Religious Culture in Late Antique Arabia: Selected Studies on the Late Antique Religious Mind 9781463206307, 1463206305

Citation preview

Religious Culture in Late Antique Arabia

Islamic History and Thought

6 Series Editorial Board Series Editor

Peter Adamson Isabel Toral-Niehoff Isabel Toral-Niehoff Ahmad Khan Manolis Ulbricht Jack Tannous Jan Just Witkam

Ahmad Khan

JackEditorial TannousBoard Advisory

Binyamin Abrahamov Konrad Hirschler Manolis Ulbricht Asad Q. Ahmed James Howard-Johnston Mehmetcan Akpinar Maher Jarrar Jan Just Witkam Abdulhadi Alajmi Marcus Milwright Mohammad-Ali Amir-Moezzi Harry Munt Massimo Campanini Gabriel Said Reynolds Agostino Cilardo Walid A. Saleh Godefroid de Callataÿ Jens Scheiner Farhad Daftary Delfina Serrano Beatrice Gruendler Georges Tamer Wael Hallaq

Islamic History and Thought provides a platform for scholarly research on any geographic area within the expansive Islamic world, stretching from the Mediterranean to China, and dated to any period from the eve of Islam until the early modern era. This series contains original monographs, translations (Arabic, Persian, Syriac, Greek, and Latin) and edited volumes.

Religious Culture in Late Antique Arabia

Selected Studies on the Late Antique Religious Mind

Edited by

Kirill Dmitriev Isabel Toral-Niehoff

gp 2017

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2017 by Gorgias Press LLC

All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. ‫ܚ‬

1

2017

ISBN 978-1-4632-0630-7

Library of Congress Cataloging-in-Publication Data A Cataloging-in-Publication record is available from the Library of Congress. Printed in the United States of America

TABLE OF CONTENTS Table of Contents ..................................................................................... v Introduction .............................................................................................. 1 I. RITUAL AND THEOLOGICAL ASPECTS.............................. 9 The Silence of the Gods. Some Observations on the Destruction of Pagan Temples, Shrines and Statues in the Late Antique East (from Constantine to Muḥammad) ........... 11 KONSTANTIN M. KLEIN

The Relationship Between Theological Teaching and Religious Practice by the East-Syrian Christians in Qatar (Sixth – Seventh Centuries) ........................................................................ 89 MARTIN TAMCKE

II. RE-INVESTIGATING THE TEXTS OF EARLY AND CLASSICAL ARABIC POETRY .......................................... 103 Glory and immortality: the motif of monumentum aere perennius by Samawʾal b. ʿĀdiyāʾ .............................................105 KIRILL DMITRIEV Tracing the Reception of the Protoevangelium of James in Late Antique Arabia: the Case of the Poetry of Umayya ibn Abī aṣ-Ṣalt and its Intersection with the Quran .............................123 CORNELIA HORN

Traders, Innkeepers and Cup-Bearers: Foreigners and People of the Book in Arabic Wine Poetry ...............................................147 BRUNO PAOLI

v

vi

RELIGIOUS CULTURE IN LATE ANTIQUE ARABIA

III. RELIGIOUS IDENTITIES IN THE CONTEXT OF POLITICAL AND SOCIAL HISTORY ................................. 163 From Mavia to al-Mundhir: Arab Christians and Arab Tribes in the Late Antique Roman East ...................................................165 GREG FISHER Inclusion and Seclusion in Late Antique and Early Medieval Settlement Strategies in Syria and Palestine ............................219 MATTIA GUIDETTI

IV. PERSPECTIVES ON THE RELIGIOUS HISTORY OF SOUTH ARABIA ............................................................. 247 Altsüdarabische Kulte im vorchristlichen Abessinien ....................249 FRANCIS BREYER Les évolutions du calendrier dans le royaume de Ḥimyar : quelques hypothèses....................................................................281 CHRISTIAN JULIEN ROBIN

INTRODUCTION KIRILL DMITRIEV & ISABEL TORAL-NIEHOFF

Peter Brown’s publication of The World of Late Antiquity in 1971 has been commonly regarded as a new starting point for the study of the era that has come to be known as Late Antiquity. Since then, the period is less apprehended under the perspective of decline and decadence, and designated by measurement against a classic norm set by Antiquity, as it was seen in the previous decades. Now it has become increasingly a temporal category having a proper constitution, defined by the binding elements of Empire and Monotheism. 1 This reconceptualization has had enormous consequences, besides that which has been called ‘an explosion of scholarship’. 2 First, the time-line has been gradually extended right into the ninth, even tenth century encompassing the Umayyad and partly the Abbasid Caliphates, that are hereby interpreted as forms of Late Antique Monotheistic Empires. 3 Second, the geographical focus has shifted See the recent surveys of the history of Late Antique Studies, cf. Cameron, (ed.) Late Antiquity on the Eve of Islam, xiii-lxviii; and especially regarding the debated inclusion/exclusion of the early Islamic period, in Fowden, Before and After Muhammad 18–48 and Al-Azmeh, The Emergence of Islam in Late Antiquity, 1–47. 2 Dijksrtra and Fischer, Inside and Out, 1. Cf. also Giardina, Esplosione di tardoantico. 3 This expansion has also been contested vehemently (cf. Giardina, Esplosione di tardoantico). The common view, however, tends now towards a time-line between 250–800 and so to include the Early Islamic period: cf., for instance, Bowersock, Brown and Grabar, eds., Interpreting late antiquity. Al-Azmeh, The Emergence of Islam In Late Antiquity has argued 1

1

2

KIRILL DMITRIEV & ISABEL TORAL-NIEHOFF

to include the areas located at the Eastern and Southern periphery of the Roman Empire, whose peoples regularly interacted with Roman-Greek culture and participated in the gradual conversion to monotheistic religions (e.g., the Sassanian Empire 4, but also the socalled ‘other’ Barbarians in Ethiopia, Abyssinia and Sudan 5; and those in Arabia 6). One of the many positive consequences of this change has been that it has helped to bridge the traditional sharp break between Classical and Islamic Studies; thus, it is now less exceptional that scholars of both fields meet, discuss, and publish together 7. Another important outcome is the modification and broadening of the scholarly view on the origins of Islam as religion, culture and society. Scholars of Ancient and Medieval History accept Islam more and more as an heir to Antiquity on a par with Byzantium and Latin Europe; and scholars in Islamic History and Culture have learnt to appreciate the Late Antique dimension of Islam. Having now left aside outdated Eurocentric arguments of “dependency” and “borrowing”, and with a stronger focus on interaction and cross-pollinations, they use the framework of Late Antiquity to extensively in favor of seeing Islamic civilisation as the ‘most successful crystallization’ of Late Antiquity (p.2 et passim). 4 For instance, see the recent sample volumes: Börm and Wiesehöfer (eds.), Commutatio et Contentio, and Wood (ed.), History and Identity, and recent publications Payne, A state of mixture, and Toral-Niehoff, al-Hira. 5 Dijksrtra and Fischer, Inside and Out. For the focus on the ‘other barbarians’ in recent scholarship cf. 1–9. 6 Fisher, Between empires and Fisher, (ed.), Arabs and empires before Islam. 7 See the publication series Studies in Late Antiquity and Early Islam (SLAEI) edited by Lawrence Conrad and Jens Scheiner, whose first volume appeared 1992 as the result of a pioneering workshop that was held by scholars from Byzantine and Islamic studies: Cameron and Conrad (eds.), The Byzantine and Early Islamic Near East. Many interdisciplinary workshops on this field have been held since then, and numerous joint publications might be mentioned here – see, for instance, Bowersock, Brown and Grabar (eds.), Interpreting Late Antiquity, including a chapter on “Islam” (by Hugh Kennedy, 219–237), and Wood, (ed.,) History and Identity.

INTRODUCTION

3

discuss their material under a fresh perspective, for example, by exploring the applicability of concepts like “Empire,” otherwise associated with Greco-Roman Antiquity, on the early Islamic Caliphate; 8 or by examining the creative re-use and re-framing of Biblical material and Judeo-Christian attitudes and religious practices in Islam. Recent publications like The Emergence of Islam in Late Antiquity, 9 Der Koran als Text der Spätantike, 10 The Bible in Arabic, 11 and alHira 12 clearly testify to this change in perception in Islamic Studies and Islamic History. In Religious Studies, the seminal The Making of the Abrahamic Religions in Late Antiquity 13 shows that Islam is increasingly understood in the framework of Late Antique religions; and the programmatic study Before and After Muhammad: The First Millennium Refocused by Garth Fowden, demonstrates that a scholar rooted in Classics might adopt a comprehensive view that includes the Classical, the Islamic and the Post-Classical World on equal terms. The present volume contributes to this dynamic field of research by focusing at the religious culture of the region, broadly defined as Arabia, that gave birth to Islam and Islamic culture. The volume adopts the framework of the new understanding of Late Antiquity with a strong emphasis on the emergence of Monotheism as its distinctive feature. The term ‘religious culture’ has often been used in relation to specific theological systems and ritual practices such as Judaism, Christianity, Manichaeism etc. The present volume provides it with a more comprehensive meaning addressing what one could call the Late Antique religious mind. It emphasizes the common religious attitudes, practices and religious world-views connected to Late Antiquity in its broader sense by including various monotheistic and other, often described as pagan, attitudes that existed and co-existed within the orbit of the emerging Islam. It also explores the many ways of their local manifestations, i.e. the Cf. e.g. the ERC-project in Hamburg ‘The Early Islamic Empire at Work,’ directed by Stefan Heidemann. 9 Al-Azmeh, The Emergence of Islam in Late Antiquity. 10 Neuwirth, Der Koran als Text der Spätantike. 11 Griffith, The Bible in Arabic. 12 Toral-Niehoff, al-Hira. 13 Stroumsa, The making of the Abrahamic religions in late antiquity. 8

4

KIRILL DMITRIEV & ISABEL TORAL-NIEHOFF

various local performances of the Late Antique global religious mind. This approach is grounded in the awareness that Islam developed through a dynamic process of negotiating a new religious culture with all of them, as well as in the fact that the differences between these religious systems and communities in the Late Antique Near East did not lead to the constitution of isolated entities which can be clearly separated from each other. This also applies to the regions of the Eastern Mediterranean and the Arabian Peninsula, which should be regarded not as a periphery of the Late Antique world, but on the contrary as the most loaded hubs at the crossroads of civilisations of the Late Antique West and East. 14 The volume takes a rather broad anthropological and cultural approach to religion, here understood as a cultural semiotic system, ethos and world-view, and assumes that these are not only conveyed by texts, but also mediated by reported religious and societal practices. Thus, it follows the well-known proposition by Clifford Geertz who defines religion as ‘(1) a system of symbols (2) which acts to establish powerful, pervasive and long-lasting moods and motivations in men (3) by formulating conceptions of a general order of existence and (4) surrounding these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic.’ 15 Therefore, the approach adopted in the volume does not restrict its scope to such phenomena as theology, religious scriptures and rituals, but extends it to include other, less overtly religious aspects, that, nevertheless, help better understand how the late antique mind approached religion and how this affected early Islam. Such a broad approach to religious culture in Arabia is quite ambitious, and obviously, it cannot be addressed comprehensively in a small volume. Thus, we have opted to leave aside the most prolific field of Quranic studies, since numerous recent publications and research projects have already considerably extended this central debate. Other gaps, for instance, the rather minor presence In contrast to al-Azmeh, The Emergence of Islam in Late Antiquity, who emphasizes the particular backwardness of Western Arabia. For the role of the Arabs at the crossroads of Late Antique Civilizations cf. ToralNiehoff, al-Hira. 15 Geertz, Religion 90. 14

INTRODUCTION

5

of Art History and Archaeology, which would have added a very important perspective, resulted from the philological and historical background of the scholars involved in the enterprise. Therefore, the following sample of studies aims rather at a preliminary, even patchy insight into the potential of this perspective for the study of Early Islam, without claiming to be exhaustive, but rather presenting illustrative case-studies. The idea for this volume originated at the international research conference ‘Religious Culture in Late Antique Arabia,’ which took place in Berlin, Germany in 2009, and was generously supported by the Fritz Thyssen Foundation. Some contributions (Breyer, Dmitriev, Horn, Klein, Paoli, Tamcke) were presented at the conference, others have been added later (Fisher, Guidetti, Robin). The first section (Klein, Tamcke) focuses more on the ritual and theological aspects of Late Antique Arabia as viewed from different angles. Konstantin Klein addresses in his contribution the topic of the destruction of idols and the conversion of pagan temples into sites of monotheistic worship from the comparative perspective of Late Antique history and draws inspiring parallels between these developments in Late Rome and in Arabia on the eve of Islam. Martin Tamcke dedicates his study to the interaction between theology and religious practices in Qatar, where a Christian community with an ecclesiastical structure flourished on the soil of the Arabian Peninsula well into the first two Islamic centuries. The second section (Dmitriev, Horn, Paoli) encompasses three articles that study the religious mind of the Late Antique Arabs by investigating the potential of the main and most debated source, namely, early and classical Arabic poetry. Problems of authenticity and the overt lack of explicit religious connotations hamper the use of this material in the study of religious history; however, a careful reading of the poetic sources can reveal their relevance for understanding the religious views of the people who populated the Arabian Peninsula on the eve of Islam. Kirill Dmitriev explores the notion of perennial glory in the poetry of Samawʾal b. ʿĀdiyāʾ, and draws parallels with Biblical conceptions of the time, arguing that the familiarity with an individual approach to immortality, as evidenced in poetry, opened the path to the Quranic message. Cornelia Horn investigates the poetry of Umayya ibn Abī aṣ-Ṣalt and its intertextual connections to the Quran and Christian

6

KIRILL DMITRIEV & ISABEL TORAL-NIEHOFF

apokrypha. She proposes a consideration of the parallels between the Quran and Umayya not in terms of dependency, but as reflecting the interaction with a third tradition, namely, the retelling of material related to Christian canonical and apocryphal stories. In addition, Bruno Paoli investigates the cultural meaning of the poetic topos of the dhimmī cupbearer in classical Arabic poetry, which, according to his argumentation, reflects the tradition of nonArabic, mainly Jewish wine-traders and waiters that remained unchanged after the rise of Islam, independently of religious precepts. The third section (Fisher, Guidetti) approaches the political aspects of conversion and religion, and their impact on the society and settlement patterns of the Arab groups that inhabited the periphery of the Roman and Sasanian Empires, and the Early Caliphate. Greg Fisher’s article draws a wide panorama of the Arab tribes dwelling in Greater Syria in Roman times, and how they interacted with Rome, focusing on the instrumentalisation of Christian religion, conversion, and sectarian alliances as a means of policy and dominance. Mattia Guidetti investigates the settlement patterns in Late Antique and Early Islamic Syria and the incidence of conversion to Islam, i.e. the dynamics between Islamisation and Arabisation that marked the social and religious history of the area. The concluding section (Breyer, Robin) contains two articles focusing on the aspects of the religious history of South Arabia. Francis Breyer approaches the question from outside, basing his study on the Abyssinian material from the Aksumite period, since it provides unique evidence for the pagan South Arabian religious culture. Christian Robin presents a meticulous reconstruction of the calendar systems in South Arabia as evidenced by inscriptions, and examines the effect of the shift towards monotheistic religions. As he shows, the change had as a consequence the gradual evolution of the pagan system towards a reformed calendar that localised the Julian calendar, thus connecting the acceptance of monotheistic religions, namely the South Arabian variant of Judaism (judéomonotheisme), and then Christianity, with the introduction of Late Antique temporalities. *** The present volume not only illustrates the variety of the religious culture in Late Antique Arabia, including the variegated local “paganisms”, but it also highlights some common lines that are related to the main connecting and connected features that define the pe-

INTRODUCTION

7

riod of Late Antiquity: Empire and Monotheism. On the one hand, we can observe the emergence of several local variants of monotheistic religions, communities, practices and organizations, and show how they operated within a framework of various polytheistic traditions and institutions that reacted to this monotheistic challenge. On the other hand, we see the strong link of these monotheisms with political power, by serving simultaneously as an ideological weapon and as a symbol of political alliances, and frequently even legitimizing imperial claims. To a certain degree, these new monotheisms were all rooted in a combination of Biblical traditions with Roman Imperialism; likewise, all of them share notions of community, scriptualism and immortality, just to mention a few striking similarities. The emerging Islamic religious culture was part of this thriving religious context, and it grew in an intensive interaction with those diverse yet cognate communities and world-views. A comprehensive understanding of the religious culture in late antique Arabia is essential for the study of the Late Antique religious mind and the origins of Islam as one of its outstanding manifestations.

BIBLIOGRAPHY Börm, H. and J. Wiesehöfer (eds.), Commutatio et Contentio. Studies in Late Roman, Sasanian, and Early Islamic Near East, Düsseldorf 2010. Bowersock, G., P. Brown and O. Grabar (eds.), Interpreting Late Antiquity, Harvard 1999. Cameron, A. and L. Conrad (eds.), The Byzantine and Early Islamic Near East. Problems in the Literary Source Material, Princeton 1992. Cameron, A. (ed.), Late Antiquity on the Eve of Islam. Farham 2013. Giardina, A., Esplosione di tardoantico, in Studi Storici 40, 157–180 (1999). Geertz, C., Religion as a Cultural System, in The Interpretation of Cultures. Selected Essays, New York 1973, 87–125. Griffith, S.H., The Bible in Arabic: The Scriptures of the ‘People of the Book’ in the Language of Islam, Princeton 2013. Fisher, G. Between empires: Arabs, Romans, and Sasanians in Late Antiquity, Oxford 2013.

8

KIRILL DMITRIEV & ISABEL TORAL-NIEHOFF

Fisher, G. (ed.) Arabs and empires before Islam, Oxford 2015. Dijkstra, J. and G. Fisher (eds.), Inside and Out. Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, Leuven and Paris 2014. Fowden, G., Before and After Muhammad: The First Millennium Refocused, Princeton 2013. Neuwirth, A., Der Koran als Text der Spätantike. Ein europäischer Zugang, Berlin 2010. Payne, R.E., A state of mixture: Christians, Zoroastrians, and Iranian political culture in late Antiquity, Oakland 2015. Stroumsa, G. The Making of the Abrahamic Religions in Late Antiquity, Oxford 2015. Toral-Niehoff, I., Al-Hira. Eine arabische Kulturmetropole im spätantiken Kontext, Leiden 2014. Wood, Ph. (ed.), History and Identity in the Late Antique Near East, Oxford 2013.

I. RITUAL AND THEOLOGICAL ASPECTS

9

THE SILENCE OF THE GODS. SOME OBSERVATIONS ON THE DESTRUCTION OF PAGAN TEMPLES, SHRINES AND STATUES IN THE LATE ANTIQUE EAST (FROM CONSTANTINE TO MUḤAMMAD) KONSTANTIN M. KLEIN

In his account De defectu oraculorum, the Greek historian Plutarch reports that once, in the days when Tiberius ruled the Roman world, an Egyptian helmsman, Thamus, steered his ship towards Italy. When passing by the Ionian Islands, a voice hailed across the waters asking him to proclaim that Πὰν ὁ μέγας τέϑνηκε – the great Pan has died. Thamus complied, and afterwards all people on deck could hear sounds of wailing coming from the shores. When the ship reached Italy, news began to spread, and even the Emperor himself had the matter investigated. A multitude of wise men present at his court speculated that the incident was real and that the immortal gods were subject to death. 1 Neither Tiberius (nor Plutarch, De defectu oraculorum 17. It has been suggested that the passengers on the ship might have misinterpreted the exclamation “the allgreat Tammuz has died” (ϴαμοῦς [ὁ] πανμέγας τέϑνηκε) for “the great Pan is dead” (ϴαμοῦς, Πὰν ὁ μέγας τέϑνηκε); cf. Reinach, La mort 5–19 and Tesslar, The death of Pan 180–3. However, this and other modern attempts of explanations (e.g. “Zan” as an older form of Zeus instead of “Pan”) have been proved linguistically impossible and unfulfilling, cf. Nock, Ὁ μέγας 164–5 and Borgeaud, The death 271–3 and 278–9. It is 1

11

12

KONSTANTIN M. KLEIN

Plutarch) paid much attention to the fact that at the same time a young man preaching in Galilee was becoming famous for driving out demons. 2 After entering Jerusalem, however, he was arrested by the Sanhedrin for blasphemy against the Mosaic Law, and was eventually tried for sedition against Rome by Pontius Pilate, the Roman prefect of Judaea, who ordered the convict’s crucifixion. For Christian readers of Plutarch’s De defectu oraculorum, the parallels between the death of the god Pan and Jesus Christ’s miraculous deeds on earth and His victory over death were obvious. Eusebius of Caesarea was the first author who explicitly linked the two events in his early-fourth century Praeparatio evangelica: But it is important to observe the time at which he [scil. Plutarch] says that the death of the daemon took place. For it was the time of Tiberius, in which our Saviour, making His sojourn

important to note that there is a distinct difference between the “death of Pan” described in Plutarch’s account and common annual dying-andrising cycles of gods known in most pantheons, be it the Akkadian deity Tammuz (or Sumerian Dumuzid, Canaanite and Greek Adonis, Etruscan Atunis, cf. Müller, Daphnis 29–36), Osiris and Horus in Pharaonic Egypt, the Mesoamerican god Quetzalcoatl, Gullveig, Heiðr, Odin and Baldur in Norse mythology or the shamanistic war-hero Lemminkäinen in Finnish oral poetry. On Near Eastern deities, cf. Mettinger, The riddle of resurrection, esp. 15–39 for a re-evaluation of the scholarly debate on dying-and-rising gods since the publication of James Frazer’s, Adonis. However, for a tomb of Zeus himself on Crete, known in classical antiquity, cf. Lucian’s mocking commentary in De sacrificiis 10: οἱ δ’ αὖ Κρῆτες οὐ γενέσϑαι παρ’ αὑτοῖς οὐδὲ τραϕῆναι μόνον τὸν Δία λέγουσιν, ἀλλὰ καὶ τάϕον αὐτοῦ δεικνύσιν· καὶ ἡμεῖς ἄρα τοσοῦτον ἠπατήμεϑα χρόνον οἰόμενοι τὸν Δία βροντᾶν τε καὶ ὕειν καὶ τὰ ἄλλα πάντα ἐπιτελεῖν, ὁ δὲ ἐλελήϑει πάλαι τεϑνεὼς παρὰ Κρησὶ τεϑαμμένος. (As for the Cretans, they not only say that Zeus was born and brought up among them, but even point out his tomb. We were mistaken all this while, then, in thinking that thunder and rain and everything else comes from Zeus; if we had but known it, he has been dead and buried in Crete this long time.), cf. Müller, Daphnis 26 with n. 7. 2 Cf. esp. Luke 8:26–33, Mark 5:9 and, in a slightly different version, Matthew 8:28–34.

THE SILENCE OF THE GODS

13

among men, is recorded to have been ridding human life from daemons of every kind: so that there were some of them now kneeling before Him and beseeching Him not to deliver them over to the Tartarus that awaited them. 3

According to Eusebius, the coming of a more powerful God signified the eternal lapse into silence of another, older deity. In a similar fashion, though much later, Ṭabarī recorded that “when Jesus was born, every idol upon the earth fell on its face” 4 – just as the same author collected similar miraculous signs that accompanied the birth of Muḥammad. If we assume with Plutarch and Eusebius that gods can die, it is worth asking whether they can also be killed. 5 And indeed, the idea of exterminating gods (or demons – Eusebius, Praeparatio evangelica 5,17,13: ἐπιτηρῆσαι δ’ ἄξιον τὸν καιρὸν ἐν ᾧ ϕησι τὸν ϑάνατον γεγονέναι τοῦ δαίμονος. οὗτος δὲ ἦν ὁ κατὰ Τιβέριον, καϑ’ ὅν ὁ ἡμέτερος σωτὴρ τὰς σὺν ἀνϑρώποις ποιούμενος διατριβὰς πᾶν γένος δαιμόνων ἐξελαύνειν τοῦ τῶν ἀνϑρώπων ἀναγέγραπται βίου· ὥστε ἤδη τινὰς τῶν δαιμόνων γονυπετεῖν αὐτὸν καὶ ἱκετεύειν μὴ τῷ περιμένοντι αὐτοὺς Ταρτάρῳ παραδοῦναι. Cf. the allusion to Luke 8:30–3. On aspects of the contemporaneity of the events and potential consequences of a death of Pan for Tiberius, cf. Borgeaud, The death 261–5. 4 Cf. also Taʾrīkh 734–5: ‫فلما ولد عيسى لم يبق في ا��رض صنم يعبد من دون الله ا�� ّ اصبح‬ ُ ّ َ ُ ‫ساقطا لوجهه‬. The narration is based on a transmission-chain going back to Ibn Masʿūd. Similar omens occurred, according to Ṭabarī and his sources, when Muḥammad was born, cf. Taʾrīkh 981, discussed below. 5 However, already in the late second century, Tertullian’s famous aphorism had proclaimed that the gods once were mere men (deos [...] homines fuisse). On the downside, this meant that pagan gods could not die or be killed, since they were already dead for aeons; cf. Tertullian, Apologeticum 10, 3 with Stroumsa, Barbarian philosophy 102–3. Tertullian argued that the pagan deities were not real, did never exist and had no power; he employed myths concerning Saturn to exemplify that the latter was once a mere man that only later became venerated as a god. On similar thoughts in Graeco-Roman religious philosophy, cf. Prinz, Die Kirche 285–6. In his treatise De idololatria (on idolatry) 3, 1, Tertullian argues that once upon a time there existed no idols (idolum aliquamdice retro non erat), and that it was the Devil who brought into the world the makers of statues, portraits, 3

14

KONSTANTIN M. KLEIN

non-human creatures common to all monotheistic religions) in order to establish a new cult at the place of their former veneration is an ancient concept. 6 This contribution aims to investigate whether these narrative traditions, which were highly popular in Late Antique Christian texts, 7 can also be found in Islamic literature describing the destruction of sanctuaries and idols during the lifetime of Muḥammad. The following pages aim to familiarize scholars of early Islam with the Late Antique material – but also to bring Late Antique historians into contact with the religious life of Late Antique Arabia. This contribution therefore has the character of a survey of the sources, and necessarily there will be a lengthy – though still not exhaustive – historical sketch of late Roman legislation concerning paganism. Moreover, certain episodes of paganChristian conflict will be highlighted, as they will serve as points of comparison with the early Islamic material. It is important to note that for both traditions the sources used for this study will be considered primarily as literary, and not necessarily as historical accounts. This is true not only for texts from the Muslim tradition, mostly written down in the early ninth century and re-imagining historical events of the time of Muḥammad, but also for the majority of the Christian texts, mostly hagiography that features protagonists that were sometimes either completely fictional or whose deeds were exaggerated to a very high extent. This turn to literary motifs is no excuse; the material with its contradictive accounts simply does not allow for a historical reconstruction of the course and every kind of representation (3, 2: at ubi artifices statuarum et imaginum et omnis generis simulacrorum diabolus saeculo intulit); cf. Romans 1:23, see also Vermander, La polémique 111–112 and van Winden, Idolum 110–114. A similar explanation for the creation of (false) gods is offered in Ibn alKalbī, Kitāb al-Aṣnām 31, 13–32, 4 which I will discuss in more detail elsewhere. 6 Saradi, Christian attitudes 56. 7 There is no space here to discuss at length the Jewish material, first and foremost Mishna ʿAvoda Sara, for this text see now Binder, Tertullian. Cf. also the seminal study by Halbertal and Margalit, Idolatry.

THE SILENCE OF THE GODS

15

of events and can therefore not be taken at face value. 8 Within the past fifty years, the term ‘hagiography’ has overcome much of its rather pejorative connotation of being uncritical works of authors who were highly reverential towards their (mostly biographical) objects of writing. As the boundaries between Late Antique hagiography and historiography were often fluid, furthermore, as many historians incorporated hagiographical material in their accounts, they both will be referred to as ‘hagiography’ in this contribution. It may seem unusual to apply the same term to some sub-genres of early Islamic historiography, however, the parallels in style and narrative technique are striking as is their focus on individuals, especially in the case of the Sīra. 9 A second caveat is necessary regarding the term ‘paganism’. While it is commonly accepted that this term is used in modern studies not in a pejorative way, the problem consists that there never was a uniform pagan party or any institutionalised paganism in Late Antiquity. The modern term thus inevitably groups together a variety of traditions, beliefs and schools. While this is long accepted for the adversaries of early Christianity, 10 the same problem exists for the Islamic sources: even though the subject has been investigated in a specific monograph quite recently, we still do not know who exactly the early Islamic ‘pagans’ – mushrikūn – were. 11 In contrast to the earlier evidence concerning the pagan-Christian conflict for which the majority of the sources Cf. Bagnall, Models 26 and Papaconstantinou, Là où le péché 244– 55 who points to the element of exaggeration in Christian sources concerning the destruction of temples. A good example for the contradictoriness of the sources is Eusebius, Vita Constantini 2, 45, 1, alluding to a law issued by Constantine I forbidding sacrifices, in contrast to Libanius, Oratio 30, 37, claiming that the same emperor did not proceed against sacrifices (οὐκ ἐπὶ τὰς ϑυσίας προῆλϑε); cf. Wallraff, Die antipaganen Maßnahmen 9–10. 9 Cf. Khalidi, Images of Muhammad 15–16 and 57–64. For a similar approach towards the Christian texts, cf. Gotter, Rechtgläubige 43–4. 10 Cf. Wallraff, Die antipaganen Maßnahmen 9; see also the edited volume by Lavan and Mulryan, The archaeology on various archaeological examples from both the Late Antique West and East. 11 Cf. Hawting, The idea of idolatry, and Bowersock, Empires 66–7. 8

16

KONSTANTIN M. KLEIN

represent Christian points of view, but for which also a certain amount of pagan voices were transmitted, the situation for early Islam is even more complicated since all our evidence derives from only one perspective as no direct sources from the other side survived.

I

LATE ROMAN TRADITIONS

I.1 The Constantinian paradigm – an archetype of destruction? In his biography written by Eusebius, the Vita Constantini, Constantine the Great appears as the paragon of fighters against paganism. Eusebius’ text contains several examples of the emperor dealing with pagan sanctuaries and statuary, the most prominent of which being Constantine’s order to have a Roman temple removed, its stones meticulously carried off and the soil purified to create the space on which the Church of the Holy Sepulchre in Jerusalem would be built: [...] but the Emperor gave further orders that all the rubble of stones and timbers from the demolitions should be taken and dumped a long way from the site. [...] But not even this progress was by itself enough, but under divine inspiration once more the Emperor gave instructions that the site should be excavated to a great depth and the pavement should be carried away with the rubble a long distance outside, because it was stained with demonic bloodshed. 12

All this according to the testimony of Eusebius, though excavations revealed that in Jerusalem – like everywhere else – spolia of the temple were reused at a large scale for the construction of the Eusebius, Vita Constantini 3, 27: [...] ἀλλὰ πάλιν βασιλεὺς αἴρεσϑαι καὶ πορρωτάτω τῆς χώρας ἀπορρίπτεσϑαι τῶν καϑαιρουμένων τὴν ἐν λίϑοις καὶ ξύλοις ὕλην προστάττει. [...] ἀλλ’ οὐδ’ ἐπὶ τοῦτο μόνον προελϑεῖν ἀπήρκει, πάλιν δ’ ἐπιϑειάσας βασιλεὺς τοὔδαϕος αὐτό, πολὺ τοῦ χώρου βάϑος ἀνορύξαντας, αὐτῷ χοῒ πόρρω που καὶ ἐξωτάτω λύϑροις ἅτε δαιμονικοῖς ἐρρυπωμένον ἐκϕορεῖσϑαι παρακελεύεται. 12

THE SILENCE OF THE GODS

17

new church. 13 Locating the spot where Christ had been buried proved not too difficult as the Romans had marked it by building over the holy place: “Once upon a time, wicked men (or rather the whole tribe of demons through them)” had constructed a “terrible and truly genuine tomb, one for souls, for dead idols, and built a gloomy sanctuary to the impure demon of Aphrodite; then they offered foul sacrifices there upon defiled and polluted altars.” 14 Eusebius, who amply used Constantinian letters and documents for his account, styles the emperor’s actions against pagan cults as an all-out quest against the old gods. However, by casting a close look at the imperial letters and orders Eusebius cites, it becomes indeed clear that Constantine tried to abolish blood sacrifice. He also legislated against temple prostitution as well as religious castration, and – perhaps connected to these – against the cult of Aphrodite with its notions of unchastity and magic. 15 Only few of the cases of destruction mentioned by Eusebius, however, appear to have been empire-wide phenomena. All together, he only mentions six temples that had been destroyed and/or replaced by churches. Eusebius’ suggestive use of wording misled not only modern scholarship into believing that there was a general pattern of architectural transformation ‘from temple to church’ but also convinced a subsequent generation of Late Antique authors into thinking that they witCf. Corbo, Il Santo Sepolcro di Gerusalemme iii, esp. visible on photographs nos 14, 16 and 17. I wish to thank Ute Verstegen (Marburg) for pointing this out to me. 14 Eusebius, Vita Constantini 26, 1.3: ἄνδρες [...] ποτε δυσσεβεῖς, μᾶλλον δὲ πᾶν τὸ δαιμόνων διὰ τούτων γένος. [...] τῆς γῆς ὕπερϑε δεινὸν ὡς ἀληϑῶς ταϕεῶνα ψυχῶν ἐπισκευάζουσι νεκρῶν εἰδώλων, σκότιον Ἀϕροδίτης ἀκολάστῳ δαίμονι μυχὸν οἰκοδομησάμενοι, κἄπειτα μυσαρὰς ἐνταυϑοῖ ϑυσίας ἐπὶ βεβήλων καὶ ἐναγῶν βωμῶν ἐπισπένδοντες. While Eusebius mentioned a temple of Aphrodite, Jerome states in Letter 58, 3 that a temple of Iupiter preceeded the Constantinian basilica in Jerusalem. Of the vast amount of literature, see for instance, Patrich, The early Church 102–5. However, for the purpose of this paper mainly concerned with literary depictions these questions are of secondary importance only. 15 Cf. Gotter, Rechtgläubige 48 and Wallraff, Die antipaganen Maßnahmen 10–15. 13

18

KONSTANTIN M. KLEIN

nessed a gradual development spanning the period from the first Christian emperor to their present times. For instance, Rufinus, who continued Eusebius’ Church History wrote that “the cult of idols, which on the initiative of Constantine and thereafter had begun to be neglected and destroyed, collapsed in his [scil. Theodosius’] reign.” 16 The Constantinian destructions had a deictic function by projecting the emperor’s own conversion into an empire-wide moral reformation and conversion, and they served as a litmus test for future emperors whose deeds were measured against those of Constantine. 17 Eusebius’ rhetorical case can arguably best be seen by how he described the overturning of pagan cult in the empire following his description of the six singular cases: In other provinces whole crowds changed sides and came to the knowledge of the Saviour; in every territory and city they got rid of the things they formerly held sacred, made of all kinds of wood, as if they were nothing. Temples and built-up precincts they demolished without orders from anyone, and building churches on their foundations they changed from their former error. 18

The fact that Eusebius wrote that (some) temples were voluntarily destroyed only suggests that there were no specific imperial orders to do so. 19

Rufinus, Historia ecclesiastica 11, 19: idolorum cultus, qui Constantini institutione et deinceps neglegi et destrui coeptus fuerat, eodem imperante conlapsus est. 17 Eusebius, Vita Constantini 3, 55, 57–8 with Gotter, Rechtgläubige 48 and 50–1. 18 Eusebius, Vita Constantini 4, 39, 2: καὶ ἐϕ’ ἑτέρων δ’ ἐπαρχιῶν αὐτόμολοι τῇ σωτηρίῳ προσιόντες γνώσει ἀϑρόοι κατὰ χώρας καὶ κατὰ πόλεις τὰ μὲν πρότερα αὐτοῖς ἱερὰ νενομισμένα, ἐν ὕλῃ παντοίων ὄντα ξοάνων, ὡσανεὶ τὸ μηϑὲν ὄντα ἠϕάνιζον, ναούς τ’ αὐτῶν καὶ τεμένη εἰς ὕψος ἠρμένα μηδενὸς ἐπικελεύοντος καϑῄρουν, ἐκκλησίας δ’ ἐκ ϑεμελίων ἀνιστῶντες τῆς προτέρας ἀντικατηλλάττοντο πλάνης. 19 Cf. Wallraff, Die antipaganen Maßnahmen 8. 16

THE SILENCE OF THE GODS

19

I.2 Violence and law in Late Antiquity Parallel to the gradual advancement of Christianity during the fourth century, culminating with its establishment as state religion in February 380 in the reign of Theodosius I, we can detect a growing amount of violence and intolerance towards paganism and heresies. 20 While a law issued by Constantius II, demanding an end to the “madness of sacrifices (sacrificiorum insania)”, left still some room for interpretation, 21 the general prohibition of public sacrifices confirmed by Theodosius I in the 380s seems to have constituted a rather vehement intervention into the religions life of fourth century pagans. 22 Between 389 and 391 a series of laws was issued, which almost completely forbade visits to temples as well as pagan celebration on feast days. 23 This had visible results on the big cities of the Late Roman Empire: In Rome the sacred fire of Vesta was no longer maintained and the Vestal Virgins were dismissed, whereas the New Rome on the Bosporus witnessed the closing of its temples on the acropolis and their subsequent conversion to Saradi, Christian attitudes 47–9 and Cameron, The violence 113– 14. On how strikingly little importance was attached to Theodosius’ edict cunctos populos by later church historians, cf. Errington, Church and State 21–72. 21 Codex Theodosianus 16, 10, 2 (a. 341). On the marginalisation of sacrifices from the times of Constantine I onwards, cf. Belayche, Partager 463–4; on the problems arising from the (perhaps intended?) vague wording of Codex Theodosianus 16, 10, 2, cf. McLynn, Pagans 575–6. On the law and the problems arising from its contextualisation, cf. Hahn, Gesetze 205– 6: the abbreviated form as given in the Codex suggests a distinct generality that the original law might not have intended. On the question whether Constantine I issued a similar law, alluded to in Eusebius, Vita Constantini 2, 45, 1, cf. Wallraff, Die antipaganen Maßnahmen 9–10. 22 Frank Trombley argues that the repetitive decrees against sacrifice issued by the sons of Constantine demonstrate that they were ineffective, cf. Trombley, Hellenic religion i, 1–97; however, the sources do not support this image of flourishing sacrificial cults until the times of Theodosius I. Cf. also Cameron, The last pagans 65–7 who suspects that in both West and East by the 380s public rituals no longer included animal sacrifice. 23 Codex Theodosianus 16, 10, 10 (a. 391). 20

20

KONSTANTIN M. KLEIN

new forms of use: the temple of Artemis was turned into a gaming room for dice players and the temple of Aphrodite into a carriagehouse surrounded by a hostel for the city’s penniless prostitutes. 24 This is the official perspective as reflected in legislation and imperial orders; we will return later to the question of what impact these laws actually had and how their content was received by the population. The literary sources penned by Christian authors, in contrast, downright celebrate the growing intolerance against pagan places of worship: Quickly formed (and sometimes similarly fast dissolving) groups of monks mixed with socially disadvantaged seasonal farmhands grouped together and wandered through the Late Antique East looting cities and the countryside. The hagiographer of one of these charismatic monks reports – not without a certain pride – that even predominantly Christian cities closed their gates to the band of Alexander Akoimetos, afraid of an escalation of violence and too many hungry mouths to feed. 25 In the middle of the fifth century, the group of Barṣawmā directed their anger and violence quite randomly towards all sorts of potentially destructible structures be it temples, Jewish synagogues or Roman administrative buildings. 26 For Barṣawmā these buildings were all On the fate of the Vestal Virgins cf. Conti, Tra integrazione 209– 22 with a prosopography of the last Vestals (214–20), on Theodosius’ conversions of the Constantinopolitan temples cf. Malalas, Chronographia 13, 39. Also in Alexandria, the temple of Tyche was turned into a tavern, cf. Palladas in Anthologia Graeca 9, 180–3 and 10, 90 with Hahn, Gewalt 95. 25 Cf. the seminal study on Alexander Akoimetos by Caner, Wandering, esp. 157 and 168; on the transitory nature of monastic bands looting the countryside, Brown, The rise 83–4 and Trombley, Hellenic religion ii, 143–5. 26 Acerbi, Il potere 301–3 and Acerbi, Terror y violencia 277–90; as well as Bowersock, Polytheism 4; cf. also the summaries of the Syriac text given by Nau, Résumé. A decree from the year 423 explicitly protected Jews and pagans who were living “quietly and attempting nothing disorderly or contrary to law.” Christians (this would apply to Barṣawmā and his followers) who proved violent against them, were compelled to restore the triple or quadruple of the property which they took away; cf. Codex Theodosianus 16, 10, 24 (a. 423); cf. Cameron, The last pagans 72, proposing a 24

THE SILENCE OF THE GODS

21

pagan. The basic statement of these accounts is that, without either official control or noteworthy administrative objection, individual groups centered on charismatic holy men took over the state’s monopoly on legitimate violence and used it according to their own interpretation. However, it is important to note that what is described in these hagiographical descriptions is far from being an officially adopted policy of the Christian State or of the Church, and should not be taken at face value; there never was any systematic state-sponsored destruction of pagan sanctuaries. 27 To some extent the fourth-century legislation went along with a seemingly declining fondness of sacrifice at least among the more intellectual circles of Late Antique paganism. Already in the second century Lucian of Samosata had stressed the “silliness” (ἀβελτερία) of sacrifices, 28 and Sallustius, a teacher of Julian the Apostate, explained that although temples may be likenesses of the celestial sphere, the gods themselves gained nothing from the existence of their shrines. However, these splendid buildings still might help the generally low level of law enforcement in both cases, destruction of pagan sites as well as punishment of those whose anti-pagan campaigns overshoot the target; cf. also Hunt, Christianising the Roman Empire, 157–8. 27 Cf. Bagnall, Models 26; Dally, Pflege 98; Papaconstantinou, Là où le péché 244–55 and Saradi, Christian attitudes 47. 28 Lucian, De sacrificiis i: Ἅ μὲν γὰρ ἐν ταῖς ϑυσίαις οἱ μάταιοι πράττουσι καὶ ταῖς ἑορταῖς καὶ προσόδοις τῶν ϑεῶν καὶ ἅ αἰτοῦσι καὶ ἅ εὔχονται καὶ ἅ γινώσκουσι περὶ αὐτῶν, οὐκ οἶδα εἴ τις οὕτως κατηϕής ἐστι καὶ λελυπημένος ὅστις οὐ γελάσεται τὴν ἀβελτερίαν ἐπιβλέψας τῶν δρωμένων (In view of what the dolts do at their sacrifices and their feasts and processions in honour of the gods, what they pray for and vow, and what opinions they hold about the gods, I doubt if anyone is so gloomy and woe-begone that he will not laugh to see the idiocy of their actions). For an extensive commentary and interpretation of Lucian’s short account De sacrificiis, cf. Berdozzo, Götter 71–94, esp. 84–6 (on the passage quoted here). On the general debate concerning the necessity of sacrifices in Greek philosophy in Roman imperial times, cf. Belayche, Sacrifice 101– 26, Stroumsa, La fin du sacrifice 108–111 and Auffarth, Götterbilder 311– 12.

22

KONSTANTIN M. KLEIN

mortals to search for (and perhaps find) union with the gods. 29 We need to be careful, however, not to equate these intellectual statements about cult practice with a general decline among the population, 30 not least as the compilers of the Theodosian Code grouped regulations on pagans, sacrifices and temples in one section, 31 which makes these three very different domains appear intrinsically interrelated to a modern audience. 32 Moreover, it is important to note that not all legislative actions against paganism were automatically directed against temples and statues too: The overall image that derives from the legislation is that the closing of temples was Sallustius, De deis et mundo 15: οἱ μὲν ναοὶ τὸν οὐρανὸν [...] μιμοῦνται. ([T]emples are a copy of heaven). Cf. also Belayche, Partager 474–6, and Stroumsa, La fin du sacrifice 113. 30 A similar scenario of decline has often been proposed for Late Antique Arabia on the eve of Islam. Michael Lecker, however, could show that idolatry in fact did not show signs of weakening e.g. in Medina and Yamāma, cf. Lecker, Idol worship 340–1, Lecker, ʿAvodat 258 and Lecker, forthcoming 1, 24 and 36–7. 31 In chronological order, Codex Theodosianus 16, 10, 2 (a. 341), 16, 10, 4 (a. 346), 16, 10, 5 (a. 353), 16, 10, 6 (a. 356), 16, 10, 7 (a. 381), 16, 10, 9 (a. 385), 16, 10, 11 (a. 391), 16, 10, 12 (a. 392), 16, 10, 13 (a. 395), 16, 10, 17 (a. 399), 16, 10, 18 (a. 399), 16, 10, 25 (a. 435) speak about sacrifices, whereas only 16, 10, 4 (a. 364) and 16, 10, 16 (a. 399) discuss the closing of temples – to these two decrees can be added 16, 10, 12, 2 (a. 392) and 16, 10, 20, 1–2 (a. 415) on the confiscation of temple property. Finally 16, 10, 6 (356) and 16, 10, 10 (a. 391) discuss the worship of idols, while 16, 10, 19, 1 (a. 408) urges the destruction of idols. 32 McLynn, Pagans 574. However, already Pliny (Epistula 10, 96, 10) speaks of templa desolata, cf. Saradi, Christian attitudes 49. Archaeological evidence shows that major temples in Egypt were re-used for administrative purposes as early as in the third century, and Roger Bagnall has put forward the proposition of an early decline of temples in Egypt: “there is no way of evading the fact that major temples like the Triphieion across from Panopolis or the temple of Luxor were taken over by the Roman government for official or military purposes,” Bagnall, Models 33, cf. also Cameron, The last pagans 64–5; Dally, Pflege 98 and McLynn, Pagans 575– 6. 29

THE SILENCE OF THE GODS

23

important, not their destruction. 33 In some cases, the old and august temples – perhaps the most prominent buildings within the cities of the Roman Empire – carried a notion of identity for the cities’ populations who cared for their preservation. They may not have been used for religious services anymore, however, the monuments with their architectural grandeur provided the inhabitants with an environment that separated them from the barbarian world around them. 34 A law issued in 399 has often been interpreted as a safeguarding provision for the temples; in it, Arcadius decreed that the ornaments of public works shall be preserved (publicorum operum ornamenta servari). 35 Another decree from the same year proclaimed the marking of temples with the sign of the cross as an easy way to avoid their destruction, consequently, a temple which was empty of ‘illicit things’ was safe. 36 The content of this law was reinforced in 435 – one of the last texts in the Theodosian Code – when it was stated that all pagan sanctuaries still in use “shall be destroyed by the command of the magistrates” – but again adding that a purification by the erection of the sign of the venerable Christian religion, i.e. the cross, was equally sufficient. 37 Cf. Hahn, Gesetze 208. Cf. the examples in Meier, Alte Tempel 370 and Kunderewicz, La protection 152–3. 35 Codex Theodosianus 16, 10, 15 (a. 399); cf. Kunderewicz, La protection 146 and Dally, Pflege 99 for further examples. On the preservation of statues as objects of art, see below. 36 Codex Theodosianus 16, 10, 18 (a. 399). Roughly at the same time, however, imperial legislation also treated the temples in a different manner: a temple in the countryside had to be demolished without disturbing the peace – while its masonry could be used for other construction works, cf. Codex Theodosianus 16, 10, 16 (a. 399) and 15, 1, 36 (a. 397). 37 Codex Theodosianus 16, 10, 25 (a. 435); cf. Hunt, Christianising the Roman Empire 157 and Kunderewicz, La protection 141. Whilst it is true that – especially in the case of the emperors of the Theodosian dynasty – the imperial fight against paganism is inextricably linked with legislative activity, we should be careful not to attribute too much weight to the passages in the church histories of this era that copiously quote the decrees and regulations: the ecclesiastical historians of the fifth century (Socrates, 33 34

24

KONSTANTIN M. KLEIN

I.3 A plea for the temples In the wake of this legislative development against pagan worship, prominent and influential pagans, however, did not just keep to the sidelines. A distinctive (but still unsuccessful) voice of dissent came from Libanius, friend of Christians and pagans alike, not least the teacher and companion of Julian, the last pagan Roman emperor. Libanius taught as professor of rhetoric in Antioch since 359 where he counted famous Christians among his disciples, John Chrysostom, for instance, or Theodore of Mopsuestia. Even after the death of Julian in 363, he remained in this prominent position and was awarded the honorific title of praefectus praetorio by Theodosius I. Libanius’ speech For the temples (Περὶ τῶν Ἱερῶν), which can be dated to 386, was directed to this emperor. 38 It is, however, important to note that we should not perceive this highly cultured rhetor’s oration as in any way representative of other Late Antique pagans – the handful of pagan statements which we have in comparison to the bulk of material from Christian descriptions, as Roger Bagnall put it, cannot be safely taken as representative of anyone but of their authors themselves. 39 Libanius based his plea for the preservation of temples – or to put it differently: against their destruction – mainly on one argument: the blessed emperor Constantine had made no juridical alteration concerning this matter. Of course, the law of Constantius II, mentioned above, was undeniable, according to Libanius, but it was Sozomen, Theodoret and Rufinus) were all continuing and to varying extents imitating Eusebius’ Historia ecclesiastica, a work that itself incorporated roughly two hundred and fifty official documents and letters in full as well as more than one hundred texts in lengthy summaries. A further reason for the significant role of imperial decrees in the church histories may be Socrates’ and Sozomen’s knowledge of Latin, the language of the decrees, and, in the latter author’s case, his own education as a lawyer; cf. Cameron, The last pagans 69. 38 Cf. Petit, Zur Datierung 54–7; for the honorary title praefectus praetorio (most likely awarded in the winter of 383/384), cf. ibid. 47–8 and Nesselrath, Libanius 27–30. 39 Bagnall, Models 25–6, cf. also, for introductionary remarks on the speech, Wiemer, Für die Tempel? 162.

THE SILENCE OF THE GODS

25

issued only because this emperor had bad advisors. 40 Also, Theodosius’ confirmation of the ban on sacrifices could not be denied, but, Libanius set forth in arguing that the emperor had not (we also could add: not yet) issued a law ordering the closure of temples or a decree that forbade entering them. To burn incense, burning fire in general or the libation of perfume, all constitute completely lawful actions that are not forbidden. 41 Employing perhaps not the best justification, Libanius argues, that no-one could really prove that pagans were still sacrificing in the new Christian Empire: did it constitute a pagan sacrifice if someone was feasting on some oxen in the temple which had been butchered at a different place? An innocent outing like this could not be forbidden, even if at the same time some incense was burnt, and if some of the drinking songs might accidentally resemble the old hymns to the gods. 42 The Christians, however, Libanius goes on, especially the monks, whom he calls a black-robed tribe (οἱ δὲ μελανειμονοῦντες οὗτοι), who eat more than elephants, turn violent against the temples: they “attack the temples with sticks and stones and bars of iron, and in some cases, disdaining these, with hands and feet.” 43 Libanius, Oratio 30, 6–7; all edicts issued by Constantine, which tolerate pagan rituals, date to the time before 324, cf. Codex Theodosianus 9, 16, 1–3 (probably all a. 319). Cf. on Libanius’ oration in general Prinz, Die Kirche 282–3. Libanius’ own religious views are summarized in Nesselrath, Libanius 54–9. 41 Libanius, Oratio 30, 8. 42 Libanius, Oratio 30, 16–8. Libanius might allued to a decree dating to 346 which calls for a preservation of temples outside the city walls because of their connection with public entertainment (“since from them is provided the regular performance of long established amusements for the Roman people”), cf. Codex Theodosianus 16, 10, 3 (a. 346). 43 Libanius, Oratio 30, 8: οἱ δὲ μελανειμονοῦντες οὗτοι καὶ πλείω μὲν τῶν ἐλεϕάντων ἐσϑίοντες, πόνον δὲ παρέχοντες τῷ πλήϑει τῶν ἐκπωμάτων τοῖς δι’ ᾀσμάτων αὐτοῖς παραπέμπουσι τὸ ποτόν, συγκρύπτοντες δὲ ταῦτα ὠχρότητι τῇ διὰ τέχνης αὐτοῖς πεπορισμένῃ μένοντος, ὦ βασιλεῦ, καὶ κρατοῦντος τοῦ νόμου ϑέουσιν ἐϕ’ ἱερὰ ξύλα ϕέροντες καὶ λίϑους καὶ σίδηρον, οἱ δὲ καὶ ἄνευ τούτων χεῖρας καὶ πόδας (But this black-robed tribe, who eat more than elephants and, by the 40

26

KONSTANTIN M. KLEIN

What follows equals complete destruction: the stripping of roofs, demolition of walls, the tearing down of statues, the overthrow of altars. The priest must either keep quiet or die. 44 After they have destroyed one temple, they hasten to the next, then to a third, Libanius remarks, trophy is piled on trophy, and all this is against the law. 45 The temples were the eyes of the cities and the souls of the countryside, Libanius claims. 46 They were essential for the supremquantities of drink they consume, weary those that accompany their drinking with the singing of hymns, who hide these excesses under an artificially contrived pallor – these people, Sire, while the law yet remains in force, hasten to attack the temples with sticks and stones and bars of iron, and in some cases, disdaining these, with hands and feet). 44 The, at least from a modern perspective, perhaps most potent argument of Libanius’ defense of temples, may be his reference to the scriptures which proclaim the centrality of Love for Christians and how this clashes with their actual deeds, cf. Libanius, Oratio 30, 21. 45 Libanius, Oratio 30, 8: καὶ τρόπαια τροπαίος ἐναντία τῷ νόμῳ συνείρεται. Indeed, the first law actually ordering the destruction of (rural) temples dates to July 399, consecutive laws clearly command that urban temples were under protection. Cultic images, however, were to be removed, cf. Codex Theodosianus 16, 10, 16 (a. 399), 16, 10, 18 (a. 399) and 16, 10, 19, 2 (a. 407) with Kunderewicz, La protection 146–67. 46 Libanius, Oratio 30, 9: χωροῦσι τοίνυν διὰ τῶν ἀγρῶν ὥσπερ χείμαρροι κατασύροντες διὰ τῶν ἱερῶν τοὺς ἀγρούς. ὅτου γὰρ ἂν ἱερὸν ἐκκόψωσιν ἀγροῦ, οὗτος τετύϕλωνταί τε καὶ κεῖται καὶ τέϑνηκε. ψυχὴ γάρ, ὦ βασιλεῦ, τοῖς ἀγροῖς τὰ ἱερὰ προοίμια τῆς ἐν τοῖς ἀγροῖς κτίσεως γεγενημένα καὶ διὰ πολλῶν γενεῶν εἰς τοὺς νῦν ὄντας ἀϕιγμένα (So they sweep across the countryside like rivers in spate, and by ravaging the temples, they ravage the estates, for wherever they tear out a temple from an estate, that estate is blinded and lies murdered. Temples, Sire, are the soul of the countryside: they mark the beginning of its settlement, and have been passed down through many generations to the men of today). On the visibility of temples in the cityscape, cf. McLynn, Pagans 577–8 and Libanius, Oratio 30, 42: μηδὲ τὸ χεῖρα μὲν ἀποκόπτειν ἀνϑρώπου δεινὸν ἡγώμεϑα, πόλεων δὲ ὀϕϑαλμοὺς ἐξορύττειν μέτριον (Do not let us think it a crime to cut off a man’s hand and a credit to gouge out the eyes of cities). As the decree law that actually ordered the destruction of temples explicitly restricts them to those in the countryside, we may assume that a

THE SILENCE OF THE GODS

27

acy of the Roman Empire, an argument in which he – unknowingly – took up Tertullian’s argument in reverse that war and true religion were mutually exclusive, that is that Rome’s greatness derived from its irreligiosity. 47 The destruction of the temples, however, had not only cultural, but also economic consequences: farmers, used to their familiar shrines, would loose all hope once the temples were demolished, and they would consequently cultivate the land in a less productive and responsible way; and Christian looting would not even make a stop at the cities. 48 The real victims were Theodosius’ subjects, and, Libanius specifies, those who spend their life working, not the idle monks. 49 The temples were the property of the emperor himself, so it was him whom the monks deprive, Libanius notes while demanding military protections for the sanctuaries. And whereas an emperor is after all a mortal being, it is ultimately the imperishable Roman Empire that is robbed of its ancient art treasures. 50 Above all, and in this we recognize the trained jurist Libanius, Christian violence had no legal basis: the emperor had not ordered these demolitions, and this is why they constitute self-administered justice. As there are no offences committed by pagans, so there cannot be verdicts, and even if there were, they could not be enforced by the monks. 51

Roman emperor such as Theodosius I – at least to a certain degree – valued the urban temples as aesthetic monuments; cf. Codex Theodosianus 16, 10, 16 (a. 399) with Meier, Alte Tempel 367–8. 47 Cf. Libanius, Oratio 30, 31–4; in a different speech, Libanius argues that the constant assistance and help by the gods was essential for military success, cf. Oratio 18, 167–9. Cf. also Wiemer, Für die Tempel 165. For Tertullian’s argument, cf. Apologeticum 25, 14–17 with Stroumsa, Barbarian philosophy 104–5. 48 Libanius, Oratio 30, 10–11. 49 Libanius, Oratio 30, 12. 50 Libanius, Oratio 30, 43 (imperial property), 14 (military protection) and 22 (art treasures made by the ancient ones); cf. also Klein, Spätantike Tempelzerstörungen 285 and 288. 51 Libanius, Oratio 30, 26.

28

KONSTANTIN M. KLEIN

I.4 Imaginary destructions in Gaza and the end of Serapis in Alexandria What Libanius classified as a judicial lapse was one some twenty years later Porphyry, Bishop of Gaza (most likely merely a literary figure 52) was careful not to commit. According to his hagiographer, Porphyry himself travelled to Constantinople to ask Theodosius’ son, the Emperor Arcadius, to grant him a decree that permitted the legitimate closure of the temple of Zeus Marnas in his episcopal town; a decree, however, that Porphyry chose to interpret quite freely: He used a mixture of pitch, sulfur and pig fat to burn down the temple. 53 Libanius could still hope in the mid-380s when he The Vita Porphyrii claims to be an eyewitness report recorded by a disciple of Porphyry, Marc the Deacon; it is in all likelihood a much later creation as evident from the lack of attestation to major protagonists, including bishop Porphyry himself, in an otherwise well-documented period of the history of Palestine in Late Antiquity. As the text in its current form quotes passages written by Theodoret of Cyrus, it cannot have been written before the mid-fifth century; cf. Bruns, Markus and Trombley, Hellenic religion i, 246–82. 53 Marc the Deacon, Vita Porphyrii 33, 40–5, 49, cf. also Porphyry’s first attempts which only resulted in the closing of smaller temples in Gaza, but not of the important Marneion, Vita Porphyrii, 26–7 with Klein, Spätantike Tempelzerstörungen 288. Together with the burning temple, the magic books of the pagans were consumed by fire (Vita Porphyrii 69– 71). Since the reign of Theodosius’ predecessor Valens (364–78) it was not allowed to read books concerning pagan magic, and people apparently were quite afraid of this kind of knowledge: John Chrysostom recounts how when he was a child in Antioch and played on the banks of the river Orontes, he found such a magic treatise, took it with him, but then threw it in the river when he saw a Roman soldier approaching; cf. Chrysostom, Homilia 38, 5 (Patrologia Graeca 60, 274–5) with Sarefield, Bookburning 290. Only the sacred books of the Jews were – at times – saved: when the Christians of Menorca persecuted their Jewish fellow-citizens in 418 and forced them to a collective conversion, the island’s synagogue equally went up in flames. All that was in the building was destroyed with the 52

THE SILENCE OF THE GODS

29

suggested that at least in Rome and in Alexandria sacrifices were still performed. Both cities were too important to the benefit of the Empire that one would risk its general safety by terminating these cults. Egypt was the gift of the Nile, and only feasts celebrated in the large Alexandrian temples, first and foremost in the magnificent temple of Serapis, induced the river to flood and bring annual fertility to the fields. Elsewhere Libanius mentions that he prays that the Serapeion may stay intact. 54 However, less than five years after he had expressed these hopes, what he feared came to pass: The Antiochian defender of temples crops up in the sources at least until 393, so he must have learned of the destruction of the Serapeion. In 391, Theodosius promulgated a decree that forbade entering and walking around in pagan temples. 55 In the same year, Theophilus I, Bishop of Alexandria obtained legal authority to reuse one of the now abandoned temples and planned to transform it into a church. During the renovations, a riot between Christians and pagans erupted, which ended with the pagans barricaded themexception of the holy books, which were, of course, also holy to the Christians. The latter consequently acted as safe keepers of the writings so that the books would not “suffer any injustice from the Jews;” cf. Severus of Menorca, Epistula Severi 13, 13: Omnia eius ornamenta, exceptis tamen libris atque argento, cum ipsa pariter ignis absumpsit. Libros enim sanctos ne apud Iudaeos iniuriam paterentur nos abstulimus. 54 Libanius, Oratio 30, 35 and 30, 44: ἤκουσα δὲ καὶ ἐριζόντων τινῶν, ἐν ὁποτέρῳ τὸ θαῦμα μεῖζον ἱερῷ, τῷ μηκέτ’ ὄντι τούτῳ ἢ ὃ μήποτε πάθοι ταὐτόν, ἐν ᾧπερ ὁ Σάραπις (I have even heard it argued which temple held the greater marvel, this [scil. probably the large temple of Edessa, closed by Theodosius in 382, cf. Codex Theodosianus 16, 10, 8] that is now no more or that of Serapis, which I pray may never suffer the same fate [of destruction]). For some considerations on the special status of the temples in Rome and Alexandria, cf. Gutsfeld, Hahn and Lehmann, Christlicher Staat 228. On Libanius’ apparently successful intervention for the conservation of a pagan sacred grove at the sanctuary of Apollo in Daphne, cf. Libanius, Oratio 1, 262 with Nesselrath, Libanius 28. 55 Cf. Codex Theodosianus 16, 10, 10 and 11 (a. 391), with Gaudement, La condamnation 597–602.

30

KONSTANTIN M. KLEIN

selves inside the Serpeion which eventually was demolished by monks and Roman soldiers with imperial toleration. 56 The pagan inhabitants of Alexandria had to learn to live without what was arguably their most important god. However, soon voices spread the rumor that Serapis was not killed by the Christian monks but rather had withdrawn from this world that had proved unworthy of his presence. 57 I.5 Statue collections and anger: the case of Gessius and Shenoute Theodosius the Great may not have ordered the destruction of the Serapeion, however, the events in Alexandria were inspired by his legislation, even if the emperor did not intend – and perhaps not foresee – its consequences. Further south in Egypt, the end of the great temple of Alexandria spurred another charismatic and resolute holy man to action: 58 Even if the ancient gods had indeed been There is no agreement on the exact date of the destruction of the Serapeion; I follow Judith McKenzie’s arguments that place it in 391 (Codex Theodosianus 16, 10, 11), the same year in which Theodosius’ law forbade entering pagan temples in general; cf. McKenzie, Gibson and Reyes, Reconstructing the Serapeum 108. It is important to note that, against Johannes Hahn’s dating attempt, there is no reason to dismiss the year 391 because of this law; neither the issue of the law nor the destruction of the temple need to be consequences of each other. With his dating of the destruction to before April 392, Hahn reaches a similar conclusion as Barnes, Ammianus Marcellinus 61–2. A useful list of all suggested dates for the destruction is compiled in Hahn, Vetustus 369–70 with n. 6. On the level of imperial influence/toleration, cf. Cameron, The last pagans 62– 3. 57 Cf. Brown, Christianization 635. 58 Shenoute was infamous for the discipline in his monastery and had once beaten a disobedient monk so hard that he died of the blows, cf. Bell, Besa 6; cf. Besa, Vita Sinuthii 6 for the saint, still a boy, beating a possessed man with a simandron until the demon left his body, and Vita Sinuthii 81–2 for him being involved in a brawl. For the Coptic tradition of Shenoute beating Nestorius with a Gospel manuscript at the Council of Ephesus, see Vita Sinuthii 129; cf. Lubomierski, Die Vita Sinuthii 177–8 56

THE SILENCE OF THE GODS

31

gradually fading away, Shenoute, the abbot of the monastery of Atripe (south of modern Sawhāj), undoubtedly helped accelerate this process. 59 His anti-pagan activities were carried out with a crowd of highly indoctrinated and aggressive monks. Shenoute, however, emphasized that all destructions for which he was responsible were undertaken only in order to avoid disorder (ⲁⲧⲁⲝⲓⲁ or ϣⲧⲟⲣⲧⲣ). 60 Among the sometimes stereotypically depicted antagonists of Shenoute, a certain Flavius Aelius Gessius stands out. He was a former praeses Thebaidis, the regional governor during the years 376–378. 61 Moreover, Gessius was a rich man and exploited his workers – two reasons which sufficed to attract Shenoute’s anger. 62 However, in this particular case, the polemic is extraordinarily strong: The Life of Shenoute, written by the abbot’s successor Besa, 63 even remarks that the protagonist saw the sinner Gessius in hell being tormented for his impiety by having his tongue bound to his big toe. 64 In Shenoute’s own writings, an even more serious reason why Gessius attracted the holy man’s condemnation is revealed: Shenoute claimed to have seen the former governor offerand Hahn, Gewalt 223. In the following passages, the Bohairic Vita Sinuthii (ed. J. Leipoldt) will be quoted. 59 Bell, Besa 18–19. 60 Emmel, Shenoute of Atripe 162–5 and 180; Behlmer, Schenute lxv. Cf. Besa, Vita Sinuthii 83–4 on the destruction of pagan temples. 61 Bagnall, Models 30–1 and Emmel, From the other side of the Nile 102–3; cf. Gessius 2, in PLRE i, 395. The episode appears also in Arabic version, where Gessius’ name is given as Kāsīyūs; cf. Besa, Vita Sinuthii (versio Arabica) 425, 11. 62 On Shenoute’s commitment to the poor farmhands, cf. Hahn, Gewalt 237–9 and Lubomierski, Die Vita Sinuthii 176–7. 63 On the author, cf. Lubomierski, Die Vita Sinuthii 156–7 and 167– 70; cf. also Bell, Besa 3–6 and Windau, Besa von Atripe. 64 Besa, Vita Sinuthii 88; cf. Brakke, Demons 104. On the different traditions concerning the cursing of Gessius, cf. Lubomierski, Die Vita Sinuthii 189; in another sermon Shenoute called Gessius “this son of pestilence” (ⲡⲉⲓϣⲏⲣⲉ ⲛⲗⲟⲓⲙⲟⲥ; cf. Shenoute, The Lord thundered 379, 9–10), whereas the community is addressed as a “tomb full of powerless demons (ⲧⲁⲫⲟⲥ ⲉⲩⲙⲉϩ ⲛⲇⲁⲓⲙⲱⲛⲓⲟⲛ ⲉϥⲙⲟⲟⲩⲧ; 365, 4–5).

32

KONSTANTIN M. KLEIN

ing libations in the already destroyed temple of Atripe and casting roses and peach branches around the ruins. Whereas Gessius, in an expression of Late Antique Egyptian syncretism, might not have seen a contradiction in paying homage to the old gods as well as to Christ, for Shenoute the case was clear: Gessius was worshipping Satan. 65 At a later passage in the text, the Life of Shenoute mentions that “one day, our father went to the city of Shmin (Panopolis) to carry off in secrecy by night the idols in Gessius’ house.” 66 Thanks to a fortunate textual find, we are relatively well informed on the framework of this incident that offers us insights in a “Late Antique real-life drama.” 67 Stephen Emmel edited and translated a fragment of a speech that Shenoute delivered when he was publicly questioned regarding his trespassing into Gessius’ house and stealing cultic statues. 68 The text postdates the death of Theodosius I (d. 395) but must have been written before the death of his son Arcadius (d. 408), most likely around 400. 69 It recounts how Shenoute went to the private Shenoute, De iudicio 91–2; cf. Brakke, Demons 97. Shenoute alludes to John 8:44: “Ye are of your father the devil, and the lusts of your father ye will do.” 66 Besa, Vita Sinuthii 125: ⲁⲥϣⲱⲡⲓ ⲇⲉ ⲟⲛ ⲛⲟⲩⲉϩⲟⲟⲩ ⲉⲧⲁⲡⲉⲛⲓⲱⲧ ϣⲉ ⲉϦⲟⲩⲛ ⲉϣⲙⲓⲛ ϯⲡⲟⲗⲓⲥ ϫⲉⲛⲧⲉϥⲓⲛⲓ ⲛⲛⲓⲓⲇⲱⲗⲟⲛ ⲉⲃⲟⲗ Ϧⲉⲛⲡⲏⲓ ⲛⲅⲉⲥⲓⲟⲥ. 67 Bagnall, Models 24; Brakke, Demons 104 speaks of a “very public drama.” 68 Cf. Emmel, Shenoute of Atripe 190–7; in what follows, the Coptic text is presented without the diacritical characters of the critical edition (such as brackets and sub-dotted letters). 69 Shenoute states that Theodosius’ descendants (ⲅⲉⲛⲟⲥ) were ruling. According to Stephen Emmel, this certainly excludes the time after the death of Honorius (d. 423), who was succeeded by Valentinian III who was not of the same γένος; he therefore narrows the terminus ante quem to 408, the death of Arcadius. However, the years between 408 and 423, when both Honorius in the West and Theodosius II (the son of Arcadius and therefore also of the same γένος) in the East were emperors, cannot be completely dismissed; cf. Emmel, Shenoute of Atripe 179. For a date around 400, cf. Emmel, From the other side of the Nile 111–113 and (2008) 180. 65

THE SILENCE OF THE GODS

33

(bed?)chamber of Gessius and removed idols of Cronus, Hecate, Zeus and a variety of specifically Egyptian objects, such as statuettes of bald-headed priests and a nilometer (“the ell, the measure of the water’s rise,” ⲡⲙⲁϩⲉ ⲡϣⲓ ⲙ̅ⲡⲙⲟⲩϩ ⲙ̅ⲡⲙⲟⲟ̅ⲩ). 70 The objects were publicly displayed to demonstrate Gessius’ shame. The fragment breaks off and we do not know how Shenoute himself accounted for what happened with the idols later on, however, Besa’s Life adds information on the end of Gessius’ statues: When they came to the pagan’s door, the doors of the house opened immediately one after another until they entered the place where the idols were. So with the brothers who were with him, he picked them up, took them down to the river, smashed them in pieces and threw them in the river. 71

Before Shenoute’s actions can be discussed under the aspect of his conception of legality, it is worth to cast a closer look upon Gessius and his statues. The governor’s career seemed to have come to an end with the death of Valens in 378, and it has been argued that he was aligned in some way with this emperor and the homoeans. A staunch supporter of Nicean Christianity (such as Shenoute) therefore would have conceived Gessius as an Arian. 72 From the edited fragments it is possible to reconstruct a verbal exchange between the monk and the official which must have preceded the events. It is likely that Gessius conceived himself not as a pagan, and he most certainly would not have admitted such a label for himself. Idolatry and paganism are accusations that only Shenoute insisted on applying to his opponent, 73 whom he asked whether he came with the Emmel, Shenoute of Atripe 169–70, Shenoute, Let our Eyes fr. 2, 5. Besa, Vita Sinuthii 126: ⲟⲩⲟϩ ⲉⲧⲁⲩϩⲱⲗ ϩⲓⲣⲉⲛⲡⲓⲣⲟ ⲛⲧⲉⲛⲓϩⲉⲗⲏⲛⲟⲥ Ϧⲉⲛϯⲟⲩⲛⲟⲩ ⲁⲛⲓⲣⲱⲟⲩ ⲛⲧⲉⲡⲓⲏⲓ ⲟⲩⲱⲛ ⲛⲥⲁⲛⲟⲩⲉⲣⲏⲟⲩ ϣⲁⲧⲉϥϩⲱⲗ ⲉϦⲟⲩⲛ ⲉⲡⲓⲙⲱⲓⲧ ⲉⲣⲉⲛⲓⲓⲇⲱⲗⲟⲛ ⲛϦⲏⲧϥ ⲟⲩⲟϩ ⲡⲁⲓⲣⲏϯ ⲁϥⲟⲗⲟⲩ ⲛⲉⲙⲛⲓⲥⲛⲏⲟⲩ ⲉⲑⲛⲉⲙⲁϥ ⲁⲩⲉⲛⲟⲩ ⲉⲃⲟⲗ ϩⲓϫⲉⲛⲫⲓⲁⲣⲟ ⲁⲩⲕⲟⲣϫⲟⲩ ⲁⲩⲁⲓⲧⲟⲩ ⲙⲫⲁϣⲓ ⲫⲁϣⲓ ⲁⲩⲥⲁⲧⲟⲩ ⲉϦⲣⲏⲓ ⲉⲫⲓⲁⲣⲟ. Cf. also Shenoute, Let our eyes fr. 2,3–4 and 1,22– 25 where the intruder claims that no servant or no gardener opened the door, but Christ made them open by themselves. 72 Bagnall, Models 30–1. 73 Bagnall, Models 28 and Brakke, Demons 104. 70 71

34

KONSTANTIN M. KLEIN

mind of Christ, or with the mind of Cronus. 74 Yet, in the report that Shenoute gives, Gessius is accused of lying: And it was through Him [scil. Christ] that we exposed them openly so as for everyone to recognize his [scil. Gessius’] contempt and his shame, for them to recognize that he is a liar for having said, “There are no idols (ⲛ̅ⲓⲇⲱⲗⲟⲛ) in my house,” when I asked him, since Jesus, in whom he does not believe, has exposed his moral failings. 75

Even though no records of Gessius’ perspective survived, one could still assume that he might not have considered himself neither a pagan nor a liar: even from Shenoute’s polemic it becomes quite clear that the idols constituted a collection of pieces art objects rather than objects of worship. 76 We may assume that the Roman governor would have drawn a distinct line between the words “idol,” a statue that served in a cultic context (εἴδωλον), and a mere “statue” (ἄγαλμα). If indeed the pieces destroyed by Shenoute formed a collection of statues and local antiquities, which Gessius judged worthy and aesthetic enough to exhibit in his house, there was no fault in his claim that he kept no idols in the house. Undoubtedly, the boundaries were fluid on whether a statue represented the deity itself, a piece of art, or a lifeless stone. However, already two-hundred years earlier, the cultic image of Apollo from the temple of Bassai was, according to Pausanias, given as a gift to Megalopolis for the adornment of the newly founded city, where the statue was positioned in the market place, i.e. not inside a temple. However, sacrifices were offered in front of the statue, while parts of the victims were then carried to the temple in Bassai outside the city of Megalopolis, where the remains were conShenoute, Let our eyes fr. 1,26. Shenoute, Let our eyes Fr. 1,25: ⲁⲩⲱ ⲉⲃⲟⲗ ϩⲓⲧⲟⲟⲧϥ ⲛ̅ⲧⲁⲛⲟⲩⲟⲛϩⲟⲩ ⲉ̅ⲃⲟⲗ ϩ͞ⲛⲟⲩⲡⲁⲣ̅ⲣⲏⲥⲓⲁ· ⲉ̅ⲧⲣⲉⲟⲩⲟⲛ ⲛⲓⲙ ⲉⲓⲙⲉ ⲉⲡⲉϥⲥⲱϣ ⲙ͞ⲛⲡⲉϥϣⲓⲡⲉ· ⲉⲧⲣⲉⲩⲉⲓⲙⲉ ϫⲉⲟⲩⲣⲉϥϫⲓϭⲟⲗ ⲡⲉ ⲛ̅ⲑⲉ ⲛ̅ⲧⲁϥϫⲟⲟⲥ ϫⲉⲙ͞ⲛⲗⲁⲁⲩ ⲛ̅ⲓⲇ ̈ ⲱⲗⲟⲛ ϩ͞ⲙⲡⲁⲏⲓ ⲛ̅ⲧⲉⲣⲉⲓϫⲛⲟⲩϥ· ⲉⲁⲓ͞ⲥ̅ ϭⲉ ⲡⲁⲓ ⲉ̅ⲧ͞ϥⲟ ⲛ̅ⲁⲧⲛⲁϩⲧⲉ ⲉ̅ⲣⲟϥ ⲟⲩⲱⲛ͞ϩ ⲉⲃⲟⲗ ⲛ̅ⲛⲉϥⲕⲁⲕⲓⲁ̅. 76 Emmel, From the other side of the Nile 107 n. 58 and (2008) 171. 74 75

THE SILENCE OF THE GODS

35

sumed. 77 The example shows that one statue could function both as cultic image and aesthetic object at the same time, however, the function and role of idols within a religious setting were hotly debated: did the statue of a deity represent his/her likeness in the way a statue of Alexander represented his (with the not unimportant difference that, of course, no ritual could invoke the hero’s spirit into the statue), was it rather a symbol of a deity whose nature could never be reflected on a graven image, or was it a representation indicating the deity in the way that smoke signifies fire? 78 Plutarch, who recounted the story of the death of Pan, noted that the gods should be defined by virtue, sublimity and magnanimity rather by anthropomorphic bodies carved by woodcutters, stonemasons and brass-founders, 79 and he ridiculed the simple men who did not understand the differences between the true existence of the deities and their mere sculptural depiction. 80 Much later Augustine wondered: Does anyone imagine that idols have any sense of perception? Yet, when they are set in lofty shrines to be honoured, and are waited on by those who pray and offer victims, dumb and lifeless as they are, they give the illusion of moving and feeling, and greatly increase the veneration of the crowd, on which their cult so greatly depends. 81

Pausanias, Graeciae descriptio 8, 14, 9; cf. Auffarth, Götterbilder 314 as well as 307–8, quoting other examples from classical antiquity, e.g. Cicero’s speech In Verrem (2, 4, 106–15); cf. also Breitner, Statuen 109–12. 78 Cf. Graf, Plutarch 253. 79 Plutarch, De superstitione 4, 166E; cf. Graf, Plutarch 259. 80 Plutarch, De Iside et Osiride 71, 379 C, mocking humans who assumed, for instance, that the Capitoline Jupiter was lost by fire in the wake of the Roman Civil War. 81 Augustine, Epistula 102, 18 (PG 33, 377): Et idola quidem omni sensu carere, quis dubitet? Verumtamen cum his locantur sedibus, honorabili sublimitate, ut a precantibus atque immolantibus attendantur, ipsa similitudine animatorum membrorum atque sensuum, quamvis insensata et exanima, afficiunt infirmos animos, ut vivere ac spirare videantur; accedente praesertim veneratione multitudinis, qua tantus eis cultus impenditur. 77

36

KONSTANTIN M. KLEIN

In his didactic poem Against Symmachus, Prudentius ridicules the process by the Romans imbibe from their infants’ days a childlike belief in statues, which, despite of all care shown to them, are ephemeral images: bronze statue collapse, the golden surfaces wear off, silver becomes tarnished, and all metals eventually oxidize. 82 Most Late Antique Christian texts converged in their discussions on the role of statues in pagan religion in emphasizing their inanimate character. In doing so they encouraged a disengagement from the concept that any godly power originated from such images. 83 While a large amount of former idols were certainly protected and saved from destruction by Christians who desacralized and estheticized them into mere works of art, 84 cases such as that of Shenoute and Gessius demonstrate that not all inhabitants of the later Roman Empire shared similar thoughts when it came to the significance of statues. A series of Late Antique imperial decrees against collecting statues from various (tomb) monuments has been read as an empire-wide context of art-dealership that was particularly flourishing in the late fourth century; however, similar legislative evidence against the use of spolia does exist from the first and second centuries as well. 85 A decree from the Codex Theodosianus ordered that – in the wake of an overall negative attitude towards temples – those temples which contained statues of aesthetic value “must be measured by the value of their art rather than by their divinity.” 86 Prudentius, Contra Symacchum 2, 750–3 (rusting statues) and 1, 203–10 (veneration in a child-like state of mind). 83 Breitner, Statuen 110 and Graf, Plutarch 255. The Vita Melaniae 44, written in the mid-fifth century, most likely by her successor as leader of her monasteries, Gerontius, contains, among other references to statutes (no of which is explicitly negative) an interesting comparison: true Christians should become as motionless as a statue, so that no insult, no stroke, no protest would be returned. 84 Cf. Meier, Alte Tempel 361–74 and Dally, Pflege 97–114. 85 Saradi, Christian attitudes 50–1 with Codex Theodosianus 15, 1, 14 (a. 365), 9, 17, 4 (a. 357) and 9, 17, 5 (a. 363). More examples from the classical period can be found in Meier, Alte Tempel 365–7; cf. also Kunderewicz, La protection 139–40. 86 Codex Theodosianus 16, 10, 8 (a. 382): artis pretio quam divinitate metien82

THE SILENCE OF THE GODS

37

While dismantling temples of their art and leaving them untouched but vulnerable to natural disasters and the passing of time, in most cases accounted for a slow and ‘natural’ decay of temples and a relocation of their statues, certain cases of violent destruction of cultic images are archaeologically attested. 87 Contemporary to the three churches in Jerusalem and Bethlehem, Constantine the Great also ordered the cleansing of a forth pagan site in Palestine, Mamre. His mother-in-law, Eutropia, had informed him that the place where Abraham had lived, was tainted: Idols fit only for absolute destruction have been set up beside it, [...] and an altar stands nearby, and foul sacrifices are constantly conducted there. 88

The emperor decreed the destruction of the sanctuary and explicitly stated that any idols that were found at the place should be consigned to the flames and the altar completely demolished. Constantine commissioned the comes Acacius to devote all possible effort and endeavor to clearing the whole area and had a basilica built on the spot. While this church was visited and mentioned by Late Antique pilgrims, the excavations carried out by Evaristus Mader in 1926–1928 unveiled surprising findings: Even though there were arguably not many pieces, the excavators discovered fragments of pagan statues and reliefs dating to the times before the destruction. A relief depicting an androgynous Hermes with female breasts and jewelry was hardly recognizable, as the figure’s face had been chisda; cf. also Bassett, The urban image 112 and Meier, Alte Tempel 367 and 371 who draws parallels to the de-sacralization of art in the wake of the French revolution: the term “monuments historiques” appears for the first time in the 1790s, when – after a relatively ephemeral iconoclastic riot – the first heritage conservation laws were decreed. 87 Bar, Continuity 288; cf. Bowersock, Polytheism. 88 Eusebius, Vita Constantini 3, 53, 1: εἴδωλα τε γὰρ πάσης ἐξωλείας ἄξια παρ’ αὐτὴν ἱδρῦσϑαι καὶ βωμὸν [...] πλησίον ἑστάναι καὶ ϑυσίας ἀκαϑάρτους συνεχῶς ἐπιτελεῖσϑαι. On the theophany at Mamre, cf. Genesis 18:1–20. The Bordeaux Pilgrim, visiting the Holy Land in 333 CE, mentions the “exceptionally beautiful” Constantinian basilica in Mamre, cf. Itinerarium Burdigalense 599.

