁnd objectionable; this may mean 119

ROMAN EDESSA

that Bardais.an’s disciples deliberately colored his teachings to make them more acceptable to the Church. What is most noticeable from the very beginning, however, is the role of Greek philosophy in the dialogue. The text opens with a description of the author’s visit to the house of a friend, at which he encountered Bardais.an and witnessed his conversation with an interlocutor named Awida, which the author then recounts. Thus far, the work directly echoes the opening of Plato’s Republic. The dialogue proper begins with an exchange in which Awida is represented as asking: If God is one, as you say He is, and He has created mankind intending you to do what you are charged to, why did He not create mankind in such wise that they could not sin, but always did what is right? Thereby His desire would have been fulﬁlled. Bardais.an then responds: Tell me, my son Awida, what do you think: The God of the Universe is not One, or He is One and does not desire man’s conduct to be good and just? (Translation Drijvers, BLC 5 [text p. 4]) By posing the initial question in terms of the oneness of God, the author also allows Bardais.an to confront the contention of the heretic Marcion, who claimed that the Old Testament God, creator of the visible universe and of humanity, was a lesser entity than the Christian God: a theory inﬂuenced by Neoplatonism and the ‘Demiurge’ of Plato’s Timaeus itself. This dualism is a feature of many of the early Gnostic systems and, by rejecting the ‘evil’ commandments of the God of the Jews, served as justiﬁcation for a sort of theological anti-Semitism. The overall thrust of the treatise, however, is a defense, by a self-declared Christian, of the concept of free will, in opposition to the idea that the stars determine one’s fate. This position is supported ﬁrst by a dialectic section reminiscent of the Platonic dialogues, and then by an ethnographic section. Here the author catalogues the varying customs of peoples across the Roman Empire and beyond, with the purpose of proving that the stars cannot predetermine human behavior.7 Finally the dialogue presents Bardais.an as dealing with the ‘new race’ of Christians, who, though they live among the peoples of the world, 120

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

are not bound by those people’s customs but by the laws of their own faith. The confrontation with Marcionism, which is one example of the attempt to apply Greek philosophy to Christian beliefs, helps set the stage for the consideration of Bardais.an’s intellectual environment. The dramatic dialogue pictured in the Book of the Laws may never have taken place. Yet Eusebius reports that Bardais.an was an adept dialectician in the confrontation with the Marcionites, which suggests that there is some validity to the thoughts presented, if not to the dramatic setting itself.8 Marcion, a thinker born in Pontus and active in the ﬁrst half of the second century, spread his ideas primarily among a Greek-speaking population. The Chronicle of Edessa has an entry recording his separation from the Catholic Church in 138 – a fact which reﬂects the thinker’s importance to the Christians of Edessa (Chron. Min. 4/5). There is no way of knowing when this entry was inserted, although it may very well have come long after the event it records. In any case, Marcion’s ideas had clearly gained currency in Edessa by early in the third century, and probably earlier. The recent suggestion by one scholar that the ‘Church of the Christians’ mentioned in the Chronicle of Edessa was a church of the followers of both Marcion and Bardais.an seems unlikely to be correct, since Bardais.an is consistently identiﬁed as an opponent of Marcionism.9 The point here, however, is that Bardais.an and his followers (and presumably those of Marcion as well) were familiar with at least some of the elements of Greek philosophy, as interpreted by early Christian theology. Bardais.an is reported to have sent his son to study in Athens, but it is unclear whether he himself had traveled west of the Euphrates, and it is not clear whether he wrote in Greek (though it seems all but certain that he at least understood it).10 In addition to his philosophical and theological work, the Edessan composed over 150 hymns in Syriac (in imitation of Greek hymnography, according to Sozomen); his opponent Ephrem later found the need to use Bardais.an’s principles in the composition of his own work to combat heretical thinking.11 In his cultivation of Classical culture Bardais.an may be compared with his two older contemporaries, also from the Syro-Mesopotamian region: the pagan satirist Lucian of Samosata (born c. 120) and the Christian theologian and supposed founder of the Encratite sect, Tatian, who returned to the East from Rome in 172. Whereas Lucian wholeheartedly embraced Greek culture and became one of the leading lights of second-century literature, Tatian made a different use of 121

ROMAN EDESSA

his Classical education. One of his best-known works is the Oratio ad Graecos, which sets out to prove the antiquity of Christianity and Judaism and their superiority to Greek culture and religion (text and translation: Whittaker 1992). Another is the Diatessaron: a gospel harmony, surviving now only in fragments, and probably originally composed in Syriac. By contrast with both of these writers, Bardais.an developed his own synthesis of Christianity and Greek philosophy with ideas derived from other sources. This synthesis is the source of the ‘fusion’ that appears in the Book of the Laws. Bardais.an is represented in the dialogue on fate as speciﬁcally stating that he has studied Chaldaean astrology, but no longer believes that the stars govern human destiny.12 Given what we have seen of the inﬂuence of Babylonian religion in Edessa, it is certainly believable that the sage was well versed in these matters. Even beyond the Near Eastern context, Bardais.an took an interest in Eastern philosophies. Tradition associates him with contacts between the Roman Empire and India; he is said to have inquired of an Indian embassy to the Emperor Elagabalus about the customs of the Indians (Porph. Peri Stygos 1.56, De Abst. 4.17; Jer. Adv. Jov. 2.14). The ethnographic section of the Book of the Laws contains descriptions of Indian customs, which may have been based upon such an interview.13 His interest in such matters, and the presence of Eastern elements alongside those derived from Classical philosophy in Bardais.an’s system, highlight the role of the Near Eastern cultural entrepôt in the numerous religious and philosophical systems that arose in the early Christian centuries. Julius Africanus, the author of a treatise on miscellaneous topics known as the Kestoi, was a Christian philosopher, chronographer, and an ofﬁcer in the Roman army who visited Edessa, probably on ofﬁcial business, early in the third century. A passage of his Kestoi dealing with archery is valuable because it is based on a personal encounter with Bardais.an; it describes him as a courtier and skillful archer in the retinue of King Abgar the Great, and refers to him as a ‘Parthian.’14 While there is no reason to believe that the author intentionally misrepresented the philosopher’s ethnicity, this report surely means no more than that Bardais.an looked and sounded like an ‘Oriental’: he spoke Syriac and probably dressed in Eastern-style clothing like his patron the king. The philosopher’s name, meaning ‘Son of the Dais.an [River],’ bolsters the tradition that he was born in Edessa itself; but according to the same tradition, his parents came from elsewhere.15 122

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

The questions of Bardais.an’s primary cultural afﬁliation, however, and even of whether or not he wrote Greek, are in a sense irrelevant. In addition to the philosophical notions already mentioned, transliterated Greek terms – for example, the Syriac NMWS for ‘law’ (from Greek ν µο) – are a common feature both of the Syriac Bible versions and of later Syriac literature. They appear throughout the Cureton manuscript containing the Book of the Laws. It is on record that Bardais.an’s followers immediately translated the dialogue on fate into Greek, in an action which preﬁgured Edessa’s emergence in the following years as one of the main centers of the Greek–Syriac translation movement.16 Syriac itself is imbued with Greek. Perhaps the most important point that can be gleaned from the remnants of Bardesanite literature, however, is the conclusion that the political barrier between the Roman and Parthian spheres was no obstacle to the interchange of ideas. Nor did the regional dominance of Syriac as both ofﬁcial language and vernacular prevent the area from being washed by the intellectual currents that still ﬂowed around what had been the Hellenistic world. To go beyond this, it will be necessary to look at the broader outlines of those movements, in particular at Gnosticism itself, and at the evidence that Ephrem Syrus provides both for Bardais.an’s thinking and for the tumultuous marketplace of ideas that was Edessa in the second and third centuries – and indeed much later.17

Heresy, Hellenism and Gnosis Arriving in Edessa after Rome’s surrender of his native Nisibis to the Persians in 363 ce, Ephrem found his new home to be a hotbed of what he considered execrable heresies.18 Christians of the strain approved by Ephrem, worshipping in a church established 50 years earlier by the orthodox Bishop Quna, were lost among a throng of followers of other teachers of a more or less ‘heretical’ bent.19 Like the heretics – Arians, Valentinians, Bardesanites, Audians – Ephrem’s Christians were known at Edessa by the name of the founder of their group: a situation that moved Ephrem to exclaim, Cursed be those who allow themselves to be called after the name of Palut. rather than that of Christ . . . Palut. himself would not have wished that people call themselves after his name! (Ephr. Syr. Contr. Haer. 22) 123

ROMAN EDESSA

In this way Ephrem declared that the followers of Quna, the ‘Palutians,’ were the true Christians, and should insist that they be called by that name, not by that of the legendary founder of the Catholic sect. His need to assert their primacy makes clear for us that ‘orthodox’ Christians, if not in the minority, were still only a shaky majority at that time in Edessa. Ephrem wrote at a time when Christianity had been tolerated by the imperial authorities for 50 years, and had been the religion of the emperors (except for Julian, 361–3) for 40. At Edessa, however, Christianity of one form or another was ﬁrmly ensconced well before that time. Records of the early period are scarce and unreliable, but one seemingly genuine narrative depicts a community whose people – pagan, Jewish and Christian – were united in support of Shmona and Guria, the city’s ﬁrst Christian martyrs, during the Diocletianic persecution.20 By the reign of Ephrem’s contemporary Julian, Edessa had gained a reputation for devout, and contentious, Christianity. This led to a serious imperial snub when the ‘Apostate’ refused to visit Edessa during his Persian campaign – instead making a point of visiting its pagan rival, H . arran, to worship at the temple of the moon-god (Sozomen 6.1; Amm. Marc. 23.3.1–2). The emperor had already addressed a letter to the city’s inhabitants, admonishing them to desist from their factional quarrels and sarcastically announcing that he was conﬁscating the property of the Edessan church to help them follow the Christian command to ‘sell all that you own and distribute the proceeds to the poor’ (Jul. Ep. 40; Luke 18:22). The quarrels at Edessa that provoked the emperor’s anger were between the Arian faction, supported by Julian’s predecessor Constantius, and the followers of the Gnostic teacher Valentinus.21 Ephrem had little affection for either of these groups, but as far as Edessa was concerned, the three factions upon which he heaped most of his scorn – proving their popularity in his day – were the followers of Marcion, Bardais.an and Mani. Ephrem wrote that Bardais.an occupied a middle position between the other two heresiarchs, which was true chronologically; yet he failed to recognize any signiﬁcant differences among their philosophies. Nearly 150 years after Bardais.an’s death, the Edessan philosopher and his ‘school’ were in the thick of the cosmological, theological, and soteriological disputes of the day. The letter of Julian, with its reference to the followers of Valentinus, shows that the Valentinian Gnostics were an important, or at least a recognized, faction at Edessa, despite the strenuous efforts of the Arians and other, more orthodox, Christians to eliminate them. 124

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

This, along with the evidence we have seen above for the inﬂuence of Marcionism, makes it advisable to outline Gnosticism in general. Gnosticism (or ‘Gnosis’) as a philosophy remains ill deﬁned, as is to be expected of an esoteric movement that, until the discovery in this century of a secret cache of texts, was known primarily through the writings of its detractors.22 Although it is misleading to speak of a unitary ‘Gnostic religion,’ it can in general be said that the Gnostics stressed salvation from the evils of existence but de-emphasized both ‘faith’ and moral behavior, or good works, as paths to it, putting their trust instead in γνωσι – insight, or the knowledge of certain mysteries known to an elect few who would uniquely be saved.23 Along with this soteriology went a dualistic cosmology and theology that, like Marcion (himself often counted among the Gnostics), tended to deprecate the world and even the heavenly bodies as the work of an inferior Demiurge. The Gnostics spoke instead of an invisible or ‘stranger’ God, indescribable and in an unseen heaven, as the true source of all good, and the one with which the adherents of Gnosis – trapped in a world of darkness and matter – must strive to be reunited. The inﬂuential Gnostic movement of Valentinus spoke of the creation as more or less a mistake, which came about through the fall of a heavenly entity (usually known as Sophia or Wisdom) from the heavenly Pleroma which emanated from the true God.24 The Gnostic movement is clearly inﬂuenced by Neoplatonism and other philosophical trends of the late Classical world; yet it seems to have had its origin in the context of Hellenistic Judaism.25 It is in the same context, of course, that Christianity was born, and certain early Christian texts such as the Gospel of John betray marked Gnostic inﬂuence.26 Jewish or pre-Christian Gnosticism quickly adapted to the new faith, and Gnostic soteriology after the ﬁrst century tended to incorporate Jesus as the savior. Yet the variety of sects grouped under the name of Gnostics soon became recognized as the ﬁrst ‘heresy’ to threaten the orthodox religion, and the First Epistle to Timothy already warns against the dangers of the ‘Gnosis falsely so called’ (1 Tim. 6:20). Some of the ﬁrst systematic works of Christian theology were written in reaction to this dangerous philosophy – particularly against the version expounded by Valentinus, the ﬁrst and greatest systematizer of Gnosis before Mani. Geographically, Gnosticism, as an amalgam of biblical traditions and Hellenistic philosophy, is to be traced to the Syro-Palestinian region; one of the ﬁrst Gnostic teachers with whom the early Christians had to deal – Simon Magus – was from Samaria (Acts 8). Other components of Gnostic philosophy, however, derive from the wider 125

ROMAN EDESSA

Near Eastern context, and its dualism is evidence of the inﬂuence of Persian religion. Some of the early literature of Gnostic Christianity was almost certainly written ﬁrst in Syriac, possibly at Edessa itself; this is the case, for instance, with the ‘Hymn of the Pearl’ or ‘Hymn of the Soul’ which is found in the apocryphal Acts of the Apostle Thomas.27 Therefore, the presence and inﬂuence of Gnosis at Edessa, well before the time when Julian’s letter provides solid evidence, seems assured. Both the dualistic theology of Marcion and the philosophy of Bardais.an (with its emphasis on darkness and matter) betray elements of Gnostic thought, and the boundaries of Gnosis are nearly ﬂexible enough for Bardais.an, like Marcion, to be included within them. In comparing Bardais.an with Gnosticism in general, we must observe, ﬁrst, that his work contained a signiﬁcant admixture of non-Greek ideas. In defending the unity of God, Bardais.an developed a cosmology of four YTY or essences, commingling to form HWL or matter (@λη), which is in accord with Classical philosophy; but he added a ﬁfth principle, darkness, which he identiﬁed as the source of evil. It is to overcome that darkness that humans are endowed with free will (BLC 14, 16). The concept of the darkness, which derives from Persian religion, became the basis for Mani’s dualism, and hence made unavoidable the ﬁerce hostility of Ephrem, for whom Manichaeism constituted the greatest threat to orthodoxy.28 Ephrem believed that Bardais.an’s philosophy challenged the unity of God, rather than defended it – and that in doing so, it ‘opened the gate’ for Manichaeism (Prose Ref. I, xc [English], 122 [Syriac]). However, while Bardais.an echoes the Gnostics in their concern for the plight of the human soul – and talks of the ways in which the material world imposes a kind of unavoidable fate upon humanity (BLC 26–7) – he does not partake of Gnostic dualism, at least not insofar as his ideas are represented by the Book of the Laws. Bardais.an’s Christology, like that of the Gnostic literature, is Docetic, denying the suffering of Jesus on the Cross and rejecting the resurrection of the body.29 These can be said to be among the Gnostic elements in the Syriac philosopher’s system. The presence of such elements is not necessarily fatal to orthodoxy – mainstream Christianity as represented by the New Testament canon and by the works of Clement of Alexandria is not without Gnostic touches – but the fusion of elements in Bardais.an bespeaks more than a touch of inﬂuence.30 A second look at Bardais.an’s work conﬁrms that he was far from an ordinary Gnostic. In addition to his defense of free will and the 126

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

unity of God against Marcion, Bardais.an displays important nonGnostic traits. Most importantly, rather than seeing the entirety of creation, and matter itself, as evil, he speaks only of a mixture of the elements that disrupted an original harmony and gave rise to evil. It would seem all but certain that, if the genuine teachings of Bardais.an as they were well known to Ephrem Syrus partook of the Valentinian heresy, Ephrem would have mentioned this – but he does not (cf. Drijvers 1966: 183–4). Nevertheless, the possible ties with, and apparent inﬂuence of, Gnosis in Bardesanite thinking conﬁrm once again the participation of Edessa in the thought-world of the Roman Empire. In adopting Gnostic theology, however, Edessa was not only taking up ideas that had been developed elsewhere; it was actually a laboratory in which the thinking of philosophers including Marcion, Bardais.an and Mani was further developed and then propagated. The peculiar character of Gnosticism – its ability to incorporate and meld together Hellenic philosophical concepts with Eastern thinking – is further evidence of Edessa’s own extraordinary situation, with feet ﬁrmly planted in both worlds.

Early Edessan Christianity The picture of theological turmoil evoked by Ephrem’s complaint concerning the Palutians ‘suggests . . . that Christianity in communion with the Great Church developed [at Edessa] as a sort of precipitate in a cloudy solution,’ in the words of the Jesuit scholar Robert Murray (Murray 1975: 7). This merely rephrased the insight of Walter Bauer, who had observed that Edessan Christians of this period would all have to be described as ‘heretical,’ and suggested that to call oneself a Christian in second-century Edessa would more likely mean that one was a Marcionite than anything else (Bauer 1934; Eng. tr. Kraft 1972). The use of the name ‘Palutians,’ to which Ephrem gives witness, seems to indicate that Palut. was an historical ﬁgure, since it is hardly believable that a Christian group would adopt the name of a ﬁctional character if it were really ﬁrst established at Edessa by someone else, such as Quna. The Doctrine of Addai suggests a date of about 200 for Palut., which is probably not far off. His faction most likely did not establish itself at Edessa much earlier, for if it had, it probably could have taken for itself the unmodiﬁed name of ‘Christians,’ which went instead to the Marcionites, the Bardesanites or both. Conceivably, the date could be later. Eusebius, however, reports (Hist. Eccl. 5.23.4) that the 127

ROMAN EDESSA

churches of Osrhoene were consulted when a controversy arose in the Church over the date of Easter, around 197. If this report is genuine, it offers conﬁrmation for the establishment of at least a small Catholic community in Osrhoene, and probably in its capital of Edessa, around 190–200.31 This community, however, was at ﬁrst a quiet minority among the Christians of Edessa, and the date of its establishment is not crucial to the attempt to understand Edessan society and culture. Of more importance is what can be observed about the practices and beliefs of the early Christians of whatever stripe, and about their non-Christian contemporaries. The strength of the monastic movement in the Syriac-speaking region, and the tenor of the writings of Christian authors like Aphrahat and Ephrem the Syrian, have been taken by many as indicating an ascetic bent in the earliest Syriac Christianity. But in fact some of the relevant writings of Ephrem have been shown to be spurious, and without them there is very little direct evidence of the monastic or proto-monastic impulse in the ﬁrst Christian centuries (Grifﬁth 1995: 221–2). On the other hand the popularity of Marcionism, for which the Bardesanite literature gives evidence, may imply that a signiﬁcant number of Edessan Christians by the end of the second century adhered to the dualistic theology which entailed an ascetic ethic along with the intellectual rejection of the God of the Old Testament. It is very likely, as well, that many Edessans were members of other Gnostic groups, either Christian or non-Christian, which almost all advocated a similar asceticism.32 A second clue is the strong preponderance of evidence showing that the ﬁrst version of the New Testament to be widely used among Syriac-speaking Christians was the Diatessaron of Tatian. A native of ‘the land of the Assyrians,’ Tatian underwent a Classical Greek education and reached a fairly high level of accomplishment as a rhetorician, traveled to Rome and studied under Justin Martyr (100–165) before returning to the Orient in 172.33 The Diatessaron, whose original language was probably Syriac, combined the gospels into one narrative by selective use of passages from all four. The full text does not survive, but reconstructions based on extracts and works of criticism by ancient writers tend to support the idea that Tatian produced a scripture that harmonized with his own ascetic ideal, including a recommendation of sexual abstinence. This is in agreement with a work by Tatian that does survive in Greek, the Oratio ad Graecos. In the midst of a series of arguments attacking various aspects of Greek culture and philosophy and setting out to prove the 128

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

superiority and antiquity of Christianity and Judaism, Tatian outlines a philosophy centering on the transcendence and unity of God, and on creation by means of the divine Logos. Of greatest relevance for us, Tatian maintains that matter, because of its association with ‘demons,’ must be rejected (Oratio 18.14). Though it does not come through clearly in these passages, Tatian eventually emerged as the ‘Patriarch of the Encratites’ or strict ascetics, criticized by St Jerome for his teaching that marriage, the ﬂesh and wine were evils to be avoided at all costs.34 Tertullian also knew of his strict asceticism (De Ieiun. 15). Tatian is said to have taken over the bans on marriage and on eating meat from Saturninus and Marcion, after his return to the East to found his own school (Eus. Caes. Hist. Eccl. 4.29). It was at this school that the Diatessaron was probably produced and propagated, a publishing effort whose success is rivalled only by the achievement of the inventors of the Abgar legend. It has been suggested that Tatian himself ﬁrst brought Christianity to Edessa, an idea which is not contradicted by the few ﬁrm bits of evidence we have; the proposal was even put forth at one point that the name of Addai in the Abgar legend masks that of Tatian (Burkitt 1924: 130; CAH 12, 493–6). This view is no longer advocated by anyone, and there is no hard evidence of any connection between Tatian and Edessa. But his gospel’s popularity in the Syriac-speaking region, along with what we know of the position of Marcionism there, is consistent with the idea that the ﬁrst Syriac Christianity had a decidedly Gnostic and ascetic ﬂavor. If there were any doubt about this, it ought to be dispelled by the Gnostic literature centering around the ﬁgure of Thomas, the ‘Apostle of India.’ The apocryphal Acts of Thomas, known since the last century, tell the story of Thomas’s mission to and travels in India spreading the gospel, with an interesting twist: beginning at the wedding of a local king’s daughter, Thomas makes a career out of persuading young married couples not to have sex (Wright 1871: 1.171–333; Ortiz de Urbina 1965: 37–41). The Acts include an allegorical Syriac work, the ‘Hymn of the Pearl,’ that is believed to be of very early origin, telling of the soul’s redemption through obtaining the ‘pearl’ of Wisdom. This Gnostic or Gnosticizing text forms a connection with the thought-world of the Gospel of Thomas, the Coptic text of which (from a Greek or Syriac original) was found in the Nag Hammadi library and published in 1959 (ed. and tr. Guillaumont 1959; NHC II.2 [Robinson 1990: 126–38]). This Thomas literature may have been produced at Edessa, but there is no 129

ROMAN EDESSA

proof of this (Ehlers 1970: 284–317). What is certain is that Thomas’s name was linked with that of Edessa; the pilgrim Egeria visited his supposed tomb there in the late fourth century, and read texts containing what she took to be the words of the apostle himself (Itinerarium Egeriae 19.2; Smelik 1974). Neither the pre-Christian nor the Christian Gnostics were of one accord as to their moral and ethical precepts. Some of them advocated (or are accused by their detractors of advocating) an amoral libertinism, in keeping with their denial of the authority of the creating and avenging Demiurge. A strict asceticism, however, was more common among the Gnostics: a practice which found theological justiﬁcation in the need to keep the souls of the elect as pure and unencumbered by matter as possible (Brown 1988: 116–18). The appearance in some Syriac Christian texts of the phrases ‘Sons of the Covenant’ and ‘Daughters of the Covenant’ (Bnay Qyama, Bnat Qyama) supports the conclusion that Syriac Christianity already showed a leaning toward asceticism, well before the emergence of true monasticism.35 Such asceticism was not necessarily ‘heretical,’ since the fourth-century orthodox writers Ephrem and Aphrahat, both of whom are very likely to have been Sons of the Covenant, advocate a celibate life within an elect community, even while allowing for marriage by the non-elite within the Church. All of these texts are at least 100 years later than the period with which we are concerned. There are indications, however, that originally the Qyama may have embraced the entire Christian community, and that baptism involved a vow of celibacy (cf. Vööbus 1958: 93–5, 175–8; Murray 1975: 59–80). It is conceivable that the earliest Christian community or communities in this area traced their roots to a Jewish form of Gnosis or to a sectarian Jewish community similar to the Qumran sect that produced the Dead Sea Scrolls, in which some similar terminology appears.36 Robert Murray has also re-emphasized the apparent links between early Syriac texts, including the Peshitta New Testament, and Judaism, as part of an argument that Nisibis was the focus of Christian activity in the early centuries. This is far from a decided question, however. Against the focus on relations with Nisibis and a Christianity coming to Edessa ‘from the East,’ Drijvers has written that ‘Ideas, like goods and travellers, move along the existing roads, and do not ﬂy through the air or come down straight from heaven. Christianity similarly arrived at this northern Mesopotamian region and the areas further east from the Syrian metropolis Antioch; with 130

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

Rome, Antioch was the main centre of early Christianity in the Roman Empire’ (Drijvers 1992: 128). This commonsensical reminder is useful as a caution against the romantic idea that Edessan Christianity represented a ‘pure’ form of the religion more in touch with its Palestinian roots than in the Hellenized empire (cf. Vööbus 1958: 4). At the same time, however, Drijvers’s argument – developed in the process of a demonstration that the Abgar–Addai myth originated in anti-Manichaean propaganda – tends to downplay the evidence for Jewish inﬂuences.37 We would do well to remember that the earliest Christianity as it developed during the Apostolic period was not uniform, even within one city. Biblical texts record the break between the Apostle Paul and others within the Church who wished to impose Jewish Law on all Christians, and Paul gives evidence not only of his own internal conﬂict over the Law, but of the fact that, while he was evangelizing cities in Asia Minor and to the west, others of a different theology were active elsewhere (Rom.; Gal. 2:11–14; see Meeks 1983: 115– 16). The fact that Antioch was the main western terminus of the road running from Edessa says nothing about the way in which Christianity reached Edessa or the form it took when it arrived there. Given the state of the evidence, it is impossible to decide from which ‘direction’ Christianity came. There are, in fact, indications that at least some early Christian communities in the region around Edessa had members of Jewish origin, and that these may have existed alongside the more ‘Greek’-oriented community of Bardais.an and others (Brock 1995: 282). In any case, it is probable that the form of Christianity dominant in Edessa by the end of the second century was based on a dualist theology of some sort, and that "γκρα´τεια or extreme asceticism of the sort advocated by Tatian held wide appeal.