38

KONSTANTIN M. KLEIN

eled off; a statue of Dionysus did not fare better: the head of the deity had been smashed with one vigorous blow. 89 All other sculptural findings, among them a Hercules with lionskin, showed similar signs of violence, and thus corroborate the execution of Constantine’s order to demolish all images in Mamre. At the same time when the comes Acacius was supervising the smashing of the idols of Dionysus, Hercules, and the Hermaphrodite in Mamre, Constantinopolitans who visited the new capital’s premier baths, those of Zeuxippos, could marvel at a large collection of statuary which was displayed there by order of the same emperor who had ordered the destructions in Mamre. An (incomplete) poem (Ekphrasis) by Christodorus of Thebes mentions about eighty exquisite pieces of art that were arranged in order to display a unique program, the grand vision of imperially shaped romanitas, i.e. the idea of a Constantinopolitan succession to the traditions of Rome, which was a product of Constantine’s own aspirations for his city. 90 From the old capital he had brought the Palladium, the statue of the armed Athena allegedly fallen from heaven and intrinsically linked with the mythical fate of the city of Troy. By exhibiting the Palladium at his forum in Constantinople, the emperor irrevocably connected the histories of three cities: Troy, Rome, and Constantinople. 91 The largest concentration of statues, however, was to be found in the Baths of Zeuxippos and in the HippoBar, Continuity 284–5 and Mader, Mambre 135–7. It is a different question, however, whether the Constantinian decrees terminated the pagan worship at Mamre: Sozomen mentions the past plurality at the spot about a hundred years later (cf. Historia Ecclesiastica 2, 4, 2) but formulates he remarks about various cult practices in the present tense, thus raising suspicions that the building of Constantine’s basilica did not necessarily change the nature of the place revered by Jews, Christians and pagans alike; cf. Wallraff, Die antipaganen Maßnahmen 11–12. 90 Bassett, The urban image 50, 56, 65–6 and esp. 75–8 for the agenda behind the Constantinian statuary. On Late Antique statues in general, cf. also the edited volume by Bauer and Witschel, Statuen, as well as the “The last statues of antiquity” database: http://laststatues.classics.ox.ac.uk/ 91 Bassett, The urban image 69 and 205–6 (cat. no. 114), cf. also Cameron, The later Roman Empire 170. 89

THE SILENCE OF THE GODS

39

drome. Among the deities and famous historical persons depicted, both Dionysus and Hercules were among the most popular images – statues of the same gods that had been violently smashed in Mamre. A visitor to the Baths of Zeuxippos, however, could also contemplate statues of Hecuba (like one of the pieces that were taken by Shenoute from Gessius’ house) and of a Hermaphrodite. Whereas such a figure had been violently chiselled off in Mamre, one century later, Christodorus, was full of praise for the exquisitely ambiguous statue in the baths: There stood a lovely Hermaphroditus, nor wholly a man, nor wholly a woman, for the statue was of mixed form: readily couldst thou tell him to be the son of fair-bosomed Aphrodite and of Hermes. His breasts were swelling like a girl’s, but he plainly had the procreative organs of a man, and he showed features of the beauty of both sexes. 92

While Christodorus, who flourished at the turn of the fifth to the sixth century, was enthused over the beauty of a “golden, naked, and all glittering” Aphrodite and admired the defined back muscles of a naked wrestler’s statue, 93 Eusebius of Caesarea, in contrast, Christodorus, Ekphrasis 102–7: Ἵστατο δ’ Ἑρμαϕρόδιτος ἐπήρατος, οὔϑ’ ὅλος ἀνήρ, οὐδὲ γυνή· μικτὸν γὰρ ἔην βρέτας· ἦ τάχα κοῦρον Κύπριδος εὐκόλποιο καὶ Ἑρμάωνος ἐνίψεις· μαζοὺς μὲν σϕριγόωντας ἐδείκνυεν, οἷά τε κούρη· σχῆμα δὲ πᾶσιν ἔϕαινε ϕυτοσπόρον ἄρσενος αἰδοῦς, ξυνῆς ἀγλαΐης κεκερασμένα σήματα ϕαίνων. On the statues of Dionysus, Heracles and the Hermaphrodite, cf. Bassett, The urban image 52. For statues in the Roman bath of Aphrodite in Ptolemais-Acco, cf. Mishna, ʿAvoda Sara 3, 4 with its famous dispute between “Proklos ben Pilosolos” and Rabban Gamliel; cf. Hoss, Die spätantike Blüte 171–2. 93 Christodorus, Ekphrasis 99–101 (Aphrodite: χρυσῆν Ἀϕροδίτην, γυμνὴν παμϕανόωσαν) and 228–40 (the naked wrestler). Similarly, Malalas is full of praise for Constantine’s public display of images: Chronographia 13, 8: ὁμοίως δὲ ἀνεπλήρωσε καὶ τὸ λεγόμενον Ζεύξιππον δημόσιον λουτρὸν κοσμήσας κίοσι καὶ μαρμάροις ποικίλοις καὶ χαλκουργήμασιν· (Likewise he completed the public bath known as the Zeuxippon, and decorated it with columns and marbles of many colours and bronze statues). 92

40

KONSTANTIN M. KLEIN

had difficulties to reconcile Constantine’s art collections with the general pious image he presented of the Christian emperor in his Vita Constantini. Consequently, Eusebius resorted to a literary presentation of the public display of statues as a sophisticated display of their ludicrousness, while – to some extent, and especially for statues that did not depict gods – still acknowledging the artistic value: In yet other cases the sacred bronze figures, of which the error of the ancients had for a long time been proud, he [scil. Constantine] displayed to all the public in all the squares of the Emperor’s city, so that in one place the Pythian was displayed as a contemptible spectacle to the viewers, in another the Sminthian, in the Hippodrome itself the tripods from Delphi, and the Muses of Helicon at the palace. The city named after the Emperor was filled throughout with objects of skilled artwork in bronze dedicated in various provinces. To these under the name of gods those sick with error had for long ages vainly offered innumerable hecatombs and whole burnt sacrifices, but now they at last learnt sense, as the Emperor used these very toys for the laughter and amusement of the spectators. 94

Eusebius, Vita Constantini 3, 54, 2–3: ἄλλων τὰ σεμνὰ χαλκουργήματα, ἐϕ’ οἷς ἡ τῶν παλαιῶν ἀπάτη μακροῖς ἐσεμνύνετο χρόνοις, ἔκδηλα τοῖς πᾶσιν ἐν ἀγοραῖς πάσαις τῆς βασιλέως πόλεως προὐτίϑετο, ὡς εἰς ἀσχήμονα ϑέαν προκεῖσϑαι τοῖς ὁρῶσιν ὧδε μὲν τὸν Πύϑιον, ἑτέρωϑι δὲ τὸν Σμίνϑιον, ἐν αὐτῷ δ’ ἱπποδρομίῳ τοὺς ἐν Δελϕοῖς τρίποδας, τὰς δ’ Ἑλικωνίδας Μούσας ἐν παλατίῳ. ἐπληροῦτο δὲ δι’ ὅλου πᾶσα ἡ βασιλέως ἐπώνυμος πόλις τῶν κατὰ πᾶν ἔϑνος ἐντέχνοις χαλκοῦ ϕιλοκαλίαις ἀϕιερωμένων, οἷς ϑεῶν ὀνόματι πλείστας ὅσας ἑκατόμβας ὁλοκαύτους τε ϑυσίας εἰς μάταιον ἀποδόντες μακροῖς αἰῶσιν οἱ τὴν πλάνην νενοσηκότες ὀψέ ποτε ϕρονεῖν ἔγνωσαν, τούτοις αὐτοῖς ἀϑύρμασιν ἐπὶ γέλωτι καὶ παιδιᾷ τῶν ὁρώντων βασιλέως κεχρημένου. Cf. Cameron and Hall, Eusebius 301–3 who ascribe difficulties to Eusebius who had “to work hard, and draw on all his linguistic resources, to turn Constantine’s beautification of his city [...] into an anti-pagan gesture,” whereas Saradi, Christian attitudes 50 suggests that Eusebius did not try to conceal “what apparently was in everyone’s mind,” the artistic value of the statues from 94

THE SILENCE OF THE GODS

41

Being much more than the object of ridicule as Eusebius classified them, the imperial statues articulated themes of kingship and dynastic legitimation. However, they were by far not the only, and arguably not the most impressive collection on public display in Constantinople: A century later, the aristocrat Lausos, a chamberlain at the court of Theodosius II, who was known for his patronage of the Lausiac History as well as for being a devout Christian, displayed his private collection in a portico along the main street of the capital, the Mese. Lausos’ collection, which he must have acquired during his tenure in office and therefore had access to statuary amassed by the closure of temples, contained extraordinary pieces. While in the late fifth century, Cassiodorus called the Romans a nation “rich in statues” (populus copiosissimus statuarum), 95 the same epithet could verily be applied to this collector: among the at least thirteen statues, all stemming from one of the great Hellenic sanctuaries, were the chryselephantine Olympic Zeus of Phidias and Praxiteles’ Aphrodite of Knidos. In contrast to earlier arrangements of statuary, the collection of Lausos was less concerned with dynastic legitimacy, but rather embraced the question of religion. In a way not too dissimilar from Eusebius’ humiliating remarks concerning the Constantinian statues, this collection demonstrated the overpowered deities, while at the same time it exemplified, by the relatively wide range of dates of production, a notion of progress in the development of art. 96 There is reason to assume that the statues provoked mixed reactions in their beholders as the Christian worldview simultaneously admired, derided, and feared free standing sculptures. However, we have to expect that a person like Lausos, god-fearing and appreciative of art at the same time, saw no contradiction between his faith and his taste. A similar attitude might have prevailed some two decades earlier in Shmin, where the former governor Gessius collected pieces of art on a much smaller scale than Lausos. Through his official position he commanded over somewhat comantiquity. Cf. also Wallraff, Die antipaganen Maßnahmen 14–15. 95 Cassiodorus, Variae epistolae 7, 13, 1; cf. Breitner, Statuen 110. 96 On the Lausos collection, cf. Bassett, The urban image 98–120 (ch. 5), esp. 98–101.

42

KONSTANTIN M. KLEIN

parable prerequisites, as the statuary of dismantled temples was considered property of the imperial state. 97 While we do not know for which purpose Gessius kept the statues in his chamber (personal admiration of art, imitation of famous collectors in the capital or pagan idolatry), the secrecy of his collection offered a point of attack for Shenoute, who based his accusations on Biblical commands that clearly stated that idols should be destroyed: 98 Is there any witness more reliable than the witness of the Scriptures? Tell me from out of them what the sin is that I have committed, or how it was wrong for me to remove idols from the house of a godless man, and in particular the image of Kronos, along with her who is called Hecate, through whom people are deceived at the oracles. 99

Shenoute, who had received an excellent education in Greek rhetoric that is evident from his texts even though it remains a mystery where he learned Greek, styled Gessius as a deviant idolater and by doing so ostracized and segregated him from the community of Shmin: Gessius was now one of those “who were alien to the flock.” 100 With several rhetorical questions Shenoute justified the events by continuously referring not only to the seriousness of Bassett, The urban image 100. Shenoute, Let our eyes fr. 1,3 suggests that the objects found in Gessius’ house all were originally derived from temples which were closed and/or destroyed at the time of Shenoute’s speech; cf. also Emmel, Shenoute of Atripe 178. 98 Cf. esp. Deuteronomy 13:6–9 and 17:2–5 or Exodus 22:20 to which Shenoute, in a free paraphrase, alluded in his speech. 99 Shenoute, Let our eyes fr. 1, 10 (Emmel’s emphasis): ⲙⲏ ⲟⲩⲛ̅ϭⲉⲙ͞ⲛ͞ⲧⲙⲛ͞ⲧⲣⲉ ⲉϥⲟ ⲛ̅ϩⲟⲧ ⲛ̅ϩⲟⲩⲟ ⲉ̅ⲧⲁⲛⲉⲅⲣⲁⲫⲏ· ⲙⲁⲧⲁⲙⲟⲓ ⲉⲃⲟⲗ ⲛ̅ϩⲏⲧⲟⲩ ϫⲉⲟⲩ ⲡ͞ⲉ ⲡⲛⲟⲃⲉ ⲛ̅ⲧⲁⲓⲁⲁϥ ⲏ̅ ⲧⲁⲓⲡⲗⲁⲛⲁ ϩ͞ⲛⲟⲩ̑ ϫⲉⲁⲓϥⲓ̅ ⲛ̅ϩⲉⲛⲉⲓⲇⲱⲗⲟⲛ ⲉⲃⲟⲗ ϩ͞ⲙⲡⲏⲓ ⲛ̅ⲟⲩⲣⲱⲙⲉ ⲛ̅ⲁⲧⲛⲟⲩⲧⲉ ⲛ̅ϩⲟⲩⲟ ⲇⲉ ⲧϩⲓⲕⲱⲛ ⲛ̅ⲕⲣⲟⲛⲟⲥ· ⲙ͞ⲛⲧⲉⲧⲟⲩⲙⲟⲩⲧⲉ ⲉ̅ⲣⲟⲥ ϫⲉⲧϩⲉⲕⲁⲧⲏ ⲧⲁⲓ ⲉ̅ⲧⲟⲩⲁ̅ⲡⲁⲧⲁ ⲛ̅ⲛⲣ̅ ⲱⲙⲉ ⲉ̅ⲃⲟⲗ ϩⲓⲧⲟⲟⲧ͞ⲥ ϩ͞ⲛⲙ̅ⲙⲁ ⲛ̅ϣⲓⲛⲉ· Cf. Eusebius, Vita Constantini 3, 54, 3 (quoted above), as well as Let our eyes fr. 1, 21. 100 Shenoute, Let our eyes fr. 1, 8: [...] ⲉ̅ⲛⲉⲧⲟ ⲛ̅ϣⲙⲙⲟ̅ ⲉⲡⲟ̅ϩⲉ. On Shenoute’s style, education and Christian attitude, titled “christuslose Frömmigkeit,” cf. Leipoldt, Shenute von Atripe 82. 97

THE SILENCE OF THE GODS

43

Gessius’ offence, 101 but also to the legal situation – at least in the holy man’s interpretation of it, which brings us back to the legal situation of the destruction of temples mentioned in the first part of this paper: Shenoute mentions the edicts of Theodosius I and his honoured descendants who decreed to demolish the temples and to dig out their foundations until no stone was left on top of the other. 102 While rhetorically defending himself from the accusation of trespassing, the holy man jumped at several conclusions, one of which is particularly remarkable and telling about his concept of legality and about how he arbitrarily expanded the imperial decrees: “If it is illegal to remove demonic images from that man’s house,” he argues, “is it also illegal to remove them from their temples?” 103 If the emperors had ordered that statues should be removed from temples, then it was also legitimate for cult objects used for the same purpose elsewhere (i.e. in Gessius’ house) to be destroyed as well. 104 In a way, Shenoute’s argument, though deriving from opposite points of view, is reminiscent to that of Libanius in his oration in defense of the temples, where the Antiochean rhetor similarly twisted the emperor’s words: everything that was not forbidden, so Libanius, must be allowed. It has been argued that the only law to which Shenoute could have referred to was the decree issued in 399 by Theodosius concerning temples in the countryside, 105 however, this law is hardly Shenoute, Let our eyes fr. 2, 5–6. Shenoute, Let our eyes fr. 1, 3. 103 Shenoute, Let our eyes fr. 2, 8: ⲉϣϫⲉⲉⲝⲉⲥⲧⲓ ⲁⲛ ⲉϥⲓ ⲛ̅ϩⲉⲛϩⲓⲕⲱⲛ̅ ⲛ̅ⲇⲁⲓⲙⲟⲛⲓⲟⲛ ⲉⲃⲟⲗ ϩ͞ⲙⲡⲏⲓ̈ ⲙ̅ⲡⲉⲧ͞ⲙⲙⲁⲩ ⲏ̅ ⲉⲝⲉⲥⲧⲓ ⲁⲛ ⲟⲛ ⲉϥⲓⲧⲟⲩ ⲉⲃⲟⲗ ϩ͞ⲛⲛⲉⲩⲣ̅ⲡⲏⲩⲉ· 104 Emmel, Shenoute of Atripe 178. On the role of laws in Libanius’ argumentation, cf. Wiemer, Für die Tempel? 170. 105 Codex Theodosianus 16, 10, 16. If the incident happened slightly later than Emmel dates it, Shenoute could also allude to the imperial decrees of 407 (Codex Theodosianus 16, 10, 19, 1–2) which ordered that idols should be torn down, while the temples in the cities and in the countryside were to be designated for public use. A date after 407 seems possible for Shenoute, who died in 465 or 466, Gessius, however, who left office around 378 would most likely then have been in his 70s (as opposed to his 101 102

44

KONSTANTIN M. KLEIN

recognizable in the way Shenoute chose to present it in his defence. Nevertheless, this misquotation should not necessarily be seen only as a deliberate perversion of justice, it rather shows how the population of the Late Antique world learned about established law: not by reading the regulations, but rather by seeing the actions resulting out of them as well as by learning about them by word of mouth. 106 The evidence from the dispute between Shenoute and Gessius suggests that the decrees of the Codex Theodosianus are normative sources, which do not necessarily reflect the degree of their application in reality. I.6 Temples and statues between destruction and imperial protection At a first glance we can discern a certain development: Constantine set the example which was then emulated by others: Shenoute of Atripe who demolished temples in Panopolis as well as the statues in the private art collection of Gessius, the Christian mob of Alexandria destroying the famous Serapeion upon the call of their bishop, and Barṣawma and Porphyry, terminating pagan worship in the Euphrates region and in Gaza respectively. Nevertheless, despite the undeniable appeal of these texts, there are altogether only few cases recording the destruction of pagan temples in the sources; these episodes, however, are sometimes narrated up to five times in different versions. The repertoire is small and only few cases are well-documented. 107 Moreover, there is an important difference 60s with a date shortly after 399 CE). 106 Cf. Bagnall, Models 26, Errington, Christian accounts 398–443 and Meier, Alte Tempel 369. However, in a sermon on the veneration of martyrs, delivered in Panopolis, Shenoute claims that every edict and decree (ⲇⲓⲁⲧⲁⲅⲙⲁ and ⲡⲣⲟⲥⲧⲁⲅⲙⲁ) issued by the emperor was implemented immediately and everywhere in the Empire, cf. Hahn, Gesetze 201–2. On the distribution of the laws, arguing for a local implementation only, cf. Cameron, The last pagans 60–2 and 72–3 and Hunt, Christianising the Roman Empire 157–8. 107 Cf. Gotter, Rechtgläubige 87 for a useful chart enumerating all destructions of temples mentioned in Eusebius’ Vita Constantini and the Church Histories written by Rufinus, Socrates Scholasticus, Sozomen and

THE SILENCE OF THE GODS

45

between Constantine’s alleged destructions and the later ones by Shenoute, Barṣawma or Porphyry: this difference is located in the nature of the texts that record the destructions – as well as in the different spheres of influence of the protagonists of these texts. The hagiographical genre puts emphasis on the anti-pagan actions of its heroes, whereas the ecclesiastical histories stress the emperors’ legislation against pagan holy sites. It is unthinkable for the later church historians of the fifth century, that Theodosius I (let alone Theodosius II) would have taken up an axe to demolish a temple or a statue. They rather may (or may not – in the latter case the Christian historians gently skip the topic and proceed to other deeds) have issued laws that regulated the destruction of temples – laws that sometimes would be astonishingly misinterpreted – but neither emperors nor their court historians wished to convey the notion that Christianity had to be established by force. On the contrary it was a case of the workings of the divine plan. 108 In historiography the motif of destruction of temples is an expression of the emperor’s piety and orthodoxy, as the emperor was not physically present and acted by means of decrees and orders. The smoother the process was, the better. If there was resistance, it was almost always by seditious humans, never by the demons believed to be dwelling in the temples, as such an occurrence could only be solved by a symbolic duel with the evil powers, a fight for which the emperor was too far away. The duel between demons and their human enemies was therefore reserved for hagiography operating with saints who were physically present at the places of destruction. 109 To my knowledge there is only one Christian text in which both aspects are combined, and this is due to its peculiar nature: In AgaTheodoret. Stephen Emmel remarks on the textual evidence from Egypt that “the list of cases of “temple-destruction” [...] is remarkably short, and only very few of these cases are at all well documented,” cf. Emmel, Shenoute of Atripe 161. 108 Cameron, The last pagans 74. 109 Gotter, Rechtgläubige 57–9. I would not agree with his statement: “Wenn [...] in der Kirchengeschichte christliche Charismatiker auftreten und pagane Heiligtümer niederreißen, heißt das im Grund, daß der Kaiser nicht fromm bzw. orthodox ist.”

46

KONSTANTIN M. KLEIN

thangelos’ History of the Armenians we encounter a rare fusion of church history and hagiography, and two early-fourth century protagonists that work hand in hand, Trdat and Gregory the Illuminator, king and holy man, the first issuing the decrees that the latter immediately executed: Gregory took counsel with the king and together they agreed to “overthrow, destroy and extirpate the scandals, to suppress them entirely [...].” 110 Together they hastened from Vałarshapat to the city of Artashat in order to destroy the altars of the deity Anahit there. On the road they first came across the shrine of the god Tir and there they “first [...] set work, and destroyed, burnt, ruined and razed it.” 111 Their collaboration is styled in the wording of a crusade against paganism: “our boast is in the cross and our joy in the glory of God.” 112

II

THE FATE OF SANCTUARIES AND IDOLS IN THE TIME OF MUḤAMMAD

Early Muslim texts describing the destruction of pre-Islamic sanctuaries in the Ḥijāz employ a similar phraseology. Reading all these Agathangelos, Historia armeniaca 777: Առնոյր այնուհետև խորհուրդ հաւանութեան ընդ թագաւորին և ընդ իշխանսն, նախարարօքն և զօրօքն հանդերձ` վասն խաղաղութեան հասարակաց, քակել, կործանել, բառնալ զգայթակղութիւնսն ի միջոյ, ի բաց կորուսանել, զի մի՛ ումեք լիցի խէթ և խութ այնուհետև ընդ ոտս անկանել և խափան առնել՝ ելանել յազատութիւնն վերին: On evil spir110

its and demons in pre-Christian Armenia and their afterlife until the present day, cf. Russell, Zoroastrianism 447–58 and Russell, Pre-Christian Armenian Religion 2689–90. 111 Agathangelos, Historia armeniaca 778: Ապա ինքն իսկ թագաւորն

խաղայր գնայր ամենայն զօրօքն հանդերձ ի Վաղարշապատ քաղաքէ՝ երթալ յԱրտաշատ քաղաք, աւերել անդ զբագինսն Անահտական դիցն, և որ յԵրազամոյն տեղիսն անուանեալ կայր: Նախ դիպեալ ի ճանապարհի երազացոյց երազահան պաշտաման Տրի դից, դպրի գիտութեան քրմաց, անուանեալ Դիւան գրչի Որմզդի, ուսման ճարտարութեան մեհեան. նախ ի նա ձեռն արկեալ՝ քակեալ այրեալ աւերեալ քանդեցին: 112 Agathangelos, Historia armeniaca 777: [...] պարծանք ի խաչն, և գովութիւն ի փառս Աստուծոյ:

THE SILENCE OF THE GODS

47

episodes together, for instance in a work such as Ibn al-Kalbī’s Kitāb al-Aṣnām, in which the author compiled all information on idols and their sanctuaries available to him, one could easily gain the impression of a systematic quest against paganism. However, a comparative look into biographical and historiographical works of the period shows that the motif of destruction of temple occupies only a few pages of these otherwise rather lengthy accounts. The following section aims to discuss what function these passages have within their larger context and what role individuals, above all Muḥammad, play in them. II.1 A threat to the Kaʿba Before we cast a look upon the destruction of pre-Islamic loca sancta in the first decade after the hijra, it is important to note that even before Muḥammad’s birth, his biography was already closely linked to a story describing a dangerous threat to the Kaʿba in connection with a relentless commitment to monotheism: Muslim tradition generally agreed on his birth in the so-called Year of the Elephant (ʿāmm al-fīl, 570 CE), named after the attempt of Abraha, the Ethiopian-born king of Saudi Arabia to destroy the Kaʿba. 113 The historians style Muḥammad’s paternal grandfather ʿAbd al-Muṭṭalib as the main opponent to Abraha’s attack and present him as the only one heroic enough to speak up against the intruder’s hybristic claim that God won’t be able to protect the Meccan sanctuary against the South Arabian military force, to which ʿAbd al-Muṭṭalib countered “that the House [i.e. the Kaʿba] has a Lord of its own who will deEarly traditions regarding Muḥammad’s birth in the Year of the Elephant are collected in Ṭabarī, Taʾrīkh 966–8. Modern scholarship assumes that the events of Abraha’s campaign happened either in 547 or in 552 and therefore hardly coincide with Muḥammad’s year of birth, cf. Conrad, Abraha and Muhammad 225–40 and Kister, The campaign of Ḥulubān 425–36. According to a tradition transmitted from Ibn al-Kalbī, Ṭabarī mentions at a different passage (Taʾrīkh 268) that Muḥammad was born in the forty-second year of the reign of Kisrā Anūsharwān, which suggests 573 as an alternative year of birth. For Q 105, cf. Neuwirth, Der Koran 112–18 and Sinai, Fortschreibung 68–74 concerning the intra-Quranic chronology of the mentions of Ibrāhīm. 113

48

KONSTANTIN M. KLEIN

fend it.” 114 On his way to Mecca, Abraha reached the nearby city Ṭāʾif. 115 In order to protect their sanctuary, the Ṯaqīf of Ṭaʾif became guilty of aiding the intruders and abetting, as they addressed the South Arabian ruler: “This house of ours (they meant [the house of the deity] al-Lāt) is not the House which you seek. You want the House which is at Mecca (they meant the Kaʿba), and we will send a man with you who will guide you.” 116 Consequently, the South Arabian army was led by a man from Ṭaʾif called Abū Righāl, with whose help they reached as far as two miles from Mecca, where the guide suddenly died – the campaign’s first omen of failure. 117 While Abū Righāl is at best a semi-legendary figure, the story about him directing Abraha to the city may have been elaborated as part of a later anti-Ṯaqafī bias, hostile to the role of prominent men from Ṭāʾif. However, it also shows that within the Ṭabarī, Taʾrīkh 939: ‫وأ ّن للبيت ربّا سيمنعه‬. ʿAbd al-Muṭṭalib affirms his trust in God by asking him to defend his sacred area (ḥimā) and to repel the enemy of the House, cf. Taʾrīkh 940. Cf. Taʾrīkh 1088 for ʿAbd alMuṭṭalib also being the first person to adorn the Kaʿba with gold. On his role in the Sīra, cf. Khalidi, Images of Muhammad 65–7. 115 Ibn Hishām, Sīra 32: ‫�ت بيت لهم بالطائف كانوا يعظّمونه نحو تعظيم الكعبة‬ ٌ ُ �� ‫( وال‬As to al-Lāt it was a temple of theirs in al-Tāʾif which they used to venerate as the Kaʿba is venerated). Cf. Crone, Meccan trade 151–6 on religious fairs in the vicinity of Ṭāʾif and Mecca. Note that Mecca itself was no pilgrimage center in pre-Islamic times, still recognizable as most of the rituals of pilgrimage are conducted outside town, cf. Crone, Meccan trade 172–7, Donner, Muhammad and the beilievers 29 and Wellhausen, Reste 79–81. On the sanctuary of the deity al-Lāt in Ṭāʾif itself, cf. the summary of the Arabic literary sources in Krone, Die altarabische Gottheit al-Lāt 372–6. 116 Ṭabarī, Taʾrīkh 937: ‫وليس بيتنا هذا بالبيت الذى تريد يعنون ال��ت ا ٕنما تريد البيت الذى‬ ‫بم ّكة يعنون الكعبة ونحن نبعث معك من يدلّك‬. Cf. Ibn Hishām, Sīra 32. 117 Ṭabarī’s narrative continues with the observation that the Arabs subsequently hurled stones at Abū Righāl’s grave in al-Mughammis until the present day. This practice must have developed early, as it is attested in a verse by the Umayyad poet Jarīr (d. 728/279 CE) who brags about enjoying to stone his enemy Farazdaq’s tomb in a similar way as the Arabs stone Abū Righal’s alleged burial place (cited by Azraqī, Akhbār Makka 93); cf. Tottoli, Abū Righāl, and Bosworth, The History 223 n. 552. 114

THE SILENCE OF THE GODS

49

historical narration the inhabitants of Ṭāʾif were styled as collaborators against the Kaʿba from a very early point onwards. This behaviour would, in the course of history, become apparent in their refusal to take Muḥammad under their protection, and would eventually be punished by the violent destruction of their sanctuary of al-Lāt through early Muslim forces. 118 Whereas Muḥammad is thus linked via his direct ancestors with a brave advocacy of the Kaʿba, it is not only Muḥammad’s family that excelled as campaigners for a monotheistic faith: the prophet’s long line of ancestors (going back to Adam) contains several other fighters against idolatry – sometimes in accordance with the biblical sources, but occasionally in distinct Late Antique versions of the Old Testament narrative. II.2 Biblical predecessors: Abraham, Moses and David Whereas the biblical text does not, for instance, include much information on Abraham’s father Teraḥ, 119 Jewish tradition presents him as a producer of idols or as a magnate of the polytheist king Nimrod. 120 The following episode mentioned in the Midrash Genesis Rabba takes place in Teraḥ’s shop: R. Ḥiyya said: Teraḥ was a manufacturer of idols. He once went away somewhere and left Abraham to sell them in his place. [...] A woman came with a plateful of flour and requested him, ‘Take this and offer it to them.’ So he took a stick, broke them, and put the stick in the hand of the largest. When

For the destruction of al-Lāt’s sanctuary in Ṭāʾif, cf. Ibn Hishām, Sīra 917–18, Ṭabarī, Taʾrīkh 1692 and Ibn al-Kalbī, Kitāb al-Aṣnām 10, 15– 16, discussed below. 119 Cf. Genesis 11:26–32, however, being a descendent of Shem, Teraḥ is marked as polytheist (cf. Genesis 11:10–26 and esp. Joshua 24:2: “And Joshua said unto all the people, Thus saith the Lord God of Israel, Your fathers dwelt on the other side of the flood in old time, even Teraḥ, the father of Abraham, and the father of Nachor: and they served other gods”). 120 Cf. Ginzberg, The legends of the Jews i, 185–98 and Ginzberg, The legends of the Jews v, 211–12 with notes 24–9. 118

50

KONSTANTIN M. KLEIN his father returned he demanded, ‘What have you done to them?’ “I cannot conceal it from you,’ he rejoined. ‘A woman came with a plateful of fine meal and requested me to offer it to them. One claimed, “I must eat first,” while another claimed, “I must eat first.” Thereupon the largest arose, took the stick, and broke them.’ ‘Why do you make sport of me,’ he cried out; ‘have they then any knowledge!’ ‘Should not your ears listen to what your mouth is saying.’ he retorted. Thereupon he seized him [scil. Abraham] and delivered him to Nimrod. 121

The Quranic references to Abraham/Ibrāhīm show narrative parallels to Genesis Rabba, one of the oldest exegetical Midrashim, with the basic alteration that Ibrāhīm’s father has a different name, Āzar. 122 The people of Ibrāhīm are styled as incorrigible idolaters just as those coming into Abraham’s shop in Genesis Rabba. According to Muslim tradition, Ibrāhīm mocked the very idols that his father produced. 123 In a clever taunt, he started breaking them up with an axe. When only the largest idol remained, Ibrāhīm tied the Genesis Rabba 38:13: ‫ חד זמן נפק‬,‫אמר ר’ חייא תרח עובד צלמים הוה‬ ‫ חד זמן אתת איתתא טעינא חד פינך דסלת‬.‫לאתר הושיב אברהם מוכר תחתיו‬ ‫ קם נסיב בוקלסה ותברהון ויהב ההוא בוקלסה‬,‫אמ’ ליה הא לך קרב קדמיהון‬ ‫ אמר ליה מה‬,‫ כיון דאתא אבוה אמר ליה מה עבר להון כרין‬.‫בידוי דרבה רבהון‬ ‫ דין‬,‫נכפור לך אתת חדא איתתא טעינא חד פינך דסלת ואמרת לי קרב קומיהון‬ ‫ קם הדין רבה נסיב בוקלסה‬.‫אמר אנא איכול קדמאי ודין אמר אנא איכול קדמאי‬ ‫ אמר ליה ולא ישמעו אזניך‬,‫ ידעין אינון‬,‫ ]אמר ליה[ מה את מפלה בי‬.‫ותברינון‬ ‫ נסביה ומסריה לנמרוד‬,‫מפיך‬. On structure, content and date of Genesis Rabba, cf. Stemberger, Introduction 300–8 who dates the final redaction of this Midrash to the first half of the fifth century. 122 Q 6:74: “[Remember] when Abraham said to his father Āzar, ‘Do you take idols as gods? I see you and your people in manifest error.” Ṭabarī, Tafsīr 6, 74 discusses whether “Āzar” is just a different name of “Teraḥ” (as “Israel” is a different name for “Jacob,” but designating the same person). 123 Cf. Ṭabarī, Taʾrīkh 256, 258 and 260 (following a different transmission-chain); cf. Q 21:58–9. Eventually, Ibrāhīm dissociated himself from his father Āzar, cf. Q 9:114: “When it became clear to him that his father was an enemy of God, he would have nothing to do with him.” 121

THE SILENCE OF THE GODS

51

axe to its hand and left them. The returning people became angry and astonished, until eventually Ibrāhīm’s doing was explained by his declaration that it was not him who broke the idols, but “[...] this chief one [...] did it. Ask them, if they can speak.” 124 Ṭabarī’s version, based on Ibn Isḥāq, adds: “He [scil. the largest idol] became very angry because you worshipped these little ones together with him while he is greater than they are, so he broke them.” 125 While the people of Ibrāhīm were subsequently confounded with shame, as they realized that their idols obviously could not speak, the episode’s problematic nature is hardly resolved when they retaliate by throwing Ibrāhīm into a burning fire from which only God’s omnipotence protects him. 126 Except for him standing up against idolatry, Ibrāhīm is, of course, credited with building the Kaʿba in Mecca. In a way, Muḥammad himself, whose direct decent from Ibrāhīm is reiterated in the entire Muslim biographical and historiographical tradition, is described as the typological equivalent of his ancestor. Ibrāhīm’s destruction of idols develops a pattern for subsequent Islamic idoloclastic endeavors. The second famous fighter against idolatry in all three Abrahamic religions is Moses rebuking the Children of Israel for worshipping the Golden Calf. However, the figure of Mūsa in the Quran and Muslim historical tradition is much more defined as a lawgiving forerunner of Muḥammad who parallels the latter by being sent by God to warn the Pharaoh. 127 The destruction of the Golden Calf, which was burnt and had its ashes scattered in the sea Q 21:62–3. Ṭabarī, Taʾrīkh 260 ‫غضب من أن تعبدوا معه هذه الصغار وهو أكبر منها فكسرهن‬. Parallels from Jewish traditions are collected in Ginzberg, The legends of the Jews i, 197–8. 126 Q 21:68–70, stating that God made the people of Ibrāhīm the “great losers.” 127 Cf. Q 7:138: “We brought the Children of Israel across the sea, and they came to a people who were devoted to some idols that they had. They said, ‘O Moses, make for us a god, as they have gods.’ He said, ‘You are a people who are ignorant.’” For an account of the destruction of the Golden Calf, cf. Ṭabarī, Taʾrīkh 1, 495. 124 125

52

KONSTANTIN M. KLEIN

according to Bible and Quran, 128 remained the recommended way for destroying idols according to the rules of the Mishna. 129 It does, however, not seem to have been formative for the later Islamic destructions in the time of Muḥammad: while idols were often burnt, the disposal of their remains does not seem to have been of particular importance. 130 Many biblical narratives in the Muslim tradition derive from the works of Wahb ibn Munabbih, a transmitter of Judeo-Christian episodes presented within an Islamic framework. 131 While most of the works written by Wahb are no longer extant, much can be deduced from passages quoted by later historians. A papyrus, now kept in Heidelberg constitutes one of the oldest of its kind in Islam and contains long passages of a Ḥadīṯ Dāwūd compiled by Wahb Q 20:97. Mishna, ʿAvoda Sara 3,3: ‫ צורת‬,‫המוציא כלים ועליהם צורת החמה‬ ‫ יוליכם לים המלח‬,‫ צורת הגרקון‬,‫הלבנה‬. (If one finds a vessel and on it a drawing of the sun or a drawing of the moon, or a drawing of a dragon, he should cast them into the Salt Sea). 130 On destruction of idols with fire, cf. on al-Lāt (Ibn al-Kalbī, Kitāb al-Aṣnām 10, 15–16), on Dhū ʾl-Khalaṣa (Kitāb al-Aṣnām 22, 18–23, 10 and Ṭabarī, Taʾrīkh i, 1763 (quoting Wāqidī)) and on Dhū ʾl-Kaffayn (Ibn Hishām, Sīra 254). On the burning of household idols, cf. Sīra 335 and Ṭabarī, Taʾrīkh i, 1244. In contrast, cf. above for the meticulous disposal of soil and stones from the destroyed Roman temple replaced by Constantine’s Church of the Holy Sepulchre. The most notable exception, i.e. a tradition according to which the early Muslims paid slightly more attention to purify the place, is Ibn al-Kalbī’s account on the destruction of the idols in the Kaʿba where the statues are destroyed but burnt at a different place, cf. Kitāb al-Aṣnām 19, 16–20, 2. Ibn ʿAbd al-Barr (Istīʿāb 2, 776) recounts that Muḥammad told members of the Ḥanīfa to destroy their (Christian) church, to rebuild it as a mosque and to purify the soil with the water of is ablution, cf. Lecker, forthcoming 1. 131 Cf. Khoury, Wahb b. Munabbih. Regarding Wahb’s authorship of Ḥadīṯ Dāwūd, cf. Khoury, Wahb b. Munabbih 185–8. Cf. Ṭabarī, Taʾrīkh i, 1763 for his identically named ancestor as the first collector of Quranic knowledge in Ṣanʿāʾ, and Dhayl al-Mudhayyal 2493–4 on Wahb’s own biography. 128 129

THE SILENCE OF THE GODS

53

and some fragments of his rendering of Muḥammad’s Maghāzī. Wahb’s narration concerning the prehistory of Dāwūd’s fight against Jālūt (Goliath) modifies the biblical tradition of the capture of the Ark of the Covenant (al-Tābūt) by the Philistines. In the biblical narrative the Philistines take the Ark to several locations in their realms and at each location misfortune befalls them (tumors, a plague of mice, and the overthrow of the idol of Dagon in the temple of Ashdod). Eventually, they remorsefully return it to the Israelites. 132 In Wahb’s version, in contrast, the capture of the Ark is more closely connected with the person of Jālūt who is presented as a leader-like figure among the Amalekites. They bring the Ark into a village of Palestine: Then they put it into the house of their idols, and their idols were overthrown. And they had a large idol made out of gold; it was the largest idol in terms of size. It had two pupils made out of (red) rubies. That idol fell down in prostration before the Ark. And its pupils descended to its cheeks, while water flowed down out of them. When the custodians of the house of their idols entered and saw that, they tore their hair. 133

Cf. 1 Samuel 4–6. Wahb b. Munabbih, Ḥadīṯ Dāwūd 3, 19–4,5: ‫فوضعوه في بيت أصنامهم‬ ‫فاصبحت أصنامهم منكوسة وكان لهم صنم كبير من ذهب هو كبير أصنامهم بالضخم قد كان له حدقتان من‬ ‫ياقوتتين حمراوتين فخر ذلك الصنم ساجدا للتابوت وانحدرت حدقتاه على وجنتيه يسيل منهما ألما فلما دخلت سدنة‬ ‫بيت أصنامهم وراوا ذلك نتفوا شعورهم‬. Cf. Khoury, Wahb b. Munabbih 213–21 for Wahb’s biblical sources; Ṭabarī’s version, although mainly based on a different work by Wahb b. Munabbih, gives the story with varied motifs, Taʾrīkh i, 547–59. In his account, the anthropomorphic idol is placed on top of the Ark; several days the Amalekites find the constellation turned upside-down when they return the next morning. They eventually proceed to nail the idol on top of the Ark, just to then find one of its hands and both feet cut off and the idol cast down under the Ark, cf. Taʾrīkh i, 552; this narrative is thus closer to the Biblical version of 1 Sam 5:2–5. Cf. Ibn Hishām, Sīra 433 as well as Ṭabarī, Taʾrīkh i, 551 and 1, 1300 for typological parallels between the David’s battle with Goliath and the Battle of Badr (according to Muslim tradition, the number of Muḥammad’s soldiers equalled that of Saul’s army). 132 133

54

KONSTANTIN M. KLEIN

Moreover, the plague of mice mentioned in the biblical narrative is transformed into a violent attack of highly aggressive rodents sent by God: Wahb narrates how a mouse killed a man, and how subsequently numerous Amalekites died thereafter. 134 Finally, they attempted to rid themselves of the Ark: they initially tried to burn or smash it, but neither fire nor sword could destroy the chest. Eventually, they put it upon an oxcart that was returned to the Israelites by angels. The biggest alteration of Wahb’s narration, however, is ascribing the capture of the Arc from the Philistines to the Amalekites. By this, he re-connected the biblical past with the pre-Islamic tradition according to which the Amalekites were a people that controlled Mecca and the Ḥijāz until Ismāʿīl, the son of Ibrāhīm, ousted them. 135 II. 3 Gods in the oven and idols in the cesspit: pre-Islamic deities before the Conquest of Mecca Despite the large space that biblical fighters against idolatry occupy in Islamic historiography, early Islam was very successful in erasing most of the knowledge about the not so distant past, the time of jāhiliyyā. Neither the Sīra nor the works of the universal historians such as Ṭabarī contain much information on the veneration of preIslamic gods; we face a veritable damnatio memoriae in the sources. The main work of reference on the pagan days is the Book of Idols Wahb b. Munabbih, Ḥadīṯ Dāwūd 4, 6–12. While Ṭabarī notes that the Amalekites suffered from neck pain; he and his sources seem unaware of Wahb’s mice episode, Taʾrīkh 552–3. The Biblical rodents have often been connected to an unsuccessful attack against Egypt made by Sanacharibos mentioned in Herodotus, Historiae 2, 141, 2. However, this episode seems to be independent from the Biblical one; the mice as saviours of Egypt can be explained by the defending pharaoh’s preference for the god Horus who was associated with the shrewmouse; cf. Grabbe, Of mice and dead men 119–40 and Lloyd, Herodotus’ account 42–3. 135 Cf. Ibn al-Kalbī, Kitāb al-Aṣnām 3–4, cf. Nöldeke, Über die Amalekiter 25 and 29–42, esp. 36–7, who regarded Ibn al-Kalbī and his father themselves as the main originators of the belief of an Amalekite occupation of Mecca; cf. also Q 2:250–1 for Jālūt in the Quran, cf. Conrad, The Biblical tradition 281–7. 134

THE SILENCE OF THE GODS

55

(Kitāb al-Aṣnām) by Hishām Ibn al-Kalbī (c. 737–819). Writing some generations after the triumphal success of Islam, in a time when knowledge about the pre-Islamic lifestyle had become fashionable again, Ibn al-Kalbī collected the remaining information on ancient Arabian deities and cults. His work is generally considered as a reliable source and was used by many later historians. 136 One of the greatest difficulties posed by the Kitāb al-Aṣnām is that it is hard to date the information it provides: the worship of some of the deities mentioned in the text may have faded away well before the advent of Islam. However, it has recently been suggested that the amount of time between the end of idol worship and the beginning of scholarly interest in it was not very large. 137 In some cases the narratives of destruction (or their absence) can give us hints: for instance, it appears likely that the deities Hubal, 138 al-Lāt and Dhū ʾl-khalaṣa were venerated until the advent of Islam, as the destruction of their sanctuaries in Mecca, Ṭāʾif and Tabāla respectively plays a vital role in Islamic historiography. In general, however, ancient deities are not mentioned very often in the historical tradition outside the Kitāb al-Aṣnām. Ibn Hishām randomly notes: Along with the Kaʿba the Arabs had adopted Ṭawāghīt, 139 which were temples which they venerated as they venerated

Cf. Ṭabarī’s praise for Ibn al-Kalbī as the transmitter of experts on pre-Islamic history and lore, Taʾrīkh 719; cf. Atallah, al-Kalbī 2. Cf. Lecker, ʿAvodat 252 and Lecker, Wadd 127 who argues that there was no deliberate effort to hide information regarding idolatry, but rather differing points of interest, i.e. mainly into heroes, famous horses and camels from pre-Islamic times. 137 Lecker, Was Arabian idol worship declining 23–4; Susanne Krone, in contrast, suggests a much larger time of about hundred years, cf. Krone, Die altarabische Gottheit al-Lāt 20. 138 On the seeming contradiction of worship to Hubal in the Kaʿba while Allāh was the alleged supreme god of Mecca, cf. Crone, Meccan trade 192–4 who convincingly refutes Wellhausen’s assumption that Hubal and Allāh were actually identical, Wellhausen, Reste 75–6. 139 In the Quran the plural “ṭawāghīt” denotes pre-Islamic deities venerated by means of idols, such as al-Lāt and al-ʿUzzā, whereas the singular “ṭāghūt” is a synonym for the Devil (cf. Q 4:60; it can, however, also 136

56

KONSTANTIN M. KLEIN the Kaʿba. They had their guardians and overseers and they used to make offerings to them as they did to the Kaʿba and to circumambulate them and sacrifice at them. Yet they recognized the superiority of the Kaʿba because it was the temple and mosque of Abraham. 140

What reads like a Kaʿba-centered henotheism and a somewhat peaceful coexistence between the sanctuary and other holy places was about to change as a result of Islam’s clear commitment to a strict monotheism that offered no space for other deities. While the polemic against idolatry found shape in words – in the Quran and in ahādīṯ – it did, however, according to the historical tradition, not translate into ‘idoloclastic’ deeds before the early Muslims’ emigration to Medina. The majority of the historians recount an episode allegedly witnessed by ʿAlī when he stayed in Qubāʾ, a locality in the south of Medina, 141 shortly after the hijra. He noticed that a man used to visit an unmarried Muslim woman in the middle of the night, knocked on her door and delivered something to her. Naturally, ʿAlī felt suspicious about him and asked the woman for the meaning of these nightly visits. She told him that the man was Sahl ibn Ḥunayf ibn Wāhib who knew that she was on her own, and so he used to break up the former household idols of his tribe at night and bring her the pieces to use as fuel. 142 It has been suggested that one reason why we find more information on household idols in narrations concerning Medina, is that the conversion

denote diviners, soothsayers or magicians); cf. Fahd, Ṭāghūt and Atallah, Jibt et Ṭāghūt 69–82. 140 Ibn Hishām, Sīra 54–5: ‫بيوت تعظّمها‬ ٌ ‫يت وهي‬ َ ‫العرب قد ات�خَ َذ ْت مع الكعبة َط َو ِاغ‬ ُ ‫وكانت‬ ٌ ‫اب وت ُْهدي لها كما ت ُْهدي للكعبة وتطوف بها ك َط َوافها بها و َت ْن َح ُر عندها وهي َت ْعرف‬ ‫ج‬ ‫وح‬ ‫ة‬ ‫سدن‬ ‫لها‬ ‫الكعبة‬ ‫كتعظيم‬ � ُ ٌ ‫بيت ابراهيم ومسجده‬ ُ ‫ف ََض َل الكعبة عليها ��ٔنّها كانت قد عرفت أنها‬. 141 Cf. Watt and Winder, al-Madīna, 994–1007; esp. 1001. 142 Ibn Hishām, Sīra 335, for a slightly expanded version with passage of direct speech, cf. Ṭabarī, Taʾrīkh 1244. Besides Sahl ibn Ḥunayf, his fellow-Medinan ʿAbdallāh ibn Jubayr is also mentioned as using idols as firewood, cf. Balādhurī, Ansāb al-ashrāf 1, 265. Cf. Mishna, ʿAvoda Sara 3, 10 that specifies that using a pagan sacred tree as fuel is allowed. On the burning of idols, cf. Wāqidī, Kitāb al-Maghāzī 2, 870–1 with Lecker, ʿAvodat 257.

THE SILENCE OF THE GODS

57

of Mecca in 630 CE was a relatively immediate and allencompassing event, whereas conversion in Medina constituted a slow process accompanied by internal strife. 143 Michael Lecker compiled a substantial list of idols mentioned, venerated and sometimes destroyed in Medina. Not surprisingly, most of the Medinan idol breakers were among the earliest and perhaps most enthusiastic supporters of Islam before the Battle of Badr. 144 It is noticeable, however, that while the biographically arranged Kitāb aṭ-Ṭabaqāt alkubra by Ibn Saʿd provides much information about both idol worship and destruction in Medina, those historians who choose a continuous narrative for their works seemed only little interested in this topic. Apart from the idols used as fuel mentioned above, Ibn Hishām judged only one other episode worthy of being integrated in his recension of Ibn Isḥāq’s Sīra: It is again about one of the early converts who were so passionate in regards to their new monotheistic faith. The protagonist is a youth called Muʿādh, the son of ʿAmr ibn al-Jamūḥ, who himself belonged to a leading family of Medina. 145 While Muʿādh was present at the first meeting at ʿAqaba, his father kept to his idolatrous life and had a wooden idol of Manāf 146 in his house. Muʿādh and his friends sneaked in at night, carried the idol away and threw it on its face into a cesspit. The next morning, ʿAmr cursed the thieves and began the search for his idol. Finding it in the privy, he washed and perfumed it and put it up again in his house. Over the next nights the story repeated itself, until one day, after having cleaned and purified it, ʿAmr fasLecker, Was Arabian idol worship declining 25; cf. 25–37 for numerous other examples of household idols, idols of noblemen and tribal groups; cf. also Lecker, ʿAvodat 256. 144 Lecker, Was Arabian idol worship declining 29. 145 On the stand of ʿAmr in Medinan society, cf. Lecker, Was Arabian idol worship declining 29–30. Ṭabarī does not mention the story, however, Muʿādh ibn ʿAmr is named as being one of those who were present at ʿAqaba, while his father is referred to as remaining in his polytheistic religion, cf. Taʾrīkh 1227. On the entanglement of tribal leadership and idol worship in Medina, cf. Lecker, Muslims, Jews and pagans 44–5. 146 Ibn Hishām gives the name as “Manāt,” for “Manāf” as the correct form, cf. Lecker, Idol worship 337; see also Fahd, Le Panthéon 122–3. 143

58

KONSTANTIN M. KLEIN

tened his sword to it, saying: “If you are any good at all defend yourself since you have this sword.” 147 The next night, Muʿādh not only took the sword from the idol’s neck, he also hung a dead dog to it where ʿAmr had previously attached the weapon. 148 The shocking discovery the next morning constituted the moment in which ʿAmr realized the impotence of Manāf and the reason for him to convert to Islam. While it is possible that ʿAmr indeed owned an idol, just as other Medinan noblemen did, 149 the narration concerning his son’s adventurous tactic to enlighten the father and his polytheistic environment, is a clever adaptation of the episode of Ibrāhīm and his father’s idols, known from allusions in the Quran and from Midrashic lore. A report of Wāqidī, quoted by Qalqashandī, mentions that the ability to read and write was rare but did exist among the pre-Islamic inhabitants of Medina, and that a Jew from the Yahūd Māsika used to teach it to the Arab children. 150 Michael Lecker has argued that the literate elite of Medina, who helped Islam to obtain a foothold in the city, was educated by Jewish teachers. 151 While it cannot be completely excluded that a similar version of the Abraham story cited above from Genesis Rabba was known in Medina via Jewish tutors, the believers and the historians recording the story certainly knew the Quranic version crediting Ibrāhīm with the destruction of his father’s idols. Only a Ibn Hishām, Sīra 304: ‫فامتنع فهذا السيف معك‬ ‫كان فيك خي ٌر‬. ْ Cf. Ṭabarī, Taʾrīkh 943, where in one tradition of the Abraha legend mentioned above, Nufayl al-khaṯʿamī not only smeared the apse of Abraha’s church with excrements, but also gathered together some putrefying animal carcasses and threw them into it. 149 Lecker, Was Arabian idol worship declining 32 n. 174 mentions a tradition recorded by Abū Nuʿaym according to which an old woman stood behind the idol of Manāf in Medina and answered on its behalf whenever people wanted to consult it; cf. Dalāʾil 311. 150 Qalqashandī, Ṣubḥ al-aʿshā 3, 14–15; cf. Lecker, Zayd B. Thābit 264–5. In the explanation to some verses by Ḥassān ibn Ṯābit, a woman named al-Shaʿṯāʾ is mentioned, whose father was the head of the Jews who were in charge of the House of Torah study, cf. Iṣfahānī, Kitāb alAghānī xvii, 169–70. 151 Lecker, Idol worship 340–1 and esp. Lecker, Zayd B. Thābit 271. 147 148

THE SILENCE OF THE GODS

59

few details had to be altered to relocate the Biblical episode to Medina: in both narratives a young man is mocking his father’s and his environment’s beliefs, the idols themselves are harmed and something is attached to them, a pickaxe in the case of Ibrāhīm, and a sword replaced by a dead dog in the case of Muʿādh. It therefore is only a logical consequence that after recognizing the falsehood of his beliefs, ʿAmr thanked God “for having delivered him from the blindness and error in which he had lived hitherto.” 152 II.4 The impotence of the gods The triumphal procession of monotheism over the pagan deities arguably only started with the Conquest of Mecca and Muḥammad entering the Kaʿba. However, the traditions regarding the exact events diverge significantly from one another. Ibn Hishām has the triumphant prophet entering the sanctuary, where “he found a dove made of wood. He broke it in his hands and threw it away.” 153 Azraqī, in contrast, speaks of 360 idols that Iblīs had strengthened with lead. Muḥammad points at them with a stick and recites: “Truth has come and falsehood has vanished. Falsehood will always vanish,” after which the idols collapsed one after the other. 154 In Ibn al-Kalbī’s version, Muḥammad recites the same āya, but he pierces the idols’ eyes with the point of his arrow and then has them broken, removed and set on fire. 155 That in this version the idols are burnt at a different location to where they stood may result from the sacred status of the Kaʿba which should not be polluted by the overpowered gods – a vague echo of Constantine’s endeavors to have the demonically contaminated soil removed from the construction site of his basilica in Jerusalem. In this regard Ibn al-Kalbī’s testimony is unique – nowhere else do we read Ibn Hishām, Sīra 304: [...]‫مى والض��لة‬ َ ‫ويشكر الله الذي انقذ ُه م ّما كان فيه من ال َع‬. Cf. Q 6:74 (I see you and your people in manifest error) and 21:65 (They were confounded, and said, ‘You know that they cannot speak’). On the importance of mockery in early Islamic tales of conversion, cf. Lecker, ʿAvodat 256. 153 Ibn Hishām, Sīra 821: ‫فوجد فيها حمام ًة من عيدان فكسرها بيده ثم طرحها‬. َ ََ 154 Azraqī, Akhbār Makka i,70; cf. Q 17:81. 155 Ibn al-Kalbī, Kitāb al-Aṣnām 19, 16–20, 2. 152

60

KONSTANTIN M. KLEIN

of a disposal separated from the act of destruction. However, I would argue that concerns like these were relatively unimportant for early Muslim historians. The essential point they repeatedly made was that the pre-Islamic idols were void and only empty names, just as the Quran had revealed it, 156 and, above all, that they were without any real power. Deaf, dumb, speechless and sightless, [an idol] can neither hurt nor benefit. It cannot help you at all to dispense with God. You decorate it with gold and silver and make it a temptation unto man. You worship it, not God, and you force humans to worship it. You call it the Lord. 157

Ṭabarī, employing material from Ibn Isḥāq and Wahb ibn Munabbih and using Quranic language, 158 puts these words, which summarize the early Muslim’s attitude towards pagan deities, into the mouth of a Christian saint: Jirjīs (St George), who in this narrative attacks the priests of Apollo for their vain idolatry. He is a figure that was (and is) venerated widely in the Near East regardless of the religion of his devotees, and who was sometimes identified with Khiḍr. 159 Ṭabarī’s narrative continues with Jirjīs being forced to sacrifice to the (Roman) idols, but in the moment when he should make his offerings, they started moving, trembled and rolled down towards the saint. Eventually, the Devil jumped out of Cf. esp. Q 7:71 (Do you argue with me about names that you named?), Q 22:62 (because what they invoke, to His exclusion, is false), Q 31:30 (what they invoke apart from Him is the False) and Q 53:23 (They are merely names which you and your forefathers have bestowed. God has sent down no authority in them.). 157 Ṭabarī, Taʾrīkh 797: ‫�� ينطق و�� يبصر و�� يسمع و�� يضر و�� ينفع و�� يغنى عنك من الله‬ ُ ّ ‫ٔجبرت عليه عباد الله ودعوتَه ربّا‬ َ ‫شي ًئا فزيّن َته بالذهب والفضّ ة لتجعله ف ْتنة للناس ثم عبدتَه دون الله وا‬. 158 Cf. Q 2:18. 159 The Arabic traditions concerning Jirjīs, based on a variety of Latin, Greek, Syriac, Coptic and Ethiopic material, is a vast field that cannot be discussed in detail here; cf. Budge, George of Lydda (mainly on the Coptic and Ethiopic material; on the Arabic versions, cf. 63–6) and Brooks, The Acts (on the Syriac Acts of St George); cf. also Franke, Begegnung mit Khidr 88 n. 201. 156

THE SILENCE OF THE GODS

61

one statue. Ṭabarī concludes that “since then Satan has not entered the inside of any idol, out of fear of collapse; nor will he ever do so, as people say.” 160 While the motif of a saint being forced – sometimes under torture – to confess his allegiance to a deity is very common in almost all Christian Martyr’s Acts as well as in several Islamic narrations, 161 Jirjīs and his self-mutilating idols share similarities to an episode in the Life of Porphyry in which the narrator recalls: When we arrived at the place where the statue of Aphrodite stood (Christians carried the precious wood of Christ, that is to say the sign of the cross), the demon who dwelled in the statue, unable to bear the sight of the dreaded sign, went out of marble with a great uproar, smashed the marble and broke it into pieces. 162

Ṭabarī, Taʾrīkh 810–11: ‫جوف صنم مخاف َة الخسف و�� يدخله‬ َ ‫فما دخل ا ٕبليس منذ يومئذ‬ ‫بعدها فيما يذكرون أبدًا‬. 161 The most famous ordeal certainly is that of the slave Bilāl, who is tortured by Umayya ibn Khalaf by being thrown on his back and having a great rock put upon his chest during the hottest hours of the day until he would deny Muḥammad and worship al-Lāt and al-ʿUzzā. According to Waraqa ibn Nawfal, Hilāl continued to utter “One, one,” thus proving his commitment to Islam; cf. Ibn Hishām, Sīra 205. 162 Marc the Deacon, Vita Porphyrii 61: ὡς ἐϕϑάσαμεν τὸν τόπον ἔνϑα ὑπῆρχεν τὸ εἰρημένον εἴδωλον τῆς Ἀϕροδίτης (ἐβάσταζον δὲ Χριστιανοὶ τὸ τίμιον ξύλον τοῦ Χριστοῦ, τουτέστιν τὸν τύπον τοὺ σταυροῦ), ἑωρακὼς ὁ ἐνοικῶν δαίμων ἐν τῇ στήλῃ, μὴ ϕέρων ἰδεῖν τὸ ϕοβερὸν σημεῖον, ἐξελϑὼν ἐκ τοῦ μαρμάρου μετὰ ἀταξίας πολλῆς, ἔρριψεν αὐτὴν τὴν στήλην καὶ συνέκλασεν αὐτὴν εἰς πολλὰ κλάσματα. On the powerful sign of the cross, cf. Agathangelos, Historia armeniaca 779: 160

Իսկ սուրբն Գրիգոր իբրև տեսեալ զայն՝ զնշան տէրունական խաչին առնոյր և դիմէր ի դուռն մեհենին. և ամենայն շինուածք մեհենին ի հիմանց դղրդեալ տապալեցան. և լուցեալ յանկարծօրէն փայտակերտն հրդեհեցաւ ի տէրունական նշանին զօրութենէ, և ծուխն ծառացեալ մինչև յամպս հասանէր: (But Saint Gregory, when he saw this [scil. the

demons’ attack] made the sign of the Lord’s cross and ran to the door of the temple. Then the whole edifice of the temple shook from its founda-

62

KONSTANTIN M. KLEIN

While Ṭabarī mentions that Satan did not enter any idols after the confrontation with Jirjīs, it is in fact the almost complete absence of deities that defines most references to idols in the Muslim historical tradition. In Medina, ʿAmr had challenged his idol of Manāf to defend itself since it had a sword 163 – but Manāf, as almost all other idols, failed this trial. Already in pre-Islamic times, entire tribes realized that they could even curse their gods without consequences. The son of the last guardian of the deity Wadd recounted how he used to drink the milk sacrifice that he should have brought to the idol, thus mocking the statue’s inability to consume the offering. 164 A certain Abū Ṭalḥa from Medina wanted to marry a woman, who rejected his proposal on the grounds that he was an unbeliever. She too blamed him for worshipping a stone that could neither harm nor do good and a piece of wood that was specially made for him – the same form of reproach Ibrāhīm used when people came to his father’s shop to buy idols the latter had manufactured. Abū Ṭalḥa, however, decided to convert to Islam, and his bride-to-be accepted this as her dowry. 165 The accusation of defenselessness towards statues is, of course, much older than Islam and not peculiar to the Muslim historical tradition; Minucius Felix had ridiculed it already in the late second century: Mice, swallows, falcons know that they [scil. the statues of the deities] have no feeling: they gnaw them, they trample on them, they sit upon them; and unless you drive them off, they build their nests in the very mouth of your god. Spiders, indeed, weave their webs over [an idol’s] face, and suspend their

tions and collapsed. Suddenly catching fire the wooden construction burned by the power of the Lord’s sign, and the smoke rose up to the clouds). 163 Ibn Hishām, Sīra 303–4. Cf. Ginzberg, The legends of the Jews i, 195–7 for a similar question in the Jewish tradition. 164 Ibn al-Kalbī, Kitāb al-Aṣnām 34, 10; cf. Fahd, Le Panthéon 183, who calls the episode a ‘préparation coranique,’ and Lecker, Wadd 126. 165 Ibn Saʿd, Ṭabaqāt 8, 425–6; cf. Lecker, Was Arabian idol worship declining 26 with n. 134.

THE SILENCE OF THE GODS

63

threads from its very head. You wipe, cleanse, scrape, and you protect and fear those whom you produce. 166

Similar accusations can be found in the writings of Tertullian, Rufinus or Prudentius. Apart from idols not being able to defend themselves, an additional topos in the Muslim sources it, that the ancient deities failed to influence the lives of humans: several stories recount how people questioned the oracles, that were often connected to important shrines, however, when they acted against the provided prophesies, they still gained success in their endeavors. 167 The general passivity of the gods therefore was, according to the Sīra, the reason for the movement of pre-Islamic monotheists (ḥunafāʾ, sing. ḥanīf): They were of the opinion that their people had corrupted the religion of their father Ibrāhīm, and that the stone they went round was of no account; it could neither hear, nor see, neither hurt, nor help. 168 The ḥanīf Zayd ibn ʿAmr ibn Nufayl is credited with verses that state that he venerated Hubal only when he was young and his wit was slow (idh ḥilmī yasīru), 169 thus equating polytheist worship with a less-developed form of religion. Shlomo Pines investigated Late Antique notions of intellectual superiority of Late Antique Jews over Arabs: whereas Isḥāq (Isaac) was born only after Ibrāhīm’s covenant with God, his older brother Ismāʿīl, the mythical ancestor of all Arabs, was born when his father was Minucius Felix, Octavius 24, 9–10: mures, hirundines, milvi non sentire eos sciunt norunt: inculcant insident ac, nisi abigatis, in ipso dei vestri ore nidificant; araneae vero faciem eius intexunt et de ipso capite sua fila suspendunt. vos tergetis mundatis eraditis et illos, quos facitis, protegitis et timetis. Cf. Windau, Minucius Felix on the date of the Octavius. 167 Cf., for instance, Ibn al-Kalbī, Kitāb al-Aṣnām 29, 7–30, 8 for the deity Dhū ʾl-Khalaṣa’s oracle proved wrong several times. 168 Cf. Ibn Hishām, Sīra 143: ‫فقال بعضهم لبعض تعل�موا والله ما قومكم على شي ٍء لقد‬ ُ ‫دين ابيهم ابراهيم ما َح َج ٌر ن َِطيف به �� يسمع و�� ُي ْب ِص ُر و�� َي ُض �ر و�� َي ْنف َُع‬ َ ‫ا�خْ َط ُؤوا‬. Cf. Rubin, Ḥanīfiyya 99–102 and passim on the ḥunafāʾ. 169 Ibn al-Kalbī, Kitāb al-Aṣnām 14, 1–5. Cf. Kitāb al-Aṣnām 12, 6–7 where Muḥammad is mentioned stating that he had offered a white ewe to al-ʿUzzā in a time “‫ وأنا على ِدين قومي‬,‫( ”لقد أ ْه َد ْي ُت لل ُع �زى شا ًة عفرا َء‬when I was [still] following the religion of my people). Cf. Ṭabarī, Dhayl al-Mudhayyal 2321 on Zayd b. ʿAmr b. Nufayl. 166

64

KONSTANTIN M. KLEIN

still in a state of ignorance. 170 While Patricia Crone and Michael Lecker demonstrated that pagan belief in pre-Islamic Arabia was far from suffering a spiritual crisis and rather prospered on the eve of Islam, 171 this is, however, precisely the imagery with which Muslim historians chose to depict the situation on the Arabian Peninsula in the decades before the arrival of Muḥammad. II.5 The destiny of the gods after the Conquest of Mecca At a first glance the accounts on the destruction of idols and their shrines in the vicinity of Mecca, where Muḥammad had already purified the Kaʿba, read very similar to the episodes mentioned so far. When the important military commander ʿAmr ibn al-ʿĀṣ came to Ruhāṭ, the place of the sanctuary of the deity Suwāʿ, its keeper asked him what the reason of his visit was. ʿAmr answered upfront that he came to destroy the idol. “You cannot destroy it,” 172 the keeper’s replied in vain, as ʿAmr suited actions to his words and proved him wrong by demolishing Suwāʿ. Seeing the impotence of his deity, the keeper adopted Islam. The Ṯaqīf of Ṭāʾif were warned not to intervene in the destruction of their idol of al-Lāt which is called worthless (hadar): Don’t help al-Lāt for God is about to destroy her. How can one who cannot help herself be helped? 173

When Khālid ibn al-Walīd set out to destroy the sanctuary of the idol al-ʿUzzā in the valley of Nakhla, he and the guardian of the idol engaged in a veritable poetical agon. The guardian, a man from the Banū shaybān of the Banū Sulaym (Ibn al-Kalbī gives his name Pines, Jāhiliyya 184–6. Cf. Crone, Meccan trade 241 and Lecker, Was Arabian idol worship declining 35–7. 172 Ṭabarī, Taʾrīkh 1648–9; cf. also Wāqidī, Kitāb al-Maghāzī 2, 870; Suwāʿ is mentioned in Ibn al-Kalbī, Kitāb al-Aṣnām 6, 11–13 and 35,11–16 albeit without reference to the idol’s destruction; on the location of Ruhāṭ, cf. Yāqūt, Muʿjam al-buldān 2, 878. 173 Ibn Hishām, Sīra 871: ‫صر‬ َ �� ‫�� َت ْن ُصروا ال‬. ُ ِ ‫�ت أن الله ُم ْهلِكُها | وكيف ُي ْن َص ُر َمن هو ليس َي ْن َت‬ For very similar verses, cf. Ibn al-Kalbī, Kitāb al-Aṣnām 11, 3–5: ‫�ت‬ َ �� ‫�� َت ْن ُصروا ال‬ ‫إ �ن الله ُم ْهلِكُها | وكيف ن َْص ُرك ُُهم َمن ليس َي ْن َت ِص ُر‬. 170 171

THE SILENCE OF THE GODS

65

as Dubayya ash-shaybānī) tried in vain to harangue the goddess proclaiming that if she would prove unable to kill Khālid, she should disappear in shame: O al-ʿUzzā, make an annihilating attack on Khālid, Throw aside your veil and gird up your train. O al-ʿUzzā, if you do not kill this man Khālid Then bear a swift punishment or become a Christian. 174

While in Ibn al-Kalbī’s version of the narrative, al-ʿUzzā indeed makes an appearance in the shape of a ghastly witch with gritting teeth, and Khālid had to resolve the dramatic scene by splitting her head and killing the guardian, Ibn Hishām’s version only mentions that the idol was demolished by Khālid who then returned to Muḥammad. While indeed the emotions attributed to the polytheists in all these narratives indicate that populations of the Arabian Peninsula in the seventh century believed in the power of their idols, 175 I would argue that in terms of narrative structures this was no longer an important theme in historical writing which by this point of the narrative had focused to describe the political triumph of Islam. Undoubtedly, notions of the new God’s superiority over multitudes of impotent pagan deities obviously continued to play a role in all texts taking place after the Conquest of Mecca. Nevertheless, this is the reason why it took Ibn Hishām’s Muḥammad to break only one single idol and through this all other deities surrendered (in a way not dissimilar to Eusebius’ assertion that the great Pan died merely because Jesus Christ walked in this world). Azraqī, by contrast, whose focus is concerned with recording the exact past of the city of Mecca, reports the ‘precise’ number of 360 idols. After the Conquest of Mecca, the narrative moves away from the objects of destruction and focuses on those who destroy them. Ibn Hishām, Sīra 839–40: ‫يا ُع �ز شُ د�ي شَ �د ًة �� شَ َوي لها | على خالد اَ ْل ِقي ال ِق َنا َع‬ ‫وشَ �مرِي || يا ُع �ز أن لم تَ ْق ُتلي المر ِء خال ًدا | ف ُبو ِءي بِ ِٕا ْث ٍم عاجلٍ او َت َن �صرِي‬. The scene is narrated in a longer and more dramatized version in Ibn al-Kalbī, Kitāb al-Aṣnām 15, 7– 17, 2 (cf. esp. 16, 6–7 for a different version of the poem: �� ‫ شُ د�ي شَ �د ًة‬,ُ‫ا� ُع �زاء‬ ِ ‫ ;)تُ َك �ذبِي | على خالد! ا� ْل ِقى‬cf. also ‫الح َما َر وشَ �مرِي! || ف ٕان ِ�ك ا ٕ ��� َت ْق ُتلِي اليوم خالدًا | َت ُبو ئِي ب ُذ �ل عاج��ً و َت َن �صرِي‬ Sīra 32 on the shrine of al-Lāt in general. 175 Lecker, Was Arabian idol worship declining 36–7. 174

66

KONSTANTIN M. KLEIN

The majority of these narratives can be read in the light of tests and expressions of loyalty – to the new faith, to Muḥammad, and to the community of believers: When Jarīr ibn ʿAbd Allāh came to Muḥammad in order to embrace Islam, the latter prompted Jarīr to destroy the idol of Dhū ʾl-Khalaṣa as his first task as a Muslim. According to the sources, the new convert had to risk his life and met strong resistance from tribes loyal to the deity, he had to kill several hundreds until his mission was completed and the stone idol of Dhū ʾl-Khalaṣa found its final resting place as the threshold of the mosque at Tabāla. 176 In the cases of Ṭufayl ibn ʿAmr ad-Dawsī and Mustawghir ibn Rabīʿa two converts were ordered to destroy the idols of their tribes, the Daws and the Banū Rabīʿā ibn Kaʿb respectively. 177 While Ibn al-Kalbī records that Ṭufayl was sent by Muḥammad to Dhūʾl-Kaffayn, Ibn Hishām’s version explicitly notes that he had pleaded with Muḥammad until he received permission to destroy the idol, possibly as a proof of his commitment. The latter version quotes poetry that Ṭufayl allegedly spoke when setting the idol on fire: Not of your servants am I, Dhūʾl-Kaffayn, Our birth is far more ancient than thine. To stuff this fire in your heart I pine. 178

Proofs of loyalty such as these did, according to the Sīra reach private households as well; the same Ṭufayl ibn ʿAmr ad-Dawsī asked his wife, who also wanted to accept Islam, to cleanse herself from the deity Dhūʾsh-sharā that she revered. 179 While within the narraIbn al-Kalbī, Kitāb al-Aṣnām xxii, 18–23, 10; cf. Ṭabarī, Taʾrīkh i, 1763 who quotes Wāqidī. 177 On Ṭufayl and the idol of the Daws, Dhū ʾl-Kaffayn, cf. Ibn alKalbī, Kitāb al-Aṣnām 24, 2–6; cf. Ibn Hishām, Sīra 252–4 on the conversion of Ṭufayl ibn ʿAmr ad-Dawsī. On Mustawghir and the idol of the Banū Rabīʿa ibn Kaʿb, Ruḍāʾ, cf. Sīra 56 and Kitāb al-Aṣnām 19, 2–9. 178 Ibn Hishām, Sīra 254: ‫ك �فين َلست من عبا ِدكَا || مي��دنا اق َدمر من ِمي��َ ِدكا | اني‬ ُ ْ ْ َ ‫يا ذا ال‬ ُ �ُ ‫حشوت النا َر في فُ َؤا ِدكا‬. ُ 179 Cf. Ibn Hishām, Sīra 253–4. On the importance of idols in daily life and the deep emotional connection with them, cf. Lecker, ʿAvodat 264. 176

THE SILENCE OF THE GODS

67

tive frameworks of the triumphal history of Islam ‘idoloclasm’ offered the authors the possibility to highlight particularly successful and devout Muslims, overzealous behavior risked achieving the opposite effect: when in Ibn Hishām’s Sīra Muḥammad sent out troops to the district around Mecca inviting the tribes to Islam, he explicitly ordered his missionaries not to fight. Khālid ibn al-Walīd, however, beheaded some people of the Banū Jadhīma and was strongly rebuked by Muḥammad. 180 Consequently, Khālid’s campaign to destroy the idol of al-ʿUzzā in the valley of Nakhlā offered him the possibility to cut a better figure and demonstrate obedience to Muḥammad who, in Ibn al-Kalbī’s version, had to approve every single step in the wake of the destruction. 181 Khālid passed this and all subsequent tests when he was sent on a second and third mission to teach the Quran, the sunna and the requirements of Islam (maʿālim al-Islām) in Najrān and further south in the Yemen. 182 The sources speak of several idols being destroyed by Khālid, and Ibn al-Kalbī reports that his father had met the last guardian of the deity Wadd, Mālik ibn Ḥāriṯa al-Ajdārī, who told him how he witnessed the destruction of the idol: “Later I saw Khālid ibn al-Walīd as he broke and smashed it into pieces.” 183

III CONCLUSION

The literary motif of the destruction of holy places shows striking similarities in Christianity and Islam (and to a lesser extent in Late Antique Judaism): There is, first and foremost, the numeric parallel – both in Christian and in Muslim tradition relatively few cases of Ibn Hishām, Sīra 833–4. Ibn al-Kalbī, Kitāb al-Aṣnām 15, 7–17, 2; cf. Ṭabarī, Taʾrīkh 1,1648 and Wāqidī, Kitāb al-Maghāzī iii, 873–4. 182 Ṭabarī, Taʾrīkh 1724–6 and 1731–2. 183 Ibn al-Kalbī, Kitāb al-Aṣnām 34, 9–11, esp. 11: ‫ثم رأيت خالد بن الوليد بع ُد‬ ُ ‫ك ََس َر ُه فجعله ُج َذاذًا‬. The narrative continues, after a short digression, with Ibn alKalbī’s father inquiring from Mālik for a description of how the idol looked like, cf. Kitāb al-Aṣnām 35, 6–9; cf. the description of the idol of Wadd transmitted by Sharqī ibn Quṭāmī, one of Ibn al-Kalbī’s teachers (cf. Ibn Abī ʾd-Dunyā, Ishrāf fī manāzil al-ashrāf 248; quoted in Lecker, Wadd 126–7). 180 181

68

KONSTANTIN M. KLEIN

destructions of sanctuaries are recorded. However, these few incidents appear to fulfill an important role since they were almost all recorded more than once, sometimes in up to five or more versions with minor differences. Just as in Late Antique Christian church histories the involvement of the emperor is explicitly stressed, in the Muslim tradition it is always Muḥammad who ordered and supervised the destructions. In both cases, these acts of violence against pagan holy places have a deictic function by projecting the individual conversion of the person destroying the sanctuary into a larger moral reformation. 184 The motif of destruction plays a lesser role in historiographical works than, for instance, in the Sīra or in biographical works. Moreover, all subsequent destructions carried out by the followers of Muḥammad after the Conquest of Mecca echo the latter’s cleansing of the Kaʿba. It is worth noting that apart from his most important act, Muḥammad otherwise is rather styled as someone who regulated destructions. All accounts put much emphasis on describing how the physical act of the destruction of a holy place is carried out by Khālid, ʿAmr, Ṭufayl, Mughīra or Abū Sufyān, but that the order for doing so exclusively came from above – in a way mirroring how the Christian Church histories stressed the emperors’ legislation and downplayed the role of individual anti-pagan fighters. For learned pagans of the Greco-Roman world the temples resembled the “eyes of the cities” 185 and had an aesthetic value that was deemed worthy of preservation even by means of laws decreed by Christian emperors. This was not the case for pre-Islamic Arabia. Even though we find an exception in the ruler Abraha, who had his cathedral in Ṣanʿāʾ decorated with marble and gilded beams, 186 it appears that in general no sanctuary in pre-Islamic Arabia was renowned for its aesthetic value. 187 While we sometimes 48.

184

Eusebius, Vita Constantini iii, 55, 57–8 with Gotter, Rechtgläubige

Libanius, Oratio 30, 42. Ibn al-Kalbī, Kitāb al-Aṣnām 34, 6. 187 One might add those accounts describing how the pre-Islamic Kaʿba in Mecca was adorned with the Muʿallaqāt poems. However, their name, meaning “suspended,” appears only from the late ninth century 185 186

THE SILENCE OF THE GODS

69

encounter grief for the killed guardians of the sanctuaries, apparently nobody felt sorry for the destroyed holy places as such. It is not surprising that our knowledge of the architectural features of pre-Islamic loca sancta in the Arabian Peninsula derives almost exclusively from these few literary sources; an exception might be the sanctuary of Dhū Samāwī excavated in Najrān, called Kaʾbatān (Kʾbtn, with an Alif instead of ʿAyn), which may be identified as the Kaʿba of Najrān of the Banū ʾl-Ḥāriṯ mentioned by Ibn al-Kalbī. 188 To stay in Najrān – excavations revealed that in Umayyad times a miḥrāb was built into an older structure, however, the alleged Kaʿba of Najrān apparently never was turned into a mosque. The cases of pre-Islamic pagan loca sancta in the Arabian Peninsula that were converted into Islamic holy places are few and seem to be restricted to holy trees: while sacred stones (with the exception of a very important one in Mecca, of course) seem to have been avoided, some trees were saved into Islamic times by ascribing traditions to them that Muḥammad had rested in their shadow. But even within these few cases one can observe that these traditions were not kept for long. Quite noticeably, the old gods were nothing to be proud of, and we are left with the image of a religiously empty Arabian peninsula, at least in the imagery with which Muslim historians chose to describe the situation in the decades before the arrival of Islam. As we have seen, the sanctuaries were as a rule never reused. The only exception might be the above mentioned idol of the deity Dhū ʾl-Khalaṣa that was integrated as a spolium in the mosque of Tabāla. 189 Here we might draw a final comparison to the destruction of the Serapeion in Alexandria about 250 years before Jarīr ibn ʿAbd Allāh destroyed Dhū ʾl-Khalaṣa. The great temple of Alexandria was built on the city’s only natural elevation. Its onwards. The claim that the poems were hung up in the Kaʿba in honour of their prizewinning authors, must be discharged as a folk-etymology. The custom is not attested in any early source, and it therefore appears more likely that the name means “esteemed;” cf. Kister, The seven odes 27–36 and Beeston, The “Muʿallaqāt” 111–13. 188 Cf. Ibn al-Kalbī, Kitāb al-Aṣnām 28, 8 with Robin, Nagrān 51–3. 189 Ibn al-Kalbī, Kitāb al-Aṣnām 22, 18–23, 10 and Ṭabarī, Taʾrīkh i, 1763.

70

KONSTANTIN M. KLEIN

destruction signified a noticeable change for the cityscape as well as a significant display of how power in the city shifted from pagans to Christians. In Alexandria, at the end of the fourth century that was so much shaped by inter-religious violence, the latter visibly emerged as the victorious party. It is possible to interpret Libanius’ oration For the temples not only as a plea to conserve the shrines but also to eventually re-use them, as auditoria, courthouses or as gathering places for the rural population. 190 Surely, the famous orator from Antioch would not have thought of a re-use as churches, but this is exactly, what for instance Gregory the Great in the West suggested: if they were solidly built, existing temples should not be torn down but rather used for the veneration of the true God. 191 Modern scholarship has repeatedly emphasized how impressive the transformation of the Serapeion into a Christian church must have been. 192 However, archaeological excavations revealed quite astonishing evidence: No church was ever erected within the Serapeion (neither in the cella nor in the temple complex’ courtyard), the new Christian building was located outside the pagan sacred precinct. As in all other Greek and Roman temples, the religious services in the Serapeion took place outside in the open. The cella of the gigantic temple complex therefore only had a width of nine meters, too small to use it as an adequate church. 193 Whether this was the only reason for not converting the temple building itself into a church, remains doubtful. Usually the Late Antique fear of demons, dwelling in a pagan building’s foundations is used as an explanation. While indeed cases for churches built on top of pagan temples reCf. Klein, Spätantike Tempelzerstörungen 290–2. Gregory the Great, Epistula 11, 56 (MGH epist. 2, 331) with Klein, Spätantike Tempelzerstörungen 292–3. 192 For instance, cf. Klein, Spätantike Tempelzerstörungen 293 and Hahn, Vetustus 369 proposing that the church signalized Christian identity. 193 Cf. McKenzie, Gibson and Reyes, Reconstructing the Serapeum 107–9 and Baldini, Problemi 97–152. A similar case can be observed for the temple of Aphrodite in Aphrodisias where the problem caused by a small cella was solved by turning the temple peristyle into the interior colonnade of the future church; cf. Ratté, New research 130–3. 190 191

THE SILENCE OF THE GODS

71

main extremely rare until the sixth century, there were, of course, exceptions, for instance the symbolically important Church of the Holy Sepulchre in Jerusalem. We have seen at the beginning of this study how much emphasis was placed by Eusebius to explain how even the soil was meticulously cleansed of demonic pollution. Epigraphic research has proved that within the area of the Patriarchates of Jerusalem and Antioch the phenomenon of churches being built on top of pagan shrines remained extremely rare, 194 and also for Egypt the case appears rather similar. Arguably, there existed a second reason for the lack of a church within the precinct of the Serapeion: By deliberately leaving the destroyed space in ruins, the defeat of paganism was evocatively visualized in the urban landscape. The Temple Mount in Jerusalem, devastated in 70 CE may serve as a comparison – even after the transformation of the Holy City into the spiritual centre of Christendom, which took its very beginning with Constantine’s construction of the Church of the Holy Sepulchre, the holy place of Judaism was left as an ostentatiously open space pointing to the supersession of the old faith by a new one: Even an illiterate pilgrim who could not read about the city’s past in the history books (and given that he or she was unaware of Christ’s prophesy that “there shall not be left [...] one stone upon another, that shall not be thrown down” 195). The same meaning might be ascribed to the ruins of the Serapeion, or – if one travelled up the river – in Hermopolis, Dendera or Luxor. Jan Assmann demonstrated that in pre-modern times the physical use of violence was often preceded by verbal escalaCf. Haensch, Le financement. A further field of study would be to investigate to what extent destroyed synagogues were re-used as churches, since their ground plans were much more apt to be transformed into Christian churches, and especially since synagogue buildings often became more monumental in Late Antique times. Even though a law issued in 423 forbade the construction of new synagogues, Late Antique Jews were not very much afraid to undertake many new building projects; a later law (a. 438) then would threaten that newly-constructed synagogues were quasi automatically turned into churches by force; cf. Rutgers, The synagogue 449–50. 195 Matthew 24:2. 194

72

KONSTANTIN M. KLEIN

tions. 196 The violent conflict in Alexandria followed on a vast amount of eloquent polemic. But also afterwards, the destruction was echoed in a multitude of church histories: not only the sheer amount of text but also the drastic display of violence within this corpus is astonishing. This was important for a society that can to a certain extent be characterized by its fears – we may think of the ubiquitous demons dwelling in the old temples’ foundation walls. The literary commemoration of the Christian victory over paganism functioned as a form of self-assurance. Comparable studies for the Western part of the late Roman Empire have showed that the authors almost always ascribe the stimulus of pagan-Christian violence to the pagans, presenting the Christians as merely reacting to provocations. At a time when the monopoly on legitimate violence was only partly resting on professional institutions – there was no ‘police’ in Alexandria – this meant that the Christians styled themselves as the preservers of general security. 197 This may result from the fact that the sources are biased – there are by far more literary accounts on such violent conflicts written by Christians, just as we do not have any pagan account from pre-Islamic times recounting the Islamic destruction of sanctuaries in the Arabian peninsula. Nevertheless, the preoccupations of Christian – and later Muslim – authors may not be the only reason for this slant in the evidence. Ultimately, the Roman legislation of the fourth century just as Muḥammad’s strict orders in the seventh century did not leave much room for pagans to argue against the destruction of their temples and shrines. Theodosius I had urged Bishop Theophilus and his monks to stop the conflict, however, he had also granted them permission to close the temple and to destroy the idols within it. The emperor must have been aware that the Alexandrian Christians might deliberately misinterpret the second part of his order. At a first glance, the case was different in the Ḥijāz where the orders left no room for interpretation and where, as we have seen, the buildings to be destroyed were of a much smaller scale. Nevertheless, in times when writing was less important than it is to us, ruined temples, sanctuaries and shrines (be it in Egypt, Jerusalem 196 197

Cf. Assmann, Monotheismus 13. Salzman, Pagan-Christian religious violence 282–5.

THE SILENCE OF THE GODS

73

or the Ḥijāz) would have functioned as a different kind of text: proclaiming the superiority of the new over the old in a language that was universally comprehensible for any adherent to the triumphant faith and thereby complemented the written word.

BIBLIOGRAPHY LACL PLRE

Lexikon der antiken christlichen Literatur, ed. S. Döpp and W. Geerlings, Freiburg 31998. The Prosopography of the later Roman Empire, ed. A. Jones, J. Martindale and J. Morris, Cambridge 1971–1992.

Primary sources All Biblical quotations are from the King James Bible, Quranic quotations are taken from Alan Jones’ translation (Exter 2007). Abū Nuʿaym, Abū Nuʿaym Aḥmad ibn ʿAbdallāh al-Iṣbahānī: Dalāʾil an-nubuwwa, ed. M. Qalʿajī and ʿA. ʿAbbās, Beirut 1986. Agathangelos, [Historia armeniaca] Agathangelos: History of the Armenians, ed. and trans. R. Thomson, Albany 1976. Augustine, Epistulae, ed. J.-P. Migne (Patrologia Latina 33), Paris 1902, coll. 61–1094. Azraqī, [Akhbār Makka] Die Chroniken der Stadt Mekka I, ed. F. Wüstenfeld, Leipzig 1858. al-Balādhurī, Ansāb al-ashrāf, ed. M. Ḥamīdullāh, Cairo 1959. Besa, [Vita Sinuthii] Sinuthii Archimandritae Vita et Opera Omnia I, ed. J. Leipoldt and W. Crum (Corpus Scriptorum Christianorum Orientalium 4. Copt. 1), Paris 1901 . Besa: The Life of Shenoute, trans. D. Bell (Cistercian Studies Series 73), Kalamazoo 1983. Besa, Vita Sinuthii (versio Arabica), ed. and trans. E. Amélineau in Monuments pour servir à l’histoire de l’Égypte chrétienne aux IVe, Ve, VIe et VIIe siècle, vol. I (Mission archéologique française au Caire. Mémoires. 4), Paris 1888–1895, 289–478. Cassiodorus, [Variae epistulae] Selected Variae of Magnus Aurelius Cassiodorus Senator, trans. S. Barnish (Translated Texts for Historians. 12), Liverpool 1992. Christodorus, Ekphrasis, ed. and trans. W. Paton in The Greek An-

74

KONSTANTIN M. KLEIN

thology, Books I-VI (Loeb Classical Library. 67), Cambridge, MA 81993, 57–91. Chrysostom, Homilia 38, ed. J.-P. Migne (Patrologia Graeca 60), Paris 1862. Cicero, [In Verrem (2)] Cicero, Orationes III, ed. G. Peterson, Oxford 1978. Cicero, [In Verrem (2)] Cicero: The Verrine Orations, trans. L. Greenwood, 2 vols, Cambridge 1928. Mommsen, T. (ed.), [Codex Theodosianus] Theodosiani libri xvi cum constitutionibus sirmondianis, Berlin 1905. Pharr, C. (trans.), [Codex Theodosianus] The Theodosian Code and Novels and the Sirmondian Constitutions: a translation with commentary, glossary, and bibliography (The Corpus of Roman Law. 1), New York 1952. Eusebius, [Praeparatio evangelica] Die Praeparatio Evangelica, ed. K. Mras (Die Griechischen Schriftsteller der ersten Jahrhunderte. Eusebius. 8.1), Berlin 1954. Eusebius, [Praeparatio evangelica] Eusebius of Caesarea: praeparatio evangelica (preparation for the Gospel), trans. E. Gifford, Oxford 1903. Eusebius, [Vita Constantini] Eusebius von Caesarea: De vita Constantini – Über das Leben Konstantins, ed. and trans. H. Schneider (Fontes Christiani. 83), Turnhout 2007. Eusebius, [Vita Constantini] Eusebius: Life of Constantine, trans. Av. Cameron and S. Hall, Oxford 1999. Theodor, J. and Albeck, C. (eds.), [Genesis Rabba] Midrash Bereshit Rabba: critical edition with notes and commentary, 3 vols, Jerusalem 1965 [reprint Jerusalem 1996]. Freedman, H. (trans.), [Genesis Rabba] Midrash Rabbah: translated into English, vol. 1–2, London 1939. Gerontius, [Vita Melaniae] Vie de Sainte Mélanie, ed. and trans. D. Gorce (Sources Chrétiennes. 90), Paris 1962. Gerontius, [Vita Melaniae] The Life of Melania the Younger: introduction, translation, and commentary, trans. E. Clark (Studies in Women and Religion. 14), New York and Toronto 1984. Herodotus, [Historiae] Herodoti historiae, ed. C. Hude, Oxford 31976. Herodotus, [Historiae] Herodotus, The Persian Wars, trans. A. Godley,

THE SILENCE OF THE GODS

75

4 vols., Cambridge 1920–1925. Ibn ʿAbd al-Barr, [Istīʿāb] Al-Istīʿāb fī maʿrifat al-aṣḥāb, ed. ʿA. alBijāwī, 4 vols., Cairo n.d. Ibn Hishām, [Sīra], Das Leben Muhammed’s nach Muhammed Ibn Ishâk bearbeitet von Abd el-Malik Ibn Hischâm: Erster Theil, Text, ed. F. Wüstenfeld, 2 vols., Göttingen 1858–1859. Ibn Hishām, [Sīra], The Life of Muhammad: a translation of Isḥaq’s Sīrat rasūl Allāh, trans. A. Guillaume, Oxford 1955, repr. Karachi 1967, 232010. Ibn al-Kalbī, [Kitāb al-Aṣnām] Das Götzenbuch: Kitâb al-Aṣnâm des Ibn al-Kalbî: Übersetzung mit Einleitung und Kommentar, ed. and trans. R. Klinke-Rosenberger (Sammlung Orientalistischer Arbeiten 8), Leipzig 1941. Ibn Saʿd, [Ṭabaqāt] Ibn Saʿd az-Zuhrī, aṭ-Ṭabaqāt al-kubrā, ed. I. ʿAbbās, 9 vols., Beirut 1985. Geyer, P. and Cuntz, O. (eds.), [Itinerarium Burdigalense] Itineraria et alia geographica (Corpus Christianorum Series Latina 175), Turnhout 1965, 1–26. Wilkinson, J. (trans.), [Itinerarium Burdigalense] Egeria’s travels, Warminster 31999, 26–34. Libanius, [Orationes] Libanii Opera III: Orationes XXVI-L, ed. R. Foerster, Leipzig 1906, repr. Hildesheim 1963. Libanius, [Orationes] Libanius: selected works in three volumes, II: Selected Orations, trans. A. Norman (Loeb Classical Library 452), Cambridge and London 1977. Lucian, [De sacrificiis] Luciani Opera II: Libelli 26–43, ed. M. Macleod, Oxford 1974, 114–20. Lucian, [De sacrificiis] Lucian in eight volumes, trans. A. Harmon, vol. 3, Cambridge 1969, 153–71. Malalas, [Chronographia] Ioannis Malalae Chronographia, ed. J. Thurn (Corpus Fontium Historiae Byzantinae 35), Berlin and New York 2000. Malalas, [Chronographia] The Chronicles of John Malalas, trans. E. Jeffreys, M. Jeffreys and R. Scott (Byzantina Australiensia. 4), Melbourne 1986. Marc the Deacon, [Vita Porphyrii] Marc le Diacre: vie de Porphyre, évêque

76

KONSTANTIN M. KLEIN

de Gaza, ed. and trans. H. Grégoire and M.-A. Kugener (Belles Lettres. Collection byzantine), Paris 1930. Minucius Felix, [Octavius] M. Minuci Felicis Octavius, ed. B. Kytzler, Leipzig 21992. Krupp, M. (ed. and trans.), [Mishna ʿAvoda Sara] Die Mischna: Textkritische Ausgabe: ʿAvoda Sara: Götzendienst, Jerusalem 2002. Paton W. (ed. and trans.), [Palladas] The Greek Anthology, Books IX (Loeb Classical Library 84), Cambridge 1917, nos 180–183. Pausanias, [Graeciae descriptio] Pausaniae Graeciae descriptio, ed. F. Spiro, Leipzig 1903. Pausanias, [Graeciae descriptio] Pausanias’ description of Greece, trans. J. Frazer, 6 vols., New York 1965. Pliny, [Epistulae] C. Plini Secundi epistularum libri novem, ed. R. Hanslik, Stuttgart/Leipzig 1958, 31992, ed. M. Schuster. Pliny, [Epistulae] Pliny the Younger, Letters, trans. B. Radice, 2 vols., Cambridge 1969. Plutarch, [De defectu oraculorum] Plutarche: Œuvres morales VI: Dialogues Pythiques, ed. and trans. R. Flacelière, Paris 1974, 83–196. Plutarch, [De defectu oraculorum], Plutarch’s Moralia in sixteen volumes, trans. F. Babbitt, vol. 5, London/Cambridge 1969, 350–501. Plutarch, [De Iside et Osiride] Plutarche: Œuvres morales, ed. and trans. C. Froidefond, vol. V.2, Paris 1988. Plutarch, [De superstitione] Plutarche: Œuvres morales, ed. and trans. J. Defradas, J. Hani and R. Klaerr, vol. 2, Paris 1985. Prudentius, [Contra Symmachum] Prudentius: contra Symmachum, ed. and trans. H. Tränkle (Fontes Christiani 85), Turnhout 2008 . Qalqashandī, [Ṣubḥ al-aʿshā] Aḥmad ibn ʿAlī al-Qalqashandī: Ṣubḥ alaʿshā fī ṣināʿat al-inshāʾ, ed. M. shams ad-Dīn, 3 vols., Cairo 1913–1919. Rufinus, [Historia ecclesiastica] Eusebius Werke 2,2: Die Kirchengeschichte, ed. T. Mommsen (Griechische christliche Schriftsteller), Leipzig 1908. Rufinus, [Historia ecclesiastica] The Church History of Rufinus of Aquileia: books 10 and 11, trans. P. Amidon, Oxford 1997. Sallustius, [De deis et mundo] Sallustius: Concerning the gods and the universe: edited with prolegomena & translation, ed. and trans. A. Nock,

THE SILENCE OF THE GODS

77

Cambridge 1926. Severus of Menorca, [Epistula Severi] Severus of Minorca: Letter on the conversion of the Jews, ed. and trans. S. Bradbury, Oxford 1996. Shenoute, [De iudicio] Schenute von Atripe: De iudicio, ed. H. Behlmer (Cataloge del Museo Egizio di Torino. 1: Monumenti e Testi 8), Turin 1996. Shenoute, [Let our eyes] Let our eyes, a “discourse” of Shenoute, ed. and trans. S. Emmel in From temple to church: destruction and renewal of local cultic topography in Late Antiquity, ed. J. Hahn, S. Emmel, U. Gotter (Religions in the Graeco-Roman world), Leiden and Boston 1976, 182–97. Shenoute, [The Lord thundered] Shenoute’s sermon “The Lord thundered”: an introduction and translation, ed. and trans. J. Timbie and J. Zaborowski in Oriens Christianus 90 (2006), 91–123. Sozomen, [Historia ecclesiastica] Sozomenos: Historia ecclesiastica, ed. and trans. G. Hansen, 4 vols., (Fontes Christiani 73), Turnhout 2004. Ṭabarī, [Dhayl al-Mudhayyal] Annales quos scripsit Abu Djafar at-Tabari, ed. M. de Goeje et al., vol. 13, Leiden 1901. Ṭabarī, [Dhayl al-Mudhayyal] The history of al-Ṭabarī: an annotated translation. XXXIX: biographies of the prophet’s companions and their successors, trans. E. Landau-Tasseron, Albany 1998. Ṭabarī, [Taʾrīkh] Annales quos scripsit Abu Djafar at-Tabari, ed. M. de Goeje et al., 13 vols., Leiden 1879–1901. Ṭabarī, [Taʾrīkh] The history of al-Ṭabarī: an annotated translation, 40 vols., Albany 1985–2007. Ṭabarī, [Tafsīr] Jāmiʿ al-bayān ʿan taʾwīl āy al-Qurʾān, ed. ʿA. b. ʿAbd al-Muḥsin at-Turkī, 28 vols., Cairo 2001. Tertullian, [Apologeticum] Tertulliani Apologeticum, ed. H. Hoppe (Corpus Scriptorum Ecclesiasticorum Latinorum 69), Vienna 1939. Tertullian, [De idololatria] Tertullianus, de idololatria: critical text, translation, and commentary, ed. and trans. J. Waszink and J. van Winden (Supplements to Vigiliae Christianae 1), Leiden and New York 1987. Wahb b. Munabbih, [Ḥadīṯ Dāwūd] Wahb b. Munabbih: Der Heidelberger Papyrus PSR Heid Arab 23, ed. and trans. R. Khoury, vol.

78

KONSTANTIN M. KLEIN

1 (Codices Arabici Antiqui 1), Wiesbaden 1972, 33–115. Wāqidī, [Kitāb al-Maghāzī] The Kitāb al-Maghāzī of al-Wāqidī, ed. M. Jones, 3 vols., Oxford 1966. Secondary sources Acerbi, S., Il potere dei monaci nei concili orientali del V secolo: il Costantinopolitano Eutiche e il Siro Bar Sauma, in Studia Historica – Historia Antigua 24 (2006), 291–313. Acerbi, S., Terror y violencia anitjudía en Oriente durante el siglo V: la biografía del archimandrita Sirio Bar Sauma, in Formas y usos de la violencia en el mundo romano, ed. G. Castañeda and R. Salinero, Madrid 2007, 277–90. Assmann, J., Monotheismus und die Sprache der Gewalt (Wiener Vorlesungen im Rathaus 116), Vienna 32007. Atallah, W., Jibt et Ṭāghūt dans le Coran, in Arabica 17 (1970), 69– 82. Atallah, W., al-Kalbī 2, EI2, iv, 495–6. Auffarth, C., Götterbilder im römischen Griechenland: vom Tempel zum Museum?, in Ritual dynamics and religious change in the Roman Empire, ed. O. Hekster, S. Schmidt-Hofner and C. Witschel (Impact of Empire 9), Leiden and Boston 2009, 307– 25. Bagnall, R., Models and evidence in the study of religion in late Roman Egypt, in From Temple to Church: Destruction and Renewal of local cultic Topography in Late Antiquity, ed. J. Hahn, S. Emmel and U. Gotter (Religions in the Graeco-Roman World 163), Leiden and Boston 2008, 23–41. Baldini, A., Problemi della tradizione sulla ‘disruzione’ del Serapeo di Alessandria, in Rivista storica dell’antichità 15 (1985), 97–152. Bar, D., Continuity and change in the cultic topography of late antique Palestine, in From Temple to Church: Destruction and Renewal of local cultic Topography in Late Antiquity, ed. J. Hahn, S. Emmel and U. Gotter (Religions in the Graeco-Roman World 163), Leiden and Boston 2008, 275–98. Barnes, T., Ammianus Marcellinus and his world, in Classical Philology 88 (1993), 55–70. Bassett, S., The urban image of late antique Constantinople, Cambridge

THE SILENCE OF THE GODS

79

2004. Bauer, F. and C. Witschel (eds.), Statuen in der Spätantike, Wiesbaden 2007. Beeston, A., The “Muʿallaqāt” problem, in Arabic literature to the end of the Umayyad Period, ed. A. Beeston, Cambridge 1983, 111–13. Behlmer, H., Schenute von Atripe: de iudicio (Cataloge del Museo Egizio di Torino. 1: Monumenti e Testi 8), Turin 1996. Belayche, N., «Partager la table des dieux» L’empereur Julien et les sacrifices, in Revue de l’histoire des religions 218 (2001), 457–86. Belayche, N., Sacrifice and the theory of sacrifice during the ‘pagan reaction’: Julian the Emperor, in Sacrifice in religious experience, ed. A. Baumgarten (Studies in the History of Religion 93), Leiden and Boston 2002, 101–126. Bell, D., Besa: Life of Shenoute (Cistercian Studies Series 73), Kalamazoo 1983. Berdozzo, F., Götter, Mythen, Philosophen: Lukian und die paganen Göttervorstellungen seiner Zeit (Untersuchungen zur antiken Literatur und Geschichte 106), Berlin 2011. Binder, S., Tertullian, On Idolatry and Mishna Avodah Zara: questioning the parting of the ways between Christians and Jews (Jewish and Christian Perspectives Series 22), Leiden and Boston 2012. Borgeaud, P., The death of the great Pan: the problem of interpretation, in History of Religions 22 (1983), 254–283. Bosworth, C., The History of al-Ṭabarī V: the Sāsānids, the Byzantines, the Lakmids, and Yemen, Albany 1999. Bowersock, G., Polytheism and monotheism in Arabia and the three Palestine, in Dumbarton Oaks Papers 51 (1997), 1–10. Bowersock, G., Empires in collision in Late Antiquity, Waltham 2012. Brakke, D., Demons and the making of the monk: spiritual combat in early Christianity, Cambridge and London 2006. Breitner, G., Statuen der Spätantike – kulturelles Erbe oder pagane Altlast?, in Wechsel der Religionen – Religion des Wechsels, ed. N. Krohn and S. Ristow (Studien zu Spätantike und Frühmittelalter 4), Hamburg 2012, 107–19. Brown, P., The rise and function of the Holy Man in Late Antiquity, in Journal of Roman Studies 61 (1971), 80–101.

80

KONSTANTIN M. KLEIN

Brown, P., Christianization and religious conflict, in Cambridge Ancient History 13: The Late Empire, a.d. 337–425, ed. A. Cameron and P. Garnesy, Cambridge 1998, 632–64. Brooks, E., The Acts of St. George, in Le Muséon 38 (1925), 67– 115. Bruns, P., Markus, Diakon, in LACL 489. Budge, E., George of Lydda: a study of the cultus of St. George in Ethiopia, London 1930. Cameron, Al., The last pagans of Rome, Oxford 2011. Cameron, Av., The later Roman Empire ad 284–430, Cambridge 1993. Cameron, Av., The violence of orthodoxy, in Heresy and identity in Late Antiquity, ed. E. Iricinschi and H. Zellentin (Texts and Studies in Ancient Judaism 119), Tübingen 2008, 102–14. Cameron, Av. and S. Hall, Eusebius: Life of Constantine, Oxford 1999. Caner, D., Wandering, begging monks: spiritual authority and the promotion of monasticism in Late Antiquity (The Transformation of the Classical Heritage 33), Berkeley, Los Angeles and London 2002. Conrad, L., The Biblical tradition for the plague of the Philistines, in Journal of the American Oriental Society 104 (1984), 281–97. Conrad, L., Abraha and Muhammad: some observations apropos of chronology and literary topoi in the early Arabic historical tradition, in Bulletin of the School of Oriental and African Studies 50 (1987), 225–40. Conti, S., Tra integrazione ed emarginazione: le ultime Vestali, in Studia Historica – Historia Antigua 21 (2003), 209–22. Corbo, V., Il Santo Sepolcro di Gerusalemme: aspetti archeologici dalle origini al periodo crociato, 3 vols. (Studium Biblicum Franciscanum. Collectio maior 29), Jerusalem 1981–1982. Crone, P., Meccan trade and the rise of Islam, Princeton 1987. Dally, O., Pflege und Umnutzung heidnischer Tempel in der Spätantike, in Die spätantike Stadt und ihre Christianisierung, ed. G. Brands and H.-G. Severin (Spätantike, frühes Christentum, Byzanz: B. Studien und Perspektiven 11), Wiesbaden 2003, 97–114. Donner, F., Muhammad and the beilievers: at the origins of Islam, Cambridge and London 2010.

THE SILENCE OF THE GODS

81

Emmel, S., From the other side of the Nile: Shenute and Panopolis, in Perspectives on Panopolis: an Egytipan town from Alexander the Great to the Arab conquest, ed. A. Egberts, B. Muhs and J. van der Vliet (Papyrologica Lugduno-Batavia 31), Leiden 2002, 95–113. Emmel, S., Shenoute of Atripe and the Christian destruction of temples in Egypt: rhetoric and reality, in From temple to church: destruction and renewal of local cultic topography in Late Antiquity, ed. J. Hahn, S. Emmel and U. Gotter (Religions in the GraecoRoman World 163), Leiden and Boston 2008, 161–201. Errington, R., Church and State in the first years of Theodosius I, in Chiron 27 (1997), 21–72. Errington, R., Christian accounts of the religious legislation of Theodosius I, in Klio 79 (1997), 398–443. Fahd, T., Ṭāghūt (A.), in EI2, x, 93–4. Fahd, T., Le Panthéon de l’Arabie centrale a la veille de l’Hégire (Institut Français d’Archéologie de Beyrouth, Bibliothèque archéologique et historique 88), Paris 1968. Franke, P., Begegnung mit Khidr: Quellenstudien zum Imaginären im traditionellen Islam (Beiruter Texte und Studien 79), Wiesbaden 2000. Frazer, J., Adonis Attis Osiris (The Golden Bough 4.1), London 31914. Gaudement, J., La condamnation des praticques paiennes en 391, in Epektasis: FS J. Daniélou, ed. J. Fontaine and C. Kannengiesser, Paris 1972, 597–602. Ginzberg, L., The legends of the Jews I: Bible times and characters from the Creation to Jacob, Philadelphia 1909. Ginzberg, L., The legends of the Jews V: notes to volumes I and II – from the Creation to the Exodus, Philadelphia 1955. Grabbe, L., Of mice and dead men: Herodotus 2.141 and Sennacherib’s campaign in 701 BCE, in ‘Like a bird in a cage’: the invasion of Sennacherib in 701 BCE, Sheffield (Journal for the Study of the Old Testament Supplement Series 363) 2003, 119–40. Graf, F., Plutarch und die Götterbilder, in Gott und die Götter bei Plutarch: Götterbilder – Gottesbilder – Weltbilder, ed. by R. HirschLuipold (Religionsgeschichtliche Versuche und Vorarbeiten

82

KONSTANTIN M. KLEIN

54), Berlin and New York 2005, 251–66. Gotter, U., Rechtgläubige – Pagane – Häretiker: Tempelzerstörungen in der Kirchengeschichtsschreibung und das Bild der christlichen Kaiser, in From temple to church: destruction and renewal of local cultic topography in Late Antiquity (Religions in the Graeco-Roman World. 163), ed. J. Hahn, S. Emmel and U. Gotter, Leiden and Boston 2008, 43–89. Gutsfeld, A., J. Hahn and S. Lehmann, Christlicher Staat und ‹panhellenische› Heiligtümer: zum Wandel überregionaler paganer Kultstätten im spätantiken Griechenland, in Antike Religionsgeschichte in räumlicher Perspektive, ed. J. Rüpke, Tübingen 2007, 228–37. Haensch, R., Le financement de la construction des églises pendant l’Antiquité Tadrive et l’évergétisme antique, in L’Antiquité Tardive 14 (2006), 47–58. Hahn, J., Gewalt und religiöser Konflikt: Studien zu den Auseinandersetzungen zwischen Christen, Heiden und Juden im Osten des Römischen Reiches (von Konstantin bis Theodosius II.) (Klio-Beihefte NF. 8), Berlin 2004. Hahn, J., Vetustus error extinctus est: wann wurde das Sarapeion von Alexandria zerstört?, in Historia 55 (2006), 368–83. Hahn, J., Gesetze als Waffe? Die kaiserliche Religionspolitik und die Zerstörung der Tempel, in Spätantiker Staat und religiöser Konflikt: Imperiale und lokale Verwaltung und die Gewalt gegen Heiligtümer, ed. J. Hahn (Millennium-Studien 34), Berlin 2011, 201–20. Halbertal, M. and A. Margalit, Idolatry, Cambridge 1992. Hawting, G., The idea of idolatry and the emergence of Islam: from polemic to history, Cambridge 2004. Hoss, S., Die spätantike Blüte römischer Thermen in Palästina, in Die antike Stadt im Umbruch, ed. N. Burkhardt and R. Stichel, Wiesbaden 2010, 165–77. Hunt, D., Christianising the Roman Empire: the evidence of the Code, in The Theodosian Code, ed. J. Harries and I. Wood, Ithaca 1993, 143–58. Khalidi, T., Images of Muhammad: narratives of the prophet in Islam across the centuries, New York, London and Toronto 2009.

THE SILENCE OF THE GODS

83

Khoury, R., Wahb b. Munabbih: Der Heidelberger Papyrus PSR Heid Arab 23, vol. 1 (Codices Arabici Antiqui 1), Wiesbaden 1972. Khoury, R., Wahb b. Munabbih, in EI2, xi, 34–6. Kister, M., The campaign of Ḥulubān: a new light on the expedition of Abraha, in Le Muséon 78 (1965), 425–436. Kister, M., The seven odes, in Rivista degli Studi Orientali 44 (1969), 27–36. Klein, R., Spätantike Tempelzerstörungen im Widerspruch christlicher Urteile, in Roma versa per aevum: Ausgewählte Schriften zur heidnischen und christlichen Spätantike, ed. R. von Haehling, Hildesheim, Zürich and New York (2002), 284–294 (originally published in Studia Patristica 24 (1993), 135–42). Krone, S., Die altarabische Gottheit al-Lāt (Heidelberger Orientalistische Studien 23), Frankfurt, Berlin and Bern 1992. Kunderewicz, C., La protection des monuments d’architecture antique dans le Code Théodosien, in Studi in onore di Edoardo Volterra, vol. 4, Milan 1971, 137–53. Lavan, L. and M. Mulryan (eds.), The archaeology of late antique ‘paganism’ (Late Antique Archaeology 7), Leiden and New York 2011. Lecker, M., Idol worship in pre-Islamic Medina (Yathrib), in Le Muséon 106 (1993), 331–46. Lecker, M., Muslims, Jews and pagans: studies in early Islamic Medina (Islamic History and Civilization: Studies and Texts 13), Leiden and New York 1995. Lecker, M., Zayd B. Thābit, “a Jew with two sidelocks”: Judaism and literacy in pre-Islamic Medina (Yathrib), in Journal of Near Eastern Studies 56 (1997), 259–73. Lecker, M., Was Arabian idol worship declining on the eve of Islam?, in People, tribes and society in Arabia around the time of Muḥammad (Variorum Collected Studies Series 812), ed. Michael Lecker, Aldershot 2005, III:1–43. Lecker, M., ʿAvodat haʾelilim bi-zfon ʿarav bi-tqufat ha-jāhiliyya (Polytheism in Northern Arabia on the eve of Islam), in Polytheism in the Land of Israel and its neighbours in Antiquity, ed. M. Kister, Jerusalem, 2008, 250–64. Lecker, M., Wadd, the weaponed idol of Dūmat al-Jandal and the

84

KONSTANTIN M. KLEIN

quṣṣāṣ, in Dieux et déesses d’Arabie: images et représentations (Orient & Méditerranée 7), ed. I. Sachet and C. Robin, Paris 2012, 123–30. Lecker, M., (forthcoming) “Idol worship in pre-Islamic Yamāma” in Festschrift (TBA). Leipoldt, J., Shenute von Atripe und die Entstehung des national-ägyptischen Christentums (Texte und Untersuchungen 25), Leipzig 1903. Lloyd, A., Herodotus’ account on Pharaonic history, in Historia 37 (1988), 22–53. Lubomierski, N., Die Vita Sinuthii: Form- und Überlieferungsgeschichte der hagiographischen Texte über Schenute den Archimandriten (Studien und Texte zu Antike und Christentum 45), Tübingen 2007. Mader, E., Mambre: Die Ergebnisse der Ausgrabungen im heiligen Bezirk Râmet el-khalîl in Südpalästina 1926–1928: Textband, Freiburg 1957. McKenzie, J., S. Gibson and A. Reyes, Reconstructing the Serapeum in Alexandria from the archaeological evidence, in Journal of Roman Studies 94 (2004), 73–121. McLynn, N., Pagans in a Christian Empire, in A companion to Late Antiquity, ed. P. Rousseau, Oxford 2009, 572–87. Meier, H.-R., Alte Tempel – neue Kulte: Zum Schutz obsoleter Sakralbauten in der Spätantike und zur Adaption alter Bauten an den christlichen Kult, in Innovation in der Spätantike (Spätantike, frühes Christentum, Byzanz: B. Studien und Perspektiven 1), ed. B. Brenk, Wiesbaden 1996, 361–74. Mettinger, T., The riddle of resurrection: dying and rising gods in the ancient Near East (Coniectanea Biblica. Old Testament 50), Stockholm 2001. Müller, H.-P., Daphnis – ein Doppelgänger des Gottes Adonis, in Zeitschrift des Deutschen Palästinavereins 116 (2000), 26–41. Nau, F., Résumé de monographies syriaques: Barṣauma [...], in Revue de l’Orient chrétien 18 (1913–1914), 270–6 and 379–89, and 19, 113–34 and 278–89. Nesselrath, H.-G., Libanius: Zeuge einer schwindenen Welt (Standorte in Antike und Christentum 4), Stuttgart 2012. Neuwirth, A., Der Koran: Handkommentar mit Übersetzung, vol. I: frühmekkanische Suren: poetische Prophetie, Berlin 2011.

THE SILENCE OF THE GODS

85

Nock, A., Ὁ μέγας Πὰν τέθνηκε, in Classical Review 37 (1923), 164– 5. Nöldeke, T., Über die Amalekiter und einige andere Nachbarvölker der Israeliten, Göttingen 1864. Papaconstantinou, A., “Là où le péché abondait, la grâce a surabondé:” sur les lieux de culte dédiés aux saints dans l’Égypte des Ve-VIIIe siècles, in Le sacré et son inscription dans l’espace à Byzance et en Occident: études comparées (Byzantina Sorbonensia 18), ed. M. Kaplan, Paris (2001), 245–9. Patrich, J., The early Church of the Holy Sepulchre in the light of excavations and restoration, in Ancient churches revealed, ed. Y. Tsafrir, Jerusalem 1993, 101–17. Petit, P., Zur Datierung von Libanios’ Rede ›Pro templis‹, in Libanios, ed. G. Fatouros and T. Krischer (Wege der Forschung 621), Darmstadt 1983, 43–67 (originally published as Sur la date du «Pro templis» de Libanius, in Byzantion 21 (1951), 286– 309). Pines, S., Jāhiliyya and ʿilm, in Jerusalem Studies in Arabic and Islam 13 (1990), 174–94. Prinz, F., Die Kirche und die pagane Kulturtradition: Formen der Abwehr, Adaption und Anverwandlung, in Historische Zeitschrift 276 (2003), 281–303. Ratté, C., New research on the urban development of Aphrodisias in Late Antiquity, in Urbanism in Western Asia Minor: new studies on Aphrodisias, Ephesos, Hierapolis, Pergamon, Perge and Xanthos, ed. D. Parrish, Portsmouth (2001), 116–47. Reinach, S., La mort du grand Pan, in Bulletin de correspondance hellénique 31 (1907), 5–19. Robin, C., Nagrān vers l’époque du massacre: notes sur l’histoire politique, économique et institutionelle et sur l’introduction du christianisme (avec un réexamen du Martyr d’Azqīr), in Juifs et Chrétiens en Arabie aux Ve et VIe siècles: regards croisés sur les sources, ed. J. Beaucampe, F. Briquel-Chatonnet and C. Robin (Collège de France – CNRS Centre de Recherche d’Histoire et Civilisation de Byzance. Monographies. 32 = Le massacre de Najrân: religion et politque en Arabie du Sud au VIe siècle 2), Paris 2010, 39–106.

86

KONSTANTIN M. KLEIN

Rubin, U., Ḥanīfiyya and Kaʿba: an inquiry into the Arabian preIslamic background of dīn Ibrāhīm, in Jerusalem Studies in Arabic and Islam 13 (1990) 85–111 (reprinted in: Muḥammad the Prophet and Arabia, ed. U. Rubin (Variorum Collected Studies Series 968, no. X), Aldershot 2011). Russell, J., Zoroastrianism in Armenia (Harvard Iranian Studies 5), Cambridge and London 1987. Russell, J., Pre-Christian Armenian Religion, in Aufstieg und Niedergang der Römischen Welt II,18,4, ed. W. Haase and H. Temporini, Berlin 1990, 2679–92 (reprinted in: Armenian and Iranian Studies, ed. J. Russell (Harvard Armenian Texts and Studies 9), Cambridge 2004, no. 30, 370–88). Rutgers, L., The synagogue as foe in early Christian literature, in “Follow the wise” (B Sanhedrin 32b): FS L. Levine, ed. Z. Weiss, Winona Lake 2010, 449–68. Salzman, M., Pagan-Christian religious violence, in Violence in Late Antiquity: perceptions and practices, ed. H. Drake, Aldershot 2006, 265–85. Saradi, H., Christian attitudes toward pagan monuments in Late Antiquity and their legacy in later Byzantine centuries, in Dumbarton Oaks Papers 44 (1990), 47–61. Sarefield, P., Bookburning in the Christian Roman Empire: transforming a pagan rite of purification, in Violence in Late Antiquity: perceptions and practices, ed. H. Drake, Aldershot 2006, 287– 95. Sinai, N., Fortschreibung und Auslegung: Studien zur frühen Koraninterpretation (Diskurse der Arabistik 16), Wiesbaden 2009. Stemberger, G., Introduction to the Talmud and Midrash, Cambridge 1991. Stroumsa, G., Barbarian philosophy: the religious revolution of early Christianity (Wissenschaftliche Untersuchungen zum Neuen Testament 112)., Tübingen 1999. Stroumsa, G., La fin du sacrifice: les mutations religieuses de l’Antiquité tardive, Paris 2005. Tesslar, V., The death of Pan: a classical instance of verbal misinterpretation, in Psychoanalytic Review 8 (1921), 180–183. Tottoli, R., Abū Righāl, in EI3 (checked 24.04.2013).

THE SILENCE OF THE GODS

87

Trombley, F., Hellenic religion and Christianization c. 370–529: vol. I (Religions in the Graeco-Roman World 115/1), Leiden, New York and Cologne 1993. Trombley, F., Hellenic religion and Christianization c. 370–529: vol. II (Religions in the Graeco-Roman World 115/2), Leiden, New York and Cologne 1994. Vermander, J.-M., La polémique de Tertullien contre les dieux du paganisme, in Revue des sciences religieuses 53 (1979), 111–23. Wallraff, W., Die antipaganen Maßnahmen Konstantins in der Darstellung des Euseb von Kaisareia, in Spätantiker Staat und religiöser Konflikt: Imperiale und lokale Verwaltung und die Gewalt gegen Heiligtümer (Millennium-Studien 34), ed. J. Hahn, Berlin 2011, 7–18. Watt, W. and R. Winder, al-Madīna, in EI2, v, 994–1007. Wellhausen, J., Reste arabischen Heidentums, Berlin and Leipzig 21927. Wiemer, H.-U., Für die Tempel? Die Gewalt gegen heidnische Heiligtümer aus der Sicht städtischer Eliten des spätrömischen Ostens, in Spätantiker Staat und religiöser Konflikt: Imperiale und lokale Verwaltung und die Gewalt gegen Heiligtümer (MillenniumStudien 34), ed. J. Hahn, Berlin 2011, 159–85. Windau, B., Besa von Atripe, in LACL 126–7. Windau, B., Minucius Felix, in LACL 504–5. Winden, J. van, Idolum and idololatria in Tertullian, in Vigiliae Christianae 36 (1982), 108–14.

THE RELATIONSHIP BETWEEN THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE BY THE EAST-SYRIAN CHRISTIANS IN QATAR (SIXTH – SEVENTH CENTURIES) MARTIN TAMCKE

The church province Qatar, which was raised to the rank of a metropolitan see quite late, was one of the most important and vigorous church regions on the Arabian Peninsula even long after the Islamic conquest. Yet as far as the Church is concerned, the name “Qatar” comprised much more than the peninsula and the state which bear this name nowadays. 1 Just like in the neighbouring bishoprics especially in the present Bahrain area, traders and pearl fishers were obviously an important element in the metropolis of Qatar, which had a determining mental influence on church life. 2 This region, already ruled by a bishop in the third century, lay at a considerable distance from the political centre of the Church in Seleucia-Ctesiphon, the capital of the Sassanid Empire, but quite close to the centre of the old Persian areas of the Church, which were mainly in opposition to the central ecclesiastical power. 3 The circumstances regarding its assignment to a jurisdiction before it became a metropolis have not been adequately explained so far, but the close relationship to Persia and the considerable distance to the Ioan, Muslime 55–9. al-Khalifa, Bahrain 349–51. 3 Tamcke, Christians 105–20; Ioan, Muslime 103–14. 1 2

89

90

MARTIN TAMCKE

patriarchs are obvious. This situation forced the patriarchs to undertake great efforts in order to keep this region under their jurisdiction. From time to time the patriarchs themselves had to set off and plead for agreement with the church leaders there. However, the relationship remained a tense one. Of course there were also forces in the region that were closer to the patriarchs than to the church hierarchy of the Persians. The sources attribute the founding of the metropolitan see of Qatar to Īshōʿyahḇ III, but his successor Giwargis might as well be the founder. 4 The Arab rule had led to a strong reduction of the East-Syrian Christianity on the Arabian Peninsula, but its continuation there can be noticed up until the eleventh century. 5 At that time, Qatar was the only metropolitan see of the Church on the Arabian Peninsula; the other see that also had jurisdiction over the peninsula was that of the metropolitan on Socotra Island. The theological academies of the Church of the East were critical for the theological scholarship within its area of activity. They were centres of academic erudition, hermeneutic methodology and exegetic precision. 6 Knowledge of the classical antiquity concentrated here and was eventually transformed into East-Syrian Culture. However, these places proved to be centres for continuing challenge when testing new paths which contrasted with the conventional piety and with the traditional self-understanding of the Church. Quite easily they earned a reputation as cradles of heretics. An example of this is the case of Henana, one of the most prominent scholars at the School of Nisibis, who kept the Church under his spell for half a century with his new approaches. Until today no such East-Syrian academy could be identified on the Arabian Peninsula, as it was the case for the Mesopotamian and Iranian homelands of the Church of the East. This is all the more astonishing, as it is quite obvious what a vast spiritual potential unfolded within the Church in this region. Isaac of Nineveh (late seventh century) descended from this area, transgressing all Young, Patriarch 98. Thomas attended the synod of 676 as metropolitan of Qatar. 5 Brock, Syriac Writers 95, Fiey, Chrétiens syriaques. 6 Fiey, Nisibe, Becker, Fear of God, Becker, Sources. 4

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE

91

the borders of the Church and leaving us a substantial spiritual literature, which had a stimulating effect on the traditions of all big Churches in the world. 7 This is also the homeland of the mystic Dadisho Qatraya (seventh century), who commented intensively on the spiritual works of the Desert Fathers and acted within the theological-philosophical traditions of Evagrius Ponticus. 8 Furthermore, the most important interpreters of the liturgy of the Church of the East originated in Qatar. Yet, no one has succeeded so far in proving the existence of a local school comparable to the one of Nisibis, although the intellectual potential of Qatar obviously spread across vast areas of the Middle East. As far as the powerful theological heritage of Qatar writers is concerned, we nowadays wonder about the circumstances under which they developed their theological conceptions and how these affected theological reflection. It is obvious that Qatar is a region characterized by monasticism. The attitude of the monks towards the Church leaders is particularly noteworthy. Not only were they committed to the policy of the Persian hierarchy, but they acted absolutely irrespective of church authorities and the local clergy. This situation placed them in the middle between the community members, who were faithful to the patriarchate, and the representatives of the local Church, who opposed the patriarchate. The monks position might be related to later features of the local monasticism, which substantially resulted from the reforms undertaken by Abraham of Kashkar, as it is illustrated by Florence Jullien in her dissertation at the University of Paris. 9 These positions of monasticism can be inferred from the widespread impact of the most prominent writers of the region, but they are even more obvious in the letters of CatholicosPatriarch Īshōʿyahḇ III (middle of the seventh century). Although it provides a unilateral perspective and lacks the letters from the Patriarch’s interlocutors, this corpus of letters illustrates their author’s struggle for influence on the monks as well as his controversy with the local Church hierarchy. Ovidiu Ioan’s recently pubTamcke, Gotteserlebnis 107–16. For this and the following cf. Brock, Syriac Writers. 9 Jullien, Le monachisme. 7 8

92

MARTIN TAMCKE

lished dissertation compiles the main material, which gives an insight into the historical processes around the Muslim conquest of this region. 10 It was clearly proven what excellent representatives of theological teaching the monks of this region were – Isaac of Nineveh and Dadisho Qatraya were both monks. However, it has hardly been explained so far how theology and science were concretely cultivated in the monasteries, or if we can possibly speak of theological education there. In contrast to this situation, during the last quarter of the sixth century even the representatives of the Church hierarchy seemed to need basic information regarding quite elementary theological matters and liturgical customs of the Church of the East, for which they used to turn to the patriarch. This state of affairs increases the urgency of the following question: how could such leading theologians and liturgical interpreters have developed during the seventh century in a region which shortly beforehand seemed to be no man’s land in respect of theology? In view of the rich material on this topic, we will approach this question from the complex perspective of the relationship between theology and religiosity. C.D.G. Müller’s dictum that the Arab Christians were “blank slates” with respect to theology and teachings and his supposition that “they must have assumed and honoured all that pervaded clothed in the garment of imposing holiness” express the special and, meanwhile, substantial significance of East-Syrian monasticism for the Arabian Peninsula. Müller has already stated that it were only the members of the Church of the East dwelling on the Arabian Peninsula that nourished serious scientific ambitions. Yet this situation, which Müller attributed to religiosity at that time, needs a closer inspection with respect to the concrete situation within the Church of the East. 11 The following first example can hardly lead beyond a rough sketch of the circumstances, but it manages to shed light upon the situation existing a few decades before the Islamic conquest. Jacob, the local bishop of Dirin Island in the neighbourhood of Qatar, turned to the patriarch, who lived in the capital of the 10 11

Ioan, Muslime. Müller, Kirche 19.

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE

93

Sassanids in Mesopotamia. The patriarch confirms to the bishop that his love for God has been revealed in his love for the people who were entrusted to him. 12 Therefore, the patriarch places the bishop’s concern not primarily in the field of religiosity, but in that of theology. The Holy Scriptures are the key to the teachings of life – the Torah for the Jews and the Gospel for the Christians. 13 A man who is always concerned with the love for teaching, an therefore endevours to gain theological knowledge – not with indifference, but thoroughly, not in a superficial way, but in truth and not only pretending to do so – such a man can please God indeed. 14 This fundamental statement confers a religious significance to theological teaching. The endeavour to gain knowledge must not be separated from religiosity, but forms a religious act itself. The patriarch explicitly regards and accepts Jacob as a theologian who reflects in this way and struggles against the heretics along with the teachers of the truth. 15 Although the bishop’s 33 questions regarding the priesthood and the liturgy do not correspond to the interest of the patriarch, who wanted to write about the church order generally, he admits the restricted and concrete inquiries. Subsequently, he underlines his own theological authority by reminding of his former teaching activity and of a theological writing about baptism and Eucharist which he had previously published. 16 It is striking to see how freely the patriarch dedicates himself to these questions. Since the 33 questions are actually 22 as far as the content is concerned, he answers them in such a manner that he goes over the bishop’s planning. 17 Some of them remain unanswered, because he judged that such an answer would be so easily comprehensible to

166.

12

Braun, Das Buch der Synhados 237, Chabot, Synodicon 424, Syrian text

Braun, Das Buch der Synhados 238, Chabot, Synodicon 425, 166. Braun, Das Buch der Synhados 238, Chabot, Synodicon 425–6, 167. 15 Braun, Das Buch der Synhados 238–9, Chabot, Synodicon 426, Syrian text 167. 16 Braun, Das Buch der Synhados 239, Chabot, Synodicon 426, Syrian text 167. 17 Braun, Das Buch der Synhados 271, Chabot, Synodicon 450, 191. 13 14

94

MARTIN TAMCKE

anyone, that it needed no further attention. 18 Such statements on the side of the patriarch could be understood as arrogance towards the bishop, but at the same time they express the trust of the Church leader in the local theological competence. This mixture of authority and partnership in the theological discourse was not at all unusual, but highly appropriate for the theology and piety of the Church of the East, as they were based on the relationship between teachers and students to a great extent. The patriarch is aware that the bishop was actually expecting short answers to his questions. He admits to Jacob that this was appropriate for the authors of canons, who combined their orders with punishments in case of disregard. He, however, prefers to follow the customs of the teachers, who in their teachings referred to evidence provided by the Scripture and by nature. They used to illustrate the essence and the characteristic of the topic they were discussing. Their aim was to teach the reader and to explain to him the facts in a way that would increase his knowledge. Consequently, the patriarch answers questions that had practical interests and asked for simple instructions for ecclesiastical life and religious accomplishment. This, however, is precisely what he does not deliver. Instead, he presents a complete didactic work that embeds the religious concern in theological reflection. It should not be judged now whether the bishop was given such an answer because the patriarch felt that he still had to be theologically instructed or that his request for hasty decisions had to be corrected in favour of a theological discourse. The decisive factor might have been the fact that the patriarch regarded the field of instruction as belonging to himself rather than to the church order. Not only his words and his explicit reference to the bishop’s interest in teaching support this assessment, but also the manner in which he conveys his didactic work to the bishop. The latter may read the writing and add what is missing. 19 The patriarch very clearly signifies that he is not making authoritative decisions on the basis of his synodal powers. Instead, he intends to act as it is Braun, Das Buch der Synhados 271, Chabot, Synodicon 450–1, 191–2. Braun, Das Buch der Synhados 450–1, Chabot, Synodicon 451, Syrian text 192. 18 19

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE

95

usual for the relationship between “disciple” and “master”. 20 It is less important whether such a teacher-student relationship between the two correspondents was the basis of their dialogue or not. In any case, the preserved didactic work documents that the theological teaching specific to the Church of the East was actually present in the region of the later metropolitan see of Qatar. The surviving document is not only an item of theological import, but it bears witness of a theological dialogue, although the patriarch considered the questions of the bishop as too pragmatic and although we only know these questions and the bishop’s position from the scanty report of the patriarch. Indeed, the questions deal with such matters as the liturgical activity of the priest, touching upon topics such as money lending and approaching even the question whether an infertile woman may get married. 21 A large section is dedicated to questions regarding disputes that influence church life. Jacob of Dirin therefore raises a variety of topics before the patriarch. Obviously, the patriarch acts as more than the church leader who issued orders; he actively seeks the theological dialogue. But history does not reveal whether Jacob met this interest and how he responded to the invitation to continue independently the piece of writing he had received. However, the fact that his theological passion is confirmed by the patriarch suggests that he might have done so. The theological dialogue could therefore have replaced the authoritative decision of a synod, substituting enforced orders by self-responsible theological action. During the seventh century, the situation in the region changed considerably. The relationship between the patriarch and the metropolitan of Persia was tense. This state of affairs caused an intricate game of changing alliances in the region as well as a struggle for influence and independence as a metropolitan seat. 22 This interval grants us deep insights into the area – for example through the letters sent by Patriarch Īshōʿyahḇ III to the representatives of the region – and only ends with the synod which PaBraun, Das Buch der Synhados 271–2, Chabot, Synodicon 451. Braun, Das Buch der Synhados 240–70, Chabot, Synodicon 428–49, 168–90. 22 Braun, Das Buch der Synhados 332–3. 20 21

96

MARTIN TAMCKE

triarch Giwargis summoned in May 676 in Dirin. 23 Beforehand, the bishops had visited various islands and places on the north-eastern shore of the Arabian Peninsula and in the Persian Gulf. An authoritative decision imposed by the canons of a synod, something that Īshōʿyahḇ I had deliberately avoided, this time became a conscious object and purpose of the negotiations. Although the decisions made at the previous synods still existed, reforms for the believers in that region were considered to be absolutely necessary. 24 The topic of theology and piety was approached several times during the synod. The particularity of this synod is also obvious in its composition. It brought together all the bishops of the Church of the East whose sees were located on the northern side of the Arabian Peninsula. The patriarch and the bishops who attended the synod made it the synodal purpose to convey the “law” to the “herds”, that is to the believers who had been entrusted to them. 25 The believers described as herds of the sea are those “who abode in south world”. 26 This illustrates the position of the patriarchate which may have been similar to the self-understanding of the bishops in the region. In geographical terms, the region represented the southern hemisphere of the Church to the church members. Even the list of the bishops itself reflects the local character of the decisions that were made at this synod: Thomas, bishop of Qatar, Īshōʿyahḇ, bishop of Dirin Island, Sergius, bishop of Trihan (this is the only see outside of the Arabian area at the Persian Gulf), Stephanus, bishop of Oman, Pusai, bishop of Hagar in the middle of Bahrain, and Shahin, bishop of Hatta, which lies on the very shore. 27 216.

23

Braun, Das Buch der Synhados 335, Chabot, Synodicon 482, Syrian text

Braun, Das Buch der Synhados 335–6, Chabot, Synodicon 482, Syrian text 216. 25 Braun, Das Buch der Synhados 335, Chabot, Synodicon 481, 216. 26 Braun, Das Buch der Synhados 335, Chabot, Synodicon 481, 216. 27 Braun, Das Buch der Synhados 335, Chabot, Synodicon 481, 216, Braun also suggested that Mrwnāyē should be corrected and replaced with Māzūnāyē, Braun, Das Buch der Synhados 335 footnote no. 2. This correction is similarly suggested for the passages in the texts of Īshōʿjahb. 24

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE

97

In his bishopric, the bishop of Oman recorded large movements of conversion from Christianity to Islam, and this already at the time of Giwargis’s predecessor. The reason was not the pressure exerted by the ruling Muslims, but the taxes that they demanded. As the patriarch himself reports, taxes imposed by the Muslims amounted to half of the overall income. Christians had to pay these sums for the right to keep their faith. 28 Unimpressed by the amount of the taxes, the catholicos criticized the conversions provoked by economic constraint as wrong decision. In his view, the believers gave up their inherited faith, which would have brought them eternal life, in order to preserve the half of their material goods. By abandoning their Christian faith – which had been won through the martyrdom of so many people in the world – they proved themselves incapable of sacrificing half of their fortune for their faith. The patriarch summoned up theology and piety in his struggle against the growing numbers of conversions. In combination with the schism that had been provoked by the Arabian bishops against the patriarchate in Mesopotamia, this phenomenon made a new order necessary, which Īshōʿyahḇ III strove to achieve in vain. He had to accept the fact that his legations were not welcome in Arabia and that the monks, who remained faithful to him, were persecuted by the bishops. The participants in the synod addressed the deplorable state of affairs and agreed that the natural virtue and love for God were consigned “to oblivion because of spiritual unsteadiness and physical distress”. 29 God gave man a temporary life and mortal body and Īshōʿjahḇ III, 11, 182/12, 251. The institution of taxes had been approved by the Patriarch himself, who unambiguously impressed this upon his believers. Those who are authorized to demand taxes should also be paid the taxes, Īshōʿjahḇ III, 1955: 268–9/194. Regarding the taxation in Oman see Hitti, The origins 118–19 (about the taxation in Bahrain with reference to the contract of the Christians, Jews and Zoroastrians and with the detail that the inhabitants had to share their possessions with the Muslims see Hitti, The origins 121). In Bahrain the half of the goods which were given away was referred to as tax on goods, grains and dates, Hitti, The origins 123. 29 Braun, Das Buch der Synhados 334; Chabot, Synodicon 480, Syrian text 28

98

MARTIN TAMCKE

He allowed the weaknesses to master him, “so that he who deeply lives in piety [could get his reward] through great effort.” 30 Laws had served this effort from time immemorial, in the beginning as warnings given to Adam and leading to fear of God, then from Noah to Abraham and Moses; back then, the laws were given through Moses “for the old people, which is a shadow of the mystery of the new one”; and eventually the appearance of God’s beloved brought the Gospel with its transcendental laws, which guide us to Heaven. 31 He was then followed by the apostles, priests and teachers of the next ages. The use of the theological topos “law” represents a comparative reference to the way the Jews understood and employed the law. The law here becomes a theological category employed in reaction to religious and ethical shortcomings. Three obvious steps concerning the tension between theology and piety result from this. First, spiritual unsteadiness is regarded as completely devastating from the dogmatic perspective, i.e. as a lack of straightforwardness and faithfulness towards one’s own position. The hesitating spirit of the residents of Qatar is regarded as a weakness, and not as adaptability or flexibility. And such a weakness is included in the religious endeavours of the believers. Therefore, this weakness needs to be fought with the help of the law. The fathers gathered at the synod did not consider the laws to be unalterable, but as something easily fit into various contexts. The alternating character of human weakness has always required reforms appropriate for relationships between peoples and countries and responsive to new necessities – reforms which should also be put down in writing. 32 The synod’s basic premise was therefore the fact that the bishops who gathered there were facing new circumstances in their sees and needed new laws in order to cope with them. This necessity for flexibility in the face of new circumstances becomes atmospherically possible when it is seen in light of the end of the world. 215. 215.

30 31 32

Braun, Das Buch der Synhados 333–4. Braun, Das Buch der Synhados 334, Chabot, Synodicon 481, Syrian text Braun, Das Buch der Synhados 334–5, Chabot, Synodicon 481, 216.

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE

99

In their observations, the members of the synod disregarded the contradiction between the measures necessary for the issuing of laws and the imminent end of the world. The hierarchs explicitly stressed that they had to exert their authority as church leaders in such difficult times – the hard times of end of the world. 33 In times of decay, Christ-loving people dwelling in the south of the world, that is in the sees of the Church in Arabia, needed revival and correct laws so that they could “stay within the limits of piety”. In this case, theology serves religiosity which seems to be endangered and needs to be preserved. The theological effort, as creative as may be, has a reactive character. But the reaction was caused by a challenge regarded as overwhelming. The responsible hierarchs of the Church in Arabia had to fight with their back to the wall against the changes provoked by the Muslim conquerors within the religious and social life of the Christians living in a Muslim world. The reorganization was primarily focused on the religious practice. Both Patriarch Īshōʿyahḇ III and the bishops gathered at the synod summoned up by the Patriarch Giwargis described the essential religious feature of Arabian Christendom as deficiency. Already Īshōʿyahḇ III had demanded a reinforcement of teaching efforts. The synod organized by Giwargis pursued the same goal. The activity in theological education increases. On every Sunday and every holiday, all priests and teachers must talk about faith in their sermons and lectures delivered to the believers. 34 This would instruct future teachers and also prepare the believers for controversies with persons of other religions, who would ask them about their faith. Since these interlocutors are described as heretics, there is no doubt that even in the first canon the Muslims were referred to as challenging opponents. 35 Anyway, the bishops considered that their task was to keep the alien customs away from their believers. Braun, Das Buch der Synhados 335, Chabot, Synodicon 481, 216. See Canon 1 of the synod, dedicated entirely to the teaching matter, Braun, Das Buch der Synhados 336, Chabot, Synodicon 482, Syrian text, 217. 35 Chabot, Synodicon 483, 217. The list of the great scholars of the region is impressive. During the Umayyad dynasty it was Isaac of Nineveh, an outstanding figure due to his spiritual writings (Baumstark, Geschichte 33 34

100

MARTIN TAMCKE

The basic tendency of the practical measures taken during the synod went towards a demarcation between the Christians and their environment. Therefore the differences were particularly emphasised, especially with respect to marriage requirements. Christian women were no longer allowed to marry non-Christian men. 36 Bigamy was forbidden. 37 The church was still flourishing, so further orders for new church buildings had to be issued. 38 Besides such practical resolutions, piety was also addressed: for example, the visible use of the cross in the cult as witness of God’s blessing during the wedding ceremony was regarded as imperative. Theology became evident in the specific Christian rite, which the assembly did not regard as protected by the local religious practice. As a proof for the continuous existence of their religious community, special importance was given to the open church attendance. 39 Neither the “private prayer” carried out without the participation of the community nor prayers uttered quickly and secretly were sufficient. Church attendance became an act of witnessing. It was a legal obligation to go to the Church and participate in the liturgy every evening and every morning. Exceptions were made only in case of absolute necessity. Silent prayers in chapels or private homes were unacceptable, since the connection to the community was not es223–5). Later, Dadisho Qatraya testifies to the monastic-mystical wisdom of Qatar (Baumstark, Geschichte 226–7). Besides, Qatar also produced important interpreters of the liturgy. The most important authors were gathered by S. Brock in Brock, Syriac Writers 85–96. 36 The marriage issues are settled in Canon 13 (Interdiction of a woman’s marriage without her parents’ consent, mediation by the cross and the benediction pronounced by the priest), Braun, Das Buch der Synhados 343–4, Canon 14 (Interdiction of marriages between Christian women and non-Christian men), Braun, Das Buch der Synhados 344, (Interdiction of bigamy), Braun, Das Buch der Synhados 345–6. 37 Canon 16 of the synod, Braun, Das Buch der Synhados 345–6, Chabot, Synodicon 489, Syrian text 224. 38 Canon 2 of the synod, Braun, Das Buch der Synhados 336–7, Chabot, Synodicon 483, 217–18. 39 Canon 14 of the synod, Braun, Das Buch der Synhados 344–5; Chabot, Synodicon 488, 224.

THEOLOGICAL TEACHING AND RELIGIOUS PRACTICE 101 sential there. People had to remain in the church until the final benediction pronounced by the priest. Jacob of Dirin had once expected authoritative decisions made by a synod, but he received only the theological dialogue between master and disciple. This time, however, the local church leaders, together with the patriarch, employed theology in the explanation of the legal measures meant to reform the religious life and to protect the survival of the Church. With this, theology unfolded its capacity to pursue changes in religiosity.

BIBLIOGRAPHY Al-Khalifa, A. and M. Rice (eds.), Bahrain through the ages. The history, London 1993. Baumstark, A., Geschichte der syrischen Literatur. Mit Ausschluß der christlich-palästinensischen Texte, Bonn 1922. Becker, A.H., Fear of God and the Beginning of Wisdom. The School of Nisibis and the Development of Scholastic Culture in Late Antique Mesopotamia, Philadelphia 2006. Becker, A.H., Sources for the Study of the School of Nisibis, Liverpool 2008. Braun, O., Das Buch der Synhados oder Synodicon orientale. Die Sammlung der nestorianischen Konzilien, zusammengestellt im neunten Jahrhundert nach der syrischen Handschrift, Museo Borgiano 82 der vatikatischen Bibliothek, Amsterdam 1975. Brock, S.P. Syriac Writers from Beth Qatraye, in ARAM 11–12 (1999–2000), 85–96. Chabot, J.B., Synodicon Orientale ou Recueil des synodes nestoriens, Notices et extraits des manuscrits de la Bibliothèque Nationale 37, Paris 1902. Fiey, J.M., Nisibe, metropole syriaque orientale et ses suffragants des origines a nos jours (CSCO 388, Subsidia 54), Louvain 1977. Fiey, J.M., Chrétiens syriaques sous les Abbassides, surtout à Bagdad (749– 1258) (CSCO 420, Subsidia 59), Louvain 1980. Hitti, Ph.Kh., The origins of the islamic state. Being a translation from the Arabic accompanied with annotations geographic and historic notes of the Kitâh futûh al-buldân, of al Imâm abu-l Abbâs Ahmad ibn Jâbir alBalâdhuri, Piscataway 1916. Ioan, O., Muslime und Araber bei Īshōʿjahb III. (649–659) (Göttinger

102

MARTIN TAMCKE

Orientforschungen 1, Syriaca 37), Wiesbaden 2009. Jullien, F., Le monachisme en Perse. La réforme d’Abraham le Grand, père des moines de l’Orient (CSCO 622, Subsidia 121), Louvain 2008. Müller, C.D.G., Kirche und Mission unter den Arabern in vorislamischer Zeit (Sammlung gemeinverständlicher Vorträge und Schriften aus dem Gebiet der Theologie und Religionsgeschichte 249), Tübingen 1967. Tamcke, M. (ed.), Christians and Muslims in dialogue in the Islamic Orient of the Middle Ages. Christlich-muslimische Gespräche im Mittelalter (Beiruter Texte und Studien 117), Würzburg and Beirut 2007. Tamcke, M. (ed.), Gotteserlebnis und Gotteslehre. Christliche und islamische Mystik im Orient (Göttinger Orientforschungen 1, Syriaca 38), Wiesbaden 2010. Young, W.G., Patriarch, shah and caliph. A study of the relationships of the church of the East with the Sassanid empire and the early caliphates up to 820 A.D. with special references to available translated syriac sources, Rawalpindi 1974.

II. RE-INVESTIGATING THE TEXTS OF EARLY AND CLASSICAL ARABIC POETRY

103

GLORY AND IMMORTALITY:

THE MOTIF OF MONUMENTUM AERE PERENNIUS BY SAMAWʾAL B. ʿĀDIYĀʾ

KIRILL DMITRIEV If one is truly to succeed in leading a person to a specific place, one must first and foremost take care to find him where he is and begin there… But all true helping begins with a humbling. Søren Kierkegaard

Early Arabic poetry is commonly considered profane. One could argue that, as in many other cultures, in its origins Arabic poetry was closely related to magical practices, one could point to the function of poetry in the cult of mourning or refer to numerous attestations of oaths in poetry and to texts containing Biblical motifs and references to Jews and Christians. Nonetheless, against the background of several other Near Eastern literatures, formed to a great extent by rich traditions of liturgical poetry, the transmitted works of early Arabic poetry do indeed appear free from any religious connotations, something which has been interpreted as indicating “that the Arab of Central Arabia, in the days before al-Islâm, interested himself little in religion of any sort”. 1 The jahiliyyah concept, that draws a clear line of demarcation between pre-Islamic and Islamic cultural history, seems to find its confirmation here and in return, allows one to exclude the early Arabic literary tradition from the context of Arab religious history. The texts of early Ara1

Lyall, Translations xxvii.

105

106

KIRILL DMITRIEV

bic poetry remain, however, the major literary source on preIslamic Arabia. If one does not subscribe to the jahiliyyah model unconditionally, but attempts to investigate the transition from the pre-Islamic to Islamic period in its historical continuity, one has to pay close attention to early Arabic poetry. 2 In other words, if one is to understand how Arabs reached the stage of their cultural history at which Islam emerged, one should start from where they did, and look at their profane pre-Islamic heritage. Arabic literary works of this time offer neither the epic pathos of Gilgamesh nor the prayerful warmth of the psalmist. One has, therefore, to be humble in approaching the sources, which do not contain anything religious, at least not at first glance. That is to say that religious themes do not only find their expression in specific liturgical genres, but can also be reflected in other forms of literary production. From this point of view, texts of early Arabic poetry, as profane as they are, provide evidence of important aspects of the social and cultural consciousness of Arabs in Late Antiquity, which are relevant also in the context of religious history. One of them is the notion of immortality. It has been commonly assumed that Arabs in pre-Islamic time did not develop a notion of hereafter and did not elaborate on the idea of immortality. This assumption seems to find its confirmation in the well known quranic verse 45:24:

‫وت َون َْح َيا َو َما يُ ْه ِل ُك َنا إِ ��� الد ْ�ه ُر‬ ُ ‫َوقَالُوا َما ِه َي إِ ��� َح َياتُ َنا ال �د ْن َيا َن ُم‬

They say, ‘There is nothing but our present life; we die, and we live, and nothing but Time destroys us’. (Trans. A.J. Arberry). 3

Some poetic texts also suggest that prior to Islam Arabs did not acknowledge the reality of the hereafter. Labīd, a famous early Arab poet, describes the fleetingness of human life in the opening line of his Rithāʾ Abrad: On the importance of Arabic poetry for the study of the Quran see Bauer, Relevance 699. 3 Arberry, Koran ii, 212. 2

GLORY AND IMMORTALITY

ِ ‫َوتَبقى‬ ‫الجبا ُل َبعدَنا َوال َمصانِ ُع‬

107

‫َبلينا َوما تَبلى ال ُنجو ُم ال َطوالِ ُع‬

We [mortals] wither and perish, but the stars that rise on high do not, and the hills and the water-towers remain after we have gone. (Trans. A. Jones). 4

Yet, in theoretical literary discourse it has been recognised that the power to glorify is inseparably linked with the power to immortalise, and that poetic preoccupation with this power is immanent in poetry. In Classical philology, for example, all Greek literature – song, poetry, prose – is commonly regarded as originating in κλέος, the act of praising famous deeds. 5 It is remarkable that in the Christian tradition the Antique notion of glory obtains a significant religious dimension and becomes a central attribute of immortality - Rom 2:7: τοῖς μὲν καθ’ ὑπομονὴν ἔργου ἀγαθοῦ δόξαν καὶ τιμὴν καὶ ἀφθαρσίαν ζητοῦσιν, ζωὴν αἰώνιον

To those who by persistence in doing good seek glory, honour and immortality, he will give eternal life. (New International Version).

Glory is here a synonym for salvation, as also in Rom 8:17–18: εἰ δὲ τέκνα, καὶ κληρονόμοι: κληρονόμοι μὲν θεοῦ, συγκληρονόμοι δὲ Χριστοῦ, εἴπερ συμπάσχομεν ἵνα καὶ συνδοξασθῶμεν. Λογίζομαι γὰρ ὅτι οὐκ ἄξια τὰ παθήματα τοῦ νῦν καιροῦ πρὸς τὴν μέλλουσαν δόξαν ἀποκαλυφθῆναι εἰς ἡμᾶς. And if children, heirs also, heirs of God and fellow heirs with Christ, if indeed we suffer with Him in order that we may also be glorified with Him. I consider that our present sufferings are not worth comparing with the glory that will be revealed in us.

4 5

Jones, Early Arabic Poetry 96. Nagy, Early Greek views 9.

108

KIRILL DMITRIEV (New International Version).

In the cultural tradition of Late Antiquity glory and immortality are closely linked together. Early Arabic poetry makes a similar connection. Praising is prevalent in Arabic poetry, and in the tribal Arab society of pre-Islamic time poetry itself was a powerful tool to glorify and to achieve immortality in the memory of future generations. It is not surprising that such immortality in the social environment of early Arab tribal society, which was so strongly constituted by the collective consciousness, was primarily a collective rather than an individual matter. Human life was perceived within a chain of generations, constantly reproducing themselves in the endless circle of time. Nevertheless, poetic sources provide evidence that Arabs in the pre-Islamic period, especially in the decades immediately preceding the rise of Islam, were increasingly concerned with the afterlife and immortality as being relevant to the individual. A poem attributed to as-Samawʾal b. ʿĀdiyāʾ provides an interesting variation on this topic. *** As-Samawʾal b. ʿĀdiyāʾ 6 was born around 560 CE to Jewish parents and lived in the region Taymāʾ, where Jewish presence was attested since at least the second century CE. 7 In or near to this oasis was located al-Ablaq 8, the castle of as-Samawʾal’s family, a place connected to the only known episode from as-Samawʾal’s life, which also made his name proverbial: as-Samawʾal became famous in the Arabic tradition as someone who was uncompromisRather scant details about as-Samawʾal’s life are reported in alIṣbahānī, al-Aghānī iv, 332–3, ix, 96–9, xxii, 116–21. For an overview of the research on as-Samawʾal’s poetry and further bibliography see: Bauer, Al-Samawʾal b. ʿĀdiyā, and Sezgin, GAS ii, 249–50. On Jews in Arabia see: Newby, A History of the Jews, and more recent articles by Hoyland, The Jews of the Hijaz; Hoyland, The Jewish Poets. 7 For an overview of ongoing archaeological research on Taymāʾ see Hausleiter, Das antike Tayma’. 8 Werner Caskel notes that the meaning of the name ‘al-Ablaq’ is ‘the spotty one’ referring to the red and grey stone of the castle, see Caskel, Lihyan, nr. 27 and Caskel, Al-Aʿshà Nr. 25, 6, 133. 6

GLORY AND IMMORTALITY

109

ingly faithful and loyal. As he was entrusted with the armour of his protégé, as-Samawʾal refused to hand it over to his enemies in exchange for the life of his own son, who was executed as a result. A reference to these events contains verse 12 of the poem number VI in as-Samawʾal’s diwan, where the owner of the armour is described as ‘the Kindī’. This attribution and the identification of asSamawʾal’s protégé with Imru’ al-Qays, the grandson of the last king of Kinda, in Kitāb al-aghānī 9 have been regarded by Werner Caskel as later inventions. 10 Judging the entire poem for that reason as a forgery by Muslim-Arab genealogists or composers of the Imru’ al-Qays Romance might, however, go too far. The reference to ‘the Kindī’ has no significant weight within the text of the poem as a whole. It could well have been added to it in Islamic time, but does not undermine the general pre-Islamic character of the poem, which deserves a closer analysis also apart from its biographical implications. The poem has 16 lines and is composed in the wāfir metre. 11 The two opening lines contain the so-called aṭlāl motif and the motif of women railing against the poet. Both motifs are conventional in the nasīb, the opening part of a qaṣīda. The following part of the poem is a fakhr, the poet’s self-praise. In lines 3 and 4 as-Samawʾal talks about the castle of al-Ablaq, constructed by his grandfather ʿĀdiyāʾ, and then refers to his grandfather’s commandment to preserve what he has built (line 5). The castle of the ancestors is a visible manifestation of asSamawʾal’s inherited authority, which he has to defend. Authority constituted the structure of early Arab society and this authority was based on prestige. Prestige was related to fame and glory, and as such it necessarily implied the recognition of others, the acknowledgement of society. Deeds themselves were not important, but rather the glory they helped to achieve for those fulfilling them, as ʿAntara b. Shaddād clearly says: al-Iṣbahānī, al-Aghānī ix, 99. Caskel, Al-Aʿshà Nr. 25,6, 136. 11 For the full text of the poem and its translation refer to the appendix. 9

10

110

KIRILL DMITRIEV

‫ثَ َنا ًء َو��َ َما ٌل لِ َم ْن ��َ َل ُه َم ْج ُد‬

‫َو��َ َمالَ إِ�� � َما ا�فا َدكَ َن ْيلُ ُه‬

There is no treasure except for that, which brings you praise and there is no treasure for someone, who has no glory. 12

Authority had not only to be achieved, but also secured over time. The logic of ʿĀdiyāʾs demand is obvious. As-Samawʾal answers this challenge with complete confidence in the following passage, beginning at line 6:

‫َو��َ خَ شَ ٍب َو َم ْج ٍد َق ْد ا� َت ْي ُت‬

ٍ‫َو َب ْي ٍت َق ْد َب َن ْي ُت بِ َغ ْي ِر ِطين‬

A good many bayt have I built without clay and without wood, and a good many honours have I won.

A magnificent castle is metaphorically replaced by a structure that does not have even the basic construction materials such as clay and wood, yet allows the poet to claim that he has reached his main goal – glory. Glory and fame manifest themselves through communication. Its major medium in early Arab society was poetry itself, and it is to the inseparable connection between poetry and glory that asSamawʾal gave special emphasis. The fist half of the quoted line contains the word bayt which carries three major meanings in Arabic: ‘tent, house’, ‘house of worship, sanctuary’, and ‘line of poetry, verse’. The ingenious play on the semantics of the word bayt underlies the metaphorical sense of the verse. The “tents” of asSamawʾal are his poetry. This interpretation is supported also by another variation of the verse:

‫َعلَى َظ ْهر ال َم ِط �ية َق ْد َب َن ْي ُت‬

ٍ‫س ِم ْن َو َب ٍر َو ِطين‬ َ ‫و َب ْي ٍت َل ْي‬

A good many bayt neither out of hair nor of clay have I built upon the back of a she-camel. 13

Explicitly excluding hair and clay, the common construction materials for tents and houses, the poet indicates that bayt refers to poet12 13

Sharḥ Dīwān ʿAntara 126. Dīwān as-Samawʾal 97.

GLORY AND IMMORTALITY

111

ry and recalls the habit of reciting it during long rides in the desert. 14 As-Samawʾal gives preference to poetry over a monumental building – a meaningful change, which introduces a new perspective on immortality. Al-Ablaq is placed in opposition to the finiteness of human life on earth. The symbolic immortality of a castle is, however, deprived of any implication of the individuality of its creator. It is understood merely in the context of collective memory. Mentioning someone who built the castle does not change anything. The name of ʿĀdiyāʾ stands not for his unique artistic talent as an architect, but for the impersonal glory and authority embodied in the castle. Both the monument and the name of the man who commissioned it to be built are indications of prestige and glory. It is clear that their ‘immortality’, like the foundation stones of the castle itself, is anchored to the earth and implies no eschatological dimension. It is therefore not surprising that for Labīd impressive monuments like water-towers become synonymous with stars and mountains, great and practically eternal, yet absolutely impersonal objects of the natural word. For ʿĀdiyāʾ the preservation of his castle is of the utmost importance, because he regards it as a powerful means of bequeathing authority on his descendants. For Labīd the perspective of such collective immortality is no longer satisfactory and his reference to man-made towers, along with stars and mountains, only increases the dramatic tension of his reflection: monuments are not symbols of the eternity of human glory, but the opposite – the undeniable and highly visible proof of the shortness and futility of human life. If it is true also for asSamawʾal, then what makes poetry different from a castle in preserving and communicating a poet’s glory? The opposition of poetry to built monuments is a well-known motif in antique and Late Antique literatures. One of its earliest attestations is provided in an Egyptian source, the famous papyrus Compare also a line of an anonymous poet quoted by Lane: ِ ‫َب َن ْي ُت ال َم ِطىِ َظ ْه ِر َعلَى َو َب ْي ٍب َي ْر ُع ُف الخَ َي ِاشي ِم َمشْ ق‬ ‫ُوق بِا� ْس َم َر‬ “Many a bayt upon the back of the camel have I constructed with a tawny thing slit in the nose and bleeding,” i.e. with the reed-pen, Lane, Lexicon i.i, 280–1. 14

112

KIRILL DMITRIEV

Chester Beatty IV 15 dated as early as 1200 BC. The text of the papyrus contains praise for writers who acquired imperishable fame through their literary work: Their portals and mansions have crumbled, Their ka-servants are [gone]; Their tombstones are covered with soil, Their graves are forgotten. Their name is pronounced over their books, Which they made while they had being; Good is the memory of their makers, It is forever and all time!

(Trans. by Miriam Lichtheim). 16

The Greek poets Pindar (522/518 – 445 BC) and Simonides (557/556 BC – 468/467 BC), and the Roman poet Ovid (b. 43 BC) also compare poetry with architectural monuments. The bestknown variation on the topic, which was later imitated in various European literatures, is the one by the Roman lyric poet Horace (65 – 8 BC). In his ode Carminum III, 30 Horace famously proclaims: Exegi monumentum aere perennius Regalique situ pyramidum altius

I have completed a monument more lasting than bronze and far higher than that royal pile of Pyramids. (Trans. by Peter Saint-Andre).

The capability of poetry to outlast its author was also reflected by Arab poets as early as pre-Islamic times, and subsequently. 17 AsSamawʾal, on the contrary, does not elaborate on this idea explicitly. Even if we assume that it is implied, the time scale for asSamawʾal is not the main point of contrast between poetry and the castle. The immortality provided by poetry remains trapped in this British Museum ESA 10684. Lichtheim, Ancient Egyptian literature 177. 17 For examples refer to the article by Gelder, Persons as Texts/Texts as Persons 237–53. 15 16

GLORY AND IMMORTALITY

113

world in the same way as the eternity claimed by a monument, even if words can be more lasting than stones. Being memorable makes poetry an excellent medium, but does not change its function in glorifying and preserving an authoritative name that is immortal only as a subject of collective memory. Other motifs in the fakhr section of as-Samawʾal’s text also follow the conventions of early Arabic poetry in depicting the poet as a hero personifying collective ideals. Line 9 touches upon the theme of the finiteness of human life: VI:9

‫َوق �َض ْي ُت الل� َبا َن َة َوٱش َت َف ْي ُت‬

ً‫َف ٕا ِْن ا� ْه ِل ْك َف َق ْد ا� ْب َل ْي ُت ُع ْذرا‬

So if I perish, yet I have given a proof of forgiveness, fulfilled the purpose and have been cured.

One might be inclined to assume that behind the notion of forgiveness and the description of other moral characteristics of the poet, in contrast to his achievements as a leader in war (line 7) and a successful lover (lines 14–15), lies the idea of a personal reward in the hereafter, which should not have been unfamiliar to asSamawʾal as a Jewish poet. Nevertheless, the words “if I perish” introduce the perspective of the future in negative terms and allow for immortality only in the form of memory. The quantitative aspect of duration in time is here of secondary importance, and this might be the reason why as-Samawʾal remains silent about poetry being more durable than monuments. Yet one can not overlook the strong opposition between the poetry and the castle in his qaṣīda. The contrast conveys the dialogic structure of the passage: as-Samawʾal puts poetry at the forefront as a direct response to his grandfather’s commandment to preserve the castle. Moreover, the polarizing context is underlined by the description of poetry as being of a provocatively intangible material – “without clay and without wood”. Poetry is not just another tool to achieve glory; it is a fundamentally different one. But if the communicative function of the castle and poetry is the same – perpetuating glory and authority – then their opposition by as-Samawʾal becomes meaningful only

114

KIRILL DMITRIEV

through the emphasis on poetry as an individual achievement, on poetry as manifestation of “the singularity of the creative initiative”. 18 Horace, as Sergej Averinzev observes, has already used the motif of monumentum aere perennius to express his individual literary achievement as “having been the first to adapt Aeolic songs into Italian measures”. 19 The distinguishing feature of the ode by Horace is its unambiguous statement about the poet as an author and, consequently, about the immortality of poetry as his individual achievement and not simply another edition of impersonal glory and collective authority made perpetual in a new format. As-Samawʾal is less explicit. His reference to poetry, along with numerous other examples of “the abundance of the metaphorical equation of texts and persons” 20 in early and classical Arabic poetry, should not be over-interpreted in the context of the idea of individual authorship in its post-modern sense. Nonetheless, his emphasis on poetry as a contrast to the heritage of his predecessors, symbolically represented by the inherited castle, provides an expression of the deeper self-consciousness of an individual. Such consciousness unfolds itself fully in the increasingly dramatic perception of death and a desire for personal immortality. It is remarkable that as-Samawʾal explicitly extends his claim for immortality beyond the symbolism of the castle. In the opening line of poem III he declares: III:1

ِ‫يت ال َمصي ِر ِسوى ا���ب َلق‬ ُ ‫َو َب‬

‫بِا���ب َلقِ الفَر ِد َبيتي بِ ِه‬

In the unique al-Ablaq is my house [bayt], but the house [bayt] of the final goal is other than al-Ablaq.

The line suggests not only a notion of the hereafter, but more explicitly denies the absolute superiority of the allegedly eternal nature of the physical world as opposed to human destiny. Such an Averinzev, Authorship and authority 111. Averinzev, Authorship and authority 111. 20 Gelder, Persons as Texts/Texts as Persons 252. 18 19

GLORY AND IMMORTALITY

115

approach reflects the increasing consciousness of a human as an individual being, which can be observed in Arabic poetry shortly before the rise of Islam. 21 An-Nābigha adh-Dhubyānī (d. 605), a contemporary of as-Samawʾal, offers, in a short poem on the death of a certain Ḥisn 22, another example of a strong poetic expression of man’s rebellion against the impersonal powers of space and time:

ِ ‫َيف بِ ِحصنٍ َو‬ ‫موح‬ ُ ‫الجبالُ ُج‬ َ ‫َوك‬ ‫حيح‬ ُ ‫السما ِء َوا���دي ُم َص‬ َ ‫نُجو ُم‬

‫ُفوس ُهم‬ َ ‫َي‬ ٌ ‫قولون ِح‬ ُ ‫صن ثُ �م َتأبى ن‬

‫َو َلم تَل ِف ِظ ال َموتى القُبو ُر َو َلم َت َزل‬

They say Ḥisn! – Then their souls refused to listen. How could this befall Ḥisn, while the mountains are still visible? The earth has not vomited up its graves, still are the stars in the heavens, and the surface of the earth is undamaged.