The making of a myth Given what we have already observed of the inﬂuence at Edessa of thought systems combining Jewish, Christian and Hellenic concepts, the issue of Christian origins only appears more complicated, and is not likely to be decided within the scope of our present inquiry. The primary piece of evidence – the purported exchange of letters between a King Abgar and Jesus – says more about the social/ religious conﬂicts of the third and fourth centuries than about the true beginnings of Edessan Christianity. Yet we may, in exploring the dimensions of the question, gain some insight into the reasons 131

ROMAN EDESSA

for the story’s great attractive power, as well as into some of the social and cultural dimensions of the coming of the new faith. The best-known version of the legend is that given by Eusebius: A copy of a letter written by Abgar the toparch to Jesus and sent to him at Jerusalem by the courier Ananias: ‘Abgar Ukkama, the toparch, to Jesus the good Savior who has appeared in the district of Jerusalem, greeting. I have heard concerning you and your cures, how they are accomplished by you without drugs and herbs. For, as the story goes, you make the blind recover their sight, the lame walk, and you cleanse lepers, and cast out unclean spirits and demons, and you cure those who are tortured by long disease and you raise dead men. And when I heard of all these things concerning you I decided that it is one of two things, either that you are God and came down from Heaven to do these things, or are the Son of God for doing these things. For this reason I write to beg you to hasten to me and to heal the suffering which I have. Moreover, I heard that the Jews are mocking you, and wish to ill-treat you. Now I have a city very small and venerable which is enough for both of us.’ The reply from Jesus to the toparch by the courier Ananias: ‘Blessed are you who believed in me, not having seen me, for it is written concerning me that those who have seen me will not believe in me, and that those who have not seen me will believe and live. Now concerning what you wrote to me, to come to you, I must ﬁrst complete here all for which I was sent, and after thus completing it be taken up to Him who sent me; and when I have been taken up, I will send to you one of my disciples to heal your suffering and give life to you and those with you.’ (Eus. Caes. Hist. Eccl. 1.13) Eusebius identiﬁes the letters as having been translated from Syriac originals in the archives of Edessa. That such documents were placed at one time in the city archives is not in itself unbelievable. The question of the story’s truth, however, need not detain us; even in Eusebius’s shortened version it contains enough elements to label it a fabrication – the phrase ‘Son of God; ‘ Jesus’s reference to his Ascension; the hint of anti-Semitic rhetoric. The whole story was, 132

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

indeed, declared a fraud by Pope Gelasius in 494. By then, however, the tale had spread far and wide, and the Letter of Jesus had grown longer with the addition of a benediction promising impregnability to Edessa. This happened before the end of the fourth century, although the expanded version apparently took some time to reach the West.38 Eusebius’s account continues by describing the dispatch to Edessa of Thaddaeus, one of the Seventy Apostles, and his healing and evangelization of Abgar and others. It ends by saying that these events took place ‘in the 340th year,’ that is, 28/9 ce. The general outlines of Eusebius’s account are very similar to those of the Syriac Doctrine of Addai, and the two tales must derive from a common ancestor, Eusebius having replaced the name of Addai with that of Thaddaeus, a biblically attested apostle. The Doctrine of Addai, however, expands on the story in several respects, for example by adding several vicious anti-Jewish passages and an account of a portrait of Jesus painted by Abgar’s messenger (Doct. Addai [Howard] 8–11, 58–9, 84–5). This portrait later gained a spiritual value equal to that of Jesus’s letter itself, and in some versions of the story became an image mystically created, like Veronica’s Veil.39 Generations of earlier scholars, faced with the unbelievability of the Abgar legend, have attempted to preserve Edessa’s reputation as the ‘First Christian Kingdom’ by locating a royal conversion elsewhere in time.40 What makes the issue appropriate for us to take up at this stage in our inquiry is the view of certain inﬂuential writers, Burkitt and Duval among them, that Edessa’s ﬁrst Christian king was Abgar VIII ‘the Great,’ Bardais.an’s contemporary and patron. A number of general considerations, and one piece of hard evidence, militate in favor of this idea. The king’s relationship with Bardais.an, who at least professed Christianity no matter how much Ephrem may have denied him the name, and the passage in the Chronicle of Edessa mentioning the Church of the Christians in 201, show that this king at least tolerated the new religion. The Doctrine of Addai states that, after the deaths of Addai and his disciple Aggai, the new leader of the Edessan Christians, Palut., was consecrated by Bishop Serapion of Antioch. Since Serapion was bishop from around 190 to 209, this is a serious chronological confusion on the part of the author; but this passage has helped to support the idea of the establishment of Edessan Christianity during the reign of Abgar VIII.41 It is difﬁcult to determine how such a confusion could have come about, seemingly telescoping more than one hundred years of history and making Palut. a contemporary of both Serapion and Aggai. 133

ROMAN EDESSA

The best explanation is probably to assume that considerations of chronology simply buckled under the strong desire of the ‘Palutian’ Christians (who did have a connection with the Antiochene church) to enhance their status within the community by the claim to apostolic authority. The hard evidence for Christianity’s establishment under Abgar VIII is the following passage in the Book of the Laws of Countries: In Syria and Edessa ( WRHY) there was the custom of selfemasculation in honor of Tar atha (Atargatis), but when King Abgar believed [HYMN, translated by Drijvers ‘came to the faith’], he ordered that every man who emasculated himself should have his hand chopped off. And from that day to this no one emasculates himself in the territory of Edessa. (BLC 58–9)42 Of relevance also is Julius Africanus’s reported description of Abgar as >ερ3ν ανδρα, ‘a holy man.’43 In the light of the other evidence, this has been taken as a reference to Christianity, but it is certainly not as clear as one would wish. If, in fact, we take up the suggestion of Sebastian Brock, that the word ‘believed’ in the Bardesanite treatise is an interpolation, there is precious little solid evidence at all for the Christianity of Abgar VIII or any other king of Edessa (Brock 1992: 227). Eusebius, in quoting this passage of Bardais.an, remarks no reference to Christianity – referring instead to the king’s order ‘at a single moment’ (Eus. Caes. Praep. Ev. 6.44; Millar 1993: 476). This supports the contention that the reference to Abgar’s belief is an interpolation; besides, even if it is not, the text is not explicit about which Abgar is meant. However, the lack of an explicit mention of Abgar’s faith in Eusebius’s version does not mean that it was not there in the original. Eusebius may have omitted the line as incomprehensible, since he did not seem to doubt the tradition that Edessa had been Christian since shortly after the Cruciﬁxion. Probability also strongly favors the conclusion that Bardais.an is referring to his own contemporary, since he gives him no epithet. If any of Edessa’s kings had converted to Christianity it might be expected to be reﬂected in the coins, which sometimes (like coinage throughout the ancient world) bear religious iconography; but no Christian symbols appear on the coins of Abgar VIII or any Edessan ruler. Instead, the kings of the Abgar line wear a Parthian tiara covered with a network of dots or stars; if there is any religious 134

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

reference here it is to the astral cult, well attested for pre-Christian Edessa. Comparanda for the adoption of Christian symbols on coins, however, are lacking, until they begin to appear on the coins of the Christian Roman Empire in the Byzantine period. There is, then, no compelling reason to believe that royal Edessa was ever ‘ofﬁcially’ Christian, or even that Christians were in the majority there before the fourth century. About Abgar VIII, it can only be said that he tolerated the new religion. This in itself might be considered remarkable, considering the city’s later history of violent doctrinal conﬂict. Since there is no evidence of suppression of the Christians before the end of the monarchy, it may be that it was Christianity itself that brought with it the urge to impose uniformity of belief by force. If the tale of the conversion of a king in Eusebius and the Doctrine of Addai has no basis in fact, whether at the time of Jesus or in the court of Abgar the Great, where did it come from? Analysis of the text reveals some clues to its purpose. The story is a genealogical myth of the kind familiar from Classical literature, which traces the origin of a community back to a mythical or divine ancestor. One might think, in this connection, of the heroic ancestors of the Athenian γ-νη, or of the supposed descent of Rome’s Iulii from Venus. As it appears in Eusebius, the tale gives no particular advantage or status to anyone besides the Edessan church as a whole. The more extensive version in the Doctrine of Addai, however, contains an important clue. As we have seen, that narrative follows up the story of the evangelization of Edessa with that of Aggai, the disciple of the original Apostle Addai (Eus.: Thaddaeus; Doct. Addai [Howard] 80–1). Aggai himself suffers martyrdom at the hands of the evil son of the ﬁrst Christian king, anachronistically named Abgar Severus, and is unable to ordain his successor Palut. (104–5). Therefore the latter must turn to Serapion at Antioch, whom the narrative names as having been himself ordained by Zephyrinus, bishop of Rome (this statement recurs in the Acts of the Martyrdom of Barsamya). Besides the chronological impossibility of the Addai–Aggai–Palut.–Serapion nexus, there is a problem with Serapion’s own pedigree as stated here. Serapion’s tenure at Antioch began in 189 or 192, and Zephyrinus’s at Rome not until 202 (cf. Burkitt 1913: 25–6; Brock 1992: 227–8). This link has been invented, and we may be able to guess why. All these connections give the community that traced its religious ancestry to Palut. a double claim to authority: as the inheritor of Jesus’s blessing via Addai and Aggai, and as the duly appointed arm of the Church of Rome. We know from Ephrem that 135

ROMAN EDESSA

the ‘Palutians’ were those who, he thought, truly deserved the name of Christians: the orthodox, anti-Marcionite, anti-Bardesanite, antiManichaean (and, in Ephrem’s day, anti-Arian) faction. There thus arises a strong presumption that the tale of Addai in the form that we have it, and possibly the whole Abgar–Jesus legend, was fabricated in the context of the struggles that ensued when Bishop Quna founded the orthodox church at Edessa in 313, thus bringing the city into the Catholic fold (Chronicle of Edessa in Chron. Min., 4/5). In a series of very convincing articles, Drijvers has connected the Addai narrative speciﬁcally to the confrontation with Manichaeism (Drijvers 1970, 1974, 1982b, 1992). The very name of Addai is the same as that of an important apostle of the Manichaeans, and Drijvers is able to point to a number of other suggestive parallels, including the portrait of Jesus in the Addai story, which corresponds to an image of Mani himself that was a central feature of some of his followers’ ceremonies. Mani had a community of followers at Edessa to whom he wrote a letter partially preserved in the Cologne Mani codex (Cameron and Dewey 1979). Just as the prevalence of Marcionism in Edessa during the earliest years provides a background against which to interpret Bardais.an, the power and popularity of Manichaeism provide the setting for the Doctrine of Addai. It is not hard to imagine that the Abgar legend may have been reworked more than once, each group modifying it for its own special purposes. In the current version the names of certain Edessan nobles appear as early converts: in particular, Abshelama and Barsimya (Doct. Addai [Howard] 70–1). These names are the same as those appearing in martyrologies which, though they pretend to date to the time of Trajan, are certainly inventions. In these fabricated texts, it would seem, we may have an attempt by members of the urban nobility to give their families a claim to priority within the Christian community as old as that of the ﬁrst genuine martyrs, who were from the surrounding villages (Brock 1992: 227–8; cf. Segal 1970: 83). It will be remembered that the family of a Barsimya was one of those that apparently claimed high social standing by means of its close association with a royal Abgar (on the basis of a recently discovered mosaic inscription; Drijvers 1982a). If our conjecture is correct, it implies that Christian faith replaced royal ideology as a ratiﬁer of social position, once the monarchy became extinct.

136

E A R LY C H R I S T I A N I T Y A N D E D E S S A N C U LT U R E

Conclusion Although Bardais.an seems to have written primarily, if not exclusively, in Syriac, given his familiarity with Greek philosophical systems, and the examples of Lucian and Tatian, it is probable that he understood Greek. Whatever languages he read or wrote in, his activities – and those of the more orthodox continuators of the Syriac literary tradition including Ephrem – suggest the existence of a tradition of Edessan literature probably extending over much of the second century and even earlier. Almost none of this hypothesized Syriac literature has been preserved, which is not surprising because much of it was probably imbued with heretical notions. Without a doubt, Edessa was in touch with, and strongly affected by, intellectual currents that would wash over the entire Roman Empire during Late Antiquity, after the arrival of ofﬁcial Christianity. But this involvement started much earlier, as we can see by the ﬁgure of Bardais.an – heretical though he may have been. Bardais.an was ‘one of the most important links in the transmission of Hermetic lore in the Near East,’ according to a recent study (Green 1992: 88; on Bardais.an and the Hermetica see especially Drijvers 1970). This assessment, based largely on the descriptions of the work of the philosopher by later cataloguers and by heresiologists, suggests that the Edessene thinker was intimately caught up, not only in transmitting the esoteric Hermetic lore, but in developing it, and imparting a Near Eastern ﬂavor to the mixture of religion, magic and science that was philosophy in Late Antiquity. The myth of the evangelization of Edessa by Addai or ‘Thaddaeus,’ in its preserved form, is well explained by the need of the orthodox Christian community to provide itself with an apostolic genealogy in the battle with Gnosticism, including its later proponents the followers of Mani. By the middle of the fourth century the Gnostics were clearly on the losing side in this struggle, as the letter of Julian shows. Two centuries earlier, however, the picture was entirely different. Christianity may very well have come to Edessa, and long held sway there, in a Gnostic or a Gnosticizing form. In addition, its success there suggests that the ground had been well prepared, perhaps by a Jewish or other form of non-Christian Gnostic teaching. The Life of Abercius cannot provide reliable details about either the deeds of that early Christian wanderer or the exact theological orientation of early Syriac Christianity. Nevertheless, the genuine inscription of Abercius conﬁrms the picture of a lively Christian 137

ROMAN EDESSA

community in Edessa and the Syriac-speaking region just at the time that we have called Edessa’s Golden Age, namely, the reign of King Abgar VIII. This community can only have been brought into being through a certain openness, and indeed continuity, between the Syriac-speaking East and the Greek-speaking West (Syria and Asia Minor). The message conveyed by the remnants of early Syriac literature and by the evidence for early Edessan Christianity is similar to that of art and archaeology: Edessa lay open to inﬂuences from all directions, and it adopted and incorporated them into a synthesis that is neither purely ‘Greek’ nor purely ‘Oriental.’ It can only be called Edessan.

138

7 CONCLUSION

By the time it reached the end of the long road toward becoming a truly Roman provincial city – ﬁrst as a Roman municipium, and later as a colonia – Edessa was already experiencing the ﬁrst shocks of the renewed conﬂict with Persia under the Sassanid kings. The ﬁrst ten to ﬁfteen years after incorporation of the former client kingdom in the empire will have featured important developments that set Edessa on the path to full Romanization. Among them were the absorption of the entire region into the Roman tax-revenue system, a process in which the cities of the empire were always instrumental; opportunities for social advancement (for example, membership in the Roman equestrian order) based on service to the municipality and to Rome; the growing presence and inﬂuence of Roman soldiers from the forces garrisoning Osrhoene, and the service in the Roman army of troops provided by Edessa (the Osrhoenian archers attested under Severus Alexander, Dio 78.14.1; Hdn. 6.7.8, 7.1.9, 7.2.1). The extent of Romanization is clear when we examine the reign of the last native Edessan ruler, Abgar X, who came to the throne in the early stages of the conﬂict with Sassanid Persia. The precise conditions of this renewal of the monarchy are beyond our retrieval; yet we have been able to reconstruct the outlines of the picture and to ﬁnd a convincing context for Abgar’s otherwise mysterious coinage. Gordian III, present at Antioch for an unknown reason early in his reign (April 239), personally appointed Abgar to an important military or administrative role pending the emperor’s return to lead the war in person, in autumn 242. The evidence for this role, however, indicates at the same time how much the circumstances had changed. Although he still wore the high tiara and the other marks of authority that derived from the monarchy of the Parthian period, Abgar X (‘Aelius Septimius Abgar’) expressed in his nomen and in his titulature a conception of authority that derived from the West – 139

ROMAN EDESSA

he had been granted the ‘consulship’ at Orhai/Edessa. After a century of clientela that had begun with Trajan, and another quartercentury of direct rule beginning with Caracalla, the realm of the Abgars had become a Roman city.

Waves of war Until Septimius Severus created the provinces of Mesopotamia and Osrhoene as a ‘bulwark’ for Syria, neither the Roman nor the Parthian authorities had any settled plan to incorporate the upper Mesopotamian region within their empire (Dio 75.3.2–3). Rather than attempts to invade and consolidate, the repeated onslaughts of the emperors and the kings were more like border raids – intended to inﬂict damage and seize wealth without establishing permanent control. It was this desultory approach on the part of the great powers that enabled the kings to pursue a semi-independent policy – seizing as much autonomy as circumstances allowed, and yielding to greater force, or the threat of it, when necessary. We have argued that the rhetoric of Abgar X’s regime indicates a change in attitude, and that the manner of his accession to the throne and the ‘consulship’ shows that the revived monarchy remained ﬁrmly rooted within the empire. Whether or not this is the case, the end of Abgar X’s reign (whether he died in the confrontation with Persia or was removed from the throne for some reason is unknown) marked the end of any attempt to retain even partial autonomy. It was also, if our explanation for the monarchy’s brief reappearance is correct, the end of Rome’s reliance on local leaders for military support in the region. The client kingdom of Armenia to the north remained in existence and would at times play a similar role; but as far as Roman Osrhoene and Mesopotamia proper were concerned, the provincial system was now ﬁrmly in place, under the protection of the Roman military. This arrangement was shaky at best. It failed to prevent repeated Persian raids in the 250s that reached as far as Asia Minor, and led to the prominence of the Palmyrene leader Odenathus and the ensuing Palmyrene attempts at establishing an independent empire. The greatest threat, however, came in 260, when Shapur I, no longer content with the cash payment he had received from Philip I, returned in force and laid siege to Edessa. The Emperor Valerian rushed to the defense of the strategic city, and in a battle under its walls, according to the inscription of Shapur at Naksh-i-Rustam, the Sassanid king took the Roman emperor captive. Shapur continued to 140

CONCLUSION

display the raider’s approach to the region, and may not have captured the city itself; but he did inﬂict serious damage on the Roman forces, and ravaged the countryside around Edessa. Conditions remained unsettled on the eastern frontier until 298, when the Emperor Diocletian signed a treaty with Narses of Persia and reorganized the frontier. These years must have been dispiriting for the people of Edessa; recurring waves of war strained the region’s resources, and in addition to demanding support for its military the Roman government engaged in religious persecution. Diocletian, remembered as an organizer and administrator, was also a persecutor – and it was in his reign that the ﬁrst genuine Edessan martyrs met their deaths, apparently with the sympathy and support even of many non-Christian Edessans (Burkitt 1913: the martyrdom account of Shmona and Guria). Even after the persecution ended, the warfare continued. The contemporary historian Ammianus Marcellinus gives an account of the Christian Empire’s confrontation with Persia that shows the importance of Edessa as a supply base and rallying point – especially in connection with the Persian war of Constantius and his successor Julian, 360–3 (Amm. Marc. 18.7.7, 19.6.12, 20.11.4, 21.7.7, 21.13.1). This war ended in Julian’s death and the agreement by his successor Jovian to cede the important fortress at Nisibis to Persia, with important consequences for Edessa. With Nisibis under Persian control, Edessa became the ﬁrst community inside the Roman Empire on the direct East–West route, and it may be that its fortunes began improving from this date, on the basis of increased trading activity. It was at this date also that Ephrem Syrus relocated from Nisibis to Edessa and began there his career of polemic to which we owe much of our knowledge of Bardais.an and the intellectual atmosphere of the time. As the home both of Bardais.an and, now, of Ephrem, Edessa was well on its way to becoming the important cultural center that it would remain until the seventhcentury Muslim Conquest and beyond.

The Edessan genius The success of Bardesanite philosophy, in terms of its longevity and its undeniable inﬂuence far beyond the Syriac-speaking region – above all through the medium of Manichaeism – bears witness to the vitality of the Edessan cultural scene during and immediately after the reign of Abgar VIII. As we have seen, the popularity at Edessa of Greek philosophical ideas – the Neoplatonic notion of the 141

ROMAN EDESSA

creating Demiurge; the concept of @λη; the use of Socratic dialectics – is not a sure guide to Edessa’s true cultural afﬁliations. For it is only insofar as they played a role in the Christian theological debates that we see those concepts used. In this way the nature of the sources continually casts a shadow over our enterprise. Even so, Bardais.an is for us an example of Edessa’s openness to the intellectual currents of the Roman world. What is more, he is not an isolated example. Two others from the Syro-Mesopotamian region are Tatian and Lucian of Samosata, both of whom worked and traveled in the Western world, and became part of the empire-wide literary and theological dialogues of their day. Neither Tatian nor Lucian was from Edessa itself. But Tatian returned from Rome to the East and spread his teachings there, not least in Osrhoene, where his Encratite philosophy received a sympathetic hearing. Lucian remained part of the Western literary scene, but from his works (especially the De Syria Dea) it is clear that he came originally from a cultural milieu not far removed from that of Bardais.an. The Edessan philosopher, too, surely could have had the opportunity to pursue a career in the wider Western literary world; it is on record that he did send his son to study in Athens. The fact that he did not himself leave perhaps says a great deal about the vitality of the Edessan cultural scene. The ﬂowering of intellectual life in Edessa and the Syriac-speaking region bore fruit in a large corpus of religious literature which is preserved for us only in those works that survived the doctrinal controversies. This means, for the most part, those works that were written after the end of the monarchy – following the victory of Christianity in Edessa and in the Roman Empire as a whole. Yet the earlier period must also have seen a great amount of activity. Although a critical look at the Abgar– Jesus legend has conﬁrmed that it has no basis in fact, and that neither Abgar V nor any later king ofﬁcially converted to the new faith, the picture that we have been able to draw of the life of the kingdom under Abgar VIII is one of prosperity and cultural vitality. It was this last phase of the monarchy that provided the conditions for the work of Bardais.an and other productive thinkers; it also seems likely that it is to this period that we should assign the impressive remnants of Edessa’s past that the pilgrim Egeria admired and associated with Jesus’s correspondent. We have dubbed this period Edessa’s ‘Golden Age’ because of its apparent surge in cultural production and because of the admiring mythology to which it gave rise in later periods – the sort of nostalgia that is a prerequisite for any Golden Age. 142

CONCLUSION

The remnants of pre-Christian religion further illuminate the question of cultural orientation and antecedents. Edessa under the monarchy revered a pantheon very similar to those of the surrounding peoples. In addition to the Syrian Goddess Tar atha/Atargatis, the best-attested of Edessa’s gods, the group included the Babylonian Bel and Nebo, the Sun, Sin the moon-god of H . arran, and the Arab gods Aziz and Mun im, πα´ρεδροι of the Sun ( Julian). The evidence for forms of devotion – the baetyl cult to which the Sumatar inscriptions and the coins of Wael testify, and the ceremony pictured in the Tripod mosaic – conﬁrms this ﬁnding. There is no positive evidence that Edessa’s people worshipped any of the ancestral Greek gods of the Seleucid colony’s founders, or that they consciously identiﬁed and subsumed any of the non-Greek deities to a Greek one – with the possible exception of the later coins of the colonia, bearing the bust of a Greek goddess that may represent the city’s Tyche. Even as it lay open to the intellectual and religious ideas coming from the West – including Christianity in its various forms – Edessa maintained its ties with its Eastern religious antecedents. Whether this should be taken as meaning the ideas and beliefs that the Abgarids brought with them when they took power or, on the other hand, traditions native to the region that survived the Seleucid colonization and thrived under the Abgars, is not known. Frustration is bound to attend the attempt to get a general picture of the Edessan cultural milieu by these means. All the preserved literature – including the Bardesanite treatise – is too vulnerable to the anti-historical impulses engendered by doctrinal controversies to allow for much conﬁdence. In any case, even if we could be sure we had a clear grasp of the ideas of the earliest Christians, it is likely that they were a small group at ﬁrst, perhaps unrepresentative of the general population. We therefore turn to the funerary mosaics, a visual record of the culture of Edessa’s people, and even of some of their names and faces. In these artworks, most of which are unfortunately no longer preserved, we have seen powerful evidence of that very openness to inﬂuences from all directions that we see in other aspects of Edessan culture. The Phoenix and Orpheus mosaics illustrate Greek mythological themes that were presumably familiar to the artists’ patrons, who nevertheless spoke and read Syriac, not Greek – as we can tell by the transliterations of the mythological names. These scenes, and those in the other pavements, are presented within ﬁgured borders that have unmistakable models in the mosaics of Hellenized Roman 143

ROMAN EDESSA

Syria. This ‘Greek’ element of the works, however, is to be set against the subjects portrayed. Their loose trousers, soft shoes, tiaras and long ﬁgured robes speak loudly, if not of Parthian origins then certainly of non-Greek and non-Roman ones. The subjects are portrayed in hieratic frontal poses or in contexts such as the so-called funerary banquet familiar from Palmyrene art, and are identiﬁed in many cases by Syriac mosaic texts. The faces of Edessa that stare out at us from these scenes speak mutely of a thriving Eastern culture (itself containing differing degrees of Parthian, Arabic and Aramaic identity) whose life, like the scenes inside the mosaic borders themselves, was carried on within a framework provided by Hellenism. This is the understanding we have gained of the life of the Syriac community among the walls, stoai and public buildings of the Seleucid colony. Although the Abgars and their subjects were more or less powerless to fend off the often destructive waves from both East and West that washed over their land in century after century, they showed that at the same time they could withstand and beneﬁt from the accompanying waves of cultural inﬂuence, in the process creating something new. That was the special genius of Edessa.

144

APPENDIX NUMISMATIC NOTES

This appendix is neither a catalogue of the Edessan coinage nor a systematic study. Its two main purposes are, ﬁrst, to provide illustrations of some of the coins that are most interesting from an historical point of view, above all those of Abgar X; and second, to present some of the evidence and argument concerning that king’s coins on which the conclusions in Chapter 4 have been based. The jumping-off point for this discussion is a 1969 article on the Abgar X coinage by Helge Gesche, the only recent numismatist to have devoted serious attention to the subject (Gesche 1969).