The life of Ḥisn, a single individual, appears to be so valuable and unique that the poet claims just the same immortality for him as for the universe. The human individual is no longer subject to the outside world, but rather the outside world to the individual. *** Poem VI by as-Samawʾal cannot be classified as a religious text. It is a self-praise by the poet and, despite some specific motifs, it remains within the conventions of the early Arabic tradition of praise poetry. Nevertheless, the text provides evidence that ideas of immortality and the hereafter were not unknown to Arabs in preIslamic times; their notion of human life was not perceived as absolutely limited to physical existence, but also implied hopes of, Jacobi, Time and Reality. Ahlwardt, Divans 166. As Renate Jacobi has pointed out to me, Ḥisn is a rare personal name. Here it refers most probably to Ḥisn ibn Ḥudhayfa, the chief of al-Fazāra, a sub-group of an-Nābigha’s own tribe of Banū Dhubyān. Zuhayr ibn Abī Sulmā, another famous early Arab poet, dedicated a poem to Ḥisn ibn Ḥudhayfa, see Sharḥ Dīwān Zuhayr 124–44. For the akhbār on Ḥisn refer to Mufaḍḍalīyyāt i, 364–5. 21 22

116

KIRILL DMITRIEV

and a strong desire to achieve, immortality. This immortality might be nothing more than the memory of descendants, with no religious connotation as such. However, as the example of asSamawʾal’s poem shows, people in Arabia in the second half of the sixth century approached the issue of death and immortality as an individual matter. The later quranic revelation and its eschatology are strongly focused on the individual character of every human being and individual moral responsibility, as Q 23:99–101 says:

‫َحت�ى إِذَا َجاء ا َ�ح َد ُه ُم ا ْل َم ْو ُت قَالَ َر �ب ا ْر ِج ُع ِون‬ ‫ْت كَ�� � إِن َ�ها َك ِل َم ٌة ُه َو قَائِل َُها َو ِمن َو َرائِ ِهم َب ْر َز ٌخ إِ َلى َي ْو ِم‬ ُ ‫َل َعل�ي ا� ْع َم ُل َصالِ ًحا فِي َما َت َرك‬ ‫ون‬ َ ُ‫ُي ْب َعث‬

‫ون‬ � ‫َف ِٕاذَا نُ ِف َخ فِي‬ َ ُ‫اب َب ْي َن ُه ْم َي ْو َم ِئ ٍذ َو��َ َي َت َساءل‬ َ ‫الصو ِر فَ��َ ا َ�نس‬

Till, when death comes to one of them, he says, ‘My Lord, return me; haply I shall do righteousness in that I forsook.’ Nay, it is but a word he speaks; and there, behind them, is a barrier until the day that they shall be raised up. For when the Trumpet is blown, that day there shall be no kinship any more between them, neither will they question one another. (Trans. A.J. Arberry). 23

Such a radical individualism might not have been accepted among the early audience of the Quran if they had not already been familiar with an individual approach towards the issue of human mortality and immortality. The poetical sources provide some evidence to support this assumption. The notion of immortality in the context of early Arabic panegyrics refers much more to this world than to the next and says more about the poets’ life on earth than their speculations about the hereafter. Nonetheless, in texts with strong liturgical connotations references to the finiteness of human life are also often expressed in a homiletic context with clear moralistic implications. Moses himself, as attested to in the biblical psalms, offers no other 23

Arberry, Koran 44.

GLORY AND IMMORTALITY

117

perspective of the hereafter but a returning to dust and is more concerned about giving guidance in earthly life than about anything that follows after it: Ps 90:1.3.12

‫ית ָלּנוּ ְבּד ֹר וָ ד ֹר׃‬ ָ ִ‫ֲאד ֹנָ י ָמעוֹן ַא ָתּה ָהי‬ ‫י־א ָדם׃‬ ָ ֵ‫אמר שׁוּבוּ ְבנ‬ ֶ ֹ ‫ד־דּ ָכּא וַ תּ‬ ַ ‫ָתּ ֵשׁב ֱאנוֹשׁ ַﬠ‬ ‫הוֹדע וְ נָ ִבא ְל ַבב ָח ְכ ָמה׃‬ ַ ‫ִל ְמנוֹת יָ ֵמינוּ ֵכּן‬

1. Lord, you have been our dwelling place throughout all generations.

3. You turn men back to dust, saying, “Return to dust, O sons of men.” 12. Teach us to number our days aright, that we may gain a heart of wisdom. (New International Version).

The finiteness of human life is a phenomenon of frontier crossing. In dealing with it, several Late Antique religious sources do not just offer utopian visions of the future, but take account of the significance that confronting death has for the present life. In this context the profane character of early Arabic poetry should not be misinterpreted. This poetry had a normative meaning and in many aspects fulfilled the function that in other cultural milieus belonged to the domain of religious literature. As James Montgomery notes in the manifesto concluding his study of the early Arabic qaṣīda, “an Arabic poetry for Arabs, like an Arabic Quran for Muslims, in the religious domain, was an expression of poetic, cultural and political autonomy and pride”. 24 It remains our task to approach early Arabic poetry in a way that will make it more accessible as a source for the study of the religious history of the Arabs.

24

Montgomery, Vagaries 259.

‫‪118‬‬

‫‪KIRILL DMITRIEV‬‬

‫‪APPENDIX‬‬

‫‪As-Samawʾal, Poem VI: 25‬‬

‫‪.1‬‬

‫َعفَا ِم ْن ا ِآل ف َِاط َم َة ٱلخُ َب ْي ُت‬

‫‪.2‬‬

‫ا� َعا ِذ َل َت �ي َق ْو َل ُك َما َع َص ْي ُت‬

‫‪.3‬‬

‫َبنى لي َعا ِد َيا ِح ْصناً َح ِصيناً‬

‫‪.4‬‬

‫ِط ِم ّراً َت ْزلُ ُق ٱ ْلع ْق َبا ُن َع ْن ُه‬

‫‪.5‬‬

‫َوا ْ�و َصى َعا ِد َيا َجد�ي بِا� ْن ��َ‬

‫‪.6‬‬

‫َو َب ْي ٍت َق ْد َب َن ْي ُت بِ َغ ْي ِر ِطينٍ‬

‫‪.7‬‬

‫ش في ُد َجى ٱل �ظ ْل َما ِء َم ْج ٍر‬ ‫َو َج ْي ٍ‬

‫‪.8‬‬

‫َو َذن ٍْب َق ْد َع َف ْو ُت لِ َغ ْي ِر َبا ٍع‬

‫َو��َ َوا ٍع َو َع ْن ُه َق ْد َع َف ْو ُت‬

‫‪.9‬‬

‫لك َف َق ْد ا� ْب َل ْي ُت ُع ْذراً‬ ‫َف ٕا ِْن ا� ْه ْ‬

‫َوق �َض ْي ُت الل� َبا َن َة َوٱش َت َف ْي ُت‬

‫‪.10‬‬

‫ِص ت َْج َت ِديني‬ ‫َوا ْ�صر ُِف َع ْن َق ْوار َ‬

‫‪.11‬‬

‫َفا�حمي الجا َر في ال ُجلّى َف ُيمسي‬

‫‪.12‬‬

‫َو َف ْي ُت بِا� ْد ُر ِع ٱ ْل ِك ْن ِد �ي إِن�ي‬

‫‪.13‬‬

‫يب‬ ‫َوقَالُوا إِن� ُه َك ْن ٌز َر ِغ ٌ‬

‫‪.14‬‬

‫س‬ ‫َو َل ْو��َ ا ْ�ن ُيقَالَ َص َبا ُع َن ْي ٌ‬

‫‪.15‬‬

‫َوقُ �ب ِة َح ِ‬ ‫اصنٍ ا�دْخَ ْل ُت َرأْ ِسي‬

‫‪.16‬‬

‫اس ِم ْن َها‬ ‫َو َدا ِه َي ٍة َي َظ �ل ٱل �ن ُ‬

‫س بِ ِه �ن َب ْي ُت‬ ‫إِلى ِٕٱ�� ْح َرا ِم َل ْي َ‬ ‫ْت َو إ ِْن َغ َو ْي ُت‬ ‫لِ َنف ِْسي إ ِْن َر ِشد ُ‬ ‫َو َع ْيناً ُكل� َما ِشئ ُْت ٱس َت َق ْي ُت‬ ‫إِذَا َما َضا َم ِني شَ ْيءٌ ا� َب ْي ُت‬ ‫ت َُض �ي َع َيا َس َم ْوا�لُ َما َب َن ْي ُت‬ ‫َو��َ خَ شَ ٍب َو َم ْج ٍد َق ْد ا� َت ْي ُت‬ ‫َي ُؤ �م بِ�� َد َم ْل ٍك َق ْد َه َد ْي ُت‬

‫َو َل ْو ا�ن�ي ا�شَ ا ُء بِ َها َج َز ْي ُت‬ ‫يت‬ ‫َعزيزاً �� يُرا ُم إِذا َح َم ُ‬

‫إِذَا ما ُذ �م ا� ْق َوا ٌم َو َف ْي ُت‬ ‫فَ��َ َوٱلل� ِه ا� ْغ ِد ُر َما َمشَ ْي ُت‬ ‫ض ٱل ُب ُي ِ‬ ‫وت َل َق ْد َص َب ْو ُت‬ ‫إ َِلى َب ْع ِ‬ ‫َو ِم ْع َص َم َها ٱ ْل ُم َوش� َم َق ْد َل َو ْي ُت‬ ‫ِق َياماً بِٱ ْل َم َحار ِِف َق ْد َك َف ْي ُت‬ ‫‪Dīwān as-Samawʾal 96–101.‬‬

‫‪25‬‬

GLORY AND IMMORTALITY 1. Al-Khubayt is effaced [abandoned] by the tribe of Fāṭima up to al-Iḥrām there is no tent.

2. Oh you two railing women! I disobey your words! It is up to me whether I do right or wrong.

3. ʿĀdiyā has built me an unassailable fortress and a source: every time I want – I draw (water out of it).

4. A high-rising, from which eagles go down. When something harms me, I do not yield.

5. ʿĀdiyā, my grandfather commanded: May you not destroy, oh Samawʾal, what I have built!

6. A good many houses have I built without clay and without wood, and a good many honours [glory] have I won. 7. A good many forces, which in the darkness of the night, heavily equipped hold the course towards the land of the king, have I led. 8. A good many faults have I forgiven someone worthless and thoughtless, and I have spared him. 9.

So if I perish, yet I have given a proof of forgiveness, fulfilled the purpose and have been cured.

10. I move away from harsh (words) provoking me and if I wish, I could requite them.

11. If I protect the neighbor in a serious situation, he becomes strong, he is not abandoned when I protect him.

12. I fulfilled the promise with the armour of the Kindī. I am, when other people are blamed, faithful. 13. They said: it is an immense treasure!, but, by God, I will not betray as long as I live [lit. walk].

14. And were it not that they would have said: a decrepit man is passionate about (visiting a women in) a tent, I would have been pas-

119

120

KIRILL DMITRIEV sionate.

15. And a good many tents of a chaste (women) have I put my head in and have banded her tattooed wrist.

16. And a good many calamities, from which people search protection escaping, have I saved from with dexterity.

BIBLIOGRAPHY

Ahlwardt, W., The Divans of the Six Ancient Arabic Poets Ennābiga, ʾAntara, Tharafa, Zuhair, ʾAlqama and Imruulqais, London 1870. Arberry, A.J., The Koran Interpreted, 2 vols., London 1955. Averinzev, S., Authorship and authority [in Russian], in Istoričeskaya poetika. Literaturnye epohi i tipy hudožestvennogo soznaniya, Moscow 1994, 105–25. Bauer, Th., The Relevance of Early Arabic Poetry for Qurʾānic Studies Including Observations on Kull and on Q 22:27, 26:225, and 52:31, in The Qurʾān in Context. Historical and Literary Investigations into the Qurʾānic Milieu, ed. by Angelika Neuwirth, Nicolai Sinai and Michael Marx, Leiden 2010, 699–732. Bauer, Th., Al-Samawʾal b. ʿĀdiyā, in EI2, vol. 8, Leiden 1995, 1041–2. Caskel, W., Lihyan und Lihyanisch, Cologne 1953. Caskel, W., Al-Aʿshà Nr. 25, 6, in: Studi Orientalistici in Onore di Giorgio Levi Della Vida, vol. 1, Rome 1956, 132–40. Dīwān as-Samawʾal, ed. by Wādiḥ aṣ-Ṣamad, Beirut 1996. Gelder, G.J. van, Persons as Texts/Texts as Persons in Classical Arabic Literature, in Conscious Voices. Concepts of Writing in the Middle East, ed. Stephan Guth, Priska Furrer and Johann Christoph Bürgel. Beirut 1999, 237–53. Hausleiter, A., Das antike Tayma’: eine Oase im Kontaktbereich der Kulturen. Neue Forschungnen an einem Zentralort der Karawanenstraße, in Roads of Arabia. Archäologische Schätze aus Saudi Arabien, Berlin 2011, 102–23. Hoyland, R., The Jews of the Hijaz in the Qur’ān and in their inscriptions, in New Perspectives on the Qur’ān. The Qur’ān in its His-

GLORY AND IMMORTALITY

121

torical Context 2, ed. G.S. Reynolds, London 2011, 91–116. Hoyland, R., The Jewish Poets of Muḥammad’s Ḥijāz, in Judaïsme en Arabie, ed. by Christian Robin (forthcoming). al-Iṣbahānī, Abū al-Faraj, Kitāb al-Aghānī, 24 vols., Cairo 1927– 1974. Jacobi, R., Time and Reality in “Nasīb” and “Ghazal”, in JAL 16 (1985), 1–17. Jones, A., Early Arabic Poetry. Select Poems, Oxford 2011. Lane, E.W., Arabic-English Lexicon, vol. 1.1, London 1863. Lichtheim, M., Ancient Egyptian literature: a book of readings, vol. 2. The New Kingdom, Berkeley and London 1976. Lyall, Ch.J., Translations of Ancient Arabian Poetry, Chiefly Prae-Islamic with an Introduction and Notes, London 1885. Montgomery, J., The Vagaries of the Qaṣīdah. The Tradition and Practice of Early Arabic Poetry, Cambridge 1997. Nagy, G., Early Greek views of poets and poetry, in The Cambridge History of Literary Criticism, ed. George A. Kennedy, Cambridge 1989. Mufaḍḍalīyyāt. An Anthology of Ancient Arabian Odes, ed. Ch.J. Lyall, 2 vols., Oxford 1918–1921. Newby, G.D., A History of the Jews of Arabia from Ancient Times to their Eclipse under Islam, Columbia 1988. Sezgin, F., Geschichte des Arabischen Schrifttums, vol. 2, Leiden 1967. Sharḥ Dīwān ʿAntara, Beirut. Sharḥ Dīwān Zuhayr ibn Abī Sulmā, Cairo 1944.

TRACING THE RECEPTION OF THE PROTOEVANGELIUM OF JAMES IN LATE ANTIQUE ARABIA: THE CASE OF THE POETRY OF UMAYYA IBN ABĪ AṢ-ṢALT AND ITS INTERSECTION WITH THE QURAN CORNELIA HORN

As one scrutinizes the poetry of pre-Islamic Arabic origins and concentrates on traditions that may stand in some parallel relationship to Biblical material, whether of canonical or apocryphal origins, the works of two poets, ʿAdī ibn Zayd and Umayya ibn Abī aṣ-Ṣalt, emerge as prominent and relevant corpora. ʿAdī ibn Zayd’s work has been the subject of multiple investigations in more recent years. 1 Interest in Umayya ibn Abī aṣ-Ṣalt’s poetry has been somewhat more restricted. 2 When considering the work of this author, See for example Dmitriev, An Early Christian Arabic Account, Toral-Niehoff, Eine arabische poetische Gestaltung des Sündenfalls; Hainthaler, Christliche Araber vor dem Islam 90–3, Hainthaler, ʿAdī ibn Zayd al-ʿIbādī, and Horovitz, ‘Adi ibn Zayd, the Poet of Hira. 2 Finkel, Old Israelitish Tradition 7–21 considered some aspects of the relationship of the work of Umayya ibn Abī aṣ-Ṣalt and Biblical literature. See also Hirschberg, Jüdische und christliche Lehren. Umayya’s works have been examined also with regard to their conception of the divine. See Borg, The Divine in the Works of Umayya b. Abî al-Ṣalt. 1

123

124

CORNELIA HORN

scholars have occupied themselves to a large extent with discussions regarding the authenticity of poems attributed to him. 3 In an exploration and analysis of the matter in 1996, Tilman Seidensticker formulated with some confidence that “there might well be some authentic material among the nearly 900 lines ascribed to Umayya.” 4 At least there is a good possibility that some of this poetry is authentically Umayya’s, independent of whether or not the precise contours of that subset can be delineated. Yet in order to reach convincing conclusions that allow one to decide for or against the authenticity of a given text, it is necessary to establish criteria on the basis of which one might be able to judge a given piece of writing. Especially relevant, it seems, is the question of the role one ought to ascribe to the level of closeness of a given poem allegedly composed by Umayya to text that is also found in the Quran. Unless one treats the Quran a priori as being essentially different from other texts and, what is most relevant here, as a text that with its words, phrases, and formulations cannot refer to texts that were available in the cultural world from and in which the Quran arose, it has to be held as a distinct possibility that parallels between the Quran and the poetry ascribed to Umayya evince a form of intertextuality. Three possible and plausible explanations for such an intertextual relationship ought to be considered. Either the poems, whether complete or preserved only in fragments, which are ascribed to Umayya depend on the Quran, or the Quran depends on them, or both the poems and the Quran depend on a third, shared tradition, which in itself may originate in an oral or written communication. That third tradition can be thought of as a tradition that is shared between the Quran and the poetry attributed to Umayya insofar as it constitutes a point of intersection between the two. Whether or not the works ascribed to Umayya in the end have in fact been composed by him or rather by someone else is a question that is of interest and some consequence for the modern researcher’s ability to situate these texts more precisely. This concern is to be taken seriously. Where the authenticity of Seidensticker, The Authenticity of the Poems, and more recently Nagel, Zur Einführung xii–xvi. 4 Seidensticker, The Authenticity of the Poems 96. 3

THE PROTOEVANGELIUM OF JAMES

125

such poems and fragments of verses can be established, the dating of the material becomes significantly more accessible. Yet it is not a question that is all-decisive when determining the value of this poetry for discerning the geographical and cultural spread of a shared tradition in reaction to or in interaction with which both parts of the Quran and of the texts and fragments ascribed to Umayya emerged. Within the limits and for the purposes of the present study, the examination of elements of such a shared background or context concern primarily the spread of apocryphal Christian traditions in the milieu of early Islam. For Umayya ibn Abī al-Ṣalt’s later life, dates in the late sixth to the first half of the seventh century CE have been established. He may have died either in or around 608 CE or 630 CE. 5 He is said to have originated from aṭ-Ṭā’if, a city located about fifty miles south-east of Mecca. 6 Umayya is thought of as having been a member of the Thaqīf tribe and related to Meccan aristocracy. 7 Muḥammad ibn Ḥabīb (d. 244/859) is credited with having been the first to collect Umayya’s poetry in a dīwān, 8 to be followed in the twentieth century by several further scholars attempting to gather Umayya’s poetry. 9 Yet a full critical examination of UmayBorg, Umayya b. Abī al-Ṣalt as a Poet 4 thinks he died around 608 CE. Huart, A History of Arabic Literature 25 had him die in 630 CE. See also Huart, Une nouvelle source 134; and Busse, Islam, Judaism, and Christianity 27. 6 For a convenient summary of Umayya’s biography see Seidensticker, The Authenticity of the Poems 88. On Umayya and his work, see also Huart, Une nouvelle source; Schultheß, Umajja b. Abi-ṣ Ṣalt; and Power, Umayya ibn Abī-ṣ-Ṣalt. 7 For a discussion of reports related to interactions between Muḥammad and Thaqīf, see Kister, Some reports. 8 Huart, Une nouvelle source 141 and 144, Seidensticker, The Authenticity of the Poems 88, and Reynolds, The Qur’ān and Its Biblical Subtext 32. 9 Schultheß, Umajja ibn Abi ṣ Ṣalt, Power, The poems of Umayya b. Abi-s-Salt; Yamut, Diwān Umayya ibn Abī al-Ṣalt; Bahja ʿAbd al-Ghafūr alḤadīthī, Umayya b. a. al-Ṣalt. Ḥayātuhu wa-Shi’ruhu, ʿAbd al-Ḥāfiẓ as-Saṭlī, Umayya b. a. al-Ṣalt; Vallaro, Umayyah ibn Abī ṣ-Ṣalt, and al-Jubaylī, Diwān 5

126

CORNELIA HORN

ya’s work is still a scholarly desideratum. That he had gained fame as an author of poetry that could be related to Jewish and Christian sacred texts, particularly the psalms, is in evidence relatively early on. A fragment by Surāqa al-Aṣghar (d. ca. 85/700) witnesses to that fact already 160 years before ibn Ḥabīb’s collection, when it says: “(Remember Labīd …) and Umayyya, the ‘sea’ (of knowledge) in whose poetry are pieces of wisdom like a detailed revelation in the Psalter.” 10 Umayya’s poems deal prominently with topics that are relevant for religious questions, for example matters related to the creation of the world, the service of the angels, the great Flood, or the creation of humankind. The frequency of such religiously relevant themes is unusual when compared to compositions of preIslamic Arab poets who are known not to have been Jews or Christians. 11 In and of itself, Umayya’s interest in such themes would not have to be considered unreasonable. Yet it might be difficult to explain had he been completely unfamiliar with and unreceptive of the Jewish and/or Christian tradition. There is, however, not much evidence that would support the claim that he was in fact a Jew or a Christian. 12 For Muslims, identifying him as a ḥanīf, and thus as a monotheist, was a plausible explanation for his identity. 13 Given the relative prominence of religious material of a Jewish or Christian character in the corpus ascribed to Umayya, this poetry invites a multi-layered analysis that pursues discerning aspects of the intertextual functions of these religious traditions. In order to situate the investigation more firmly at the intersection of Christian and Late Antique Arabic traditions, the present focus is on figures familiar from the New Testament. Among the poems ascribed to Umayya ibn Abī al-Ṣalt. 10 (wa-dhkur Labīdan …) wa-Umayyata l-baḥra lladhī fī shi’rihī ḥikamun ka-waḥyin fī l-zabūri mufaṣṣalī. See Seidensticker, The Authenticity of the Poems 96 (with modification). 11 See Seidensticker, The Authenticity of the Poems 87. For some evaluation of religious sentiment in pre-Islamic poetry, see Sayuti, The Concept of Allāh. 12 Seidensticker, The Authenticity of the Poems 87. 13 See the comments in Frank-Kamenetzky, Untersuchungen 48; and Nicholson, A Literary History of the Arabs 149–50.

THE PROTOEVANGELIUM OF JAMES

127

Umayya, one finds a section of text—seventeen verses in length— that deals mainly with Jesus’ mother Mary and in a less prominent manner also with Jesus as a child. Some aspects of Umayya’s presentation of the story call to the mind of a reader what she or he may have heard from the Gospel narratives concerning Mary’s conception of and giving birth to Jesus. Yet the identification of details with what a reader can find in the canonical text remains incomplete. Many aspects do not correspond to the biblical text. Rather, the verses reveal elements of popular traditions known from other ancient texts about Jesus and his mother. Quite independent of the question of the authenticity of this text as part of a work by Umayya, the verses witness to the spread of apocryphal Christian infancy-of-Jesus traditions in ancient Arabia. In 1911 Friedrich Schultheß published the poem that is here being investigated as number 38 of his collection of what was preserved of Umayya’s work. 14 Whereas many items which one finds in Schultheß’ study are fragments of larger poems that have not come down to modern times as a whole, no elements of text 38 indicate that only a portion of a larger work is preserved in this instance. Upon closer examination, the reader notices that the text contains some parallels to segments of verses in various sūras in the Quran. This very question of parallels between the Quran and the poetry ascribed to Umayya was the subject of a 1911–Königsberg doctoral dissertation by Israel Frank-Kamenetzky. On three pages, FrankKamenetzky collected his evidence for parallels in the Quran to material in poem 38. 15 He concluded that in two verses, 10 and 15, one sees direct borrowing from the Quran while for the rest of the verses text 38 offers a poetic retelling of sūra 19. 16 In the end, he was inclined to group poem 38 with those texts, for which Umayya’s authorship was to be discounted. 17 While valuable as a basis from which to work, FrankKamenetzky’s study presents the reader with some problems and displays limitations. In light of new directions in Quranic studies Schultheß, Umajja ibn Abi ṣ Ṣalt 48–9. Frank-Kamenetzky, Untersuchungen 20–2. 16 Frank-Kamenetzky, Untersuchungen 38 and 46. 17 Frank-Kamenetzky, Untersuchungen 47–8. 14 15

128

CORNELIA HORN

that take the relevance of the cultural context of pre-Islamic Arabia as a factor that shaped the text of the Quran, carefully revisiting and reevaluating the results of Frank-Kamenetzky’s study has become a necessity. One of the problems with his analysis rests with his neglect to consider to what extent it was possible that at a given instance at which one might discern a parallel between one or several of the verses of Umayya’s or pseudo-Umayya’s poetry and the Quran, both texts, potentially independent of one another, could have drawn on or reacted to a common third tradition. Such a tradition, with which they could have interacted, would therefore constitute a point of intersection for Umayya’s poetry and the Quran. The following observations revisit some aspects of FrankKamenetzky’s study of the material in poem 38 in light of that possibility. In their analysis they go beyond the considerations of a shared source, which was advanced by Joshua Finkel and then by Joachim Hirschberg in his 1939–Cracow dissertation. 18 In several instances the Quran and Umayya’s poem 38 share specific vocabulary. Yet in most cases parallels in content are not articulated by way of employing strictly identical terminology. For instance, Umayya 38:3 and Q 66:12 (and 21:91) express Mary’s chastity through the noun “farju.” Yet poem 38 uses this noun in a prepositional phrase whereas the Quran employs it as object of a verbal expression meaning “to guard.” The Quran merely tells that “Mary guarded her chastity.” Poem 38 on the other hand offers a much more detailed formulation, emphasizing that “she did not strive to get married and did not engage with men, neither sexually nor in conversation.” With its comments on Mary refraining from orally communicating with a man, which are not found in the Quran, the text of poem 38 could have suggested to a reader who was aware of apocryphal Christian traditions that Umayya’s poem shaped its presentation in parallel to the depiction of Mary in the context of the annunciation as found in the Protoevangelium of James. 19 In that apocryphal Christian text, an Arabic text of which is See Finkel, Old Israelitisch Tradition; and Hirschberg, Jüdische und christliche Lehren 37–8. 19 For access to the Greek text of the Protoevangelium of James, accompanied by a French translation, see de Strycker, La forme la plus ancienne. An 18

THE PROTOEVANGELIUM OF JAMES

129

evidenced in manuscript form (MS Sinai Arab. 436) dating to the tenth century CE at the latest, 20 the reader encounters Mary trembling when she first heard the voice at the well talking to her. The reader also notices that later on, when the angel addressed his words to her, Mary responded with great restraint. 21 The apocryphal Protoevangelium of James and Umayya’s poem 38 share details of the presentation of Mary with one another that are not found in the Quran. At the same time, concern for and confirmation of Mary’s chastity are significant themes advanced by the Protoevangelium of James. Poem 38 may have received inspiration for its own emphasis on that theme from its acquaintance with the same concern expressed in the Protoevangelium of James. The likelihood of such a familiarity with that apocryphal text also is supported through observations that arise from other sections of the poem. A comparison of Umayya’s verse 38:4 to possible Quranic parallels reveals that the shared vocabulary between verse 38:4 and Q 19:16–17 amounts to two nouns, namely “hijāb” and “ahl,” as well as the prepositional phrase “min dwn.” It is instructive to analyze more deeply the usage and context of the noun hijāb, “screen,” here. In contrast to Q 19:17, which simply employs the noun hijāb and leaves open what kind of a screen is envisioned, poem 38:4 offers a further identification when it speaks more specifically of edition of the Greek, which is largely but not exclusively based on de Strycker’s edition and which is accompanied by an English translation, is also found in Hock, The Infancy Gospels of James and Thomas 32–77. 20 The critical edition of the Arabic text of the Protoevangelium of James is not completed. For the publication of a section comprising chapters 1– 15 see Garitte, ‘Protevangelii Jacobi’. Using mss. Sinai Arab. 535 and 436, Garitte based his transcription of the text on manuscript evidence from the tenth and twelfth centuries. For further comments see Horn, Apocryphal Gospels in Arabic 589–92. For discussion of the reception of the Protoevangelium of James in the Coptic- and Arabic-speaking milieux in Egypt, see Horn, Mary between Bible and Qur’ān. For the reception of the text into the milieu in which Islam was emerging, see Horn, Intersections. 21 See Protoevangelium of James 11.2 (ed. and tr. de Strycker, La forme la plus ancienne 114–17).

130

CORNELIA HORN

the “screen of the house (ḥijāba l-bayti).” If one compares these references to a screen that functions to separate Mary from her people with the description of Mary’s earliest years at home as found in the Protoevangelium of James one notices in the apocryphal text a scene featuring Mary when she was only half a year old. 22 At that point, in anticipation of the girl’s later stay in the Temple her mother decided to create a sanctuary for her daughter in her own house. The Greek reads: ἐποίησεν ἁγίασμα ἐν τῷ κοιτῶνι αὐτῆς, “She made a sanctuary in her bedroom,” whereas the Arabic has: wajaʿalat la-hā bayta qudsi fī baytihā, “She set up for her a sanctuary in her house.” This was done so that the little girl’s foot would not touch anything impure again. If viewed against such a scene in the background, the reference to a “screen of the house” in Umayya’s poem 38 strikes one as being more directly suggestive of a parallel to the imagery used in the Protoevangelium of James than what one finds in the Quran. A further shared element on the level of vocabulary between the Quran and Umayya’s poem 38 consists of the reference to “her people” from whom Mary is being separated. The commonality of expression used for the description of the scene of separation however does not extend to the verbs that are used to speak of the separation from “her people.” Both passages feature the place to which Mary goes, yet those places are not identified in the same manner. While the Quran speaks of “a place in the East,” Umayya 38:4 refers to the “desert of Damdam.” In addition, one observes that whereas the Quran speaks of Mary withdrawing to a remote place after she had conceived the child, it never identifies such a place—neither the “place in the East,” nor the “remote place” (reached after conception)—as a desert. In contrast to this, Umayya’s poem 38:4 speaks of her going to the desert area before the conception has taken place and does not again speak of a desert place. In this case, the Quran’s description may be closer to an account of events found for example in the Protoevangelium of James. There, after her conception of a child, one finds Mary in the desert, having been sent to it as part of the test of her chastity through the ritual of drinking the water of the curse (see Numbers 5:11–31). Protoevangelium of James 6.1 (ed. and tr. de Strycker, La forme la plus ancienne 90–1, ed. and tr. Garitte, ‘Protevangelii Iacobi’ 386, lines 4–5). 22

THE PROTOEVANGELIUM OF JAMES

131

Further expansions that are part of Umayya’s poem 38, for instance when in verse 5 that text offers even an additional verse with comments on the quality of the desert to which Mary withdrew, could be seen as evidence that poem 38 participated in its own right in the project of creating apocryphal traditions. They support the observation that the author of poem 38 was confident to shape his own thought or the material he reworked in line with what was important to him. Summarizing the results for the comparison of Umaya’s verse 38:4 to possible Quranic parallels, one notes that beyond the shared vocabulary of two nouns and a prepositional phrase, verbs in the respective contexts are not identical. Also, the specific order in which actions are described is different. Nevertheless, the outcome of events, namely that at the end there was a screen between Mary and her people, and that she found herself in a different place, are the same. One can detect noteworthy parallels to apocryphal Christian traditions. Yet in this passage the number of potential parallels to the Protoevangelium of James seems to be evenly distributed between Umayya’s poem 38 and the Quran. One might conclude that in this instance the data is insufficient to claim that poem 38 and the Quran depend on one another, no matter in which direction. It could very well be the case that both developed their own account of events on the basis of a story that had existed already in their milieu, to which they had access, and the knowledge of which they could presuppose among their audience. While the reader may detect parallels between the Quran and Umayya’s poem 38 on the level of the choice of identical vocabulary used in specific text passages, at some instances those parallels do not extend to the grammatical form of a given root. The only direct parallel between Q 19:17 and Umayya’s verse 38:6 is that the Quran speaks of God sending someone to Mary whereas verse 38:6 uses rasūl as noun to designate the one sent to her. In this set of evidence, the Quran seems to be interested in formulating that the one sent, referred to as ruhana, “our spirit,” was in everything like a human being. This detail may suggest a familiarity on the part of the intended audience of the Quran with traditions witnessed to in apocryphal Christian texts. In the Protoevangelium of James, Mary

132

CORNELIA HORN

asked the angel whether her conception of a child would be like to the normal manner in which a woman conceived. 23 Given the setting, a reader may well imagine at this point that here Mary was thought of as having felt threatened and that she anticipated that the angel might approach her physically, engaging her in a sexual act. In the Quran, God’s spirit is featured precisely like such a normal human being, even suggesting that the human being was of a shapely appearance. Sūra 19:17 formulates “and we sent to her our spirit (rūḥana) and it appeared to her as a human being of regular built (basharan sawīyan).” With regard to the nature and characterization of the figure, who was sent to Mary, greater common ground seems to exist between the presentation found in the Quran and in the Protoevangelium of James. Yet one also notices that Umayya’s verse 38:6 places relatively more emphasis on the eloquence of the “messenger” than the Quran. In this regard, Umayya’s poem is closer to the role assigned to voice that can be detected in the Protoevangelium of James especially in those places, at which direct references to a voice are featured as well as in the several places at which Mary is presented as hearing these voices. Next to connections on the level of vocabulary between Umayya’s poem 38 and the Quran, 24 one may find identical or very similar ideas being expressed by way of employing variant terminology. Umayya 38:10, which offers as a conditional clause “if you are a believer,” may be regarded as a parallel to Q 19:18, a verse that formulates a similar idea in the conditional clause “if you are one who fears [God].” A possible parallel that Frank-Kamenetzky did not identify and analyze fully is to be found in the relationship between Umayya 38:13–14, which speaks of Mary’s surroundings reacting to her negatively as she was about to give birth, and Q 19:27–28, which has her bring her baby to her people who then offer their criticism. Further comments on that parallel are offered below. Protoevangelium of James 11.2 (ed. and tr. de Strycker, La forme la plus ancienne 114–17). 24 See also Umayya’s verse 38:13 and Q 3:36, which may stand in parallel to one another on the basis of the use of the word waḍʿ. 23

THE PROTOEVANGELIUM OF JAMES

133

When comparing the poetry ascribed to Umayya with the Quran one also finds some similarities in the structure of the expression of a given thought. Umayya 38:8 and Q 19:19 are one example of such structural parallelism in so far as a good portion of the material in these verses is similar in arrangement. In addition, these verses also share some vocabulary. Umayya 38:8 has the visitor to Mary say, “I am a messenger from the merciful one,” and Q 19:19 offers “I am the messenger of your lord.” Whereas Umayya 38:8 employs the term ibn, “son,” to speak of what the messenger announces to bring to Mary, Q 19:19 reads ghulāman, “a young man.” For this passage, no immediate parallel to an apocryphal text can be discerned. The content, nevertheless, is close to Luke 1:30– 31, which has “and the angel said to her: … you will give birth to a son (huios).” When set against the background of the two verses from Luke, one might conclude that Umayya’s ibn is closer to huios than the Quran’s ghulām. As one considers the construction of the scene of the annunciation to Mary as presented in Umayya 38 one may add an additional observation that Frank-Kamenetzky did not formulate, possibly because he did not consider potential relations of the material he studied to apocryphal texts. In Umayya 38:7, the messenger tells Mary not to think that angels that come from the Lord of ‘Ād and Jurhum are liars. The specification of a plural for “angels” may be related to the multiple times that either only a voice or an angel speaks to Mary in the course of the annunciation scene in the Protoevangelium of James. 25 In his study of the relationship between the Quran and the poetry ascribed to Umayya, Clément Huart observed that in at least some instances where the two texts seem to share material, the poetry ascribed to Umayya offers a surplus of information. 26 One such case is to be found at Umayya 38:9, which has Mary wonder how she should have a child since she was “not a whore, not pregnant, and not a married woman.” At Q 19:20 (= Q 3:47) one reads, “no man has touched me and I am not unchaste.” Such cases See Protoevangelium of James 11.2 (ed. and tr. de Strycker, La forme la plus ancienne 114–17). 26 See for example Huart, Une nouvelle source 160–1. 25

134

CORNELIA HORN

might indicate that the Quran shortened what was available in the fuller or longer version of the story that was ascribed to Umayya. A similar line of argument which presumes that a shorter presentation of a story is the later and derivative one also underlies at least one line of solutions to the so-called synoptic problem in New Testament studies. There scholars argue that insofar as Matthew and Luke’s versions of individual episodes are found to be shorter than what Mark has to offer, they support the argument of Markan primacy. 27 Whether or not this argument is convincing is a different matter. When comparing Umayya 38:9 and Q 19:20 (= Q 3:47) with canonical and apocryphal Christian texts, one may note that they share the added concern for Mary’s chastity. They do so over and against Luke 1:34 and a passage in the History of the Blessed Virgin Mary. The History of the Blessed Virgin Mary, a Syriac apocryphal text, the roots of which go back at least to the early sixth century, has Mary wonder how she could “bear a son” given that she has “known no man.” 28 In this text Mary is convinced that “a thing such as this, that is to say, virgins giving birth to children without union with a man, has never been heard of from the days of the beginning of the world.” 29 Umayya 38:9 with its comment that This standard argument in discussions of the synoptic problem has also entered the realm of the introductory New Testament textbook. See for example Harris, The New Testament 112–13. 28 History of the Blessed Virgin Mary (ed. and tr. Budge, The History of the Blessed Virgin Mary 20 [Syriac] and 22 [English]). 29 History of the Blessed Virgin Mary (ed. and tr. Budge, The History of the Blessed Virgin Mary 20–1 [Syriac] and 23 [English], translation modified). Quite a similar concern is expressed in the words placed into Mary’s mouth in a Syriac dialogue poem featuring Mary and the angel announcing her conception to her. That sogitha, sometimes attributed to Narsai, sometimes to Ephraem the Syrian, likely dates to the fifth or sixth century. In the dialogue poem, Mary keeps trying to show the futility of the angel’s message, declaring to him that “nature itself can well convince me that virgins do not ever give birth” (strophe 24), that “a son in a virgin is not to be seen, and no one has ever slept with me” (strophe 32), and that “no man has ever known me, nor any ever slept with me. How can this 27

THE PROTOEVANGELIUM OF JAMES

135

Mary wondered how her conception of a child might happen since she was not married stands in parallel to this passage. Where the History of the Blessed Virgin Mary presents Mary as stating that it has not been heard that virgins give birth to children without union with a man, the Marian character’s self-characterization also implies that she herself was a virgin and not married. Comments above have already indicated that with regard to the manner of Mary’s conception, the Protoevangelium of James may be connected to the thought of the Quran when in the Protoevangelium of James Mary actively reflects about the manner of conception and, while scrutinizing the angel’s words, says, “Shall I conceive from the Lord the living God the same way that every woman becomes pregnant?” 30 This emphasis of the Protoevangelium of James may be in the background of the Quran laying stress on the prominently human character of the spirit coming to Mary. Israel Frank-Kamenetzky proposed to place Umayya 38:14 in parallel to the text found at Q 19:27. 31 Although some ideas may be parallel in these two passages, one does not encounter parallel terminology. On the level of ideas, verse 38:14 connects the moment at which Mary is giving birth with comments concerning the frustration over criticism of her and the admonition that she should do be, what you have said, for without such union there will not be a son” (strophe 34). See Brock, tr., Bride of Light 114–15. See Brock, Bride of Light 13–14 for comments regarding the dating and authorship of the work. 30 Protoevangelium of James 11.2 (ed. and tr. de Strycker, La forme la plus ancienne 114–117). For the early reception of this material in the Syriac tradition, see the Syriac version of the Protoevangelium of James (ed. and tr. Smith Lewis, Apocrypha Syriaca 10 [Syriac] and 5 [English]: “Shall I conceive and bring forth from the Lord, as all women bring forth?” The palimpsest manuscript evidence for this text dates to the second half of the fifth or the early part of the sixth century. See Smith Lewis, Apocrypha Syriaca x. For the Arabic witness, dated to the twelfth century, see Garitte, ‘Protevangelii Iacobi’ 391. 31 Frank-Kamenetzky, Untersuchungen 21 refers to Q 19:28 by number, but offers the text of Q 19:27. The numbering for Quran verses used in the present article follows the edition provided in ʿAlī, The Meaning of The Holy Qur’ān.

136

CORNELIA HORN

penance for having done something wrong. Towards the end of the verse, Q 19:27 features Mary bringing her new-born child to her people, and they in turn comment on the situation. The remarks uttered by “her people” in the Quran are less negative, given that the text has “her people” speak of the child as an “amazing thing.” Frank-Kamenetzky did not note the parallel between Umayya’s verse 38:13 and Q 19:27, where both texts unveil Mary’s trouble when she has to deal with criticism of her actions. For the most part, Frank-Kamenetzky only lists the text of verses that might be parallel, but does not provide any analysis. At verse 38:14 in the poem attributed to Umayya one reads that those of her surroundings criticized Mary for walking upon reprehensible or forbidden paths, an action for which the law would have had her to be shamed and stoned. A similar idea may be in the background of Q 19:28, which implies the characterization of Mary, the “sister of Aaron,” 32 as having done evil and acted unchastely, quite in contrast to her parents. Here the Quran merely reports what people thought Mary’s condition was when she acted. Yet when Umayya 38:14 has those of her surroundings suggest that she should be shamed and stoned, these remarks go further and draw what might be regarded as legal consequences of her actions. In addition, Umayya 38:14 could express the idea of punishment required for Mary’s action that also underlies the judgment brought upon her at the hand of the high-priest in the Protoevangelium of James. 33 In that text the high-priest required of both Mary and Joseph to undergo the test of the water of the curse, which, if Mary had been guilty, allegedly would have brought bitter pain upon her. 34 One might On the seeming confusion of various figures named Mary / Miriam in the Biblical text as well as in the Quran, see the comments in Horn, Frühsyrische Mariologie 329–32. 33 Protoevangelium of James 16.1–2 (ed. and tr. de Strycker, La forme la plus ancienne 138–9). 34 Numbers 5:27–8. For discussions of the passage in Numbers 5 that deals with the test of adultery, see for example Miller, Another look; Briggs, Reading the sotah text (Numbers 5:11–31); Haberman, The suspected adulteress; von Nordheim, Das Gottesurteil; Frymer-Kensky, The 32

THE PROTOEVANGELIUM OF JAMES

137

formulate that Umayya 38 and the apocryphal Christian text share an interest in justice being served that is not found in the Quran. Both passages feature negative reactions on the part of others to Mary being pregnant or giving birth to a child based on the assumption that Mary had had illegitimate intercourse. In both instances, knowledge of a tradition of criticism Mary experienced regarding her pregnancy from the side of the Jewish priesthood with the accompanying testing of the woman suspected of adultery could stand in the background. Given that the Protoevangelium of James presents the scene of criticism and testing as occurring before Mary delivered her child, 35 that Umayya 38:13 shows her as experiencing the criticism in immediate connection with giving birth to the baby, whereas Q 19:27–28 envisions a scene at which the delivery of the baby is already an event of the recent past, it might be possible to argue also on the basis of this observation that Umayya’s verse is closer to the portrayal found in the Protoevangelium of James than what is offered in the Quran. Toward the end of the text of Umayya 38, a pronounced parallel to Q 19:30 in motif and vocabulary is to be noted. Both introduce the terminology of prophethood, with the text ascribed to Umayya having God “letting a speaking prophet say the truth” and Q 19:30 having Jesus proclaim that God “has … made me a prophet.” Subsequently, the speaking prophet in Umayya 38:16 declares, “he [God] has equipped me with knowledge,” while in Q 19:30 that prophet-child announces that God “has given me a revelation.” Beyond commenting on the shared ideas of prophecy and being equipped with revealed knowledge, it is noteworthy that in both instances the prophet in question speaks, and does so in what one might call a rather miraculous manner. For Frank-Kamenetzky, the reference to “a speaking prophet” in Umayya 38:15 presumes a somewhat forced move of the text to stay close to the content of Strange Case; Milgrom, Case of the Suspected Adulteress; and Fishbane, Accusations of Adultery. For a study of aspects of the medieval interpretation of the passage, see for example Zemler-Cizewski, Rupert of Deutz and the law of the stray wife. 35 Protoevangelium of James 15–16 (ed. and tr. de Strycker, La forme la plus ancienne 130–41).

138

CORNELIA HORN

the Quran which at 19:29 also features the speech of a child (prophet). 36 Yet it might be more constructive to consider the text of a Christian apocryphon that might lie behind both texts. Examining the Arabic Apocryphal Gospel of John has already proven to be a fruitful undertaking in the search of witnesses to shared traditions between Late Antique Christian and early Islamic materials. Elsewhere, a limited range of aspects of the parallel between the Quran’s presentation of the child Jesus speaking as a newborn in defense of his mother Mary and the depiction of a rather similar scene in the Arabic Apocryphal Gospel of John was the subject of investigation. 37 The Arabic Apocryphal Gospel of John, a littleknown apocryphal gospel that is now preserved only in Arabic, has a lost Syriac Vorlage which dates to pre-Islamic late antiquity. 38 The earliest manuscript witness to the copying of the Arabic text of the work, MS Sinai Arabic 441, dates to the late twelfth century CE. Based on internal evidence in the text, especially in the form of data on creative apocryphal traditions and missionary activities of various apostles, Michel van Esbroeck proposed that the Arabic witness preserves traditions that go back to at least the eighth century. 39 The preliminary study of the complete manuscript witnesses that are at hand for this text lead to the proposition of dating the production of the Arabic translation from the Syriac original to ca. 800 CE, that is, within reasonably close proximity to the compilation and redaction of the Quran as well as the compilation of Umayya’s dīwān. In the Arabic Apocryphal Gospel of John, reference to a list of miracles or “signs” related to Mary and/or Jesus is a prominent feature. Umayya’s poem 38 displays great interest in the idea of “sign.” In fact, the poem is framed by an introductory verse that Frank-Kamenetzky, Untersuchungen 22. See Horn, Syriac and Arabic Perspectives 289–90. 38 See Moraldi, Vangelo Arabo apocrifo 21–2; Horn, Apocryphal Gospels in Arabic 605–6; Horn, Syriac and Arabic Perspectives 287–8, and Horn, Editing a Witness. A critical edition and translation of the Arabic Apocryphal Gospel of John is in preparation. 39 Van Esbroeck, A propos de l’Évangile apocryphe arabe attribué à Saint Jean. 36 37

THE PROTOEVANGELIUM OF JAMES

139

announces a “prophesying sign (‘āyatun munabbi’atun) from the Lord of Mary.” The “speaking prophet” of Umayya 38:15 identifies himself as a “sign of God” in Umayya 38:16. 40 The “sign” speaks to defend Mary in Umayya 38:16–17, stating that he was not sent in shame and sin. Whereas the verses of Q 19:27–32 in their reference to accusations against Mary and the words of the infant speaking are more diffuse and rather vaguely deal with the need of offering a defense of Mary, the verses of Umayya 38:16–17, in their explicit approach to the need of a defense of Mary and the related sanctity of the child-prophet, fit quite closely with the presentation found in the Arabic Apocryphal Gospel of John. In that apocryphal gospel, the newborn child referred to his birth as a fulfillment of the prophecy of Isaiah 7:14, offered a defense of his mother’s virginity, and explained that the goal of his incarnation from the pure Lady Mary was the redemption of humankind. 41

CONCLUSIONS

The preliminary analysis of the possible interplay between the Quran, passages from the poetry ascribed to Umayya ibn Abī aṣṢalt, and ancient Christian apocryphal texts and traditions allows one to reach at least some preliminary conclusions. The Protoevangelium of James is one of the most popular Christian apocrypha and was widely received in late antiquity. 42 It was available in Syriac For Frank-Kamenetzky, Umayya 38:16 stands in relationship to Q 19:21, that has the messenger announce that the child to be received by Mary is a sign. Yet Frank-Kamenetzky does not exhaust the possible range of explanations for the relationship between these two passages. In particular, he does not consider Christian apocrypha, like the Arabic Apocryphal Gospel of John. 41 Arabic Apocryphal Gospel of John 5.7 (ed. and tr. Galbiati, Iohannis evangelium apocryphum arabice 33–4 [Arabic] & 30–1 [Latin], tr. Moraldi, Vangelo arabo apocrifo 54–5 [Italian]). 42 The literature on the Protoevangelium of James is quite voluminous and cannot be presented in any greater detail here. An important gap in the research and scholarly treatment of this text is the lack of a sustained commentary on the work. Toepel, Protevangelium has provided a new survey and assessment of the literature. Moreover, several relatively recent 40

140

CORNELIA HORN

translation and then came to be incorporated in a larger Life of Mary, a composite apocryphal text, likely from the fifth century. 43 Arabic translations of the Protoevangelium of James are attested in the manuscripts kept at the Monastery of Saint Catherine’s in Sinai. Those manuscripts that contain relevant text of the Protoevangelium of James date back at least to the tenth century CE (the earliest being MS Sinai Arab. 436). 44 That in several passages in Quran 3 and 19 there are parallels to the Protoevangelium of James has been clearly documented in research by Suleiman Mourad and the present author. 45 It should not come as a surprise then that also pre-Islamic or early Islamic Arabic poetry displays some acquaintance with this text, manifestly parallel to, but also beyond the parallels between the Quran and the Protoevangelium of James. Neither may one be astounded that parallels can be made likely between early Arabic poetry and other Christian apocrypha, like the History of the Blessed Virgin Mary and the Arabic Apocryphal Gospel of John, both of which are familiar with the Protoevangelium of James. The parallels that can be detected between Umayya 38 and the Quran do not suffice to establish a clear dependency either of poem 38 upon the Quran or of the Quran on Umayya 38. If there existed a presumably Christian retelling of material related to canonical and apocryphal stories of the annunciation and the birth of Jesus, even if it was available only orally in Arabic, it is not evident that Umayya 38 would have to depend on the Quran. The Quran does not ever literally correspond with what the poem ascribed to Umayya has to say and so does not have to have been Umayya’s publications have presented ample evidence of the spread and reception of the Protoevangelium of James in the Late Antique Christian East. See Horn, Mary between Bible and Qur’ān, and Horn and Phenix, Apocryphal Gospels in Syriac 531–7. 43 For the discussion of some of the cultural aspects that accompanied and shaped the document, see Horn, From Model Virgin to Maternal Intercessor. 44 For a brief description of the manuscript, see Atiya, A hand-list of the Arabic manuscripts 12–13. 45 See Mourad, On the Qur’ānic Stories about Mary and Jesus; and Horn, Intersections.

THE PROTOEVANGELIUM OF JAMES

141

source or in turn have been dependent upon Umayya. The evidence discussed in this study does suggest that apocryphal Christian traditions pertaining to the annunciation to Mary and the birth and earliest childhood of Jesus were present in and widely received in the cultural milieu, upon which early Arabic poetry drew, quite independent of whether said verses ever were recognized to be canonical or not.

BIBLIOGRAPHY

ʿAbd al-Ghafūr al-Ḥadīthī, B., Umayya b. a. al-Ṣalt. Ḥayātuhū waShiʿruhū, Baghdad 1975. ʿAbd al-Ḥāfiẓ as-Saṭlī, Umayya b. a. al-Ṣalt, Damascus 1977. ʿAlī, ʿAbd. Y., The Meaning of The Holy Qur’ān, Beltsville, Maryland 2001. Atiya, A.S., A hand-list of the Arabic manuscripts and scrolls microfilmed at the library of the Monastery of St. Catherine, Mount Sinai, Baltimore 1955. Borg, G., The Divine in the Works of Umayya b. Abî al-Ṣalt, in Borg, G. and E. de Moor (eds.), Representations of the divine in Arabic poetry, Amsterdam 2001, 9–23. Borg, G., Umayya b. Abī al-Ṣalt as a Poet, in Vermeulen, U. and D. De Smet (eds.), Philosophy and Arts in the Islamic World. Proceedings of the Eighteenth Congress of the Union Européenne des Arabisants et Islamisants held at the Katholieke Universiteit Leuven, September 3—September 6, 1996 (Orientalia Lovaniensia Analecta 87), Leuven 1998, 3–13. Briggs, R.S., Reading the sotah text (Numbers 5:11–31): holiness and a hermeneutic fit for suspicion, Biblical Interpretation 17.3 (2009), 288–319. Brock, S. (tr.), Bride of Light. Hymns on Mary from the Syriac Churches, Baker Hill, Kottayam, Kerala, India 1994. Budge, E.A.W., The History of the Blessed Virgin Mary and the History of the Likeness of Christ Which the Jews of Tiberias Made to Mock at, 2 vols., London 1899. Busse, H., Islam, Judaism, and Christianity: Theological and Historical Affiliations, translated from the German by A. Brown (Princeton Series on the Middle East), Princeton, 1998.

142

CORNELIA HORN

De Strycker, É., La forme la plus ancienne du Protévangile de Jacques. En appendice: Les versions arméniennes traduites en Latin par Hans Quecke (Subsidia Hagiographica 33), Bruxelles 1961. Dmitriev, K., An Early Christian Arabic Account of the Creation of the World, in Neuwirth, A., N. Sinai and M. Marx (eds.), The Qur’ān in Context. Historical and Literary Investigations into the Qur’ānic Milieu (Texts and Studies on the Qur’ān 6), Leiden and Boston 2010, 349–87. Finkel, J., Old Israelitish Tradition in the Koran, Proceedings of the American Academy for Jewish Research 2 (1930–31), 7–21. Fishbane, M., Accusations of Adultery: A Study of Law and Scribal Practice in Numbers 5:11–31, Hebrew Union College Annual 45 (1974), 25–45. Frank-Kamenetzky, I., Untersuchungen über das Verhältnis der dem Umajja b. Abi ṣ Ṣalt zugeschriebenen Gedichte zum Qoran, Ph.D. thesis, Königliche Albertus-Universität zu Königsberg in Preussen, Kirchhain N.-L. 1911. Frymer-Kensky, T.S., The Strange Case of the Suspected Sotah (Numbers 5:11–31), in Vetus Testamentum 34.1 (1984), 11–26. Galbiati, I. (ed. and tr.), Iohannis evangelium apocryphum arabice, Mediolani 1957. Garitte, G., ‘Protevangelii Jacobi’ versio arabica antiquior, in Le Muséon 86 (1973), 377–96. Haberman, B.D., The suspected adulteress: a study of textual embodiment, in Prooftexts 20.1–2 (2000), 12–42. Hainthaler, Th., Christliche Araber vor dem Islam. Verbreitung und konfessionelle Zugehörigkeit. Eine Einführung (Eastern Christian Studies 7), Leuven, Paris and Dudley 2007. Hainthaler, Th., ʿAdī ibn Zayd al-ʿIbādī, the Pre-Islamic Christian Poet of al-Hīra and his Poem nr. 3 written in Jail, in Parole de l’Orient 30 (2005), 157–72. Harris, S.L., The New Testament. A Student’s Introduction. Sixth Edition, Boston, New York, and elsewhere 2009. Hirschberg, J.W., Jüdische und christliche Lehren im vor- und frühislamischen Arabien. Ein Beitrag zur Entstehungsgeschichte des Islams / Zydowskie i chrześcijańskie nauki w Arabii pogańskiej i w pierwszym okresie islamu (Prace Komisji Orientalistycznej—Polska

THE PROTOEVANGELIUM OF JAMES

143

Akademia Umiejętności 32), Kraków 1939. Hock, R.F., The Infancy Gospels of James and Thomas (The Scholars Bible 2), Santa Rosa, California 1995. Horn, C.B., From Model Virgin to Maternal Intercessor: Mary, Children, and Family Problems in Late Antique Infancy Gospel Traditions and Their Medieval Trajectories (forthcoming). Horn, C.B., Editing a Witness to Interactions between Christian Literature and the Qur’ān: Status Quaestionis and Relevance of the Arabic Apocryphal Gospel of John, in Parole de l’Orient 37 (2012), 87–103. Horn, C., Apocryphal Gospels in Arabic, or Some Complications on the Road to Traditions about Jesus, in J. Frey and J. Schröter, with the assistance of J. Spaeth (eds.), Jesus in apokryphen Evangelienüberlieferungen. Beiträge zu außerkanonischen Jesusüberlieferungen aus verschiedenen Sprach- und Kulturtraditionen (Wissenschaftliche Untersuchungen zum Neuen Testament 254), Tübingen 2010, 583–609. Horn, C.B., Syriac and Arabic Perspectives on Structural and Motif Parallels regarding Jesus’ Childhood in Christian Apocrypha and Early Islamic Literature: The ‘Book of Mary,’ the Arabic Apocryphal Gospel of John, and the Qur’ān, Apocrypha 19 (2008), 267–91. Horn, C.B., Mary between Bible and Qur’ān: Soundings into the Transmission and Reception History of the Protoevangelium of James on the Basis of Selected Literary Sources in Coptic and Copto-Arabic and of Art Historical Evidence Pertaining to Egypt, Islam and Christian—Muslim Relations 18.4 (2007), 509– 38. Horn, C.B., Intersections: The Reception History of the Protoevangelium of James in Sources from the Christian East and in the Qur’ān, in Apocrypha 17 (2006), 113–50. Horn, C.B., Frühsyrische Mariologie: Maria und ihre Schwestern im Werk Aphrahats des Persischen Weisen, in M. Tamcke and A. Heinz (eds.), Die Suryoye und ihre Umwelt. 4. deutsches SyrologenSymposium in Trier 2004. Festgabe Wolfgang Hage zum 70. Geburtstag (Studien zur Orientalischen Kirchengeschichte 36), Münster 2005, 313–32. Horn, C.B., and R.R. Phenix, Apocryphal Gospels in Syriac and

144

CORNELIA HORN

Related Texts Offering Traditions about Jesus, in J. Frey and J. Schröter, with the assistance of J. Spaeth (eds.), Jesus in apokryphen Evangelienüberlieferungen. Beiträge zu außerkanonischen Jesusüberlieferungen aus verschiedenen Sprach- und Kulturtraditionen (Wissenschaftliche Untersuchungen zum Neuen Testament 254), Tübingen 2010, 527–55. Horovitz, J., ‘Adi ibn Zayd, the Poet of Hira, in Islamic Culture 4 (1930), 31–69. Huart, C., Une nouvelle source du Qorān, in Journal Asiatique 10.4 (1904), 125–67. Huart, C., A History of Arabic Literature, New York, 1903. Kister, M.J., Some reports concerning al-Ṭā’if, Jerusalem Studies in Arabic and Islam 1 (1979), 1–18. Reprinted in M.J. Kister, Studies in Jāhiliyya and Early Islam (Variorum Collected Studies Series 123), Aldershot, Great Britain, and Burlington, Vermont 1980, reprinted 2002, ch. XI. al-Jubaylī, J.S., Dīwān Umayya ibn Abī al-Ṣalt, Beirut 1998. Milgrom, J., Case of the Suspected Adulteress, Numbers 5:11–31: Redaction and Meaning, in R. Friedman (ed.), in The Creation of Sacred Literature: Composition and Redaction of the Biblical Text (University of California Publications. Near Eastern Studies 22), Berkeley 1981, 69–75. Miller, D., Another look at the magical ritual for a suspected adulteress in Numbers 5:11–31, in Magic, Ritual, and Witchcraft 5.1 (2010), 1–16. Moraldi, L., Vangelo Arabo apocrifo dell’Apostolo Giovanni da un Manoscritto della Biblioteca Ambrosiana (Biblioteca di Cultura Medievale), Milan 1991. Mourad, S.A., On the Qur’ānic Stories about Mary and Jesus, in Bulletin of the Royal Institute for Inter-Faith Studies 1 (1999), 13–24. Nagel, T., Zur Einführung: Der Koran im spätantiken Vorderasien, in T. Nagel, unter Mitarbeit von E. Müller-Luckner (eds.), Der Koran und sein religiöses und kulturelles Umfeld (Schriften des Historischen Kollegs. Kolloquien 72) München 2010, vii-xxiv. Nicholson, R.A., A Literary History of the Arabs, Cambridge 1956, Power, E., The poems of Umayya b. Abi-s-Salt: additions, suggestions and rectifications, in Mélanges de la Faculté Orientale de

THE PROTOEVANGELIUM OF JAMES

145

l’Université Saint-Joseph 2, Beirut 1912, 145–95. Power, E., Umayya ibn Abī-ṣ-Ṣalt, Mélanges de la Faculté Orientale de l’Université Saint-Joseph 1, Beyrouth 1906, 197–222. Reynolds, G.S., The Qur’ān and Its Biblical Subtext (Routledge Studies in the Quran), London and New York 2010. Sayuti, N., The Concept of Allāh as the Highest God in pre-Islamic Arabia (A Study of Pre-Islamic Arabic Religious Poetry), M.A. thesis, The Institute for Islamic Studies, McGill University, Montreal 1999. Schultheß, F., Umajja ibn Abi ṣ Ṣalt. Die unter seinem Namen überlieferten Gedichtfragmente, Leipzig and Baltimore, 1911. Schultheß, F., Umajja b. Abi-ṣ Ṣalt, in C. Bezold (ed.), Orientalische Studien. Theodor Nöldeke zum siebzigsten Geburtstag (2. März 1906) gewidmet, 1, Gieszen 1906, 71–89. Seidensticker, T., The Authenticity of the Poems ascribed to Umayya Ibn Abī al-Ṣalt, in J.R. Smart (ed.), Tradition and Modernity in Arabic Language and Literature, Richmond 1996, 87– 101. Smith Lewis, A. (ed. and tr.), Apocrypha Syriaca. The Protevangelium Jacobi and Transitus Mariae with Texts from the Septuagint, the Corân, the Peshiṭta, and from a Syriac Hymn in a Syro-Arabic Palimpsest of the Fifth and Other Centuries (Studia Sinaitica XI), London 1902. Toepel, A., Das Protevangelium des Jakobus. Ein Beitrag zur neueren Diskussion um Herkunft, Auslegung und theologische Einordnung (Frankfurter Theologische Studien 71), Münster 2014. Toral-Niehoff, I., Eine arabische poetische Gestaltung des Sündenfalls. Das vorislamische Schöpfungsgedicht von ʿAdī b. Zayd, in D. Hartwig, W. Homolka, M.J. Marx and A. Neuwirth (eds.), “Im vollen Licht der Geschichte.” Die Wissenschaft des Judentums und die Anfänge der kritischen Koranforschung (Ex Oriente Lux. Rezeptionen und Exegesen als Traditionskritik 8), Würzburg 2008, 235–56. Vallaro, M., Umayyah ibn Abī ṣ-Ṣalt nella seconda parte del ‘Kitāb az-Zahrah’ di Ibn Dāwūd al-Iṣfahānī (manoscritto di Torino), Atti della Accademia Nazionale dei Lincei. Memorie. Classe di Scienze morali, storiche e filologiche. Serie 8. 22.4, Rome 1978, 423–80. Van Esbroeck, M., A propos de l’Évangile apocryphe arabe attri-

146

CORNELIA HORN

bué à Saint Jean, in Mélanges de l’Université Saint-Joseph 49 (Mélanges offerts au R. P. Henri Fleisch, S.J.) (1975–1976), 597–603. Von Nordheim, E., Das Gottesurteil als Schutzordal für die Frau nach Numeri 5, in R. Bartelmus, Th. Krüger and H. Utzschneider (eds.). Konsequente Traditionsgeschichte: Festschrift für Klaus Baltzer zum 65. Geburtstag (Orbis Biblicus et Orientalis 126), Fribourg and Göttingen 1993, 297–309. Yamut, B., Diwān Umayya ibn Abī al-Ṣalt, Beirut 1934. Zemler-Cizewski, W., Rupert of Deutz and the law of the stray wife: anti-Jewish allegory in the De sancta trinitate et operibus eius, in Recherches de Théologie et Philosophie Médiévales 75.2 (2008), 257–69.

TRADERS, INNKEEPERS AND CUPBEARERS: FOREIGNERS AND PEOPLE OF THE BOOK IN ARABIC WINE POETRY BRUNO PAOLI

It is not the slightest paradoxe of the medieval Mulim society, where the consumption of alcoholic drinks was officially prohibited, that to have allowed the growth of an autonomous wine poetry which knew its highlight between the end of the eighth century and the middle of the ninth, with poets such as Abū Nuwās (d. 814 or 815), Muslim b. al-Walīd (d. 823) or Ibn al-Ḍaḥḥāk (d. 864). The classical wine poem (khamriyya) contains a number of recurring themes, motifs and images, usually treated as part of stories featuring the poet and his companions, the wine merchant or the inn-keeper, the cup-bearer and the singer in a tavern, a monastery, a garden or a notable’s house. The story of a trip to the tavern (ḥāna, ḥānūt or bayt khammār) is undoubtedly one of the classics of the genre. It is the occasion for the poet to describe the innkeeper’s (khammār) colorful character. In this registry, Abū Nuwās was surely the most gifted of all wine poets. As Bencheikh says, “the portraits he gives us are among the strongest, the most colorful, the best of those that animate his poems” 1. Here is a very representative example 2:

1 2

Bencheikh, Poésies bachiques d’Abū Nuwās 53. Dīwān Abī Nuwās 61, lines 1–4.

147

148

BRUNO PAOLI

ِ ‫إِلى َب‬ ‫يت خَ ّما ٍر َن َزلنا بِ ِه ظُ ْه َرا‬

‫َظ َننّا بِ ِه خَ يراً َف َظ �ن بِنا شَ �را‬

‫ض ُم ْز َو �را َوقالَ َلنا ُه ْج َرا‬ َ ‫َفا�ع َر‬ ِ ‫المكنون ِمن ُه َل َك الخَ ْت َرا‬ ‫َو َيض ِم ُر في‬

ٍ ‫تيان ِص‬ ِ ِ‫َوف‬ ‫فت َم ِط �ي ُهم‬ ُ ‫دق قَد َص َر‬

ً‫يس ُمس ِلما‬ َ ‫َف َل ّما َحكى ال ُزنّا ُر ا ْ�ن َل‬

‫سيح بنِ َمر َي ٍم‬ ِ ‫َفقُلنا َعلى دينِ ال َم‬ ً‫َو َل ِك ْن َيهو ِد �ي يُ ِح �ب َك ظا ِهرا‬

Of how many faithful companions did I divert the camel Toward an innkeeper’s house, on the stroke of noon. His belt spoke for him: he was not a Muslim. We thought the best of him and him, of us, evil. We asked him if his religion was that of the Messiah, son of Mary. He looked askance at us and talked to us sharply. Because he was Jewish, of these people who seem good, But who, in secret, only aims at betraying you.

The innkeeper is Jewish (yahūdī). But it could just as well have been Christian (naṣrānī), as evidenced by the hesitation of the poet, who believes, first, that his religion is that of “the Messiah, son of Mary”. The fact that he wears a belt (zunnār) is the sign that he is not a Muslim. Indeed, this belt was at the time, the hallmark of dhimmī, whether Jews, Christians and Zoroastrians, but we do not know the exact date when this rule was established. According to Tritton, the tradition which dates it back to the Caliph ʿUmar b. alKhaṭṭāb (634–644) is of doubtful authenticity, so that it should be situated later, probably during the reign of ʿ Umar b. ʿAbd al-ʿAzīz (718–720) 3. More recently, Levy-Rubin aimed at proving that the normalized version of the ‘Covenant of ʿUmar’, although it cannot be dated prior to al-Mutawwakil, is the offshot of a number of earlier regionally diversified attempts that went back to the treaties of surrender signed during the Islamic conquests, and were probably inspired by comparable Byzantine and Sasanian rules, restrictions and habits. 4 This question I will discuss later. Tritton, Zunnār 1313. Levy-Rubin, Non-Muslims with special attention to ch. 3 which is devoted to the restrictions regarding the public dress, appearance and 3 4

TRADERS, INNKEEPERS AND CUP-BEARERS

149

Elsewhere, a woman holds the cabaret, as this Umm Ḥuṣayn who knows how to relieve the poet… of his savings 5:

ِ ‫ض‬ ٍ ِ‫ب‬ ِ‫حاح َو َعين‬ ِ ‫مال ِم َن البي‬ ِ ‫الص‬

‫ِبت إِلى قُط َرب� ٍل َفا�تَي ُتها‬ ُ ‫َطر‬

‫َمانين ديناراً ِجياداً َذخَ رتُها‬ َ ‫ث‬

ِ‫ِبت بِدَين‬ ُ ‫َفا�نفَق ُتها َحتّى شَ ر‬ ِ‫عت رِدا ًء ُمع َل َم ال َط َرفَين‬ ُ ِ‫َو ب‬ ِ‫ُم َه �ذ َب ٍة تُكنى بِا� �م ُح َصين‬

ِ‫فَ�� بُ �د ِمن تَقبي ِل َي الشَ َف َتين‬

ِ‫بِا�م َر َد كَالدينا ِر فاتِ َر َعين‬

ِ‫َضيض ر ِاج ُح ال َك َف َلين‬ ٌ ‫ا�غ �َن غ‬ ِ‫س ِمن ِم َئ َتين‬ ِ ��‫س في ِٕا��ف‬ ُ ‫ا� َق ْر ِط‬

‫عت قَميصاً سابِ ِريّاً َو ُج �ب ًة‬ ُ ِ‫َو ب‬ ‫دين ابنِ ِعمر َان دي ُنها‬ ُ ‫لِخَ ّما َر ٍة‬

ٍ‫ُلت َلها إِن َلم تَجودي بِنائِل‬ ُ ‫َوق‬ ‫وى‬ ً ‫فَقا َلت ف ََهل تَرضى بِغَي ِر ِهما َه‬

‫فَجا َءت بِ ِه كَال َبد ِر َيش ُرقُ َوج ُه ُه‬ ِ ‫عسراً غَي َر ُم‬ ِ ‫حت َعنها ُم‬ ‫وس ٍر‬ ُ ‫َف َر �و‬

I languished for Quṭrabbul 6, so I went there, With a nest-egg of gold and silver.

Eighty fine dinars that I first saved, Then spent until the last, to finally drink on credit. I sold my shirt and my tunic, I sold my coat embroidered edge to edge,

To an inn-keeper whose religion was that of Ibn ʿImran, A pure and sincere [woman] called Umm Ḥuṣayn. I told her: “If you’re not generous with me, Let me at least kiss your lips!”

behavior of non-Muslims. According to the author, it is the result of the adoption of Iranian dress codes. 5 Dīwān Abī Nuwās 86, lines 1–8. 6 According to Yāqūt al-Ḥamawī, Quṭrabbul “is a village between Baghdad and ʿUkbarā whose wine holds the name. It is a place where the idle go for a walk and where we find inns and wine shops. It is often evoked by the poets,” Kitāb al-buldān iv, 133.

150

BRUNO PAOLI “Do you not prefer, she said, the passion of a youngster, Smooth as a new penny, with languid eyes?”

And she brought him to me, his face shining like the moon, His foreign accent, his freshness, his ass that waddles. I dashed away without a penny in my pocket Poorest of two hundred dinars.

Umm Ḥuṣayn is also Jewish: her religion, says the poet, is that of the son of ʿImran, which is none other than Moses. Other taverns were kept by Christians, as the so-called Bishra, wife of a priest from Tall ʿAzāz, near Raqqa, in Syria 7, or that other, anonymous, mentioned by Ibn al-Muʿtazz (d. 908) in this line 8:

ِ‫بح غَي ُر ُمبين‬ ُ ‫َط َر‬ ُ ‫قت َو َضو ُء‬ ِ ‫الص‬

‫سيح بِدي ِنها‬ ُ ‫َوخَ ّما َر ٍة يُعنى ال َم‬

A [female] wine seller whose religion is that of the Messiah, Whom I visited before the [first] light of dawn.

Others were kept by Zoroastrians, like this “daughter of a mage” (min banāt al-majūs), recalled by the same Ibn al-Muʿtazz 9:

��‫تَرى ال �د �ن في َبي ِتها شائ‬ ��‫فَكا َل ْت َلنا َذ َهباً سائ‬

ِ ‫وخَ ّمار ٍة ِمن َب‬ ‫س‬ ِ ‫نات ال َمجو‬ ِ ً‫َو َزنّا َلها َذ َهبا‬ ً‫جامدا‬

A wine merchant, one of those daughters of a mage, In the house of whom you can see the jars elevated. We weighed for her solid gold, So that she measures for us liquid gold 10.

The ethnic or religious identity of wine sellers is not always specified. But when it is, and there are many examples, they always are non-Muslims, Jews, Christians and Zoroastrians, whose name al-Sirrī, Al-muḥibb wa-l-maḥbūb iv, 343–5. Shiʿr Ibn al-Muʿtazz ii, 248 (n°814, line). 9 Line quoted by al-Sirrī, Al-muḥibb wa-l-maḥbūb iv, 151–2. 10 “Solid gold” refers to golden coins, whereas “liquid gold” refers to a golden wine, a wine with yellow gleams (ṣahbāʾ). 7 8

TRADERS, INNKEEPERS AND CUP-BEARERS

151

sometimes betrays their non-Arab origin, as Sarjīs (Serge) of Ṭīzanābād, or Anāhīd and Khīq from Sijistān 11. In many cases, however, they adopt an Arabic nickname. Thus, the first innkeeper quoted replied that his name was Samawʾal (Samuel), a Jewish name if any, but that he was commonly called Abū ʿAmr, which is an Arabic kunya 12 :

‫َعلى ا�ن�ني ا�كنى بِ َعم ٍر َو�� َعمرا‬

‫فقُلنا َل ُه ما ِٕا��س ُم قالَ َس َموا� ٌل‬

‫�كس َبتني �� َسنا ًء َو�� فَخرا‬ َ ‫َو�� ا‬

‫َوما شَ �رفَتني كُن َي ٌة َع َر بِ �ي ٌة‬ ‫َو َل ِكن�ها خَ ف�ت َو َقل�ت ُحروفُها‬ ِ ‫َفقُلنا َل ُه ُعجباً بِ َظ‬ ‫رف لِسانِ ِه‬

‫َو َل َيست َكا�خرى إِن�ما خُ ِلقَت َوقرا‬ ‫دت ا�با َعم ٍر ف ََج �و ْد َلنا الخَ مرا‬ َ ‫ا َ�ج‬

We asked: “What is your name? “; he said: “Samuel” “But they call me Abū ʿAmr; it’s up to you.”

Although an Arabic nickname is a poor honour, As it does not guarantee the nobility nor gives the right to boast, We had to admit its lightness and its brevity, Not like others that are full of heaviness.

But to our amazement, he spoke a pleasant language: “Well spoken, Abū ʿAmr! Now, serve us some wine!”

The fact that Abū Nuwās recognizes Abū ʿAmr’s merit to speak a pleasant language certainly means that Arabic is not his native tongue. In admitting his surprise, he recognized that it is more common to deal with innkeepers speaking Arabic with a foreign accent. In the second poem quoted, the boy that Umm Ḥuṣayn leads to the poet is said to be aghannu, name derived from the verb ghanna which literally means “speaking through the nose, with a nasal voice”, and according to Arab lexicographers, it applies in particular to the Persians’ nazalisation of vowels followed by a nūn. al-Sirrī, Al-muḥibb wa-l-maḥbūb iv, 333 (Khīq); 335 (Anāhīd); and 343–5 (Bishra). 12 Dīwān Abī Nuwās 61, lines 5–8. 11

152

BRUNO PAOLI

The same word, as we shall see shortly, was employed by one preIslamic poet at least, al-Aswad b. Yaʿfur (d. ca. 590), to describe the innkeeper’s accent. In the monograph that she devoted to the wine poetry of Abū Nuwās, Mary Bonnaud rightly stressed that the medieval Muslim authors had little interest in the technical aspects of the making and preservation of wine, leaving it “to those who had the right of drinking and selling it, that is Jews and Christians” 13. “The wines that were drank in Baghdad [in the time of Abū Nuwās], she says, were made by Jews and Christians” 14 or came from Persia. Right though these statements are, they may give the impression that the Islamic ban suffices to explain why the wine occupations, during the first centuries of Islam, were generally held by non-Muslims. We will see that it is not the case, because the religious factor only overlapped an older cultural factor whose importance was not denied after the advent of Islam; because, as Mary Bonnaud says, “all these dhimmī who had the right to drink and trade wine, had inherited Greek, Iranian or Mesopotamian processes”. 15 The reading of wine poetry attributed to ancient poets will show that it was the tradition, before Islam as well, that wine merchants, innkeepers and cup-bearers be non Arabs (ʿajam), as evidenced by this first example from a poem by Muraqqish al-Aṣghar (d. ca. 570) 16:

‫تُ َعل�ى على الن�ا ُجو ِد َط ْوراً وتُ ْقد َُح‬ ‫يُطا ُن عليها َق ْر َم ٌد وتُ َر �و ُح‬ ِ ‫الس‬ ‫وق ُم ْر بِ ُح‬ � ‫يُدْنيها من‬

‫من الل� ْي ِل َب ْل فُوها ا� َل �ذ وا�ن َْص ُح‬ َ

‫وما ق َْه َو ٌة َص ْهبا ُء كال ِم ْس ِك ري ُحها‬ ‫ثَ َو ْت في ِسبا ِء ال �د �ن ِعشْ ر َِين ِح �ج ًة‬

‫ي��ن‬ َ ‫َسباها رِجا ٌل من َي ُهو َد تَبا َع ُدوا لِ ِج‬ ً‫با� ْط َي َب ِم ْن فيها إذا جئ َْت طارِقا‬

There is no golden yellow wine, the scent of musk, Poured into a large cup and then ladled from cup to cup, Bonnaud, La poésie bachique d’Abû Nuwâs 101. Bonnaud, La poésie bachique d’Abû Nuwâs 101. 15 Bonnaud, La poésie bachique d’Abû Nuwâs 101–2. 16 Dīwān al-Muraqqishayn 88–89 (n°1, lines 8–11). 13 14

TRADERS, INNKEEPERS AND CUP-BEARERS

153

A captive who dwelt for twenty years in a jar Sealed with a coating and well ventilated,

And that some Jewish people bought to make profit, Away from the Gilan and approaching the market, hoping to good benefit.

[No wine] better than her mouth, to whom visits her at night. Yes, her mouth, at night, is even more delicious and pure.

The figure of the Jewish merchant is also mentioned in these four lines by ʿAdī b. Zayd (d. ca. 600) 17:

َ ‫ِف ت‬ ‫يق‬ ٌ ‫ُريك الق ََذى ُك َم‬ ُ ‫يت َر ِح‬ ‫يق‬ ُ ‫َفا� ْذكَى ِم ْن نَشْ رِها الت� ْع ِت‬

ُ‫وحانت من اليهو ِد �ي ُسوق‬ ‫�ن‬ ْ ‫أ ْر َي ِح �ي غ َم ْن َد ٌر ِغ ْرني ٌق‬

‫با َك َر ْت ُه �ن قَرق َُف َك َد ِم ال ُجو‬

ِ �‫صانَها الت‬ ِ‫اج ُر ال َي ُهو ِد �ي َح ْو َل ْيـ ـن‬ ِ ‫ُض‬ ِ ‫الختا ُم عن‬ ‫حاج ِب الد‬ � ‫ث ّم ف‬ ‫فاستباها أشَ �م ِخ ْرقٌ كري ٌم‬

[The waitresses] were eager to serve to us a fresh wine, [red] As the blood of the womb, the deposits visible, tan-colored nectar. [A wine] kept by the Jewish merchant for two years, Whose ageing has made his scent even more penetrating.

Then the stamp that sealed the amphora was broken, Sign that the time had come for the Jew to go to the market,

So that the most proud and generous man buys it, Eager to give, young man with a gentle face and complexion of milk. 18

According to these lines, the Jewish trader (tājir) purchases wine, probably wholesale, to let it age for years in jars, and then resell it at ʿAdī b. Zayd 77 (n°13, lines 9–12). The meaning of ghamandar, which does not appear in any dictionary, remains uncertain. One should probably read ghamaydar (“young, tender”), as suggested by the editor of the poet’s dīwān. This is the choice I did. 17 Dīwān 18

154

BRUNO PAOLI

retail in making the best possible price. Al-Akhṭal (d. 710) admits that he has paid a high price for this wine 19:

‫َفا�ط َل َعها إِلى ال َع َر ِب ال ِتجا ُر‬ ‫كون َلها َيسا ُر‬ َ ‫َتا�ب�ى ا�و َي‬

‫َتوا َعدَها ال ِتجا ُر إِلى إِناها‬ ‫َفا�ع َطينا الغَ�� َء بِها َوكانَت‬

The merchants made an appointment with each other when it came At maturity, and presented it to their Arab [customers]. We [were willing to] pay a high price for this wine, and yet He refused itself to us because they asked for an exorbitant price.

In these two lines, the Arabs are in the role of customers, and good customers, if we are to believe al-Akhṭal, who is ready to pay a high price for a good wine. As for the merchants who sell their wine “to the Arabs”, it is implicitly admitted that, by deduction, they are not Arabs. The innkeepers were also, often, Jews. In Medina itself, a member of the tribe of Banū Qurayza, one of the Jewish tribes of the city, was said to be the owner of a tavern. Salām b. Mishkam, who was his neighbor, once hosted the Meccan Abū Sufyān, an opponent of the Prophet who lately converted to Islam. He kept him for three days and three nights, not allowing him to take leave before the reserves of the innkeeper had been emptied 20. The preislamic poets have not failed to mention the figure of the innkeeper, as in this passage of a poem by al-Aswad b. Yaʿfur 21:

‫زجت بما ِء َغوادي‬ َ ‫بس�� َف ٍة ُم‬ ُ ِ‫وافى بها لدرا ِهم ا��ٔسجاد‬ ‫من الفُرصا ِد‬ َ ‫ق َنا�ت أناملُ ُه‬ ‫مشين بِا���رفا ِد‬ َ ‫ونواع ٌم َي‬

‫هوت وللشباب لذاذ ٌة‬ ُ ‫ولقد َل‬ ‫من خَ مر ذي نَطف أغ �َن ُمنطق‬ ‫َيسعى بها ذو تُومتين ُمشَ �م ٌر‬ ‫البيض تَمشي كالبدُو ِر وكالد�مى‬ ُ ‫و‬

Dīwān al-Akhṭal 110 (n°73, lines 3–4). al-Sirrī, Al-muḥibb wa-l-maḥbūb iv, 329. 21 al-Mufaḍḍal al-Ḍabbī, Al-mufaḍḍaliyyāt 218 (n°44, l. 22–5). 19 20

TRADERS, INNKEEPERS AND CUP-BEARERS

155

Yea, once I enjoyed the sweetest flavour of youth, and delighted myself Of a clear wine, mingled with the water of the clouds,

Sold to foreigners in exchange for drachmas,

By an innkeeper with a twang in his speech, wearing earrings and belt.