Abgar X: the restoration coinage1 The bronze coins of Abgar X all bear obverse portraits of Gordian III, laureate or radiate (Fig. App.1a). The most common reverse type among these coins, and the one with the greatest historical and iconographical interest, is that on Abgar’s large bronzes depicting what some historians have called an ‘investiture’ scene and we have chosen simply to call a ‘Presentation’ (Fig. App.1b). It is described by the Catalogue of the British Museum as follows:2 Gordian receiving Abgar; on l., Gordian, laureate, wearing toga, seated r. on sella curulis on suggestus, holding sceptre in l.; on r., Abgar, bearded, wearing diademed tiara, kandys and trousers, r. holding ﬁgure of Nike, l. on hilt of short sword. Two other reverse types are known in this denomination. On one, Gordian stands facing right, holding a globe and another object, while Abgar stands facing him with a wreath in his raised right hand (Fig. App.2).3 On this coin, as on the preceding one, Abgar 145

ROMAN EDESSA

Figure App.1a Abgar X, the ‘Presentation’ coinage (obverse), bearing a portrait of the Emperor Gordian III (238–44 ce) and thus acknowledging the sovereignty of Rome despite the brief resurrection of the Edessan monarchy. Photo by the author; used by permission of the American Numismatic Society.

wears the diademed tiara, trousers and kandys or long-sleeved caftanstyle Persian robe, while Gordian, in contrast to the previous coin, presents a decidedly military aspect in paludamentum and cuirass. Both of these ﬁrst two coins have reverse legends naming both rulers: ΑVΤΟΚ ΓΟΡ∆ΙΑΝΟC ΑΒΓΑΡΟC ΒΑCΙΛΕVC. A third reverse type on ﬁrst-denomination coins depicts the Edessan ruler, dressed as in the previous two cases, on horseback facing right (Fig. App.3; BMC Arabia, etc. 114, nos. 141–3). On some specimens the king’s left arm can be discerned, extending over the pacing horse’s neck as if raised to acknowledge the cheers of observers (see, e.g., Gesche 1969: pl. 3 no. 4). Here the legend names only the king: ΑΒΓΑΡΟC ΒΑCΙΛΕVC. Although a complete die-study on these coins is yet to be concluded, there is at least a fairly high degree of die linkage among the ﬁrst-denomination coins, especially those with the ﬁrst reverse type. The British Museum Catalogue notes ﬁve cases of obverse die linkage 146

N U M I S M AT I C N O T E S

Figure App.1b Abgar X, the ‘Presentation’ coinage (reverse). King Abgar X of Edessa, standing, greets Gordian III and presents him with a winged Victory ﬁgure, possibly indicating that Abgar played a military role in the defense of the empire against the Sassanid Persians. Photo by the author; used by permission of the American Numismatic Society.

among nine specimens, including a link between coins of the ﬁrst and second reverse types.4 Among specimens with the ﬁrst reverse type in the American Numismatic Society collection, nine obverse dies were used to strike fourteen coins. The largest group, with obverse bust of Gordian laureate with paludamentum and cuirass, shows the most die linkage: seven coins produced with three obverse dies. This high level of linkage, and the observation that Edessan coins seem to have dominated circulation in the region during these few years, support a picture of a steeply accelerated mint operation during the short reign of Abgar X.5 The coins of this ﬁrst denomination have a diameter of 31–3 mm and vary in weight from 15.70 to 22.99 gm, with a cluster of specimens weighing approximately 18–19 gm. In addition, the Edessan mint produced coins of two or three other denominations, all of which bear portrait busts of the two rulers on opposite faces, with no 147

ROMAN EDESSA

Figure App.2 On the reverse of another coin of Gordian III, Abgar hails the standing emperor, who holds objects symbolizing his universal sovereignty. Photo by the author; used by permission of the American Numismatic Society.

other type. The second, and most abundant, denomination, which bears a single star on each face as a subsidiary type or symbol, most probably represented a unit of currency; coins in this denomination measure 22–5 mm across and weigh 6.54–11.70 gm. Finally there are the fractions, third- and fourth-denomination coins weighing 4.28–6.10 gm and 2.81–3.64 gm respectively.6 It seems that this system may have been partially assimilated to the standard of Antioch, which produced no bronze during the reign of Gordian III, but whose third-century ‘colonial’ coins roughly approximate these in weight (Butcher 1988: 73–4, Tables I and II).

Lordship and victory: the Presentation scene The imagery of the Presentation scene eloquently expresses the relationship between the rulers, beginning with the details of their respective costumes. While Gordian wears the quintessentially Roman civilian garment, the toga, the sword-bearing Abgar appears 148

N U M I S M AT I C N O T E S

Figure App.3 In a reverse coin type suggestive of ‘Adventus’ imagery, King Abgar X is depicted on horseback, possibly after having had his throne conferred (in person?) by Gordian III, whose portrait appears on the obverse. Photo by the author; used by permission of the American Numismatic Society.

in clothing that could hardly be less Roman: the kandys or longsleeved, knee-length Iranian garment, with a row of buttons down the front, over the loose trousers.7 Abgar’s costume is especially reminiscent of that of Moqimu, the clan leader in the ‘Family Portrait Mosaic’ (Fig. 5.7), which should caution us against reading too much symbolism into what may be merely a realistic portrayal of the king’s customary dress. On the other hand, the king’s tall tiara, as a characteristically royal piece of headgear, necessarily carries some symbolic freight. The upright tiara – tiara recta or τια´ρα ρθ – was the prerogative of royalty in the Persian Empire and its successor states (Xen. Cyr. 8.3.13; Suda s.v. τια´ρα: κ σµο "πικεφα´λαιο· Aν ο> βασιλει µ νοι ρθ)ν "φ ρουν παρα` Π-ρσαι, ο> δ4 στρατηγο κεκλιµ-νην). The tiara appears in Syriac literature as well; in the Doctrine of Addai the ﬁrst Edessan disciple is a craftsman whose specialty was to make 149

ROMAN EDESSA

the tiara and silken garments of the king (Doct. Addai [Howard] 70–1). The particular form of tiara worn by Abgar X (and by earlier Edessan kings on their coinage) is unique; it has no exact parallels among the depictions of earlier Eastern rulers. Abgar’s tiara – on all his coins, but most of all on the ﬁrst-denomination pieces, especially those with the ‘Presentation’ scene – is higher than the normal royal headpiece, which is usually semicircular in shape. In the ‘Presentation’ scene, it is so elongated as to double the height of the king’s head, while the long ribbons of its diadem reach down to the level of his shoulder blades. The top ridge of the tiara on the coins bears a row of projections; these can be seen on the obverse busts of the second denomination as hooks, or some other type of curved appendage. The greater detail of the second-denomination busts also makes it possible to discern the decoration on the tiara, which consists of a network of dots connected by lines or, in some examples, a rosette of dots around a central dot. The closest parallel to this tiara is presented by those of Parthian kings from Vologaeses II (c. 105/6–47) to Artabanus V (c. 213–27). Although these tiaras usually do not stand up as high as Abgar’s, and generally carry a different form of decoration, they do have the hook-like attachments along the crest.8 On almost all the Parthian tiaras, a prominent feature is a set of lappets or ﬂaps hanging down from the side or back, or both. Numismatists to date have not noted any such feature on the tiara of the Edessan king’s coin portrait, and indeed it is difﬁcult to see any on the ﬁrstdenomination coins. Here again, however, the second-denomination units allow for greater certainty. Although by no means as prominent as the neck-covering ﬂaps of some of the Parthian kings’ portraits, a small lappet, no larger than an earﬂap, is clearly present on these reverse portraits.9 Thus, the coin portrait of the Edessan king ﬁts squarely into the tradition of Parthian royal portraiture, even as the monarch seems to swear allegiance to the ruler from Rome. In the left-hand portion of the scene, a striking element is the emperor’s seat, which – the BMC description notwithstanding – is deﬁnitely not a curule stool.10 As a matter of fact, this stool or chair is unlike any other known ancient piece of furniture – Greek, Roman, or Near Eastern. The legs of Gordian’s seat, of which two (and sometimes part of a third) are visible, differ in exact shape and spacing from one specimen to another. In all cases, however, they have curved upper portions resembling the arms of a lyre, above a less rounded lower portion in which the legs either remain parallel 150

N U M I S M AT I C N O T E S

or curve away from each other, ending in splayed feet. The legs support a thick cushion on which the emperor is seated. In no case do they touch or cross. By contrast to this, the legs of chairs, thrones and stools depicted on ancient coins and works of art – and those of actual pieces recovered by excavation – seem always to be crossed, as in the curule stool and its more ancient predecessors; straight and parallel, like the legs of Zeus Nikephoros’s throne on the well-known Hellenistic reverse type; or to splay away from each other in a continuous curve, like the legs of the ‘Chair of Hegeso’ on an Athenian gravestone.11 Chairs and thrones in Ancient Near Eastern art tend to conform to the same pattern.12 It is perhaps signiﬁcant that Gordian is seated on an atypical seat, rather than the curule stool that would be expected. The sella curulis had great symbolic weight, being the prerogative of a certain class of ofﬁcials, the ‘curule’ magistrates.13 Although other types of seat are known, the sella curulis is by far the most common seat on coin types which emphasize the emperor’s ofﬁcial status or his performance of his duties. On pieces which depict the emperor in a military context, the curule stool is replaced by the sella castrensis or camp stool, a less elaborate version of the folding stool with straight, crossed legs. Among the coin types which feature this form of seat are the rex parthis datvs and rex armeniis ‘Investiture’ types of Trajan and Lucius Verus, the closest parallels in Roman numismatics for the Gordian III/Abgar X pieces under discussion. A literary account of the ceremony in which the Armenian King Tiridates abased himself before Nero and received the diadem (in a political situation that was in some ways analogous to that of Abgar X under Gordian) is quite explicit in identifying the emperor’s seat during this ceremony as a sella curulis (Suet. Nero 13.1). In the face of these observations, the form of seat seen on the Edessan coins may justly be considered an anomaly that calls for explanation. In the only attempt at an explanation to date, Gesche (1969: 62) conjectures that the Edessan die-cutter transformed Gordian’s Roman ‘throne’ into one that was customary in the Parthian– Hellenistic region, and hence familiar in Osrhoene. This happened, she says, by a ‘more or less conscious’ (‘mehr oder weniger bewußten’) process (ibid.). For comparison, she cites the throne of Datames on fourth-century bce coins of Tarsus (BMC Lycaonia etc., Tarsus nos. 31–4; pl. 29, 11–13). This seat, however, although backless, has straight ‘turned’ legs similar to those of Zeus’s throne on the wellknown Hellenistic ‘Zeus Nikephoros’ coin type. It is difﬁcult to see 151

ROMAN EDESSA

how this could have served as the model for Gordian’s seat, unless we assume that the style of this throne was itself inﬂuenced by the Roman sella. In other words, the two visible legs, which are straight and parallel in the Parthian-style throne, could have been deformed and made to approach each other, resulting in a seat that could, on cursory examination, be taken for a curule or camp stool. This transformation – whether conscious or unconscious – would require the artist to have a degree of familiarity with the shape of the Roman stool, and perhaps with coin types on which it is depicted. This raises the further art-historical question of the appropriate background against which to view the imagery of the Edessan coin. We shall return to this question after completing our detailed description of the Presentation scene. The British Museum Catalogue describes Gordian as holding a sceptre. Often such a sceptre would be topped by an eagle, as is seen on some obverse types from Edessa.14 It is, however, quite clear from several specimens that this is not the case here; instead of an eagle, the coins display a spearpoint or arrowhead, sometimes quite a large one, on the tip of the shaft. This attribute, which Gesche identiﬁed as an arrow, was the focal point of her discussion of these coins. It must be said, however, that the identiﬁcation of the object in Gordian’s hand is not certain. While on some specimens it appears no longer than a normal arrow, on others its point reaches beyond the top of the seated emperor’s head, while it is unclear how far below his hand the shaft extends.15 A ﬁnal, and crucial, element in the Presentation scene is the object presented: the small Victory ﬁgurine which Abgar holds in his outstretched hand. The overall composition of the scene has the effect of focusing attention on this small ﬁgure, poised as it is on the verge of passing from one ruler to the other. In this transaction there may be some important clues to the two rulers’ relationship and to the message of the coin. In addition to the undeniable importance of military success as the ultimate ratiﬁer of royal and imperial standing, Nike herself was well established as a Greek and a Roman coin type – not least on Parthian coins and those of Alexander and the Seleucid kings, whose heritage was still strongly felt in this part of the world (Bellinger and Berlincourt 1962). The winged goddess appears frequently as a small ﬁgure in the hand of a god or a ruler, as the patroness who herself actively presents a symbolic object – typically a palm branch or a wreath – to the human who claims her blessing, and in a great variety of other postures. Nevertheless, scenes involving two human 152

N U M I S M AT I C N O T E S

ﬁgures and a Victory are relatively rare. In the Seleucid–Parthian tradition, such scenes seem to be completely lacking; the closest parallels to our type may be those on which the deity herself crowns a ruler.16 For instances of one historical personage – be he an emperor, king, or other individual – presenting another with such a Victory ﬁgure, we must turn to the Roman numismatic tradition; yet even here, they are rare enough. As Gesche observes, acclamation scenes in which a personiﬁcation of the Roman army presents the emperor with a Victory are of little use; here the symbolic content is entirely different.17 To this could be added the presentation of a Victory to the emperor by the goddess Spes (Hope), on a coin of Severus Alexander (BMC VI 204–5, nos. 926–8; pl. 30, no. 928). In this case the gesture is not an acclamation, but presumably represents the hope of a successful outcome to the emperor’s campaign. The same idea may be represented by a coin of Tarsus on which the Koinoboulion, the personiﬁcation of the regional council, presents a Nike ﬁgure to the Emperor Caracalla, although this is not so clear (Levante 1993: pl. 75, no. 1490). The point is that all of these presentation scenes necessarily convey a different message from those in which one historical individual presents a Victory ﬁgure to another. The few relevant parallels in Roman coinage, supplemented by an even smaller number of similar scenes in other media, are presented by Gesche (1969: 63–4). The group comprises medallions of Marcus Aurelius and Lucius Verus; Maximinus and Maximus; Philip I and Philip II; Valerian and Gallienus; Carus and Carinus, and Carus and Numerianus.18 Certain other reverse types probably represent the joint erection of a statue of Victory rather than a presentation scene.19 In all these cases, the two principals are either (actual or potential) colleagues in the imperial power, like Marcus Aurelius and Lucius Verus – the ﬁrst two Roman emperors to hold the title of Augustus simultaneously (161–9) – or a sitting emperor and his designated or presumed successor.20 In some cases, the individual who presents the Victory is the subordinate ruler or successor; in others, he is the supreme Augustus. In some cases, the Victory ﬁgure may represent military activity by the subordinate, who by this gesture shows or acknowledges that his success in the ﬁeld belongs to the ‘Oberkaiser’ under whose auspices it was attained. The symbolism of this type of scene can therefore be connected with the practice whereby the emperor celebrated triumphs on behalf of his generals, and with the concept of the indivisibility of the Victoria Augusti (Augusta).21 Since Abgar was neither a Roman emperor nor a presumed or potential successor to the purple, it is far from certain what 153

ROMAN EDESSA

relevance these Roman coins have for the scene in which a friendly king presents the emperor with a Victory. It was Gesche who connected the symbolism of this scene with the later episodes in which local Near Eastern leaders assumed a role in the defense of the empire, and either assumed or were granted imperial honors. With these incidents as background, Gesche (1969: 69–70) would explain the Victory-presentation scene between Abgar and Gordian as the expression of just such a partnership, although she leaves rather vague the precise dimensions of this partnership and whether it involved any actual military activity by Abgar. Raised to the status of a ‘partner’ with Gordian, she says, Abgar acknowledges and pays homage to his Oberkaiser, and reafﬁrms his loyalty – making plain that his enthronement presents no threat to the imperial order.22 With the help of Syriac Document A from the Euphrates archive, it is now possible to get a somewhat clearer picture of Abgar’s role and relationship with Rome. The document’s prescript tells us that he was ‘honored with the consulship,’ a phrase that is very puzzling but seems, on the basis of a comparison with slightly later documents from the same archive, to be a governorship of some sort. It is also now possible to get a nearly precise idea of the point at which Abgar took up this ofﬁce: it must be somewhere near April of 239, at which time the emperor was brieﬂy present at Antioch. The overall picture of Abgar’s regime leaves the impression that he was not a ‘partner’ of the emperor so much as a loyal servant – and it is in this sense that the presentation gesture can be convincingly interpreted. Without any additional evidence it would be hazardous to afﬁrm that Abgar’s governorship gave him a place in the Roman command structure, but it is a possibility. Another possibility is that he was able to command the loyalties of local forces, which may have been able to help stem the tide of the Persian advance, and that it is to this activity that the ﬁgure of Victory refers. This is purely speculative. In any case, whatever services Abgar provided, in the presentation gesture he lays them all at the feet of his emperor. Gesche’s interpretation of the gesture as that of a subordinate partner to his senior forms the second part of an argument whereby the object in Gordian’s left hand is seen as an arrow, a Persian and ancient Near Eastern symbol of lordship. Investigating the iconography of our scene in the Hellenistic, Persian and ancient Middle Eastern context, Gesche shows that the arrow often appears as a symbol of divine and royal lordship, as do the associated images of bow and quiver. This iconographical tradition can be traced back through the Parthians to the Seleucid kings, the Persian Empire and 154

N U M I S M AT I C N O T E S

to earlier civilizations.23 On the basis of Gesche’s reconstruction, it can be provisionally admitted that Parthian royal ideology took over the image of at least the bow, and presumably the arrow as well, from the Persian and perhaps earlier traditions, transmitted through the Seleucid kingdom. This pedigree for the imagery of the arrow enabled Gesche to postulate that the coin portrait of the Roman emperor with this attribute amounted to an acknowledgement of his sovereignty, and perhaps his divinity – albeit an acknowledgement of whose symbolic signiﬁcance the artist may have been only partially conscious (60 n. 76). It remains for us to observe, however, that many of the comparanda Gesche cites have to do, not with an arrow or arrows, but with bows and quivers. It may be doubted that when a Parthian coin depicts the king holding a bow, the idea of an arrow as a symbol of lordship is directly implied.24 Moreover, not one of the comparanda portrays an enthroned ruler holding a single arrow in sceptre-like fashion. On the coin of Tarsus proposed as a parallel for Gordian’s throne and pose, the ruler holds what Gesche calls an arrow, not upright, but horizontally. The closest approximation to Gordian’s gesture is the relief from Nineveh depicting, it is thought, the investiture of an Assyrian king (Gesche 1969: pl. 4 no. 11). This enthroned king holds in his left hand a bow, and in his right a bunch of arrows pointed upwards: an image that is still rather far, iconographically as well as chronologically, from that seen on our coin. It is perhaps forgivable, therefore, to have some reservations about the message of lordship embodied in Gordian’s arrow, which may not be an arrow after all. Whether or not those reservations are justiﬁed, however, Gesche’s approach provides a salutary example of the need to remain aware of Edessa’s multiple cultural and political afﬁliations. With this in mind we shall redirect our attention to the Presentation image one last time, looking at the message conveyed by the overall composition ﬁrst against the background of comparable imperial Roman coinage, and then in the Parthian numismatic context. The image of the emperor seated on the suggestus while engaged in various transactions is well established as a feature of a number of Roman coin types. For our scene, the best parallels are the coins of the emperors Trajan and Lucius Verus whose reverse types record the disposition of Eastern kingdoms as the result of successful campaigns by those emperors.25 Trajan’s regna adsignata and rex parthis datvs types portray the emperor seated on the suggestus, in the ﬁrst case receiving three Armenian kings whose kingdoms he 155

ROMAN EDESSA

assigned in 115 (Dio 68.18), and in the second placing a diadem on the head of King Parthamaspates, his newly installed client on the Parthian throne (Dio 68.30 – on this piece the emperor also extends his hand to the kneeling ﬁgure of Parthia on the far left). A closely related type, rex parthvs, shows the abasement of Parthamaspates’s predecessor, Parthamasiris, before the victorious emperor (Dio 68.19–20: a graphic description of the scene, down to Trajan’s seat on the tribunal). On all of these pieces the kings are on the ground before the emperor’s tribunal, facing him or (in the case of Parthamaspates) facing away while receiving the diadem. Chronologically closer to our piece are the reverse types of Lucius Verus, rex armeniis datvs (rex armen dat), c. 164. Here the emperor sits on a camp stool on suggestus facing left, crowning the Armenian king who, like Parthamaspates under Trajan, stands on the ground before him, also facing left. These scenes show a formal similarity to our Edessan coin type, with the emperor seated on a stool on top of a platform receiving an Eastern ruler. It is the differences, however, that are more striking. To begin with, the positions of king and emperor are reversed: whereas on all the Roman ‘Investiture’ types the emperor on his platform occupies the right-hand portion of the scene (or, in the case of L. Verus, sits at the center facing left), on the coin of Abgar it is the Edessan king who stands at the right, in the position of power. Equally signiﬁcant is his posture: whereas the subject rulers on the coins of Trajan and Verus are either physically smaller than their emperors, or are portrayed in poses of submission – or both – Abgar presents an imposing ﬁgure, drawing himself up to his full height. With the addition of the exaggeratedly tall tiara, his height thus equals or exceeds that of the seated emperor. The angular lines of Abgar’s pose – with his left arm on the prominent acinaces – make him an even more forceful ﬁgure. Finally, while the action in the Roman ‘Investiture’ types proceeds from emperor to king with the presentation of the diadem, here the king – already wearing his emblems of royalty – extends his hand to offer Gordian the ﬁgure of Victory, herself charged with signiﬁcance. Of course, the fact that this scene between king and emperor appears on a coin of Edessa, not one of Rome, has in itself a certain signiﬁcance. Along with the other points of contrast, it underlines the fact that, whatever the emperor’s role may be, the focus of the scene is on the king and his gesture. Likewise, whatever the precise import of that gesture, one effect – perhaps the primary effect – of the coin’s rhetoric is forcefully to associate the king of Edessa with 156

N U M I S M AT I C N O T E S

imperial authority. This message is underlined by the coins of smaller denominations, which bear portraits of the king and the emperor on opposite faces, with no apparent attempt to distinguish between obverse and reverse, and by the other ﬁrst-denomination types, particularly the scene in which Abgar presents a wreath to a standing Gordian in military dress. In our search for parallels to Abgar’s coin we earlier limited ourselves to scenes in which two actual historical personages engaged in a similar transaction involving a Victory ﬁgure, and found some parallels that were not only few in number but also rather far removed from the context of our coin. If we remove that limitation and look instead for scenes of a similar overall nature, we are able to reap a rich harvest among the Parthian coins. Many Parthian coins bear such images: usually they portray a goddess (Tyche) presenting a seated king with the tiara or a palm branch.26 Clearly there are differences between these and our coin; the object presented is different, and the king’s seat is always on the ground rather than raised on a platform, besides the obvious fact that it is a goddess who is doing the presenting rather than a human. It is, nevertheless, hard to avoid making this comparison, for in the cultural context of Edessa one must assume that its people, the intended target of any of the coins’ messages, were at least as familiar with the coinage of the kingdom that until quite recently had ruled a large expanse of a region quite close to the city. What we have seen of Edessa’s cultural background indicates that its people were open to both Eastern and Western inﬂuences, so it looks likely that they would see Abgar’s coin with the Parthian models in mind. If so, what is the signiﬁcance of the confrontation between the two rulers? It is scarcely believable that Abgar aspired to be seen as a divinity; but for those who were accustomed to seeing the Parthian reverse image, the appearance of the king in the place in the composition that was normally held by the goddess might very well have given added weight to his message of legitimacy and power. The implications of such a juxtaposition of visual ideas, however, are difﬁcult to puzzle out with precision. Putting the king in the goddess’s position certainly would not interfere with, and would probably strengthen, the message that we have already identiﬁed as the primary one – legitimation of Abgar through association with the emperor. The coin’s presumed Edessan ‘audience,’ however, lived in a state that had received successive waves of Roman interest and involvement – including the presence and presumed commercial 157

ROMAN EDESSA

activity of soldiers in the Roman army – since at least 165. It is very hard to say whether Roman or Parthian coins would have been more familiar to them as ‘models’ at this particular time. At the very least, the need to understand this coin in its confusing context underscores the impossibility of seeing ancient Edessa as primarily either Western or Eastern in its cultural orientation. Similar observations apply to the vexed question of Gordian’s seat in the Presentation scene. If a process such as the one imagined by Gesche did occur, there are Parthian as well as Roman models for a cross-legged stool as the ruler’s seat (Shore 1993: 158, no. 435, of Vologaeses IV, crudely represented). By far the most common seats among the Parthian coins, however, are the straight-legged thrones and the Omphalos. Only the Roman curule stool is prominent and common enough to have served as the model for the deformation of the true royal seat, so something like the process Gesche proposes may have occurred. Beyond this, there is one ﬁnal possibility, and it may be the best: the stool of Gordian may represent an idiosyncratic seat on which the emperor sat when he received Abgar in 239: a piece of furniture whose symbolic signiﬁcance, if any, is irretrievable.

The colonia renewed When Edessa’s mint resumed the striking of colonial coinage after the departure of Abgar X (under circumstances that are completely unknown to us), it returned to the reverse image of the city goddess, or Tyche, that had provided its standard reverse type since the founding of the colonia (Figs. App.4a–4b). As on other colonial coinages of the region, the image used here was derived from the Tyche of Antioch, a statue carved by Eutychides for the Syrian metropolis c. 296–293 bce (Pliny HN 34.51; Malalas 201.1–2, 276.6–9; Pausanias 6.2.7).27 The statue was enormously inﬂuential and was used by some 44 cities as a coin type, among them Carrhae, Edessa, Rhesaina and Nisibis in the Mesopotamian region.28 The goddess’s veiled, turreted bust began to be used at Edessa as a reverse type as early as the reign of Caracalla (BMC Arabia, etc. 97–9, nos. 39–54, pl. 14, nos. 11–15). Under Elagabalus, however, with the city’s promotion to colonia, ﬁrst-denomination coins appear bearing her ﬁgure in full, seated on a rock or mound of rocks with a ﬁgure representing the city’s river, arms outstretched, at her feet (BMC Arabia, etc. 99–101, nos. 55–68; pl. 14, nos. 17–18; pl. 15, nos 1–3). This remained, until Abgar’s accession, the standard reverse type of the large bronzes, and some of the smaller 158

N U M I S M AT I C N O T E S

Figure App.4a Obverse of a coin of the revived Colonia Edessa, portraying Gordian III and possibly minted in the last years of that emperor’s reign. Photo by the author; used by permission of the American Numismatic Society.

denominations (the reverses of the smaller coins generally portray only the veiled and turreted bust, facing right). After his departure, the type of the large pieces had changed. The ﬁrst-denomination coins of the restored colonia, successors to the ‘Presentation’ coins of Abgar X and their companion pieces, bear a large veiled and turreted bust facing left, in a noticeably different style from earlier busts. On these coins Tyche’s veil hangs down the back of her head, leaving her with a long uncovered neck (in contrast to the veil on the earlier busts, which wraps around the front of her neck). Her portrait is rendered in a cruder style, without the careful modeling of earlier representations. She faces a small ﬁgure on a pedestal which carries an object on its shoulder, possibly a local divinity or, as has also been suggested, a representation of the zodiac sign under which the colonia was founded.29 In between Tyche and the ﬁgure is a small representation of a pedimented temple, which appears as a 159

ROMAN EDESSA

Figure App.4b Colonia Edessa under Gordian III (reverse), showing city’s Tyche (a motif inspired by the Tyche of Antioch and used by dozens of cities across the Roman Empire). Photo by the author; used by permission of the American Numismatic Society.

main type on some of the earliest coins of Edessa and was used as a mint mark on the silver tetradrachms which Edessa produced for Caracalla. Edessa’s renewed colonial coinage and adoption of a new reverse type would not be particularly interesting were it not for the observation that its sister city and sometime rival, Carrhae, was striking remarkably similar coins just at this time (BMC Arabia, etc. 89, nos. 54–8; pl. 13, nos. 1–2). Carrhae’s Tyche under Gordian III has the same attributes and faces a small ﬁgure on a pedestal, in the same manner as that of Edessa; moreover, the two reverse types in this period are rendered in an identical style. In addition to the identifying legend – ΜΗΤ ΚΟΛ ΚΑΡΡΗΝWΝ, as opposed to ΜΗΤ ΚΟΛ Ε∆ΕΣΣΗΝWΝ on Edessa’s coins – the Carrhene pieces are distinguished by the presence of a crescent moon above the Tyche’s head. Still, the similarity cannot be denied; some of the Carrhene pieces even have the little 160

N U M I S M AT I C N O T E S

temple that is considered characteristic of Edessa (misidentiﬁed in the BMC, ibid. no. 54, as a small altar). This similarity suggests that, in the period after Abgar’s departure from the throne, the mint of Edessa was striking for Carrhae, and perhaps for the other Mesopotamian cities as well. Such an arrangement makes sense when we recall that none of the other cities had minted since the reign of Severus Alexander. It was Edessa, which had resumed minting with the installation of Abgar X, that was best prepared to produce a regional coinage upon the departure of the Sassanid occupiers. An attempt to ﬁnd obverse die-links between the Gordian III coinage of Edessa and that of Carrhae has not yet met with success, but it is to be expected that a systematic study of both coinages, and perhaps of the whole region during this reign, will produce interesting results bearing on the relative standing of Edessa and the other cities.