And dealt [to the clients] by a cup-bearer who wears two earrings, His sleeves rolled up and his fingertips as stained with mulberry juice, While [singers] with clear faces, like the full moon Or like statues, soft to the touch, carried cups around.

In these verses, the innkeeper (khammār) and the cup-bearer (sāqin) are not named but referred to by epithets both built with the relative (dhū/dhī): both wear earrings (nuṭafa or naṭafa, pl. nuṭaf; and tawma, in the dual form here). The former, the poet observes, has a strong accent. He uses the same word (aghannu) which Abū Nuwās used to describe striplings. This suggests that al-Aswad’s innkeeper is not an Arab, and his dress, earrings and belt (niṭāq or minṭaq, whence munaṭṭaq, “girt”) tends to confirm it. According to Tritton, the minṭaq must be distinguished from the zunnār, which is thicker 22. The term zunnār is employed by Abū Nuwās. But it is not attested in older poetry. On the contrary, it seems that the minṭaq, before Islam, was mainly worn by nonArabs, as evidenced by this example from the dīwān of Imruʾ alQays (first half of the sixth century) in which he used, as in that of al-Aswad b. Yaʿfur quoted above, the past participle munaṭṭaq 23:

ِ‫ِقيا َم ال َعزي ِز الفار ِِس �ي الم َن �طق‬

ِ ‫ص إِذ َي‬ ‫خضبو َن ُه‬ ِ ‫َوقا َم ِطوالَ الشَ خ‬

And he stood up to full height, while being dyed, As stands up the Persian governor, his waist girt. 22 23

Tritton, Zunnār 1312. Dīwān Imri’ al-Qays 175 (n°30, line 31).

156

BRUNO PAOLI

In this verse, which is part of a horse description, it is a Persian with his waist girt (al-fārisī al-munaṭṭaq) who serves as a comparison. It is later employed by Harītha b. Badr (d. 684) in a bacchic verse 24 :

‫كف الوصيف المنطّق‬ ّ ‫تخايل في‬

‫تراها إذا ما الماءُ خالط جسمها‬

[A wine that] you see, when water penetrates its body, Swinging proudly in the hand of the servant girt at the waist.

Other terms are occasionally used to describe the belt. As for an example, Ḥassān b. Thābit (d. 660) uses the word ḥizām 25:

ِ ‫ُمح َتل َُق ال ِذفرى شَ دي ُد‬ ‫الحزام‬

‫س‬ ٍ ُ‫يسعى بِها ا�ح َم ُر ذو بُرن‬

Served by [a cup-bearer] with a white skin, wearing a burnous, The temples razed and a robust waist under his belt.

The subject of this line (aḥmar) should not be understood in its usual modern sense (red). In the classical Arabic language, it could be used to designate men with a white and fair skin, as opposed to aswad (black), which means those whose skin is mat and dark. Hence, these two terms were used metaphorically to distinguish between the dark-skinned (sumra) Arabs (ʿarab) and the paleskinned (bayāḍ) non-Arabs (ʿajam) The figure of the cup-bearer is often introduced by the formula yasʿā bi-hā (“Served by...”) and generally follows a description of wine. We saw an example of this formula in the third verse of the passage of al-Aswad b. Yaʿfur cited above, where the wine is served by a cup-bearer wearing two earrings (dhū tawmatayni). The same formula is used by al-Musayyab (died ca. 570), whose cup-bearer has a single earring (dhū tawmatin) 26:

‫َيسعى بِها ذو تو َم ٍة َل ِبق‬

‫رف ُم َعت� َق ٌة‬ ٌ ‫عانِ �ي ٌة ِص‬

Ḥammūdī al-Qaysī, Ḥāritha b. Badr al-Ghudānī 168 (n°33, line 3). Sharḥ dīwān Ḥassān b. Thābit 381, line 9. 26 Geyer, Kitāb al-ṣubḥ al-munīr 356 (n°14, line 10). 24 25

TRADERS, INNKEEPERS AND CUP-BEARERS

157

A wine from ʿĀna, a pure wine that has been allowed to age, Served by [a cup-bearer] who wears an earring, elegant.

It is found again in this line of al-Aʿshā, who employs an epithet to refer to the cup-bearer (dhū zujājātin), who, once again, wears earrings (la-hū nuṭafun) 27:

ِ ‫ص ا�س َف َل ال ِس‬ ‫ربال ُمع َت ِم ُل‬ ٌ �‫ُم َقل‬

ٍ ‫ف‬ ‫َيسعى بِها ذو ُز‬ ٌ ‫جاجات َل ُه نُ َط‬

Served by [a cup-bearer] who wears earrings, The bottom of his shirt rolled up, requested from all sides.

The poets of the Abbasid era continued to use the same formula. Abū Nuwās, for example, employed it no less than seven times. I only quote the following line, without earrings but with the expression dhū ghunnatin which is the equivalent of aghannu, employed before Islam by al-Aswad b. Yaʿfur, and later by Abū Nuwās himself about the Jewish innkeeper, to describe the strange accent of the cup-bearer 28:

ِ ِ‫ظات ب‬ ِ ‫ُم َت َك �ح ُل الل ََح‬ ‫السح ِر‬

‫َيسعى بِها ذو ُغن� ٍة َغنِ ٌج‬

Served by [a cup-bearer] with a nasal voice, who darts glances, His eyes shadowed with kohl, as if by magic.

Even before Islam, merchants, innkeepers and cup-bearers were, by tradition, non-Arabs, mostly Jewish, and this tradition perpetuated itself during the first centuries of Islam, and this has nothing to do with the precepts of the new religion and the prohibition of alcoholic beverages. Indeed, whether non-Arabs or followers of a religion other than Islam, or often both, they are almost always described in the sixth century as well as in the ninth, with the same three basic attributes, that is their foreign accent (a nasal voice, aghannu or dhū ghunnatin), their earrings (tawmatayn or nuṭaf) and their belt (zunnār, minṭaq or ḥizām). In his presentation of the edict of ʿUmar b. ʿAbd al-Azīz, Ibn ʿAbd al-Ḥakam (d. 871) uses minṭaq, rather than zunnār, to desig27 28

Dīwān al-Aʿshā al-Kabīr 59 (n°6, line 41). Dīwān Abī Nuwās 99, line 3; and 76, line 8.

158

BRUNO PAOLI

nate the belt of the dhimmī 29. As for Abū Yusūf (d. 798), he does not make use of zunnār except for the edict of ʿUmar b. al-Khaṭṭab, and he employs the plural zunnārāt, rather than zanānīr which will become afterwards the most common form. But in his presentation of the edict of ʿUmar b. ʿ Abd al-ʿAzīz, he only makes use of minṭaq 30. Despite these fluctuations, it would seem that the term zunnār, of Greek origin and probably passed into Arabic through Aramaic, is gradually entered into the official use and progressively replaced the Arabic minṭaq which was used before Islam to refer to the belt worn by non-Arabs, and which continued to be used, particularly by poets. Despite the distinction sometimes drawn between zunnār and minṭaq, it is likely that the official dress code which applied to Muslim-dominated dhimmī has endorsed a preexisting fashion. As Tritton says, “at the time of the conquest, it was not necessary to require Christians to dress differently from the Arabs because they already did so 31“. And it is only later, when non-Muslims could be tempted to dress as Muslims, and at a time when the dhimmī were often suspected of complicity with the enemy, either Byzantine, Crusader or Persian, that the authorities felt the need to legislate. This legislation had as its main objective, as the Arab authors say, to prevent dhimmī to dress as Muslims, and particularly to wear the turban (ʿimāma) or other head coverings (qalansuwa, ʿiṣb), as well as the taylasān, a veil worn over the turban, and the kubā, a short jacket of Persian origin. They were also prohibited, at the time of ʿUmar b. ʿAbd al-ʿAzīz, to dress as Muslim soldiers and to saddle their mounts other than wooden saddles. Finally, they had to wear a belt (zunnār or minṭaq) and to have their hair cut short on the sides. This brings us back to the line of Ḥassān b. Thābit previously cited, in which the cup-bearer is designated by an adjective (aḥmar) whose meaning is “one who has white skin”. He wears a burnous and a belt, and the projecting parts of his skull, above the ears, are See Tritton, The Caliphs and their Non-Muslim Subjects 117, who refers to Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr. 30 Tritton, The Caliphs and their Non-Muslim Subjects 117, after Abū Yūsuf Yaʿqūb, Kitāb al-kharaj. 31 Tritton, The Caliphs and their Non-Muslim Subjects 115. 29

TRADERS, INNKEEPERS AND CUP-BEARERS

159

shaved. He has the significant attributes of the perfect dhimmī. Ḥassān b. Thābit died shortly after the caliph ʿUthmān (m. 656), and this verse raises the question of whether the dress code applicable to dhimmī is prior to ʿUmar b. ʿAbd al-ʿAzīz or not and, therefore, whether we can bring credence to information that date its origin back to the time of the caliph ʿUmar b. al-Khaṭṭāb (r. 634– 644). It is said that Ḥassān lived two diametrically different existences: court poet before Islam, and as such going in for wine poetry, he then became the champion of the Prophet and renounced the pleasures of libertinage and light poetry of his first life to enter the service of the emerging religion. According to this “hagiographic” version of his life, the line quoted above, like the many wine lines that he composed, should date from its pre-Islamic period, its first life. But there must be far from this biography to reality: how can one explain, otherwise, the insertion of a wine sequence in a poem of praise dedicated to the Prophet 32 Although Ḥassān swears God that he had composed the poem before Islam, it is difficult to believe. Unless if the poet, being out of inspiration, had reused old materials. But the little poem which the line quoted belongs to seems rather old 33. It consists of twenty-one lines divided into three sequences: a nasīb (“love prologue”) of fifteen lines, a short raḥīl (“crossing the desert on a camel’s back”) of three lines and, finally, three lines of tribal fakhr (praise of the tribe). The wine section, which occupies no less than nine lines, is included in the nasīb, inside a description of a female djinn whose appearance held the poet awake all night. The taste of the mouth of this “djinnette” is compared to that of a wine melted with pure water, the water of a stream in this particular case. The process involving the insertion of a wine development inside a description of the mouth of the beloved is ancient, already attested in the poetry of many preIslamic poets 34. This poem was therefore in all probability, composed before Islam, during the first life of Ḥassān, and it provides, therefore, an additional proof that the dress code imposed on the dhimmī by the Muslim authorities did endorse preexisting practices, Sharḥ dīwān Ḥassān b. Thābit 3. Sharḥ dīwān Ḥassān b. Thābit 380–2. 34 See Paoli, Mémoire sur la poésie bachique arabe 31–4. 32 33

160

BRUNO PAOLI

with the main objective to prohibit non-Muslims to dress as Muslims. 35 This short presentation of wine professionals in the Arab and Muslim world has left many gray areas and remaining questions: a further study should examine all the attributes, clothing or other, of these characters as well as those of their Arab clients, and a moment away from the taverns, to make a tour of Christian monasteries, another residence of choice for wine lovers. But it might have helped to show the interest, for cultural, social and economic history of the Arabs and people of the Middle East, of literary sources that are too often neglected by specialists. The wine professionals, generally non-Arab (ʿajam), often Jews or Christians, sometime Zoroastrians, dhimmī after the advent of Islam, are an interesting sample of non-Arab and/or non-Muslim community, of which only a broader perspective can tell us if it can be considered representative or, conversely, marginal or atypical. The observations made here speak for the first hypothesis. The preserved Arabic wine lines and poems actually contain a lot of very valuable information about ethnic, religious and social groups and their relationships in the pre-Islamic and Muslim Middle East, but also about wine culture, all information that is highly valuable for archaeological and historical research, if one attaches to it the importance it deserves.

BIBLIOGRAPHY 1. Poetical works Dīwān Abī Nuwās, ed. A.ʿA. al-Ghazzālī, Cairo 1953. Dīwān ʿAdī b. Zayd al-ʿIbādī, ed. M.J. al-Muʿība, Baghdad 1965. Dīwān al-Akhṭal, ed. K. Ṣādir, Beyrouth 1999. Dīwān al-Aʿshā al-Kabīr, ed. M. Ḥusayn, n.p., n.d. One might argue that the genuineness of many of Ḥassān’s poems is considered doubtful. But in most cases, the question just cannot be resolved, so that I personally follow Blachère, Histoire who considers that the important point is wether they can be considered representative of the ancient Arabic poetical tradition or not. And as far as Ḥassān’s given poem is concerned, the answer is positive. 35

TRADERS, INNKEEPERS AND CUP-BEARERS

161

Dīwān Imriʾ al-Qays, ed. M.A. Ibrāhīm, Cairo 1958. Dīwān al-Muraqqishayn, ed. K. Ṣādir, Beirut 1998. Geyer, R., Kitāb al-ṣubḥ al-munīr, London 1928. Ḥammūdī al-Qaysī, N., Ḥāritha b. Badr al-Ghudānī, in Mağallat almağmaʿ al-ʿilmī al-ʿirāqī 25 (1974), 142–85. al-Mufaḍḍal al-Ḍabbī, al-Mufaḍḍaliyyāt, ed. A.M. Shākir and ʿA.M. Hārūn, Cairo 1976. Sharḥ dīwān Ḥassān b. Thābit, ed. ʿA. al-Barqūqī, Cairo 1929. Shiʿr Ibn al-Muʿtazz, ed. Y.A. al-Samarrāʾī, Baghdad 1978. al-Sirrī b. Aḥmad al-Raffāʾ, Al-Muḥibb wa-l-maḥbūb wa-l-mashmūm wal-mashrūb, ed. M.Ḥ. Al-Dhahabī, Damascus 1987. 2. Other sources Abū Yūsuf Yaʿqūb, Kitāb al-kharaj, in Fī al-turāth al-iqtiṣādī al-islāmī, Beirut 1990, 96–372. Bencheikh, J.E., Poésies bachiques d’Abū Nuwās. Thèmes et personnages, in Bulletin d’études orientales 18 (1963–1964), 1–54. Blachère R., Histoire de la littérature arabe, vol. 1, Paris 1952. Bonnaud, M., La poésie bachique d’Abû Nuwâs, Bordeaux 2008. Ibn ʿAbd al-Ḥakam, Futūḥ Miṣr, ed. C. C. Torrey, The History of the Conquest of Egypt, North Africa and Spain known as Futūḥ Miṣr of Ibn ʿAbd al-Ḥakam, New Haven 1922. Levy-Rubin, M., Non-Muslims in the Early Islamic Empire: From Surrender to Coexistence, Cambridge 2011. Paoli, B., Mémoire sur la poésie bachique arabe et sur le rôle des poètes préislamiques d’al-Ḥīra et de sa région dans la formation du genre, http://cle.ens-lyon.fr/arabe/genese-de-la-poesie-bachiquearabe48365.kjsp?STNAV=&RUBNAV=&RH=CDL_ARA110300. Tritton, A. S., Zunnār, in EI1, iv, 1312–3. Tritton, A.S., The Caliphs and their Non-Muslim Subjects, London 1930. Yāqūt al-Ḥamawī, Kitāb al-buldān, ed. F.ʿA. al-Jundī, Beirut, n.d.

III. RELIGIOUS IDENTITIES IN THE CONTEXT OF POLITICAL AND SOCIAL HISTORY

163

FROM MAVIA TO AL-MUNDHIR: ARAB CHRISTIANS AND ARAB TRIBES IN THE LATE ANTIQUE ROMAN EAST 1 GREG FISHER INTRODUCTION

In 2009, construction crews improving road access to Queen Alia International Airport uncovered a church complex at Tall alʿUmayrī East, in the suburbs of Amman. A rescue excavation revealed a substantial basilica-style structure containing a Greek inscription, found in front of the apse, which demonstrates the connection of the church to the cult of St. Sergius, and calls for the protection of ‘the great majesty’ Alamoundaros. 2 The chance discovery of the church provides further evidence for the link between the family of Alamoundaros, the Jafnids, the leading dynasty of Ghassān in the sixth century, and the rural churches and monasteries of Syria and Arabia. As supporters of these mostly monophysite communities, the Jafnids became connected to the decision by Research for this paper was supported by the Social Sciences and Humanities Research Council of Canada’s Standard Research Grants program. The support of the Council is gratefully acknowledged. I am also very grateful to Elizabeth Key Fowden and Fergus Millar for reading earlier drafts of this paper and offering critical comments and suggestions, and to both, and Maurice Sartre and Pierre-Louis Gatier, for kindly sharing work in progress. 2 Initial publication: al-Shami, Archaeological excavation project. For the inscription: Bevan, Fisher, and Genequand, Late antique church, and Gatier, Les Jafnides 193–5. 1

165

166

GREG FISHER

the imperial authorities to provide new bishops to the monophysites following the persecutions of Justin I, and worked to mediate ecclesiastical disputes on behalf of successive emperors in Constantinople. 3 By doing so, the Jafnids deftly exploited the political opportunities which were packaged together with becoming Christian, and which had been utilized by their predecessors in the fourth and fifth centuries. The narratives of the ways that Arabs were Christianized along the periphery of the Empire, and the careers of individual Arabs, such as Mavia, Aspebetos, and Amorkesos, reveal the role played by Christianity in establishing alliances, and repairing broken treaties. But by providing new opportunities, and new hierarchies, Christianization threatened to reorder the world of new converts. To some extent, the Jafnids were subjected to these pressures, as they became subsumed into the highly-politicized world of late Roman religious affairs. The other face of Jafnid Christianity, however, was an expansion of the traditional functions of tribal leadership to the imperial arena—particularly mediation—and a predominantly rural focus for their public expressions of Christian piety which avoided an overt association with the mostly urban Chalcedonian hierarchy. Becoming Christian, but staying tribal, the Jafnids provided powerful models of Arab élites on the eve of the Muslim invasions, openly connect to a vigorous monotheism which linked the rural communities of Syria and Arabia, and the Arabs, and others, who lived within them.

BECOMING CHRISTIAN

The victory of Constantine ensured the primacy of a powerful universal monotheism in Roman politics, and gave the Roman state a new form of religious identity. 4 From Nicaea to Chalcedon and The term is used here (as with Jafnid) for the sake of familiarity and convenience, although it must be noted that the monophysites never referred to themselves as such. See Millar, The evolution of the Syrian Orthodox Church 43–92, who prefers the term ‘the orthodox’, used by the monophysites to describe themselves. Millar offers a comprehensive examination of the terminological issues and urges caution over the use of the term ‘miaphysite.’ 4 The emergence of a Christian commonwealth is still best expressed 3

FROM MAVIA TO AL-MUNDHIR

167

afterwards, defining Christian orthodoxy now became the responsibility of the Emperor, as divisions between different communities—Nestorians, Chalcedonians, Donatists, monophysites, and others—threatened efforts at imperial stability and unity. 5 From the perspective of interstate politics, the Romans looked east to a Sasanian Empire which now possessed a more sharply-defined religious posture than its Parthian predecessor, even while it tolerated expressions of religious difference. The Sasanians placed Zoroastrianism at the centre of a multi-religious community which also included Jews, Christians, Buddhists, and others, forming a counterpoint to the Roman Christian state. 6 Over time, though, the Roman Empire, driven by an increasing focus on state-sponsored monotheism, came to infuse a greater emphasis on universal religion in its political and military stance towards its neighbour. If the fifth century witnessed only minor confrontations between the eastern Empire and Sasanian Persia, even these, especially that of 421/2, contained prominent religious subcurrents. 7 Between the fifth and the seventh centuries, competition between the two states for religious influence as far afield as Armenia, Axum, the Caucasus, and Ḥimyar, as well as the ongoing Roman concerns over the status of Christians in the Sasanian Empire, ensured that Roman political decisions, sometimes more so than Sasanian ones, were ever more influenced by religious concerns. 8 by Fowden, Empire to Commonwealth; see too Fowden, Varieties of religious community; Drake, The impact of Constantine. 5 Allen, Definition and enforcement, and van Rompay, Society and community. 6 On the complexity of Sasanian religious identity, see Shaked, Religion in the late Sasanian period; Daryaee, Sasanian Persia 69–99. See too Walker, Legend of Mar Qardagh, and Walker, From Nisibis to Xiʾan. 7 War of 421: Blockley, East Roman Foreign Policy 52–9; Rubin, Diplomacy and war; Decret, Les conséquences sur le Christianisme; on interstate relations, Haarer, Anastasius I; Dignas and Winter, Rome and Persia; Canepa, The Two Eyes of the Earth. 8 Greatrex, Byzantium and the East; Howard-Johnston, Witnesses; Sarris, Empires of Faith; for disputes over Caucasian and Armenian border areas, Dignas and Winter, Rome and Persia.

168

GREG FISHER

Against this background, the choice of religious affiliation became infused with political overtones, and Christianization became, for the Romans, a powerful tool through which entire communities could be politically subsumed into the Christian oikoumenē and denied to the Sasanians. 9 Bonds of trust and obligation between allied states were strengthened, when political power in those states was centralized around Christian leaders. 10 A shared religion brought common cultural and political ties; the speech which Agathias placed in the mouth of the pro-Roman and Christian Lazican elder, Phartazes, as he argued against the desirability of an alliance with the Sasanians, underscored this important Late Antique political reality. 11 Showing an interest in the new religion could also liberate: Evagrius relates how Persarmenian envoys came to Justin II in secret, pleading to become vassals of a Christian Roman Empire, to free them from the restraints placed on them by the Sasanians. 12 The role of Christianity in defining political affiliation, and binding peripheral ‘outsiders’ to the state, was quickly recognized by ancient authors, who added it to their well-developed lexicon of civilising influences. Deceptively one-sided narratives described ‘conversion’, emphasising both the barbarity of those who had not embraced the new religion, while showing the benefits of ‘turning’ away from worshipping idols and false gods. Language of rebirth and redemption emphasized passing from one state to the next. 13 Arabs played a role in such stories, where they were used to show how remaining ignorant of Christianity could be correlated with barbarism; the literary models deployed were not necessarily new to Christian authors. 14 Ps.-Nilus’ Narrationes lamented the ‘bestial and Maas, “Delivered from their ancient customs”. E.g. Haas, Mountain Constantines. 11 Agath. Hist. 3.12.7–8. 12 Evag. HE. 5.7. 13 The literature on conversion is extensive. The classic account remains Nock, Conversion. For discussion see Lim, Christian triumph and controversy; MacMullen, Christianizing the Roman Empire; Brown, Power and Persuasion in Late Antiquity. 14 E.g. Amm. 14.4.1–4, for the Arabs. For an engaging discussion, Shaw, “Eaters of flesh, drinkers of milk”. 9

10

FROM MAVIA TO AL-MUNDHIR

169

bloodthirsty life’ of the Arabs, who worshipped not Christ, but the ‘Morning Star’; worse, the writer claimed, they liked to sacrifice children at dawn! Only conversion could save them. 15 Ps-Nilus also tells us of savage raids on holy men by Arab barbarians, St. Sabas implores the Emperor Anastasius to build a fort to protect his laura, and the pilgrimage story of Egeria refers to a fort at Clysma, constructed specifically to defend against Saracen raids. These stories reflect a common concern that the holy might easily become the prey of the lightning-fast and savage swoops from the desert which such barbarians perpetrated. 16 Narratives of this sort highlighted the barbarity of the non-Christian Arabs, as well as the corresponding holiness of monks, nuns, and others who were frequently taken captive and sold, sacrificed, or ransomed. 17 Even if they were sometimes overblown in the ecclesiastical literature, raids of this sort were a Near Eastern reality, and not always targeted at hapless ascetics. An inscription from 334 from near Azraq, in Jordan, records the construction of a reservoir, needed to save Roman soldiers the risk of Arab ambush while out foraging for water. 18 Authors writing in Syriac recall similar ‘atrocities’ perpetrated by non-Christian Arabs against Christians. 19 Stories of sacrifice to barbarian deities also appear in the writings of Procopius, in his description of the activities of the Sasanian Arab ally, al-Mundhir Ps.-Nilus, Narrationes, 3.1. For Ps.-Nilus, Ammonius, and others, see now Caner, History and Hagiography and Ward, The Mirage of the Saracen. 16 Ps-Nilus, Narrationes, 4.1–5; Cyr. Scyth. Vit. Saba, 72; Egeria, Travelogue, 6 (= Caner, History and Hagiography, 212). Lenski, Captivity and slavery 254–6, on fortifying against Saracens, and Caner, History and Hagiography 273–86, discussing testimony of Procopius, Theophanes, and others. 17 For others, see Ammonius, Relatio 3–5; Mich. Syr. Chron. 2.422. See Lenski, Captivity and slavery, and Ward, The Mirage of the Saracen. 18 AÉ, 1948, 136. See Kennedy, Roman Army in Jordan 72; Parker, Romans and Saracens 144. 19 E.g. Isaac of Antioch, 1.208; Ps-Zach, HE 8.5 on al-Mundhir sacrificing four hundred virgins from ‘the assembly of the apostle Thomas’ at Emesa (Homs), Syria. For what is still the best survey of Syriac writings on Arabs, see Segal, Arabs in Syriac literature. 15

170

GREG FISHER

(504–554). 20 Jerome’s biography of St. Malchus took delight in describing the hapless saint’s capture by savage Arab raiders, who compelled him to drink camel milk, and eat barely cooked meat. To make matters worse, Malchus lost his clothes, and assumed the stereotyped lifestyle of the wandering shepherd. Nothing, Jerome’s audience would easily have understood, could be farther from Christian civilization. 21 As a didactic, moralising message, Christian authors provided numerous stories of hermits, monks, and priests, who encountered the whole spectrum of barbarians—including Arabs—and through their divinely-inspired actions, produced startling results: the childless conceived, nomads ceased their travels and built homes, evil spirits were banished, and the ignorant learned how to bake bread, farm, and engage in other pursuits vital to civilization. 22 Healing miracle stories, grounded in the repudiaton of earlier and false faiths, were particularly popular. Theodoret’s biography of Symeon the Stylite, for example, describes how an Arab leader begged Symeon to heal a man paralysed in his legs. Symeon ordered the irrevocably damaged man to give up his impiety, and then, when he was suddenly made well, to pick up the Arab who had brought him to the saint. Astonished, the newly-healed duly hoisted the Arab, ‘a large man’, proving the saint’s power to the astounded multitude who witnessed the miracle. 23 Other examples of this story type exist in the literature, including the famous tale of Zokomos, who was unable to father children, and so sought the healing powers of those who promised a divine solution. In exchange, he arranged for his subjects to be baptized and, tellingly, to provide soldiers for Rome. 24 Proc. BP 2.28.12, on al-Mundhir’s sacrifice of a son of the Jafnid leader al-Ḥārith (active 528–69) to Aphrodite. 21 Jer. Vit. Sanct. Malch. 4–5. For the most recent discussion of this episode, see Lenski, Captivity and slavery 237–9, and Fisher, Arabs and Empires Ch. 6 (G. Fisher). 22 E.g. Cyr. Scyth. V. Euth. 15; Soz. HE 6.38; Jer. Vit. Hil. 1–12, 25; Antonius, V. Sym. 22; Theod. V. Sim. 21. 23 Theod. V. Sim. 16. 24 Soz. HE. 6.38. 20

FROM MAVIA TO AL-MUNDHIR

171

Such categorical stories, intended to entertain, moralize, and teach, naturally obscured the complexities of ‘conversion.’ While Christianising did of course involve change, adopting Christianity did not necessarily mean the irrevocable loss of earlier, nonChristian identities; nor did the introduction of a monothestic religion immediately jeopardize the survival of paganism, in a world where non-Christian forms of monotheism existed. 25 Rather, across the spectrum of possibilities, there was room for a surprising range of individual and group responses. Looking, then, beneath the binary structure of conversion narratives, we find subtexts which show the benefits, and pitfalls, faced by new converts, including Arabs, of manifesting even a superficial interest in Christianity.

TEREBON, ASBEPETOS, AND AMORKESOS: SOCIAL

STRATIFICATION AND POLITICAL OPPORTUNISM

Against the background of a common ‘healing story’ template, Cyril of Scythopolis tells the story of Terebon, the son of an Arab potentate named Aspebetos, a vassal of the Persians. A persecution of Christians had convinced Aspebetos to switch sides, a decision assisted by Terebon’s illness, which the Sasanian magi had been unable to heal. These two facts—his rather convenient (or opportune) sympathy towards Christians, and his need for a ‘true’ healer— highlighted the suitability of Aspebetos to receive Christianity. Sure enough, when he arrived in the Roman Empire, the magister militum Anatolius engaged Aspebetos as phylarch, a title used for Arabs serving under Roman commanders in a range of functions, including frontier defence. 26 While Aspebetos went about his duties,

Athanassiadi and Frede, Pagan Monotheism in Late Antiquity; Mitchell and van Nuffelen, One God. Cf. Bowersock, Polytheism and monotheism 6: ‘In Caesarea sat a Christian bishop, while pagan gods were cultivated alongside the Talmudic investigations of rabbis… At Petra, amid the rock tombs of ancient Nabataean worthies, and virtually adjacent to a Nabataean temple, stood a Christian church within earshot of the annual celebration of the birth of the indigenous god Dusares…’ 26 For the best comprehensive discussion of the position of phylarch, see Grouchevoy, Trois ‘niveaux’ de phylarques. 25

172

GREG FISHER

Terebon had a dream, which led his father, and those with him, to St. Euthymius. 27 The saint cured Terebon’s illness; Cyril narrates that ‘the barbarians, astounded at so total a transformation and so extraordinary a miracle, found faith in Christ.’ In a fine example of the popular framework of a conversion narrative, emphasising rebirth and renewal, the Arabs with Terebon and Aspebetos immediately abandoned their old ways, and were ‘transferred through baptism from slavery to freedom.’ 28 This idea reappears in another part of the story, where Cyril explains that the Arabs of Aspebetos had ‘formerly been wolves of Arabia, but had then joined the rational flock of Christ.’ 29 The finality of the ‘conversion’ which took place was designed to underscore the precious attainment of membership in a Christian world, obtained only by decisively forsaking one’s previous existence. Elsewhere, those who flocked to see Symeon the Stylite gave up idols and chose God, 30 while the leader of the pro-Persian Naṣrid Arabs at al-Ḥīra melted down (sometime after 580) a golden Aphrodite, choosing baptism, once and for all giving up his ‘savage’ former way of life. 31 The healing of Terebon may have convinced some that the God worshipped by Euthymius was more powerful than their own, and thus offered a wise choice; in a world with a range of deities, adding another was hardly new, especially when that deity offered a quantifiable advantage. Whether or not this was genuine commitment is impossible to know, and the question is, in any case, a circular one. What is more interesting, and measurable, is the range of benefits and changes which accrued to the followers of Aspebetos, and particularly Aspebetos himself, as a result of the events narrated by Cyril. The newly-baptized Arabs built churches, ovens for baking bread, cisterns, and settled down, ceasing their wandering. 32 These events may simply reflect the typecast benefits of membership in the Christian oikoumenē, advertised by those who promoted Cyr. Scyth. V. Euth. 10. Ibid. Translations are those of R.M. Price. 29 Cyr. Scyth. V. Euth. 15. 30 V. Sym. Syr. 77. 31 Evag. HE. 6.22. 32 Cyr. Scyth. V. Euth. 15. 27 28

FROM MAVIA TO AL-MUNDHIR

173

the activities of holy men like St. Euthymius. On the other hand, the existence from this time of the Palestinian ‘Parembole’ or ‘Encampment’—created by Euthymius, and whose first bishop was Aspebetos—does suggest that social changes were taking place as a consequence of Christianization. Recently discussed-evidence for the sedentarization of Arabs in the Eastern Desert suggests that similar changes may have also been taking place across the Sinai. 33 Hermitic lauras elsewhere in Palestine attracted Saracens, as is related in the Life of St. Sabas (439–531), himself a friend of Euthymius. 34 Simeon the Stylite offered an immobile holy resource (Theodoret exclaims that for a time, he even chained himself to a rock!) and represents another locus around which would-be converts could congregate. 35 Religiously-associated fixed points, some of which were probably encouraging settlement, did exist, and could be people, as well as objects or places. 36 Changes resulting from the introduction of Christianity to would-be converts were not confined to Palestine. They could also be more overt in their pressures to settle or in the way that they reordered the worlds of the newly-Christian. The Caucasian Tzani, when Christianized, found their territory punctuated by a different sort of fixed points—forts and other military installations—which sought forcibly to orient their attention around expressions of imperial control. 37 Sometimes a more subtle approach was used. The Life of the bishop of Tikrīt, Aḥūdemmeh (d. c.575) narrates his attempts to reach out to the Arabs of the Iraqi steppe and desert, and to focus their religious activities around static centers of worship. 38 While Aḥūdemmeh did not attempt to sedentarize them, this more Power, The material culture and economic rationale of Saracen settlement; Ward, The Mirage of the Saracen. 34 Cyr. Scyth. V. Saba. 13. 35 Theod. V. Sim. 10. 36 Cf. Soz. HE 6.38.10. The importance of fixed points, especially in rural areas, is examined in Fowden, Rural converters, and Fisher Arabs and Empires Ch. 6 (G. Fisher). 37 Proc. BP 1.15.26; Aed. 3.6.6–8, 3.6.9–14. For discussion see Maas, “Delivered from their ancient customs”. 38 Hist. Ahud. 4; Hainthaler, Christliche Araber 106–7. 33

174

GREG FISHER

inconspicuous form of control exercised over the movements of the desert Arabs was itself an important form of social influence. 39 In less ancient times, the introduction of Christianity in the colonial Americas involved pressures to adopt new social frameworks, including the introduction of settlements organized around churches and church missions. 40 The use of such fixed places was understood by observers, both ancient and modern, to help with converting peripheral, nomadic, or ‘unsettled peoples’, by reordering space, and traditional lifeways. While the followers of Aspebetos may have undergone social change, the bulk of the benefits—particularly the political profit— went not in the majority to them, but to Aspebetos, whose relationship with his people dramatically changed when he transitioned from tribal chief to a bishop, under the authority of a church hierarchy. This divergence of the élite stratum and the group into a new hierarchy reflects a common process, stratification, catalyzed by the introduction of external pressures. Tribes are, and were, particularly susceptible to stratification: as socio-political organizations arranged around family groups and under the leadership of a chief, they are more decentralized and egalitarian than states, which tend to be hierarchically-organized and use established institutions to control resources. 41 When examining older tribal societies, we are largely dependent on modern anthropological studies for context, which can present difficulties. Yet while it would be erroneous to assume that ancient tribal societies necessarily functioned in precisely the same way as modern ones, applying the results of anthropological research to ancient problems has proved productive. 42 Moreover, the characteristics of tribal leadership, outlined Fowden, Barbarian Plain 120–9. Radding, Wandering Peoples; Braatz, Surviving Conquest. 41 For definitions, see most recently, Salzman, The meeting of the twain 83–4; Szuchman 2009; references in n. 42. 42 Salzman, The meeting of the twain, and Lancaster and Lancaster, Concepts of leadership. See also, for ancient and modern perspectives, Tapper, Introduction; Lancaster and Lancaster, Tribal formations; Lapidus, Tribes and state formation 25–6; Tapper, Anthropologists, historians, and tribespeople 50. 39 40

FROM MAVIA TO AL-MUNDHIR

175

here, are highly relevant to the careers of the Arab individuals discussed in this paper. Chiefs of tribes depended not on coercive control for their position, but on the consent of the families whom they led. Their élite status came from their ability to mediate disputes within the tribe and maintain peace, and, especially when states became involved, as is the case for the fifth- and sixth-century examples examined here, through acting as representatives for tribal interests, delivering benefits to the tribe. 43 In other words:

or:

the functions of tribal leaders, those of reputation, good men…are as spokesmen, negotiators, advisors, consultants, instigators of defence, and initiators of enterprises…leaders have a basic function for tribespeople and others of solving problems, which they do by advice, generosity, and mediation. 44

the smallish, close-knit yet brittle tribe as the autonomous center; the sayyid, ‘spokesman’ and leader, effective through the assumption of responsibilities, hence prestige and influence rather than circumscribed prerogatives. 45

These functions of the chief, specifically that of providing benefits to the tribe, could be adapted into an imperially-controlled context because the Romans favored centralizing communication around prominent leaders. This was partly accomplished by creating the office of phylarch, used in Arabia to access the resource of the tribe. Elsewhere, for example, in North Africa, Roman authorities The role of the chief as outlined here is clearly explained in Salzman, Pastoralists esp. 77–101. See too Salzman, Tribal chiefs as middlemen. 44 Lancaster and Lancaster, Concepts of tribe. The consensual position of chief is underscored by Beck, Qashqa’i 218: ‘the nature of the gathering [of ilkhani, or chief, and tribal members] was relatively egalitarian, and the decision-making process was often informal. People presented their problems in this open forum, presided over by the ilkhani and his advisors, and decisions could be reached in the course of group discussion without the ilkhani’s having made any kind of definitive statement’. 45 von Grunebaum, The nature of Arab unity 11. 43

176

GREG FISHER

designated the leaders of tribes as principes, dealing with them to maintain peace and stability in tribal communities. Occasionally, as in Spain, early imperial Syria, and north Africa, the authorities used Roman officials as prefects to provide the same conduit between tribe and state, but by the late Empire, prefects in north Africa were drawn from tribal leaders, reinforcing the position of a tribal chief as the legitimate point of contact between tribe and state. This general scenario was not confined to the Roman Empire: in the kingdom of Mari, representatives of the tribe were appointed to negotiate transactions between tribe and state, and in more recent times, the Iranian government focused on the position of sardar (tribal leader) as the way to access the tribe and negotiate with it. 46 Two important and connected points arise from this. Firstly, because the chief did not possess any innate authority, but ‘worked for the tribe’, his prestige was maintained through success in his position, primarily through mediation and accessing state resources on behalf of the tribe. Imperially-designated offices such as phylarch were thus more of a help than a hindrance. They gave tribal leaders access to the resources of the Empire, and, in some cases, the Emperor himself. This meant that high-profile political opportunities stemming from holding the phylarchate, such as arbitration at the invitation of the Emperor, which buttressed authority within the tribe through association with high-prestige people, places, and events, were highly desirable. A chief such as al-Ḥārith the Jafnid could, as we shall see below, point to such personal connections as proof of his ability to fulfill his role. Secondly, ‘hierarchical political institutions’ within the tribe tended to be generated by dealings with states, rather than by evolving by themselves. 47 The development of such hierarchies in the tribe, characterized by the elevation of the chiefs by imperial support (political, moral, or financial) and the access of the people Modéran, Les Maures, for principes and prefects in North Africa. For prefects in Syria and Spain, Sartre, Middle East Under Roman Rule 56; Cotton, Some aspects. For Mari, Matthews, Pastoral Nomadism in the Mari Kingdom 139–40, and for the sardar, see Salzman, Tribal chiefs 205. 47 Irons, Political stratification 362, discussed by Salzman, Pastoralists 101. 46

FROM MAVIA TO AL-MUNDHIR

177

under them to new sources of food, shelter, and opportunities as soldiers, monks, and so on, could all be causes of stratification. Changes in the position of the chief could have significant consequences. In 1836, for example, political power in Jabal Shammar (in central-northern modern Saudi Arabia), hitherto separated between the mostly sedentary population of Hail, and the nomadic sections of Shammar, was centralized around a single chief. The man who accomplished this, ʿAbdallāh Ibn Rashīd, did so in 1836 by deposing his cousin, Ibn ʿAlī, the ruler of Hail. Ibn Rashīd, a tribal chief of the Abde section of Shammar, thus found that not only had he retained control of a nomadic part of Shammar, but that he now also had to assume the functions of a town ruler. His position was transformed, and subsequent events—the raising of a standing military force, an organized method of extracting taxes, and so on—reveal that under his leadership, hierarchical mechanisms associated with the state, rather than the tribe, were emerging. Ibn Rashid was able to mitigate some of the problems all of this posed to his relationship with the tribe by acquiring a reputation for generosity. As a man who could find (financial) solutions to issues faced by tribesmen, Ibn Rashīd found a way to keep his prestige intact, and as a military leader who could deliver victory to his soldiers, he provides a good example of the Shammar ideal of al-amīr sayf wa manṣaf—a generous and victorious tribal leader. 48 This vignette reveals the importance of maintaining a continuity of benefits to the tribe as a way to manage changes at the top, the same solution, to the same problem, faced by the Jafnids in the sixth century. Not all chiefs could maintain their position when faced with the challenges and intrusions presented by the state. The chief and his coterie had always represented an élite stratum in the tribe, but Christianization into a world of religiously-influenced imperial politics could further, and more dramatically, stratify tribes into élite levels, connected to and with stakes in the hierarchy of the Empire, and non-élites, whose traditional relationship with tribal leaders was now threatened. 49 (It should be noted that such stratification was 48 49

al-Rasheed, The process of chiefdom-formation 35–6. Cf. van Dam, Becoming Christian 50. For modern parallels, see

178

GREG FISHER

not necessarily an official policy, but, rather, a potential consequence of adopting the state religion). With Christianization and baptism, then, the position of Aspebetos changed. He took the name Peter, and was ordained as bishop, with the consent of the patriarch of Jerusalem, Juvenal. Later, he took part in the council of Ephesus in 431, moving now in the same circles as those arbitrating the heresy of Nestorius. Peter’s subscription is found in the acts of the Councils, and from the debates and arguments of the Council, Cyril states, Peter reported to none other but St. Euthymius himself. 50 What had happened to Aspebetos’ position as chief? Cyril has cause to aggrandize the position of Aspebetos, in order to show the power that came with Christianization, but even if parts of the story are exaggerated, the role of bishop accorded to Aspebetos engendered a signal change from his mundane earlier job—guarding the frontiers against Christian escapees, on the orders of the Sasanian magi. While the most dramatic examples of stratification of Arab converts are provided by the Jafnids, who attained audiences in Constantinople, took Roman titles, patronized building construction, and were celebrated on church mosaics and monastery inscriptions, Cyril’s story provides an early glimpse of some of the social and political changes which might have been occurring as Arab tribes accepted Christianity. Aspebetos found advancement and a new role in the religious hierarchy of the Empire. Other Arab newcomers also found opportunities through becoming Christian, while the Roman state found that accepting new Christian allies could relieve pressure on the Empire during difficult times. A story related by Malchus illustrates some of the variables involved and underscores once again how the position of tribal leaders could be altered by engagement with the state. In 474, an Arab adventurer named Amorkesos sent a priest, Peter (not to be confused with Aspebetos) to treat on his behalf with the emperor Leo. Amorkesos was, like Aspebetos, an émigré from the Sasanian Empire, and had obtained for himself the Adelman and Aron, From borderlands to borders 830–1; Weber, Spanish Frontier 105–6; Barber, Social Stratification. 50 Subscription: ACO, 1, vol. 1/7, 85 (‘of Peter of the Parembole’) and 115 (‘Peter bishop of the Parembole’); Cyr. Scyth. V. Euth. 20.

FROM MAVIA TO AL-MUNDHIR

179

island of Iotabe in the Gulf of ʿAqaba. Here, he expelled the Roman tax officials and installed himself in their stead, becoming wealthy from imposing tariffs on the trade which flowed through the region. 51 Probably realizing that he was about to attract the wrong sort of Roman attention, he used Peter to obtain an alliance with Leo to legitimize his position. Malchus says that Leo received him, invited him to dinner, and seated him amongst the patricians present. He also gave Amorkesos public funds, gifts, and, above all, allowed him to retain Iotabe. Malchus is hostile to Leo, and castigates him for accepting Amorkesos, via Peter’s entreaties, as one who had been ‘persuaded’ to become a Christian. Leo may have had ulterior motives for allowing Amorkesos to keep his position: military levies were heavily depleted after the loss of the Roman expeditionary force at Cape Bon in 468, and Leo faced tensions with both the Ostrogoths and the Persians. 52 Amorkesos, for his part, succeeded in staving off the expected armed Roman response, which did not materialise until the reign of Anastasius, in 498. 53 The agreement between Leo and the ‘barbaric’ Amorkesos would probably have been unthinkable without the veneer of respectability provided by the priest, Peter, and Amorkesos’ claim to have become a Christian. If Amorkesos had gained an advantage, though, it was something of a false economy. He had won a reprieve, and presumably maintained his position by translating the tax wealth into a reputation for generosity (as with Abdullah Ibn Rashid). But he had also politically subordinated himself to a Christian Emperor, within a hierarchy which was not of his choosing. In effect, he was not dissimilar to Aspebetos in adopting a Roman identity, expressed via an open connection to the Christian religion, which placed him firmly within Roman, not tribal power structures. The Christianization of Arabs thus contained both benefits and liabilities for would-be converts. Christianization offered a form of acceptance into the oikoumenē, and offered new roles for those who took advantage of it: bishops, military and community Malchus, fr. 1. For Iotabe, see Mayerson, The island of Iotabe. Priscus, fr. 37 (Goths); Blockley, East Roman Foreign Policy 74–5 (Persians). 53 Theoph. Chron. 141. 51 52

180

GREG FISHER

leaders, soldiers, and parishioners. The careers of Aspebetos and Amorkesos show how significant and tangible benefits could accrue to the chiefs, as tribal élites, but their positions also changed as a result. During the time period of these stories, the Romans pursued very much an ‘ad hoc’ form of alliance with Arabs along the south-eastern frontiers of the Empire, but by the sixth, the policy was altered to support a single family dynasty, the Jafnids. Christianity continued to play an important role in the relationship between Rome and the Arabs in the sixth century, as it had in the fourth and fifth. Between the support of the Emperor and the opportunities fostered through their religious activities, the Jafnids emerged as high-profile Christian Arab leaders. But even with such close links to the Empire, the Jafnids show that whatever else being Christian might offer to the Arabs, it was entirely compatible with remaining tribal, and the conduct of the family remained grounded in the essential chiefly duties of mediation and communication between tribe and state. The Jafnids, St. Sergius, and the monophysites Changing relationships Between the third and sixth centuries, the relationship between the Romans and the Arabs who lived along the southern and eastern frontiers of the Empire grew steadily more complex in ways which parallel frontier relationships in the west: the recruitment of barbarians into the Roman military, agreements with individual barbarian leaders for military assistance, and the appearance of the formal language of treaties, as barbarian foederati became part of Rome’s military hierarchy. 54 Early evidence for Roman efforts to build friendly relations with Arabs along the southern frontier comes from the Greek/Nabataean Aramaic bilingual inscription, from a temple at al-Ruwāfa, now in Saudi Arabia, dating to the reign of For the west, see Heather, The late Roman art of client management; Heather, Foedera and foederati; Heather, Empires and Barbarians; Halsall, Barbarian Migrations 417–54; James, Europe’s Barbarians. For comparison between east and west, Fisher, Between Empires 72–127. 54

FROM MAVIA TO AL-MUNDHIR

181

Marcus Aurelius. 55 An inscription from Ḥegrā/Madāʾin Ṣāliḥ, southeast of al-Ruwāfa, confirms that Roman soldiers were present in the area during the same period. 56 A century and a half later, the famous Arabic language inscription from al-Namāra, in Syria, demonstrates the importance of imperial support in underwriting Arab political authority. 57 The Romans were actively extending their reach across the southern frontiers, as they had along the Rhine and the Danube in the west, and by the late fourth or early fifth century, units designated Saraceni indiginae were part of the late Roman army. 58 Arabs offered military services of Julian during his ill-fated expedition, 59 while the use of the term hupospondoi, sg. hupospondos ‘those/one under treaty’, and the equivalent of the western foederati, 60 appears in a variety of contexts describing alliances with Arabs, suggesting that the Romans were attempting to apply a formal administrative or legal framework to their increasing use of Arab militia. 61 For an entirely new edition of these inscriptions by Michael Macdonald, see Fisher (ed.), Arabs and Empires, Ch. 1. For older versions, see Milik, Inscriptions grecques et nabatéenes; Macdonald, Quelques réflexions sur les Saracènes. 56 al-Talhi and al-Daire, Roman presence in the desert. 57 For a new edition of the inscription from al-Namāra by Michael Macdonald, see Fisher (ed.), Arabs and Empires, Ch. 7. For older versions, see: Bordreuil et al; Macdonald, Old Arabic 469. For a recent discussion of the inscription, Fisher, Between Empires 138–44. 58 E.g. ND Or. 28, 32; see Amm. 22.15.1–2: ‘et Scenitas [tentdwelling]…Arabas, quos Sarracenos nunc appellamus’. The term ‘Saracen’ (Syr. ṭayyāyē) appears from the third/fourth century to describe the Arabs of the desert, although the reasons for the change are not clear. For a summary of possibilities, see Macdonald, On Saracens; Graf, The Saracens; Hoyland, Arabia and the Arabs 235; Segal, Arabs in Syriac literature 101–5. 59 Amm. 23.3.8. 60 Malchus, fr. 15, makes this equivalence. 61 E.g. Malchus fr. 1 (Amorkesos); Soc. Schol. HE 4.36, on Mavia (discussion in this section); Soz. HE 6.38, 7.1; Cyr. Scyth. V. Euth. 10 on Aspebetos (above); Soc. Schol. HE 7.10, concerning Alaric. 55

182

GREG FISHER

Even before the arrival of Aspebetos and Amorkesos, Arab military allies were expected to be Christian. The story of Mavia (fl. 375) is an illustration of how that expectation might also cut both ways. 62 Mavia was a queen, married to an unnamed Arab leader, who had been allied with Valens. When the alliance was cancelled after the death of her husband, Mavia took umbrage and rebelled, coming close to worsting in battle a very senior Roman commander who was saved only at the last gasp by the subordinate who had initially summoned his assistance. Peace was agreed when the Romans acceded to her demand that a non-Arian bishop, Moses, would be provided for her and her people (Moses may even have been one of her tribe, but this is not clear). On her return to alliance with Rome, it seems that Mavia’s forces defended Constantinople in the wake of Valens’ death at Adrianople, 63 and Mavia sealed her new status and pact with the Empire by marrying her daughter to another senior military commander, Victor. 64 Valens’ attempt to ‘force’ Arianism on Mavia seems to have been part of the reason for the rebellion, and while the religious angles of the story might have been exaggerated by the numerous church writers who report it, the details underscore the importance of a common, shared religious outlook in creating ties of trust. Indeed, the Arian/Orthodox tensions in the story of Mavia (if they are not inventions of Rufinus) 65 suggest that while the Romans saw value in demanding religious conformity from their allies, at least the same might be said about the Arabs themselves—this was not a passive receipt of Christianity, but an active interest, and the tale suggests that others beyond the tribal élite found the religion relevant. The Christian outlook of Mavia helped not just in dealings with imperial authorities, but was also of relevance to her people. As a tribal For a detailed examination of this episode, see Lewin, ʿAmr ibn ʿAdī, Mavia, the phylarchs and the late Roman army; Schmitt, Mavia, die Königin der Sarazenen; Bowersock, Mavia, Queen of the Saracens; Mayerson, Mavia, Queen of the Saracens, who warns against over-interpreting the story as it is reported by ecclesiastical historians. 63 Soz. HE. 7.1. 64 Soz. 6.38; Soc. Schol. HE. 4.36; Theod. HE. 4.20. 65 Mayerson, Mavia, Queen of the Saracens 131. 62

FROM MAVIA TO AL-MUNDHIR

183

leader, her position required her to maintan her prestige and deliver benefits to the tribe. Military victory against a formidable enemy, and extracting concessions from the Emperor, did just that, and it is clear from the subsequent defence of Constantinople, that she maintained her positions both of Roman ally and tribal chief. We might thus imagine that the provision of a bishop for her people, in which she effectively delivered state resources to the tribe, did for Mavia’s reputation what it would do for that of al-Ḥārith the Jafnid, nearly seven generations later. Delivering benefits thus helped tribal leaders to maintain their position and their traditional roles as chiefs, a process helped by the Roman policy of dealing directly with tribal representatives, rather than the tribe as a corporate entity. This ensured that Mavia, not the Romans, could portray herself as the benefactor of her people. Mavia was probably also helped by not working too closely with the state itself. Although she married her daughter to the magister militum Victor, there is no evidence that she played the kind of high-level political roles enjoyed by the Jafnids. The Jafnids The Jafnid family was the beneficiary of the Empire’s decision, after the fifth century, to give up a fragmented system of alliances of hupospondoi and concentrate, instead, on one single alliance, as a counterpoint to the Sasanian Empire’s reliance on the Naṣrid dynasty. 66 (While debate continues over the usefulness of terms such as ‘Jafnid’, for the sake of convenience, and because of the familiarity of such terms, we shall treat here the family leadership of Ghassān, comprised of the descendants of Jabala, as ‘Jafnids’). 67 Between 500 and 528, the Romans turned their attention to the northern part of the Arabian Peninsula. Disturbances in c.500 involved a man named Jabala, who seems to have concluded an agreement with the Emperor Anastasius after being defeated by Romanus, the commander in the region. 68 At the same time, the Romans intensified

Outlined by Proc. BP. 1.17.40–7. A concise overview of the terminological problems is given by Hoyland, Insider and outsider sources 267–74. 68 Theoph. Chron. 141. See discussion in Shahîd, Byzantium and the 66 67

184

GREG FISHER

diplomatic efforts to win over the Ḥujrid family of Kinda, former vassals of the kingdom of Ḥimyar, which resulted in an alliance in 502/3. 69 But by 527/8, the Ḥujrid alliance collapsed with the death of its leader in battle, and the Emperor Justinian shortly afterwards turned to the son of Jabala, al-Ḥārith, known as ‘Arethas’ in Roman sources, and placed him in a position of authority over ‘as many clans as possible.’ 70 Al-Ḥārith, and his son, al-Mundhir, retained the confidence of the Empire until 582, providing forces for frontier defences and playing a role in several important engagements, including at Callinicum in 531, attacks on the Naṣrid base at al-Ḥīra in Iraq, and an attempted invasion of the Sasanian Empire across the Euphrates in 580. 71 They also played a considerable role in ecclesiastical politics, becoming increasingly active over the course of the sixth century as advocates of the monophysites. This choice to support the monophysites, opposed for much of the sixth century to the orthodox Chalcedonian position desired by the Emperor, can perhaps be explained by the correspondence between the timing of the entry into the Empire of the Jafnids and their followers, Ghassān, and the presence on the throne of the pro-monophysite Emperor Anastasius. For new allies wanting to be aligned with the dominant form of religious belief in the early sixth century, monophysitism was the obvious choice, and with the death of AnArabs in the Sixth Century, ii/ii 10–11. 69 The outline of events is explained by the career diplomat Nonnosus, preserved by the later compiler Photius, Bib. 3; see Fisher, Arabs and Empires Ch. 5 (G. Fisher); Bowersock, Throne of Adulis 135–43; Millar, Rome’s Arab allies 207–9; Beaucamp, Le rôle de Byzance en mer Rouge 200–3. The Roman alliance with the Ḥujrids is inferred from Photius as well as from Theoph. Chron. 144 and Proc. BP 1. 20. 9–10. For discussion of this and the Ḥujrid alliance with Ḥimyar, see Robin, Les Arabes de Ḥimyar; Robin, Le royaume Ḥujride; Gajda, ‘Ḥuǧr b. ʿAmr ; Fisher, Kingdoms or dynasties. 70 Proc. BP 1.17.47. 71 Jafnid/Ghassān on Roman operations: Proc. BP. 2.16.1–6 (AD 541); local defence: Malalas, 446–7; Callinicum: Proc. BP 1.18.35, with discussion in Greatrex, Rome and Persia at War 200–7; attack on al-Ḥīra: Joh. Eph. HE 3.6.3–4; CE 580: Joh. Eph. HE 3.6.16–17.

FROM MAVIA TO AL-MUNDHIR

185

astasius almost a generation later in 518, it is distinctly possible that monophysite Christianity had now become the preference of the Arabs who had entered the Empire along with Jabala. It is not clear if Jabala played any political role in the Empire. By contrast, his son, al-Ḥārith, quickly emerged as a supporter of the monophysites. Whether al-Ḥārith did so for personal political gain or out of religious conviction is hard to measure, but great political prestige came in 542 when he became associated with a request to Justinian’s wife, Theodora, for two new bishops for the monophysites in Syria, who had suffered heavily under the persecutions of Justin I (r. 518–527). Jacob Baradeus and Theodore were made available, and quickly began, themselves, to consecrate a new batch of monophysite bishops to repair the damage. 72 Via his association with this event, al-Ḥārith applied the function of tribal chief to a wider setting, for it was apparent that even if the bishops were initially intended only for Ghassān, the wider monophysite population would be the beneficiaries. 73 The Jafnid leader thus assumed the position of chief not only of his tribe, but also a position of patronage over the monophysites in the communities of Syria and Arabia. He had thus become a Roman patron, but framed in tribal terms, and his son, al-Mundhir, would continue in the same fashion. Evidence of the respect won by al-Ḥārith and his family is found scattered across modern-day Syria and Jordan. The epigraphic evidence for the Jafnids places them within the late Roman aristocracratic hierarchy, with references to the titles of paneuphēmos (famosissimus) endoxotatos (gloriosissimus), and megaloprepestatos (magnificentissimus) and the honorific positions and titles of komes, stratēlatos (magister militum), and patrikios. 74 Greek inscriptions from Qaṣr alḤayr al-Gharbī near Palmyra log a visit by al-Ḥārith just before his death in 569, recording an archimandrite who used the tenure of alḤārith (‘Flavius Arethas, patrikios’) as phylarch to date his own time Joh. Eph. Vitae (PO 19, 153–4); Menze, Justinian and the Making of the Syrian Orthodox Church 222–3. 73 Hoyland, Late Roman Provincia Arabia 128. 74 See Bevan, Fisher, and Genequand, Late antique church for a table of Jafnid leaders and titles. 72

186

GREG FISHER

in office. 75 Elsewhere, it is al-Ḥārith who should likely be identifed with the the mosaics from the church of St. Sergius at Nitl, near Madaba, close to the newly-discovered church of St. Sergius at Tall al-ʿUmayrī East, which acclaim ‘Erethas, the son of al-Arethas’. 76 The Sergius cult was especially popular with the Arabs in antiquity, and the church at Nitl is an early example of the association between it and the Jafnids, a link assiduously exploited by al-Ḥārith’s son, al-Mundhir. 77 Meanwhile, al-Ḥārith’s brother, Abū Karib, appears in a monophysite Syriac codex from a monastery near Palmyra, where he carries the title mlk, ‘king’, a title also given to alḤārith on the Arabic inscription from Jabal Says in southern Syria. 78 A lintel from Sammāʾ in the Ḥaurān, based on Psalm 120, appeals for the protection of the ‘endoxotatos phylarch Abū Karib’, 79 and he also appears giving a gift of land to Justinian 80 and, significantly, mediating a dispute recorded in P. Petra inv. 83. 81 A possible link to the Jafnids comes from the region of Maʿarrat al-Nuʿmān in Syria, where a bronze plaque describes ‘Naaman’ as ‘endoxotatos’, phylarch, and ‘stratēlatos (magister militum)’. This may be the same individual as al-Nuʿmān, a son of al-Mundhir, arrested shortly after his father in (?)583. 82 The overall impression is one of regional IGLS 5.2553 b,d. For discussion see Genequand, Some thoughts, and Gatier, Les Jafnides 188–93. 76 Piccirillo, The church of Saint Sergius; Gatier, Les Jafnides 205. 77 Theoph. Sim. 5.1.7, and Sev. Ant. Hom. cath. (PO 4, 93), both make the link between the Arabs and the cult. See Fowden, Barbarian Plain, 98. 78 Abū Karib as mlk: Millar, A Syriac Codex; al-Ḥārith at Jabal Says: al-ʿUsh, Unpublished Arabic texts 320; Macdonald, A note on new readings; for the inscription itself, see the edition by M.C.A. Macdonald with commentary in Fisher, Arabs and Empires, Ch. 7, and Robin and Gorea, Un réexamen de l’inscription arabe préislamique. 79 Sartre, Deux phylarques arabes 151; Gatier, Les Jafnides 203. 80 Proc. BP. 1.19.8–14. 81 Kaimio, P. Petra inv. 83. 82 IGLS 4.1550; Gatier, Les Jafnides 200–1. The identification is suggested by Millar, Rome’s Arab allies 218. 75

FROM MAVIA TO AL-MUNDHIR

187

prominence, and the preponderance of Christian contexts underscores the public links between the Jafnids and Christianity. The site of Qaṣr al-Ḥayr al-Gharbī has been tentatively identified with the monastery of ‘Haliarum’, mentioned in a letter to Jacob Baradaeus confirming the faith of Arabian monophysite clergy, in the face of the Tritheist heresy. The 137 subscriptions to the letter, which contain a wealth of information on places and individuals, have recently been translated into English for the first time by Fergus Millar. 83 The letter itself is part of a larger collection of correspondence (BL Add. 14602), edited with a Latin translation by J.-B. Chabot. 84 Traces of the Jafnids appear both in the confirmation of faith (º41) and in correspondence in the wider collection. In º39 al-Ḥārith is described as the ‘Christ-loving and glorious patrikios’, working to find common ground between the letter’s authors and two Tritheist bishops, Conon of Tarsus and Eugenius of Seleucia, apparently in or around 568/9. 85 The mediatory function is significant, especially in its anticipation of the much wider intercessive role played by al-Mundhir in religious disputes. Either alḤārith or al-Mundhir should also be seen as the unnamed phylarch urging mediation alongside of Peter of Callinicum. This event took place, significantly, at a church of St. Sergius at Gabitha, a location which appears under signature 24 in letter º41 (‘the monastery of Beth mar Sergius of Gabitha’). 86 Gabitha is the Syriac rendering of Millar, Christian monasticism; van Roey and Allen, Monophysite Texts, usefully number the letters in the collection, a sequence followed by Millar and also here. For Haliarum: Letter º41 (=van Roey and Allen, Monophysite Texts 270), subscription º119 = Millar, Christian monasticism 112. See Gatier, Les Jafnides, and Genequand, Some thoughts 70, on Haliarum. 84 The letter (º41) was originally published by Lamy, Profession de foi, and the collection as a whole by Chabot, Documenta ad origines. For a general discussion of the documents see Millar, Christian monasticism, and van Roey and Allen, Monophysite Texts. 85 Letter º39 (=van Roey and Allen, Monophysite Texts 270). Date of the letter : Millar, Christian monasticism 106, citing Honigman, Évêques et évêchés monophysites 185. 86 Letter º41 (=van Roey and Allen, Monophysite Texts 270), subscrip83

188

GREG FISHER

(al-) Jābiya, a site frequently connected with the Jafnids by Muslim authors, but as yet unidentified. 87 Al-Ḥārith is also apparently the author of letter º23 in the collection, addressed to Jacob Baradaeus, although the main issue being discussed is not clear. 88 Al-Ḥārith’s son, al-Mundhir, also appears in the signature 121 in letter º41, which identifies Mar Eustathios as presbyter of ‘the church of the glorious and Christ-loving patrikios, Mundhir’. 89 This phrase presumably refers to a physical structure, not an organisation or hierarchy (below, n. 96). Throughout these examples, we find actions known from other contexts: involvement in mediation, a connection to St. Sergius, and an overall elevated level of prestige indicated by a personal connection with Jacob Baradaeus. The association with the provision of Jacob and Theodore to the monophysites was a supreme political coup for the Jafnids, and set the tone for the activities of the family until their demise in 582. That they supported monophysites rather than Chalcedonians seems to have mattered less than it might seem, especially under the tenure of Justinian and Theodora. Yet even when subsequent emperors returned to persecution, the religious landscape remained sufficiently ambigious for much of the sixth century, allowing even high-profile monophysites to also maintain allegiance to the Emperor. The latter years of Justinian’s reign bore witness to the ascendancy of the Jafnid family: in March of 548, al-Ḥārith and his brother sent envoys to Abraha, king of Ḥimyar. 90 In 562, a peace treaty agreed between Rome and the Sasanians changed the status of the Jafnids and their allies from hupospondoi to summachoi, recognizing the greater contribution in defending frontier areas and also in maintaining stability in Syria and Arabia. Although it is clear tion º24 = Millar, Christian monasticism 109. For Peter, see Mich. Syr. Chron. 2.384 and Millar, A Syriac codex 30. 87 Millar, A Syriac codex 30. On al-Jābiya, see Sartre, Trois études 177– 89, and Fisher, Arabs and Empires Ch. 4 (D. Genequand). 88 Letter º23 (=van Roey and Allen, Monophysite Texts 269). 89 Letter º41 (=van Roey and Allen, Monophysite Texts 270), subscription º121 = Millar, Christian monasticism 112. 90 CIS 4. 541, and newly-translated by C.J. Robin in Fisher, Arabs and Empires Ch. 3. See too Hoyland, Arabia 55; Smith, Events in Arabia 440.

FROM MAVIA TO AL-MUNDHIR

189

from Procopius’ discussion of foederati (hupospondoi) that the term, by his day, engendered greater equality with imperial authorities, the change to summachoi was significant, in that this term was also used in contemporary authors (including Procopius) to make a specific appeal to an independence of action which set summachoi apart from hupospondoi. 91 Al-Ḥārith died in 569. His son al-Mundhir succeeded him, and continued the trends begun by his father, particularly in support for the monophysites. The tenure of al-Mundhir also witnessed a greater emphasis on the cult of St. Sergius, an increase in public connections to rural churches, and the establishment of buildings. Through these actions al-Mundhir staked a claim to membership in the Roman Christian élite. Even as he slowly became more identified with this role, though, he avoided relinquishing his primary role of tribal leader. Muslim sources attributed a large number of buildings to the Jafnids. Modern consensus is that the lists of these structures, particularly that of Ḥamza al-Iṣfahānī (b. 280/893, d. after 349/961), are greatly exaggerated, but a very small number of buildings connected with the Jafnids have been identified. 92 They include Qaṣr al-Ḥayr al-Gharbī and Nitl, linked to al-Ḥārith, but the rest are associated with the reign of al-Mundhir: a martyrium near Damascus, no longer in existence, 93 and a house at al-Ḥayyat, in the heart of the Ḥaurān, with a Greek inscription commemorating the building of the structure ‘under the paneuphēmos Alamoundaros, patrikios’. 94 The newly-discovered church of St. Sergius at Tall al-ʿUmayrī Men. fr. 6.1; Proc. BP. 3.11.1–3. See for discussion Heather, Foedera and foederati 57–74 and Fisher, Between Empires 116–24. 92 For the most recent assessment, see Fisher, Arabs and Empires Ch. 4 (D. Genequand) and Ch. 8 (H. Munt). See for older assessments: Genequand, Some thoughts 78; Sartre, Trois études 182; Shahîd, Ghassānid religious architecture. 93 Wadd. 2562c, originally thought to be a tower. See Gatier, Les Jafnides 195–7 for the amended reading. 94 Wadd. 2110 = IGLS 16.628. Gatier, Les Jafnides 197–8. The author is grateful to Maurice Sartre for kindly sharing his new reading of this inscription prior to its formal publication in IGLS 16. 91

190

GREG FISHER

yielded an inscription with the invocation ‘our Lord Jesus Christ, God of Saint Sergius, protect the megaloprepestatos (magnificentissimus) Almoundaros, the komes’. 95 The reference to the ‘church of alMundhir’ in the collection of letters, discussed above, likely reflects another, unknown location. 96 The most prominent building connected with al-Mundhir is the extramural structure at the site of alRuṣāfa, the location of the major shrine of St. Sergius in the Roman Empire. The building, which is still standing, contains an inscription reading ‘the fortune of al-Mundhir triumphs’. 97 There has been a long debate over the exact purpose of the building, and whether or not it can (on the basis of the inscription) be accepted as having been built by or on the orders of the Jafnid leadership. In her pivotal study of the Sergius cult, Elizabeth Key Fowden suggested that the building possessed both religious and secular functions. Close to the ornate northern gateway of the city, it provided a highly visible locus where the Jafnid leadership might interact with pilgrims visiting the shrine, and for al-Mundhir to exhibit the position of authority which he enjoyed as an imperial Christian ally. 98 Several aspects of the building underscored its Roman, Christian connections: its plan mimics that of the baptistry of al-Ruṣāfa’s ‘basilica A’ (the church of St. Sergius) and it is stylistically similar to other sixth-century churches in Syria, 99 while the richly-decorated interior, featuring friezes, relief sculptures, and crosses, boasted of the the wealth and sophistication of alMundhir, via the inscription over the apse, ‘within the standard repertoire of fifth- and sixth-century church decoration’. 100 CuriSee the readings of the text in Bevan, Fisher, and Genequand, Late antique church, and Gatier, Les Jafnides 193–5. 96 Millar, Christian monasticism 114. 97 SEG 7. 188 and Gatier, Les Jafnides 198–200. 98 For discussion about the purpose of the building, see: Sauvaget, Les Ghassanides et Sergiopolis; Brands, Die sogenannte Audienzsaal des al-Mundir; Fowden, An Arab building; Fowden, Barbarian Plain 167–70, and the discussion by Genequand, Some thoughts 78. 99 Fowden, Barbarian Plain 153. 100 Fowden, An Arab building 307. For a later comparison, Fowden and Fowden, Studies on Hellenism 145, on the Q. ʿAmra frescoes: ‘they at95

FROM MAVIA TO AL-MUNDHIR

191

ously, however, the building was placed a considerable distance outside the northern walls of al-Ruṣāfa, which distanced it from the overwhelmingly imperial context of the fortified city. If al-Mundhir or the Jafnids did indeed use this space, it seems that they were seeking to appeal to as wide a base as possible—pilgrims passing through the northern gateway; shepherds and nomads of the steppe; and Roman agents and officials. Mark Whittow conceptualized the building as ‘the equivalent of a great shaykh’s seven-pole tent, but built in stone and in a Roman idiom… [which] lay outside the city, isolated like an immovable tent from any other structures’. 101 Developing this expression of the structure’s essentially tribal character, Fowden has argued that the building also fulfilled the function of a ḥaram, a ‘neutral ground where conflicting parties could meet and seek resolution of disagreement’, backed, in this case, by the authority and prestige of St. Sergius, whose shrine lay several hundred metres away, but whose blood from his own martyrdom was said to have spilled on the very ground where alMundhir may have held court. 102 Whether or not al-Mundhir ever set foot inside the structure, his invocation on the inscription, together with the proximity of the major shrine of the Sergius cult (linked as it was with the Jafnids) provides another prominent Jafnid connection to a site located for meeting and communication, deliberately disassociated from an urban context. It is hard to think of a more appropriate location for the tribal leader’s role of representing the tribe to the outside world, mediating disputes which threatened peace, and presiding over discussions between members of the tribe. To underscore this interpretation of the building’s function, it is necessary also to consider the other side of alMundhir’s role as a late Roman, but also a tribal, élite—as powerbroker and arbitrator for the monophysites in the Empire. In 579/80, a dispute began amongst the monophysites, threatening attempts which had continued intermittently throughout the sixth century to negotiate a compromise with the Chalcetest to a rather advanced state of Mediterranean acculturation on the part of their Umayyad patron and his immediate circle’. 101 Whittow, Rome and the Jafnids 222. 102 Fowden, An Arab building 314–15; Fowden, Barbarian Plain 98–9.

192

GREG FISHER

donians. After 542, the consecration of new bishops by Jacob and Theodore promised a new company of monophysite clergy which might eventually match the Chalcedonian hierarchy, but now, divisions between monophysites endangered the unity so crucial to any future success. Failure to provide a single, unified voice might cause loss of face, and also of ground in negotiation. What was worse, the schism potentially threatened the ability of the movement to defend itself against future persecution. At stake for the Emperor, meanwhile, was nothing less than the ongoing effort to negotiate religious peace and stability in the eastern provinces. The row had begun when Jacob’s authority was challenged through the consecration of Paul, as patriarch of Antioch, and become more complicated by 580, when two successive Alexandrian patriarchs, Peter and Damian, involved themselves: Peter had deposed Paul, and as Jacob had supported Peter in his struggle with Paul, those following Paul naturally turned on him. The end of Peter’s tenure in 578, when he was replaced by Damian (patriarch 578–606), did little to assuage the wounds which had opened. 103 Yet another complication emerged part of the way through when Jacob died in 578, and the contoversy continued to burn. Ironically, at the top level of the monophysite leaders, the bitter arguments had come to resemble a tribal feud, which urgently required mediation. Even if some in the Chalcedonian camp may have watched these developments with a certain sense of satisfaction, they were overruled by the Emperor, Tiberius II, who invited al-Mundhir to Constantinople as arbitrator. 104 Tiberius reinforced his expectations by providing al-Mundhir with a ‘royal crown’ (or diadem) and his sons with titles. 105 Given the prestige of the Jafnids amongst the monophysites, and the previous role of al-Ḥārith as mediator between the monophysites and Trithesists, the summons was logical. For al-Mundhir, though, problems loomed: in the 570s, he had crossed swords with Justin II, who had ordered an assassination van Rompay, Society and community 252, Allen, Evagrius 32–4, and Frend, Monophysite Movement 324–9 briefly cover this complex affair. 104 Joh. Eph. HE 3.4.38–41. 105 Ibid., 3.4.39. See the discussion of this episode in Fisher, Arabs and Empires Ch. 6 (G. Fisher). 103

FROM MAVIA TO AL-MUNDHIR

193

attempt, botched in the actual event, and a dangerous enemy was emerging in the person of Maurice, from 574–582, the Count of the Excubitors. 106 With the political credibility of the Jafnids thus on the line, al-Mundhir was surely wary of how his tribal supporters, who depended on him to deliver the benefits of the Roman alliance, would view any failure in Constantinople; in the event, alMundhir disappointed. The negotiations collapsed when Damian made an agreement with al-Mundhir, and then reneged after it had been presented as a fait accompli to the Emperor. John of Ephesus’ vitriolic discussion of the event screams the betrayal he felt on behalf of his Jafnid hero. 107 With al-Mundhir’s standing and integrity in tatters in Constantinople, and presumably with some among the tribe questioning his leadership, it was only a short while before alMundhir found, in a spectacular piece of bad luck, that his enemy, Maurice, had become the next Emperor, and al-Mundhir was deposed and exiled in short order. While the Romans acted first, it seems likely that there may have been elements among the tribe who saw in their leader’s failure an opportunity to depose him, but the clamour of his supporters at Bostra after his arrest shows that he still retained a great deal of support. 108 Furthermore, even though these events signalled the formal end of the Jafnid alliance, mediation remained a function of Arab leaders. A certain Jafna, in the 590s, appears as a broker in yet another dispute between monophysites. The argument once again involved the troublesome Damian, and, again, he seems to have scuppered proceedings. 109 Jafnid support for monophysitism was complex. The idea that al-Ḥārith and his son possessed a singular zeal for monophysitFor the curiously Shakespearean way in which the assassination was botched, see Joh. Eph. HE 3.6.3. For Maurice as comes excubitorum, PLRE IIIB, ‘Mauricius 4.’ 107 Joh. Eph. HE 3.4.43. 108 Ibid., 3.3.41–2. 109 Mich. Syr. Chron. 2.368; cf. Chron. 1234 p. 215, where Abū Jafna Nuʿmān b. al-Mundhir appears as a go-between for Maurice and Khusrau II. Hoyland, Late Roman Provincia Arabia 129 and Fisher, Arabs and Empires Ch. 5 (G. Fisher). 106

194

GREG FISHER

ism 110 is easily deduced from the partisan history of John of Ephesus, who had suffered personally in the persecutions of Justin II, and from later stories such as that related by Michael the Syrian (West Syrian Patriarch, 1166–99), whose Chronicle causes al-Ḥārith to engage in debate with Ephrem, a member of the Chalcedonian clergy. Ephrem puts the merits of Chalcedonianism, but al-Ḥārith refuses to be taken by the position and snubs Ephrem’s offer of bread, claiming that by coming from Chalcedonian hands, it is as tainted to the Jafnid as camel meat would be to his opponent. 111 The appearance of ‘camel meat’ as a negative element in the story recalls the stereotypes of traditional Roman views of the peoples of the desert, and this was, too, a very late story when the divisions between Chalcedonians and monophysites had hardened considerably. Further, as van Ginkel has shown, much of the Chronicle was designed to show how monophysites withstood Chalcedonian pressure; in a parallel fashion, much of John’s work was intended to demonstrate that monophysitism, not Chalcedonianism, was the true orthodoxy. 112 Such positions required champions, and the Jafnids provided highly-visible actors whose ultimate betrayal at the hands of perfidious Chalcedonians (or ‘polluted’ monophysites) only served to highlight their own moral superiority. Even in exile the Jafnids could be champions for John’s purposes: Maurice, after toppling al-Mundhir, demanded the latter’s adherence to Chalcedonian orthodoxy through his son al-Nuʿmān. Al-Nuʿmān refused, stolidly averring that the Arab tribes were already orthodox, adding that he would be killed if he wavered. John’s portrayal of alNuʿmān’s fortitude provided an unmistakeable message. 113 It Most recently, Shahîd, The peninsular Arab presence; Frend, Monophysite Movement 284, 297; cf. discussion in Fowden, Empire to Commonwealth 130. 111 Mich. Syr. Chron. 2.246–8, discussed in Fowden, Barbarian Plain 142–3; van Ginkel, Making history 357. 112 van Ginkel, John of Ephesus; cf. Millar, The evolution of the Syrian Orthodox Church (n. 3) on the views of the monophysites as ‘the orthodox’. 113 Joh. Eph. HE. 3.3.56. See discussion of this episode in Millar, The evolution of the Syrian Orthodox Church 55–6. 110

FROM MAVIA TO AL-MUNDHIR

195

should also be remembered that the process by which Chalcedonians and monophysites divided was neither linear nor quick, and despite periods of violent persecution, serious attempts were made throughout the sixth century to find a compromise. Politicallymotivated missionary work along the frontiers, during the reign of Justinian, was every so often carried out using imperiallysanctioned monophysites—including John of Ephesus—and such events served only to inject further opacity into the relationship between Chalcedonians and monophysites in the sixth century. 114 Al-Mundhir’s role as mediator, carried out in Constantinople, failed, but it does not obscure the fact that he had managed to protect his position as tribal chief, even while he engaged in high-level imperial ecclesiastical politics. Both he and his father, al-Ḥārith, had done this by a structured disengagement from the Empire. By supporting monophysitism to such a degree, they retained a certain distance from the centre of Chalcedonian politics, but profited from the ambiguities in the sixth-century relationship between orthodoxy and heresy. When they did engage with popular expressions of Christian piety, such as monasteries, they chose, on the basis of Haliarum and the codex from Palmyra, those of the monophysites. Since monophysitism also had yet to acquire a hierarchy of patriarchs and bishops that could match that of the Chalcedonians, supporting the monophysites also avoided overt identification with the Chalcedonian hierarchy, which itself was strongly linked to imperial power. It presumably also gave the Jafnids the opportunity to assume more of a leadership position than would have been possible amongst the Chalcedonians, where they would likely have been shut out of the ecclesiastical chain of command. The rural context was also important: all of the Christian sites connected with the Jafnids, including the otherwise-urban environment of al-Ruṣāfa, are in the country. The 137 signatures to letter º41 discussed above also indicate overwhelmingly rural locations. As Millar notes, the list ‘mentions not a single bishop, and refers at the most to one or two cities…it is a vivid and detailed reflection of the life of monasteries located in villages’. 115 While al-Ḥārith and al114 115

van Rompay, Society and community 250–1. Millar, Christian monasticism 113.