Edessan mint production Although there may be no completely satisfactory method for determining the volume of coinage produced at the Edessan mint – whether in absolute or relative terms – the record of coin hoards found in the Dura-Europos excavations probably offers as good a picture as we are likely to get of circulation, which in itself should reﬂect mint volume (Bellinger 1949: 165–91). From the analysis of 21 hoards discussed in the ﬁnal excavation report, 12 of which contained coins from the Edessan mint (ranging in date from Septimius Severus to Trajan Decius), it seems that the reign of Gordian III was the mint’s most active period by far. The ﬁnds of Edessan coins in Dura hoards can be classiﬁed as follows:30 Septimius Severus Elagabalus Severus Alexander Gordian III Trajan Decius Total

12 40 255 828 1 1,136

Of the 828 coins of the reign of Gordian, 620 were minted under Abgar X, and 208 – nine of them for Tranquillina – under the colonial administration. The eighth and ninth Dura hoards (actually a single hoard artiﬁcially divided through having been excavated in two stages) probably give a better picture of circulation in the 161

ROMAN EDESSA

mid-third century (Bellinger 1939). Of the 2,847 coins in this collection, covering the period from Domitian to Trajan Decius, 896 are from Edessa, and of those, 642 are of Gordian III’s reign. These can be classiﬁed as follows: 475 pieces of Gordian and Abgar; 159 of Gordian and the colonial Tyche, and eight of Tranquillina and the Tyche (Bellinger 1939: 25–8, with table facing p. 14). If this reﬂects the volume of mint production with anything near accuracy, the Edessan mint under Abgar produced three times as many coins for Gordian as it did under the colonia, even though the time periods involved are roughly equal. This observation provides one last index by which to judge the importance of the king who held the πατεα on Rome’s behalf, and held the line against the enemy during a period of some considerable anxiety.

162

NOTES

INTRODUCTION 1 Translation after Segal (1970: 19). Like pas.griba, nuhadra is an ofﬁcial title echoing Parthian usage; it occurs also in the inscriptions at Sumatar Harabesi near Urfa, perhaps with the meaning ‘military governor.’ The inscription is number 27 in the collection of H.J. Drijvers (Drijvers 1972). 2 Prof. Brown expressed this desideratum in a conversation with the author, February 1996.

1 2 3

4

5

6

1 THE EARLIEST EDESSA Cf. R. Payne Smith, Thesaurus Syriacus (Oxford: Clarendon Press, 1879), vol. 1, p. 38 s.v. DM , citing the lexicons of Bar-Ali, Bar-Bahloul and George Karmsedinoyo. Byzantine historians put the city’s foundation date at 304 bce: Eusebius, Excerpta 179; Cedrenus 292; Syncellus 520. Pliny, HN 5.21.1: ‘Arabia supra dicta habet oppida Edessam, quae quondam Antiochia dicebatur . . .’. It has been suggested that Pliny may have confused Edessa with Syrian Antioch, which was also surnamed Callirhoe – Duval 1892: 109–10. On the possibility of the name having been changed, Tscherikower 1927: 88–9. H . arran was populated by ‘Macedonians’ in this period (Dio 37.5.5); at Seleucia on the Eulaeus, an inscription of 21 ce shows the Parthian king dealing with the leading citizens in Greek (Welles 1974: 299–301, no. 75). Of the names in the Edessan king list preserved by the Chronicle of Zuqnin, the majority have endings in ‘-u’ that echo Nabataean usage; most names, including Ma nu, Bakru, Abdu, Sahru, and Gebar u, are actually found in Nabataean inscriptions. The Nabataeans were closely related to the Arabs; another Edessan royal name, Wael, is still in common use as a given name in Arabic today. These circumstances have led some scholars to assert that Edessa was ruled by a Nabataean tribe (Duval 1892: 110–12; Segal 1970: 16). This is the ﬁrst entry after the ﬂood of 201, which for an unknown reason is placed at the beginning of the chronicle. On the Era of Abraham and the Chronicle of Zuqnin’s list of kings, see Chapter 2.

163

ROMAN EDESSA

7 Among the names of the ﬁrst kings of Edessa, according to the Chronicle, were Orhai (the eponymous founder), Abdu, Phradasht the son of Gebar u, Bakru, Ma nu and Abgar ‘Peqa.’ 8 BMC Roman Empire I 110 nos. 679–82 (signis parthicis receptis); cf. Res Gestae Divi Augusti 29.2: ‘Parthos trium exercituum Romanorum spolia et signa reddere mihi supplicesque amicitiam populi Romani petere coegi.’ CAH X (1996) 158–63. 9 H. J. W. Drijvers, underlining the importance of Hellenism at Edessa, has in more than one place referred to the city’s designation as the ‘Athens of the East’ (Drijvers 1980: 10; 1992: 126; 1996: 175; Drijvers and Healey 1999: 38). I have been unable to locate the source or sources of this quotation. 10 Amida, not Edessa, would normally have formed a pair with Nisibis. To read ‘Amida’ for ‘Edessa’ in the text cited also eliminates an awkward doublet, for the passage ends by saying that next after Nisibis came Edessa of Osrhoene, ‘et ipsa civitas splendida.’ With or without the emendation, Edessa is ranked highly among the trading cities of this part of Mesopotamia. 11 As if to illustrate the lack of certainty over this question, we have statements from scholars as widely differing as that of Charlesworth, who described the route through Edessa merely as a ‘variant’ – fourth in a list of possible routes to the East – and, more recently, of Drijvers: ‘Edessa was a main station on the silk road to China’ (Charlesworth 1926: 100–1; Drijvers 1992: 138). 12 Although the purported date of the events in the Doctrine of Addai is shortly after the death of Jesus, the King Abgar who is described as converting to Christianity (and whose son was named Abgar Severus) must, on the contrary, be Abgar the Great, who ruled approximately from 176 to 211. This leaves aside the question of the historicity of the conversion narrative itself, to be taken up in Chapter 6. 13 Tabula Peutingeriana: Codex Vindobonensis 324, Facsimile edition (Graz: Akademische Druck-u. Verlagsanstalt, 1976). 14 See Dillemann’s reconstruction of the Peutinger routes on the presentday map of the region (Dillemann 1962: 148). 15 Despite the advantages of the fortiﬁcation system, it would seem that the hillside approach to the Citadel remained a weak point until the Emperor Justinian enlarged and strengthened the wall which brought it within the city enclosure (Procop. Aed. 2.7.13–16). 16 The springs appear in the Chronicle of Edessa’s ﬂood account and in Pliny’s statement, HN 5.21.1, that Edessa was formerly known as Antioch Callirhoe, a fonte nominata. 17 The belief in this ‘blessing’ is claimed as the reason for the Persian Chosroes’s desire to capture the city in 540: Procop. Wars 2.12.31. The blessing may have been absent from the text of Jesus’s letter known to Egeria before her visit to Edessa, since she remarks that the version she heard there was longer than the one she knew: vere amplius est, quod hic accepi, Itin. 19.19 (Arce). 18 The aqueducts are mentioned in the Acts of the Edessan martyrs Shmona and Guria in a seemingly genuine account from the time of the Diocletianic persecution (Burkitt 1913: 106 [Syriac 20] – in the same account

164

NOTES

19

20 21

22

23 24 25 26

27

the Roman governor of the city is described as presiding in the ‘Basilica [BSLYQ ] by the Winter Baths’: ibid. 101 [Syriac p. 15]). By the sixth century two aqueducts led into the city from the north; they were reconstructed in 505 by the governor Eulogius (Wright 1882: 69 [Syriac 81–2]). Edessa’s closeness to Armenia was such that the Armenian historian Moses of Khorene, in his desire to give an early pedigree to the Armenian church, created a family tree portraying close ties among the royal families of Edessa, Armenia and Adiabene (Moses of Khorene 2.26, 36; Duval 1892: 203–4). Edessa may also have successfully resisted Shapur during the invasion of 241 that was eventually turned back by Gordian III, but this is even less certain than its fate in 260. The etymology given in this passage, which would connect Osrhoene to the Persian King Chosroes, is certainly to be rejected. The Syriac name of Edessa – Orhai – and the name of the kingdom and later province (which also appears in inscriptions as Orrhoene) are almost certainly related both to each other and to the tribal name cited by Pliny (HN 6.31), ‘Arabes Orroei.’ All early Syriac sources including inscriptions talk of Orrha¯i and not Osrhai; ‘Osrhoene’ (Syr. ZRYN ) does not appear in any Syriac source until after the reign of Constantius (292–306). The dux of Osrhoene had command of nine units of equites including the Equites sagittarii indigenae primi Osrhoeni at Rasin; the Legio IV Parthica at Circesium; six alae and two cohorts. The following castella are listed in Osrhoene: Gallaba, Callinicum, Dabana, Banasam, Syna Judeorum, Oroba, Thillazamara, Mediana, Rasin, Circesium and Apatna. Also under the dux’s command, among other units, was an Ala Prima Parthorum at ‘Resaia’ (scr. Resaina): l. 30. This cannot be the important fort of Resaina/Theodosiopolis, for this is listed as a castellum under the Dux Mesopotamiae: Or. 36. It may be necessary, however, to correct the Notitia here and place it in Osrhoene – if only for reasons of geographical logic (Dillemann 1962: 107). Contra Dillemann 1962: 102 – relying on Pliny HN 5.21 – and Duval 1892: 90 n. 4 (Singhara and Carrhae independent of but surrounded by Osrhoene). See also the works of J. B. Segal (1953; 1963a). The Araboi are to be distinguished from the fully nomadic peoples – the Bedouin – who are called the T.ayyaye by Syriac writers, and Scenites or Saracens by the Greeks and Romans. The term rb appears in similar context in a group of Hatran inscriptions, some of them naming an ofﬁcial, the mlk dy rb or ‘king of Arab’ who may have performed a function similar to that of the shallita in the Sumatar texts (Segal 1986: 63–5). This Arab kingdom, therefore, provides a parallel for Edessa’s attempt to exert its control in the countryside in the gaps left by the large imperial powers. This Arab is probably not the ‘Arvastan’ of Armenian authors, designating the entire region east and west of Nisibis; nor does this phrase seem to refer to the widespread province of ‘Arbayestan’ or ‘Beth Arbaye,’ mentioned in Shapur I’s description of his empire and in

165

ROMAN EDESSA

Syriac ecclesiastical lists of a later period (Marquart 1901: 162–3; cf. Maricq 1965: 46–7 n. 5: ‘Il est clair que Be¯th Arba¯ye¯ est un calque de nom iranien ofﬁciel de la province . . . Arab est le nom indigène de l’ensemble de cette région; il peut désigner . . . aussi bien sa partie perse que sa partie romaine.’ References in the sixth-century ce chronicler ‘Joshua the Stylite’ to the Arab are in a rhetorical context that shows he is speaking of a western district under Roman control, not the entire northern Mesopotamian region ( Joshua the Stylite [ed. Wright 1968]: 27, 39 [Syriac pp. 33, 46]). Again, in the work of the Syriac historian pseudo-Zacharias Rhetor (around 569), the ‘region of Arab’ ( TR D RBY ) is a western region needing to be protected from the incursions of the Persians and the Tayyaye or Bedouin (CSCO Scr. Syr. ser. 3 vol. 6, pp. 24, 63; Syriac pp. 35, 92). 28 For the coins, BMC Arabia, etc. 81, nos. 1–2 (Anthemusia = Batnae under Caracalla, legend ΑΝΘΕΜΟΥΣΙΩΝ); 82, no. 1 (Carrhae: already ΚΑΡΗΝΩΝ ΦΙΛΟΡΩΜ[ΕΩΝ] under Marcus Aurelius), 83 no. 5 (Carrhae under Septimius Severus: ΚΟΛΩΝΙΑ . . . ΚΑΡΩΝ); 119 nos. 1–4 (Nisibis under Macrinus, Severus Alexander: ΣΕΠ[ΤΙΜΙΑ] ΚΟΛ[ΟΝΙΑ] ΝΕΣΙΒΙ; 125, nos. 1–2 (Rhesaina under Caracalla; cf. 127, nos. 10–14, Trajan Decius: ΣΕΠΤΙΜΙΑ ΚΟΛΟΝΙΑ ΡΗΣΑΙΝΗΣΙΩΝ). 29 This identiﬁcation, which ﬁts with the generally ‘Hellenized’ character of Document B as compared with Document A, was already suggested by Teixidor (1990: 156, with n. 26), and is also supported by the Syriac version of the Acts of Chalcedon, where ‘Haikla da-S.ida’ takes the place of Marcopolis in the Greek version. 30 Later Church sources – the Acts of the Ecumenical Councils and the Notitia Antiochena of 570 – are more explicit; it appears that the bishops of Carrhae, Birtha, Maratha, Constantia, Marcopolis, Batnae, Tell Mahre, Circesium, Dausaron, Callinicum, Hemerias and Nea Valentia pertained to the district of Osrhoene (Honigmann 1925: 60–88). In 458, ﬁve of the nine Osrhoenian bishops who wrote a letter to the Emperor Leon were those of Edessa, Callinicum, Circesium, Carrhae and ‘Balien[a]’ – possibly a misspelling of Batnae: Concilium Universale Chalcedonense (ed. Schwartz) II.5, 40, 41. Cf. ibid., II.6, 96. Osrhoene is here, again, conceived as extending from the Euphrates as far east as Tella/Constantia, and as far south as the eastward bend of the Euphrates. 2 THE COMING OF ROME 1 Cf. Trajan’s arabia adq(uisita) coins, BMC Roman Empire III 96–7, nos. 474–7; 207, nos. 977–81; 211, no. 997 et al. The Nabataean kingdom was annexed as the Province of Arabia in 106; see in general Bowersock 1983. 2 See Brock 1992a: 10–13; Ortiz de Urbina 1965: 211–12; Witakowski 1987. Chronological questions pertaining to the later Edessene kings have recently received a satisfactory treatment by Millar (1987: 559–62). 3 Chron. Zuq. 56/44: The birth of Jesus in Abraham 2015 also given as Year 309 ‘of Alexander of Macedon’; on the same page, the statement

166

NOTES

4

5 6

7

8 9 10

1

2

3

4

that the Edessans count the years from Year 1706. The death of Caracalla (217 ce) is recorded under the Year 2133, which with the subtraction of 1706 gives 216/17 ce. Σεουηρο δ- . . . κατα` των βαρβα ´ ρων "πιθυµ!6 δ ξη "στρα´τευσε, των τε Ορροηνων κα των Αδιαβηνων κα των Αραβ!ων (Dio 75.1.1). Cf. 48.17.1 (on Trajan): Μετα` δ4 ταυτα "στρα´τευσεν "π Αρµεν!ου κα Πα´ρθου, πρ φασιν µ4ν Bτι µ) τ3 δια´δηµα π ατου ε+λφει, αλλα` παρα` του Πα´ρθων βασιλ-ω, = των Αρµεν!ων βασιλε, τC δ αληθε!6 δ ξη "πιθυµ!6. On ‘Assyria,’ which may be an invention of Late Roman propaganda, see Lightfoot 1990. Gutschmid 1887: 30; cf. Duval 1892: 210–12. By the evidence of countermarks, it has been suggested that Roman coins were in use at Edessa before these ﬁrst local coins were minted (Howgego 1985: 109, 250, nos. 26, 725). But if the Syriac (Estranghelo) ‘M’ accompanying this countermark stands for MLK , as Howgego allows, it could just as well be the Parthian as the Edessan king that is portrayed. A scribal error is responsible for giving this 12-year reign to ‘Ma nu the son of Izates’; in fact the ruler is the same Ma nu the son of Ma nu who reigned before Wael, as the chronicle makes clear by specifying that he ruled a total of 36 years. For the rest of this paragraph I follow the interpretation and arrangement of events as re-created by Anthony Birley (Birley 1987: 121–30). SEG 2.817 = Cumont 1926: 410, no. 53: inscription of the epistate¯s Aurelius Heliodoros in honor of L. Verus. Cf. Lucian Hist. Conscr. 20, 27–8: the ‘Battle of Europus.’ Chronological uncertainty is deepened by the fact that Marcus Aurelius is not the only emperor who could have given Aurelius H . afsai his nomen and his status as ‘freedman of Antoninus.’ Those could just as easily have come from Caracalla, who in fact had a closer connection with this region than M. Aurelius. To accept this hypothesis, however, would place Bar Nahar more than 40 years after the other rulers of Arab, which seems unlikely. 3 FROM KINGDOM TO PROVINCE The description of Abgar as ‘king of the Persians’ by the much later Historia Augusta suggests that Edessa may have been still viewed as alien and potentially hostile: HA Severus 18.1, ‘Persarum regem Abgarum subegit.’ Hasebroek 1921: 75; Birley 1988: 115 (‘Septimius had already formed a serious, long-term design to extend the eastern frontier’); Mitford 1980: 1206 (‘Severus was determined to retain control of the southern approaches to Armenia’); cf. Millar 1993: 124. According to Dio, Severus’s campaign against the Osrhoenians, Adiabenians and Arabians took place during the siege of Byzantium. The paragraphs describing the reasons for this campaign and its aftermath (Dio 75.1.2–3) are found in a separate extract, Exc. UG 69. Dio 75.1.2–3, tr. Cary (Loeb edition). The Greek text reads: < Οτι ο> Ορροηνο κα ο> Αδιαβηνο αποστα´ντε κα Ν!σιβιν πολιορκο υντε, κα ττηθ-ντε 2π3 Σεουρου, "πρεσβεσαντο πρ3 ατ3ν µετα` τ3ν του Ν!γρου θα´νατον, οχ Bπω E

167

ROMAN EDESSA κα Fδικηκ τε παραιτοµενοι, αλλα` κα εεργεσ!αν απαιτουντε E κα 2π4ρ ατου τουτο πεποιηκ τε· το' γα`ρ στρατι,τα τα ` του Ν!γρου φρονσαντα Gλεγον "κε!νου Hνεκα "φθαρκ-ναι. κα τινα κα δωρα ατI Gπεµψαν, το τε α+χµαλ,του κα τα` λα´φυρα τα` περι ντα αποδ,σειν . ο µ-ντοι οKτε φρουρο' λαβειν ;θελον , 2π-σχοντο αλλα` κα το' λοιπο' "ξαχθηναι "κ τη χ,ρα Fξ!ουν. δια` ταυτα = π λεµο ο2το συν-στη.

5 Drijvers 1965: 56–7; for this interpretation, see Bowersock 1983: 79n.; cf. also Kennedy 1987: 65. It seems unlikely, however, that this refers (as Kennedy suggests) to the annexation of the rest of Osrhoene by Caracalla; Syriac authors do not normally call Osrhoene itself ‘Arabia.’ 6 As, e.g., Mommsen, R. Gesch. 5, 409, n. 1; Hasebroek 1921: 75; J. Sturm, RE 17, col. 736 s.v. Nisibis; Birley 1988: 115; Drijvers 1977: 877–8 (the last, however, should not be considered as viewing these events ‘through the lens of Roman history’). 7 Two of the earliest modern historians of Edessa, taking note of the statement in HA Severus 18.1, considered it certain that Osrhoene and Adiabene were acting with the backing of Parthia in a revolt against Pescennius Niger: Gutschmid 1887: 34; Duval 1892: 213–14. Segal (1970: 13–14) also speaks of a ‘pro-Parthian uprising in Mesopotamia’ following Pertinax’s death, and of Osrhoene and Adiabene as probably hoping ‘that the Roman hold had weakened.’ 8 Laodicea received the ius Italica ‘ob belli civilis merita,’ Dig. 50.15.1.3. For Antioch as Niger’s mint see BMC Roman Empire vol. 5 (Pertinax to Elagabalus), cvii–cviii. During the conﬂict with Niger, denarii were minted in the name of Severus at Laodicea and one other Syrian mint which has been traditionally identiﬁed as Emesa. Similarities between the output of ‘Emesa’ and Niger’s own coinage, however, are so extensive as to make some scholars hypothesize that Niger’s mint was taken over by Severus at some point during the conﬂict: Buttrey 1992; Kevin Butcher, Keeper of Roman Coins, Fitzwilliam Museum (private communication). 9 C. Iulio Pacatiano [v.e.], proc. | Augustorum nostrorum, militiis | equestribus perfuncto, proc. provinc. | O[sr]hoenae, praefecto legionis Parthi | cae, pr[o]c. Alpium Co[t]tiarum, adlecto | inter comit[es A]uggg. nnn., procurator. | pro legato provinc. Mauretaniae Tingi | tanae, col[o]nia Aelia Aug. Italica | p[atr]ono merentissimo. CIL 12.1856 = Dessau 1353. See also CIL 3.865 (Dacia); 6.1642 (Rome); Pﬂaum 1960: 605–10, no. 229. 10 Magie (1950: 1544), modiﬁes the common view by suggesting that the province was created during Severus’s second Parthian campaign, around 197, and that at this time Abgar was allowed to retain a small portion of his kingdom. 11 Whether the Province of Mesopotamia had yet been created this early is still unknown, but it is possible that Dio’s statement, quoted above, about the status of Nisibis (Dio 75.3.2) reﬂects that province’s creation simultaneously with Provincia Osrhoena. It was certainly in place by 198, the date preferred by some authors (e.g. Kennedy 1979: 262). Although Pacatianus may have been originally based at Nisibis, it seems unlikely that the city was actually in Osrhoene, as Kennedy suggests.