196

GREG FISHER

Mundhir arbitrated for the Empire in Constantinople, their prominence was maintained amongst the rural monophysite communities, far from metropolitan sees and the Chalcedonian hierarchy. Around these locations—monasteries, churches, martyria— elements of Arab Christian identity coalesced, insulated from the urban focus of the Chalcedonian hierarchy. 116 With the Jafnid connection to the cult of St. Sergius, lines were blurred further. The cult was popular amongst the Arabs, but not exclusively so, and it was available to both Chalcedonians and monophysites, who both celebrated the feast day of the saint on the same date. 117 At al-Ruṣāfa, the most public of all the connections to St. Sergius, the Jafnids based themselves in a building clearly designed for communication, patronage, and mediation free of overtly-Roman associations. The al-Mundhir building is not a qaṣr in the Umayyad sense, but is a remarkable avatar for those structures in its ability to transmit a particular image of power, framed in the terms of the state through the frescoes, inscription, decoration, and design, and in the terms of the tribe via its distancing from the city and through its function as an audience hall connecting desert tribes and state-sanctioned political power. 118 Al-Ruṣāfa lay on major communication routes in the region and because it held an honoured place in Syria as the home of the Sergius cult, it formed a nexus for the exchange of people, news, and intelligence. 119 At the same time, it was sufficiently distant from places such as Bostra, Damascus, Antioch, and Constantinople, to allow the Jafnids to do business on much their own terms. One is reminded of Garth Fowden, Rural converters, discussing Tall al-ʿUmayrī, Nitl, and other rural locations, in the context of ‘the pattern adopted by rural converters, in which Christian architecture played a significant role in establishing and affirming a Christian identity among the Arabs’. 117 Fowden, Barbarian Plain 156. 118 The communicative function of the Umayyad quṣūr is discussed by Genequand, Umayyad castles 3–6, and stressed by King, Settlement patterns in Islamic Jordan, and Grabar, Holod, and Trusdale, City in the Desert i, 156–65. 119 Fowden, Barbarian Plain 67. See on the site most recently Brands, City and territorium of Ruṣāfa in late antiquity. 116

FROM MAVIA TO AL-MUNDHIR

197

Fowden’s characterization of the site of al-Ḥumayma in southern Jordan, ‘at once remote from the intensely Umayyad atmosphere of Damascus and its satellite quṣūr, and well-placed, right by the road that linked the port of Ayla…by way of Maʿan to Damascus, for the gathering of news and gossip’. 120 In this light, it is also curious to note that the rural sites linked to the Jafnids were never far from the desert, where some manpower of Ghassān was probably still based. These locations were also out of the cities of the Empire, while within easy reach of major communication routes. Such places thus favoured communication between desert and settled lands, between imperial Chalcedonian centre, and monophysite periphery. We can note, too, parallels with forms of tribal leadership in the post-Roman west—that of the Ostrogoth Theoderic, for example: Theoderic took great care to maintain regular face-to-face contact with his chief military retainers and their retinues. This he achieved through a network of lavishly-decorated royal palaces…Theoderic and his court travelled between these palaces and throughout the Ostrogothic encampments holding sumptuous feasts, distributing donatives and rewards, and hearing grievances…Theoderic’s kingship was thus military in tone, itinerant in nature, and rooted in the traditions of face-to-face lordship [and] he further bound his Ostrogothic followers to him by supporting the Arian church. 121

The multitude of sites connected to al-Ḥārith and al-Mundhir which were designed around public meeting spaces, especially churches, public spaces par excellence, reinforces how the Jafnids retained their tribal functions. And so, whether as imperial arbitrator, present amongst church congregations at Nitl or Tall alʿUmayrī, or at al-Ruṣāfa receiving delegations during the pilgrimage to the shrine of St. Sergius, the tribal roles of mediation, with a stress on communication, and representing the tribe to the various agents of the state, and vice versa, were always at the centre of the Fowden, Qusayr ʿAmra 282; on Ḥumayma, see Oleson et al Preliminary report. 121 Sarris, Empires of Faith 353. 120

198

GREG FISHER

way that the Jafnids framed their relationship with the Christian Roman state. A Christian Arab identity? The Arabs castigated by Ps.-Nilus or Ammonius for being barbarous, savage, and cruel, could be redeemed by joining the commonwealth; this was part of the promise of conversion. Even so, some authors, such as Procopius, still saw even the Christian alḤārith through the range of ethnographic assumptions about barbarians—perfidious, treacherous, an echo of Ammianus’ famous statement that Arabs were good ‘neither as friends or enemies’. 122 Even Arab Christians, then, could still be barbarians, but what Christianity did do was open up a range of political, cultural, and social possibilities within the Empire that might otherwise have remained closed. Some of these have been explored here. Becoming Christian provided ambitious Arabs like al-Mundhir with the chance to play a role usually associated with late Roman élites— taking honorific titles such as patrikios, acting as benefactors, and working with high-ranking officials in the Empire. Thus, while it might not make Arabs or other barbarians into Romans, Christianity provided access to a collective membership, which made Arab converts culturally and politically compatible with other Christians in the Empire. How can the effect of Christianity on any developing Arab identity prior to the emergence of Islam be measured? It would be misleading, perhaps, to think that there was such as thing as panArab identity (in the modern sense) in the 500s; examinations of the term ‘Arab’ in antiquity reveal a perplexing range of meanings and understandings, and it is instead more likely that the identities of tribes, rather than the Arabs as a discrete people, dominated. 123 The poetry of the pre-Islamic era, and the stories of the ‘battle days Proc. BP. 1.17.48; Amm. 14.4.1. Macdonald, Arabs, Arabias, and Arabic, offers a comprehensive discussion of the different meanings applied to the word ‘Arab’. See generally for discussion of the problem of Arab identity: Fisher, Between Empires; Fisher, Kingdoms or dynasties?; Hoyland, Arab kings; Hoyland, Epigraphy and the emergence of Arab identity. 122 123

FROM MAVIA TO AL-MUNDHIR

199

of the Arabs’ focus on tribal identities, not ‘Arab’ ones. Al-Akhnas, for example, focused on the attributes of individual tribes, saying, in part of one poem: And Ghassān – their strength, all know, is other than in their kin – for them fight the legions and the squadrons of mighty Rome. 124

The Jafnids, closely linked with Ghassān, became the most ‘Roman’ of any of the Arabs in pre-Islamic late antiquity because of their close link with the Empire, which underwrote their power. The development of that power, from defeat at the hands of Romanus in 497/8, to categorization as summachoi in 561, and then to the gift of a ‘crown’ by Tiberius II to al-Mundhir on his ill-fated visit to Constantinople in 580, was accompanied by their association with the Christian religion. I have argued here that being Christian and being tribal were compatible, and that the Jafnids exploited their Christian affiliations to retain their tribal functions. What then, of others? The experience of the Jafnids seems to have had a knock-on effect of sorts elsewhere in Syria and Arabia. A series of martyria from the fifth and sixth centuries, all from rural areas on communications routes, and the appearance of Arabic language and script inscriptions from sixth-century Christian contexts, show how the experience of the Jafnids, while prominent, was not an isolated example. The martyria and inscriptions show again how Arab élites were behaving like late Roman community élites, but the (apparently) sudden appearance of Arabic, and its connection to these Christian monuments, raises questions. The Arabic inscriptions did not necessarily represent an attempt at ethnic differentiation on the part of their authors, but I would suggest that they really only make sense within the context of the emergence of Arab Christian élites within the Roman Empire (specifically the Jafnids), cultivating places of convergence and communication. 125 Christianity was an Lyall, Mufaḍḍalīyāt ii, 149–51. Fowden, ‘Rural converters’: ‘it is not surprising that the new script appeared in these contexts, since the accommodation of the Christian religion and exploitation of ecological knowledge [referring to the 124 125

200

GREG FISHER

overarching unifier which could at times cut across the different divides between individual tribes, as it also cut across other divides, such as the gulf between Roman and non-Roman, civilized, and barbarous. Membership in the polyglot Christian commonwealth permitted the maintenance of tribal identities even while being Christian required, at least on the surface, allegiance to Rome. The inscribed martyria in question are all from Syria and its environs. Of the five, the earliest offer only tenuous connections between Arabs and Christianity. The first is a martyrion of St. Thomas, from Anasartha in northern Syria, dating to 425/6, dedicated by a Mabia/Mavia. This has attracted attention mostly because of speculation that the Mavia here might be the same Mavia from the story described above, but it is almost certainly too late for that to be the case. 126 Furthermore, the Arabic name is not conclusive proof of ethnic identity. 127 The second martyrion, also from Anasartha and from the same period, offers a few more clues. It is dedicated by a lamprotatos (clarissimus), Silvanus, and includes a Greek text featuring a number of ‘Homeric echoes’; Feissel interprets the reference to Silvanus’ power in ‘Erembois’ as an analogy to the Odyssey, referring to Arabs. 128 The martyrion also honours Silvanus’ daughter, who is described as a wife of a phylarch, and since the latter term is very commonly used to describe Arab allies, questions have been asked about Silvanus’ identity. Was he a Roman officer, or perhaps even a phylarch himself? Unfortunately, despite strenuous attempts, the question cannot be answered with certainity. 129 provision of water at Jabal Says] were absolutely central to the Arabs’ maintenance of local and regional power.’ 126 Mouterde, Le limes de Chalcis, i, 194–5; Feissel, Les martyria d’Anasartha 205–9. 127 Cf. Liebeschuetz, Nomads, phylarchs and settlement 144. On problems of associating names and ethnicity, see Macdonald, Personal names in the Nabataean realm. 128 IGLS 2.297. Feissel, Les martyria d’Anasartha 213–14. 129 Mouterde, Le limes de Chalcis 193; Jalabert and Mouterde in IGLS 2, 168–70, see him as a dux Arabiae; Feissel, Les martyria d’Anasartha 213: ‘Titulaire d’une dignité romaine, Silvanos n’était cependant pas un fonc-

FROM MAVIA TO AL-MUNDHIR

201

The third (and weakest) example comes from al-Ramthāniye in the Golan. A martyrion from 377, sponsored by a certain Flavius Naʿamān, suggests, via the name (but with the same caveats) an Arab connection. Dauphin, who published the find, connected the site with Ghassān. While this cannot be certain (especially if Ghassān did not enter the Empire until the late fifth century) the martyrion is close to the generally-presumed location of al-Jābiya, and, more convincingly, analysis of the site has suggested that it might have offered a seasonal focus for nomads. 130 These three examples thus suggest, but do not prove, a connection between public expressions of devotion of the sort usually associated with élite benefactors, and Arabs. With the fourth example, a martyrion of St. John from Ḥarrān in Syria, containing an inscription in Arabic, this connection becomes far more certain. 131 The date of the Ḥarrān structure, 568, a year before al-Ḥārith died, is significant. Over a generation had passed since the monophysites had won Jacob and Theodore, and by now the Jafnids had provided highly visible models of Arab leadership, supported by no less an individual than the Emperor. It seems reasonable to think that this development had encouraged acceptance, and greater engagement of Arabs in similar activities. Thus, while models of Christian Arab leaders had existed in the fourth and fifth centuries, such as Mavia and Aspebetos, they never possessed the level of prominence, or, vitally, the depth of Roman support, which the Jafnid family enjoyed. This explains why it is at Ḥarrān that, after the ambiguities of the fifth century, a clear, public connection is tionnaire impérial: son autorité sur les Arabes (appelés en style homérique…), qui plus est une autorité perpétuelle, ne peut être que celle d’un chef indigène, autrement dit un phylarque arabe’. 130 Dauphin, Pèlerinage ghassanide; see the discussion in Fowden, Barbarian Plain 147–9. 131 Wadd. 2464 = IGLS 15.261. See Dussaud and Macler, Rapport sur une mission scientifique 726;see also Sartre, Trois études 177; Genequand, Some thoughts 80; Littmann, Osservazioni sulle iscrizioni 193–8; Robin, La réforme de l’écriture arabe 332–6. An updated translation of the Ḥarrān inscription can be found in Fisher, Arabs and Empires Ch. 6 (Greek: G. Bevan) and Ch. 7 (Arabic: M.C.A. Macdonald).

202

GREG FISHER

made between a non-Jafnid Arab leader and a Christian site, and made not just in Greek, but also in Arabic, attesting that a phylarch, Sharaḥīl, paid for the structure. Earlier Arabic inscriptions, such as that at al-Namāra, were written in the Arabic language but in scripts, such as Aramaic, usually associated with other languages. Here, at Ḥarrān, the inscription is in both the Arabic language and script; indeed, it is one of the very earliest extant examples of this phenomenon and one of only three from the preIslamic era. While one of those three is from a secular context— the Jabal Says inscription mentioned above—the other is also, like Ḥarrān, from a clearly Christian setting. A martyrion of St. Sergius, from Zabad in northern Syria, whose Arabic addition to the Greek and Syriac texts is a prayer for a number of people. Aside from everything else, it is another indication of the link between Arabs and the Sergius cult. 132 The Arabic part of the Ḥarrān inscription was done first, which, Hoyland argues, implies ‘that the Arabic was more important to the phylarch who commissioned the work, presumably in some way an important aspect of his identity’. 133 Addressing the same issue, Millar states that ‘there is no sufficient reason’ to connect the appearance of Arabic purely to ‘the power or influence of any one dynasty of ‘Arab’ phylarchs’. 134 Yet the emergence of the Jafnids as Christian Arab leaders, prominent in rural areas such as the location where the Ḥarrān martyrion is found (and for which the martyrion of Flavius Naʿamān had perhaps provided an avatar, as a rural node for Christian worship) is key to the puzzle. 135 As Sachau, Eine dreisprachige Inschrift aus Zebed; Sachau, Zur trilinguis Zebedea; Littmann, Osservazioni sulle iscrizioni 193–8. Recently: Hoyland, Epigraphy and the emergence of Arab identity 232; Fisher, Between Empires 144–53; Robin, La réforme de l’écriture arabe 336–8. An updated translation of the Zabad inscription with commentary can be found in Fisher, Arabs and Empires Ch. 6 (Greek and Syriac: G. Bevan) and Ch. 7 (Arabic: M.C.A. Macdonald). 133 Hoyland, Late Roman Provincia Arabia 133. 134 Millar, A Syriac codex 32. 135 Cf. Millar, Rome’s Arab allies 219: ‘the role of the major sixthcentury dynasty of phylarchs was thus compatible with the contempora132

FROM MAVIA TO AL-MUNDHIR

203

Jafnid power rested on Roman support, so we might imagine that other phylarchs, such as Sharaḥīl, took at least some of their cues from the Jafnids. The Ḥarrān example is, then, as Hoyland suggests, a statement of identity, but perhaps not in the way we might expect. While it may indicate a progressive interest in being linguistically different—Greek was the usual prestige language for public inscriptions of this sort—it also reflects a crucially important trend: the increasing participation of Arabs in a monotheist commonwealth. 136

CONCLUSION

The Christianization of the Arabs between the third and sixth centuries changed the status of the tribes and the chiefs who led them. It gave them access to the oikoumenē, and provided previously unparalleled political opportunities within the Roman political and ecclesiastical hierarchy. Christianization posed obstacles, such as the threat of stratification, but I have stressed here how the actions of the Jafnids, in particular, mitigated those risks by framing their participation as Roman allies in tribal terms: as mediators, conduits, and providers of benefits to the tribe. Some, on the pattern of an Aspebetos, probably decided to make their careers in the Roman hierarchy and leave the tribal milieu, but it is clear from the experience of the Jafnids that tribal identities did not disappear with Christianity. While the Sasanian-allied Naṣrids have not been discussed in detail here, it is noteworthy that while they avoided too open an affiliation with Christianity, this never prevented the flourishing of the religion at al-Ḥīra, their base. 137 It was not just that tribal leaders found ways to fit Christianity into their social strucneous activity of (apparently) lesser figures.’ 136 Cf. Fowden, Rural converters, commenting that ‘Arab adoption of church-building activity as a means of identification with the Christian faith as attested in our growing epigraphical evidence shows a tangible degree of integration.’ 137 Trimingham, Christianity Among the Arabs 189–90; Fisher, Between Empires Ch. 2; most recently, Toral-Niehoff, The ʿIbād of al-Ḥīra; ToralNiehoff, Al-Ḥīra; Fisher, Arabs and Empires Ch. 6 (P. Wood) and Ch. 8 (H. Munt and I. Toral-Niehoff).

204

GREG FISHER

tures, but also that Christianity was compatible with the tribe. 138 We have seen here how tribal leaders arbitrated Christian disputes, but that the reverse could also be true is underscored in a vignette from Theodoret’s Life of Symeon Stylite. Symeon arbitrated a quarrel between Arabs in a most unholy way, ‘hurling threats at them from above and calling them dogs’, but it is clear that his prestige as a holy man was interpreted as of equivalent rank to the prestige and influence of a tribal sheikh. 139 Likewise, it was the cachet of St. Sergius which underpinned the authority of al-Mundhir at alRuṣāfa. While Christianization could bring stratification, it was not inevitable. Indeed, it was as the engagement of the Jafnid family with Christianity grew more prominent that Jafnid power was strengthened. This suggests that the growing Christian ‘identity’ of the Jafnids actually strengthened the tribal power of the leadership. Like other successful tribal leaders throughout history, the Jafnids, until 582, keenly understood the ‘ambiguities and dilemmas’ of their relationship with the Empire, and ‘balanced their conflicting aspects’ accordingly. 140 While al-Mundhir did not become a holy man, he assumed some of the prestige and esteem accorded to such figures by having access to bishops, priests, and St. Sergius. By displaying these connections prominently, usually at fixed points designed for communication and convergence, the Jafnids ‘reaffirmed and maintained’ their position as community leaders and tribal chiefs. 141 When the followers of The Prophet invaded the Roman Near East in the seventh century, they were not the first examples of Arabs in Syria and Jordan under powerful leaders, adhering to a potent, universalizing religion which could unify and bind together the different tribes. The spread of Christianity among the Arabs anticipated, at least in part, what was to come. In what we now know retrospectively as the ‘eve of the Muslim invasions’, the JafCf. Debié, La Christianisation des Arabes 19, on the adaptability of Christian ceremony away from ‘settled’ church environments, geared towards nomads. 139 Theod. V. Sim. 15. 140 Paraphrased from Beck, Tribes and the state 218. 141 Fowden, ‘Rural converters.’ 138

FROM MAVIA TO AL-MUNDHIR

205

nids wielded considerable power by cultivating the interface between orthodoxy and heresy, centre and periphery, desert and settled lands, and tribe and state, all connected by the cultural bond provided by Christianity. One cannot help but speculate that the followers of The Prophet would not have been quite as successful, in such a short space of time, had they not already been shown the way by powerful tribal leaders, following, and exploiting, a religion of the Book.

BIBLIOGRAPHY ACO = Schwartz, E. (ed.), Acta conciliorum oecumenicorum, Strasbourg 1914. AÉ = Académie des inscriptions et belles lettres, Année Épigraphique, Paris 1888–. CIS = Corpus inscriptionum semiticarum, Paris 1881–. IGLS = Jalabert, L., Mouterde, R. et al (eds.), Inscriptions grecques et latines de la Syrie, Paris 1929–. Nd. Or = Seeck, O. (ed.), Notitia Dignitatum accedunt Notitia Urbis Constantinopolitanae et Laterculi Provinciarum, Berlin 1876. PLRE = Jones, A.H.M., Martindale, J.R., and Morris, J. (eds.), The Prosopography of the Later Roman Empire, 3 vols., Cambridge 1971–92. PO = Patrologia Orientalis, Paris 1907–66; Turnhout 1968–. SEG = Supplementum epigraphicum graecum, Leiden/Amsterdam, 1923–. Wadd. = Inscriptions grecques et latines recueillies en Grèce et en Asie Mineure, eds. P. Le Bas and W.H. Waddington, 3 vols., Paris, 1853–70. Editions and collections of primary sources Agathias, Agathiae Myrinaei Historiarum libri quinque, ed. R. Keydell, Berlin 1967. Aḥūdemmeh = Histoires d’Aḥoudemmeh et de Marouta, trans. F. Nau, Paris 1903. Al-Akhnas = Lyall, C.J. (ed.), The Mufaḍḍalīyāt. An Anthology of Ancient Arabian Odes Compiled by Al-Mufaḍḍal son of Muḥammad, ac-

206

GREG FISHER

cording to the recension and with the commentary of Abū Muḥammad Al-Qāsim ibn Muḥammad al-Anbārī, 3 vols., Oxford 1918. Ammianus Marcellinus. Ammiani Marcellini Rerum gestarum libri qui supersunt, ed. V. Gardthausen, 2 vols., Leipzig 1874–5. Ammonius = History and Hagiography from the Late Antique Sinai, trans. D.F. Caner, with S. Brock, R.M. Price, and K. van Bladel, Liverpool 2010. Antonius = Das Leben des Heiligen Symeon Stylites, ed. H. Lietzmann, Leipzig 1908. Chabot, J.-B., Documenta ad origines Monophysitarum illustrandas, Leuven 1952. Chronicle of 1234, Chronicon anonymum ad annum Christi 1234 pertinens, eds. and trans. J.-B. Chabot and A. Abouna, 2 vols., Paris and Leuven 1916–74. Cyril of Scythopolis, Kyrillos von Skythopolis, ed. E. Schwartz, Leipzig 1939. Cyril of Scythopolis, The Lives of the Monks of Palestine, trans. R.M. Price with introd. and notes by J. Binns, Kalamazoo 1991. Evagrius, Kirchengeschichte, eds. and. trans. A. Hübner and J. Bidez, 2 vols., Turnhout 2007. Isaac of Antioch, S. Isaaci Antiocheni Opera Omnia, ed. and trans. G. Bickell, Giessen 1873. Jerome, Trois vies de moines. Paul, Malchus, Hilarion, trans. P. LeClerce with introd. by E. Morales, Paris 2007. John of Ephesus, Iohannis Ephesini Historiae ecclesiasticae pars tertia, ed. and trans. E.W. Brooks, 2 vols., Paris, 1935–36. John of Ephesus, Lives of the Eastern Saints, ed. and trans. E.W. Brooks, Paris, 1923–25. Malalas, Chronographia, ed. J. Thurn, Berlin 2000. Malchus = The Fragmentary Classicising Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus, and Malchus, ed. and trans. R.C. Blockley, 2 vols., Liverpool 1981–83. Menander, The History of Menander the Guardsman, ed. and trans. R.C. Blockley, Liverpool 1985. Michael the Syrian, Chronique, ed. and trans. J.-B. Chabot, 4 vols., Paris 1899–1910.

FROM MAVIA TO AL-MUNDHIR

207

(Ps.-) Nilus = History and Hagiography from the Late Antique Sinai, trans. D.F. Caner, with S. Brock, R.M. Price, and K. van Bladel, Liverpool 2010. Photius, Bibliothèque, ed. and trans. R. Henry, 9 vols., Paris 1959–91. Priscus = The Fragmentary Classicising Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus, and Malchus, ed. and trans. R.C. Blockley, 2 vols., Liverpool 1981–83. Procopius, Procopii Caesariensis opera omnia, ed. J. Haury, 4 vols., Leipzig 1962–64. Severus of Antioch, Les Homiliae Cathedrales de Sévère d’Antioche, trans. R. Duval, Paris 1906. Socrates Scholasticus, Kirchengeschichte, ed. G.C. Hansen, Berlin 2011. Sozomen, Kirchengeschichte, ed. J. Bidez, rev. G.C. Hansen, Berlin 1995. Theodoret, Histoire des moines de Syrie: Histoire Philothée, eds. P. Canivet and A. Leroy-Molinghen, 2 vols., Paris, 1977–79. Theodoret, Historia Ecclesiastica, ed. L. Parmentier, rev. G.C. Hansen, Berlin 1998. Theophanes, The Chronicle of Theophanes Confessor: Byzantine and Near Eastern history, A.D. 284–813, trans. C. Mango and R. Scott, Oxford 1997. Theophanes, Chronographia, ed. C. de Boor, 2 vols., Hildesheim 1963–54. Theophylact Simocatta, The History of Theophylact Simocatta, ed. and trans. M. Whitby and M. Whitby, Oxford 1986. (Ps.-) Zachariah. The Chronicle of Pseudo-Zachariah Rhetor. Church and War in Late Antiquity, ed. G. Greatrex, trans. R. Phenix and C. Horn, with introd. S. Brock and W. Witakowski, Liverpool 2011. Secondary sources Adelman, J. and S. Aron, From borderlands to borders: empires, nation-states, and the peoples in between in North American history, in AHR 104/3 (1999), 814–41. Allen, P., The definition and enforcement of orthodoxy, in CAH, xiv, 811–34.

208

GREG FISHER

Allen, P., Evagrius Scholasticus, The Church Historian, Leuven 1981. Athanassiadi, P. and M. Frede (eds.), Pagan Monotheism in Late Antiquity, Oxford 1999. Barber, B., Social Stratification. A Comparative Analysis of Structure and Process, New York 1957. Beaucamp J., Le rôle de Byzance en mer Rouge sous le règne de Justin: mythe ou réalité? in J. Beaucamp, F. BriquelChatonnet, and C.J. Robin (eds.), Juifs et Chrétiens en Arabie aux Ve et VIe siècles. Regards croisés sur les sources, Paris 2010, 197– 220. Beck L., The Qashqa’i of Iran, New Haven 1986. Beck, L., Tribes and the state in nineteenth- and twentieth-century Iran, in P.S. Khoury and J. Kostiner (eds.), Tribes and State Formation in the Middle East, Berkeley 1990, 185–225. Bevan, G., G. Fisher and D. Genequand, The Late Antique Church at Tall al-ʿUmayrī East: New Evidence For the Jafnid Family And the Cult of St. Sergius in Northern Jordan, in BASOR, in press. Blockley, R., East Roman Foreign Policy. Formation and Conduct from Diocletian to Anastasius, Leeds 1992. Bordreuil, P., A. Desreumaux, C. Robin and J. Teixidor, 205. Linteau inscrit: AO 4083, in C. Robin and Y. Calvet (eds.), Arabie heureuse, Arabie déserte. Les antiquités arabiques du Musée du Louvre, Paris 1997, 265–9. Bowersock, G.W., Mavia, Queen of the Saracens, in W. Eck, H. Galsterer, and H. Wolff (eds.), Studien zur antiken Sozialgeschichte. Festschrift Friedrich Vittinghoff, Vienna 1980, 477– 95. Bowersock, G.W., Polytheism and monotheism in Arabia and the Three Palestines, in DOP 51 (1997), 1–10. Bowersock, G.W., The Throne of Adulis. Red Sea Wars on the Eve of Islam, Oxford 2013. Braatz, T., Surviving Conquest. A History of the Yavapai Peoples, Lincoln NE. 2003. Brands, G., City and territorium of Ruṣāfa in late antiquity and early Islam, in A. Borrut, M. Debié, A. Papaconstantinou, D. Pieri, and J.-P. Sodini, (eds.), Le Proche-Orient de Justinien aux Ab-

FROM MAVIA TO AL-MUNDHIR

209

bassides. Peuplement et dynamiques spatiales, Turnhout 2011, 59–76. Brands, G., Die sogenannte Audienzsaal des al-Mundir in Reṣāfa, in DaM 10 (1998), 237–41. Brown, P., Power and Persuasion in Late Antiquity. Towards a Christian Empire, Madison 1992. Canepa, M., The Two Eyes of the Earth. Art and Ritual of Kingship Between Rome and Sasanian Iran, Berkeley 2009. Cotton, H., Some aspects of the Roman administration of Judaea/Syria-Palestine, in W. Eck and E. Müller-Luckner (eds.), Lokale Autonomie und römische Ordnungsmacht in den kaiserzeitlichen Provinzen vom 1. bis 3. Jahrhundert, Munich 1999, 75–91. van Dam, R., Becoming Christian. The Conversion of Roman Cappadocia, Philadelphia 2003. Daryaee, T., Sasanian Persia. The Rise and Fall of an Empire, London 2009. Dauphin C., Pèlerinage ghassanide au sanctuaire byzantin de Saint Jean-Baptiste à Er-Ramthaniyye en Gaulantide, in E. Dassman and J. Engeman (eds.), Akten des XII. Internationalen Kongresses Für Christliche Archäologie, Bonn 22.-28. September 1991, ii, Münster 1995, 667–73. Debié, M., ‘La Christianisation des Arabes nomades de la mer rouge à l’Euphrate’, in Dossiers d’Archéologie et Sciences des Origines 309 (2005/6), 16–23. Decret F., Les conséquences sur le Christianisme en Perse de l’affrontement des empires romain et sassanide, in ReAug 14 (1979), 91–152. Dignas B. and E. Winter, Rome and Persia in Late Antiquity. Neighbours and Rivals, Cambridge 2007. Drake, H.A., The impact of Constantine on Christianity, in N. Lenski (ed.), The Cambridge Companion to the Age of Constantine, Cambridge 2006, 111–36. Dussaud, R. and F. Macler, Rapport sur une mission scientifique dans les régions désertiques de la Syrie moyenne, in Nouvelles archives des missions scientifiques et littéraires 10 (1902), 411–744. Feissel, D., Les martyria d’Anasartha, in V. Déroche (ed.), Mélanges Gilbert Dagron, Paris 2002, 201–20.

210

GREG FISHER

Fisher, G. (ed.), Arabs and Empires Before Islam, Oxford in press. Fisher, G. Between Empires. Arabs, Romans, and Sasanians in Late Antiquity, Oxford 2011. Fisher, G. Kingdoms or dynasties? Arabs, history, and identity before Islam, in JLA 4/2 (2011), 245–67. Fowden, E.K., An Arab building at al-Rusāfa-Sergiopolis, in DaM 12 (2000), 303–27. Fowden, E.K., The Barbarian Plain: Saint Sergius between Rome and Iran, Berkeley 1999. Fowden, E.K. Rural converters among the Arabs. In D. Schwartz, N. McLynn, and A. Papaconstantinou A. (eds.) Conversion in Late Antiquity: Christianity, Islam, and Beyond, Farnham, forthcoming. Fowden, G., Empire to Commonwealth: Consequences of Monotheism in Late Antiquity, Princeton 1993. Fowden, G., Qusayr ʿAmra. Art and the Umayyad Elite in Late Antique Syria, Berkeley 2004. Fowden, G., Varieties of religious community, in G. Bowersock, P. Brown and O. Grabar (eds.), Interpreting Late Antiquity. Essays on the Postclassical World, Cambridge Mass. 2001, 82–106. Fowden, G. and E.K. Fowden, Studies on Hellenism, Christianity and the Umayyads, Paris 2004. Frend, W. H. C., The Rise of the Monophysite Movement. Chapters in the History of the Church in the Fifth and Sixth Centuries, Cambridge 1972. Gajda I., Ḥuǧr b. ʿAmr roi de Kinda et l’établissement de la domination Ḥimyarite en Arabie central, in PSAS 26 (1996), 65–73. Gatier, P.-L., Les Jafnides dans l’épigraphie grecque au VIe siècle, in D. Genequand and C.J. Robin (eds.), Les Jafnides, des rois arabes au service de Byzance (vie siècle de l’ère chrétienne). Paris 2014, 185–214. Genequand, D., Some thoughts on Qasr al-Hayr al-Gharbi, its dam, its monastery and the Ghassanids, in Levant 38 (2006), 63–84. Genequand, D., Umayyad castles: the shift from late antique mili-

FROM MAVIA TO AL-MUNDHIR

211

tary architecture to early Islamic palatial building, in H. Kennedy (ed.), Muslim Military Architecture in Greater Syria: From the Coming of Islam to the Ottoman Period, Leiden 2006, 3–25. van Ginkel, J., John of Ephesus. A Monophysite Historian in SixthCentury Byzantium, D. Litt. Thesis, Rijksuniversiteit Groningen 1995. van Ginkel, J., Making history: Michael the Syrian and his sixth century sources, in R. Lavenant (ed.), Symposium Syriacum VII: Uppsala University, Department of Asian and African Languages, 11– 14 August 1996, Rome 1998, 351–8. Grabar O., R. Holod and W. Trusdale, City in the Desert. Qasr alHayr East, 2 vols., Cambridge Mass. 1978. Graf, D., The Saracens and the defense of the Arabian frontier, in BASOR 229 (1978), 1–26. Greatrex, G., Byzantium and the East in the sixth century, in M. Maas (ed.), The Cambridge Companion to the Age of Justinian, Cambridge 2005, 477–510. Greatrex, G., Rome and Persia at War, 502–532, Leeds 1998. Grouchevoy, A.G., Trois ‘niveaux’ de phylarques. Étude terminologique sur les relations de Rome et de Byzance avec les Arabes avant l’Islam, in Syria 72 (1995), 105–31. von Grunebaum, G.E., The nature of Arab unity before Islam, in Arabica 10/1 (1963), 5–23. Haarer, F. K., Anastasius I. Politics and Empire in the Late Roman World, Cambridge 2006. Haas, Ch., Mountain Constantines: the Christianization of Aksum and Iberia, in JLA 1/1 (2008), 101–26. Hainthaler, T., Christliche Araber vor dem Islam, Leuven 2007. Halsall, G., Barbarian Migrations and the Roman West, 376–568, Cambridge 2007. Heather, P., Empires and Barbarians: Migration, Development, and the Birth of Europe, Oxford 2010. Heather, P., Foedera and foederati of the fourth century, in W. Pohl (ed.), Kingdoms of the Empire. The Integration of Barbarians in Late Antiquity, Leiden 1997, 57–74. Heather, P., The late Roman art of client management: imperial

212

GREG FISHER

defence in the fourth century west, in W. Pohl, I. Wood, and H. Reimitz (eds.), The Transformation of Frontiers: from Late Antiquity to the Carolingians, Leiden 2001, 15–68. Hefner, R.W. (ed.), Conversion to Christianity. Historical and Anthropological Perspectives on a Great Transformation. Berkeley 1993. Honigman, E., Évêques et évêchés monophysites d’Asie antérieure au VIe siècle, Leuven 1951. Howard-Johnston, J., Witnesses to a World Crisis. Historians and Histories of the Middle East in the Seventh Century, Oxford 2011. Hoyland, R.G., Arabia and the Arabs. From the Bronze Age to the Coming of Islam, London 2001. Hoyland, R.G., Arab kings, Arab tribes, Arabic texts and the beginnings of (Muslim) Arab historical memory in late Roman inscriptions, in H. Cotton, R. Hoyland, J. Price, and D. Wasserstein (eds.), From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East, Cambridge 2009, 377–400. Hoyland, R.G., Epigraphy and the emergence of Arab identity, in P. Sijpesteijn, L. Sundelin, S. Tovar, and A. Zomeño (eds.), From Andalusia to Khurasan: Documents from the Medieval Islamic World, Leiden 2007, 219–42. Hoyland, R.G., Insider and outsider sources: historiographical reflections on pre-Islamic Arabia, in J. Dijkstra and G. Fisher (eds.), Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers, Leuven 2014, 267–80. Hoyland, R.G., Late Roman Provincia Arabia, Monophysite monks and Arab tribes: a problem of centre and periphery, in SemClas 2 (2009), 117–39. Irons, W., Political stratification among pastoral nomads, in Équipe Ecologie et Anthropologie des Société Pastorale (ed.), Pastoral Production and Society, Cambridge 1979, 361–74. James, E., Europe’s Barbarians, AD 200–600, Harlow 2009. Kaimio, M. P., Petra inv. 83: a settlement of dispute, in I. Andorlini, G. Bastianini, M. Manfredi, and G. Menci (eds)., Atti del XXII Congresso Internazionale di Papirologia. Firenze, 23–29 agosto 1998, ii, Florence 2001, 719–24. Kennedy, D., The Roman Army in Jordan, London 2004. King, G.R.D., Settlement patterns in Islamic Jordan: the Umayyads

FROM MAVIA TO AL-MUNDHIR

213

and their use of the land, in A. Hadidi (ed.), Studies in the History and Archaeology of Jordan, iv, Amman 1992, 369–75. Lamy, Th.-J., Profession de foi adressé par les Abbés d’Arabie à Jacques Baradée’, in Actes du XIe Congrès des Orientalistes, section sémitique, Paris 1898, 117–37. Lancaster, W. and F. Lancaster, Concepts of leadership in bedouin society, in J. Haldon and L. Conrad (eds.), The Byzantine and Early Islamic Near East, vi: Elites Old and New in the Byzantine and Early Islamic Near East, Princeton 2004, 29–61. Lancaster, W. and F. Lancaster, Tribal formations in the Arabian peninsula, in Arab. Arch. Epig. 3 (1992), 145–72. Lapidus, I.M., Tribes and state formation in Islamic history, in P.S. Khoury and J. Kostiner (eds.), Tribes and State Formation in the Middle East, Berkeley 1990, 25–47. Lenski, N., Captivity and slavery among the Saracens in late antiquity (CA. 250–630 CE), in AnTard 19 (2011), 237–66. Lewin, A., ʿAmr ibn ʿAdī, Mavia, the phylarchs and the late Roman army: peace and war in the Near East, in A. Lewin and P. Pellegrini (eds.), The Late Roman Army in the Near East from Diocletian to the Arab Conquest. Proceedings of a Colloquium held at Potenza, Acerenza, and Matera, Italy (May 2005), Oxford 2007, 243– 62. Liebeschuetz, J.H.W.G., Nomads, phylarchs and settlement in Syria and Palestine, in A.S. Lewin and P. Pellegrini (eds.), Settlements and Demography in the Near East in Late Antiquity. Proceedings of the Colloquium, Matera 27–29 October 2005, Pisa 2007, 131–46. Lim, R., Christian triumph and controversy’, in in G. Bowersock, P. Brown and O. Grabar (eds.), Interpreting Late Antiquity. Essays on the Postclassical World, Cambridge Mass. 2001, 196–218. Littmann, E., Osservazioni sulle iscrizioni di Harran e di Zebed, in Revista degli studi orientali 4 (1911), 193–8. Maas, M., “Delivered from their ancient customs”: Christianity and the question of cultural change in early Byzantine ethnography’, in K. Mills and A. Grafton (eds.), Conversion in Late Antiquity And The Early Middle Ages, Rochester, NY 2003, 152–88. Macdonald, M.C.A., A note on new readings in line 1 of the Old Arabic graffito at Jabal Says, in SemClas 2 (2009), 223–5.

214

GREG FISHER

Macdonald, M.C.A., Arabs, Arabias, and Arabic before Late Antiquity, in Topoi 16/1 (2009), 277–332. Macdonald, M.C.A., Old Arabic (Epigraphic), in K. Versteegh and M. Eid (eds.), Encyclopedia of Arabic Language and Linguistics, iii, Leiden 2006–8, 464–77. Macdonald, M.C.A., On Saracens, the Rawwāfah inscription and the Roman army = Article VIII in Macdonald, M.C.A., Literacy and Identity in Pre-Islamic Arabia. Farnham 2009. Macdonald, M.C.A., Personal names in the Nabataean realm: a review article, in JSS 44/2 (1999), 251–89. Macdonald, M.C.A., Quelques réflexions sur les Saracènes, l’inscription de Rawwāfa et l’armée romaine, in H. Lozachmeur (ed.), Présence arabe dans le Croissant Fertile avant l’Hégire : actes de la table ronde internationale organisée par l’unité de recherche associée 1062 du CNRS, Études sémitiques, au Collège de France, le 13 novembre 1993, Paris 1995, 93–201. MacMullen, R., Christianizing the Roman Empire (A.D. 100–400), New Haven 1984. Matthews, V.H., Pastoral Nomadism in the Mari Kingdom (ca. 1830– 1760 BC), Cambridge Mass. 1978. Mayerson, P., Mavia, Queen of the Saracens – a cautionary note, in IEJ 30 (1980), 123–31. Mayerson, P., The island of Iotabe in the Byzantine sources, in BASOR 287 (1992), 1–4. Menze, V., Justinian and the Making of the Syrian Orthodox Church, Oxford 2008. Milik, J., Inscriptions grecques et nabatéenes de Rawwafah, in University of London. Bulletin of the Institute of Archaeology 10 (1971), 54–58. Millar, F., Christian monasticism in Roman Arabia at the birth of Mahomet, in SemClas 2 (2009), 97–115. Millar, F., Rome’s Arab allies in late antiquity. Conceptions and representations from within the frontiers of the empire, in H. Börm and J. Wiesehöfer (eds), Commutatio et Contentio. Studies in the Late Roman, Sasanian, and Early Islamic Near East, Düsseldorf 2010, 199–226. Millar, F., A Syriac codex from near Palmyra and the ‘Ghassanid’

FROM MAVIA TO AL-MUNDHIR

215

Abokarib, in Hugoye 16/1 (2013), 15–35. Millar, F., The evolution of the Syrian Orthodox Church in the preIslamic period: from Greek to Syriac, in JECS 21/1 (2013), 43–92. Mitchell, S. and P. van Nuffelen (eds.), One God: Pagan Monotheism in the Roman Empire, Cambridge, 2010. Modéran, Y., Les Maures et l’Afrique Romaine (IVe-VIIe siècle), Rome 2003. Mouterde, R., Le limes de Chalcis: organisation de la steppe en haute Syrie romaine: documents aériens et épigraphiques, 2 vols., Paris 1945. Nock, D., Conversion: The Old and the New in Religion from Alexander the Great to Augustine of Hippo, Oxford 1933. Oleson, J.P., K. ʿAmr, R. Foote, J. Logan, B. Reeves and R. Schick, Preliminary report of the al-Ḥumayma Excavation Project, 1995, 1996, 1998, in ADAJ 43 (1999), 411–50. Parker, S.T., Romans and Saracens. A History of the Arabian Frontier, Winona Lake 1986. Piccirillo, M., The church of Saint Sergius at Nitl. A centre of the Christian Arabs in the steppe at the gates of Madaba, in Liber Annuus 51 (2001), 267–84. Power, T., The material culture and economic rationale of Saracen settlement in the eastern desert of Egypt, in A. Borrut, M. Debié, A. Papaconstantinou, D. Pieri, and J.-P. Sodini (eds.), Le Proche-Orient de Justinien aux Abbassides. Peuplement et dynamiques spatiales, Turnhout 2011, 331–44. Radding, C., Wandering Peoples: Colonialism, Ethnic Spaces and Ecological Frontiers in Northwestern Mexico, 1700–1850, Durham N.C. 1997. al-Rasheed, M., The process of chiefdom-formation as a function of nomadic/sedentary interaction. The case of the Shammar nomads of North Arabia, in Cambridge Anthropology 12/3 (1987), 32–40. Robin, C.J., La réforme de l’écriture arabe à l’époque du califat médinois, in Mélanges de l’Université Saint-Joseph 56 (2006), 319–64. Robin, C.J., Le royaume Ḥujride, dit , entre Ḥimyar et Byzance’, in CRAI (1996), 665–714.

216

GREG FISHER

Robin, C.J., Les Arabes de Ḥimyar, des et des Perses (IIIe-VIe siècles de l’ère chrétienne), in SemClas 1 (2008), 167–208. Robin, C.J. and M. Gorea, Un réexamen de l’inscription arabe préislamique du Ğabal Usays (528–529 É. Chr.), in Arabica 49 (2002), 503–10. van Roey, A. and P. Allen (eds.), Monophysite Texts of the Sixth Century, Leuven 1994. van Rompay, L., Society and community in the Christian east, in M. Maas (ed.), The Cambridge Companion to the Age of Justinian, Cambridge 2005, 239–66. Rubin, Z., Diplomacy and war in the relations between Byzantium and the Sassanids in the fifth century AD, in P. Freeman and D. Kennedy (eds.), The Defence of the Roman and Byzantine East: Proceedings of a Colloquium held at the University of Sheffield in April 1986, ii, Oxford 1986, 677–95. Sachau, E. C., Eine dreisprachige Inschrift aus Zebed, in Monatsberichte der Königlichen Preussische Akadamie der Wissenschaften zu Berlin (1881), 169–90. Sachau, E.C., Zur trilinguis Zebedea, in ZDMG 36 (1882), 345–52. Salzman, P., The meeting of the twain: tribe and state, in J. Dijkstra and G. Fisher (eds.), Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers, Leuven 2014, 83–90. Salzman, P., Pastoralists. Equality, Hierarchy, and the State, Boulder 2004. Salzman, P., Tribal chiefs as middlemen: the politics of encapsulation in the Middle East’, in Anthropological Quarterly 47/2 (1974), 203–10. Sarris, P., Empires of Faith. The Fall of Rome and the Rise of Islam, 500– 700, Oxford 2011. Sartre, M., Deux phylarques arabes dans l’Arabie byzantine, in Le Muséon 106 (1993), 145–54. Sartre, M., The Middle East Under Roman Rule, trans. C. Porter and E. Rawlings, Cambridge Mass. 2005. Sartre, M., Trois études sur l’Arabie romaine et byzantine, Brussels 1982.

FROM MAVIA TO AL-MUNDHIR

217

Sauvaget, J., Les Ghassanides et Sergiopolis, in Byzantion 14 (1939), 115–30. Schmitt, O., Mavia, die Königin der Sarazenen, in T. Herzog and W. Holzwarth (eds.), Nomaden und Sesshafte – Fragen, Methoden, Ergebnisse, Halle-Wittenberg 2003, 163–79. Segal, J.B., Arabs in Syriac literature before the rise of Islam, in JSAI 4 (1984), 89–124. Shahîd, I., Byzantium and the Arabs in the Sixth Century, 2 vols., Washington 1995–2010. Shahîd, I., Ghassānid religious architecture, in P. Canivet and J.-P. Rey-Coquais (eds.), Mémorial Monseigneur Joseph Nasrallah, Damascus 2006, 115–38. Shahîd, I., The peninsular Arab presence in Oriens (Bilād Al-Shām) in Byzantine and Umayyad times, in K.G. Holum and H. Lapin (eds.), Shaping the Middle East. Jews, Christians, and Muslims in an Age of Transition, 400–800 C.E., Bethesda 2011, 173–80. Shaked, S., Religion in the late Sasanian period: Eran, Aneran, and other religious designations, in V.S. Curtis and S. Stewart (eds.), The Idea of Iran, iii: The Sasanian Era, London 2008, 103– 118. al-Shami, A., Archaeological Excavation Project at Tall al-ʿUmayrī East—first season 2009 [in Arabic], in ADAJ 54 (2010), 35– 42. Shaw, B.D., “Eaters of flesh, drinkers of milk”, The ancient Mediterranean ideology of the pastoral nomad.in AncSoc 13/14 (1982), 5–31. Smith, S., Events in Arabia in the 6th century A.D., in BSOAS 16/3 (1954), 425–68. Szuchman, J., Integrating approaches to nomads, tribes, and the state in the ancient Near East, in J. Szuchman (ed.), Nomads, Tribes, and the State, Chicago 2009, 1–14. al-Talhi, D. and M. al-Daire, Roman presence in the desert: a new inscription from Hegra, in Chiron 35 (2005), 205–17. Tapper, R., Anthropologists, historians, and tribespeople on tribe and state formation in the Middle East, in P.S. Khoury and J. Kostiner (eds.), Tribes and State Formation in the Middle East, Berkeley 1990, 48–73.

218

GREG FISHER

Tapper, R., Introduction, in R. Tapper (ed.), The Conflict of Tribe and State in Iran and Afghanistan, London 1983, 1–82. Toral-Niehoff, I., Al-Ḥīra. Eine arabische Kulturmetropole im spätantiken Kontext, Leiden 2014. Toral-Niehoff, I., The ʿIbād of al-Ḥīra: an Arab Christian community in late antique Iraq, in A. Neuwirth, M. Marx, and N. Sinai (eds.), The Qur’an in Context – Entangled Histories and Textual Palimpsests, Leiden 2010, 323–47. Trimingham, J.S., Christianity Among the Arabs in Pre-Islamic Times, New York 1979. al-ʿUsh, M. A. el-Faraj, Unpublished Arabic texts in Jabal Usais’ [in Arabic], in Al-Abhath 17/3 (1964), 227–316. Walker, J., From Nisibis to Xiʾan: the church of the east in late antique Eurasia, in S.F. Johnson (ed.), The Oxford Handbook of Late Antiquity, Oxford 2012, 994–1052. Walker, J., The Legend of Mar Qardagh. Narrative and Christian Heroism in Late Antique Iraq, Berkeley 2006. Ward, W., The Mirage of the Saracen. Christians and Nomads in the Sinai Peninsula in Late Antiquity, Berkeley 2014. Weber, D.J., The Spanish Frontier in North America, New Haven 1992. Whittow, M., Rome and the Jafnids: writing the history of a 6th-c. tribal dynasty, in J. Humphrey (ed.), The Roman and Byzantine Near East: Some Recent Archaeological Research, iii, Portsmouth R.I. 1995–2002, 207–24.

INCLUSION AND SECLUSION IN LATE ANTIQUE AND EARLY MEDIEVAL SETTLEMENT STRATEGIES IN SYRIA AND PALESTINE MATTIA GUIDETTI LATE ANTIQUE ARAB SETTLEMENTS IN SYRIA AND PALESTINE

Since the fourth century onwards, large groups of Arabs moved from the Arabian Peninsula to the north, then settling in Mesopotamia, Syria, and along the Mediterranean shore. These migrations obviously changed the identity of both the newcomers and the local people, but also established a series of remarkable patterns in the settlement policy, which will appear especially noteworthy when we analyze the outcome of the seventh-century conquest of Syria and Palestine and the subsequent Islamic rule. The late antique Arab groups, who were often united in bigger tribal confederations, came out of Arabia to live alongside the Semitic groups that had already settled in the area, like the Nabataeans. The latter had assimilated with the Roman Empire since the early third century (Constitutio Antoniniana), and were counted as Rhomaioi. The Arab groups who had migrated from the peninsula, instead, were not likely assimilated to the Roman Empire, but rather integrated at different levels. In Syria and Palestine, for instance, the elite of the Jafnids joined the Byzantine Empire at a political level by obtaining the rank of foederati. In the sixth century one of the leaders was even conferred the title of phylarch of the

219

220

MATTIA GUIDETTI

Arab foederati. 1 Culturally, being largely Christians, the great part of these Arab groups started to adopt Greek alongside Arabic; they managed to have their own monasteries and churches, but also to attend extant Christian sacred places and to intermingle with the non-Arab population. 2 Al-Ruṣāfa on the verge of the Syrian steppe is a case in point: the Jafnid al-Mundhir ordered an inscription invoking his name through a hippodrome motto to be carved in Greek on the eastern portion of an extra-muros structure. The Christian-Arabs were involved in the urban life of this pilgrimage center restored by Justinian, because of their devotion to St Sergius, whose relics were preserved in the sanctuary located in the south-eastern sector of the city. 3 Notwithstanding that the ethnical, cultural, political, and administrative trajectory of the Nabataeans was very different in comparison with that of later Arab groups, their common Semitic roots favored the confusion between them on the eve of the Muslim conquest. 4 The Arab groups which had migrated from the peninsula since the fourth century were sometimes confused with the settled Aramaic population, also because of their sedentariness and means of subsistence (sheep farming and agriculture in contrast to camel breeding). 5 As suggested by Shahid, the mutual influence between these two groups might perhaps help to explain the inscription in the church of St George on the Mount Nebo. Here, in the Province of Arabia, in the southern sacristy of the church, we find a mosaic dated to 536 displays a palm tree flanked by two oryx and by two inscriptions: on the right side it reads in Greek the name of the benefactor (Saul); and on the left side, in reads in Arabic the benevolent expression “in peace”. Influenced by the Ara1 2

291–6. 29.

3

Fisher, Between Empires 194–210. Shahid, Byzantium and the Arabs i, part 2, 824–42; 949–89; ii, part 1, Fowden, Qusayr ʿAmra 129–31; Fowden, The Barbarian Plain 101–

On Nabatean ethnic identity, see Retsö, The Nabateans. Toral-Niehoff, The ‘Ibad of al-Hira 6–7. See the case of Umm alJimal (in today Jordan): De Vries, Umm el-Jimal 232–6. 4 5

INCLUSION AND SECLUSION

221

maic script, the latter would be one of the earliest evidences of the use of Arabic in a mosaic work. 6 Despite several situations of assimilation between the Arabs and the local Semitic and Greek populations, in many cases, the first established new settlements. One term used in the sources to describe the late antique Arab settlements is ḥāḍir. In Ibn Manẓūr’s late 13th century Lisān al-ʿarab we find an explanation of the term if used referring to individuals: “Whoever settles in areas where spring waters are plentiful and leave neither in summer nor in winter, is ḥāḍir. They settle in villages, cultivated lands, and built areas, or they erect tents in proximity to a spring. Hence they fix their own residence and pasture their flocks.” 7 The passage continues by distinguishing between ḥāḍirūn and the aʿrāb, the latter being semi-nomadic people who stop by springs only during summers. A ḥāḍir is the opposite of a traveler (huwa ḍidd al-musāfir), so that it is clear that by the concept of ḥāḍir Arabic sources mean sedentariness in semi-fertile areas outside towns and cities. Ḥāḍir is also the term used to designate the settlements that appear often to be in connection with the movement of Arab groups from Arabia to Syria in the late antique period. This is the case, for instance, of Aleppo: “Close by the city of Aleppo stood a settlement called the Ḥāḍir Ḥalab in which different Arab tribes including Tanūkh lived. Abu-’Ubaidah made terms with them in which they agreed to pay poll-tax. Later they embraced Islām and lived with their descendants in the same place until a little after the death of ar-Rashīd.” 8 The Tanukhids were likely a group of Arab tribes who moved out from the Arabian Peninsula in the fourth century. Here it is worth stressing the closeness of this settlement to the city of Aleppo. Yāqūt in the early 13th century describes the same settlement: “As regards to the ḥāḍir we should say it is among the largest neighborhoods outside the walls of Aleppo. Its buildings are only a Piccirillo, The Mosaics of Jordan 178–9; Shahid, Byzantium and the Arabs ii, part 2, 297–302. 7 Ibn Manẓūr, Lisān iii, 214–9. 8 al-Balādhurī, The origins 224. 6

222

MATTIA GUIDETTI

stone’s throw away from the city walls in the south-western direction. …its inhabitants are for the most part arabized Turcomans, family members of the garrison soldiers; in the ḥāḍir one can find a fine Friday mosque where the sermon is pronounced and several markets where all kind of goods are on sale.” 9 Hence, in the 13th century, the ḥāḍir of Aleppo has been transformed into one of the extra-muros quarters of the city. It retained however one important feature of the late antique foundation: its inhabitants, although not Tanukhids but Turcomans, were newcomers. 10 Another case which might help to understand the nature of the ḥāḍir is Raqqa: “The van of ʿIyād’s army arrived in ar-Raqqah and made a raid on (the ḥāḍir in) its environs, where the Beduin Arabs (sic) were encamped with a group of peasants, carrying off much beauty. Those who escaped took to flight and entered the city of ar-Raqqah.” 11 Hitti’s translation does not include ḥāḍir in the English text and misinterprets the word “ʿarab” (translated as Bedouin Arabs) as designating the Arab people inhabiting Ḥāḍir Raqqa at the moment of the conquest. “ʿArab”, not to be confused with “aʿrāb”, meant a group of sedentary farmers and traders of Arabic origin, whereas the peasants (fallāḥūn) were of Syriac culture. According to Retsö, the term “ʿarab” was used in order to differentiate in Syria between the late antique Arabs and the seventh-century ArabMuslim conquerors (called muhājirūn). Arabs and peasants furthermore played a social role that was different from the one played by the badw (Beduins) or the a’rāb, the latter being seminomadic or nomadic groups. 12 We can go a step further in the comprehension of how the ḥāḍir looked like by looking at al-Balādhūri’s passage on Qinnasrīn: “Thus the Moslems effected the conquest of the land of Qinnasrin with its villages. The Ḥāḍir Qinnasrin had been settled by the Tanūkh tribe since they came to Syria and pitched their tents in it. Yāqūt, Muʿjam al-buldān ii, 185. Shahīd, Byzantium and the Arabs 369, 418–22; Sauvaget, Alèp 62–4. 11 al-Balādhurī, The origins 270. 12 Retsö, The Arabs 99–101; Kaegi, Byzantium 171–3. 9

10

INCLUSION AND SECLUSION

223

They later built their houses in it. These, Abu-’Ubaidah summoned to Islām. Some of them accepted it, but the banū-Saliḥ ibn-Ḥulwān ibn-’Imrān ibn-al-Ḥāfi Ibn Qudā’ah remained Christian.” 13 In this passage we find two more important points that help to understand the nature of ḥāḍir and the role they played in the aftermath of the seventh-century conquest. One can assume a twostage development of the settlement: in a first phase, late antique Arabs encamped in this area (4km far from the walled city of Chalcis), whereas in a second phase they transformed the tents into permanent buildings. After the area was conquered in the seventh century, part of the Arab population converted to Islam whereas others kept their Christian belief. 14 In recent years the site and nature of Ḥāḍir Qinnasrīn and the early Islamic stratum of Chalcis ad Belum have been archeologically investigated. The most compelling reconstruction is that the citadel of Chalcis might have been occupied by the seventh-century conquerors and that a series of satellite settlements were created around it at the same time, and that this paralleled the development of Late Antique Ḥāḍir Qinnasrīn (whose population partially converted to Islam in the aftermath of the conquest) and the establishment of the village today-known as Hadir. 15 In historiography the term “ḥāḍir” has been often assimilated to the concept of “ḥīra”, that, according to Sourdel-Thomine’s updating of Lammens’ article in the Encyclopedia of Islam, means an “encampment where nomads settled down, a jumble of tents and buildings”. 16 This term was used for a settlement such as al-Jābiya, a center located south-west of Damascus. Al-Jābiya was not built outside a “classical” city, such as the case of the ḥāḍirs mentioned above. This center, with Jalliq, Resafa/Sergiopolis, and Ḥuwwārīn, was among the towns associated in later written sources with late antique Arab settlers and more specifically with the Ghassanids, a later, loose denomination for the Arab groups lead by the Jafnids. al-Balādhurī, The origins 223. See, on this passage, Whitcomb, Hadir Qinnasrin. 15 Compare Whitcomb, Hadir Qinnasrin and Archaeological Research with Rousset, Hadir. 16 Lammens, al-Djābiya. 13 14

224

MATTIA GUIDETTI

Although not excavated, al-Jābiya has been explored at the beginning of the 20th century, and the descriptions question the “encampment-like” appearance of the site. Dussaud explains the dissemination in the outskirts of architectural elements pertaining to the ancient site: “Depuis des siècles ces ruines sont exploitées par les habitants des villages voisins qui s’y fournissent de pierre à bâtir. Aujourd’hui ils sont réduits à fouiller les fondations.” He notices the quality of the stone-cutting and points out the courtly character of the settlement: “On distingue encore des traces de murs en pierres bien taillées. L’ensemble est du type général employé dans le Ḥaurân. Comme toutes les installations ghassànides, alDjabiya n’était pas une grande ville, mais une agglomération autour des constructions royales.” 17 Caetani links the observation of the remains with some information available in the sources, and draws a sketch of the life in the settlement: “... We will limit ourselves to observe that the ruins of alǦābiyah illustrate the nature of the place. It should not have been a populous village, as in the case of many other centers around it, but a small one, developed around a copious water source springing at the foot of the isolated hill known as alǦābiyah; from the top of such hill the view dominates the entirety of the plain of Nuqrah. ...the ruins are not vast, but from the size and quality of the cutting of the stones it is clear that the buildings now in ruins were once of extreme luxury, satisfying rich people who were used to appreciate ostentation and that sojourned there for amusement. ... The Ghassanid estate, perhaps surrounded by gardens, was located at the foot of the hill; whereas on its top the hill might have been crowned by a guard tower, given that the ruins of an old fortress are still visible.” 18

This passage emphasizes the palatial features of the ruins and the existence of a guard tower over the hill; in a successive passage Caetani points out the probable presence of empty areas around 17 18

Dussaud, Mission 46. Caetani, Annali 927–8.

INCLUSION AND SECLUSION

225

the palaces where the troops and the ordinary people organized their settlements. The Christian character of the city, inferable from the written sources, has been highlighted by Shahid after Nöldeke. Al-Jābiya should have had a monastery and a church dedicated to St George, a saint particularly venerated by Christian Arabs (as was the case of St Sergius). 19 Hence, in late antique Syria and Palestine, within a process of increasing sedentariness the Arabs settled in centers provided with permanent and non-ephemeral structures; while the settlement policy in some cases was dictated by strategic locations overlooking traffic routes and the vast landscape, in other cases it was characterized by the proximity to the “classical cities”.

AFTER THE CONQUEST: EARLY MEDIEVAL SETTLEMENT POLICIES

One of the features of the 630’s conquest to be taken into account is the rather small size of the conquering army. 20 This had an impact in the post-conquest settlement policies in the new territories. It is likely the conquerors set up military camps outside the conquered cities to plan the successive steps of their advance and to organize the territory under their control. One place which, according to written sources, appealed to the conquerors since the very early period was al-Jābiya, the abovementioned Christian-Arab center that flourished in the sixth century. Written sources mention the settlement in the context of important political events of the early history of the Muslim caliphate. Because the troops would have chosen al-Jābiya as a strategic point during the years of the conquest, it is from here that in 638 ʿUmar is told to have given the so-called “Khuṭbat al-Jābiya,” a speech in which he laid the foundations for the administrative organization of the conquered lands. Still in 684 it was in al-Jābiya where all the Arab-Muslim chiefs Noeldeke, Topographie 430; Shahid, Byzantium and the Arabs ii, part 1, 96–104, part 2, 263–7; Shahid, Arab Christianity 233. 20 Kennedy, Conquest 1–32; on the historiographical problems concerning Muslim accounts of the Islamic conquest, see Robinson, Empire and Elites 1–32. 19

226

MATTIA GUIDETTI

gathered in order to proclaim Marwān as the new caliph to counterattack the threats posed by Ibn Zubayr. 21 The second caliph of the Marwanid line of the caliphate, ʿAbd al-Malik, although resident in Damascus, often stopped in al-Jābiya with his court on the way back to the capital from his winter residence of al-Ṣinnabra (on the shore of Tiberias Lake). 22 Hence, al-Jābiya constituted a sort of favored gathering place for the representatives of the scattered tribes that supported the early caliphate, permanently hosting a certain number of the ArabMuslim armies. Although there is no information yet about the physical presence of early Muslims in al-Jābiya (coins, religious architecture), it seems its importance should have lasted until the late seventh – early eighth century. From that time onwards Damascus became the political center of the caliphate, whereas at the beginning of the eighth century, the army was moved north to al-Dābiq, in anticipation of new campaigns against Constantinople. 23 Aside from al-Jābiya, early Muslims also founded new settlements with a military shape but clearly characterized also by buildings expressing the new religion and the new authority. This is the case of urban settlements such as Ayla and ‘Anjar. They share several features: the presence of massive walls with towers giving to these sites a late antique fortress-like appearance, the orthogonal planning of the roads with, in the case of ‘Anjar, a tetrapylon elevated at the crossing of the two main avenues, and the hierarchical division of the inner space between the quarter of the authorities, the market areas, and the residential units. There are also differences: ‘Anjar’s size (310x370m) is approximately double of Ayla’s (145x170m) and whereas the latter appears to have been built in front of a late antique settlement still populated in the meanwhile Ayla was erected, the rationale of the foundation of the former was probably the road networks connecting Damascus to the Mediterranean coast (though it is not unlikely that the area around it was populated during late antiquity). The dates of founding are also important: Ayla was founded very early after the seventh-century aṭ-Ṭabarī, Taʾrīkh i, 2108, 2360, ii, 474–6. al-Balādhurī, Ansāb 200. 23 aṭ-Ṭabarī, Taʾrīkh ii, 1315. 21 22

INCLUSION AND SECLUSION

227

surrender of Aylana (the nearby late antique settlement), whereas ‘Anjar appears to have been founded (but also very soon abandoned to be later repopulated in the Middle Ages) by the caliph alWalid around the year 715. 24 Both these settlements show “classical” features that were reinterpreted and revitalized during late antiquity. As for the ensemble of cardo, decumanus and tetrapylon, in fact, it is worth mentioning that these were urban features reenacted through sixthcentury imperial commissions in cities such as Jerusalem, Scythopolis, and Zenobia. 25 At the same time, however, both Ayla and ‘Anjar also have a mosque in their center (in both cases built in connection with potential government palaces, though the one excavated in Ayla is dated to the early Abbasid time), a feature which underlines the presence of the new power and new religion. Modern scholars have described such kind of settlements through the notion of “miṣr”, a term whose meaning shifts from a “garrison-town” provided with urban features to an “encampmentlike foundation.” 26 In the written sources miṣr often appears as a Whitcomb, The Miṣr; Finster, Researches in ʿAnjar; Finster, ʿAnjar; on Ayla and its role as the harbor for the nearby mineral exploitation and in the trade with the Indian Ocean region, see Power, The Red Sea 106–109. 25 Kennedy, City planning; Lauffray, Halabiyya-Zenobia 157, fig. 5; Sodini, La contribution 145. 26 Whitcomb, Amsār: “The early Muslims must have found this a conductive environment when they entered Syria and founded their amsar. The amsar were more than camps, however, but founded as urban administrative centers upon Arab urban models. The sites of ʿAnjar in Lebanon or Aqaba in southern Jordan may contain reflections of these models and their adaptations” (Whitcomb, Hadir); cfr. Kennedy, The Armies of the Caliphs 7. Creswell sees the miṣr in opposition to the classical city and as much as the “chaotic labyrinth” is concerned as the foreshadowing of the medieval “madina”: “In the foundation of these miṣrs, which were really little more than permanent encampments, facilitating for the Arabs the slow and difficult transition from a nomad life to a settled existence, no directing authority intervened beyond the marking-out (ikhtiṭāṭ), never binā’ – building, or ‘imāra – construction or repair – of the mosque 24

228

MATTIA GUIDETTI

sort of title or qualification given to urban centers without any clear-defined criteria for the nature of the urban fabric. In the Lisān al-’arab one finds this definition: “Al-Layṯ said that in Arabic miṣr denotes each locality in which Qur’anic punishments are applied and the incomes and the alms revenues are managed…” 27 In this context is important to underline that Damascus is also described by al-Muqaddasī as “miṣr”: “Damascus is the miṣr of Syria and the seat of the sovereign in the age of the Umayyads.” 28 And the same al-Muqaddasī offers a glimpse on the variety of the uses of the term according to different individuals: “There are several differences between the concepts of amṣār, the experts of law define them as populated cities in which justice is executed, the governor lives, the incomes are managed, and from where the state is ruled (as for example ʿAṭṭar, Nābulus, Zūzan); the experts of language define them as a sort of boundary line between two sides (as for example al-Baṣra, al-Raqqa and Arjān); according to the population the amṣār are any big and populated centres (as for example al-Rayy, al-Mawṣil, al-Ramla). As for us, we consider miṣr any place in which the supreme authority is located, the administrative departments have their seats, the administrators are given authority... (as for example Damascus, Qayrawān and Shirāz).” 29 The account of the foundation of al-Ramla in Palestine (early eightcentury) provides further details on the process of creation of a new urban centre and of establishing a miṣr: “Al-Walīd ibn ʿAbd al-Malik made Sulaymān ibn ʿAbd al-Malik governor of the province of Palestine. He moved to Ludd and later founded the city (madīna) of al-Ramla and made it a miṣr. The first thing he built was his palace and a building known as the “house of the dyers” in the middle of which a cistern was erected. Later he planned the mosque and had it built…” 30 This passage makes clear that miṣr is and a few main roads, hence the chaotic labyrinth of lanes and blind alleys, of tents and huts alternating with waste ground, cemeteries and mound of filth.” (Creswell, Early Muslim Architecture i, 39). 27 Ibn Manẓūr, Lisān al-ʿarab xiii, 121. 28 al-Muqaddasī, Aḥsan al-taqāsīm 156. 29 al-Muqaddasī, Aḥsan al-taqāsīm 47. 30 al-Balādhurī, The origins of the Islamic state 143.

INCLUSION AND SECLUSION

229

something distinct from planning or founding. It seems to be something which is added to an existing settlement. Perhaps its meaning is “to fortify”, 31 but the above mentioned passages and the story of al-Ramla suggest instead that miṣr here means that the city is qualified as an administrative center in which the governor had settled (with his court and soldiers). According to the sources, al-Ramla was founded nearby Ludd, because the local elite of the latter refused to provide the new governor with a plot of land to have his new palace built. 32 Early Islamic urban foundations such as Ayla, al-Ramla, and ‘Anjar, which might be interpreted as amṣār, not because of a specific urban layout, but because of the potential presence of Muslim authority in them, show the tendency to seclusion, one typical feature of the early Islamic society. 33 Although these settlements show different features, the overall evidence suggests that the new ruling elite chose to separate itself from the local conquered society through the foundation of distinct, separate settlements, while the local’s social and material life kept going on under the new rule. On the other hand, the settlement policies of late antique and post seventh-century conquest Arabs show obvious parallels. In both cases, the extant settlements attracted the newcomers, whose settling choice in turn somehow depended on the continuity of the late antique centers. In the case of late antique ḥāḍir, the commercial and farming activities of the newcomers made necessary to exploit pre-existing urban areas, whereas in the case of the post conquest settlements, the need of the minority to rule, and the continuity granted to the conquered cities resulted in the foundation of relatively small palatial and residential settlements close to the extant al-Balādhurī, The origins of the Islamic state 220, note 1. Rosen-Ayalon, The First Century; for a resume of the archaeological surveys, see: Luz, The construction. 33 The tendency to add a new urban settlement nearby an existing one, as a way in order to expand it, has been explained by Benet (The ideology of Islamic urbanization 212; Cuneo, Storia dell´urbanistica 92) through the concept of the parallax (or parallactic system) as opposed to the idea of palimpsest. I thank Dr. Ignacio Arce for sharing with me this reference. 31 32

230

MATTIA GUIDETTI

towns. In the first case, the main factor of differentiation was ethnicity (the newcomers were Arabs from the peninsula), whereas in the latter one, the additional factors were religion and the exercise of authority. The separateness between the satellite new city in the early Islamic time and the nearby Late Antique town is embodied by the prominence of the still-functioning Late Antique cathedrals in the latter and by the erection of the “mosque and palace” in the former. More controversial and difficult to assert is the occupation of the late antique ḥāḍirs by the conquerors. The pivotal role of alJābiya up to the last decades of the seventh century shows that probably because of the Arabic roots of their population, the late antique ḥāḍir attracted the new coming seventh-century Arabs. 34 As in the case of Ḥāḍir Qinnasrīn, the early medieval phase of the ḥāḍir is still under investigation, although it seems that the main bulk of the new elite settled in the citadel of the nearby city of Chalcis ad Belum. 35 In Aleppo we can attest to a certain interest in the late antique ḥāḍir, since Sulaymān b. ‘Abd al-Malik, who was the builder of the Great Mosque within the city in the first quart of the eighth century, is said to have commissioned a lavishly decorated palace in the extra-muros ḥāḍir. 36 According to the sources, it seems plausible that, since part of the ḥāḍir population converted to Islam and many others remained Christians, Christian and Muslim Arabs shared the same urban space. The intermingling between “locals” and “newcomers”, however, was not only due to the shared ethnic roots of the population, as, during the same period, it also started in the conquered cities. In the latter, the presence of small Muslim communities is evidenced by the presence of the earliest Muslim houses of worship. Actually, Arabness as an element facilitating the interaction among Christian and Muslim Arabs is often highlighted by al-Jāḥiẓ in his Treatise in Refutation of the Christians (Risāla, 14–15, 19). 35 Compare Whitcomb, Hadir and Archaeological Research with Rousset, Hadir. 36 Ibn Shihna, Al-durr 51–4. Compare Whitcomb, Hadir and Archaeological Research with Rousset, Hadir. 34

INCLUSION AND SECLUSION

231

the existence of seventh-century mosques is not supported by any archaeological remains, but only attested by descriptions in the written sources. In the seventies of the seventh century Adamnán describes the tiny and modest mosques of Jerusalem and Damascus. The mosque of Jerusalem is also mentioned in an anonymous Georgian description of al-Ḥaram al-Sharīf, whereas in Amida (Diyarbakir) the presence of a masjid is reported by Theodotus. 37 The relative small size of early Muslim communities can be deduced from the argument we usually find to support the enlargement of several mosques in the successive decades: often, the sources refer to the increasing number of believers that made necessary replanning the mosques (see the cases of the mosque of ʿAmr at Fusṭāṭ, the al-Aqṣā in Jerusalem, and the Great Mosque of Cordoba in al-Andalus). The pressure exerted on the urban fabric by the growing Muslim community is illustrated at best by the process in Damascus, where the Muslim authorities asked for, claimed, and finally obtained by force the whole precinct of the former temenos, which had been previously occupied by the Great Church of the city, flanked in a later stage by the small mosque founded straight after the conquest. The size of the earliest mosques was probably adequate to the size of the early Muslim communities. 38 Hence, in the early days after the conquest, different settlement processes were underway: while the conquered cities were not rejected, it seems that most conquerors preferred to settle outside the extant towns, or were forced to do so, because of the dynamics of conquest. Sometimes, these settlements took the form of satellite garrison cities planned following the model of late antique castra, and perhaps also imitating the relationship that late antique ḥāḍirs had with the Hellenistic-Roman towns; in other cases, they dwelled on the extant, late antique Arabic settlements in Syria and Palestine, probably by virtue of the ethnic familiarity with their population. However, with the consolidation of the Umayyad caliphate, and especially with the access to power of the Marwanid Guidetti, The Byzantine Heritage, 4. al-Balādhurī, The origins of the Islamic state 191–3; Creswell, Early Muslim Architecture i, 180–96, figs. 99–100. 37 38

232

MATTIA GUIDETTI

lineage of the dynasty, the attraction exerted by the conquered cities of ancient foundation would increase noticeably.

THE UMAYYADS: THE EXPLOSION OF THE CONQUERED CITIES

With the establishment of the Umayyad house lineage (661–750), Syria and Palestine grew in importance. At the same time, the size and nature of the Muslim community started to change: Islam fully consolidated as a distinct monotheistic religion and non-Arabs and Christian Arabs slowly merged into the Muslim community through conversion. 39 All this impacted settlement policies. The Umayyad rulers commissioned the refurbishment of the conquered cities by applying a clear-cut urban program. The transformation may be summarized in three points: the refurbishment or the construction of marketplaces, the building of a monumental Muslim core (with the mosque-palace symbiosis), and the reorganization of the “minorities” (especially the Christians, still representing the numerical majority of the inhabitants in the Syro-Palestinian cities). 40 Market areas are described by written sources in Damascus, Ḥimṣ, and al-Raqqa. 41 Beyond the newly-established settlements, archaeological surveys brought to light important evidence of market areas datable to the Umayyad time in Palmyra, al-Ruṣāfa, Jerash, and Scythopolis/Baysān. 42 Levtzion, Conversion to Islam 289–94; on the definition of conversion see, more recently, also: Papaconstatinou, “Introduction”; Carlson, Contours of Conversion. 40 The theme of the transformation of classical cities into Islamic ones has a long bibliography. Perhaps, the most representative article at this regard is Kennedy, From Poleis to Medina. 41 Yāqūt, Muʿjam al-buldān iv, 108; Madelung, W., Apocalyptic Prophecies 170–2; al-Balādhurī, The origins of the Islamic state 271. The market in al-Raqqa would later be moved and expanded once the nearby twin city of al-Rāfiqa will be founded (al-Balādhurī, The origins of the Islamic state 280). 42 al-Asʿad and Stepniowski, The Umayyad Suq; Khamis, Two WallMosaic Inscriptions; Tsafrir and Foerster, Urbanism 138–40; Ulbert, Die 39

INCLUSION AND SECLUSION

233

The market of Baysān is particularly important because it shows the imperial agenda behind its refurbishment as well as how the commercial area was meant as a place of inclusion of the newcomers into the late antique society. The area was dedicated to commerce since the late antique period with the transformation of the Roman portico and the partial occlusion of the street known as “Silvanus Street.” The Umayyads stepped into the area through the edification of a passage across the line of the shops marked by two monumental gates. On the south-eastern one, on both sides of the portal, two mosaic inscriptions have been found. Mounted into two stone frames, the wall mosaics, made of tiny glass tesserae, have an inscription with a gold Kufic script and a blue background. The choice of the wall glass mosaics and the presence of gold tesserae testify to the imperial commitment in the creation of such program. The content of the inscriptions confirms this: on the right side, the text reads the Muslim declaration of faith, whereas on the left side, under the protection of the basmala, the work is declared as having been commissioned by the caliph al-Hishām (724–743) through his governor in Jordan, Isḥāq bin Qābisa. 43 The mosaics in Baysān are extraordinarily important as they make clear how Umayyad caliphs considered the market central for their economic exploitation and appropriation of the conquered cities. Furthermore, the findings of Baysān harmonize with a passage stating that Umm Ḥākim, the wife of the caliph al-Walīd (705–715), the eldest brother of al-Hishām had a market built in Damascus and alRaqqa. 44 The involvement of the elite as displayed on the gate walls had repercussions also in the market transactions. Among the findings in the market area, one object attests the transition from the Byzantine to the Islamic rule: it is a circular weight of ca. 84g. made of a copper alloy. 45 As shown by Miles and Khamis, this weight exemBasilika. 43 Khamis, Two Wall-Mosaic Inscriptions. 44 Yāqūt, Muʿjam al-buldān iv, 108; Madelung, W., Apocalyptic Prophecies 170–2; al-Balādhurī, The origins of the Islamic state 271; alBalādhurī, The origins of the Islamic state 280. 45 Khamis, A Bronze Weight.

234

MATTIA GUIDETTI

plifies the Muslim adoption (shape, typology) and adaptation (erasure, substitution of formulae) of the late antique Byzantine-style weights. 46 Around a margin the Kufic script tells the traders the basmala and the shahāda, whereas within the inner circle, in the place in which a cross surmounted the Greek letters indicating the value of the weight, an Arabic inscription declares the Muslim authority who ordered the manufacturing. This transformation of the official tools for commercial transactions parallels the reform of the administration and coinage, which had a tremendous impact on the early Islamic Syrian and Palestinian society. In the conquered people’s view the appearance of Qur’anic formulae within the frame of the daily life activities replacing Christian formulae and authorities – choices which went together with a restriction of liberties of non-Muslims in the public sphere – inevitably marked a point of no return in the transition from the Greek to the Arab authority. 47 The encounter between the newcomers and the “local population,” however, took also place on a more practical level: three balances recovered in Baysān’s Umayyad stratum (before the earthquake of the year 749) show notches marking the weight both in Greek and in Arabic, suggesting that “business was carried out with both Greek-speaking and Arabic-speaking customers.” 48 Hence, the proliferation of the markets during the eighth century in Syria and Palestine shows how the conquered cities were by then included in the settlement process of the Muslims, and that, with the passing of time, they were even preferred to new settlements: “rather than rejecting the already existing urban centers (the predominant practice in Iraq, Egypt, and North Africa), Greater Syria poleis were continually animated both commercially and industrially during at least the first four centuries of Islamic rule.” 49 George C. Miles, A Byzantine Bronze Weight; Khamis, A Bronze Weight 148–54. 47 Donner, Muhammad and the beilievers 205–11; see also Griffith, Images. 48 Tsafrir and Foerster, Urbanism 144. 49 Foote, Commerce 38; Raḥal, Tāʾrīkh 166–87; Walmsley, Production 276–83. 46

INCLUSION AND SECLUSION

235

The establishment, refurbishment, or expansion of a market in the conquered cities in some cases accompanied the foundation of a mosque. In Jerash, the Great Mosque that was built at the crossroads of the cardo and south decumanus stands near to the late antique and early medieval market area. The shops were carved out by converting sections of the porticoes of the two main roads and by adding shop-units to the circular facade of the tetrakionion (the circular square where usually a tetrapylon is located). 50 If the hypothesis that an administrative center was located west of the mosque is correct, the early Islamic rule gravitated around the tetrakionion, whereas the Christian quarter was located north of it, around the monumental cathedral still used by the Christians. Something similar – although the mosque appears less monumental – occurs in Palmyra. Here, in a still predominantly Christian city, the mosque was built over an abandoned Roman building located a few meters south-east of the tetrakionion, nearby the so-called “Umayyad market.” 51 The Umayyads are also credited with a new market built in alRuṣāfa adjacent to St Sergius’ shrine. 52 The market should have been located west of the ecclesiastic complex, where we find the main gateway to the northern courtyard of the complex. While it is not clear if its construction occluded the view of the church complex from the public street, 53 it is likely that this intervention came along with the construction of the Great Mosque by Hishām b. ʿAbd al-Malik on the northern side of the courtyard of the basilica. This was part of a strategy that aimed at locating the mosque close to the holiest Christian shrine in al-Ruṣāfa and that included the construction of a vast residential area extra-muros for the caliph and his court. 54 Simpson, Market Buildings; Walmsley and Damgaard, The Umayyad Congregational Mosque; Walmsley, The Friday Mosque. 51 Gawlikowski, Palmyra; Denis Genequand, An Early Islamic Mosque; al-Asʿad and Stepniowski, The Umayyad Suq. 52 Ulbert, Beobachtungen. 53 Cr. Foote, Commerce 30. 54 Guidetti, Sacred Topography 351–2. 50

236

MATTIA GUIDETTI

While the pair “market plus mosque” was in some cases completed by the administrative complex, 55 the case of al-Ruṣāfa introduces the element of the local church as an element of gravity attracting the building activities of the new elite. The contiguity between the mosque established for the Muslim community and the main church attended by the Christian community embodies the urban, daily contact between different individuals in the conquered cities under the new rule. It seems likely that this pattern was not followed shortly after the conquest, but rather later on, in the first decades of the eighth century, when most great mosques were built by the Umayyad rulers. 56 Among the cases recently brought to light, al-Bara’s mosque is worth mentioning. If the interpretation recently offered by Charpentier and Abdulkarim is correct, a great mosque was built in the center of the city during the Umayyad period. A bathhouse complex divided the mosque from the main church of the town. The mosque, probably provided with a pavement mosaic and presenting well-cut dressed stone walls, was almost of the same size as the main church of al-Bara, to be eventually reduced during its medieval reoccupation. 57 The edification of the mosque in the same area as the main church, or even contiguous to it, was pursued in several cities such as Mosul, Edessa, Diyarbakir, Aleppo, Ḥimṣ, downtown Amman, Shivta (in addition to al-Ruṣāfa and al-Bara). Whereas in al-Ruṣāfa the Christian and Muslim believers had all access to the courtyard in between the two religious buildings, in Aleppo, in which the relationship the great mosque had with the main church for over 400 Bacharach, Marwanid Umayyad Building. See, in addition, the case of Tiberias, where we can attest the possible existence within the late antique city of an early Islamic compound (provided with a Friday mosque, a palace, and a bath complex) that is under investigation: Cytryn-Silverman, The Umayyad Mosque. 56 Walmsley and Damgaard, The Umayyad Congregational Mosque, 372–6. 57 Charpentier and Maamoun, Une première campagne. 55

INCLUSION AND SECLUSION

237

years is still observable, the two buildings were spaced out only by a narrow street west of the mosque. 58 This intense program of construction, which impacted the commercial areas, the place for an administrative complex, and the extant churches, evidences that the Muslim communities started to monumentalize their presence in the conquered cities during the early eighth century. An important aspect is the effort to take possession of the pivotal sectors of urban areas, which did not correspond necessarily to the center of the classical city, but reflect instead the late antique urban developments. The seizure of buildings and plots of land by the new rulers was probably part of the negotiations that took place at the time of the seventh-century conquest between the conquerors and the local communities, an agreement aiming at preserving properties, goods, and communitarian rights in exchange of taxes and loyalty. Through the consequent circulation of different individuals in the markets, in the city roads, and in the areas where the religious buildings were located, Muslims interacted on a daily bases with the local communities. The process of conversion, moreover, blurred the distinction between “local” and “not-local” as a mere religiousfocused issue. 59 At the same time, besides the extensive and bombastic building programs in the conquered towns, the Umayyads founded also new settlements. Abovementioned al-Ramla and ʿAnjar date to the early eighth century. The former was built ex-novo near to the late antique provincial capital of Ludd, supposedly because the Muslim prince and future caliph, the governor Sulaymān b. ʿAbd al-Malik, encountered difficulties to build his administrative-religious complex within the late antique city. Ludd’s population, composed of Guidetti, The Byzantine Heritage 4–7, and Guidetti, In the Shadow of the Church 41–62. 59 Parallel to these social processed are the reactions expressed by Christian theologians and Muslim early jurisprudence warning the dangers of the disappearance of boundaries between different communities. This last issue is currently under investigation by the present author; see Hoyland, Seeing Islam, 147–9, 161–3, 381–3, 604; al-Jāḥiẓ, Risāla, 14, 19; Griffith, The Church, 147–51; ; Guidetti, In the Shadow, 79–86. 58

238

MATTIA GUIDETTI

Greeks, Jews, and Arabs, is said to have been moved to al-Ramla, which therefore seems a foundation made by the new elite but not designed exclusively for it. As a matter of fact, later medieval sources describe Ludd as having a “large mosque” and a “marvelous church,” whereas al-Ramla (whose construction was supervised by a Christian from Ludd) as provided with a monastery, a great mosque and (perhaps later in the Middle Ages) a considerable Jewish community active in commercial trade. 60 The case of the city of ʿAnjar was perhaps different, since its urban structure shows a palatial core and a series of standardized residential units (at least during the Umayyad phase), that were planned perhaps to host a certain number of families of the ruler’s narrowest military units and courtiers. The same might be the case of Mshatta, a late Umayyad foundation, left unfinished, which also displays an inner core linked to palatial activities surrounded by standardized residential units. 61 These sites can be compared to the palatial areas built by the Umayyads in Amman and Jerusalem, which did not substitute the city around them, but whose purpose was most likely to host the rulers and the close courtly circles. Finally, it is worth mentioning Qaṣr al-ḥayr al-sharqī, whose inscription (today disappeared) dates the foundation to 728–9 and defines the settlement a madīna. Here we find a multi-unit settlement provided with an elite residential area, residential units for the lower members of the ruling elite, a vast agricultural enclosure, and an area for industrial activities scattered through a vast area in the Syrian steppe. 62 The people of Ḥimṣ, who according to the inscription were involved in the construction, do not seem to have settled in this new town, which apparently was a distinctively Arab-Muslim al-Balādhurī, The origins of the Islamic state 220–1; al-Muqaddasī, Aḥsan al-taqāsīm 176; Luz, The construction 48; Chelo, The roads from Jerusalem 138. Consider also the case of Helwan, the satellite city built by the Egyptian governor ʿAbd al-ʿAzīz b. Marwān, where also several Christians had monasteries and churches built: al-Armanī, Churches and Monasteries 154–7. 61 Carver, Transitions to Islam; Hillenbrand, ʿAnjar. 62 Genequand, From ‘desert castle’. 60

INCLUSION AND SECLUSION

239

town – founded in connection with the itinerant habit of early eighth century Umayyad rulers.