168

NOTES

12 Severus suspended his eastern campaigning because of the need to eliminate all rivals: πρωτον δ4 πασαν τ)ν αρχ)ν 5 Ρωµα!ων " Lαυτ3ν κα το' παιδα µεταγαγειν κα βεβαι,σασθαι Fθ-λησε, Hdn. 3.5.1. 13 Severus is already Arabicus and Adiabenicus on the boundary inscription quoted above. Additional evidence for these titles, and for ‘Parthicus,’ comes from such inscriptions as CIL 8.306 (= Dessau 417) and 6.1033 (= Dessau 425). 14 Cf. Birley 1988: 116. Severus’s armies did not, in fact, engage in any signiﬁcant confrontation with Parthia at this time. The inscription of Severus’s general Ti. Claudius Candidus, however (Dessau 1140), proves that this campaign was called the ‘Parthian War.’ It would have been politic for Severus to close this phase of his campaigning with a declaration of victory over the Parthians before turning his attention temporarily toward the struggle with Albinus. 15 Hdn. 3.9.2, tr. Whittaker (Loeb ed.). The Greek text reads: προσ-φυγε δ4 ατI κα = Οσροηνων Βασιλε' ΑKγαρο, το τε παιδα =µηρεειν " ασφα´λειαν π!στεω "ξ-δωκε, τοξ τα τε πλε!στου συµµα´χου ;γαγεν. 16 Magie 1950: 1542, approved by Drijvers 1977: 877–8. There is no need, however, to assume with Magie that this episode, and not a visit to Rome, is the basis of Dio’s remark (79.16.2) about the pomp surrounding Abgar. 17 There may be epigraphic evidence for the sons whom Abgar left as hostages in the inscription IG 14.1315 (IGR 1.179 = CIG 6196), found in Rome, which relates in elegiac verse the death of Abgar, age 26, and the erection of his tomb by his brother Antoninus. Both were sons of ‘the former king Abgar’ (= πρν Βασιλε' Αβγαρο, l. 6), which suggests that Abgar’s sons remained in Rome after his death. The lack of context for this inscription, however, leaves it open to various interpretations, and it has formerly been taken as memorializing the Abgar who was deposed by Caracalla, among other candidates. Herodian describes the emperor’s hostage policy, particularly as it related to the Eastern provinces: Hdn. 3.2.5. 18 Dio 75.2.3, tr. Cary (Loeb). The Greek text reads: µετα` δ4 ταυτα " τ)ν Ν!σιβιν = Σεου ηρο "λθMν ατ3 µ4ν "νταυθα 2π-µεινε, Λατεραν3ν δ4 κα Κα´νδιδον κα Λα βαρβα´ρου αλλον αλλC απ-στειλε, κα "περχ µενοι ιτον " το' προειρηµ-νου ο2τοι τν τε χ,ραν των βαρβα´ρων "δOουν κα τα` π λει "λα´µβανον. 19 Severus crossed the Euphrates, almost without a doubt, at Zeugma, almost directly west of Edessa (so also Hasebroek 1921: 75; Birley 1988: 115). From here the only road shown on the Late Roman Peutinger Tablet led through Rhesaina, southeast of Edessa: Tab. Peut. p. 10; Dillemann 1962: 171–5. 20 AE 1930, 141: milestones of 195 mentioning Syria Phoenice, hence showing that Syria had been divided by then. Literary sources on this division are scarce, but Dio 55.23.2 mentions Legio III Gallica as being stationed in Phoenice in his day; Ulpian lists the coloniae of Syria Phoenice, Dig. 50.15.1.1. See also Downey 1961: 239 n. 17; Ingholt 1932: 282–6. 21 The early date for Pacatianus’s procuratorship of Osrhoene lends credence to the conjecture of Pﬂaum (1960: no. 229), that he was the ofﬁcer mentioned by Herodian, 3.6.10, who was charged with holding

169

ROMAN EDESSA

22

23 24

25

26

27

28

29

30

the Alpine passes against an advance on Rome by Albinus. After Osrhoene, Pacatianus’s cursus lists ‘praefecto legionis Parthi | cae, pr[o]c. Alpium Co[t]tiarum’, CIL 12.1856, ll. 4–5. On Pacatianus as prefect and on the formation and stationing of the three Parthian legions, see Kennedy 1987: 59–62. AE 1984, no. 18: trib pot xiii, imp xii, cos iii. Ianuarius was already known as a procurator of Osrhoene from CIL 2.4135 (= Dessau 1365): Pﬂaum 1960: no. 342, pp. 892–4, giving the name as Aurelius Januarius. Κ.α´νδιδο Pρθωσ-ν µ.ε. [τ] | 3ν Qγγι ν ποτε βωµ.ο.ν. [Φ]ο!βου Απ λλωνο µα. | ντεµασιν αθανα´τοισι.ν.· η]µ δ4 Ζην3. αγαλµ.[α] | ".ρ.ι.σ.θ.-νεο βασ!ληο.. Mitford 1972: inscr. no. 77. An undated inscription from a legion at Aphrodisias in Caria mentions Singhara ("ν Σιν[γ]α´ | ροι τη Μεσοποτα | µ!α πρ3 τI Τ! | γρει ποταµI), bearing the epithet Antoniniana, which is known to have been carried by Legio Parthica I: Dessau 9477; cf. Amm. Marc. 20.6.8. An entry in the Notitia Dignitatum seems to place Prima Parthica at Nisibis in the fourth century (Not. Dig. Or. 26.29), but this is not good evidence for its original location. Castelin 1946: passim; BMC Arabia, etc. cx, 126–33. Coins are not conclusive, however; see Kennedy 1987: 61; Isaac 1980/1: passim. Millar (1993: 126, cf. 146) identiﬁes the two legions of Mesopotamia as I and II Parthica, apparently by an oversight. Ofﬁcials mentioned in inscriptions as having charge of the province include C. Iulius Pacatianus himself, who may have returned to serve as prefect of the province before retiring (Pﬂaum 1960: no. 229); Titus Claudius Subatanius Aquila, ‘ﬁrst prefect of Mesopotamia’ (Kennedy 1979: passim); an unidentiﬁed proc. Mesopotamiae (Pﬂaum 1960: no. 281) and a praepositus Mesop. (Fitz 1969: 140). Cf. Magie 1950: 1544. Nisibis: BMC Arabia, etc. 119–24 (rev. legends ΣΕΠ. ΚΟΛ. ΝΕΣΙΒΙ, etc., beginning with Macrinus); Rhesaina: ibid. 127–33 (rev. ΣΕΠ. ΚΟΛ. ΡΗΣΑΙΝΗΣΙΩΝ, etc., reign of Trajan Decius); Singhara: ibid. 134–6 (rev. ΑΥΡ. ΣΕΠ. ΚΟΛ. ΣΙΝΓΑΡΑ, Gordian III and Tranquillina). Cf. Magie 1950: 1544; Jones 1971: 220–1. Carrhae does not seem to have adopted Septimius’s name, but cf. BMC Arabia, etc. 83, no. 5 (a coin of Severus with reverse legend ΚΟΛΩΝΙΑΣ – ΜΗΚΑΡΩΝ). [The athlete Zoticus, during the reign of Elagabalus] "µηνθη Aurelius τε ατI 2π3 των ταυτα "ξεταζ ντων , κα "ξα!φνη "κ των αγ,νων αναρπασθε ανχθη τε " τ)ν 5 Ρ,µην 2π3 ποµπη απλ-του κα Bσην οKτε ΑKγαρο [VC; αγβαρο cod. Peir.] "π του Σεουρου οKτε Τιριδα ´ τη "π του Ν-ρωνο Gσχε . . . λπην τε αζµιον κα χαρα`ν ακερδη, Persian Wars 2.12.19. The epigram could reﬂect the Christian hostility to the vanity of the Roman spectacles, expressed frequently by, among others, Edessa’s own Archbishop Rabbula. ΗπατηκM γα`ρ τ3ν βασιλ-α των Οσροηνων ΑKγαρον E δ) παρα` φ!λον ατ3ν Aκειν, Gπειτα συλλαβMν Gδησε, κα τ)ν Οσροην)ν ο@τω αβασ!λευτον οσαν λοιπ3ν 2 "χειρ,σατο – Dio 78.12.1 .

31 Edessa, like the Greek cities of Syria, followed the ‘Macedonian’ system by which the Seleucid year was reckoned as beginning in October. 32 A. R. Bellinger (Bellinger and Welles 1935: 143) reckons the ﬁrst year

170

NOTES

33

to have been 213/14, an apparent error which I followed (Ross 1993: 189), but which is corrected by Millar (1993: 476).

< Οτι Αγβαρο = των Οσροηνων βασιλε' "πειδ) α<παξ "ν κρα´τει των =µοφλων "γ-νετο, οδ4ν B τι των δεινοτα´των το' προ-χοντα ατων οκ "ξειργα ´ σατο. λ γI µ4ν " τα` των 5 Ρωµα!ων ;θη µεθ!στασθαι Fνα´γκαζεν, GργI δ4 τη κατ ατων "ξουσ!α απλστω "νεφορειτο – Dio 78.12.1a.

34 Gutschmid 1887: 36; Duval 1892: 221; Millar 1993: 144. Segal (1970: 14) sees Abgar’s cruelty as the ‘pretext’ used for his removal by Caracalla. 35 ‘Antoniana’ for ‘Antoniniana’: J. and L. Robert, Bull. Epigr. 1974, no. 577. 36 Examples of such misspellings include: Dio 40.20.1 ( Αβγαρο); 68.18 (αλβαρο); 68.21 (ΑKγαρον); 78.12.1a ( Αγβαρο); 79.16.2 (ΑKγαρο – some mss. αγβαρο). Many of these, however, surely result from the copyists’ confusion between β and υ, which in later Antiquity were

pronounced identically (as ‘v’).

37 Chron. Min. 282/211. The much later historian Michael the Syrian reproduces this section of Jacob very closely (Chabot 1963: vol. 4, 77–8 [translation: vol. 1, 120]). 38 Coins of Abgar Severus: BMC Arabia, etc. ciii, pl. XIV.10 (legend ΣΕΟΥΗ ΑΒΓΑΡΟΣ); Babelon 1893: 260–4 (‘barbaric’ legends). 39 Ross 1993: 194–5. I wish now to retract this position after fuller consideration of the Syriac sources. Cf. Teixidor 1990: 160–1. 40 ’Ma nu n’a donc eu de la royauté que le titre sans la puissance’ (Duval 1892: 223). 41 Not directly relevant, but interesting, is the inscription from Hatra that seems to arise from the aftermath of just such a confrontation between city dwellers and tribespeople: Dijkstra 1990. Hatra, like Edessa, tried to administer some of the surrounding area, the ‘Arab, by appointing a governor: Segal 1986, 63–5. 42 Syncellus 440–1. Dodgeon and Lieu (1991: 9–15) conveniently collect the scattered sources for these events, as for many others during the ﬁrst phases of the Sassanid–Roman conﬂict. 43 ‘Terras interamnas, Mesopotamia scilicet, neglectas ab impura illa . . .’, HA Sev. Alex. 56.6. Macrinus’s defeat and change of policy: Dio 79.26.2–6; 27.1–3. 44 Hdn. 6.6.6. According to the Historia Augusta, Alexander defeated Ardashir in a great battle: HA Sev. Alex. 55. 45 Zosimus 1.12; Syncellus 1.674–5 (Bonn). Other accounts talk of a pretender named Taurinus, possibly a mistake for Antoninus: Epit. de Caes. 24; Mommsen, Chron. Min. 1.521. Cf. Jardé 1925: 66. Attempts to make Emesa, not Edessa, the location of this rebellion – on the basis of the coinage of ‘Uranius Antoninus’ from the former city – stumble on a detailed analysis of this coinage which shows that it cannot be any earlier than 248 (Baldus 1971: 85, 166–8). 46 Dio and Herodian both make note of the importance of the Osrhoenian archers in the force with which Alexander and Maximinus invaded Germany: Dio 78.14.1; Hdn. 6.7.8, 7.1.9, 7.2.1; cf. HA Sev. Alex. 61.8, Max. Duo 11.1, 11.7.8. For inscriptions, CIL 11.3104 (= Dessau 2765) (praeposit[o sagittari]is Orrhoenis); Dessau 2540, funerary inscription of

171

ROMAN EDESSA

Barsemis Abbei ‘ex numero Hos | roruorum’. Also CIL 11.3104, 13.6677a ([Osrhoe]nor[um]: possible reading of rasura). 47 Hdn. 7.1.9; HA Max. Duo 11.1; Gord. 26.6; BMC Arabia, etc. cv. It is, however, possible that Edessa was occupied by Sassanid forces at this time, along with Nisibis and Carrhae; if true, this would account for why it minted no coins for Maximinus. 48 See the translations of the dating prescripts of these documents in Chapter 4.

1

2

3 4 5 6

7

8

9

10

4 A KING IN ROME’S SERVICE Abgar’s Romanizing name appears in one of the recently discovered parchments from the Middle Euphrates, which also allows us to set the years of his rule at 239–42. On his coins he is only ΑΒΓΑΡΟΣ ΒΑΣΙΛΕΥΣ. Gutschmid and others, noting the inscription Dessau 857 (CIL 6.1797) from Rome of ‘Abgar | Prahates | ﬁlius, rex | principis Orrhenoru,’ assign the surname Phrahates to Abgar X. This inscription, however, is enigmatic and difﬁcult to date correctly; besides, the contemporary legal document – which might have been expected to use the king’s full nomenclature – argues against this identiﬁcation. Cf. BMC Arabia, etc. cv–cvi. In this section I am heavily indebted to the work of past scholars, most notably Xavier Loriot (1975), E. Kettenhoffen (1983), and Dr Roger Bland, from whose unpublished dissertation on the mints of Antioch and Caesarea under Gordian III (1991) I have beneﬁted immeasurably. HA Gord. 15–16, 22; Max. et Balb. 14. For Gordian’s age, Loriot 1975: 725 n. 525. For the correct name of Gordian’s father-in-law, CIL 13.1807 = Dessau 1330. On his personality and policies, Loriot 1975: 735–43. HA Gord. 26.4; Loriot 1975: 767. For Gordian’s land route through Asia Minor, Kettenhoffen 1983: 152–3; map p. 171. Among the large literature on this subject the following are particularly relevant: Sprengling 1937 (ed. princ.); Sprengling 1953; Honigmann and Maricq 1953; Maricq 1965; Rostovtzeff 1943; and for our present purposes, Loriot 1975, Kettenhoffen 1983. Syncellus P361, p. 681; Zonaras 12.18; Loriot 1975: 759–60; Kettenhoffen 1983: 155, dating the Persian advance on Mesopotamia to 236. Shapur’s inscription is silent on these developments, beginning its narration with the invasion under Gordian. HA Gord. 27.6: A ‘letter’ from Gordian to the Senate indicating Ctesiphon as his goal. Although this letter is surely forged, the fact that the ultimate battle with Shapur occurred in ‘Asuristan’ (Babylonia), lower Mesopotamia, shows that the emperor did attempt to push on farther south. Res Gestae divi Saporis ll. 3–4 (trans. Frye 1984: 371, from Middle Persian and Parthian supplemented by Greek). Asuristan or ‘Assyria’ in this inscription is Babylonia, while Peroz-Shapur, ‘victorious (is) Shapur,’ is Tell Anbar in southern Iraq: Honigmann and Maricq 1953: 112–18; Maricq 1965: 94–7. Res Gestae divi Saporis l. 12 (Frye): ‘We made prisoner ourselves with our

172

NOTES

11

12

13

14

15 16

own hands Valerian Caesar and the others, chiefs of that army, the praetorian prefect, senators; we made all prisoners and deported them to Persis.’ HA Gord. 29–30; Zosimus 1.18–19. Differing in details but all laying stress on Philip’s treachery are: Aur. Victor 27.7–8; the Epitome de Caesaribus 27.1–3, Eutropius 9.2, Syncellus 443 and Zonaras 12.18. Cf. Orac. Sib. 13.13–20, Amm. Marc. 23.5.17. George Monarchus, Chronicon (Teubner) 461; George Cedrenus, Historiarum Compendium (Corpus Scriptorum Historiae Byzantinae 24, ed. I. Bekker, Bonn, 1838) 29; Chronicon Paschale, Corpus Scriptorum Historiae Byzantinae 9.1 (ed. L. Dindorf, Bonn, 1832) 501; the Synopsis Sathas (36.16); Zonaras 12.17. Bland (1991: 510) observes that this account is ‘banal enough to have the ring of truth about it.’ On the other hand, death after a fall from a horse was a topos of classical literature; we will probably never know the true cause of Gordian’s death. Gutschmid 1887: 44–6; Babelon 1893: 286–92; Loriot 1975: 768–9 nn. 822–3; Kettenhoffen 1983: 154; Bland 1991: 504; Gesche 1969: 72 n. 129. Already in 1935 A. R. Bellinger proposed an alternative chronology for Abgar, which as it turns out is much closer to the correct version as we can now reconstruct it: Bellinger and Welles 1935: 144–6. The evidence: Dura Parchment 28, dated to Iyyar of the Seleucid year 554 (May 243) and to ‘Year 31 of the freedom of the illustrious Antoniana Edessa, Colonia Metropolis Aurelia Alexandria’; P. Euphr. Document A, from December of Seleucid 552 (December 240), ‘the second year of King Aelius Septimius Abgar, son of Ma nu the pas.griba’; P. Euphr. Document B, September of Seleucid 553, ‘in the Year 30 of the freedom of the renowned Edessa Antoniana Colonia Metropolis Aurelia Alexandria.’ Bellinger and Welles, 1935; Teixidor 1989: 219–22; Ross 1993: 187–206. Sebastian Brock (1991: 265) summarizes these ﬁndings. Teixidor 1990: 147, 150; ll. 1–7. The text given here follows Teixidor’s reading and translation; possible differences of interpretation will be discussed in what follows. Prescript of Document B: ‘In the ﬁfth year of the Autokrator Caesar Marcus Antonius Gordianus Eusebes Sebastos, in the consulship of Vettius Atticus and of Lepidus Praetextatus, in the month of September of the year 553 of the former reckoning; and in the year 30 of the freedom of the renowned Edessa [Syr. DYS’] Antoniana Colonia Metropolis Aurelia Alexandria. This document was written in Marcopolis Thera, during the priesthood of Marcus Aurelius Antahyrus (?), son of Aggai [Brock: ‘Marcus Aurelius BM (?), hiereus, son of KY (?)’], and during the archonship of Marcus Aurelius Alexandros, son of Severus, and of Baratha son of Shalamsin, on the ﬁrst day of the month’ (Teixidor 1990: 147, 150; cf. Brock 1991: 264). Prescript of P. Dura 28: ‘In the sixth year of the Autokrator Caesar Marcus Antonius Gordianus Eusebes Eutyches Sebastos, in the consulship of Annius Arrianus and

173

ROMAN EDESSA

17 18

19 20

21 22

23 24 25

26

27

Cervonius Papus; in the month Iyyar of the year 554 of the former reckoning, and in the year 31 of the freedom of the renowned Antoniana Edessa [Syr. DS ], Colonia Metropolis Aurelia Alexandria; in the priesthood of Marcus Aurelius Antiochus, eques Romanus, son of Belsu, and in the second term as strategos of Marcus Aurelius Abgar, eques Romanus, son of Ma nu, grandson of Agga, and of Abgar, son of H . afsai, grandson of Bar-KMR; on the ninth day of the aforesaid month’ (Bellinger and Welles 1935: 96; Goldstein 1966: 2, 6; Brock 1991: 264 n. 32). Brock 1991: 261 n. 12, with the reading ‘in Edess(a), the great city, the mother of all the cities of Mesopotamia.’ Bellinger and Welles 1935: 146; Duval 1892: 224–5; Millar 1993: 151. Gutschmid (1887: 44) says only that Gordian, on his arrival in the Near East, ‘einen Sprossen des alten Königsstammes wieder in Osroëne als König einsetzte’ – a conclusion which cannot stand, as we have seen. BMC Arabia, etc. 112 nos. 133–5 (Edessa); 121 nos. 14–16 (Nisibis); 135–6 nos. 7–15 (Singhara). Bellinger (1935: 147) makes this point clearly enough, but it has not dissuaded more recent scholars from placing Abgar’s coinage after that of the colonia under Gordian: for example Dodgeon and Lieu 1991: 355 n. 6. Prof. Fergus Millar (private communication). In addition to the adventvs avg aureus, RIC 4.3, 34, no. 176, reverse type concordia avg (Concordia standing l., holding patera over altar in r. hand and cornucopiae in l. hand), is thought to be from Antioch. It is described in the old catalogue of the Paris collection, no. 516, but was apparently stolen in 1831. Bland (1991: 92) shows that a third aureus sometimes attributed to Antioch (RIC 4.3, 38 no. 226A, obverse imp caes m ant gordianvs avg, reverse pax avgvsti, Pax advancing left holding cornucopiae and sceptre) is incorrectly described, and is actually from the mint of Rome. RIC 4.3, 33 no. 174: reverse p m tr p ii cos p p, emperor riding l. on horseback with transverse sceptre. Bland (1991: catalogue no. 34) lists eight specimens. Weder 1982: 53–71. At the same time, Bland rightly rejects Weder’s attempt to impose a drastic rewriting of the overall chronology of Gordian’s reign. Sync. P361, p. 681; Zonaras 12.18 (Gordian’s recapture of Nisibis and Carrhae); Oates 1955: 39–43; Maricq 1965: 17–25. On the date of Hatra’s fall: Loriot 1975: 761–2; Kettenhoffen 1983: 152; Henrichs and Koenen 1970: 125–32. HA Gord. 26.3 is quite clear on the fact that Gordian did not open the gates of the Temple of Janus until 242. In this case, the information probably derives from an annalistic compiler, perhaps Dexippus, and should be believed. The reported revolt of one Sabinianus in 240 (HA Gord. 23.4) may have been one such event, but since the HA states that this was completely put down by the governor of Mauretania, we cannot assume that it was this crisis which required Gordian’s presence so urgently. Alternatively, the urgency may have related to the pressure on the empire’s northern

174

NOTES

28 29

30

31 32 33

34

1 2

3 4

5 6

borders; as we have seen, the campaign against the Carpi and Goths in Moesia the next year delayed the renewal of the eastern campaign. Potter 1990: 389–90; Seyrig 1963: 160–1; Gawlikowski 1985: 254–5; cf. Ross 1993: 199–200. Teixidor (1989: 220–2), in his ﬁrst rendition, translated the phrase MYQR BHPTY B RHY ‘honoré comme consularis à Édesse,’ seeing Abgar as a regional governor of some sort. In the second and more complete edition, however (1990: 161–2), he opted to translate it as ‘honoré du consulat à Orhai,’ and accepted the idea of ornamenta consularia. Besides the supposed grant of honors to Odenathus, Dio 60.8.2–3 records Claudius’s grant of consular ornamenta to the Jewish King Agrippa I, and praetorian honors to his brother Herod, in 41; Syncellus 717.20 records a similar award to the Herulian leader Naulobatos in 268. Istvan Borzsák, RE 18:1 s.v. Ornamenta, cols. 1110–22. Welles 1959: no. 97 – Laetianus, vir egregius procurator, in charge of approving men and mounts in April and May 251. Feissel and Gascou (1995: 105) retract their earlier tentative suggestion of a higher imperium for Laetianus, and view him as the governor’s successor. Priscus, it should be noted, became prefect of Mesopotamia even though one of his brother’s ﬁrst acts as emperor was to surrender that province to Shapur I: Res Gestae divi Saporis l.4; HA Gord. 26.3–30.9; Zosimus 1.18–19; Zonaras 12.18–19. The comments of Ernest Will on Abgar’s titles (Feissel and Gascou 1989: 166), to the effect that local ‘dynasts’ could hold two titles – one for their own subjects and one to correspond to their position in the Roman system – appear to have some justiﬁcation, although the comparison with Odenathus, who was certainly a senator, may not be apt. The story of Odenathus’s conquests and his family’s subsequent bid for power is told most fully in HA Val. 4.2–4, Gall. 1.1, 3.1–5, 10.1–8, 13.1–5; Trig. Tyr. 15–18, 27–8, 30; Aurel. 22, 25–6, 28, 30–3, 38.1. 5 A ‘GOLDEN AGE’? Linguistic and artistic aspects of the question have been studied by Geo Widengren (Widengren 1960: 6–24). Cf. the statement of H. J. W. Drijvers (1992: 139): ‘Edessa . . . was as Hellenized as the rest of Syria and the Roman Empire in general,’ an exaggeration probably inspired by Edessa’s later importance as a center of the translation and transmission of Greek classics. First published Cureton 1855, English translation 41–51, 70–6. Cureton 1855: 40, 44. The variant that Kutbai was Arab ( RBYT ) rather than Hebrew ( BRYT ) is represented by the Liber Scholiorum of Theodore Bar Koni: CSCO 66, vol. 2 p. 287. Bakru is the name of the fourth king of Edessa according to the Chronicle of Zuqnin. Segal (1970: 43) takes this as an indication that the people of Edessa associated the art of writing with the Jews. Howard 1981; Martin 1875: 107–47; Julian, 150 C–D; Bidez and Rochefort 1924–63: 2.2.128. The oration of Julian is commonly taken

175

ROMAN EDESSA

7

8 9

10 11 12 13 14

15

16 17 18

19 20 21

as referring to Emesa, not Edessa, but the reasons for correcting the text in this manner are not conclusive. On the possible polemical purpose of the Doctrine of Addai, see Chapter 6. Jacob’s homily, as a Christian text on pagan religion, was by nature hostile, while the anti-Christian intentions of Julian the Apostate need no demonstration. Diod. Sic. 2.4 (Ascalon); Bruneau 1970: 467; cf. Goossens 1943: 62; Drijvers 1980: 79. The multiple attributes described by Lucian – sceptre, spindle, rays, a tower, and a girdle – may well have their origin in Atargatis’s composite nature, being a compound of the Semitic goddesses Asˇtarte, Anat and possibly Asˇerah (Oden 1977: 58–107). For Derketo = Atargatis cf. Strabo 16.4.27; Pliny HN 5.19; Oden 1977: 69–70. BMC Arabia, etc. 97–101, nos. 39–68; pl. 14, nos. 11–15, 17–18; pl. 15, nos. 1–3. The identiﬁcation is suggested by Drijvers 1980: 84–5. Drijvers 1980: pl. XXVI, with text p. 123. Drijvers believes that a rock portrait near the summit depicts the god himself, but this seems unlikely. Drijvers (1980: 137) speculates that the worship of Sin may have been a special enthusiasm of King Wael, based on the Sumatar evidence and the coins of Wael showing the temple with baetyl inside. BMC Arabia, etc. 101, no. 68 (Elagabalus): (rev.) two city Tychai (H . arran and Edessa?) facing each other, small temple beneath; 103 no. 79 (Elagabalus and Severus Alexander): (obv.) two emperors facing, temple beneath. On the temple as Edessa’s mint symbol see Bellinger 1940: 51. Bel occurs, for instance, in the names BLBN , BLY, BLSW, and ZLBL (Drijvers 1972: nos. 24, 25, 46, 29); also in the name ofˇ the legendary pagan priest Sharbel who supposedly converted to Christianity at a very early date – ‘Acts of Sharbel’ in Cureton 1864: 41–72. Nebo: BRNBS (Drijvers 1972: no. 46), and the priest Abednebo in the Doctrine of Addai (Howard 68–9). Segal 1970: pl. 2 (the ‘Funerary Couch Mosaic’); pl. 25 a, b (tomb reliefs). Seyrig 1951: 32–9. Segal 1970: pll. 43, 44. A similar artistic style is seen in the ‘Animal Mosaic,’ ibid. pll. 17b–20. Prof. Segal has very kindly pointed out to me that there are many examples in later Syriac literature of the phrase ‘evil latter end’ (private communication). Alongside the words about the ‘goodly’ end, this weakens the argument of the main text against deﬁnite expectations of the afterlife. It should also be noted that the latest treatment of this inscription comes down in favor of interpreting it as promising ‘a happy afterlife’ (the main difﬁculty is in the translation of the phrase h.ryt , literally, ‘end’ or ‘posterity’): Drijvers and Healey 1999: 173–5. The original text has ‘western wall’: cf. Duval 1892: 216 n. 1. Contrast the tale of the evangelization of Edessa, supposedly taken from those archives as well: Chapter 6. Drijvers (1980: 14): ‘The story of his [Abgar VIII’s] conversion to Christianity . . . should be considered apocryphal.’

176

NOTES

22 Procopius, in his account of the repair work of Justinian, also mentions the Church of the Christians (A . . . των Χριστιανων "κκλησ!α) as having been damaged and then repaired. This is probably the same structure, or at least the same location, as that mentioned in the Chronicle of Edessa. 23 Duval (1892: 218) recognizes this point and takes it as evidence that Abgar’s supposed conversion to Christianity took place after a visit to Rome. 24 ‘(The city’s bishop) duxit me primum ad palatium Aggari regis et ibi ostendit michi archiotepam ipsius . . . Item perintrauimus in interiore palatii, et ibi erant fontes piscibus plene, qualis ego adhuc nunquam uidi,’ Itinerarium Egeriae (Arce 1980) 19.6–7. 25 The governor’s seat of judgment was also called by the name δικαστριον (D YQSTRYWN), Shmona and Guria 24. 26 For the city’s gates and many of its features as described in late sources see Duval 1892: 100, nn. 6–10. 27 Humphrey 1986: 455–504: second- or third-century dates proposed for monumental hippodromes at Antioch, Caesarea Maritima, Beirut, Bostra, Gerasa etc. There is virtually no evidence for the date of construction of Edessa’s hippodrome: ibid. 533–4. 28 Drijvers and Healey (1999: 187) defend the identiﬁcation of this Abgar as Abgar the Great, with the suggestion that the portrait may depict him in the period before he took the throne in 176. This, however, is to displace the mosaic too far in time from other known similar works, especially if we adopt the approach advocated by Colledge (1994: 196–7), which would place all the mosaics within the last generation of the Edessan monarchy. 29 Rostovtzeff 1935: 293–9; Schlumberger 1960: 133–5; Balty 1981: 387–90, n. 266; Drijvers 1982a: 168, 177–8; Perkins 1973: 121–4. 30 Drijvers (1982a: 184–7) gives some examples. 31 Levi 1947: e.g. 158, ﬁg. 60; 205, 219, ﬁg. 82. Note that Levi reproduces (p. 248) a third-century Edessan mosaic with similar geometric borders, using it to illustrate Eastern styles of headwear. 32 Four known Edessan mosaics bear dates: The ‘Four-pointed Star’ mosaic (224), the Orpheus mosaic (227/8), the Phoenix mosaic (235/6), and the Funerary Couch mosaic (probably 277/8). 33 BMC Arabia, etc. 91, no. 2; 92, nos. 3–4. Cf. Duval 1892: 210–12. Edessan countermarks on early Roman coins are also in Syriac characters – Howgego 1985: nos. 26, 504, 695, 696. 34 The change to Greek may not have come immediately; whether the Syriac-legend coins of Ma nu fall before or after the break in Ma nu’s reign is an open question: see Duval 1892: 210–12; Gutschmid 1887: 30; BMC Arabia, etc. xciv–c. 35 Drijvers 1992: 139. Cf., however, Drijvers 1977: 894 – ‘Die Kultur von Edessa soll aber charakterisiert werden als im wesentlichen mesopotamisch–arabisch, mit daneben griechischen und parthischen Einﬂüssen.’