CONCLUSIONS

Although archaeological excavations are continuously bringing to light new evidence related to the seventh-eighth century Bilād alShām, it is possible to delineate some tendencies with regard to the settlement policies of the newcomers. In the very early years after the conquest, the late antique Arab presence in Syria and Palestine served probably as a sort of lighthouse to guide those who migrated from the Arabian Peninsula. This implied that several places attracted and hosted the Arab-Muslims (al-Jābiya first, al-Ruṣāfa later), and that, in addition, the ḥāḍir, first conceived as a satellite settlement close to but distinct from late antique cities, became a model for the relationship that several early Islamic foundations came to establish with the conquered cities. In a later period, the Umayyad elite undertook the construction of settlements with military, palatial and agricultural functions (ʿAnjar, Qaṣr al-ḥayr alsharqī, Mshatta). At the same time, with a pace that was determined by the end of the hostilities, the reinforcement of the Muslim identity, and the conversions of “local” individuals and groups to Islam, Muslims started to settle with more confidence within the urban area of the conquered cities (Aleppo, Ḥimṣ, Hamā, Damascus, Amman, Jerash, and so on). An analysis of the location that was chosen by Muslims for the construction of their mosques (often associated to administrative buildings) discloses the importance of finding a strategic point in the urban area. Market areas and church complexes were often considered as being the most effective sites in order to control and expand commercial activities, as well as to provide an adequate arena for the display of the new rule and faith in what were still largely non-Muslim cities.

BIBLIOGRAPHY

al-Armanī, A.S., Churches and Monasteries of Egypt and some Neighbouring Countries, transl. B.T.A. Evetts, Oxford 1969. al-Asʿad, K. and F.M. Stepniowski, The Umayyad Suq in Palmyra, in Damaszener Mitteilungen 4 (1989), 205–23. Bacharach, J.L., Marwanid Umayyad Building Activities: Speculations on Patronage, in Muqarnas 13 (1996), 27–44.

240

MATTIA GUIDETTI

al-Balāḏurī, The origins of the Islamic state, transl. P. Khuri Hitti, New York 1988. al-Balāḏurī, Ansāb al-ashrāf, ed. W. Ahlwardt, Greifswald 1883. Benet, F., The Ideology of Islamic Urbanisation, in International Journal of Comparative Sociology 4 (1963), 211–26. Caetani, L., Annali dell’Islām, vol. 3, Milan-Rome 1905–1926. Carlson T., Contours of Conversion: The Geography of Islamization in Syria, in Journal of the American Oriental Society 135 (2015), 791–816. Carver, M.O.H., Transitions to Islam: Urban Rôles in the East and South Mediterranean, Fifth to Tenth Centuries AD, in Towns in Transition. Urban Evolution in Late Antiquity and Early Middle Ages, ed. N. Christie and S.T. Loseby, Aldershot 1996, 184– 212 Charpentier, G. and A. Maamoun, Une première campagne d’étude sur la mosquée d’al-Bara?, in Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, ed. K. Bartl and A. al-Razzaq Moaz, Leidorf 2008, 45–56. Chelo, I., The roads from Jerusalem, in Jewish Travellers in the Middle Ages, ed. E. N. Adler, London 1930, 130–150. Creswell, K.A.C., Early Muslim Architecture, vol. 1., Oxford 1932. Cuneo, P., Storia dell’urbanistica. Il mondo islamico, Bari 1986. Cytryn-Silverman, K., The Umayyad Mosque of Tiberias, in Muqarnas 26 (2009), 37–61. De Vries, B., Toward a History of Umm el-Jimal in Late Antiquity, in Umm el-Jimal. A Frontier Town and its Landscape in Northern Jordan, vol. 1, ed. R. de Vries, Journal of Roman Archaeology 26 (1998). Dussaud, R., Mission dans les régions désertiques de la Syrie moyenne, Paris 1903. Finster, B., Researches in ʿAnjar. I. preliminary report on the architecture of ʿAnjar, in BAAL 7 (2003) 209–44. Finster, B., ʿAnjar: spätantik oder frühislamisch?, in Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, ed. K. Bartl and A. al-Razzaq Moaz,

INCLUSION AND SECLUSION

241

Leidorf 2008, 229–52. Fisher, G., Between Empires: Arabs, Romans and Sasanians in Late Antiquity, Oxford 2011. Foote, R.M., Commerce, Industrial Expansion, and Orthogonal Planning: Mutually Compatible Terms in Settlements during the Umayyad Period, in Mediterranean Archaeology 13 (2000), 25–38 Fowden, E.K., The Barbarian Plain. Saint Sergius between Rome and Iran, Berkeley 1999. Fowden, G., Qusayr ʿAmra. Art and the Umayyad Elite in Late Antique Syria, Berkeley 2004. Gawlikowski, M., Palmyra in Early Islamic Times, in Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, ed. K. Bartl and A. al-Razzaq Moaz, Leidorf 2008, 89–96. Genequand, D., An Early Islamic Mosque at Palmyra, in Levant 40 (2008), 3–15. Genequand, D., From ‘Desert Castle’ to Medieval Town: Qasr alHayr al-Sharqi (Syria), in Antiquity 79 (2005), 350–61. Griffith, G., The Church in the Shadow of the Mosque. Christians and Muslims in the World of Islam, Princeton 2008. Griffith, S., Images, Islam and Christian Icons, in La Syrie de Bysance à l’Islam, VIIe-VIIIe siècles, ed. P. Canivet, J.-P. Rey-Coquais, Damascus 1992, 121–38. Guidetti, M., In the Shadow of the Church. The Building of Mosques in Early Medieval Syria, Leiden 2016. Guidetti, M., Sacred Topography in Medieval Syria and Its Roots between the Umayyads and Late Antiquity, in Umayyad Legacies. Medieval Memories from Syria to Spain, ed. A. Borrut and P. M. Cobb, Leiden 2010, 339–64. Guidetti, M., The Byzantine Heritage in the Dār al-Islām: Churches and Mosques in al-Ruha between the Sixth and the Twelfth Centuries, in Muqarnas 26 (2009), 1–36. Hillenbrand, R., ʿAnjar and Early Islamic Urbanism, in The Idea and Ideal of the Town between Late Antiquity and the Early Middle Ages, ed. G.B. Brogiolo and B. Ward-Perkins, Leiden 1999, 59–98.

242

MATTIA GUIDETTI

Hoyland, R., Seeing Islam as Others Saw It. A Survey and Evaluation of Christian, Jewish and Zoroastrian Writings on Early Islam, Princeton 1997. Ibn Shihna, Al-durr al-muntakhab fī ta’rīkh mamlakat Ḥalab, ed. ʿA. alDarwish, Damascus 1984. Ibn Manẓūr, Lisān al-ʿarab, Beirut 1988. al-Jāḥiẓ, Risāla fī radd ʿalā al-naṣāra, ed. J. Finkel, Three essays of Abu ‘Othman ‘Amr b. Bahr al-Jahiz, Cairo 1926. Kaegi, W.E., Byzantium and the Early Islamic Conquests, Cambridge 1992. Kennedy, H., The Great Arab Conquest. How the Spread of Islam Changed the World We Live In, London 2007. Kennedy, H., The Armies of the Caliphs. Military and Society in the Early Islamic State, London and New York 2001. Kennedy, H., City Planning from Classical Antiquity to Early Islam, in Sciences sociales et phénomènes urbains dans le monde arabe, edited by M. Naciri and A. Raymond, Casablanca 1997, 23–8. Kennedy, H., From Polis to Madina – Urban Change in Late Antique and Early Islamic Syria, in Past and Present 106 (1985), 3– 27. Khamis, E., Two Wall-Mosaic Inscriptions from the Umayyad Market Place in Bet Shean/Baysān, in BSOAS 64 (2001), 159– 76. Khamis, E., A Bronze Weight of Sa’īd b. ʿAbd al-Malik from Bet Shean/Baysān, in The Journal of the Royal Asiatic Society 12 (2002), 143–54. Lammens, H., al-Djābiya, in EI2, ii, 360. Lauffray, J., Halabiyya-Zenobia. I, Paris 1983. Levtzion, N., Conversion to Islam in Syria and Palestine and the Survival of Christian Communities, in Indigenous Christian Communities in Islamic Lands Eighth to Eighteenth Centuries, ed. M. Gervers and R. J. Bikhazi, Toronto 1990, 289–311. Luz, N., The construction of an Islamic city in Palestine. The Case of Umayyad al-Ramla, in Journal of the Royal Asiatic Society 7 (1997), 27–54. Madelung, W., Apocalyptic Prophecies in Hims in the Umayyad

INCLUSION AND SECLUSION

243

Age, in Journal of Semitic Studies 31 (1986), 141–85. Miles, G.C., A Byzantine Bronze Weight in the Name of Bishr b. Marwān, in Arabica 9 (1962), 113–18 al-Muqaddasī, Aḥsan al-taqāsīm fī ma’rifat al-aqālīm, ed. M.J. De Goeje, Leiden 1967. Nöldeke, Th., Zur Topographie und Geschichte des Damascenischen Gebietes und der Haurângegend, in ZDMG 29 (1876), 419–44. Papaconstatinou, A., Introduction, in Conversions in Late Antiquity: Christianity, Islam and Beyond, ed. A. Papaconstantinou, N. McLynn and D.L. Schwartz, Farnham, xv-xxxvii. Piccirillo, M., The Mosaics of Jordan, Amman 1993. Power, T., The Red Sea from Byzantium to the Caliphate, Cairo 2012. Raḥal, ‘Ā., Tā’rīkh bilād al-shām al-iqtiṣādiyy fī al-ʿaṣr al-umawī, Beirut 2000. Retsö, J., The Arabs in Antiquity. Their History from the Assyrians to the Umayyads, London 2003. Retsö, J., The Nabateans. Problems of defining ethnicity in the ancient world, in Visions of Community in the Post-Roman World: The West, Byzantium and the Islamic World, 300–1100, ed. W. Pohl, C. Gantner, R. E. Payne, Burlington 2012, 73–79. Robinson C., Empire and Elites after the Muslim Conquest: The Transformation of Northern Mesopotamia, Cambridge 2000. Rosen-Ayalon, M., The First Century of Ramla, in Arabica 43 (1996), 250–63. Rousset, M.-O., Hadir, Hadir Qinnasrin, Qinnasrin, que sait-on de la capitale de la Syrie du Nord au début de l’Islam?, in Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, ed. K. Bartl and A. al-Razzaq Moaz, Leidorf 2008, 357–75. Sauvaget, J., Alèp. Essai sur le développement d’une grande ville syrienne, des origines au milieu du XIX siècle, Paris 1941. Shahid, I., Arab Christianity in Byzantine Palestine, in ARAM 15 (2003), 227–37. Shahid, I., Byzantium and the Arabs in the Sixth Century, 2 vols., Washington 1995–2009.

244

MATTIA GUIDETTI

Shahīd, I., Byzantium and the Arabs in the 4th Century, Washington 1984. Simpson, I., Market Buildings at Jarash: Commercial Transformation at the Tetrakionion in the 6th to 9th c. C.E., in K. Bartl and A. al-Razzaq (eds.), Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, ed. K. Bartl and A. al-Razzaq Moaz, Leidorf 2008, 115–24. Sodini, J.-P., La contribution de l’archéologie à la connaissance du monde byzantine (IVe-VIIe siècles), in Dumbarton Oaks Papers 47 (1993), 139–84. aṭ-Ṭabarī, Ta’rīkh ar-rusūl wa-l-mulūk, ed. M.J. De Goeje, Leiden 1879–1901. Toral-Niehoff, I., The ʿIbad of al-Hira: an Arab Christian Community in Late Antique Iraq, in The Qur’an in Context – Entangled Histories and Textual Palimpsests, ed. A. Neuwirth, N. Sinai, and M. Marx, Leiden 2009, 323–47. Tsafrir, Y. and G. Foerster, Urbanism at Scythopolis-Bet Shean in the Fourth to Seventh Centuries, in Dumbarton Oaks Papers 51 (1997), 85–146. Ulbert, Th., Beobachtungen im Westhofbereich der Grossen Basilika von Resafa, in Damaszener Mitteilungen 6 (1992), 403– 16. Ulbert, Th., Die Basilika des Heiligen Kreuzes in Resafa-Sergiupolis, Mainz am Rhein 1986. Yāqūt, Kitāb mu’jam al-buldān, 6 vols., ed. F. Wüstenfeld, Leipzig 1866–1873. Walmsley, A., and K. Damgaard, The Umayyad Congregational Mosque of Jarash in Jordan and its Relationship to Early Mosques, in Antiquity 79 (2005), 362–78. Walmsley, A., The Friday Mosque of Early Islamic Jarash in Jordan: The 2002 Field Season of theDanish-Jordanian Jarash Project, in The Journal of the David Collection 1 (2003), 111–31. Walmsley, A., Production, Exchange and Regional Trade in the Islamic East Mediterranean: Old Structures, New Systems?, in The Long Eighth Century. Production, Distribution and Demand, ed. L. Hansen and C. Wickham, Leiden 2000, 265–344.

INCLUSION AND SECLUSION

245

Whitcomb, D., Hadir Qinnasrin Excavations. 1998–99 Annual Report, http://oi.uchicago.edu/research/pubs/ar/98– 99/qinnasrin.html and Archaeological Research at Hadir Qinnasrin, 1998, Archéologie Islamique 10 (2000). Whitcomb, D., The Miṣr of Ayla: Settlement at al-ʿAqaba in the Early Islamic Period, in The Byzantine and Early Islamic Near East II. Land Use and Settlements Patterns, ed. A. Cameron and G.R.D. King, Princeton 1994, 155–70. Whitcomb, D., Amsār in Syria? Syrian Cities after the Conquest, in ARAM 6 (1994), 13–33.

IV. PERSPECTIVES ON THE RELIGIOUS HISTORY OF SOUTH ARABIA

247

ALTSÜDARABISCHE KULTE IM VORCHRISTLICHEN ABESSINIEN FRANCIS BREYER

Über die Christianisierung Äthiopiens ist bereits sehr viel geschrieben worden, die vorangehende Phase der äthiopischen Religionsgeschichte hat bislang jedoch vergleichsweise wenig Aufmerksamkeit erfahren. Ein Grund dafür dürfte sein, dass Näheres über die altsüdarabischen Kulte selbst erst in den letzten Jahrzehnten wirklich bekannt wurde, nicht zuletzt durch die Freilegung entsprechender Tempelanlagen etwa in Marib oder Ṣirwāḥ. Wie vielerorts müssen bei der Betrachtung der altsüdarabischen Kulte in Abessinien zwei Ebenen unterschieden werden: der offizielle Kult und der Volksglaube. Um die Einflüsse der altsüdarabischen Kulte in Abessinien wirklich herausarbeiten zu können, ist es gleichzeitig vonnöten, sich Gedanken über die aksumitische Religion vor der Christianisierung als Ganzes zu machen, namentlich die zugrunde liegenden autochthonen ‚afrikanischen’ Glaubensvorstellungen. Bislang wurde allzu oft unbedacht aus dem altsüdarabischen Raum Bekanntes auf Abessinien übertragen und solchermaßen die Unterschiede zwischen ursprünglich rein abessinischen Bräuchen und fremden Kultpraktiken verwischt. 1 In diesem Zusammenhang möchte ich einen Horizont herausarbeiten, den man im weitesten Sinne als ‚sudanisch’ beschreiben könnte. Ferner spielen sicherlich nicht kontaktinduzierte Tendenzen mit in das Bild hinein, aber auch Einflüsse aus dem mediterranen Raum. 1

Munro-Hay, Aksum; Kobishchanov, Aksum.

249

250

FRANCIS BREYER

Beides ist in Altsüdarabien ebenso greifbar und sollte letztlich zur Etablierung des christlichen Monotheismus in Äthiopien führen. Nachdem bereits von Äthiopien und Abessinien die Rede war, sollten einleitend einige Worte über die Verortung der folgenden Betrachtungen verloren werden. Wenn ich von Äthiopien spreche, meine ich nicht den heutigen Staat gleichen Namens, sondern die entsprechende Kulturlandschaft; man vergleiche dies mit den überlappenden Begriffen ‚Nubien’ (kulturell) und ‚Sudan’ (geopolitisch). Um allen möglichen Verstimmungen hinsichtlich der Animositäten zwischen Äthiopiern und Eriträern zu begegnen, empfielt es sich, den heute außer Gebrauch gekommenen Terminus Abessinien wieder zu aktivieren, der politisch – zumindest im Deutschen – nicht vorbelastet ist. 2 Geographisch bewegen wir uns auf dem Nordteil der äthiopischen Hochebene in der heutigen Provinz Tigray in Äthiopien bzw. in Eriträa und Süd-Sudan. Der zeitliche Rahmen ist das spätantike Reich von Aksum und sein altsüdarabisch geprägter Vorläufer, das sabäische Reich von DʿMT 3, genauer gesagt die Zeit vor der Christianisierung Abessiniens unter dem aksumitischen Negus ʿEzānā gegen 330 n. Chr. 4 An Quellen stehen uns zur Verfügung Steleninschriften in sabäischer Sprache und Schrift, solche in sog. Pseudo-Sabäisch und in Gǝʿǝz sowie griechische Inschriften. 5 Hinzu kommen altäthiopische Graffiti und vor allem die äthiopische Überlieferung aus späteren Epochen. Insbesondere Hagiographien geben versteckte Auskünfte über die heidnische Religion. Nicht zu unteschätzen sind ferner die Möglichkeiten, durch eine Art historischer Ethnoarchäologie etwa aus dem nubischen Raum auf vergleichbare Phänomene im abessinischen zu schließen. Bisher haben diese Im Italienischen ist dies anders: abessinia hat dort vor allem seit Mussolini eine pejorative Konnotation. 3 Munro-Hay, Arabia. 4 Zusammenfassung der Diskussion bei Munro-Hay, Aksum, Kapitel 10.2. 5 Bernand, Drewes and Schneider, Recueil des Inscriptions; Sima, Die “sabäische” Version. 2

ALTSÜDARABISCHE KULTE

251

Verbindungen allzu wenig Beachtung gefunden, um nicht zu sagen – gar keine. Zentrales Anliegen ist es demnach, jene kulturgeschichtlichen Verflechtungen mit den sudanischen Kulturen wenn nicht erschöpfend zu behandeln, so doch erstmals aufzuzeigen und ferner, deutlicher zwischen sabäisch und genuin aksumitisch zu trennen. Dass dies nicht immer möglich sein wird, versteht sich beinahe von selbst. Ebenso die Schwierigkeiten, vom Tenor der offiziellen Verlautbarungen weg zu kommen und Informationen über die Volksfrömmigkeit jener Zeit zu erlangen. Dass es zwischen beiden Ebenen eine teils beträchtliche Diskrepanz gegeben haben muss, zeigt nicht zuletzt die durchaus lange Zeit, welche das Christentum brauchte, um sich im aksumitischen Reich fest zu etablieren. 6 Nun also zuerst zu dem Teilaspekt der Religion, den man als Staatskult bezeichnen könnte, dem offiziell propagierten Kult der herrschenden Schicht. Verschiedentlich wurde aufgrund der prädominanten sabäischen Elemente in der Kultur des wenig erforschten Reiches von DʿMT postuliert, es müsse sich um hierbei um einen Ableger der altsüdarabischen Kultur gehandelt haben. Die Verwendung der sabäischen Sprache und Schrift sprechen dafür, wie auch der Gebrauch altsüdarabischer Architekturformen und weiterer kultureller Eigenheiten. Manche Forscher vertreten sogar die Ansicht, DʿMT sei nicht weniger als eine sabäische Kolonie auf dem afrikanischen Kontinent gewesen. 7 Für die vorliegende Fragestellung ist dies insofern von Bedeutung, als bei einer Okkupation in irgendeiner Form die Machthaber eben nicht Abessinier waren, sondern Südaraber. Wenn man der Ansicht vieler Linguisten folgt, wonach die äthiosemitischen Sprachen durch Wanderung größerer Gruppen von Sprechern semitischer Sprachen über das Rote Meer nach Afrika gelangt seien, so ändert sich das Bild entscheidend. Nun sollte sich die Bevölkerung Abessiniens zumindest kulturell vergleichsweise homogen darstellen und folglich müssten auch die Glaubensvorstellungen relativ einheitlich Ausführlicher dazu Haas, Mountain Constantines. Zusammenfassend erläutert bei S.C. Munro-Hay, Arabia. Relations in ancient times, in EncAeth i, 294–300. 6 7

252

FRANCIS BREYER

sein. Sie sind es jedoch nicht. So gibt es viele Gottheiten, die eindeutig altsüdarabischer Herkunft sind und die in aksumitischer Zeit gar nicht mehr in Erscheinung treten. Darunter ist sogar ein Gott, welcher seinerseits aus Mesopotamien ‚importiert’ wurde: der Mondgott Sîn. Mit anderen Worten: die Beobachtung von Unterschieden zwischen der sabäischen und der aksumitischen Kultschicht könnte einen Beitrag zur Frage nach den Ursprüngen und der Natur des Reiches von DʿMT leisten. Betrachten wir also zunächst die Gottheiten der altsüdarabischen Inschriften aus Abessinien näher. Der Befund entspricht allem, was von der altsüdarabischen Religion bekannt ist: An der Spitze eines Götterkreises steht der männliche Astralgott ʿAṯtar, nicht etwa ein ʾl (ʾEl oder Allāh). Daneben finden wir bedeutende Sonnengötter, allen voran Šams (wie in Qatabān und im Ḥaḍramawt). In Altsüdarabien sind jedoch die Mondgötter sehr präsent, stellen sie doch in vielen Fällen eine Art Nationalgott (s²ym) dar. Sabaʾ

Maʿīn

〈ʾlmqḥ〉

Qataban

〈wd〉

ʾAusan

Ḥaḍramawt

〈wd〉 〈ʿm〉

〈s1yn〉

*ʾAlmaqaḥ

unklar

*Wadd

,Liebe’

*Wadd

*ʿAmm Sîn

,Liebe’

,Onkel’ unklar

Aufgrund der wenigen Quellen ist in Abessinien weder der Patronatsgott eines byt: m(n)ḍḥ bzw. mnḍḥt. bekannt, noch der persönliche Schutzgott einer bestimmten Dynastie ,Sonne’ (s2ms), bzw. im Plural ʾs2ms. 8 In Abessinien erscheinen die Gottheiten fast immer nur auf Votivgaben (Räucheraltären). Der Typus dieses Objektes ist sehr altsüdarabisch und lässt auf entsprechend vergleichbare Kultpraktiken schließen. Folgende Götter erscheinen in Texten aus D’MT: 9 Beeston, Sayhadic. Vgl. Sima, Almaqah; Frantsouzoff, Ḏāt-Himyʾm; Frantsouzoff, HWBS. 8 9

ALTSÜDARABISCHE KULTE *ʾAlmaqah ʾlmqh

253

RIE 1:5; RIE 5B:2; RIE 6:3, RIE 20:2, RIE 22:2, RIE 26:3, RIE 27:2f., RIE 28:2, RIE 29:1, RIE 30:3, RIE 31:3, RIE 32C, RIE 37:1, RIE 39:4f., RIE 56:2

RIE 8:2, RIE 10:14

ʾlmqhy

RIE 9:6 (Schreib- bzw. Kopierfehler?)

ʾlmqy

Wurfholz-Szepter 10

lmq *Ḏāt-Himyām ḏ.ṯ-Hmym ḏ.ṯ-Hmn *Ḏāt-Baʿdan ḏ.ṯ-bʿḏn *Hūbas

RIE 1:6, RIE 5B:2, RIE 10:14, RIE 53, I:2

RIE 9:6f., RIE 53 II:1, RIE 71:2, RIE 72:2, RIE 74:1, RIE 75:2, RIE 77:2, RIE 158, RIE 163 RIE 1:6, RIE 2:2, RIE 5B:2; RIE 9:7, RIE 10:14, RIE 61:2 RIE 6:3, RIE 9:6, RIE 35b

hwbs

RIE 1:4, RIE 5B:2, RIE 10:14, RIE 62:2, RIE 70:2, RIE 89 I:2, RIE 153; RIE 155:1

hbs ʿAstär

RIE 1:5; RIE 5b:2; RIE 10:13f.; RIE 13:4; RIE 33 II; RIE 106:2; RIE 152:2; RIE 153; RIE 155:2

ʿš1ṯr ʿTtr

RIE 39:4; RIE 55b; RIE 56:2; RIE 58

šrqn

RIE 63A:2f.

Sonstige

vgl. śäräqä “aufgehen (Stern, Sonne, Mond)” vgl. śäräq “Neumond”

RIE 32:2

s1n

RIE 33 II

nrw

*/Sîn/

*/Narū; Nūru/?

vgl. nar “Feuer” bzw. nur “Licht” 10

Jamme, Ethiopia.

254

FRANCIS BREYER Wurfholz-Szepter 11

ʿrg

*/ʿArg; ʿErg/?

Leider haben sich keine mythologischen Texte erhalten, d.h. um die Bedeutung der Gottheiten im Einzelnen und ihre Verbindung zueinander zu ergründen, ist man auf Indizien angewiesen. Um dies zu bewerkstelligen wurden bislang zwei Wege beschritten, die Etymologisierung ihrer Namen sowie eine Evaluation der Reihenfolge, in welcher die Gottheiten genannt werden. Sie soll Einblicke in die Ordnung des Pantheon geben. Paradebeispiel für die etymologische Herangehensweise ist das Götterpaar ḎTḤMYM und ḎTBʿDN. Ihre Namen wurden als *Ḏāt-Ḥimyam ,Die des Strahlens’ bzw. *Ḏāt-Baʿdan ,Die der Ferne’ gedeutet. 12 Der zweite Ansatz basiert auf einer Theorie von Detlev Nielsen, nach der das Hauptcharakteristikum der semitischen Religionen eine Gliederung der Götterwelt in eine Trias von Astralgottheiten bestünde. 13 Diese seien Venus, Mond und Sonne. Folglich wurde jede Reihung von Götternamen in dieses dreiteilige Schema gepreßt, z.B.: Venus ①

Hūbaš

ʾAlmaqah

Ḏāt-Ḥimyam Ḏāt-Baʿdan

Mond ②

Sonne ③

Im vorliegenden Fall mussten fünf Numina auf drei Aspekte verteilt werden, folglich nahm man für vier der fünf an, sie repräsentierten nur einen Teilaspekt des astralen Wesens. Entsprechend wurden sie aufgespalten in Gegensatzpaare wie Vollmond – Neumond oder Sommersonne – Wintersonne etc. Polemisch könnte man diese etwas naives Vorgehensweise als Theorie von »Sonne, Mond und Sterne« bezeichnen und es verwundert nicht, das sie heute nicht mehr allgemein akzeptiert ist. Jamme, Ethiopia. Höfner, Orts- und Götternamen in Südarabien. 13 Nielsen, Der dreieinige Gott, bes. 312 une 321. 11 12

ALTSÜDARABISCHE KULTE

255

Erstens ist sie viel zu schematisch und zweitens ist das Konzept einer einheitlichen, sog. »altsemitischen Trias« bereits an sich fragwürdig. Das Problem ist nun, dass die Aussagen über Wesen und Funktion jener Götter, welche auf dieser Grundlage vorgenommen wurden, sich in der Forschung fest etabliert haben und immer und immer wieder fortgeschrieben werden. Diese Schieflage betrifft die aksumitische Götterwelt besonders, da sich hier mittlerweile eine Fächergrenze aufgetan hat. Götter in verschiedenen Reihungen wurden miteinander geglichen, einzig und allein aufgrund ihrer Stellung in der Liste. So gelangte man für die aksumitischen Gottheiten zu teils recht willkürlichen Gleichungen wie Hūbas mit Bəḥēr 14 oder Ḏāt-Ḥimyam mit Mədr bzw. Γῆ 15. Sehr viel aussagekräftiger als diese Positionsspielereien sind m.E. Beobachtungen über die Häufigkeiten der Nennungen einzelner Gottheiten in den Texten. So können nämlich eher Aussagen über die Verbreitung der jeweiligen Kulte gemacht werden. Aber nicht nur das. Was das Reich von DʿMT angeht, ist nämlich immer noch umstritten, woher die Träger jener Kultur denn nun genau kamen. Wenn es sich wirklich um eine altsüdarabische ‚Kolonisation’ handelt, von welcher Region Südarabiens stammten die Herrscher? Die Inschriften könnten hierzu einen Hinweis liefern. Der Mondgott Sîn wird nämlich in den DʿMT-Texten genannt und er war vor allem im Ḥaḍramawt populär – in Sabao war bereits ein anderer Mondkult etabliert. Dies würde auf eine Besiedlung aus Gegenden Südostarabiens weisen. Dagegen spricht jedoch eindeutig die Tatsache, daß der mit Abstand populärste Gott im Abessinien jener Zeit der sabäische Nationalgott ʾAlmaqah war – Heiligtümer für diesen Gott wurden in Mälazo sogar archäologisch nachgewiesen. 16 Neben der Häufigkeit ist es auch eine genaue Analyse von Graphien, die zusätzliche Informationen liefern können. Beispielsweise findet sich auf dem Zeremonial-Wurfholz aus ʿAddi Gälämo (Region Aṣbi & Darʾa) 17 eine interessante Graphie: LMQ. Nach Frantsouzoff, HWBS. RIE 185, ii, 21; 188:26. Frantsouzoff, Ḏāt-Himyʾm. RIE 185, i:21; 185 bis i:22f.; 185 bis iic:35f.; 188:25f. bzw. RIE 270 bis 30. 16 Kobishchanov, Aksum 48f. und 224f. 17 Nebes, Sabäische Texte 2 336, Anm. 37. 14 15

256

FRANCIS BREYER

Jamme ist sie in Altsüdarabien charakteristisch für das 2. bzw. 3. Jhd. n. Chr. 18 Weitere Distributionen zwischen den verschiedenen Graphien für ʾAlmaqah lassen sich jedoch nicht feststellen. 19 Für die Göttin ḎT-HMYM gilt ähnliches wie für Sîn: auch sie wurde vornehmlich im Wādī Ḥaḍramawt verehrt, vor allem in Raybūn. 20 Im Falle von HWBS könnte eine anders geartete Beobachtung wichtig sein: in sabäischen Inschriften wird diese Gottheit zwar sehr früh genannt, allerdings wird sie erst relativ spät Teil des offiziellen sabäischen Pantheons, in der zweiten Hälfte des 1. Jts. v. Chr. 21 HWBS steht in DʿMT immer an zweiter Stelle direkt nach ʿAstär und vor*ʾAlmaqah (RIE 1:5f.; RIE 5B:2; RIE 6:2; RIE 9:6f.; RIE 10:13f.), was auf eine besondere Verehrung in Abessinien schließen lassen könnte. Islamische Autoren berichten, HWBS sei eine Mondgottheit, doch in Gobo Fənṣəḥ wurde sie in einer Grotte verehrt (RIE 153; RIE 155) und dürfte demnach wohl eher chtonische Aspekte in sich vereinen. Zahlreiche Tempel (byt) altsüdarabischer Götter sind in Abessinien inschriftlich bezeugt: Yəḥa

RIE 85 I

Bei Dibdib

RIE 69; RIE 70

Mäṭära

Tokondaʿ

RIE 69; RIE 70 RIE 85 I

Nach dieser Zusammenschau über die Kulte von DʿMT wenden wir unseren Blick auf die frühaksumitische Zeit und hier speziell auf die Gottheiten der ‚heidnischen’ Inschriften. Als Erstes können wir beobachten, dass nun Götter genannt werden, welche vorher nicht in Erscheinung getreten waren, wie etwa Bəḥēr. Besonders prägnant ist jedoch ein weiterer Einschnitt: ʾAlmaqa wird nicht mehr verehrt, zumindest nicht mehr unter diesem Namen. Seine Rolle wurde ganz offensichtlich eingenommen von einem gewissen Zur Datierung von Gadurat vgl. Pirenne, L’inscription “Ryckmans 535” 1f.; Loundine and Ryckmans, Nouvelles données 411. 19 Sima, Almaqah. 20 Frantsouzoff, Ḏāt-Himyʾm. 21 Frantsouzoff, HWBS. 18

ALTSÜDARABISCHE KULTE

257

Mäḥrəm (Ἄρης). Die dritte Veränderung betrifft die mögliche Funktion einzelner Numina im Pantheon – es scheint keine Götter mehr gegeben zu haben, die dezidiert mit Sonne und Mond in Verbindung gebracht werden. Vereinte Mäḥrəm vielleicht diese Aspekte in sich? 22 Bezeugt sind in den aksumitischen Inschriften aus vorchristlicher Zeit folgende Gottheiten: Mäḥrəm (Ἄρης)

RIE 6:2, 18, 26; RIE 7: 3, 19, 21, 25; RIE 8:4; RIE 9:4; RIE 10:29f. RIE 185 I:2; 18 II:3; 19; 21; 25 bzw. RIE 185bis I:3; 19, 23, II:3f.; IIC:26, 36, 44; RIE 186:4f.; RIE 187:4; RIE 188:5, 29–30.

ʿÄstär

RIE 7:21; RIE 10:26

Mədr

RIE 7:21; RIE 10:26

Bəḥēr

RIE 185 II:21; RIE 185bis IIC:35; RIE 188:25

RIE 6:21; RIE 10:25

RIE 185 I:20f.; II:21; RIE 185bis IIC:35f.; RIE 188:25f.; RIE 277:39f.

Da sich unsere Kenntnisse zumeist auf die bloße Namensnennung beschränkt, bleiben zahlreiche Fragen offen. Wie genau muss man sich die Funktion des Maḥrəm vor allem in den Inschriften ʿĒzānās vorstellen? Ist in der aksumitischen Zeit ʿAstar noch der Hochgott oder warum wird sein Name noch bis weit in die christliche Ära hinein als Gottesbezeichnung gebraucht? Worin besteht der Unterschied zwischen den Gottheiten Mədr und Bəḥer? 23 Wenden wir uns zunächst dem Maḥrəm zu. 24 Ganz eindeutig vereinigt er in sich einige Funktionen, welche zuvor dem sabäischen Nationalgott ʾAlmaqah zugesprochen wurden. So kann er getrost als Stammesgott der Aksumiten bezeichnet werden bzw.

Kobishchanov, Aksum 229. DAE 6&7 (S. 13); Kobishchanov, Aksum 229 zitiert Vycichl (kultiviertes Land vs. Land an sich) 24 Kobishchanov, Aksum 230. 22 23

258

FRANCIS BREYER

als Gott des herrschenden Clans. So heißt es in der aksumitischen Inschrift aus Abba Ṗänṭälewon von ihm, er sei der »Gott der Aksumiten«. Bei ʿĒzānās ist er gar der »Herr des Himmels und der Erde«. Damit noch nicht genug. Die ʿĒzānās-Inschriften weisen ihn dezidiert als kriegerischen Gott aus, dem Gefangene und Beute aus Feldzügen geweiht bzw. geopfert werden. In dieser Funktion wird er mit dem griechischen Kriegsgott Ares geglichen ‒ ʿĒzānās stilisiert sich zum Sohn des »unbesiegbaren Ares« (ἀνικητος). Damit wird ein wesentlicher Zug der aksumitischen oder vielleicht auch nur der ʿĒzānās-Inschriften deutlich: das Gottkönigtum, welches der altsüdarabischen Kultur fremd war. Maḥrəm ist der Ahne und göttliche Vater des Königs. Besonders interessant ist eine Wendung auf der Adulitana, wo es im Bezug auf den König heißt, Maḥrəm sei »sein größter Gott«. Dies zeigt, wie sehr die persönliche und die politische Ebene miteinander verbunden sind: der König ist der Staat, sein persönlicher Gott ist der Nationalgott, der Gott des Clans, der Gott aller Aksumiten. Aussagen dieser Art fehlen in den altsüdarabischen Inschriften von DʿMT. Hat sich demnach ein grundlegender Wandel im Gottesverständnis vollzogen oder kommt hier nicht vielmehr ein bislang in den Texten nicht repräsentiertes Stratum zum Vorschein, eine genuin aksumitische Schicht? Beweisen läßt es sich nicht, doch meine ich, letzteres dürfte der Fall sein. Aspekte wie Göttkönigtum oder Gottessohnschaft sind nicht gerade typisch für den altsüdarabischen Raum, jedoch sehr wohl Charakteristika nordostafrikanischer Kulturen. Wir finden sie im pharaonischen Ägypten genauso wie im benachbarten spätantiken Meroë, wie im gesamten sudanischen Raum bis in die Neuzeit hinein. Fassen wir zusammen: die herausgehobene Stellung des ʾAlmaqah im Reich von DʿMT erschließt sich uns heute nur noch durch seine häufige Nennung relativ am Anfang von Götterreihungen. Demgegenüber wird Maḥrəms dominante Stellung im Pantheon durch zahlreiche Aussagen sehr viel expliziter deutlich gemacht: 1. Kosmokrator

»Herr des Himmels und der Erde«

2. Kriegsgott

»des unbesiegbarern Ares«

3. Göttlicher Vater des Königs

»ʿĒzānā ... Sohn des unbesiegbaren Ares«

ALTSÜDARABISCHE KULTE 4. Persönlicher Schutzgott des Königs

»sein größter Gott«

5. Stammesgott der Aksumiten

»Gott der Aksumiten«

259

Erinnern wir uns an die sabäischen DʿMT-Inschriften: an der Spitze des Pantheons stand hier – wie überhaupt im antiken Südarabien – der im Gegensatz zu seinem mesopotamischen Pendant Ischtar männliche ʿAstar. Nun begegnet uns dieser Gott zwar nicht mehr in den aksumitischen Inschriften, sein Name lebt jedoch bis in christliche Zeit fort und zwar als allgemeine Gottesbezeichnung ʿastär . Die äthiopische Ecclesiasticus-Übersetzung etwa bezeichnet den christlichen Gott noch im sog. ‚Mittelalter’ als ʿAstär. 25 Dies ist auch der Befund im äthiopischen Sirach. 26 Die Verbindung zwischen diesen späten Handschriften und den früheren aksumitischen Inschriften kann hergestellt werden durch eine Passage in der Stele RIE 10, wo ʿaśtar eineuitig »Himmel« bedeutet (RIE 10:25). Im aksumitischen Gəʿəz ʿaśtar liegt also ein Sprachgebrauch vor, der heute noch im nordäthiopischen Tigrə bekannt ist: ʿastär heißt dort eben immer noch »Himmel«. Diese Bedeutung läßt sich sogar in einem äthiosemitischen Lehnwort im kuschitischen Bilin greifen (ʿastär »Himmel«). 27 Wir haben es also nicht nur sprachlich, sondern auch kultur- und religionsgeschichtlich mit einem Relikt aus der altsüdarabischen Periode der äthiopischen Geschichte zu tun. Der Name des sabäischen Hochgottes wird aufgrund seiner Eigenschaft als Astralgottheit zum Begriff für den Himme schlechthin, auch und gerade als seine Rolle im Kult durch den neuen, genuin aksumitischen Maḥrəm eingenommen wird. Solchermaßen läßt sich durch die Etymologie nicht nur ein kurzer Blick auf die Glaubensvorstellungen im voraksumitischen Abessinien erhaschen, sondern auch auf den Prozess der Verdrängung durch die aksumitische Gedankenwelt. Ein weiteres Problem der aksumitischen Epigraphik ist die Unterscheidung zwischen Mədr und Bəḥer. Sie können aufgrund

Cerulli, Storia 24. Dillmann, Biblioa 117. 27 Leslau, Comparative Dictionary 73. 25 26

260

FRANCIS BREYER

der Textduplikate in den Trilinguen König ʿĒzānā unterschiedlich miteinander in Beziehung gesetzt werden: 28

aÄth.1 w⸗ṣḥf⸗n | [z.t | ] ṣḥf.t | w⸗śm⸗n | w⸗ʾ:(21)[b]ʾ⸗n | l-ʿstr | w⸗l⸗mdr | w⸗l-mḥrm | z⸗wld⸗n

aÄth.2 w⸗(C32)ṣḥf{⸗k} | z⸗ṣḥ(C33)f.ṯ | w⸗śm{⸗k} (C34) w⸗ʾ:mḥḍn{⸗k} (C35) l-ʿšṯr | w⸗b(C36)ḥr | w⸗l-mḥrm (C37) z⸗wlḏ⸗

pSab.1 w]. ṣḥf⸗n | z.t | (20) ṣḥ.f.tm | w⸗šm⸗n | w⸗[ʾ:bʾ⸗nm | ] l-ʿstrm | [w⸗](21)l-bḥrm | w⸗[l-mḥrmm | z⸗wld⸗nm | pSab.2 w⸗ʾṣḥf⸗n | z[ṣ](22)ḥf.t | w⸗šm⸗nm | w⸗ʾmḥḍ⸗nm | l-ʿṯtrm | w⸗l[b](24)ḥrm | w⸗l-mḥrm | l-z⸗ld⸗nm

Grie.1

——

Grie.2 καὶ ἀνέθηκα ταύτυν τὴν στήλην καὶ παρε(30)θέμην αὐτὴν τῷ Οὐρανῷ καὶ τῇ Γῇ tῷ μαι (31) γεννήσαντι ἀνικήτω Ἀρέει.

Wie lassen sich die teilweise gegensätzlichen Informationen miteinander in Einklang bringen? Dazu gibt es mehrere Möglichkeiten, je nachdem, von welchen Ordnungskriterien man ausgeht:

1. In allen Versionen beider Duplikate wird derselbe Gott genannt. 2. Beide pseudosabäischen Versionen nennen einheitlich einen, beide äthiopischen einheitlich einen anderen Gott. 3. Die Nennung hängt vom Aufstellungsort der Stele ab: am Ortsausgang Richtung Meer stand der Meeresgott Bəḥēr, in Richtung Landesinnere der chthonische Gott Mədr.

Keine der drei Vorschläge geht vollständig auf, es muss demnach irgendeine Form von Kontamination im Sinne der Textkritik vorliegen. Betrachten wir und daher die beiden Gottesnamen näher. Im Äthiopischen bedeutet bəḥer üblicherweise „Land, Region”, während in den übrigen semitischen Sprachen baḥr „Meer, Fluss” heißt. Eine solche Deutung könnte sich aus der Aduiltana ergeben, wo Poseidon an prominenter Stelle erscheint. Freilch stellt DAE 6&7 (13); Kobishchanov, Aksum 229 zitiert Vycichl (kultiviertes Land vs. Land an sich) 28

ALTSÜDARABISCHE KULTE

261

der Verweis auf die Adulitana lediglich eine indirekte Textparallele dar. Der Blick in die Wörterbücher ist allerdings teilweise eher verwirrend als klärend. So erfahren wir, dass die beiden Lexeme bəḥer und mədr zur Bildung von Ausdrücken verwendet werden, die das Wetter betreffen. 29 Dabei werden die entsprechenden Syntagmen im Gəʾəz mit bḥr, im Amharischen mit mədr und im Arabischen mit dunyā “Welt” gebildet: Gəʿəz

masya bəḥer

es ist Abend

ṣəḥəw bəḥer

gutes Wetter

ṣabḥa bəḥer Amharisch

Arabisch

mədr mäšä

es ist Morgen

dunkel/Abend werden

mədr läqäqä

hell/Tag werden

ʾamsat ʾaddunyā

es ist dunkel

Wirklich weiter bringt uns diese Beobachtung allerdings kaum, es sei denn man wertet sie als Hinweis auf die lexikalische Austauschbarkeit der Lexeme, womit ihr Status als Synonyme erwiesen wäre. Immerhin gibt es noch einen weiteren Hinweis gegen die Interpretation von Baḥer als reinem Meeresgott. Der arabische Autor Ibn alFaqih al-Hamadhani (10. Jhd.) meint nämlich über die zwischen Nil und Rotem Meer lebenden Bēğa: „sie sind Anbeter von Idolen... [und] nennen Gott Baḥer” 30

Meines Erachtens ist mit bəḥer und mədr ein und derselbe Gott gemeint, nur werden offenbar unterschiedliche Aspekte angesprochen. Vielleicht handelt es sich um eine Amalgamierung des Numinosen: im Landesinnern verehrte man ursprünglich die Kane, Amharic-English Dictionary 323; Leslau, Comparative Dictionary, s.v. bahr. 30 Mikawy, Short Note 183f, vgl. Mousʾad, Al Maktaba 30. 29

262

FRANCIS BREYER

»Erde« und an der Küste das »Meer«. Im Zuge der Herausbildung des aksumitischen Reiches und der Inkorporierung von Adulis in den aksumitischen Machtbereich wurde der dortige Meeresgott ins eigene Pantheon integriert und mit dem eigenen Erdgott geglichen, was aufgrund der durchaus vorhandenen Namensähnlichkeit vergleichsweise einfach zu bewerkstelligen war. Gerade diese lautliche Ähnlichkeit bei trotzdem distinkten Unterschieden inhaltlicher Natur ist es möglicherweise, warum die beiden Aspekte des Hochgottes nie ganz zu einem verschmolzen. Ein typisch altsüdarabischer Aspekt der religiösen Praxis ist das häufige Auftreten bestimmter Symbole und Symboltiere, allen vorann der charakteristische sabäische Steinbock. Dieser findet sich auch häufig auf DoMT-zeitlichen Monumenten aus Abessinien. Das bekannteste darunter ist der Tempel von Yəḥa, wo sich Bruchstücke eines halbplastischen Steinbock-Frieses fanden. Die Schmalseiten des Kultschreines von Ḥawəlti sind mit Reliefs verziert, die aus Reihen von knienden Steinböcken bestehen. Weitere Symboltiere sind weniger gut bezeugt, etwa eine Stierdarstellung von Gobochela mit der Widmung an den Gott ʾAlmaqaḥ. 31 Vielleicht sogar noch charakteristischer als der Steinbock ist ein sehr spezielles Mondsymbol, bei dem ein Vollmond auf einer liegenden Mondsichel ruht. Im Gegensatz zu den Symboltieren, die sich in aksumitischer Zeit nicht mehr nachweisen lassen, war das Mondsymbol noch bis zur Christianisierung verbreitet. Wie die Stele von Mäṭära zeigt, konnte es oben auf monumentalen Stelen angebracht sein (RIE II:142; II:24). 32 Früher dachte man, alle Monumente des Stelenfeldes von Aksum hätten ursprünglich eine Plakette mit Mondsymbol getragen, später ein Kreuz. Nach dem Fund einer solchen Plakette mit einem Kopf in Frontalansicht in einem der monumentalen Gräber von Aksum, ist diese Ansicht widerlegt. Spannend ist vor allem die Belegsituation zum Mondsymbol in der Numismatik. Die aksumitischen Münzbilder zeigen den Kopf des regierenden Königs, über dessen Kopf das Mondsymbol oder ein Monogramm seines Namens angebracht ist – man denke hier an Leclant, Haoulti-Melazo Taf. XXXIX-XL; Drewes, Les inscriptions de Melazo 95–7. 32 Höfner, Ausgrabungen 221. 31

ALTSÜDARABISCHE KULTE

263

die Münzen Konstantins des Großen mit dem ChristusMonogramm. In der Tat wird die Mondsichel in der Regierung des Königs ʿĒzānā mit dessen Bekenntnis zum Christentum durch das Kreuz ersetzt. Solchermaßen kann erschlossen werden, dass es sich bei dem Mondsymbol um das zentrale Zeichen des aksumitischen Hochgottes Maḥrəm handeln dürfte. Damit sind wir bereits an einem Punkt angelangt, der bei Studien zur Christianisierung Äthiopiens zur Sprache kam: monotheistische Tendenzen lange vor diesem Wendepunkt der Geschichte Abessiniens. Bereits in der sabäischen Zeit sind Spuren von Monotheismus greifbar, etwa im gemischten äthiopisch-sabäischen Onomastikon. Sie werden nämlich teilweise mit einer allgemeinen Gottesbezeichnung gebildet, die bereits an einen absoluten Gebrauch gemahnen: 33 〈ʾl-ʾs1〉

〈ʾl-ʿgḏ〉

〈ʾl-ʿqb〉

〈lḥyʿṯṯr〉

*/ʾIl-ʾauš/

(RIE 27:3)

*/ʾil-ʿaqab/

(RIE 27:4)

*/ʾIl-ʿagaḏ/

*/ʾIly-ʿAṯtar/

(RIE 27:4)

(RIE 41)

Ein weiteres Indiz ist eine sehr auffällige Wendung in der aksumitischen Königsinschrift RIE 11, wo Maḥrəm als »Herr des Himmels« bezeichnet wird, als ʾegziʾa samāy (RIE 11:1). Die Wortfügung ist deshalb so interessant, weil mit ʾəgziʾab(ə)ḥer heute noch der christliche Gott gemeint ist. Der Befund im aksumitischen Abessinien ist damit vergleichbar mit demjenigen im gleichzeitigen Altsüdarabien. Dort haben sich die monotheistischen Tendenzen derart verstärkt, daß nicht mehr klar ist, ob man überhaupt noch von Polytheismus sprechen kann. Ausgehend von dem typischen Epitheton »der Barmherzige«, wie er noch im Islam lebendig ist, wurde in jüngster Zeit für diese Phase der altsüdarabische Religion der Begriff »Rahmanismus« geprägt. 34 Die Parallelen sind in der Tat Schneider, Documents Épigraphiques de l’Éthiopie V; Drewes, Inscriptions de l’Éthiopie antique 9 und 101; Schneider, Deux inscriptions. 34 Nebes, Sabäische Texte 2, 356ff., besonders 358f., Anm. 200f. Zu den Epithata vgl. Robin, Le judaïsme de Ḥimyar 114.; Müller, Himyar. 33

264

FRANCIS BREYER

frappierend. Man vergleiche etwa die bereits erwähnte Inschrift RIE 11 35

(1) [ba-]ḫayla | ͗əgzi͗a | samāy | [za⸗ba] | samāy | wa⸗mədr | mawāʾi | lita | [Bei] der Macht des Himmelsherrn, [der im] Himmel und auf Erden Macht hat über mich!

mit einer Inschrift (J 1028) des jüdischen Königs von Himyar Yūsuf ʾAsʾar Yaṯʾar genannt Ḏū Nuwās aus dem Jahre 523, wo es vom jüdischen Gott ganz entsprechend heißt: 36 mrʾ / smyn / w-ʾrḍn

„Herr des Himmels und der Erde”

Mit anderen Worten: ʿĒzānā war zu Beginn seiner Regierung sicherlich nicht ein Anhänger des Christentums; Monotheist könnte er jedoch durchaus bereits gewesen sein. Wenden wir uns nun dem offiziellen Kult zu, seinen Ausprägungen und seine gesellschaftlichen Verankerung, d.h. Fragen nach Heiligtümern und ihren Installationen, Kultfunktionären und Kulthandlungen. Die altsüdarabischen Kulte sind hier eindeutig besser erforscht als die nachfolgenden äthiopischen, insbesondere, was die Tempel angeht. Insgesamt ist eine Konzentration der Kultplätze auf die Region um Yəḥa and Ḥawəlti-Mälazo festzustellen, weitere Fundplätze wären Gobochela und Ǝnda Čärqos. Bei den Heiligtümern handelt es sich fast ausnahmslos um rechteckige Einraumtempel (Ḥawəlti, Yəḥa, Gobolcha; Wuqro), nur selten ist ein Temenos festzustellen (RIE, Wuqro). Mehrere der Kultplätze weisen eine sehr große Kontinuität aus, von der altsüdarabischen Zeit über die aksumitische bis hin zur Nutzung als christliche Kirche (Abba Ṗänṭälewon, Yəḥa, Ǝnda Čärqos). Kobishchanov, Aksum 235f. Nebes, Sabäische Texte 2, 356ff., besonders 358f., Anm. 200f. Zu den Epithata vgl. Robin, Le judaïsme de Ḥimyar 114. 35 36

ALTSÜDARABISCHE KULTE

265

Über das Kultinventar sind wir trotz der geringen archäologischen Aktivität auf jenem Gebiet vergleichsweise gut unterrichtet: da wären Statuen zu nennen (Ḥawəlti, ʿAddi Gälämo; RIE) sowie ein ganzer Schrein (Ḥawəlti). Der locus classicus für die aksumitische Zeit ist eine Passage aus den ʿĒzānā-Trilingue: aÄth.1 w=ʾ:bʾ=n | ʾ:k(19)t.t | l=mḥrm | z=wld=n | msl | z=wrq | ʾḥd | 1 | w=z=brr | ʾ(20)[ḥ]d | 1 | w=z=ṣrq | śls.t | 3 |

aÄth.2 w=ʾ:b(C25)ʾ=n | ʾ:kt.t | l=(C26)mḥrm | z=wl(C27)d=n | msl | z=w(C28)rq | ʾḥd | 1 | (C29) w=z=brr | ʾḥ(C30)d | 1 | w=z=ṣrq (C31) śls.t | 3 |

pSab.1 w=ʾ:bʾ(18)[=n]m | l-mḥrmm | ḏ=wld=n[m] | [ʾkt]tm | msl | ḏ-wrqm | (19) w=z=brwr | [ʾ]ḥd[m | w=z=ṣrqm | šls.tm |

pSab.2 w=ʾ:bʾ=nm | l-mḥrm=nm | l-z!=(20)wld=nm | ʾkttm | mṯl | ḏ=wrq | 1 | w=ḏ-br[w] (21)r | 1 | w=ḏ=ṣryq | šls.t | Grie.1 ὑπερ δε ε[ὐ](29)χαριστίας τοῦ μαι γεννήσαντος ἀνικήτου Ἄρεως (30) ἀνέθηκα αὐτῷ ἀνδριάντα χβυσοῦ α κ(αὶ) ἀργύ(31)ραιον ἔνα κ ξαλκοῦς 3, ἐπ᾽ἀγαθῷ. Grie.2 ὑπερ δε εὐχαριστίας τοῦ μαι γεννήσαντος (27) ἀνικήτου Ἄρεως ἀνέθηκα αὐτῷ ἀνδρι(28)άντα χβύσαιον ἓνα κ ἀργύραιον ἔνα κ χαλκαίοθς (29) τρὶς »Und nachdem wir Mahrem, der uns erschaffen hat, Lobpreis entgegengebracht hatten (in Form von) einer (Statue) aus Gold, einer aus Silber und drei aus Kupfer ...«

An Votivgaben sind Tierstatuetten und Häusermodel belegt (Ḥawǝlti), daneben sind als Installationen Ritualbänke (Yəḥa, Ḥawǝlti) und Podien (Yəḥa; Wuqro) bezeugt, aber auch Ritualbecken mit Abflussrinne (Wuqro) und vor allem eine große Zahl an meist kleineren Altären (Ḥawǝlti; Wuqro, Yəḥa,). Bemerkenswert ist vor allem der Befund von Gobo Fənṣəḥ, einer Grotte. Über das Personal in jenen Tempels wissen wir wenig: so ist nicht einmal bekannt, ob es eine eigene Priesterklasse gab. In Altsüdarabien übte der mukarrib priesterliche Funktionen aus – ob das bei ihrem abessinischen Pendant auch der Fall war, entzieht sich unserer Kenntnis.

266

FRANCIS BREYER

Die Kulthandlungen selbst sind – zumindest in aksumitischer Zeit – wieder relativ gut fassbar – hier ist neben dem Ahnenkult der tägliche Gottesdienst zu nennen, aber auch besondere Dankesoder Bittopfer, meist in Form von Votivgaben oder Gefangenen. Besonders die Inschriften des ʿĒzānā von Aksum geben uns Einblicke in die großen Opfer zum Dank für einen erfolgreichen Feldzug: da werden Statuen geweiht und Rinder geopfert, aber auch Menschen. Aus ethnoarchäologischen Analogien kann ferner die Funktion der großen Stelen von Aksum erschlossen werden: sie dienten wahrscheinlich dem Ahnenkult eines Klans, der im Wahlkönigtum einen Herrscher gestellt hatte. Sie sind in Form von Palastfassaden gestaltet und weisen eine Scheintür auf; vor dieser befinden sich Mulden als Opferbecken in den Deckstein eingelassen, teilweise mit einer Abflussrinne. All dies spricht dafür, dass beim Ahnenkult ein ähnlicher Versorgungsaspekt eine Rolle spielte, wie in der altägyptischen Religion. Auch dort finden wir eine Scheintür, durch die der Geist des Toten mit den Lebenden Kontakt aufnehmen kann, auch dort steht die Opferplatte direkt vor dieser Scheintür und auch dort ist das Grab als »Haus der Ewigkeit« gestaltet, d.h. es imitiert die Form eines Wohnhauses. Die Mulden sind ein recht eindeutiger Hinweis auf eine Form des Libationsopfers, die Rinnen könnten Hinweise auf ein Blutopfer sein – immerhin wurden den Inschriften nach Rinder geopfert. Brandopfer ist nicht direkt bezeugt – Paribeni hat in Adulis im sog. »Sonnentempel« verkohlte Knochen gefunden 37, jedoch erwies sich der vermeintliche Tempel als Kirche und der Fundkontext ist nicht bekannt. Außerdem sind m.W. aus dem aksumitischen Bereich keine Brandaltäre bekannt. Anders verhält es sich jedoch mit dem Räucheropfer – hier ist der archäologische Befund geradezu überwältigend, denn Räucheraltäre gehören zu einem der besonders charakteristischen Fundgattungen gerade für die altsüdarabische Periode. Neben dem Opfer gehörte das Stiften von Votivgaben zu den vielleicht beliebtesten Formen der Kultausübung. Die grausamste war das Menschenopfer, das in Aksum ohne jeden

37

Paribeni, R., Ricerche.

ALTSÜDARABISCHE KULTE

267

Zweifel belegt ist. 38 Die ʿĚzānā-Inschriften sind in dieser Beziehung ziemlich eindeutig (RIE 10:29f.; RIE 188; S. 261f.): ወቀፍ | ኡአኰቴተ | ለመሕ(30)ርም | ዘወለዶሙ | ለህም | ፻ | ወፄዋ | ፶|

ʾakoteta | la=maḥ(30)rəm | za=waladomu | lahm | 100 | waḍewā | 50 | „Und (der König) weihte ein Dankopfer dem Mahrem, der ihn gezeugt hat: Rindre 100 und Gefangene 50.”

Ob es im antiken Südarabien Menschenopfer gegeben hat, ist heftig umstritten. Immerhin gibt es wenige Hinweise darauf (Misʿal 7, 9; Jidhma 2–3.), doch ist der Befund alles andere als sicher. So streitet man sich, ob an den entsprechenden Stellen von »Tötung« oder nur von »Dienerschaft« die Rede ist. 39 In Meroë hingegen wird mit Sicherheit die Tötung von Gefangenen durch Pfählen dargestellt; man vergleiche dazu auch das sog. »Erschlagen der Auf der Feinde« auf altägyptischen Tempelreliefs. 40 archäologischen Ebene sind Gefolgschaftsbestattungen in nubischen Herrschergräbern seit dem 2. Jahrtausend v. Chr. (Kerma, C-Gruppe) bis hin zum 8. Jhd. n. Chr. nachgewiesen (Ballana; Qustul; X-Gruppe). Auch hier lohnt ein Vergleich zu noch in jüngster Zeit belegten Bräuchen bei verschiedenen kuschitischen Ethnien im südlicheren Äthiopien. Dort finden wir nämlich eine Verbindung zwischen der Opferung eines Menschen beim Herrschaftsantritt eines Königs und Klanheiligtümern mit monumentalen Stelen: 41 »Die dubuscha genannten heiligen Opfer- und Versammlungsplätze sind in Dita nicht wie in Dorse den Gauen, sondern den Klanen zugeordnet. Häufig handelt es sich bei den dubischa-Plätzen um imponierende megalithische Anlagen. Sie

Kobishchanov, Aksum 231f. Beeston, Miscellaneous (bes. 11–14 zu den Texten von Mis’al und Gidma, in denen von Menschenopfern die Rede ist); Henninger, Neuere Untersuchungen. 40 Zach, Ware Mensch. 41 Frobenius, Die Völker Süd-Athiopiens 223. 38 39

268

FRANCIS BREYER bestehen aus einer mannshohen, breiten Steinmauer, die den Platz im Kreis oder Oval umgibt. Auf dieser Mauer stehen im Abstand von einigen Metern mittlehohe Menhire. In der Mitte des Innenraumes, der auf der Nordseite durch einen Eingang zugänglich ist, befindet sich die Opferstelle, auf der vom Klanführer die Opferschlachtung vorgenommen werde.«

Fassen wir also zusammen: die Kultpraxis der Aksumiten weist zahlreiche Unterschiede zu den früheren altsüdarabischen Bräuchen auf; gleichzeitig bestehen Parallelen bei eben diesen zu vergleichbaren Phänomenen in zeitgenössischen Nachbarkulturen und aus der Ethnoarchäologie. Somit können diese Aspekte wie Ahnenverehrung und Totenkult an Steinstelen oder Menschenopfer getrost als »sudanische« Elemente der aksumitischen Religion gewertet werden. Der letzte größere Themenkomplex der altsüdarabischen Religion in Abessinien betrifft den sog. »Volksglauben«, d.h. die Äußerungen der Volksfrömmigkeit. Darunter fallen Anklänge an Naturreligionen wie Astral- und Schlangenkulte, aber auch Bestattungssitten, Ahnenkult und Jenseitsvorstellungen. Bereits Enno Littmann hat Astralmythen Nordäthiopiens (Tigray) gesammelt 42 und sie mit den aus aksumitischen und altsüdarabischen Quellen bekannten Bräuchen verglichen. 43 Besonders ist hier wie dort die Verehrung des Mondes: • Bei einer Mondfinsternis stand alles still, um Gott anzurufen, den Mond nicht sterben zu lassen. • Beim Ende der Finsternis wird die Genesung des Mondes ausgelassen gefeiert. Es gibt auch den Glauben, dass nach einer solchen Mondfinsternis eine wichtige Person stirbt. • Die Mondphasen spielen ebenfalls eine wichtige Rolle: der zunehmende Mond bringt Glück und bei Neumond werden magische Praktiken ausgeführt, danach heißt es, »Arro« bzw. »Helial« habe sich wieder erholt.

42 43

Littmann, Sternensagen, vgl. Littmann, Abessinien 55f. Littmann, Abessinische Parallelen; Littmann, Abessinien 55f.

ALTSÜDARABISCHE KULTE

269

Zur Rekonstruktion der vorchristlichen Glaubensvorstellungen lassen sich jedoch nicht nur Erzählungen und Märchen heranziehen, sondern auch Hagiographien. So hat Carlo Conti-Rossini aus diesen den Kult eines vorchristlichen Mondgottes namens Gad herausgearbeitet. 44 Auf ähnliche Art und Weise lässt sich ein vorchristlicher Schlangenkult in Abessinien rekonstruieren. 45 Der Hauptzeuge (KBr ngST Kəbrä nägäÈt) dafür ist die wahrscheinlich ursprünglich kuschitische 46 Legende vom Schlangenkönig Arwe (ʾarwē »Schlange«) 47, die Vorläufer in der altägyptischen Geschichte des Schiffbrüchigen aufweist. 48 Außerdem existieren verschiedenen Äußerungen klassischer Autoren über die Verehrung von Schlangen an der Südwestküste des Roten Meeres. Zwar meint Heinzgerd Brakman, die Evidenz für einen abessinischen Schlangenkult in vorchristlicher Zeit seien insgesamt zu dürftig 49, dem vermag ich jedoch nicht zuzustimmen. Wie zuvor bei den Kultstelen und den Menschenopfern zeigt sich auch hier eine »sudanische« Verbindung. So wird in der ʿĒzānā-Inschrift RIE 11 die Gefangennahme eines nubischen »Schamanen« (kujur) 50 bechrieben, die typischerweise als Insignien ihres Ranges einen Schlangenring tragen. Nicht umsonst wird dieser Ring ausdrücklich erwähnt - direkt nach der Schilderung von der Gefangennahme zweier Späher und des ʾəngabenāwe werden die gefallenen Anführer der bekriegten Noba aufgeführt (Z. 26–27): wa=ʾəlla | mōt{ō} | magabt (26) danōkwə | 1 | dagale | 1 | ʾanakwə | 1 | ḥawāre | 1 | karkārā | 1 | māri=hōmu | 1 | [ʾaqwə](27)salu | wa=salabə=wō | qədāda | bərur |

Conti-Rossini, Gad e il dio Luna 53. Hyatt, The church in Abyssinia 27f. 46 Haberland, Untersuchungen 132. 47 Littmann, La lagenda; DAE I:39. 48 Lanczkowski, Aethiopia; Maksimov, O vzaimodeistvii religii i fol’klora. 49 Brakman, Axomis (Aksum), bes. 740. 50 Nadel, A Shaman Cult; Hakesworth, Description; Kauczorm, The Afitti Nuba. 44 45

270

FRANCIS BREYER wa=ḥəqata | warq | kōnu | magabt | ʾəlla | (28) mōtu | 5 | wa=māri | 1 |

»Diejenigen, die gefallen sind, waren die Anführer: Danokwe (1x), Dagale (1x), Anakwe (1x), Haware (1x), Karkara (1x) (und auch) ihr Priester (1x), (meine Männer) hatten (ihn) verwundet und ihm den Silberschmucke und einen (/den?) goldenen Ring abgenommen. Es waren (also) die Anführer, die starben, 5 (an der Zahl) und ein ›Magier‹.«

Dies wiederum erinnert wieder an die zuvor erwähnte Legende von ʾArwe, denn dieser erste (göttliche) König Äthiopiens wird von einem Helden mit Namen Agäbo, d.h. von einem solchen ʾəngabenāwe! Weitere Indizien stammen aus der Ethnographie. In dem bereits erwähnten Werk von Leo Frobenius über die kuschitischen Völker Süd-Äthiopiens begegnet uns mehrfach die Verehrung von Schlangen 51, etwa bei den Dorse (S. 178) »Der demutsa-Priester schachtet ein Schaf, dessen Schädel zur Bambusstange zeigen muß und dessen Blut von den heiligen Schlangen aufgeleckt wird, die in dem kleinen Opferhein ihre Schlupfwinkel haben. [...]«

oder den Djandjero (S. 348):

»Die Djandjero waren des Glaubens, daß der Thronfolger während seine Aufenthalts in dem Zabga Wäldchen von einer Python-Schlange (dĕrbo) und von Löwen und Leoparden aufgesucht wurde, aus deren Verhalten zu entnehmen war, ob die Wahl auf eine geeignete oder eine ungeeignete Persönlichkeit gefallen war. Im ersten Falle ringelte sich die Schlange um den Thronfolger und die Feliden leckten ihm die Hände und legten sich zu seinen Füßen schlafen. War der gewählte Kandidat aber ein verworfener Mensch, so wurde er von der Schlange gebissen und von den Löwen und Leoparden aufgefressen.” [...] Wenn der Thronfolger nach der Waschung aus der Quelle herausgestiegen war streifte man ihn den goldenen Armring über und setzte ihm die Krone aufs Haupt.«

51

Frobenius, Die Völker Süd-Athiopiens 223.

ALTSÜDARABISCHE KULTE

271

Es scheint auch in Südarabien Schlangenkulte gegeben zu haben, 52 allerdings scheint es kein besonders charakteristischer Zug der klassischen altsüdarabischen Religion gewesen zu sein. Andererseits ist in besonders schlangenreichen Regionen immer damit zu rechnen, dass man den oft für Mensch und Tier gefährlichen Schlangen ein gewisser Respekt, wenn nicht sogar Verehrung, entgegen gebracht wird. Mit anderen Worten: auch die Sabäer könnten im Alltag Schlangen verehrt haben. Aber selbst wenn dies der Fall war, heißt das noch lange nicht, das der abessinische Brauch aus Südarabien stammt, genauso wenig wie aus dem Alten Ägypten, wo es zahlreiche Schlangengottheiten gab. Ebenfalls naheliegend und keinesfalls ein Hinweis auf altsüdarabischen Einfluss sind die zahlreichen Hinweise auf die vorchristliche Verehrung von Bergen und Felsen. Bestes Beispiel dafür ist die Präferenz, Kultstätten auf Bergen zu errichten, wie sie in Tigray vorherrscht – man denke nur an das Kloster Däbrä Damo. 53 Abba Ṗänṭälewon bei Aksum kann sogar auf eine Kontinuität von der südarabischen Zeit bis heute vorweisen. In Däbrä Libanos wird der dort verehrte Heilige Livianus (Lobanos) auf »Herr von Däbrä Mata« genannt, d.h. mit dem zuvor dort wahrscheinlich angesiedelten Kult für den personifizierten Berg Mata identifiziert. Ein weiteres Indiz auf vorchristische Bergheiligtümer liegt in den Vorstellungen von einem »heiliger Berg« (däbr qəddus), auf dem die gefallenen Engel vor der Sintflut lebten. 54 Den zumeist wichtigsten Aspekt religiöser Vorstellungen bilden Bestattungssitten und Jenseitsvorstellungen, worunter auch der Ahnenkult zu zählen ist. Wie schon seit langem als gesichert gilt, repräsentieren die Stelen der aksumitischen Könige Grabpaläste im wörtlichen Sinne des Wortes, d.h. die monumentale Umsetzung eines Hausmodells für den Grabinhaber oder – altägyptisch gesprochen – ein »Haus der Ewigkeit« (= Grab). 55 Benoist, An Iron Age; Potts, Revisiting the Snake Burials. Im Koptischen bedeutet “Berg” ebenfalls gleichzeitig auch “Kloster” (ΤΟΟΥ < altäg. ḏw “Berg”). 54 Leslau, Comparative Dictionary 121. 55 Daum, From the Queen of Saba 11f; Lewcock, The medieval architecture of Yemen; Manzo, Considerazioni. Gänzlich abwegig ist ein 52 53

272

FRANCIS BREYER

Die Parallelen zwischen diesem Aspekt der aksumitischen Religion und altägyptischen Glaubensvorstellungen sind auffällig. So gab es auch in Ägypten Formen von Grabbauten, die Wohnhäuser imitierten, angefangen von den Sarkophagen in der Form einer Palastfassade über die sog. Mastaba mit Nischengliederung wie bei frühzeitlichen Palästen bis hin zur Scheintür als zentralem Kommunikationspunkt zwischen Totem und Hinterbliebenen. Es dürfte kein Zufall sein, daß auch die aksumitischen Stelen und Gräber solche Scheintüren aufweisen. Ganz frappierend ist jedoch eine Stele vom Elitefriedhof in Aksum, auf der sogar die ägyptische Hieroglyphe für »Leben« angebracht worden war. 56 Ob man aus diesem Umstand jedoch weiterreichende Schlüsse ziehen darf, erscheint mir zumindest zweifelhaft – direkte Verbindungen zwischen den aksumitischen und den altägyptischen Jenseitsvorstellungen wird es wohl kaum gegeben haben. Wohl aber ist anzunehmen, daß diese Vorstellungen in einem gemeinsamen Kulturraum entstanden sind und für den gesamten ›sudanischen‹ Raum Nordostafrikas charakteristisch sind. Das kulturgeschichtliche Bindeglied zwischen dem Niltal und den Horn von Afrika ist die Kultur Meroës und auch gehört die Scheintür zur üblichen Ausstattung der königlichen Grabanlagen. 57 Wenn wir uns nun die aksumitischen Scheintüren genauer betrachten, so ist vor allem die beim Umstürzen in mehrere Teile zerbrochene größte aller aksumitischen Stelen von Interesse. Sie ist auf Vorder- und Rückseite mit dem Palastfassaden-Dekor versehen und jeweils mit einer Scheintüre. Wahrscheinlich bei dem Versuch, dem Monolith aufzurichten, war sie zerbrochen und auf ein bereits bestehendes Monument gestürzt. Dabei fiel sie auf eine der Scheintüren, die demnach nicht mehr ›zugänglich‹ war. Da jedoch die Stele offensichtlich in zerbrochenem Zustand ihren Zweck als ›Grabpalast‹ nicht mehr erfüllen konnte, wurde der Griff der noch zugänglichen Scheintüre abgearbeitet und somit rituell unbrauchbar gemacht. Vergleich mit indischen Pagoden, wie diskutiert bei Schoff, The Periplus of the Erythraeoan Sea 61–6. 56 Anfray, L’archéologie. 57 Dunham, The royal cemeteries of Kush (RCK) IV; Dunham, The royal cemeteries of Kush (RCK) V.

ALTSÜDARABISCHE KULTE

273

Wie in Ägypten waren die Scheintüren sicherlich der Ort, an welchem dem Toten geopfert wurde, nicht umsonst findet man vor ihnen auf den Bodenplatten Installationen zum Opferkult wie Abflußrinnen und eingearbeitete Libationsbecken. Vor allem das nach der Scheintür benannte tomb of the false door zeigt, welche Bedeutung diesem Bauteil zugemessen wurde: er ist zentrales Element der gesamten Anlage, deren zugänglichen Teile mit der Scheintüre übrigens nicht in Verbindung stehen, sondern – wie in Meroë - über einen Gang, der vor dieser nach unten führt. Das Prinzip der Scheintür bedarf eigentlich keiner Erklärung. Sie ist ausschließlich für den Totengeist durchlässig; über sie kann er Kontakt mit der Welt der Lebenden aufnehmen, die vor ihr das Totenopfer darbringen. Wenn hier nun die königlichen Grabanlagen in Aksum als Argument für die Ergrüngung der Jenseitsvorstellungen herangezogen wird, so unter der Annahme, daß diese für König, Elite und Bevölkerung grundsätzlich vergleichbar waren. Ob dies für einen weiteren Aspekt dieser Nekropole gilt, kann bezweifelt werden. Die Stelenanlagen in Aksum waren nämlich wahrscheinlich keine Individualgräber, sondern wohl eher Kultstätten für den königlichen Ahnenkult, d.h. Begräbnisplätze derjenigen Clans, welche den König stellten. Für das Prinzip der Kollektivoder Mehrfachbestattung gibt es zwar keinen definitiven archäologischen Beweis – die Bestattungen sind zu gestört, um Definitives sagen zu können, allein die Anzahl der unterirdischen Kammern und Särge spricht dafür. Auffällig ist ferner das Fehlen jeglicher namentlicher Kennzeichnung der Gräber. Kurioserweise wurden sie in der Moderne mit fiktiven Zuweisungen an die glorreichen Herrscher Aksums belegt. Die Reste von Grabbeigaben 58 sprechen ihre eigene Sprache: wie im Diesseits soll es auch im Jenseits dem König an nichts mangeln, weder an Trinkgefäßen, noch an einem Thron aus Elfenbein! 59 Aufgrund eines Zusammenhang zwischen den sog. ›Bisi-Namen‹ (bəʿəsi »Mann (von...)«) des Königs mit Namen von Truppenteilen einerseits und ReWendoski, Ziegert, Nosnitzin and Uhlig, Eine Grabbeigabe. Zu den Beigaben vgl. ausführlich die Publikation der Ausgrabungen Phillipson, Archaeology at Aksum; Phillipson, Ancient Ethiopia.; MunroHay, Excavations at Aksum. 58 59

274

FRANCIS BREYER

gionsbezeichnungen andererseits, kann postuliert werden, daß die Clans der aksumitischen Oberschicht ihre Basis in verschiedenen Landesteilen hatten. 60 Damit spiegelt sich in den Clananlagen die politische Struktur des Reiches und weniger die Begräbnissitten. Doch ganz lassen sich beide Bereiche nicht trennen. Im Reich von Meroë sind die Nekropolen von Meroë und Napata wahrscheinlich auch verschiedenen Dynastien zuzuordnen. 61 Auf der politischen Ebene bestanden nun Gemeinsamkeiten zwischen Aksum und Meroe, die ich im Abgrenzung an vorderasiatische oder ägyptische Traditionen widerrum als typisch ›sudanisch‹ werten möchte, das Wahlkönigtum. Hier wie dort gab es wahrscheinlich einen Kreis von Elektablen, unter denen ein primus inter pares gekürt wurde. 62 Nicht nur die ›Bisi-Namen‹ der Könige von Aksum weisen für diesen Staat in eine solche Richtung, 63 sondern auch das spätere äthiopische Herrschertum (nəguśa nägäśt ›König der Könige‹) mit seinen Regionalkönigen. Als ethnoarchäologische Parallele dienen erneut ethnographische Zeugnisse von den Djandjero, bei denen sich ebenfalls Steinkreise als Orte der königlichen Ahnenkultes finden. 64 Ähnliche Stelenkreise fand R. Fattovich bei Grabungen in der Nähe von Kassala. 65 Im Ganzen scheint es sich um eine Tradition zu handeln, die in Nordostafrika weit verbreitet war. 66 Der Zusammenhang zwischen Stelen und Ahnenkult wird erwiesen durch die Inschrift auf der Stele von Mäṭära (RIE I, Nr. 34), wo es zu Beginn heißt 67: 1 z-ḥwlt | z ʾgbr 2 ʾgz | l-ʾbwh. etc.

“Dies hier ist die Stele, die ʾAgaz 68 für (das Gedenken an) seine

De Blois, Clan-names. Hormann, Beiträge. 62 Vinogradov, Diodorus. 63 De Blois, Clan-names. 64 Frobenius, Die Völker Süd-Athiopiens 223. 65 Fattovich, The stelae of Kassala. 66 Fattovich, Some Remarks; Phillipson, The Significance; Joussaume, Les cultures mégalithiques. 67 Kropp, Die Stele von Mäṭära. 68 Zu gawz “Parfum”?, vgl. hebr. ʾegoz, dazu Noeldecke, Neue Beiträ60 61

ALTSÜDARABISCHE KULTE

275

Väter hat anfertigen lassen...”

Weniger klar ist ein Hinweis auf die Kultpraxis im Zusammenhang mit den Stelen, der aus der Etymologie des äthiosemitischen Wortes für »Stele« resulitert, ḥawəlt (pl. ḥawlät; ʾaḥwälät), so belegt in den Inschriften von Mäṭära (RIE I, Nr. 34:1 zḥwlt) 69 und Anza [Zeile 2]) 70. Maria Höfner und Wolf Leslau haben dieses Lexem mit einer Wurzel ḥwl »umrunden« 71, und damit inhaltlich mit dem arab. ṭawāf in Verbindung gebracht. Letzteres hat seinen Ursprung in der Pilgertradition Altsüdarabiens. Doch sei dieser Strang nicht überbewertet! Was also läßt sich über die altsüdarabischen Kulte in Abessinien sagen? Der Staatskult in Abessinien zeigt zunächst ein gänzlich altsüdarabisches Gesicht. Mit der Emanzipation Aksums stellt sich allerdings heraus, daß dies quellenbedingt verzerrt ist, da nach Abblättern der altsüdarabischen Überlagerungsschicht eigene, genuin abessinische Gottheiten in Erscheidnung treten. Die Beziehungen zum südarabischen Raum bleiben gleichwohl sehr eng, bis hin zum Rahmanismus bzw. der Christianisierung. Die abessinische Kultpraxis läßt sich vergleichsweise gut rekonstruieren, allein Informationen über Kultpersonal sind kaum vorhanden. Trotz der Präferenz bisher vorgelegter archäologischer und inschriftlicher Quellen zu Aksum läßt sich viel über die Volksfrömmigkeit aussagen, über die Verehrung von Astalnumina, Schlangen und Bergen. Diese finden sich in ähnlicher Form auch in Südarabien, ohne daß freilich ein direkter Zusammenhang bestehen muß. War dies noch durch eine Extrapolierung aus den späteren äthiopischen Quellen ergründet worden, so basiert die Rekonstruktion von Jenseistvorstellungen vornehmlich auf ethnoarchäologischen Parallelen in den Nachbarkulturen und dem damit einhergehenden Posge 43, Anm. 5; Koehler and Baumgartner, Hebräisches und aramäisches Lexicon 10. Dagegen spricht, daß man in dieser frühen Zeit nicht von einer Schreibung aus ausgehen würde. 69 Kropp, Die Stele von Mäṭära. 70 Kropp, Monumentalised Accountancy. 71 Zitiert bei Kobishchanov, Aksum 226.

276

FRANCIS BREYER

tulat eines gemeinsamen ›sudanischen‹ Kulturhorizontes in Nordostafrika.

BIBLIOGRAPHY AE = Annales d’Ethiopie EncAeth = Encyclopaedia Aethiopica IBAES = Internet-Beiträge zur Ägyptologie und Sudanarchäologie PSAS = Proceedings of the Seminar for Arabian Studies RIE = Bernand, E., A.J. Drewes and R. Schneider, Recueil des Inscriptions de l’Ethiopie des Périodes Pre-Axoumite et Axoumite, Paris 1991ff. RSE = Rassegna di Studi Etiopici TUAT NF = Texte aus den Umfald des Alten Testaments Neue Folge Anfray, F., L’archéologie d’Axoum en 1972, in Paideuma 18 (1972), 60–78. Beeston, A.F.L., Sayhadic divine designations, in PSAS 21 (1991), 1–5. Beeston, A.F.L., Miscellaneous epigraphic notes II, in Raydan 5 (1988) 5–32. Benoist, A., An Iron Age II Snake Cult in the Oman Peninsula: Evidence from Bithnah (Emirate of Fujairah), in Arabian Archaeology and Epigraphy 18 (2007), 34–54. Brakman, H.. Axomis (Aksum), in Reallexikon für Antike und Christentum, Supplement, 1992, 718–810. Cerulli, E., Storia della letteratura etiopica, Capitolo I. Il periodo aksumita, Mailand 1956. Conti-Rossini, C., Gad e il dio Luna in Etiopia, in Studi e materiali di storia delle religioni 4, Rom 1928, 49–57 Daum, W., From the Queen of Saba to a modern state, in W. Daum (ed.), Yemen: 3000 Years of art and civilisation in southern Arabia, Innsbruck 1988, 9–31. De Blois, F., Clan-names in ancient Ethiopia, in Die Welt des Orients 15 (1984), 123–5. Dillmann, A., Biblioa V.T, Aethiopica 5 (1894), 117.