177

ROMAN EDESSA

1

2

3

4 5

6

7

6 EARLY CHRISTIANITY AND EDESSAN CULTURE The hagiographical Life of Abercius tells the same story in much more detail, including among other things an account of the missionary’s encounter with Bardais.an. This has been taken as valid evidence for Edessan Christianity by some, but the authenticity of the Life is very questionable; it may have been based on the text of the inscription and the published works of Bardais.an himself (Bundy 1990, 1992; cf. Drijvers 1967: 170–1). Other notable early Syriac texts include the Diatessaron of Tatian, almost entirely lost, and the ‘Hymn of the Soul’ embedded in the apocryphal Acts of Thomas; of lesser literary value but certainly of early date is the lengthy extract from the Chronicle of Edessa. The other two texts can possibly be dated as early as the second century ce. The Chronicon Paschale, p. 484, gives a date of 169 for Meliton; Cureton, assuming the emperor addressed is M. Aurelius, guesses as early as 160. Apparent references in the Letter of Mara Bar Serapion to a Roman attack on Samosata give possible dates between 72 (Vespasian) and 165 (L. Verus). Cf. Ortiz de Urbina (1965: 41), who assigns (pseudo-) Meliton to the reign of Caracalla, 211–17. The dates of the two texts, however, are not secure. A case has recently been made that Mara Bar Serapion’s letter is a school exercise of the fourth century or later (McVey 1990), and similar reasoning could be applied to ‘pseudoMeliton.’ Both of these texts, then, might well postdate both the Edessan ‘renaissance’ under Abgar VIII, and the emergence of ofﬁcial Christianity as an empire-wide cultural force. Drijvers 1965: text (even-numbered pages) with English translation (odd-numbered pages). Cited as BLC. On the philosophy of Bardais.an see also Drijvers 1966; Drijvers 1975. Ephr. Syr. Hymnen Contra Haereses (ed. Edmund Beck, CSCO 169–70, Scr. Syr. 76–7); Prose Refutations (Mitchell 1912–21). For a convenient guide to the Ephremic corpus: Kronholm 1978: 15–34; for polemical works, 28–33. Nearly all the anti-heretical writers have something to say about Bardais.an, although like Hippolytus in his Philosophoumena (who calls him Ardesianes), they often take him incorrectly for a Valentinian. In general, the early Greek authors are more reliable sources for the Edessan’s work than most of the Syriac literature with the exception of Ephrem. Works written in Arabic after the Islamic Conquest, most notably the Fihrist of Al-Nadim, also treat with Bardais.an as an exponent of Hellenistic philosophy; these, however, are not dealt with in the current chapter (see Drijvers 1966: 200–7). Miscellaneous sources such as the courtier Sextus Julius Africanus, an eyewitness, and the Church historian Sozomen, give tidbits of information about the philosopher’s life; the Armenian historian Moses of Chorene claims that the Edessan philosopher ﬂed to Armenia after 216 and stayed there to write a history of his new home. BLC 40–61. It is this section, which was known to Eusebius and other later writers, that earned the title Book of the Laws of Countries for the

178

NOTES

8 9 10

11 12 13 14 15 16

17

18 19 20

work. F. Boll has shown that the ideas expressed here originate in the second-century bce work of Carneades: Drijvers 1966: 18–20. Admittedly, Eusebius’s testimony on this subject cannot be conﬁdently called independent, since he was familiar with the dialogue and may have based his description of Bardais.an on it. Teixidor (1992: 49): ‘L’église de 201 aurait été celle des adeptes de Marcion et de Bardesane . . .’. Epiphanius (Panarion 56) states explicitly that Bardais.an was bilingual, but against this we have such statements as that of Eusebius indicating that he always wrote in Syriac. Bardais.an’s son in Athens: Theodoret, Haeret. Fabul. Comp. 1.22 (Migne 83, col. 372). Sozomen Hist. Eccl. 3.16 (with the claim that it was Bardais.an’s son ‘Harmonius’ who wrote the hymns); Ephrem Contr. Haer. 55. Cf. Brock 1985: 77–81. Eusebius (Praep. Evang. 6.9.32) conﬁrms Bardais.an’s astrological learning even while borrowing his arguments against the astrologers. BLC 42–3. It is noteworthy in this connection that the name of the Apostle Thomas, known as the ‘Apostle to India,’ is connected with Edessa, and that his remains were believed to have been interred there. Julius Africanus Kestoi 29, p. 300 (Thévenot). This picture is conﬁrmed by Epiphanius, Panarion 56, who says the king and the philosopher were actually brought up together. Theodore bar Koni, Liber Scholiorum, ed. A. Scher, CSCO 55, 69, Scr. Syr. 19, 26; cf. the Chronicle of Michael the Syrian (Chabot 1963: 109–10). Eus. Caes. Hist. Eccl. 4.30. Ephrem Syrus himself was credited with establishing at Edessa a school, possibly the ‘School of the Persians,’ which became proliﬁc in the production of Greek works in Syriac; Barhadbshabba Arbaya, Cause de la Fondation des Écoles (Scher 1908: 381–6). Cf. Vööbus, CSCO Subsidia 26, 7–47. Bardais.an is identiﬁed by some Christian writers, including Eusebius, as a Valentinian or a former Valentinian; and some modern students of patristics include Bardais.an, along with Marcion, Tatian and others, among the Gnostic heretics. Hippolytus Philosophoumena (ed. P. Wendland) 26.6.35; Epiphanius Panarion 56; Eus. Caes. Hist. Eccl. 4.30. Cf. also Quasten 1950: vol. 1, 263–4; for the ‘Christian Gnostics’ as a whole, 256–77. Drijvers (1966: 223–4) reviews points of similarity and dissimilarity between Bardais.an’s philosophy and Gnostic systems, and concludes that Bardais.an cannot, technically, be called a Gnostic. Other writers unequivocally speak of Bardais.an’s ‘Gnosis’; cf. Haase 1910; Hilgenfeld 1864. More recently, A. Dietrich, Synkretismus im syrisch– persischen Kulturgebiet (Abh. Ak. der Wissenschaften z. Göttingen 96) 123–43 argues for Bardesanian Gnosis. Ephrem himself memorializes the fall of Nisibis and connects its loss with the paganism of the Emperor Julian in his hymns Contra Iulianum, CSCO 174–5, Scr. Syr. 78–9. Bishop Quna: Chronicle of Edessa in Chron. Min. 4/5. Burkitt 1913: text 1–23, trans. 90–110. Among the strongest arguments in favor of this being a genuine account is the fact that the martyrs seem truly to suffer, and do not continue to spout lengthy

179

ROMAN EDESSA

21

22

23

24

25

26

27 28

29 30

professions of faith when subjected to torture; ibid. 19–23. The date of the events depicted is apparently 309.

ο> δ4 τ η Αρειανικη "κκλησ!α 2π3 του πλοτου τρυφωντε "πεχε!ρησαν τοι απ3 του Οαλεντ!νου κα τετολµκασι τοιαυτα κατα` τ)ν Εδεσσαν, ο>α οδ-ποτε "ν ενοµουµ-νC π λει γ-νοιτ αν (‘The members of the Arian church, reveling

in their wealth, laid hands on the followers of Valentinus, and have dared to commit outrages of a sort that could never have happened in a well-governed city’), Julian Ep. 40. All of the early anti-heretical theologians offer material for the study of Gnosis; prominent among them are Irenaeus of Lyons (especially useful for his description of Valentinianism), Tertullian, Clement of Alexandria and Epiphanius of Salamis. K. Rudolph (1983) remains a good introduction to Gnostic philosophy as a whole; good recent treatments of Gnostic mythology are to be found in Filoramo (1990) and Stroumsa (1984). For the Nag Hammadi cache of Gnostic literature in Coptic: Robinson 1990. This is not to say that there was no ethical element to Gnosis; most of the prominent sects were rigorously ascetic. Although the terms ‘Gnosis’ and ‘Gnosticism’ are more or less interchangeable, I will use the former when referring to the philosophical principles involved, and the latter to denote Gnosticism as a social and religious phenomenon. Irenaeus, Adv. Haer. 1.1, 11; Clem. Alex. Ex Theodoto 33.3–4. The outlines of Gnostic cosmogony as delineated by the heresiologists are conﬁrmed in great, almost bewildering, detail and variety by the Nag Hammadi documents. This is apparent from the fact that the so-called ‘Sethian’ strand of Gnosis, for example, even while rebelling against the traditional order, betrays its grounding in biblical traditions by reference to Genesis and other Biblical texts; see, for example, such Gnostic works as the Paraphrase of Shem, the Three Steles of Seth, Zostrianos and Allogenes: NHC VII.1, VII.5, VIII.1, XI.3. The beginning of the Gospel of John, with its conception of Christ as the pre-existent Logos (John 1:1–5, 9–12), is the clearest example of Gnostic thinking in the New Testament; other examples include the emphasis on knowledge vs. ignorance (John 8:32) and Gnosticizing passages in the epistles, Eph. 5:14 and Phil. 2:6–8. The epistles attributed to John have also been called the work of a ‘Christian Gnostic.’ Earlier scholars believed that Bardais.an himself might have been the author of the Hymn of the Soul, but this is no longer held possible. Cf. Burkitt 1904. Contra Haer. 3, 14, 41, 53. Ephrem’s Fifth Discourse to Hypatius (Prose Ref. I, xcii–cxix [English], 124–85 [Syriac]) seizes upon @λη as ‘a sign set upon all of them’ (i.e. Marcion, Bardais.an and Mani) to show the wickedness of their teachings. On the importance of Bardesanite thinking and Syrian–Mesopotamian Gnosis generally as the background to Mani, see Widengren 1985. This aspect of Bardais.an’s teaching is mentioned frequently, in works including letters nos. 114, 126, 145 and 151 of Theodoret (Migne 83). Clem. Alex. Stromata 7.104.2 stresses perfect knowledge of truth as the

180

NOTES

31 32 33 34 35

36 37

38 39

40 41 42

43

true ‘Gnosis’; in other ways too he often adopts the terminology of the foes against whom he writes. Doubts on the report’s value: Vööbus 1958: 3–4. Epiphanius, Panarion 40.1.4, 45.2.1; Irenaeus Adv. Haer. 1.24.2. On the generally ascetic ethic of most Gnostic sects, and almost all the early ones, see Rudolph (1983: 257–63). Ortiz de Urbina 1965: 35–7. ‘Land of the Assyrians’ ( των Ασσυρ!ων γη): Tatian Oratio ad Graecos 42. Jerome In Amos 2, 12 6.247; In Ep. ad Tit. pref. 7.686; In Ep. ad Gal. 6, 8 7.526; Whittaker 1992: 82–3. Shmona and Guria (Burkitt 1913: 70) – the earliest appearance of the Bnat Qyama; Aphrahat, Demonstrations 6. On the early ‘Covenant’ see the works of Murray (1975: 11–18; 1982: 11–16) and Sebastian Brock (1992), proposing that this early phase of Syrian asceticism be called ‘proto-monasticism.’ See Murray 1975, with, however, some words of caution against pushing the evidence too far. Also Gibson 1963–5: 24. It is not quite fair to say that ‘The idea . . . of a substantial role of the Jews in the Christianization of Edessa is ill-founded, since it is based only on a particular interpretation of the Doctrina Addai’ (Drijvers 1992: 128). The Spanish pilgrim Egeria, on her visit to the city in about 380, remarked that the version current there was longer than the one with which she was familiar: Itinerarium Egeriae 19.19. For the role of these icons in Edessa’s later doctrinal feuds, including the promotion of rival ‘genuine’ portraits, see Segal 1970: 77–8. On the image itself, its religious signiﬁcance and development of the myth of the mandylion: Cameron 1983. An exception is Gutschmid, who believed that ‘Abgar, the Great King’ in the Chronicle of Edessa’s ﬂood narrative was Abgar V. Contra Duval 1892: 202. On this see Burkitt 1904: 26–7; Duval 1892: 218, concluding that Abgar converted after his ceremonial visit to Rome. Self-emasculation among the galloi of the Atargatis cult is independently attested by Lucian, Dea Syr. 51–3. Despite the supposed abolition of this practice, it apparently persisted into the ﬁfth century, to judge by the legislation against it enacted by Bishop Rabbula of that era: Canon 55 of Rabbula, ed. Vööbus 1960: 49. Quoted by George Syncellus (Teubner ed.) 439. A similar description of Abgar is to be found in Epiphanius Panarion 56: =σι,τατο κα λογι,τατο.

APPENDIX 1 BMC Arabia, etc. 113–14, nos. 136–43. 2 BMC Arabia, etc. 113. Duval (1892: 224) speaks of ‘l’investiture du nouveau roi par l’empereur,’ while Hill (BMC Arabia, etc. cv) more accurately describes the entire series only as ‘commemorating the restoration of the kingdom.’ 3 BMC Arabia, etc. 114, no. 140. The catalogue describes Gordian as

181

ROMAN EDESSA

4 5 6

7

8

9

10 11 12

‘holding globe in l., mappa in r.,’ while Gesche (1969: 71) describes him merely as ‘Globus und einen weiteren Gegenstand in der Hand haltend.’ It is, in fact, difﬁcult to ascertain the nature of the object held in Gordian’s right hand, although on the basis of at least one specimen a consular mappa or rolled ‘napkin’ is feasible (Gesche’s pl. 3, no. 5 — apparently the same specimen as in Bellinger 1939: pl. 3, no. 55). BMC Arabia, etc. 113–14. The total of nine specimens includes a coin illustrated by Babelon 1893: 8, no. 6. Bellinger and Welles (1935: 146): ‘The joint coinage of Abgar and Gordian alone supplied the Mesopotamian cities with bronze in the years just before 242.’ Weight and size ranges are based on measurement of specimens in the American Numismatic Society collection and on published ﬁgures in the British Museum Catalogue. The fourth denomination is represented by only one specimen from each source, and it may be doubted that this truly represents a separate denomination. Although the scale of the ﬁrst-denomination coins is too small to allow for much detail, the portrait bust of Abgar on the smaller units shows that he is also wearing a necklace or beaded collar of some sort. His garment can have one, two, or three rows of buttons as well: BMC Arabia, etc. 114–16; cf. specimens in American Numismatic Society collection. Sylloge Nummorum Graecorum Copenhagen, Phoenicia etc., pl. 7, nos. 215–17, 220–1 (Vologaeses II); 240–1 (Vologaeses III); 246 (Vologaeses V); 247 (Artabanus V). Similar to the tiara of Abgar X in shape and decoration, but chronologically distant, are the diademed tiaras on the coins of Phraates III, 70–57 bce: ibid. pl. 3, nos. 70–6. BMC Arabia, etc. pl. 16, nos. 10, 11. This feature is easy to miss because the lines of the lappet, on even a slightly worn coin, can easily merge into the broad band of the tiara; cf. ibid. pl. XVI.9. For a lappet of similar size on a Parthian coin, SNG Copenhagen, Phoenicia etc., pl. 7, nos. 215–17. Cf. Gesche 1969: 61: ‘Bei genauerem Zusehen wird jedoch deutlich, daß diese Bezeichnung nicht trifft, denn die römische sella curulis sieht grundsätzlich anders aus; in der Regel ist sie als Klappstuhl dargestellt.’ Richter 1926; Wanscher 1980; Baker 1966; Malachowski 1919: 22–3. Tripods, common in antiquity and sometimes used as seats, have been excluded from this survey. Salonen (1963: pls. 4, 33.1) illustrates two pieces which have legs displaying a broken curve something like that of Gordian’s seat, although in each case the convex curve faces away from the middle and not toward it as on the Edessan coin. It is just conceivable that the die engraver (or the designer of the actual seat used by Gordian) was inﬂuenced by this tradition. In the classical tradition, the only furniture legs displaying such a broken curve appear to be bronze or marble animalleg tables and chairs, such as those found at Herculaneum (Richter 1926: ﬁgs. 322–4 [facing p. 133]; Budetta and Pagano 1988: nos. 3, 4, 6). These, however, end in animal feet, which are not observed on the legs of Gordian’s seat. These latter legs, moreover, vary in the amount of

182

NOTES

13

14

15

16 17 18 19

20

21 22 23

curvature displayed in the lower section, so that the similarity with the animal-leg furniture is distant, at best. Among the many ancient authorities attesting to the signiﬁcance of the curule stool, the following are sufﬁcient to establish the point: Livy 1.8.3, 29.37.2, 30.15.11; Dion. Hal. 3.61; Diod. 29.32; and, for the stool as the special attribute of the emperor, Tac. Hist. 2.59, Ann. 1.75; Suet. Nero 13.1, Galba 18.3; Pliny Panegyricus 59.2. Kübler 1923: cols. 1310–15. BMC Arabia, etc. 116, no. 158. Babelon (1893: 287–8) in fact assumes that this is an eagle-topped sceptre, while other numismatists (Bayer, Gutschmid) describe the object as a hasta or spear. Gesche 1969: 47, nn. 4, 5, 6. While it is difﬁcult to ﬁnd specimens in a good enough state of preservation to make this determination, in photos from the British Museum collection, two of the best-preserved specimens clearly show what appears to be a long lance rather than an arrow in Gordian’s hand: BM catalogue nos. GC28 P113.137, GC28 P113.138. I owe thanks to the Museum for copies of these photos, which I received with compliments from John Ornstein on behalf of Dr Ute Wartenburg. Gesche 1969: 62–3, n. 84; E. Babelon, Les Rois de Syrie pl. 1, no. 8; pl. 1, no. 4; pl. 12, nos. 8–12; BMC Parthia pl. 14, no. 2; pl. 22, no. 8. Shore 1993: 137–8, nos. 315–25. Gesche 1969: 63, n. 86. Examples of such scenes: RIC 5.1 276, no. 56; 279, nos. 126 and 127; 285, no. 184; 311, no. 407; RIC 5.2 49, no. 310. RIC 4.2 152, no. 120; cf. BMC VI, pl. 39, no. 183; pl. 110, no. 1; pl. 112, nos. 9–10; RIC 5.2 167, no. 225; RIC 5.2, pl. 7, no. 17; pl. 8, no. 17. Gesche 1969: 63, n. 87: Severus, Caracalla and Geta. As nonnumismatic examples of this type of scene, Gesche cites the one on the ‘Sword of Tiberius,’ interpreted as the acknowledgement of the emperor’s ultimate responsibility for victory by Germanicus, and a similar scene on a silver cup from Boscoreale. Brilliant 1963: 62, ﬁg. 2.32; 75–6, ﬁg. 2.64; Gesche 1969: 64–5, n. 95 (‘die Anerkennung der obersten Sieghaftigkeit des Kaisers’). Here we are very far from the art-historical context of Abgar’s coin; moreover, the Boscoreale cup shows, not a subordinate commander, but the goddess Venus presenting Victory to Augustus. Cf. Kuttner 1995: 25–6, 28–31. Maximinus Thrax ruled jointly with his son Maximus as Caesar from 235–8; Philip I named his son Philip II Caesar in 244, and they ruled as joint Augusti from 247 to 249; Valerian ruled jointly with his son Gallienus from 253 until the father’s capture by Shapur I in 260; Carinus and Numerianus served as Caesars under their father Carus from 282 until his death in 283. Tac. Ann. 2.41; Aur. Victor 16.4; Paneg. 2.11.4; Gagé 1933: 1–43; Gagé 1930: 1–35. Gesche 1969: 64, n. 91: ‘Einheit und Unteilbarkeit’ of the Victoria Aug. Gesche (1969: 70): ‘Abgar den römischen Kaiser als Oberherrn anerkennt und akzeptiert.’ Gesche cites and illustrates the well-known Persian ‘kneeling archer’

183

ROMAN EDESSA

24 25

26

27

28

29 30

coinage (p. 48, pl. 3 nos. 7–12); Parthian reverse types showing the enthroned Arsaces with bow (p. 49, pl. 4 nos. 1–2), as well as a series of reverses showing bow, arrow or quiver by themselves; an Assyrian investiture relief on which the king holds a bow and several arrows (p. 55, pl. 4 no. 11); a silver bowl portraying a Sassanid king holding bow and arrow (pl. 4 no. 12). For the arrow or bow as an attribute of Apollo, Mithra or the sun-god, Seleucid and Bactrian coins showing Apollo, seated or standing, holding an arrow and sometimes a bow (pp. 49–50, pl. 4 nos. 3–9); a Mithraic relief showing the birth of the god with various attributes including bow and arrow (p. 57, pl. 5 no. 1); a relief of Tiglath-Pileser on which a hand holding a bow emerges from a sun-disk (p. 54, pl. 5 no. 2). This point deserves more argumentation than Gesche’s assertion, p. 48, that bow and arrow ‘naturgemäß eine Einheit bilden.’ Trajan: BMC III, nos. 103, 215, pl. 40.8.; no. 588, pl. 19.19; nos. 1045–9, pl. 43.1; nos. 1043, 1044, pl. 42.10; Lucius Verus: BMC IV, cxxxv, nos. 1099–1106, pls. 75.8, 76.1; nos 1125, 1126, pl. 76.6A. Brilliant 1963: 108–9; ﬁgs. 3.10, 3.11, 3.12, 3.122. For examples that are chronologically relatively close to the Edessan coin see Shore (1993: 159–60, nos. 442–7, 450: Vologaeses V (191– 208). A similar type is found as early as Orodes II, 57–38 bce: 121 nos. 210–11. Downey 1961: 73–5. On the statue, the best copy of which is thought to be a marble statuette now in the Vatican (Galleria dei Candelabri IV.49), see Stewart (1990: 201–2); Toynbee (1934: 131–3); Richter (1941: 295). Bosch (1935: 254–6) lists cities in Cilicia, Lycaonia, Pamphylia, Pisidia, Caria, Phrygia, Bithynia, Galatia and Cappadocia as well as in Syria, Antioch, Samosata, Ptolemais, and Syria Provincia under Trajan. Interestingly, Antioch does not seem to have adopted this type until the reign of Tigranes, 83–69 bce: BMC Seleucid Kings of Syria 103, nos. 1–8, 104–5, nos. 9, 13–17; pl. 27, nos. 5, 6, 8, 10. Macdonald 1905: 303 n.; BMC Arabia, etc. xciii–xciv (the ﬁgure on coins of Carrhae), cvi (Edessa). In addition to the bronze pieces of various denominations listed in this table, eight silver tetradrachms from Edessa’s mint were also found: two from the reign of Caracalla and six of Diadumenian. These date from the period of Roman involvement on the frontier under Caracalla, when local mints became, in effect, ateliers of the mint of Rome. For this episode see Bellinger 1940.

184

BIBLIOGRAPHY

Allison, F. (1886) ‘Pseudo-Ionism in the Second Century A.D.,’ AJP 7: 203–17. Arce, A. (1980) Itinerario de la virgen Egeria, Biblioteca de Autores Cristianos, Madrid: Editorial Catolica. Astour, M. C. (1989) review, Journal of the American Oriental Society 109, 4: 686–8. Babelon, E. (1893) ‘Numismatique d’Éadesse en Mésopotamie,’ Mélanges Numismatiques 2, 209–96, Paris: C. Rollin and Feuardent. Baker, H. (1966) Furniture in the Ancient World: Origins and Evolution, London: Jarrold and Sons. Baldus, H. R. (1971) Uranius Antoninus: Münzprägung und Geschichte, Antiquitas 3: 11, Bonn: Rudolf Habelt. Balty, J. (1981) ‘La mosaïque antique au Proche-Orient, I: Des origines à la Tétrarchie,’ Aufstieg und Niedergang der römische Welt 2.12.2: 347–429. Bauer, P. V. C. (ed.) (1933) The Excavations at Dura-Europos: Preliminary Report of the Fourth Season of Work, Oct. 1930–March 1931, New Haven: Yale University Press. Bauer, W. (1934) Rechtglaubigkeit und Ketzerei im ältesten Christentum, Beiträge zur historischen Theologie 10. Beck, E. (1957) ‘Des Heiligen Ephraem des Syrers Hymnen Contra Haereses,’ in Corpus Scriptorum Christianorum Orientalium 169, Scr. Syr. 76. Bellinger, A. R. (1939) The Eighth and Ninth Dura Hoards, New York: American Numismatic Society. —— (1940) The Syrian Tetradrachms of Caracalla and Macrinus, Numismatic Studies 3, New York: American Numismatic Society. —— (1949) The Coins: Excavations at Dura-Europos, Final Report, v. 6, New Haven: Yale University Press. Bellinger, A. R. and Berlincourt, M. A. (1962) Victory as a Coin Type, New York: American Numismatic Society. Bellinger, A. R. and Welles, C. B. (1935) ‘A Third-century Contract of Sale from Edessa in Osrhoene,’ Yale Classical Studies 5: 95–155.

185

ROMAN EDESSA

Bidez, J. and Rochefort, G. (eds) (1924–63) L’Empereur Julien: oeuvres complètes, Paris: C. Lacombrade. Birley, A. (1987) Marcus Aurelius: A Biography, 2nd ed., New Haven: Yale University Press. —— (1988) Septimius Severus: The African Emperor, rev. ed., New Haven: Yale University Press. Bland, R. (1991) The Mints of Antioch and Caesarea under Gordian III (unpublished dissertation), London: University College London. Bosch, C. (1935) Die kleinasiatischen Münzen der römischen Kaiserzeit II.1: Bithynia, Stuttgart: W. Kohlhammer. Bowersock, G. W. (1983) Roman Arabia, Cambridge, MA: Harvard University Press. Breasted, J. H. (1922) ‘Peintures d’époque romaine dans le désert de Syrie,’ Syria 3: 177–213. Brilliant, R. (1963) Gesture and Rank in Roman Art, Copenhagen: Bianco Luno. Brock, S. (1991) ‘Some New Syriac Documents from the Third Century AD,’ Aram Periodical 3: 259–67. —— (1992a) ‘Syriac Historical Writing: A Survey of the Main Sources,’ in Studies in Syriac Christianity 2, London: Variorum Reprints. —— (1992b) ‘Eusebius and Syriac Christianity,’ in Attridge, H. and Hata, G. (eds), Eusebius, Christianity, and Judaism, Detroit: Wayne State University Press. —— (1995) ‘A Palestinian Targum Feature in Syriac,’ Journal of Jewish Studies 46: 271–82. Brown, P. (1988) The Body and Society: Men, Women, and Sexual Renunciation in Early Christianity, New York: Columbia University Press. Bruneau, Ph. (1970) Recherches sur les cultes de Délos à l’époque hellénistique et à l’époque impériale, Paris: E. de Boccard. Budetta, T. and Pagano, M. (1988) Ercolano: legni e piccoli bronzi: Testimonianze dell’arredo e della suppellettili della Casa Comana, Roma: Soprintendenza Archeologica di Pompeii. Bundy, D. (1990) ‘The Life of Abercius: Its Signiﬁcance for Early Syriac Christianity,’ The Second Century 7.3: 163–76. —— (1992) ‘Christianity in Syria,’ in Freedman, D. N. (ed.), Anchor Bible Dictionary, New York: Doubleday. Burkitt, F. C. (1904) Early Eastern Christianity: St. Margaret’s Lectures, 1904, on the Syriac-speaking Church, London: J. Murray. —— (1913) Euphemia and the Goth, with the Acts of Martyrdom of the Confessors of Edessa, London and Oxford: Williams and Norgate. —— (1924) Appendix to Note by C. Winkworth, Journal of Theological Studies 25: 403. Buttrey, T. V. (1992) ‘The Denarii of Pescennius Niger,’ unpublished Royal Numismatic Society Presidential Lecture. Cameron, A. (1983) ‘The History of the Image of Edessa: The Telling of a Story,’ Harvard Ukrainian Studies 7: 80–94.

186

BIBLIOGRAPHY

Cameron, R. and Dewey, A. (1979) The Cologne Mani Codex ‘Concerning the Origin of His Body’, Missoula, MT: Scholars Press. Castelin, K. O. (1946) Coinage of Rhesaena in Mesopotamia, New York: American Numismatic Society. Chabot, J.-B. (ed.) (1963) Chronique de Michel le Syrien, Paris: Académie des Inscriptions et Belles-Lettres. Charlesworth, M. P. (1926) Trade-routes and Commerce of the Roman Empire, Chicago: Ares Publishers (reprint 1974). Colledge, M. (1994) ‘Some Remarks on the Edessa Funerary Mosaics,’ in Darmon, J.-P. and Rebourg, A. (eds), La Mosaïque Gréco-Romaine IV, Paris: Association Internationale pour l’Étude de la Mosaïque Antique. Cramer, W. (1980) ‘Irrtum und Lüge: Zum Urteil des Jakob von Sarug über Reste paganer Religion und Kultur,’ Jahrbuch für Antike und Christentum 23: 96–107. Cumont, F. (1926) Fouilles de Doura-Europos (1922–1923), Paris: P. Geuthner. Cureton, W. (1855) Spicilegium Syriacum, Lexington, KY: American Theological Library Assciation (reprint 1965). —— (1864) Ancient Syriac Documents Relative to the Earliest Establishment of Christianity in Edessa and the Neighbouring Countries, Amsterdam: Oriental Press (reprint 1967). Dijkstra, K. (1990) ‘State and Steppe: The Socio-political implications of Hatra Inscription 79,’ Journal of Semitic Studies 35: 81–98. Dillemann, L. (1962) Haute Mésopotamie orientale et pays adjacents, Paris: P. Geuthner. Dodgeon, M. H., and Lieu, S. N. C. (1991) The Roman Eastern Frontier and the Persian Wars, AD 226–363: A Documentary History, London: Routledge. Downey, G. (1961) A History of Antioch in Syria: From Seleucus to the Arab Conquest, Princeton: Princeton University Press. Drijvers, H. J. W. (1965) The Book of the Laws of Countries: Dialogue on Fate of Bardaisan of Edessa, Assen: Van Gorcum & Co. —— (1966) Bardaisan of Edessa, Dissertation, Groningen. —— (1970) ‘Edessa und das jüdische Christentum,’ Vigiliae Christianae 24: 4–33. —— (1972) Old-Syriac (Edessean) Inscriptions, Semitic Study Series 3, Leiden: E. J. Brill. —— (1974) ‘Mani und Bardaisan: Ein Beitrag zur Vorgeschichte des Manichäismus,’ in Mélanges H. Ch. Puech, Paris: Presses Universitaires de France, 124–46. —— (1975) ‘Bardaisan von Edessa als Repräsentat des syrischen Synkretismus im 2. Jahrhundert n. Chr.,’ in Dietrich, A. (ed.), Synkretismus im Syrisch–persischen Kulturgebiet, Abhandlungen der Akademie der Wissenschaften in Göttingen, Philologisch-historische Klasse 96.

187

ROMAN EDESSA

Drijvers, H. J. W. (1977) ‘Hatra, Palmyra und Edessa,’ Aufstieg und Niedergang der römische Welt 2.8: 799–906. —— (1980) Cults and Beliefs at Edessa, Leiden: E. J. Brill. —— (1982a) ‘A Tomb for the Life of a King: A Recently-discovered Edessene Mosaic with a Portrait of King Abgar the Great,’ Le Muséon 95 (1982): 167–89. —— (1982b) ‘Facts and Problems in Early Syriac-speaking Christianity,’ The Second Century 2.1: 157–75. —— (1992) ‘Syrian Christianity and Judaism,’ in Lieu, J., North, J. and Rajak, T. (eds), The Jews Among Pagans and Christians in the Roman Empire, London and New York: Routledge. —— (1996) ‘Early Syriac Christianity: Some Recent Publications,’ Vigiliae Christianae 50: 159–77. Drijvers, H. J. W. and Healey, J. (1999) The Old Syriac Inscriptions of Edessa and Osrhoene: Texts, Translations and Commentary, Leiden: Brill. Du Mesnil du Buisson, R. (1962) Les Tessères et les monnaies de Palmyre, Paris: E. de Boccard. Dussaud, R. (1903) ‘Notes de mythologie syrienne 2,’ Revue Archéologique 124–48, 347–82. —— (1955) La Pénétration des Arabes en Syrie avant l’Islam, Paris: P. Geuthner. Duval, R. (1892) Histoire politique, religieuse et littéraire d’Édesse jusqu’à la première croisade, Paris. Ehlers, B. (1970) ‘Kann das Thomasevangelium aus Edessa stammen?,’ Novum Testamentum 12: 284–317. Feissel, D. and Gascou, J. (1989) ‘Documents d’archives romains inédits du Moyen Euphrate (IIIe siècle après J.-C.),’ Comptes Rendus de l’Académie des Inscriptions et Belles-Lettres: 535–61. —— (1995) ‘Documents d’archives romains inédits du Moyen Euphrate,’ Journal des Savants: 65–119. Filoramo, G. (1990) A History of Gnosticism, tr. Anthony Alcock, Cambridge, MA: Blackwell. Frye, R. N. (1984) The History of Ancient Iran, Handbuch der Altertumswissenschaft 3.7, Munich: C. H. Beck. Gagé, J. (1930) ‘La Victoria Augusti et les auspices de Tibère,’ Rev. Arch. 32, 1–35. —— (1933) ‘La Théologie de la victoire impériale,’ Rev. Historique 171, 1–43. Garsoïan, N. G., Mathews, T. F. and Thomson, R. W. (eds) (1982) East of Byzantium: Syria and Armenia in the Formative Period, Dumbarton Oaks Symposium 1980, Washington: Dumbarton Oaks. Gawlikowski, M. (1985) ‘Les Princes de Palmyre,’ Syria 62: 251–61. Gesche, H. (1969) ‘Kaiser Gordian mit dem Pfeil in Edessa,’ Jahrbuch für Numismatik und Geldgeschichte 19: 47–77. Gibson, J. C. L. (1963–5) ‘From Qumran to Edessa,’ Annual of the Leeds University Oriental Society 5: 24–34.

188

BIBLIOGRAPHY

Gingras, G. E. (trans.) (1970) Egeria: Diary of a Pilgrimage, New York: Newman Press. Goetze, A. (1953) ‘An Old Babylonian Itinerary,’ Journal of Cuneiform Studies 7: 51–2. Goldstein, J. (1966) ‘The Syriac Bill of Sale from Dura-Europos,’ Journal of Near Eastern Studies 25.1: 1–16. Goossens, G. (1943) Hiérapolis de Syrie: Essai de monographie historique, Louvain: Bibliothèque de l’Université. Grégoire, H. (1955–7) ‘Bardesane et S. Abercius,’ Byzantion 25–7: 363–8. Green, T. M. (1992) The City of the Moon God: Religious Traditions of H . arran, Leiden: Brill. Grifﬁth, S. (1995) ‘Asceticism in the Church of Syria: The Hermeneutics of Early Syrian Monasticism,’ in Asceticism, New York: Oxford University Press. Guey, J. (1937) Essai sur la guerre parthique de Trajan, Bucharest: Imprimerie Nationale. Gutschmid, A. von (1887) Untersuchungen über die Geschichte des Königsreichs Osroene, St Petersburg: Acad. Imp. Sc. de St Pétersbourg. Haase, F. (1910) ‘Zur bardesanischen Gnosis: Literarkritische und dogmengeschichtliche Untersuchungen,’ Texte und Untersuchungen zur Geschichte der altchristlichen Literatur (Leipzig) v. 3 no. 4. Harrak, A. (1992) ‘The Ancient Name of Edessa,’ Journal of Near Eastern Studies 51: 209–14. Hasebroek, J. (1921) Untersuchungen zur Geschichte des Kaisers Septimius Severus, Heidelberg: Carl Winter. Henninger, J. (1976) ‘Zum Problem der Venussterngottheit bei den Semiten,’ Anthropos 71: 129–68. Henrichs, A. and Koenen, L. (1970) ‘Ein griechischer Mani-codex,’ ZPE 5: 125–32. Hill, G. F. (1900) Catalogue of the Greek Coins of Lycaonia, Isauria and Cilicia, London: British Museum. —— (1922) Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia, London: Oxford University Press. Honigmann, E. (1925) ‘Studien zur Notitia Antiochene,’ Byzantinische Zeitschrift 25: 60–88. —— (1930) ‘Urfa keilschriftlich nachweisbar?,’ Zeitschrift für Assyriologie und verwandte Gebiete n.s. 5 (39): 301–2. —— (1942–3) ‘The Original Lists of the Members of the Council of Nicaea, the Robber-Synod and the Council of Chalcedon,’ Byzantion 16: 20–80. Honigmann, E. and Maricq, A. (1953) Recherches sur les Res gestae divi Saporis, Brussels: Académie Royale de Belgique. Howard, G. (trans.) (1981) The Teaching of Addai, Ann Arbor: Society of Biblical Literature. Howgego, C. J. (1985) Greek Imperial Countermarks: Studies in the Provincial Coinage of the Roman Empire, London: Royal Numismatic Society.

189

ROMAN EDESSA

Humphrey, J. H. (1986) Roman Circuses: Arenas for Chariot Racing, London: B. T. Batsford. Ingholt, H. (1932) ‘Deux Inscriptions Bilingues de Palmyre,’ Syria 13: 278–92. Isaac, B. (1980–1) ‘Roman Colonies in Judaea: The Foundation of Aelia Capitolina,’ Talanta 12/13. Jardé, A. (1925) Études critiques sur la vie et le règne de Sévère Alexandre, Paris: De Boccard. Jones, A. H. M. (rev. by Michael Avi Yonah et al.) (1971) The Cities of the Eastern Roman Provinces, Oxford: Oxford University Press. Kennedy, D. L. (1979) ‘Ti. Claudius Subatianus Aquila, “ﬁrst Prefect of Mesopotamia”,’ ZPE 36. —— (1987) ‘The Garrisoning of Mesopotamia in the Late Antonine and Early Severan Period,’ Antichthon 21: 57–66. Kettenhoffen, E. (1983) ‘The Persian Campaign of Gordian III and the Inscription of Sa¯hpuhr I at the Ka be-ye Zartost,’ in Mitchell, S. (ed.), Armies and Frontiers in Roman and Byzantine Anatolia, BAR International Series 156: 151–71. Kirsten, E. (1959) ‘Edessa,’ in Klauser, T. (ed.), Reallexikon für Antike und Christentum 4, Stuttgart: Anton Hiersemann. —— (1963) ‘Edessa: Eine Römische Grenzstadt des 4. bis 6. Jahrhunderts im Orient,’ Jahrbuch für Antike und Christentum 6: 144–72. Kronholm, T. (1978) Motifs from Genesis 1–11 in the Genuine Hymns of Ephrem the Syrian, Stockholm: CWK Gleerup. Kübler, B. (1923) Sella Curulis, RE 2A, 2 cols. 1310–15. Kuttner, A. (1995) Dynasty and Empire in the Age of Augustus: The Case of the Boscoreale Cups, Berkeley: University of California Press. Lammens, H. (1919) ‘Le Culte des bétyles et les processions religieuses chez les Arabes préislamites,’ Bulletin de l’Institut Français d’Archéologie Orientale du Caire 17: 39 ff. —— (1928) L’Arabie occidentale avant l’Hégire, Beirut: Impr. Catholique. Lepper, F. A. (1948) Trajan’s Parthian War, London: Oxford University Press. Levante, E. (1993) Cabinet des Médailles: Cilicie, Sylloge Nummorum Graecorum France, Paris: Bibliothèque Nationale. Levi, D. (1947) Antioch Mosaic Pavements, Princeton: Princeton University Press. Lieu, J., North, J., and Rajak, T. (eds) (1992) The Jews among Pagans and Christians in the Roman Empire, London: Routledge. Lightfoot, C. S. (1990) ‘Trajan’s Parthian War and the Fourth-century Perspective,’ Journal of Roman Studies 80: 115–26. Loriot, X. (1971) ‘Itinera Gordiani Augusti, I. Un Voyage de Gordien III à Antioche en 239 après J.-C.?,’ Bulletin de la Société Française des Numismates 26: 18–21. —— (1974) ‘Les Fasti Ostienses et le dies imperii de Gordien III,’ in Mélanges d’histoire ancienne offerts à William Seston, Paris: E. de Boccard.

190

BIBLIOGRAPHY

—— (1975) ‘Les Premières Années de la grande crise du IIIe siècle: De l’avènement de Maximin le Thrace (235) à la mort de Gordien III (244),’ Aufstieg und Niedergang der römische Welt 2.2: 657–787. Macdonald, G. (1905) Catalogue of Greek Coins in the Hunterian Collection v. 3, Glasgow: University of Glasgow. McVey, K. (1990) ‘A Fresh Look at the Letter of Mara Bar Serapion to his Son,’ in V Symposium Syriacum 1988, Rome: Pont. Institutum Studiorum Orientalium. Magie, D. (1950) Roman Rule in Asia Minor to the End of the Third Century After Christ, Princeton: Princeton University Press. Mango, C. (1986) ‘A Late Roman Inn in Eastern Turkey,’ Oxford Journal of Archaeology 5: 223–31. Mango, M. M. (1982) ‘The Continuity of the Classical Tradition in the Art and Architecture of Northern Mesopotamia,’ in Garsoïan, N. G., Mathews, T. F. and Thomson, R. W. (eds), East of Byzantium: Syria and Armenia in the Formative Period, Washington: Dumbarton Oaks. Maricq, A. (1965) Classica et Orientalia: Extrait de Syria, 1955–1962, Paris: P. Geuthner. Marquart, J. (1901) ‘E¯ra¯nsˇahr nach der Geographie des Ps. Moses Xorenac ı¯,’ Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen, Philologisch-historische Klasse, v. 3, no. 2, Berlin: Weidmann. Martin, J. P. (1876) ‘Discours de Jacques de Saroug sur la chute des idoles,’ Zeitschrift der Deutschen Morgenländischen Gesellschaft 29: 107 ff. Mattingly, H., Sydenham, E., and Sutherland, C. H. V. (eds) (1949) Gordian III–Uranius Antoninus, Roman Imperial Coinage 4.3, London: Spink and Son. Meeks, W. A. (1983) The First Urban Christians: The Social World of the Apostle Paul, New Haven: Yale University Press. Millar, F. (1987) ‘Empire, Community and Culture in the Roman Near East: Greeks, Syrians, Jews and Arabs,’ Journal of Jewish Studies 38.2: 143–64. —— (1993) The Roman Near East: 31 B.C.–A.D. 337, Cambridge, MA: Harvard University Press. Milne, J. G. (1982) Catalogue of Alexandrian Coins, New York: Stanford J. Durst. Mitchell, C. W. (1912–21) S. Ephraim’s Prose Refutations of Mani, Marcion and Bardaisan, London: Text and Translation Society. Mitford, T. B. (1972) The Roman Frontier Based on the Valley of the Upper Euphrates from the Black Sea to Samosata, unpublished dissertation, Oxford University. —— (1980) Cappadocia and Armenia Minor: Historical Setting of the Limes, Aufstieg und Niedergang der römische Welt 2.7.2: 1169–1228. Mouterde, R. (1959) ‘Cultes antiques de la Coelésyrie et de l’Hermon,’ MUSJ 36: 76–8. Murray, R. (1975) Symbols of Church and Kingdom: A Study of Early Syriac Tradition, Cambridge: Cambridge University Press.

191

ROMAN EDESSA

—— (1982) ‘The Characteristics of the Earliest Syriac Christianity,’ in Garsoïan, N. G., Mathews, T. F. and Thomson, R. W. (eds), East of Byzantium: Syria and Armenia in the Formative Period, Washington: Dumbarton Oaks. Nashef, Kh. (1987) ‘Rekonstruktion der Reiserouten zur Zeit der Altassyrischen Handelsniederlassungen,’ in Beihefte zum Tübingen Atlas des Vorderen Orients, Wiesbaden: Ludwig Reichert. Naster, P. (1968) ‘Les Monnaies d’Édesse révèlent-elles un dieu Elul?’, Revue Belge de Numismatique 114: 5–13. Neymeyr, U. (1989) Die Christlichen Lehrer im zweiten Jahrhundert, Leiden: Brill. Nöldeke, Theodor (1908) ‘Syrische Inschriften,’ Zeitschrift für Assyriologie und verwandte Gebiete 21: 151–61. Oates, D. (1955) ‘A Note on Three Latin Inscriptions from Hatra,’ Sumer 11: 39–43. —— (1968) Studies in the Ancient History of Northern Iraq, London: Oxford University Press. Oden, R. A. (1977) Studies in Lucian’s De Syria Dea, Missoula, MT: Scholars Press. Ortiz de Urbina, I. (1965) Patrologia Syriaca, 2nd ed., Rome: Pont. Institutum Orientalium Studiorum. Parlasca, K. (1984) ‘Neues zu den Mosaiken von Edessa und Seleukeia am Euphrat,’ in Campanati, R. (ed.), III Colloquio Internazionale sul Mosaico Antico, Rome: Edizioni del Girasole. Parthey, G. and Pinder, M. (eds) (1848) Itinerarium Antonini Augusti et Hierosolymitanum, Berlin: F. Nicolai. Payne Smith, J. (1903) A Compendious Syriac Dictionary Founded Upon the Thesaurus Syriacus of R. Payne Smith, D.D. Oxford: Clarendon Press (reprint 1967). Payne Smith, R. (1879) Thesaurus Syriacus, Oxford: Clarendon Press. Perkins, A. (1973) The Art of Dura-Europos, Oxford: Clarendon Press. Pﬂaum, H.-G. (1960) Les Carrières procuratoriennes équestres sous le haut-empire romain, Paris: P. Geuthner. Pognon, H. (1907) Inscriptions sémitiques de la Syrie, de la Mesopotamie et de la région de Mossoul, Paris: LeCoffre and Gabalda. Potter, D. S. (1990) Prophecy and History in the Crisis of the Roman Empire: A Historical Commentary on the Thirteenth Sibylline Oracle, Oxford: Oxford University Press. Quasten, J. (1950) Patrology, Utrecht: Spectrum (paperback ed. 1990, Westminster, MD: Newman Press). Regling, K. (1901–2) ‘Zur historischen Geographie des mesopotamischen Parallelogramms,’ Klio 1: 443–76. Richter, G. (1926) Ancient Furniture: A History of Greek, Etruscan and Roman Furniture, Oxford: Oxford University Press. Robinson, J. (ed.) (1990) The Nag Hammadi Library in English, 3rd ed., New York: HarperCollins.

192

BIBLIOGRAPHY

Ross, Steven K. (1993) ‘The Last King of Edessa: New Evidence from the Middle Euphrates,’ ZPE 97: 187–206. Rostovtzeff, M. I. (1935) ‘Dura and the Problem of Parthian Art,’ Yale Classical Studies 5: 155–301. —— (1943) ‘Res Gestae divi Saporis and Dura,’ Berytus 8: 17–60. Rudolph, K. (1983) Gnosis: The Nature and History of Gnosticism, Edinburgh: T. & T. Clark. Salonen, A. (1963) Die Möbel des alten Mesopotamien nach sumerisch– akkadischen Quellen, Helsinki: Suomalainen Tiedeakatemia. Sartre, M. (1984) ‘Le Dies imperii de Gordien III: une inscription inédite de Syrie,’ Syria 71: 49–61. Scher, A. (1908) Cause de la fondation des Écoles, Patrologia Orientalia 4. Schlumberger, D. (1960) ‘Descendants non-Méditerranéens de l’art grec,’ Syria 37: 131–66. Schulthess, Friedrich (1908) ‘Die syrischen Kanones der Synoden von Nicaea bis Chalcedon nebst einigen zugehörigen Dokumenten herausgegeben,’ Abhandlungen der Königlichen Gesellschaft der Wissenschaften Zu Göttingen, Philologisch-historische Klasse 10.2. Schwartz, E. (ed.) (1932) Acta Conciliorum Oecumenicorum, II.4–6, Concilium Universale Chalcedonense, Berlin: Walter De Gruyter. Seeck, O. (ed.) (1876) Notitia Dignitatum: Accedunt Notitia Urbis Constantinopolitanae et Laterculi Provinciarum, Berlin: Weidmann. Segal, J. B. (1953) ‘Pagan Syriac Monuments in the Vilayet of Urfa,’ Anatolian Studies 3: 97–120. —— (1954) ‘Some Syriac Inscriptions of the 2nd–3rd Century A. D.,’ Bulletin of the School of Oriental and African Studies 16: 13–31. —— (1959a) ‘New Mosaics from Edessa,’ Archaeology 12: 150–7. —— (1959b) ‘New Syriac Inscriptions from Edessa,’ Bulletin of the School of Oriental and African Studies 22: 23–40. —— (1960) ‘A Note on a Mosaic from Edessa,’ Syria 60: 107–10. —— (1963a) ‘The Sabian Mysteries: The Planet Cult of Ancient H . arran,’ in Bacon, E. (ed.), Vanished Civilizations, London: Thames and Hudson. —— (1963b) Edessa and H . arran, Inaugural Lecture, London: School of Oriental and African Studies. —— (1967) ‘Four Syriac Inscriptions,’ Bulletin of the School of Oriental and African Studies 30: 293–7. —— (1970) Edessa: The ‘Blessed City’, Oxford: Oxford University Press. —— (1985) ‘Abgar,’ in Yarshater, E. (ed.) Encyclopaedia Iranica, London: Routledge and Kegan Paul. —— (1986) ‘Arabs at Hatra and the Vicinity: Marginalia on New Aramaic Texts,’ Journal of Semitic Studies 31: 57–80. Seyrig, H. (1951) ‘Le Repas des morts et le “banquet funèbre” à Palmyre,’ Annales Archéologiques de Syrie 1: 32–9. —— (1963) ‘Les Fils du roi Odainat,’ Annales Archéologiques de Syrie 13: 160–1. —— (1971) ‘Bêl de Palmyre,’ Syria 48: 85–114.

193

ROMAN EDESSA

Seyrig, H., Amy, R. and Will, E. (1975) Le Temple de Bel à Palmyre, Paris: P. Geuthner. Shore, F. (1993) Parthian Coins and History: Ten Dragons Against Rome, Quarryville, PA: Classical Numismatic Group. Smelik, K. A. D. (1974) ‘Aliquanta Ipsius Sancti Thomae,’ Vigiliae Christianae 28: 290–4. Sprengling, M. (1937) ‘A New Pahlavi Inscription,’ American Journal of Semitic Languages and Literature 53.2: 126–44. —— (1953) Third-century Iran: Sapor and Kartir, Chicago: Chicago University Press. Stewart, A. (1990) Greek Sculpture: An Exploration, New Haven: Yale University Press. Stroumsa, G. (1984) Another Seed: Studies in Gnostic Mythology, Nag Hammadi Studies no. 24, Leiden: E.J. Brill. Strugnell, J. (1959) ‘The Nabataean Goddess “Al-Kutba” and her Sanctuaries,’ Bulletin of the American Schools of Oriental Research 156: 29–36. Tabula Peutingeriana: Codex Vindobonensis 324, facsimile, Graz: Akademische Druck-u. Verlagsanstalt, 1976. Teixidor, J. (1961) ‘New Evidence on the North Arabian Deity AktabKutbâ,’ Bulletin of the American Schools of Oriental Research 163: 22–6. —— (1967–8) ‘The Kingdom of Adiabene and Hatra,’ Berytus 17: 1–11. —— (1989) ‘Les Derniers Rois d’Édesse d’après deux nouveaux documents syriaques,’ Zeitschrift für Papyrologie und Epigraphik 76: 219–22. —— (1990) ‘Deux documents syriaques du IIIe siècle après J.-C., provenant du Moyen Euphrate,’ Comptes Rendus de l’Académie des Inscriptions et Belles-Lettres 144–66. —— (1992) Bardesane d’Édesse: la première philosophie syriaque, Paris: Éditions du Cerf. Toynbee, J. M. C. (1934) The Hadrianic School: A Chapter in the History of Greek Art, Cambridge: Cambridge University Press. Tscherikower, V. (1927) Die hellenistischen Städtegründungen von Alexander dem Grossen bis auf die Römerzeit, Philologus Supplementband 19. Volterra, E. (1971) ‘Una misteriosa costituzione dell’ imperatore Gordiano,’ in P.J. Verdam (ed.), Uit het Recht (Deventer: Kluwer): 207–18. Vööbus, A. (1958) History of Asceticism in the Syrian Orient I: The Origin of Asceticism, Early Monasticism in Persia, CSCO v. 184, Subsidia 14. —— (ed.) (1960) Syriac and Arabic Documents Regarding Legislation Relative to Syrian Asceticism, Stockholm: Estonian Theological Society in Exile. Wagner, J. (1983) ‘Provincia Osrhoenae: New Archaeological Finds Illustrating the Military Organization Under the Severan Dynasty,’ in Mitchell, S. (ed.), Armies and Frontiers in Roman and Byzantine Anatolia, BAR International Series 156: 103–29. —— (1985) ‘Die Römer am Euphrat und Tigris’, Die Antike Welt, Sondernummer. Wanscher, O. (1980) Sella Curulis: The Folding Stool, Copenhagen: Rosenkilde and Bagger.

194

BIBLIOGRAPHY

Weder, M. (1982) ‘Seltene Münzen der Sammlung Dattari – Neuerwerbungen des Britischen Museums,’ Numismatische Zeitschrift 96: 53–71. Welles, C. B. (1951) ‘The Population of Roman Dura,’ in Coleman-Norton, P. R. (ed.), Studies in Roman Economic and Social History in Honor of Allan Chester Johnson, Princeton: Princeton University Press. —— (1970) ‘The Gods of Dura-Europos,’ in Stiehl, R. and Stier, H. E. (eds) Beiträge zur alten Geschichte und deren Nachleben: Festschrift für Franz Altheim, Berlin: Walter de Gruyter. —— (1974) Royal Correspondence in the Hellenistic Period: A Study in Greek Epigraphy, 2nd ed., Chicago: Ares Publishers. Welles, C. B., Fink, R. O. and Gilliam, J. F. (1959) The Parchments and Papyri: The Excavations at Dura-Europos, Final Report, vol. 5, part 1, New Haven: Yale University Press. Whittaker, M. (ed. and trans.) (1992) Tatian: Oratio ad Graecos and Fragments, Oxford: Clarendon Press. Widengren, G. (1960) Iranische–semitische Kulturbegegnung in parthischer Zeit, Köln: Westdeutscher Verlag. —— (1985) ‘Bardesanes von Edessa und der syrisch–mesopotamische Gnostizismus,’ in Borgen, P. (ed.), The Many and the One, Trondheim: Tapir. Will, E. (1987) ‘Qu’est-ce qu’une baris?’, Syria 64: 253–9. Witakowski, W. (1987) The Syriac Chronicle of Pseudo-Dionysius of Tel-Mahre, Uppsala: Uppsala University. Wright, W. (1871) Apocryphal Acts of the Apostles, London: Williams and Norgate (reprint Amsterdam: Philo Press, 1968). —— (trans.) (1882) The Chronicle of Joshua the Stylite, Amsterdam: Philo Press (reprint 1968). Wroth, W. (1903) Catalogue of the Greek Coins of Parthia, London: British Museum.

195

INDEX

Abdu 163 Abednebo 176 Abercius Inscription 117, 137 Abercius, Life of 137, 178 Abgarid dynasty 3, 9, 12, 19, 22–3, 57, 63, 69, 84, 90, 143 Abgar ‘Peqa’ 164 Abgar Prahates 172 Abgar, son of Abgar (in inscription at Rome) 169 Abgar, son of Ma nu (in an Edessan mosaic) 111, 113 Abgar V Ukkama 24, 117, 142, 181 Abgar VII 44 Abgar VIII (‘the Great’) 13, 24, 28, 39, 41, 43, 46, 48, 52, 56–7, 60–5, 67, 70, 81, 83–4, 109–11, 114, 116, 133–5, 138, 141–2 Abgar IX Severus 43, 57–8, 60–4, 68 Abgar X 1, 27–8, 68–9, 72–3, 75–6, 82, 111–12, 115, 139–40, 145, 172 Abraham, Era of 29, 30, 36 Absamya, son of Adona 39 Abshelama 136 Acilius Aviola 77 Acts of Thomas 126, 129 Addai 60, 89–90, 100, 118, 127, 129, 131, 133, 135–7 Adiabene 10, 20, 23, 26, 35, 46, 48–9, 53, 165 Adme ( DM /Admi/Admum) 5 Aelius Ianuarius 54 Asˇerah 176

Afranius 10, 20 Aggai 17, 133, 135 Agrippa I 175 Ala Prima Parthorum 165 Alexander the Great 6, 33, 59, 152 Al-Nadim 178 Alps 55 Al-Rahha, Arabic name of Orhai 5 Amida 15 Ananias 132 Anat 176 Annius Arrianus 58 Anthemusia 16, 27, 34, 38, 166 Antigonus 7 Antioch 7–8, 12, 31, 34, 37, 49, 51, 60, 71, 77, 79–80, 83, 88, 115–16, 130–1, 133, 135, 139, 148, 154, 158, 168, 172, 177 Antipater 7 ‘Antiphoros’ 110 anti-Semitism 120, 132 Antonine Itinerary 17 ‘Antoninus Caesar’ 41, 85 Antoninus, Roman turncoat 21 Antoninus, son of Abgar (in inscription at Rome) 60 Antoninus, Syrian usurper 67 Apamea 7 Apatna 165 Aphrahat 128, 130 Aphrodisias 170 Aphrodite 88, 101 Aphtuh.a 1–2 Apollo 88–9, 94, 101, 184 Arab 25–6, 28, 39, 42–3, 91 Arabarchos 25

197

INDEX

‘Arabia’ 34, 48 Arabia, Province 29, 33, 48, 62 Arabic 9, 12, 63 Arabs 46, 48, 53, 62–3, 85, 87; Orroei 23, 26; Praetavi 23; Scenitae 23 Aramaic 2, 8, 12 Arbayestan 165 archers, Osrhoenian 52, 67, 139 Ardashir 65–6 Ares 88–9 Arians 123–4, 136 Armenia 10–11, 13, 15, 20, 22, 24, 27, 29, 31–3, 37–8, 44, 52, 56, 140, 151, 155–6, 165 Arsaces 184 Artabanus 65, 150 Artaxata 37 Artemis 88, 94, 101 Arvastan 165 Asia Minor 5, 7, 131, 138, 140 Assyria (Trajanic province) 33 astrology 122 Asuristan 71 Atargatis 85–7, 89, 90, 94, 101–2, 109, 115, 134, 143 Asˇtarte 176 Athens 116, 121, 142, 164 ‘Audians’ 123 Augustus 56, 110, 183 Aurelianus, son of H . afsai 43, 68 Aurelius H . afsai 25, 41, 43 Aurelius Heliodoros 167 Aurelius Zoticus 57, 170 Avidius Cassius 38 Awida 120 Aziz and Mun im 88 Baalbek 88 Ba alshamin 91–2, 94, 101 Babylon 7 Babylonia 16, 172 Bactria 7 baetyls 3–40, 42, 91–2, 101, 143 Bakru 85, 163–4 Balbinus and Pupienus 70, 78 Balien[a] 166 Balikh River 14, 16 Banasam 165 Bar-Bahloul 5

Bardais.an 13, 48, 62, 84, 114, 118–19, 120–4, 126–7, 131, 133–4, 136–7, 141–2, 178 Bardesanites 123, 127, 141 baris 73–4 Bar Nahar 41–3 Barsamya, Martyrdom of 135 Barsemias of Hatra 48 Barsimya 111–13, 116, 136 Bath Nical 87 Batnae 14–16, 22–4, 27–8, 32, 34, 86, 166 Beirut 177 Bel 87–8, 90, 92, 94, 101, 143 Belai Bar Gusai 111 Beth Arbaye 165 Beth Phouraia 58 Beth Sah.raye 106, 108 Beth Tabara 106, 108 Birecik 24, 28, 166 Bithynia 184 Book of the Law of Countries 62, 85–6, 89, 96, 119, 121–3, 126, 134 Boscoreale 183 Bostra 177 BWDR (budar) 39–42, 91 Byzantine Empire 117 Byzantium 47 Caesarea Maritima 177 Callinicum 22, 165–6 Callirhoe 163–4 Cappadocia 37, 55, 184 Caracalla 17, 55, 57, 59–60, 83, 140, 153, 158, 160, 162; assassination 17, 66; foreign policy and wars 32, 64, 66; removal of Abgar of Edessa 57–60 Caria 170, 184 Carinus 153 Carpi 175 Carrhae 5, 10, 14, 17, 18, 22–4, 51, 71, 75, 78, 81, 158, 160, 162, 163; see also H . arran Carus 153 Cassander 7 Cendere 55 Cervonius Papus 58, 174

198

INDEX

Chabina River 55 ‘Chair of Hegeso’ 151 Chalcedon, Acts of 166 Charax Sidou 27 China 15, 164 Chosroes 164, 165 Christianity 22, 117–19, 122, 125–6, 129, 131; at Edessa 3, 12, 22, 60, 100, 108–9, 117, 118, 123–4, 127–38, 142–3; in Roman Empire 13, 100, 124, 131, 142 Chronicle of Edessa 9, 24, 28, 57, 104, 106, 110, 117, 121, 133 Chronicle of Joshua the Stylite 86, 100, 110, 166 Chronicle of Zuqnin 9, 29–30, 34–6, 44, 60–1, 117 Church of St. Stephen 104 Cilicia 184 Circesium 22, 54, 72, 165–6 circumcision 64 Citadel 1, 3–4, 11, 18–19, 89, 102, 107–9, 115 C. Julius Pacatianus 49, 50, 54 Classical culture 3, 84, 115, 121–2 Clodius Albinus 54, 56 Codex Justinianus 77 coinage of Alexandria 78; of Antioch 77; of Edessa 36, 38–9, 44, 46, 49–50, 57, 60–1, 68–70, 72–3, 75–7, 79, 81–2, 90, 92, 94, 96, 101, 114–15; of Hierapolis 90; of Mesopotamian cities 26, 55–6; Parthian 114; Roman 31, 33, 76, 79 Commagene 22, 55 Commodus 27, 39, 46, 50, 70 Constantine the Great 13 Constantius II 124, 141 Constitutio Antoniniana 63 Corbulo 11 20 Ctesiphon 16, 21, 32–3, 38, 44, 56, 71

Diadumenian 184 Dialogue on Fate see Book of the Laws of Countries Diatessaron 122, 128–9, 178 Diocletian 109, 141 Dionysius of Tell-Mahre see Chronicle of Zuqnin Doctrine of Addai 16, 60, 83, 86–8, 90, 100–1, 104, 108, 127, 133, 135, 136, 149, 164, 176 ‘Documents A and B’ (from Euphrates Papyri archive) 27–8 Domitian 164 Dura-Europos 12, 25, 33, 38, 43, 58–9, 73, 78, 80–1, 83, 94, 101 Eagle (Arabian deity) 87 Easter Controversy 27, 128 Edessa

Dacia 29 Dais.an 14, 102, 105–6, 122; see also Kara Koyun Daphne 37, 100 Derik Kale 55

199

Basilica 109; baths 19, 104, 109, 110; becomes Roman municipality 12, 57, 58, 60–1, 67, 83, 115; Byzantine 117; Christianity 3, 12, 22, 60, 100, 108–9, 117, 118, 123–4, 127–30; ‘Church of the Christians’ 105, 108–9, 117, 121, 123, 133; city plan 102, 104, 107, 108, 109, 110; Crusader period 18, 102; in cuneiform records 5, 14–15; ethnic make-up 8–9, 23; ﬂoods 24, 57, 104–9, 116–17; foreign policy 10–11, 20–2, 33, 34–8, 40, 42–53, 56–7 64, 66–70, 75, 79; fortiﬁcations 1, 16, 18–19, 89, 101–2, 104–5, 107, 110, 164; Hellenism 8–9, 11–13, 68–9, 84, 94, 96, 99, 101, 110, 113–16, 118–20, 122–3, 131, 137, 141, 144; Hellenistic foundation 3, 8, 11, 102; languages 2, 7, 12–13, 83, 116, 123, 143; Parthian overlordship 3, 9, 11, 20, 25, 35–7, 40, 44, 83, 107, 115, 123, 139; religion 3–4, 25, 40, 84– 90, 94, 96, 99, 100–1, 116, 122, 124–6, 131, 133, 135, 137, 143, 176; role in international trade 3,

INDEX 14–16, 19, 27–8, 116; Roman colonia 57–9, 65, 67, 73–5, 90, 94, 114, 139, 143, 158–9, 161, 163–4; royal palaces 102, 105–11, 116; territory 22–7, 49–52, 64; water sources 14, 18–19, 27, 89, 102, 104

Gotarzes 10 Goths 175

Egeria 19, 89, 109, 130, 142, 164, 181 Egypt 7–8 Elagabalus 57, 59, 64, 66–7, 122, 158, 163, 168, 170, 176 Elegeia 31 Elul, presumed deity 92 Emesa 88, 168, 171, 176 Encratites 129 Ephesus 88 Ephrem Syrus 119, 123, 127, 141, 179 Epiphanius 179–81 Eski Hissar 54–5, 65 Eski Serug 27 Eulogius 165 Euphrates 20, 27, 33 Euphrates Papyri 73, 79 Eutychides 158 First Triumvirate 10 ﬁsh 14, 89–90, 102, 104, 109 Flavius Heraclio 67 Fronto 37 Galatia 184 Gallaba 165 Gallienus 153, 183 galloi 181 Gebar u 163–4 Gelasius 133 Gerasa 177 Germanicus 183 Germany 68, 171 Geta 55, 183 Gnosticism 118, 120, 123–30, 137 Gordian I 70 Gordian II 70 Gordian III 27, 58, 66, 68–72, 75–9, 81–2, 112, 139, 145–8, 151, 154, 163 Gospel of John 125, 180

Hadad 89–90, 94, 101 Hadrian 29, 33, 35 H . arran 5, 14, 18–19, 28, 42, 87, 91, 101, 124, 143; see also Carrhae Hatra 17, 32, 48, 78, 113, 171, 174 Helios 87–8, 92, 101, 143 Hellenism 2, 8–9, 11–13, 113–15, 118–19, 123, 144, 164 Hemerias 166 Hera 99 Herculaneum 182 heresies 118, 120–1, 123, 125, 127, 137 Hermes 88–9 Hermetic corpus 118, 137 Herod 175 Hierapolis 16–17, 24, 85, 87, 89–90, 99–100, 109 hippodrome 56, 110, 177 Hippolytus 178–9 ‘House of Eternity’ 96, 99, 113 ‘Hymn of the Pearl’ 126, 129 India 7, 15, 122, 129, 179 Islamic Conquest 20, 84, 141 Issus 47, 53 Italy 55 Iulii 135 ius Italica 59, 168 Jacob of Edessa 60, 63, 68, 110 Jacob of Serug 86–7, 90, 96, 101 Jerome 129, 181 Jerusalem 117, 132 Jesus 19, 56, 87, 109, 117–18, 125–6, 131–3, 135–6, 142, 164, 166 Jews 16, 120, 124, 130–2 John Chrysostom 100 Julia Domna 66 Julius Africanus 122, 134, 178–9 Julius Priscus 80 Julius Proculus 80 Justinian 22, 102, 107, 164, 177 Justin Martyr 128

200

INDEX

Kara Koyun 14, 102; see also Dais.an Khabur 16–17, 22, 26, 28 Koinoboulion 153 Kutbai 85, 175 Kızılburç 50 Laetus 52 Lake Van 31 Laodicea 7, 49, 168 Lateranus 52 legions 67; Egyptian and Syrian 38, 66–7; IV Scythica 54–5, 65; Parthian 54–5, 116, 165 Leon 166 Lepidus Praetextatus 58, 173 Lex Romana Wisigothica 77 Liber Junioris Philosophi 15 Lucian 37–8, 48, 64, 86, 89–90, 94, 99–100, 121, 137, 142, 167, 176, 181 Lucius Verus 20–1, 29, 36–7, 41, 46, 48, 54, 83, 115, 151, 153, 155–6, 167, 178, 184 Lusius Quietus 21, 33, 35, 44 Lycaonia 151, 184 Lysimachus 7 Macedonians 7, 94, 163 Macrinus 66, 166, 170, 171 Maioumas 100 Malalas 8, 31, 35, 158 mandylion 181 Mani 124–7, 136–7, 180 Manichaeism 126, 136, 141 Ma nu 23; bar Ma nu 24, 36; III Saphlul 16; of Singhara 34; ‘Philorhomaios’ 36, 39, 44, 45, 111, 114–15; son of Izates 30, 35, 36, 38; son of Moqimu 41; the pas.griba 1, 2, 61, 65, 69, 109, 112; on tomb at Serrin 24 Mara bar Serapion 119, 178 Maratha 166 Marcion 120–1, 124–7, 129, 179, 180 Marcionites 108, 121, 127 Marcopolis 27, 74, 166, 173 Marcus Aurelius 29, 36–7, 39, 41, 68, 153, 166–7, 173–4

Marcus Aurelius Abgar 68, 174 Marcus Junius Philippus (Philip I) 71 Marcus Licinius Crassus 10 Mauretania 174 Maximinus 68, 78, 153, 171–2, 183 Maximus 153, 183 Mediana 165 Meherdates 10, 20, 24, 83 Meliton 85–6, 119 Mesopotamia 2, 4–11, 13–14, 18, 20–3, 25, 27, 29–33, 38, 44, 48, 49, 54, 56, 64, 66–7, 71, 75, 81, 84–5, 87, 89, 100, 115, 118, 130, 158, 163; Roman province 15, 22, 35, 44, 55, 57, 62, 64, 71, 78, 80, 84, 140 Mettolbaesos 25 Misikhe 71 Mithra 184 Moesia 175 monasticism 130 Monimos and Azizos see Aziz and Mun im Moqimu in Sumatar inscription 41; in Family Portrait Mosaic 149 mosaics 57, 83, 91–6, 99, 111–13, 115–16, 119, 136, 143–4, 149 municipium 59, 64–5, 81, 110, 139 Mygdonius 26 Mysianus 109 Nabataeans 163 Nag Hammadi 129, 180 Nah.ay 91–2, 114 Naqsh-e Rustam 21, 71, 140 Naulobatos 175 Nea Valentia 166 Nebo 85, 87–90, 92, 94, 101, 143, 176 Neoplatonism 120, 125 Nero 11, 51, 151, 183 Nicopolis 17 Nike 145, 152–3 Nineveh 155 Nisibis 14–15, 17, 20–1, 23, 26, 28, 31–2, 34, 38, 42–3, 46–9, 51–3, 55–6, 71, 75, 78, 81, 117, 123, 130, 141, 158, 164–6, 168, 170, 172, 174, 179

201

INDEX

Notitia Antiochena 27, 166 Notitia Dignitatum 22, 54, 170 Numerianus 153, 183 NWHDR 39, 41, 42 Odenathus 79, 81, 140, 175 Omphalos 158 Oratio ad Graecos 122, 128, 181 Orhai 5, 8, 22, 23, 73–4, 140, 164–5, 175 Ornamenta consularia 79, 80 Oroba 165 Orodes II 184 Orphism 96 orthodoxy 118–19, 126–7 Osrhoene 1, 3, 13–15, 20, 22–4, 27–8, 34, 38, 41, 46, 48–9, 51–2, 55, 58, 60, 65–7, 81, 91, 116, 128, 139, 142, 151; kingdom 28, 52; Roman province 22, 26, 44, 47, 49, 54, 58, 62, 84, 140 Pacatianus 55 Pacorus 37 pas.griba 1–2, 61, 65, 69, 73–4, 83, 163, 173 Palmyra 12, 79, 87, 94, 101, 113 Palut. 123, 127, 133, 135 Palutians 124, 127, 136 Pamphylia 184 Parapotamia 25 Parthamasires 31 Parthamaspates 33, 35, 156 Parthia 3, 9, 11, 20, 25, 34–5, 46, 52, 55, 65, 71, 83, 84, 107, 114–15, 140, 150, 154–6; confrontation with Rome 10–11, 17, 20–1, 26–7, 30, 32–3, 35–8, 40, 42, 44, 48, 55, 64, 66, 78, 123; culture 2, 9, 11, 83, 112, 115, 144 Paul (Apostle) 131 P. Dura 28 (Edessan document discovered at Dura-Europos) 58 Peroz-Shapur 71, 172 Persia 5, 7–9, 15, 19, 22, 149, 154; Achaemenid dynasty 7, 65; confrontation with Rome 15, 21, 27, 56, 65, 67, 70–2, 75,

78, 81, 123–4, 139–41, 154; religion 126; Sassanid dynasty 65 Persian Gulf 33, 71 Pertinax 47, 168 Pescennius Niger 47, 168 Peshitta 130 Peutinger Tablet 17, 169 Philip I (‘the Arab’) 60, 80, 140, 153, 183 Philip II 153 philosophy 12–13, 114, 118–22, 125–6, 128–9, 137, 141–2, 178–80 Phoenicia 90, 182 Phraates III 182 Phradasht 164 Phrygia 117, 184 Pisidia 184 Plato 120 Pleroma 125 Pliny the Elder 8 Pompeianus 70 Pompey 10, 20, 23 Pontus 121 Pope Victor 27 Praetorian Guard 76 ‘proto-monasticism’ 181 Ptolemais 184 Pupienus 70, 78 Qumran 130 Quna 123–4, 127, 136, 179 Rabbula 170, 181 Rasin 54, 165 Res Gestae divi Saporis 71, 172 Rhesaina 18, 22, 25–6, 165 Roman Empire 2–4, 9–11, 19, 23, 30, 34–6, 38–40, 42–3, 45–6, 51, 55, 56, 60, 62–70, 75, 78–9, 81, 91, 94, 114–16, 123, 139, 154, 164; confrontation with Parthia 20, 78; confrontation with Persia 21–2, 27; foreign policy 10–11, 15, 29, 31, 42, 46–7, 59, 140 Romanization 61, 68, 101, 116, 139 Rome 27, 31, 33, 51, 56–7, 60,

202

INDEX

78–9, 88, 121, 128, 131, 135, 142, 156 Sabinianus 174 Sahru 36, 40, 43, 92, 101, 114, 163 Samaria 125 Samosata 89, 121, 142, 178, 184 S¸anlıurfa see Urfa Sapor see Shapur I Saracens 62, 165 Sardis 5 Sasan 65 Saturninus 129 Scenae 16 Scenitae 16, 23 ‘School of the Persians’ 179 Seleucia-in-Pieria 7 Seleucia on the Eulaeus 83 Seleucia on Tigris 16, 38 Seleucid Era 7, 29, 30, 58, 73 Seleucid Kingdom 3, 7–9, 11, 83, 152, 154, 155 Seleucus I Nicator 7 self-emasculation 61, 64, 85, 101, 134 sella castrensis 151 sella curulis 145, 151, 182 Septimius Severus 20, 22, 26, 29, 32, 38, 41–3, 46–57, 64–7, 83, 104, 111, 116, 140, 163 Serapion 119, 133, 135, 178 Serrin 24, 28, 92 Severianus 37 Severus Alexander 59, 66–7, 78, 139, 153, 163, 166, 176 shallita de- Arab 25, 26, 39 Shapur I 21, 65, 71–2, 75, 140 Sharbel 176 Shelmath 1–4, 102 Shmona and Guria 109, 124, 141, 164, 177, 181 silk 16–17, 27, 164 Silk Road 27, 164 Simon Magus 125 Sin (moon-god) 17, 39–40, 42–3, 91–2, 101, 120, 143, 176 Singhara 17–18, 21, 23, 34, 55–6, 165, 170, 174

Skirtos River see Dais.an; Kara Koyun Sons/Daughters of the Covenant 130 Sophia 125 Spes 153 Sporaces 24 Statius Priscus 37–8 stoai 110, 115, 144 Sumatar Harabesi 25–6, 39, 68, 83, 91, 113, 163 Susa 5 SˇWDR 41 Syna Judeorum 165 Syria 2, 7, 10–11, 15, 17, 20–1, 24, 33, 37–8, 47, 53–5, 60, 62, 66, 78, 80, 84–5, 100, 113, 115–16, 134, 138, 140 Syriac 2, 5, 8, 12, 22, 24, 68, 74, 84, 88, 92, 114, 119, 121–3, 128, 137–8, 141–3; Christian literature 16, 29–30, 57, 60, 68–9, 86, 118–19, 123, 126, 129–30, 133, 138; culture 2, 11; inscriptions 1, 24, 25, 36, 94, 96, 111, 113, 144 T. Claudius Subatanius Aquila 170 Tarsus 151, 153, 155 Tatian 121, 128–9, 131, 137, 142, 178–9, 181 Taurinus 171 Tektek Dagh 24–5, 28 Tella 25–6, 166 Tell Mahre 166 Temple of Janus 70, 174 Tertullian 129, 180 Thaddaeus 133, 135, 137 Theodosius 100 Thillazamara 165 Thomas (Apostle) 129, 130 Throne of Nimrud 1 tiara 83, 111–12, 114, 134, 139, 145–6, 149–50, 156 Tiberius 183 Ti. Claudius Candidus 55, 169 Tiglath-Pileser 184 Tigranes 184 Tigris 13, 17, 23, 26, 33 Timesitheus 70–1, 75–6

203

INDEX

Tiridates of Armenia 51, 57, 151 Tiridates son of Adona 39–42, 45 Tower of the Persians 107 trade routes 3, 15–16, 19, 28, 116 Trajan 4, 21, 23–4, 46, 71, 83, 136, 140, 151, 155–6, 163–4, 166–7, 170, 184 Trajan Decius 163–4, 166, 170 Tranquillina 70, 75–6, 163–4, 170 Tur Abdin 24 Turkey 1 Tyche 90, 115, 143, 157–62, 164 Ulu Camı 104 Uranius 67, 171 Urfa 1, 5, 8, 14, 18, 27, 50, 90, 102, 104, 163 Valentinianism 180 Valentinians 123 Valerian 21, 72, 140, 153, 173, 183

Venus goddess 135, 183; planet 88–9

Veronica’s Veil 133 Vettius Atticus 173 Victoria Augusti 153 Virans¸ehir see Tella Vologaeses II 150, 182 Vologaeses III 36, 114, 182 Vologaeses IV 158 Vologaeses V 182, 184 Wael, son of Sahru 25, 36, 38, 40, 42–4, 92, 101, 114, 143 Wael, son of Wael 40–1, 43 Worod 27 Zaitha 72 Zephyrinus 135 Zeugma 16, 18, 54, 169 Zeus 55, 88–9, 90, 94, 99, 101, 151 Zoroastrianism 22

204