ALTSÜDARABISCHE KULTE

277

Drewes, J.A., Les inscriptions de Melazo, in AE 9 (1959) 83–100. Drewes, J.A., Inscriptions de l’Éthiopie antique, Leiden 1962. Dunham, D., The royal cemeteries of Kush (RCK) IV, Royal tombs at Meroe and Barkal, Boston 1957. Dunham, D., The royal cemeteries of Kush (RCK) V, The west and south cemeteries at Meroe, Boston 1963. Fattovich, R., The stelae of Kassala: a new type of funerary monuments in the Eastern Sudan, in Archeologie du Nil Moyen 3 (1988), 55–63. Fattovich, R., Some Remarks on the Origins of the Aksumite Stelae, in Annales d’Ethiopie 14 (1987), 43–69. Frantsouzoff, S., Ḏāt-Himyʾm in EncAeth, ii, 107f. Frantsouzoff, S., HWBS in EncAeth, iii, 94f. Frobenius, L., Die Völker Süd-Athiopiens, Frankfurt a. M. 1912. Haas, Ch., Mountain Constantines: The Christianization of Aksum and Iberia, in JLA 1/1 (2008), 101–26. Haberland, E., Untersuchungen zum äthiopischen Königtum (Studien zur Kulturkunde 18), Wiesbaden 1965. Hakesworth, D., Description of a ceremony by which a Nuba Cheif became a kujur, in Sudan Notes and Records 23 (1932), 345–7. Henninger, J., Neuere Untersuchungen über Menschenopfer bei semitischen Völkern, in R. Stiegner (ed.), Al-Hadhud, Graz 1981, 65–78. Höfner, M., Orts- und Götternamen in Südarabien, in A. Leidlmair (ed.), H. von Wissmann-Festschrift, Tübingen 1962, 181–5. Höfner, M., Ausgrabungen in Abessinien, in Archiv für Orientforschung 18 (1957), 221–2. Hormann, I., Beiträge zum meroitischen Königtum, Brüssel 1971. Hyatt, H.M., The church in Abyssinia, London 1928. Jamme, A., Ethiopia, in Bibliotheca Orientalis 15 (1957), 79. Joussaume, R., Les cultures mégalithiques de l’Éthiopie, in X. van der Stappen (ed.), Aethiopia, Tervuren 1996, 58–75. Kane, T.L., Amharic-English Dictionary, Wiesbaden 1990. Kauczorm, D., The Afitti Nuba of Gebel Dair and their rela-

278

FRANCIS BREYER

tionspih to the Nuba proper, in Sudan Notes and Records 6 (1923), 1–34. Kobishchanov, Y.M., Aksum, Philadelphia 1979. Koehler, L. and W., Baumgartner, Hebräisches und aramäisches Lexicon zum Alten Testament, Leiden ³1967. Kropp, M., Die Stele von Mäṭära, in S. Wenig et al. (eds.), In Kaiserlischem Auftrag. Die Deutsche Aksum-Expediton 1906 unter Enno Littmann (Forschungen zur Archäolgie Außereuropäischer Kulturen 3.1), Aichwald 2006, 321–8. Kropp, M., Monumentalised Accountancy from ancient Ethiopia: The stele of Maryam ʿAnza, in W. Smidt and S. Wenig (eds.), Proceedings of the Second Internatinal Littmann Conference, Aksum January 2006 (im Druck). Lanczkowski, G., Aethiopia, in Reallexikon für Antike und Christentum, Supplement I, 1958, 94–134. Leclant, J., Haoulti-Melazo (1955–1956), in Annales d’Ethiopie 3 (1959), 43–82. Leslau, W., Comparative Dictionary of Gəʿəz, Wiesbaden 21991. Lewcock, R., The medieval architecture of Yemen, in: W. Daum (ed.), Yemen: 3000 Years of art and civilisation in southern Arabia, Innsbruck 1988, 204–211. Littmann, E., La lagenda del dragone di Aksum in lingua tigrai, in RSE 6 (1947), 42–5. Littmann, E., Abessinien, Hamburg 1935. Littmann, E., Sternensagen und Astrologisches aus Nordabessinien, in Archiv für Religionswissenschaft 11 (1908) 298–319. Littmann, E., Abessinische Parallelen zu einigen altarabischen Gebräuchen und Vorstellungen (Beiträge zur Kenntnis des Orients 6), Berlin 1908. Loundine, A. and J. Ryckmans, Nouvelles données sur la chronolgie des rois de Saba et Ḏu Rayḏān, in Le Muséon 77 (1964), 411. Maksimov, E.N., O vzaimodeistvii religii i fol’klora na primere obraza volshebnono zmeya iz “Skazki o poterpevshem korablekrushenie” [On the interaction of religion and folklore, from the example of the magic snake in the “Story of the Shipwrecked Sailor”], in Drevnii Vostok, Sbornik 2, Ak. Nauk

ALTSÜDARABISCHE KULTE

279

SSSR, Moskau 1980, 120–26. (Abstract in Französisch 212). Manzo, A., Considerazioni sull’architettura dell’Etiopia antica, in RSE 39 (1995), 155–72. Mikawy, F., Short Note abut Aksumite Influence on Beja Cult in the Middle Ages, in JES 23 (1974), 183f. Mous’ad, M., al-Soudania al ‘Arabia (Bibliothéque Soudano-Arabe), Kairo 1972. Müller, W., Himyar, in Reallexikon für Antike und Christentum 15 (1991), 303–31. Munro-Hay, S.C., Arabia. Relations in ancient times, in EncAeth, i, 294–300. Munro-Hay, S.C., Aksum, an African Civilization of Late Antiquity, Edinburgh 1991. Munro-Hay, S., Excavations at Aksum: an account of research at the ancient Ethiopian capital directed in 1972–1974 by the late D. Naville Chittick, London 1989. Nadel, S.F., A Shaman Cult in the Nuba Mountains, in Sudan Notes and Records 24 (1941), 85–112. Nebes, N., Sabäische Texte 2, in TUAT NF, ii, Gütersloh 2005. Nielsen, D., Der dreieinige Gott in religionshistorischer Beleuchtung I. Die drei göttlichen Personen, Kopenhagen 1922. Noeldecke, Th., Neue Beiträge zur semitischen Sprachwissenschaft, Strassburg 1910. Paribeni, R., Ricerche nel luogo dell’antica Adulis, in Monumenti Antichi, Reali Accademia dei Lincei 18 (1907) 437–572. Phillipson, D.W., Archaeology at Aksum, Ethiopia, 1993–7, 2 vols., London 2000. Phillipson, D.W., Ancient Ethiopia. Aksum, Its Antecedents and Successors, London 1998. Phillipson, D.W., The Significance and Symbolism of Aksumite Stelae, in Cambridge Archaeological Journal 4 (1994), 189–210. Pirenne, J., L’inscription “Ryckmans 535” et la chronologie sudarabe, in Le Muséon 69 (1956)б 166–80. Potts, D.T., Revisiting the Snake Burials of the Late Dilmun Building Complex on Bahrain, in Arabian Archaeology and Epigraphy 18 (2007), 55–74.

280

FRANCIS BREYER

Robin, C.J., Le judaïsme de Ḥimyar, in Arabia 1 (2003), 97–172. Schneider, R., Deux inscriptions sudarabiques de Tigré, in BiOr 30 (1973) 385–7. Schneider, R., Documents Épigraphiques de l’Éthiopie V, in AE 10 (1976) 81–93. Schoff, W.H., The Periplus of the Erythraeoan Sea, New York 1912. Sima, A., Almaqah, in EncAeth I, 207. Sima, A., Die “sabäische” Version von König oĒzānās Trilingue RIE 185 und RIE 185bis, in Archiv für Orientforschung 50, 2003– 2004, 269–84. Zach, M.H., Ware Mensch: „Heiliger Akt” oder profaner Tod? Menschenopfer im religiösen Leben Meroes?, in M. Fitzenreiter (ed.), Das Heilige und die Ware, IBAES VII, London (2007), 193–202. Vinogradov, A., Diodorus and the election ot kings of Meroë, in Meroitica 10 (1989), 353–64. Wendoski, M., H. Ziegert, D.Nosnitzin and S. Uhlig, Eine Grabbeigabe aus Aksum (Bərit oAwdi), in Aethiopica 4 (2001), 191–4.

LES ÉVOLUTIONS DU CALENDRIER DANS LE ROYAUME DE ḤIMYAR : QUELQUES HYPOTHÈSES 1

CHRISTIAN JULIEN ROBIN

Le royaume de Ḥimyar, dont la résidence royale se trouvait à Ẓafār, dans les montagnes du Yémen, a dominé l’ensemble de l’Arabie du Sud à partir de 300 de l’ère chrétienne (désormais è. chr.), puis une portion plus ou moins vaste de l’Arabie déserte entre le milieu du 4e s. et 565 environ. Il a donc été la puissance dominante en Arabie pendant l’Antiquité tardive, contrecarrée occasionnellement par les souverains naṣrides d’al-Ḥīra dans la vallée de l’Euphrate. Après le rejet du polythéisme vers 380, la religion officielle du royaume de Ḥimyar a été tout d’abord un monothéisme fortement teinté de judaïsme, que nous avons proposé d’appeler « judéomonothéisme » 2, puis, de 530 à 570 / 575 (sous les règnes de Sumūyafaʿ Ashwaʿ, puis d’Abraha et de ses fils), le christianisme. L’objet de cette contribution est d’examiner si ces changements d’orientation religieuse ont eu une incidence sur le calendrier. Nous allons traiter successivement du calendrier traditionnel de Ḥimyar (mois, décompte des années et intercalation), puis des indices qui suggèrent de possibles réformes à l’époque judéo-monothéiste et sous Abraha. Nous conclurons avec un bref examen des autres Abréviations : è. chr., « de l’ère chrétienne » ; ḥim., « de l’ère ḥimyarite » ; maḍḥ., « de l’ère de Maḍḥàm » ; radm., « de l’ère de Radmān » ; av., « avant » ; ap., « après » ; s., « siècle ». Pour les toponymes et ethnonymes, se reporter à la Carte 1, Le Yémen antique. 2 Robin 2015 c. 1

281

282

CHRISTIAN JULIEN ROBIN

calendriers d’Arabie méridionale qui conduit à la conclusion que tous ces calendriers, à une époque donnée, suivaient un même modèle et étaient probablement identiques dans les grandes lignes, à l’exception des noms de mois. Cette contribution à la chronologie de l’Arabie ancienne, en l’honneur de Michael Macdonald, rend naturellement hommage à un chercheur qui a œuvré toute sa vie pour fonder l’épigraphie arabique sur des bases solides, alliant méthode et réflexion. A. LA DÉSIGNATION DES MOIS ET LA DÉTERMINATION DES ANNÉES DANS LE ROYAUME DE ḤIMYAR Pour reconstruire les calendriers de l’Arabie méridionale avant l’Islam, nous ne disposons que des dates relevées dans les inscriptions 3. Aucun document explicatif ou plus fourni, comme un décret réglementant le calendrier ou une séquence de dates, ne nous est parvenu. Tout au plus dispose-t-on, dans deux inscriptions ḥimyarites (al-Miʿsāl 2 4 et ʿAbadān 1 5) et dans une bilingue sabéonabatéenne (DAI Ṣirwāḥ 2004-12B + 2004-6 + 2002-83 + 1993/3–x, inédit 6), de datations se référant à deux systèmes différents. La bibliographie sur les calendriers et les ères utilisés dans les inscriptions de l’Arabie méridionale est assez brève : voir principalement Beeston 1956, Lundin 1959, Beeston 1974, Pirenne 1974, Robin 1981, Robin 1998, Nebes 2004, Robin 2012 a, al-Iryānī & ʿAbd Allāh « alTaqwīm al-ḥimyarī ». 4 al-Miʿsāl 2 est datée du mois de madhraʾān de l’année 179 du comput d’Abīʿalī b. Ratʿ et de l’année 363 du comput de Mabḥūḍ ibn Abḥaḍ : voir ci-dessous, p. 15. La vocalisation de l’anthroponyme Ratʿ, qui n’est pas attesté dans la tradition arabo-islamique, est arbitraire. 5 ʿAbadān 1 est datée de dhu-madhraʾān 470 dans le calendrier et l’ère ḥimyarites [comput de Mabḥūḍ (appelé ici erronément Yabḥuḍ) ibn Abḥaḍ] et dans le calendrier et l’éponymat du Ḥaḍramawt : voir cidessous, p. 16. 6 Le texte sabéen est daté avec le mois (dhu-Sabaʾ) et l’éponyme sabéens, tandis que le texte nabatéen l’est avec le mois (ṭbt) et l’année de règne du souverain (l’an 3 de Ḥārithat roi de Nabaṭ bien-aimé de son peuple). 3

LES ÉVOLUTIONS DU CALENDRIER

283

Les dates donnent en général le mois et l’année. Pour identifier cette dernière, les Sudarabiques se sont d’abord référés à des personnages qu’on appelle « éponymes » parce qu’ils donnent leur nom à l’année ; puis, à partir du début de l’ère chrétienne, ils ont commencé à se servir d’une ère. Plusieurs ont été en usage, avant que celle de Ḥimyar ne finisse par s’imposer. Concernant les noms de mois, les inscriptions en mentionnent une centaine. Regrouper ces mois en « calendriers », c’est-à-dire en listes de douze mois formant une année, n’est pas toujours aisé. En croisant la provenance, la langue, les divinités, les institutions et le nom des éponymes, on sait reconnaître aujourd’hui une dizaine de calendriers différents. Mais c’est seulement pour ceux des grands royaumes, Sabaʾ et Ḥimyar, mais aussi, de manière moins assurée, Qatabān et Maʿīn, qu’on possède une liste complète des douze mois de l’année 7. Quand on a isolé la liste des mois propres à un royaume, il reste à classer ces mois et à déterminer lequel est le premier de l’année. En général, les données épigraphiques ne le permettent pas. Heureusement, dans un cas, celui du calendrier de Ḥimyar, on dispose de textes médiévaux (des 13e et 14e s.) qui, traitant du calendrier agricole yéménite en se référant aux mois juliens solaires du calendrier syrien, mentionnent les mois correspondants dans le calendrier ḥimyarite 8. Grâce à ces textes médiévaux, il est possible de restituer l’ordre des mois. On sait également que, en 548, le début de l’année (ou le Jour de l’An) était le 1er dhu-ḥullatān I. La liste des mois ḥimyarites dans leur ordre antique, avec leur équivalent syrien solaire, était donc : I. ḏ-ḥltn, dhu-ḥullatān = shubāṭ (février) En 548, ce mois, le premier de l’année, est redoublé (dhu-ḥullatān I et dhu-ḥullatān II) II. ḏ-mʿn, dhu-maʿūn = ādhār (mars) III ḏ-ṯbtn, dhu-thābatān = nīsān (avril) IV. ḏ-mbkrn, dhu-mabkarān = ayyār (mai)

Les autres calendriers sont ceux du Ḥaḍramawt, de Radmān, de Maḍḥàm, de Samʿī, de Haramum et (peut-être) de Nashshān. 8 Voir ci-dessous, pp. 27–29. 7

284

CHRISTIAN JULIEN ROBIN V. ḏ-qyẓn, dhu-qiyāẓān = ḥuzayrān (juin) VI. ḏ-mḏrʾn, dhu-madhraʾān = tammūz (juillet) VII. ḏ-ḫrfn, dhu-khirāfān = āb (août) VIII. ḏ-ʿln, dhu-ʿallān = aylūl (septembre) IX. ḏ-ṣrbn, dhu-ṣurābān = tishrīn I (octobre) X. ḏ-mhltn, dhu-muhlatān = tishrīn II (novembre) XI. ḏ-ʾln, dhu-ālān = kanūn I (décembre) XII. ḏ-dʾwn, dhu-diʾāwān = kanūn II (janvier) 9

Cette liste, dont la construction sera justifiée plus loin, nous servira de référence pour mettre en évidence les évolutions 10. Pour la désigner, nous utiliserons l’adjectif « classique ». 1. Le cadre chronologique Ḥimyar apparaît dans les inscriptions et le monnayage vers le milieu du 1er s. av. è. chr. Dès les origines, la résidence royale est le palais Raydān à Ẓafār (à 125 km au sud de Ṣanʿāʾ). Le territoire de Ḥimyar, qui s’étend alors de Yaslaḥ (la passe au nord de Ḍāf) à Sumāra (la passe au nord de Ibb) et à Radāʿ, n’occupe que la partie méridionale de la chaîne yéménite. Il est possible que la fondation du royaume soit un peu plus ancienne puisque l’inception (ou Les textes médiévaux qui utilisent le calendrier syrien situent le début de l’année en tishrīn I (octobre) ; de ce fait, leur liste des mois ḥimyarites commence avec dhu-ṣurābān (arabe dhū ʾl-ṣurāb, dhū ʾl-ṣurāb alawwal ou dhū ṣirāb ; Robin 2012 a, p. 129). 10 Avant la découverte des listes médiévales, les identifications étaient : — selon Beeston 1956 (p. 24), ḏ-ṯbtn (mars ou avril) ; ḏ-qyẓn (mai ou juin) ; ḏ-ḫrf et ḏ-mḏrʾn (de juillet à septembre) ; ḏ-dʾwn, ḏ-ṣrbn et ḏ-mʿn (octobre, novembre et décembre, ou novembre, décembre et janvier) ; — selon Lundin 1959 (p. 211), ḏ-ʾlʾlt et ḏ-ṯbtn (mars et avril) ; ḏ-qyẓn (juin) ; ḏ-ḫrf (juillet ou août) ; ḏ-mḏrʾn (septembre) ; ḏ-dʾwn (octobre) ; ḏ-ṣrbn (novembre) ; ḏ-mʿn (décembre) ; ḏ-ḥgtn et ḏ-mhltn (janvier et février). Noter qu’on lisait alors ḏ-ḥgtn (et non ḏ-ḥltn), ce qui était permis par la graphie très semblable du gīm et du lām après le début de l’ère chrétienne ; de ce fait, on croyait qu’un mois ḥimyarite avait le même nom qu’un mois du calendrier musulman, dhū ʾl-ḥijja (12e mois du calendrier al-muḥarram-dhū ʾl-ḥijja de Makka). 9

LES ÉVOLUTIONS DU CALENDRIER

285

commencement) de l’ère ḥimyarite est 110 av. è. chr. 11 Cette date pourrait signaler le début de l’indépendance de Ḥimyar qui était rattaché précédemment au royaume de Qatabān. Deux communes 12 voisines, Radmān et Maḍḥàm, se détachent également de Qatabān vers la même époque ; elles conservent une certaine autonomie, avant d’être annexées par Ḥimyar vers le fin du 3e s. è. chr. pour la première et vers 225 pour la seconde. Le nom de « Ḥimyar » a plusieurs acceptions parce qu’il peut désigner la partie et le tout. Jusque vers 300, il signifie le noyau initial, apparemment les communes qui avaient en partage le culte du dieu ʿAthtar 13. L’ensemble des populations qui reconnaissent l’autorité du roi de Ḥimyar est alors appelé « dhu-Raydān », d’après le nom de la dynastie qui lui-même dérive de Raydān, le palais royal de Ẓafār. Après l’annexion de Sabaʾ et la conquête du Ḥaḍramawt, vers 275–300, Ḥimyar désigne soit l’ancien royaume (devenant ainsi synonyme de dhu-Raydān), soit l’ensemble des populations de culture sudarabique par opposition aux Arabes ; dans les sources externes, ce nom est appliqué à l’ensemble de l’Empire, qui peut s’étendre jusqu’aux marges de Byzance et de la Perse. La rupture entre Qatabān et Ḥimyar se traduit de diverses manières. Les princes de Ḥimyar, dont on ignore le titre exact au 1er s. av. è. chr. 14, affirment tout d’abord leur autonomie en frappant monnaie. Ils accèdent bientôt à la dignité royale, en tout cas au plus tard vers le début de l’ère chrétienne quand le royaume voisin Cette date présente une petite incertitude (quelques années au maximum) parce qu’elle repose en dernière analyse sur une unique source manuscrite (Beaucamp et alii 1999). Mais c’est sans conséquence pour notre propos. 12 Traduction du sabaʾique shaʿab (s²ʿb), qui désigne les tribus urbanisées et organisées de l’Arabie du Sud, mais aussi, parfois, de l’Arabie déserte. Le terme shaʿab s’oppose à ʿashīrat (ʿs²rt), l’appellatif des tribus autarciques et marginales de la steppe et du désert. 13 Ces communes sont Yaḥṣib (celle de Ẓafār), Maytamum, Muhaqraʾum, Muhaʾnifum, Dhamarī, Shaddādum et, probablement, Yuhabshir, Dhubḥān et Wuṣāb (voir Robin 2013, carte 3). 14 Les inscriptions les appellent simplement « Mubāhil dhu-Raydān » et « Shammar dhu-Raydān ». 11

286

CHRISTIAN JULIEN ROBIN

de Sabaʾ, tout en conservant une grande autonomie, se soumet à eux. Leur titre officiel est désormais « roi de Sabaʾ et de dhuRaydān » 15. La langue officielle de Ḥimyar, qui était le qatabānique, est rapidement remplacée par le sabaʾique. Dans le domaine religieux, aucun panthéon collectif ni aucun temple fédéral ne semblent succéder à ceux de Qatabān : pour la première fois en Arabie méridionale, l’allégeance à un souverain n’entraîne pas automatiquement l’adhésion à certaines pratiques rituelles. Enfin, un nouveau calendrier et un nouveau système de datation apparaissent. Vers le début du 2e s. è. chr., le royaume de Sabaʾ reconquiert son indépendance. Les guerres continuelles qui opposent dès lors les princes de Ḥimyar (qui conservent le titre de « roi de Sabaʾ et de dhu-Raydān ») et les souverains de Sabaʾ, sans parler de dissidences diverses, ouvrent la voie à une conquête aksūmite (ou abyssine) du Yémen occidental qui dure près de 100 ans. Le roi ḥimyarite Yāsirum Yuhanʿim parvient à expulser les Aksūmites d’Arabie vers 270. La première conséquence de cette victoire est l’annexion du royaume de Sabaʾ qui devient désormais une simple dépendance. Peu avant 300, sous le règne de Shammar Yuharʿish, fils de Yāsirum Yuhanʿim (roi de premier rang de 287 à 311 environ), Ḥimyar s’agrandit encore avec la conquête du Ḥaḍramawt. Pour la première fois de son histoire, l’Arabie méridionale, à savoir le Yémen actuel plus la région de Najrān en Arabie séoudite et le Ẓafār omanais 16, est réunie sous une même couronne. À la suite de ces bouleversements politiques, on observe Un texte cursif unique mentionne le titre de « mukarrib de Sabaʾ et de dhu-Raydān » (L026 / 1-2). Dans deux inscriptions du 2e s., des Ḥimyarites mentionnent ou invoquent encore leur prince sans lui donner de titre royal. En [2]47 ḥim. (137–138 è. chr.), un personnage se dit « officier de Dhamarʿalī Yuhabirr dhu-Raydān » (mq(2)twy Ḏmrʿly Yhbr ḏ-(3)Rydn, KhUmayma 1 / 2–3). En 267 ḥim. (157–158 è. chr.), l’auteur fait une offrande après avoir été au service de « Dhamarʿalī dhu-Raydān » (Zubayrī al-ʿAwd 1 / 6-7 et 17–18). 16 Il importe de distinguer Ẓafār, la capitale antique de Ḥimyar, aujourd’hui un village qui porte encore ce nom, de la province méridionale de l’Oman qui a reçu ce nom au Moyen Âge, à l’époque rasūlide, et dont le nom antique était Saʾkalān. 15

LES ÉVOLUTIONS DU CALENDRIER

287

trois évolutions importantes dans les textes épigraphiques qui nous sont parvenus : une unification linguistique immédiate, la généralisation progressive du calendrier ḥimyarite à l’ensemble de l’Arabie méridionale et le choix d’une nouvelle religion, inspirée du judaïsme, qui devient officielle vers 380. Radmān et Maḍḥàm perdent alors ce qui leur restait d’autonomie. Le rejet du polythéisme et la conversion officielle à une nouvelle religion inspirée du judaïsme peuvent être datés vers 375–380, avec une faible marge d’erreur. La première manifestation officielle précisément datée de cette conversion remonte à janvier 384 : elle se trouve dans deux inscriptions que le roi Malkīkarib et ses deux fils 17 ont fait graver pour commémorer la construction de deux palais à Ẓafār 18. Dans ces inscriptions, les souverains remplacent la traditionnelle invocation à une série de divinités païennes par une formule entièrement neuve : « avec le soutien de leur seigneur, le Seigneur du Ciel » 19. L’inscription polythéiste datée la plus tardive qui soit connue à ce jour a été découverte dans le Grand Temple de Maʾrib, consacré au dieu Almaqah ; elle remonte à 379–380 20. On peut donc situer aux alentours de 380 la fin des cérémonies officielles dans ce temple et, de manière plus générale, l’interdiction de toute manifestation publique d’une adhésion au polythéisme. Une autre donnée doit être prise en compte, la date à laquelle Malkīkarib succède à son père Thaʾrān. Cette date est certainement Ce sont Abīkarib Asʿad et son frère Dharaʾʾamar Ayman. RES 3383 et Garb-Bayt al-Ashwal 2, datées de dhu-diʾāwān (ḏdʾwn) 493 ḥim. Pour la traduction julienne des mois ḥimyarites, nous utilisons de façon conventionnelle les équivalence des listes médiévales, même si elles sont évidemment approximatives (voir ci-dessous, notamment pp. 45–47). Les deux palais mentionnés dans ces deux textes s’appellent Shawḥaṭān et Kilānum ; un troisième, nommé Hirgāb, est construit par Shuriḥbiʾīl Yaʿfur fils d’Abīkarib Asʿad en décembre 462 (ḏ-ʾln 572 ḥim.) (ẒM 1). 19 b-mqm mrʾ-hmw Mrʾ S¹myn. 20 Cette date se trouve dans l’inscription inédite MB 2004 I-147 / 7– 8, datée de 489 ḥim. (379–380 è. chr.). Elle ne mentionne pas de roi dans la partie conservée. 17 18

288

CHRISTIAN JULIEN ROBIN

antérieure à 377 puisque, cette année-là, une inscription invoque les rois « Abīkarib Asʿad et Dharaʾʾamar Ayman » 21. Or, ces deux rois sont associés au trône par Malkīkarib, mais non par Thaʾrān 22. On peut ajouter que Malkīkarib accède au trône à une date sensiblement postérieure à 360 23, ce qui permet de retenir 375 environ pour le début du règne de Malkīkarib. Deux scénarii sont donc possibles. Dans le premier, le roi Malkīkarib proclame son adhésion au monothéisme dès son accession au trône (vers 375), mais n’ordonne pas l’abandon immédiat des temples polythéistes qui continuent à être visités pendant quelques années (jusque vers 380). Dans le second scenario, les deux événements — l’adhésion officielle au monothéisme et la fermeture des temples païens — interviennent à la même date, vers 380. Le judéo-monothéisme reste la religion officielle de Ḥimyar pendant près de 120 ans. Vers le début du 6e s., les rois de Ḥimyar deviennent tributaires du roi d’Aksūm (ou d’Abyssinie) qui est chrétien, pour des raisons et selon des modalités inconnues. Désormais, le prince qui accède au trône est choisi par le négus qui écarte manifestement les fils du roi précédent. Trois rois se succèdent. L’orientation religieuse du premier, Marthadʾilān Yanūf, n’est pas connue ; le deuxième, Maʿdīkarib Yaʿfur (mort peu avant juin 522) est chrétien ; quand au troisième, qui s’appelle Joseph (Yūsuf Asʾar Yathʾar), c’est plus compliqué. Joseph est placé sur le trône de Ḥimyar par le Shibām Kawkabān 1 de juin 377 (ḏ-qyṣn 487 ḥim.). Shibām Kawkabān est une petite ville, à une trentaine de kilomètres au nord-ouest de Ṣanʿāʾ. 22 Les corégents de Thaʾrān Yuhanʿim sont : 1. son fils Malkīkarib (Yuhaʾmin) ; 2. ses fils Malkīkarib Yuhaʾmin et Dhamarʿalī Yuhabirr. Concernant la chronologie et les orientations religieuses du règne de Thaʾrān, se reporter à Robin 2014 b. 23 En 360 (ʿAbadān 1), Thaʾrān Yuhanʿim ne semble pas avoir de corégent. Il règne ensuite avec un corégent (Ja 669, 670 et 671 + 788), puis avec deux (Ẓafār 10~002 ; DhM 204 ; Khaldūn-al-Jarasha 1 ; MQMinkath 1), sans doute de façon durable puisque les attestations sont relativement nombreuses. 21

LES ÉVOLUTIONS DU CALENDRIER

289

négus Kālēb 24 vers juin 522 ; on ignore s’il est alors officiellement chrétien. On sait seulement qu’il se révolte à l’automne 522 et se réclame dès lors du judaïsme. Le conflit qui éclate alors entre Ḥimyar et Aksūm prend des dimensions considérables. Les populations chrétiennes favorables à Aksūm sont massacrées. Les chrétiens de Nagrān entrent en dissidence et sont bientôt férocement châtiés (novembre 523). Une expédition aksūmite de représailles réussit à débarquer en 525 (ou l’une des années suivantes) et à vaincre Joseph qui meurt au combat. Kālēb, qui parvient jusqu’à Najrān, conquiert alors l’ensemble de l’Arabie méridionale, massacre systématiquement les juifs et apporte son soutien à l’église ḥimyarite, en renforçant sa hiérarchie et en construisant de nouveaux lieux de culte. Dès lors, le Yémen est officiellement chrétien et le reste jusque vers 570–575, d’abord sous l’autorité d’un roi vassal (Sumūyafaʿ Ashwaʿ), puis sous celle d’un général aksūmite, Abraha, qui s’est emparé du trône après 531. Abraha, après avoir réussi à stabiliser son régime en 548, reprend le contrôle de la totalité de l’Arabie déserte en 552 25. Il entreprend peu après la construction de la Grande Église de Ṣanʿāʾ, surnommée al-Qalīs, dont l’achèvement peut être daté de 560 26. D’après la Tradition arabo-islamique, Abraha aurait voulu faire d’al-Qalīs un grand centre de pèlerinage pour toute l’Arabie. C’est la raison pour laquelle il aurait lancé une expédition contre Makka, afin d’anéantir la Kaʿba qui attirait déjà de nombreux fidèles. C’est lors de cette expédition, qui se termine en débacle, qu’il aurait placé à la tête de son armée un éléphant. La question de savoir si cette expédition est historique est débattue : certains supposent que les récits qui en traitent ne sont qu’une amplification de l’énigmatique sourate 105, « l’Éléphant ». En faveur de l’historicité de l’expédition, on peut invoquer le fait que Makka (une bourgade isolée, misérable et peu peuplée) a joui d’une certaine primauté en Arabie à la fin du 6e s., qui peut être expliquée par une victoire mecquoise contre Abraha. Concernant l’éléphant d’Abraha, il est plausible qu’il est également historique Martyrion grec § 27. Robin & Ṭayrān 2012. 26 Robin 2015 b. 24 25

290

CHRISTIAN JULIEN ROBIN

parce la poésie préislamique le mentionne avec des précisions qui ne se trouvent pas dans le Coran ; un autre argument réside dans les trois gravures représentant un éléphant avec son cornac, sans doute difficiles à dater, mais certainement antiques, qui ont été découvertes sur des rochers à une centaine de kilomètres au nord de Najrān, sur la route vers Makka 27. Ces gravures représentent très vraisemblablement un éléphant qui accompagnait Abraha dans ses déplacements. En premier lieu, on sait que l’éléphant était un animal emblématique du pouvoir royal à Aksūm et, de manière plus générale, en Afrique orientale ; en deuxième lieu, si on fait la liste des souverains non sudarabiques ayant assurément circulé au nord de Najrān, on ne trouve guère qu’Abraha 28 ; enfin, on sait qu’Abraha a lancé au moins cinq campagnes en direction de l’Arabie déserte, en passant par Najrān 29. Un ultime argument, qui ne concerne pas directement l’expédition contre Makka, mais qui présente l’intérêt de confirmer une donnée de la même tradition, peut être avancé : la structure d’al-Qalīs, avec son bâtiment principal (bayt), sa cour à arcade (īwān) et sa coupole (qubba) a une structure qui s’accorde bien avec un lieu de pèlerinage. On peut aisément supposer que la qubba accueillait de saintes reliques, probablement celles d’un martyr 30. Abraha échoue devant Makka, probablement à cause d’une épidémie (en relation avec la peste de Justinien) qui frappe son armée. Il ne survit pas au désastre, mourant sur le chemin du retour. Deux de ses fils se succèdent brièvement sur le trône. Leur pouvoir s’effondre vers 570 si on se fonde sur les sources byzantines ou vers 575 d’après les sources arabes. Les Perses sāsānides étendent alors leur domination sur l’ensemble de l’Arabie, mais sans cherRobin et alii 2014, pp. 1060–1062 et fig. 62 (p. 1116) ; Robin 2015 a, pp. 36–48. 28 Si on sait que Kālēb parvient à Najrān, il est peu vraisemblable qu’il se soit aventuré plus loin vers le nord dans une région désolée. 29 De ce fait, il n’est pas sûr que l’éléphant des gravures soit nécessairement celui de l’expédition contre Makka. L’itinéraire normal des campagnes ḥimyarites vers l’Arabie déserte passait par Najrān, Ḥimà et Tathlīth, comme le montrent les expéditions de 521 et de 552. 30 Robin 2015 b. 27

LES ÉVOLUTIONS DU CALENDRIER

291

cher à la contrôler directement. C’est seulement à Ṣanʿāʾ et à Hajar dans le golfe Arabo-persique que l’on trouve des gouverneurs représentant le roi. Ailleurs, l’autorité est déléguée à une multitude de roitelets et de chefs locaux de sorte que l’anarchie gagne la totalité de la péninsule. Ce survol historique met en évidence deux périodes : une polythéiste jusque vers 380, suivie par une autre monothéiste (380– 570), d’inspiration juive tout d’abord (380–500), puis chrétienne (530–570). Comme les données sur le calendrier ḥimyarite sont peu abondantes, nous allons nous fonder sur ces deux périodes : avant et après la conversion au judéo-monothéisme (vers 380). Une telle distribution n’est pas illogique, puisque les conversions religieuses entraînent fréquemment des changements dans le calendrier. 2. Le calendrier et l’ère de Ḥimyar avant la conversion au judaïsme (c. 380) Pour la période antérieure à 380 (490 ḥim.), on dispose d’une cinquantaine de textes comportant une date : une trentaine pour Ḥimyar au sens strict et une vingtaine pour Radmān et Maḍḥàm. Ces deux dernières communes sont incluses dans notre inventaire parce que, comme nous allons le voir, leur calendrier est très semblable (et peut-être même identique) à celui de Ḥimyar. Pour la période postérieure à 380, ce sont un peu moins de cinquante textes. Dans ce décompte, nous n’avons pas inclus les textes incomplets dont la date est incertaine. Ḥimyar, Radmān et Maḍḥàm utilisent trois ères différentes qui sont parfois distinguées par l’ajout d’un nom propre, notamment quand il y a un risque d’ambiguïté ou de confusion. Les textes appellent ces ères « années de Mabḥūḍ ibn Abḥaḍ », « années de Abīʿalī ibn Ratʿ » et « années de Nabaṭum ibn Kharīf ». Cependant, pour les désigner, il me semble plus simple de se référer à la commune qui les utilisait, ḥim. (pour Ḥimyar), radm. (pour Radmān) et maḍḥ. (pour Maḍḥàm). Les noms de mois présentent quelques particularités intéressantes. On observe tout d’abord que la manière de les citer n’est pas encore normalisée. Ils peuvent être mentionnés soit sous la forme nue (par exemple ʿln dans MQ-dhū-Wayn 1, 2 et 9) soit avec le pronom de relation ḏ- (par exemple ḏ-ḫrf dans Kh-al-Harūj 1 / 4). On connaît treize noms de mois. Onze sont identiques à ceux qu’on trouve dans le calendrier ḥimyarite classique :

292

CHRISTIAN JULIEN ROBIN ḏ-ʿln (variante ʿln) (VIII) 31 ; dwʾ (XII) 32 ; ḏ-ḫrfn (VII) 33 ; ḏ-ḥltn 34 ; ḏ-mʿn (variante mʿn et ḏ-mʿwn) (II) 35 ; ḏ-mbkrn (IV) 36 ; ḏ-mḏrʾn (variante mḏrʾn et mḏrʾ) (VI) 37 ; ḏ-mhltn (X) 38 ; ḏ-qyẓn (variante qyẓn, ḏqyṣn) (V) 39 ; ṣrbn 40 (IX) ; ḏ-ṯbtn (III) 41.

Khaldūn-ʿAbd Allāh, zabūr = Mus. Univ. Dhamār 34 / 2 (ḏ-ʿl[n], 396 ḥim.) ; ThUM 34 / 2 (ḏ-ʿl[n], 399 ḥim.) ; al-Miʿsāl 6 / 11 (ḏ-ʿln sans mention d’année, mais datant du règne de Yāsirum Yuhanʿim) ; probablement Gr 11 / 3 (ʿln, [..]5 ḥim.). 32 Kh-Jarf al-Naʿīmiyya 12 / 5 (ṣ[yd] w-dwʾ, 242 ḥim.). 33 ẒM 2263+2262+2264 / 4 (…]wrḫnhyn ḏ-mḏrʾnyn w-ḏ-ḫrfny[n … ; mention ailleurs de 347 ḥim.) ; YM 1950 / 5 (ḏ-ḫrfn, [47]3 ou [48]3 ḥim.) ; Kh-al-Harūj 1 / 4 (ḏ-ḫrfn, 274 ḥim) ; Kh-al-Harūj 2 / 5 [(ḏ)-ḫrfn (?), 482 ḥim.]. 34 Ẓafār 08~235 (…wr](ḫ)n ḏ-ḥltn ḏ-s¹[…(8)…]w-ṯlṯ mʾtm[ bn ḫrf (9) Mbḥ](ḍ) bn ʾbḥḍ) ; la date, assurément antérieure à 400 ḥim. (290 EC), est incertaine. Le mīm qui est la dernière lettre du nom du roi et la graphie suggèrent le règne de Yāsirum Yuhanʿim (vers 265–287) et de restituer soit 386 ou 387 soit 396 ou 397. Kh-Ḥammat al-Ḍabʿ 5 (b-wrḫn (5) ḏ-ḥltn) ; la date de ce graffite peu soigné est incertaine ; une datation antérieure à 380 est retenue parce qu’il appartient à un ensemble polythéiste, qui paraît ancien. 35 Kh-Nūna 1A = Gl 1594 / 13 (ḏ-mʿwn, 389 ḥim.) ; YMN 13 = alMiʿsāl 18 / 12 (ḏ-(m)ʿn, 409 ḥim.) ; probablement Ghūl YU 100 / 5 (b-wrḫn mʿn, date perdue). 36 CIH 448 + Hakir 1 / 5 (bn wrḫn ḏ-mbkr[n ... ; texte daté de ḏ-qyẓn 396 ḥim.) ; 37 ẒM 2263+2262+2264 / 4 (…]wrḫnhyn ḏ-mḏrʾnyn w-ḏ-ḫrfny[n …, mention ailleurs de 347 ḥim.) ; Av Būsān 4 (…]|mḏrʾn, [40]7 ḥim.) ; ʿAbadān 1 (ḏ-mḏrʾ[n], 470 ḥim.) ; Kh-Jarf al-Naʿīmiyya 10 / 3 (mḏrʾn, sans mention de l’année, mais avec celle de Yhqm, souverain sabéo-raydānite de la fin du 1er s. è. chr.). 38 CIH 46 = Gl 799 / 6 (ḏ-mhltn, 385 ḥim.). 39 CIH 448 + Hakir 1 / 6 (ḏ-qyẓn, 396 ḥim.) ; Shibām Kawkabān 1 (ḏn qyṣ , 487 ḥim.) ; Khaldūn-Umayma 1 / 5 (qyẓn, [2]47 ḥim.). Voir aussi, ciaprès, les trois textes dans lesquels ce mois est qualifié de « premier » (qdm). 40 Kh-Jarf al-Naʿīmiyya 13 / 5-6 (ṣrbn, 242 ḥim.). 31

LES ÉVOLUTIONS DU CALENDRIER

293

Deux mois portent un nom différent, ḏṯʾ (variante ḏ-dṯʾn) 42 et ṣyd 43. Les deux attestations du mois dṯʾ ou ḏ-dṯʾn se trouvent dans des graffites sans indication d’année, mais qui sont certainement antérieurs à la conversion au judéo-monothéisme. Ces graffites mentionnent en effet l’un des trois rois appelés Dhamarʿalī Yuhabirr, qui ont régné respectivement aux 1er s. av. è. chr., 2e ap. et 4e ap. Entre les trois, la graphie (très originale, avec des formes de lettres sans précédent pour le rāʾ ou le fāʾ) semble donner la préférence au second, mais sans certitude 44. Le rang du mois ṣyd peut être restitué avec une certaine vraisemblance. Un graffite, que les difficultés de lecture ne permettent pas de comprendre parfaitement à cet endroit, mentionne un événement qui se déroule « au mois de ṣ[yd] et dʾw » (b-wrḫn (5) ṣ[yd] wdʾw) 45. Il suggère donc que le mois de ṣyd précède celui de dwʾ (XII, janvier). Or, dans le calendrier ḥimyarite classique, le mois précédant dʾw est ʾln (XI, décembre) qui, précisément, n’est pas attesté avant 380 46. On peut donc faire l’hypothèse que, avant 380, le calendrier ḥimyarite avait un mois ṣyd (XI, décembre) qui est remplacé après 380 par ʾln. Le nom de ṣyd dérive sans doute du substantif qui signifie la « chasse » (arabe ṣayd). Mais il est difficile d’établir si la chasse était alors une activité d’hiver. L’almanach de ʿUmar note que la saison de la chasse aux oiseaux migrateurs commence le 14 Kh-Ḍāf 1 / 6 (ḏ-ṯbtn, 392 ḥim.) ; peut-être Kh-Sayya 3 / 5 (ḏ-ṯbtn, sans mention d’année). 42 Kh-Jarf al-Naʿīmiyya 16 / 2 (dṯʾ, sans mention de l’année) ; KhJarf al-Naʿīmiyya 15 / 2 (ḏ-dṯʾn, sans mention de l’année). 43 Kh-Jarf al-Naʿīmiyya 7 / 4 (ṣyd, 200 ḥim.) ; Kh-Jarf alNaʿīmiyya 12 / 5 (ṣ[yd] w-dwʾ, 242). 44 On peut ajouter qu’un officier du roi Dhamarʿalī Yuhabirr (appelé ici Dhamarʿalī Yuhabirr dhu-Raydān) est l’auteur du graffite Kh-Umayma 1 daté de qyẓn [2]47 ḥim. Le site d’Umayma (à 1 km à l’est de Hakir) est voisin de celui de Jarf al-Naʿīmiyya (à 1 km au nord de Hakir). 45 Kh-Jarf al-Naʿīmiyya 12 / 5 (242 ḥim.). 46 Les occurrences sont ẒM 1 A / 14 et B / 8 (572 ḥim.) ; RobinNajr 1 / 3 (597 ḥim.) ; Murayghān 1 = Ry 506 / 9 (662 ḥim.) ; et RES 4157 / 4 (probablement du règne de Marthadʾīlān Yanūf au début du 6e s. è. chr.). 41

294

CHRISTIAN JULIEN ROBIN

aylūl (septembre) : « premier départ du faucon ; chasse avec des oiseaux de proie » 47. Quant à la chasse rituelle à l’ibex, elle se pratique au Ḥaḍramawt en toute saison, de la plus froide à la plus chaude, sans relation avec une phase de la lune ou avec un mois, mais seulement de jour ; certains, cependant, estiment qu’il est erroné de chasser pendant les fortes chaleur parce qu’il est alors facile de surprendre le gibier aux points d’eau 48. Le cas de dṯʾ est plus compliqué. Trois possibilités peuvent être envisagées. La première, qui paraît la plus plausible, est que dṯʾ est le second nom d’un mois du calendrier ḥimyarite, éventuellement propre à une région. Une deuxième possibilité est que dṯʾ est un nom ancien, remplacé par un autre au 2e ou au 3e s. ; on pourrait penser à ḏ-ṯbtn d’une part parce que sa première attestation datée remonte à 392 ḥim. 49, d’autre part parce que dṯʾ signifie « printemps » 50. Une troisième possibilité serait de supposer que dṯʾ est le nom d’un mois intercalaire. Il faut encore noter que, trois fois, le mois ḏ-qyẓn est qualifié de « premier ». Deux inscriptions sont datées de 308 et 327 ḥim. 51 et une troisième ne donne pas le nombre d’années 52. Cette dernière présente la particularité très exceptionnelle dans les graffites d’être écrite en boustrophédon, c’est-à-dire de droite à gauche à la première ligne, puis de gauche à droite à la deuxième et ainsi de suite, à l’imitation du parcours du bœuf qui laboure un champ. Dans les inscriptions monumentales, l’usage du boustrophédon est un critère de grande antiquité puisqu’il n’est plus attesté après 300 av. è. awwal khurūj al-badrī wa-ṣayd jawāriḥ al-ṣayd (Varisco 1994, pp. 39 et 59). L’almanach fait une seconde allusion à la chasse sous l’entrée « 3 kanūn al-thānī (janvier) » (Varisco 1994, pp. 27 et 46). 48 Serjeant 1976, pp. 25 et 27. 49 Noter cependant que le mois de ḏ-ṯbtn est attesté dans deux bâtonnets (X.SBS 10A / 2 ; X.SBS 11 / 2 et 20) de style graphique Ryckmans IV a, antérieur au 4e s. è. chr. 50 Voir ci-dessous, p. 37. 51 Kh-Jarf al-Naʿīmiyya 18 / 9 (qyẓ qdmn, 308 ḥim.) ; KhUmayma 5 / 7 (qyẓ qdmn, 327 ḥim.). Noter que 308 et 327 sont séparés par 19 ans (durée d’un cycle classique d’intercalation : voir pp. 32–33). 52 Kh-Umayma 25 / 2, b-wr(ḫ) qyẓn qdm. 47

LES ÉVOLUTIONS DU CALENDRIER

295

chr. environ. Pour les graffites, il est plausible que cette date soit également valide, au moins de façon approximative 53. On aurait donc, dans l’inscription Kh-Umayma 25, la plus ancienne mention d’un mois ḥimyarite. Le fait qu’un mois soit qualifié de « premier » suggère que le même mois peut être appelé « le dernier » (ʾḫrn) : dans les calendriers de Qatabān et de Sabaʾ, on relève en effet des mois appelés ḏ-Brm, ḏ-Brm qdmn et ḏ-Brm ʾḫrn ou encore ḏ-Ns¹wr, ḏ-Ns¹wr qdmn et ḏNs¹wr ʾḫrn 54. C’est un premier indice soutenant l’hypothèse qu’un mois ḥimyarite pouvait être redoublé. Un second indice se trouve peut-être dans les fragments d’une inscription monumentale provenant de Ẓafār, qui commémore la construction d’éléments de l’enceinte 55. À la ligne 4, on lit une date formulée de façon inhabituelle : « aux deux mois ḏ-mḏrʾnyn w-ḏ-ḫrfny[n … … des années de Mab]ḥūd ibn Abḥaḍ », …]wrḫnhyn ḏ-mḏrʾnyn w-ḏ-ḫrfny[n … … bn ḫryf Mb]ḥḍ bn ʾbḥḍ

Dans la lacune, il y avait assurément un nombre d’années comme l’indique la mention de [Mab]ḥūd ibn Abḥaḍ. Les deux noms de mois (ḏ-mḏrʾn et ḏ-ḫrfn, VI et VII, qui correspondent à juillet et août) sont munis d’une désinence -yn. Walter Müller l’interprète comme une forme adjective (nisba) et traduit « der ḏūmaḏraʾānische » et « der ḏūḫirāfānische ». La désinence -yn peut également noter le duel, ce qui donnerait une formulation également énigmatique, « les deux mois, les deux ḏ-mḏrʾn et les deux ḏ-ḫrfn ». Si cette dernière interprétation était retenue, elle pourrait être mise en relation avec le redoublement de deux mois, peut-être à l’occasion d’une réforme du calendrier. On notera en effet que l’inscription présente également la particularité d’être la plus ancienne à attribuer l’ère ḥimyarite à Mabḥūḍ ibn Abḥaḍ, aussi bien dans la date de la ligne 4 (cidessus) que dans celle de la ligne 3 : Voir le graffite en boustrophédon de la grotte de Ḥôq (Suquṭra) et le commentaire dans Robin 2012 b, pp. 93–94. 54 Voir ci-dessous, pp. 50–51. 55 ẒM 2263+2262+2264. 53

296

CHRISTIAN JULIEN ROBIN « …] trois cent quarante-sept des années de [Mabḥūḍ ibn A]bḥaḍ », ... ]ḏ-l-s¹bʿt w-ʾrbʿy w-ṯlṯ mʾtm bn ḫryf[ Mbḥḍ bn ʾ]bḥḍ

Le redoublement d’un mois peut être interprété comme l’intercalation d’un treizième mois, comme dans les calendrier babylonien ou juif, afin d’aligner l’année lunaire sur le rythme des saisons 56. L’hypothèse que ce soit deux des douze mois de l’année qui portent le même nom, comme les deux tishrīn ou les deux kanūn du calendrier syrien, ou les deux rabīʿ et les deux jumādà du calendrier islamique, peut être écartée à cause du nombre des noms de mois qui atteint douze 57. L’année ḥimyarite est comptée d’après une ère qui, dans un premier temps (de 50 à 220 è. chr. environ), n’est pas nommée. On écrit simplement « dans l’année qui a (le rang) un-tel ». Une dédicace indique ainsi : « cette promesse (c’est-à-dire l’offrande faite à la suite de cette promesse) eut lieu en l’an cent soixante-deux » (w-kwn ḏn s²ftn b-ḫ(10)rfn ḏ-l-ʾḥd w-s¹ṯy w-mʾtm) 58. Les nombres d’années qui sont attestés sont : 102, 111, 146, 161, 167, 200 (quatre fois), 204, 242 (deux fois), 247, 267, 274, 305, 308, 327 et 330 ḥim. 59. À partir de 230 è. chr. environ (340 ḥim.), la norme change : désormais, pendant près d’un siècle, les auteurs d’inscriptions précisent de façon presque systématique quelle est l’ère qui est utilisée. On l’observe quel que soit l’auteur — le roi ou une personne privée — et quelle que soit la provenance — la capitale ou les communes. L’ère ḥimyarite est appelée « les années de Mabḥūḍ ibn Abḥaḍ ». Ce personnage est inconnu, mais non Abḥaḍ, un nom de lignage Voir ci-dessous, p. 32. La même observation vaut également pour les mois sabéens et qatabānites qui sont redoublés (ci-dessous, pp. 50–51). 58 Höfner AF 3 = Munich 91-315336. 59 La plupart des inscriptions portant ces nombres ont été découvertes récemment par Khaldūn Nuʿmān qui les a réunies dans sa thèse (inédite), avec le sigle Kh, déjà rencontré, suivi d’un nom de site (2012). Voir Kh-al-Harūj 1 ; Kh-Ghawl al-ʿAjmāʾa 1 ; Kh-Ḥammat al-Ḍabʿ 19 = Ja 3199 ; Kh-Jarf al-Naʿīmiyya 4, 5, 6, 7, 12, 13 et 18 ; Kh-Umayma 1, 2, 3, 4, 5 et 6 ; Garb.-Hakir 2 ; Höfner AF 3 = Munich 91-315336 ; Zubayrī-alʿAwd 1. 56 57

LES ÉVOLUTIONS DU CALENDRIER

297

dont on a deux autres attestations. Une stèle funéraire provenant de la nécropole de la capitale de Qatabān porte l’inscription : « mémorial de S²mnr ibn Abḥaḍ à Riṣāfum le ẓafārite », mʿmr S²mnr bn ʾbḥḍ b-Rṣfm ẓfryn (Ja 358).

D’après la graphie, ce texte pourrait dater du 1er s. av. è. chr. L’indication que ce S²mnr est originaire de Ẓafār, qui sera plus tard la capitale de Ḥimyar, permet de considérer qu’il appartient bien au même lignage que Mabḥūḍ ibn Abḥaḍ. Le nom de lignage Abḥaḍ se retrouve au 2e s. è. chr. dans un graffite rupestre, qui a pour auteur : « Bahīl ibn Abḥaḍ, officier de Dhamarʿalī Yuhabirr dhuRaydān », Bhl bn ʾbḥḍ mq(2)twy Ḏmrʿly Yhbr ḏ-(3)Rydn (Kh-Umayma 1).

Pour autant qu’on puisse en juger, les banū Abḥaḍ appartenaient à la petite noblesse ḥimyarite. Les noms Mabḥūḍ et Abḥaḍ ne sont pas communs. La Tradition arabo-islamique ne les connaît pas. Ce n’est guère surprenant puisque la racine BḤḌ n’est pas attestée en arabe. La vocalisation « Mabḥūḍ », est arbitraire ; ce pourrait être aussi bien Mibḥāḍ que Mubaḥḥiḍ etc. Celle de Abḥaḍ, en revanche, est plausible puisqu’elle s’insère dans la longue série des noms formés sur le schème Afʿal, comme Aḥmad, Asʿad etc. Les inscriptions précisant que l’ère utilisée est celle de Mabḥūḍ ibn Abḥaḍ sont au nombre de dix ; elles portent les millésimes 347, 363, 385, 389 ; 392 ; 396, 3[.](6 ou 7) ; [40]7, 42[0], 470 ḥim. 60 Sur les dix, sept datent des règnes de Yāsirum Yuhanʿim et de son fils Shammar Yuharʿish. Deux sont quelque peu antérieures. La plus ancienne, celle de 347 ḥim., commémore un agrandissement ou une ẒM 2263+2262+2264 (Ẓafār) ; al-Miʿsāl 2 ; CIH 46 = Gl 799 (Yakār, village à quelques kilomètres au nord-est de Ḍāf) ; Kh-Nūna 1A = Gl 1594 (Būsān, à mi-chemin entre Baynūn et Ḍāf) ; Kh-Ḍāf 1 ; CIH 448 + Hakir 1 (inscription royale) ; Ẓafār 08~235 (…wr](ḫ)n ḏ-ḥltn ḏs¹[…(8)…]w-ṯlṯ mʾtm[ bn ḫrf (9) Mbḥ](ḍ) bn ʾbḥḍ ; restituer probablement soit 386 ou 387 soit 396 ou 397 [voir n. 34]) ; Av Būsān 4 ; YM 1695 (Baynūn) ; ʿAbadān 1. 60

298

CHRISTIAN JULIEN ROBIN

réfection de l’enceinte de la capitale ; son auteur (dont le nom a disparu) est probablement le roi. Il est paradoxal que ce soit dans un tel texte qu’on trouve la première référence à Mabḥūḍ ibn Abḥaḍ : dans une inscription de la capitale, sans doute royale, il semble inutile de préciser quelle est l’ère utilisée. La seconde inscription, datée de 363 ḥim., a un statut un peu particulier. Elle provient d’une commune, Radmān, qui, alors, n’est pas ḥimyarite, mais seulement une alliée de Ḥimyar. Le prince de Radmān qui publie ce texte à Waʿlān date logiquement son inscription dans l’ère locale 61, mais, ce qui est remarquable, également dans celle de Ḥimyar : « Cette mission fut une réussite et un gage de sécurité, au mois de madhraʾān cent soixante-dix-neuf de l’ère de Abīʿalī b. Ratʿ (15), (soit) dans le comput de Ḥimyarum, trois cent soixantetrois à partir de l’année de Mabḥūḍ ibn [Abḥaḍ] » 62.

L’ajout du millésime ḥimyarite ne répond probablement pas à des considérations pratiques, comme celle de se référer à une ère plus répandue. Il a plutôt une signification politique : le prince de Radmān reconnaît de façon implicite son allégeance au souverain de Ḥimyar. Quant au dernier texte, ʿAbadān 1, qui est postérieur aux règnes de Yāsirum et de Shammar, nous reviendrons bientôt sur sa signification. Les inscriptions qui ne précisent pas l’ère sont au nombre de six 63. L’une 64 est particulièrement significative parce qu’elle proVoir ci-dessous, p. 19. al-Miʿsāl 2 / 14-15 : w-kwn ḏn bltn ṣdqm w-ʾmntm w-b-wrḫn mḏrʾn ḏ-l-ts¹ʿt w-s¹bʿhy w-mʾt ḫryftm bn ḫrf ʾbʿly bn Rtʿ (15) w-t(ḫ)rf Ḥmyrm ḏ-l-ṯlṯt w-(s¹)ṯhy wṯlṯ mʾtm bn ḫrf Mbḥḍ b(n) [ʾbḥḍ]. 63 Ce sont trois inscriptions monumentales (YMN 13 = al-Miʿsāl 18, 409 ḥim. ; ẒM 704, 41[3-9] ḥim. ; DJE 25, 434 ḥim. ; Shibām Kawkabān 1, 487 ḥim.) et deux graffites rupestres (Khaldūn — ʿAbd Allāh / zabūr = Mus. Univ. Dhamār 34, 396 ḥim., et Kh-al-Harūj 2, 482 ḥim.). Il n’y a pas lieu de retenir les textes dans lesquels la fin de la datation est perdue (Gl 1541, 403 ḥim. ; MQ-Maʾrib 1, 461 ḥim. ; YM 1950, 473 ou 483 ḥim.). 64 YMN 13 = al-Miʿsāl 18. 61 62

LES ÉVOLUTIONS DU CALENDRIER

299

vient de Waʿlān, une ville où l’on datait d’après l’ère de Radmān comme nous l’avons vu. On peut supposer que, au moment de sa rédaction, mars 299 (ḏ-mʿn 409 ḥim. 65), il n’était plus nécessaire de préciser l’ère utilisée puisque seule l’ère ḥimyarite était désormais en usage. Effectivement, toutes les inscriptions postérieures 66 cessent d’identifier l’ère, à l’exception d’une seule (ʿAbadān 1), qui est singulière à plusieurs égards. C’est un très long texte dans lequel, en juillet 360 è. chr., un prince du Ḥaḍramawt occidental célèbre ses exploits militaires (ainsi que ceux de ses fils et petits-fils) et ses réalisations architecturales, agricoles ou autres. Le lignage de ce prince, dhu-Yazʾan, était déjà connu précédemment, mais il n’acquiert alors une position dominante. Il est très vraisemblable que cette promotion résulte du ralliement des Yazʾanides à Ḥimyar quand, un peu avant 300, ce royaume s’est lancé à la conquête du Ḥaḍramawt. On peut le déduire de la fortune des Yazʾanides, mais aussi d’un petit détail mentionné incidemment : « [Ils ont] achevé en outre (des travaux) dans leurs palais et leurs terres alors qu’ils reconstruisaient leur ville ʿAbadān, après que le Ḥaḍramawt l’eut détruite » 67.

Si le Ḥaḍramawt a détruit ʿAbadān, la résidence principale des Yazʾanides, c’est soit parce que ces derniers s’étaient rebellés soit parce qu’ils s’étaient engagés dans une coalition hostile. L’inscription ʿAbadān 1 est doublement datée, en mois et ère ḥimyarite et en mois et éponyme du Ḥaḍramawt : « au mois du dhū-madhraʾā[n] de l’année quatre cent soixantedix de (44) l’ère de Yabḥuḍ 68 ibn Abḥaḍ, soit, dans le calendrier ḥaḍramite, le mois (?) Shamsḥayy d’Aslamum .... » 69

Il est assuré qu’il s’agit de l’ère de Ḥimyar (inception en 110 av. è. chr.) et non de celle de Radmān (inception en 74 è. chr.) du fait du roi mentionné et du nombre d’années. 66 DJE 25 (434 ḥim.), Kh-al-Harūj 2 (482 ḥim.), Shibām Kawkabān 1 (487 ḥim.). 67 ʿAbadān 1 / 32 : … w-w]zʾw hqḥ b-ʾbt-hmw w-ʾrḍ-hmw k-ṯwbw hgr-hmw n ʿbd bʿdn k-dhr-hw Ḥḍrmt. 65

300

CHRISTIAN JULIEN ROBIN

On peut déceler dans ce choix une intention politique, celle d’afficher clairement le ralliement à Ḥimyar, tout en rappelant, dans le même temps, une spécificité qui implique une certaine autonomie. Cette revendication d’autonomie transparaît aussi dans le fait que le souverain est invoqué de façon anonyme, sans être nommé personnellement 70. Elle se confirme dans les textes postérieurs : si le premier invoque encore le souverain, toujours de façon anonyme 71, les suivants s’en dispensent 72. Nous traitons à part deux derniers documents qui sont datés de manière un peu différente, avec un nombre d’années et le nom d’un personnage qui est probablement un éponyme. Le premier est un graffite rupestre découvert par le Yéménite Aḥmad Nājī Sārī dans la région de Sinabān entre Dhamār et Radāʿ, à une dizaine de kilomètres à l’est de Hakir, puis revu par Albert Jamme et Khaldūn Nuʿmān 73. Il enregistre un accord (à propos d’un action dont la nature n’est pas éclaircie 74) avec les habitants de la ville ḥimyarite de Hakirum (aujourd’hui Hakir, à 30 km au nord-est de Ẓafār) concernant le petit bétail. Le document se conclut par la phrase : 68

Le lapicide, mal informé ou inattentif, a gravé Ybḥḍ à la place de

ʿAbadān 1 / 43-44 : wrḫ-hw ḏ-mḏrʾ[n] ḏ-l-s¹bʿy w-ʾrbʿ mʾtm b(44)n ḫrf Ybḥḍ bn ʾbḥḍ w-kn b-tḫrf Ḥḍrm(t) (w)r(ḫ) S²ms¹ḥy ʾs¹lmm (bn )[..]n. 70 « Avec l’aide et le soutien de ʿAthtar Shāriqān et des divinités de leur pays et avec le soutien de leurs seigneurs, les maîtres de Raydān », brdʾ w-mqm ʿṯtr S²r(qn) w-ʾl(ʾl)t ʾrḍ-hmw w-b-mqm ʾmrʾ-hmw ʾbʿl Rydn (ʿAbadān 1 / 42). 71 RES 5085 / 8, octobre 450 (z-ṣrbn 560 ḥim.) : « et avec l’aide de leurs seigneurs, les maîtres de Rēdān », w-b-rdʾ ʾmrʾ-hmw ʾmlkn ʾbʿl Rdn. 72 MAFRAY-Abū Thawr 4 (juin 486, ḏ-qyẓn 596 ḥim.) ; MAFYS-Ḍuraʾ 3 = RES 4069 (août 488, ḏ-ḫrfn 598 ḥim.) ; BR-Yanbuq 47 (avril 515, ḏ-ṯbtn 625 ḥim.). 73 Kh-Ḥammat al-Ḍabʿ 19 = Ja 3199 = Nājī 1. 74 Elle est exprimée par les verbe ḥffy (Rbḥm bn Ṯ(2)wrm mʾdby bn(3)y ʿwyfm (4) ḥffy qny-hmy ḥwry (5) hgrn Hkrm), ḥff et ḥfy qui se trouvent dans les graffites de deux sanctuaires, Ḥammat al-Ḍabʿ (à 3 km à l’est de Sinabān) et alNaʿīmiyya (à moins d’un kilomètre au nord de Hakir) (Nuʿmān 2012, index, p. 204) Mbḥḍ.

69

LES ÉVOLUTIONS DU CALENDRIER

301

« A été (conclu) cet engagement 75 en l’année cent quarante-six, pendant l’année de Bakrān 76 b. ʿAmmīrataʿ, que Qawlān a établi », w-kw(6)n 77 ḏn mʿhnn b-ḫrfn (7) ḏ-l-s¹ṯt w-ʾrbʿy w-m(8)ʾtm b-ḫrf Bkrn bn ʿ(9)mrtʿ ḏ-s²m Qwln

Il n’est guère douteux que l’inscription provient du territoire ḥimyarite, plus précisément de la commune de Maytamum, comme le suggèrent la mention de Hakirum (chef-lieu de Maytamum) et de Qawlān (lignage des princes de Maytamum) 78 : le mode de datation est donc ḥimyarite. S’il n’en résulte pas nécessairement que l’ère utilisée soit celle de Ḥimyar (c’est-à-dire celle de Mabḥūḍ ibn Abḥaḍ), puisqu’il a pu exister des ères locales, cela paraît vraisemblable : d’une part, on ne connaît aucune ère locale à l’intérieur de Ḥimyar ; d’autre part, la graphie, caractéristique de la fin du 1er s. av. è. chr. ou du début du 1er s. ap., s’accorde bien avec la date obtenue, 36–37 è. chr., si on considère que le nombre 146 se réfère à l’ère ḥimyarite. L’hypothèse que l’ère utilisée soit celle de Radmān (ce qui donnerait 220–221 è. chr.), comme l’a proposé Walter Müller 79, semble infondée puisque rien ne suggère que Sinabān ait dépendu de Radmān. L’expression ḏ-s²m Qwln a été traduite de diverses manières 80. C’est parce qu’on avait pas encore reconnu dans Qwln est le nom du lignage des princes de la commune Maytamum, souvent mentionné sans le bn qui est de règle ailleurs 81. Voir l’arabe ʿahana, « contracter, prendre un engagement ; conclure un pacte » 76 En arabe, Bakr et Bakrān ne prennent jamais l’article. 77 Il est vraisemblable qu’il faut corriger ainsi le kw(6)ḥ de la copie. 78 Voir aussi les mentions de la ville de Hakirum (dans Kh-Ḥammat al-Ḍabʿ 14 / 1 et 17 / 2) et de la commune Myt(m)[m] (3) ḏ-Hkrm (dans KhḤammat al-Ḍabʿ 15 / 2–3). Mais il est vrai qu’on a également une mention de la commune Yuhabshir dans Kh-Ḥammat al-Ḍabʿ 1 et 16. 79 Müller 2010, p. 18. 80 Le site DASI traduit ḏ-s²m Qwln : « That the qayl set up ». Albert Jamme propose « him of (the clan) Shaym, the ruler ». 81 Les princes de Maytamum sont les bnw Qwln w-ḏ-ʿrw : voir MAFRAY-Hajar Ṣabāḥ 1 ; Kh-Umayma 2 / 4 (mʾdbt Qwln), 3 / 3-4 (ʾbʾr 75

302

CHRISTIAN JULIEN ROBIN

Quant à Bakrān b. ʿAmmīrataʿ, c’est probablement le nom d’un éponyme donnant son nom à l’année. C’est d’autant plus vraisemblable que, dans un autre graffite de la même région, on retrouve la même formulation, avec un éponyme et un nombre : « dans l’année de Mlḥm, celle de deux cents quarante-deux », b-ḫ(r)fm (7) ḏ-Mlḥm ḏ-l-ṯny [w-](8)ʾrbʿy w-ṯty [mʾtm] (Kh-Jarf alNaʿīmiyya 12).

On a également, vers la même époque, une datation par éponyme dans le pays de Radmān 82. Ces documents font entrevoir que la datation par éponyme — le système ancien — n’a pas été remplacée instantanément par la datation d’après une ère : il y a eu une période plus ou moins longue pendant laquelle les deux systèmes ont été employés simultanément. 3. Le calendrier et l’ère de deux communes de la mouvance ḥimyarite, Radmān et Maḍḥàm Deux communes, Radmān et Maḍḥàm, qui jouxtent Ḥimyar à l’est et au sud-est sur les Hautes-Terres, mais débordent aussi vers le Bas-Pays dans le wādī Markha (grand wādī entre Bayḥān et Ḍuraʾ) ou au sud de la barrière du Kawr (orientée est-ouest, juste au sud de Bayḍāʾ-Ḥaṣī), doivent être examinées à part parce que, jusque vers le milieu du 3e s., elles n’ont pas eu exactement la même histoire. Comme Ḥimyar, ces deux communes ont appartenu au royaume de Qatabān, ont rédigé leurs inscriptions en langue qatabānique et ont vénéré ʿAmm, le grand dieu de Qatabān. Comme Ḥimyar, elles se sont séparées de Qatabān, mais à une date moins assurée, entre la fin du 2e et le milieu du 1er s. av. è. chr., et sont passées de la langue qatabānique à la langue sabaʾique. Il faut ajouter que, dans un passé très reculé, vers la fin du 8e s. av. è. chr., Radmān avait été un royaume. Cela peut contribuer b(4)n Qwln, 204 ḥim.), 4 / 3 (mqtwy S²rḥbʾl (3) bn S¹mhs¹mʿ w-Qwln), 13 / 3 (Ṯw(3)rm Qwln), 16 / 3 (mqt(3)wyy Rfʾ Yhṣbn Qwln) ; Jarf al-Naʿīmiyya 14 / 2 (mʾdb Qwln) ; Kh-Hakir 1 / 1 (Krb Yhṣbn Qwln (2) w-ḏ-ʿrw). 82 Voir ci-dessous, p. 20.

LES ÉVOLUTIONS DU CALENDRIER

303

à expliquer son aspiration à l’indépendance. On ignore si Maḍḥàm a eu des rois à une époque très ancienne. Radmān et Maḍḥàm, (re)devenues indépendantes, datent leurs inscriptions selon une ère qui leur est propre. Dans le passé, on a supposé qu’elles avaient également leur propre calendrier. On peut dire aujourd’hui que leur calendrier est probablement le même que celui de Ḥimyar. Pour Radmān, seuls quatre noms de mois sont attestés. Trois sont partagés avec Ḥimyar (ʿln 83, mḏrʾn 84 et ṣyd 85) ; quant au quatrième, ṣrb qdmn (« ṣurāb le premier ») 86, c’est également un mois du calendrier de Ḥimyar, mais avec l’ajout du qualificatif. Neuf inscriptions donnent les millésimes. Ce sont 24 (avec l’indication qu’il s’agit de l’année de Abīʿalī b. Ratʿ 87) ; 72 88 ; 81 et 85 89 ; 144 90 ; 144 91 ; 146 92 ; 148 93 ; et 179 (également avec l’indication qu’il s’agit de l’année de Abīʿalī b. Ratʿ) 94. La double datation d’al-Miʿsāl 2 / 14-15 95 se comprend mieux si le mois mḏrʾn vaut aussi bien pour l’année 179 d’Abīʿalī b. Ratʿ que pour l’année 363 de Mabḥūḍ ibn Abḥaḍ. Elle s’accorde avec l’idée que ce mois porte le même nom dans les deux calendriers. Il est moins sûr, mais assez vraisemblable, que Ḥimyar et Radmān ont une année ayant le même début et la même durée. Il n’est donc pas MAFRAY-Sāriʿ 5 / 3 et al-Miʿsāl 6 / 11. al-Miʿsāl 2 / 14, déjà cité ci-dessus, p. 15. 85 RES 3958 / 14 et al-Miʿsāl 4 / 12, auxquelles on peut ajouter YMN 15 (voir ci-après) 86 UAM 327 / 2. 87 UAM 327 / 2, « En son mois ṣurāb le premier, en l’année vingtquatre de l’année d’Abīʿalī b. Ratʿ », wrḫ-s¹ ṣrb qdmn b-ḫrfn ḏ-l-ʾrbʿt w-ʿs²ry bn ḫrf ʾbʿly bn Rtʿ. 88 MAFRAY-Sāriʿ 6. 89 MAFRAY-al-Ḥijla 1. 90 RES 3958. 91 YMN 10. 92 al-Miʿsāl 16. 93 al-Miʿsāl 4 / 12. 94 al-Miʿsāl 2 / 14. 95 Voir ci-dessus, p. 15. 83 84

304

CHRISTIAN JULIEN ROBIN

déraisonnable de conclure que Radmān et Ḥimyar ont le même calendrier et la même année, mais que chacune des deux communes a son propre système de décompte des années. Grâce à cette double datation, il est possible de situer l’inception de l’ère de Radmān en 74 è. chr. Concernant Abīʿalī b. Ratʿ, qui donne son nom à cette ère, on ne sait rien, ni sur lui ni sur son lignage. On peut supposer qu’il est à l’origine de la réforme du décompte des années. On peut adjoindre à ce petit corpus un texte de la région de Qāniya (l’antique Qāniʾatum), à quelque 25 km au nord-nord-est d’al-Miʿsāl 96. La région, précédemment qatabānite 97, est annexée par la commune de Radmān vers le milieu du 1er s. è. chr. On ne saurait dire si ce texte est antérieur ou non à cette annexion, mais il en est très proche, si on se fonde sur la langue (qui conserve quelques caractères propres au qatabānite 98) et la graphie. Après l’invocation aux divinités, le texte donne comme date : « le mois de ṣyd, l’année de Hawfāʿat b. ʿĀmilān 99, le premier », wrḫ-hw ṣyd ḫrf Hwfʿt bn ʿmln qdmn.

Le mois de ṣyd n’est pas qatabānite. C’est donc bien une datation locale avec une année identifiée par un éponyme, comme à Qatabān et dans les deux inscriptions ḥimyarites examinées cidessus 100. L’adjectif « premier » (qdmn) ne se rapporte pas au mois ṣyd, mais à l’éponyme dont c’est la première année de charge. On YMN 15 (Biʾr al-ʿAyl, Ḥawrān). Le texte, connu par une copie, se lit : Hwfʿm bn Qḥlwm brʾ w-s¹ʿs²q w-nʾy w-qyḥ w-nqb (2) bʾr-hw ʾmhrth wyn-hw Brkn w-nḫl-hw Rbṯ (3) tḥtn wynn b-bḍʿ Qnʾtm b-rdʾ ʿm ḏ-Mb[r](4)qm w-ʿm ḏ-Mnḫm wʾlh-s¹ww w-b-rdʾ s²ms¹-h[w] (5) wrḫ ṣyd ḫrf Hwfʿt bn ʿmln qdmn. Il présente apparemment quelques erreurs de copie, notamment ʾmhrth wyn-hw, à corriger peut-être en ʾmhrtn l-wyn-hw. 97 On ignore à quelle commune qatābānite elle se rattache. Ce peut être à la commune de Maryamatum (wādī Ḥarīb) dans les Basses Terres ou à une commune non identifiée de la montagne. 98 Ce sont le verbe s¹ʿs²q et pronom suffixe -s¹ww dans ʾlh-s¹ww. 99 Cet anthroponyme, inconnu de la tradition arabo-islamique, peut être rapproché de ʿĀmila. 100Voir pp. 17–18. 96

LES ÉVOLUTIONS DU CALENDRIER

305

peut donc supposer que, comme à Qatabān, cet éponyme était en fonction pendant deux années. La date à laquelle Radmān se sépare de Qatabān n’est pas connue précisément. La numismatique offre cependant un repère chronologique intéressant qui fait remonter cette sécession aux décennies qui suivent 50 av. è. chr. Dans le monnayage qatabānite, le nom du palais royal, Ḥarīb, apparaît à l’exergue des monnaies d’argent vers 50 av. è. chr. Une monnaie unique, Huth 376 101, qui est une imitation de ces premières séries mentionnant Ḥarīb remplace ce nom par Hirrān, le palais des princes de Radmān : c’est donc une monnaie qui a été frappée par un prince de Radmān. Comme la frappe des monnaies d’argent est une affirmation de grande autonomie, on peut donc en déduire que Radmān est déjà séparé de Qatabān à une date antérieure au début de l’ère chrétienne, peut-être proche de 50 av. Les mois utilisés dans la commune de Maḍḥàm portent également des noms semblables à ceux de Ḥimyar. Ce sont ʿln 102, ḏmḏrʾn 103, qyẓ 104 et ṣyd 105. Les millésimes attestés sont 49 (de l’année de Nabaṭum dhu-Kharīf 106), 100 107, 102 108, 120 109, 162 110, 163 111, Huth 2010, pp. 122–123. Ja 889 / 2 ; MQ-dhū Wayn 1 / 3 ; MQ-dhū Wayn 2 / 2 ; MQdhū Wayn 9 / 3. 103 MQ-Ḥayd Mūsà 1 = RES 4196 / 4 ; Shirjān 12 = Ja 2356 a / 9 ; MAFRAY-Ḥaṣī 5 / 6. 104 MQ-al-Ḥāṭ 1 / 6. 105 Buraʿ al-Aʿlà 1 / 6. 106 Ḥaṣī 8, « En l’année quarante-neuf ans de l’année de Nabaṭum dhu-Kharīf », b-(ḫrf) (3) ts¹ʿt w-ʾrbʿhy ḫyftn bn (4) [ḫr](f) Nbṭm ḏ-H̲rf. La vocalisation « Kharīf » est arbitraire. L’« anthroponyme » Khārif que l’on trouve dans les généalogie arabes n’est pas significatif parce qu’il dérive du nom d’un canton de Hamdān. 107 Ja 889 / 2 ; MQ-dhū-Wayn 1 / 3. 108 MQ-dhū Wayn 2 / 2. 109 MQ-dhū Wayn 4. 110 MQ-Ḍarāwiʿ 3. 111 MQ-Ḍarāwiʿ 2. 101 102

306

CHRISTIAN JULIEN ROBIN

172 112, 275 113, 301 114, 316 (à nouveau dans le comput de Nabaṭum dhu-Kharīf 115), 345 116, [3]65 117 et 388 118. L’inception de l’ère de Maḍḥàm n’est pas connue avec précision. Mais la mention de deux rois ḥimyarites dans des inscriptions datées permet de la situer vers 33 av. è. chr., avec une marge d’erreur de l’ordre de 10 ans. De Nabaṭum dhu-Kharīf, on ne sait rien sinon que le nom de lignage dhu-Kharīf apparaît dans deux textes comme vassal des banū Haṣbaḥ, les princes de Maḍḥàm 119. L’examen des calendriers de Ḥimyar et des communes de la mouvance ḥimyarite, avant 380 è. chr., permet de tirer quelques conclusions générales. Il est inutile de s’attarder sur la longue liste de nos ignorances, notamment la durée de l’année et du mois. Je vais me concentrer uniquement sur ce que nous savons. En premier lieu, il est vraisemblable désormais que, si les trois communes de Ḥimyar, Radmān et Maḍḥàm comptent différemment les années, elles utilisent le même calendrier, c’est-à-dire la même liste de mois. Or, de la fin du 1er s. av. è. chr. jusqu’aux premières décennies du IIIe è. chr., ces trois communes sont désunies et souvent impliquées dans des coalitions antagonistes. Leur calendrier commun remonte donc à une époque antérieure, pendant laquelle toutes trois étaient unies, probablement au sein du royaume de Qatabān. Une inscription non datée, mais relativement ancienne

RES 4197 bis. MQ-al-Jifjif 1. 114 MQ-al-Ḥāṭ 1. 115 MQ-Ḥayd Mūsà 1 = RES 4196, « Au mois de ḏ-mḏrʾn de (l’année) trois cent seize, en années de l’ère de Nabaṭum dhu-Kharīf », b-wrḫn ḏ-mḏrʾn ḏ-l-s¹ṯt-ʿs²r w-ṯlṯ mʾtm ḫryftm bn ḫryf N(5)bṭ(m )ḏ-H̲rf 116 Shirjān 12 = Ja 2356 a. 117 Ḥaṣī 5. 118 Buraʿ al-Aʿlà 2. 119 Voir CSAI II, 12 = CIAS 95.11/o2 n° 2 = Ja 878 / 1 ([Ḍrn ḏ](H)ṣbḥ w-H̲rf w-…rn w-Gmlm .[..]) et Shirjān 27 = Ja 1819 / 3 (… bnw Hṣbḥ w-Yʿgf ḏ-ʿms²bm w-Ḥ(3)ṭn w-H̲rf w-ḏ-Bdtm w-Rḥbtm etc.). 112 113

LES ÉVOLUTIONS DU CALENDRIER

307

puisqu’elle est en boustrophédon, confirme l’ancienneté de ce calendrier 120. Si Ḥimyar, Radmān et Maḍḥàm ont les mêmes mois, il est très vraisemblable qu’elles ont aussi la même année, avec le même début, la même durée et les mêmes divisions. Deux mois sont attestés soit sans qualificatif soit avec le qualificatif « le premier », qyẓ qdmn et ṣrb qdmn. Aucun mois n’est qualifié de « second » ou de « dernier », mais il a dû en exister puisque, à Qatabān, comme à Sabaʾ, de tels mois sont connus 121. L’ajout occasionnel du qualificatif « le premier » (et hypothétiquement « le dernier ») signifie que certains mois pouvaient être redoublés : on ajoutait donc de temps en temps un mois intercalaire pour corriger la différence de presque 11 jours entre l’année lunaire de 12 mois et l’année solaire. On notera encore que, dans les calendriers de Ḥimyar et de Radmān, le mois qui semble redoublé est différent : à Ḥimyar, ce serait le mois ḏ-qyẓn (dhu-qayẓān, V) et peut-être les mois ḏ-mḏrʾn (dhu-madhraʾān, VI) et ḏ-ḫrfn (dhu-khirāfān, VII), tandis que, à Radmān, ce serait ṣrb (ṣurāb, IX). Chaque commune aurait donc pratiqué l’intercalation à sa manière. Le décompte des années, en revanche, est différent. Chacune des trois communes a créé sa propre ère, en rupture avant l’antique et incommode système des éponymes. Il faut ajouter que ces trois communes sont les seules en Arabie méridionale à avoir fait ce choix. Dans les grands royaumes traditionnels, Sabaʾ, le Ḥaḍramawt et Qatabān, on a conservé la datation par éponymes jusqu’à la fin. 122 Le choix de la datation par une ère trahit probablement un moindre conservatisme et une plus grande ouverture aux nouveautés du monde extérieur. La pratique de dater d’après une ère remonte à l’extrême fin du 1er s. av. è. chr. Sa plus ancienne attestation vient de Ḥimyar, où Kh-Umayma 25 (voir ci-dessus, p. 12). Les premières attestations de mois de ce calendrier apparaissent ensuite dans le courant du Ier è. chr. (YMN 15, encore daté par éponyme; Ja 889 et MQ-dhū-Wayn 1 et 2, vers 75 è. chr.). 121 Voir ci-dessous, pp. 50–51. 122 Voir ci-dessous, pp. 24–25. 120

308

CHRISTIAN JULIEN ROBIN

Khaldūn Nuʿmān a découvert une inscription datée de 102 ḥim., soit 8–7 av. è. chr. La première inscription datée dans l’ère de Maḍḥàm remonte à 49 maḍḥ. soit vers 15–20 è. chr. Quant à l’ère de Radmān, sa première attestation, qui est nettement plus tardive, porte le nombre de 24 (98–99 è. chr.). La première date en ère ḥimyarite est postérieure de près d’un siècle à l’inception de cette ère (110 av. è. chr.). On peut également observer que la datation par une ère apparaît simultanément dans Ḥimyar et à Maḍḥàm. On peut en déduire que la création de l’ère ḥimyarite est sans doute très postérieure à 110 av. è. chr. La date de l’inception ne se référerait donc pas à la création de l’ère, mais à un événement fondateur reculé dans le temps. C’est d’autant plus plausible que l’existence même de Ḥimyar n’est pas attestée avant 70 av. è. chr. environ, que ce soit dans l’épigraphie ou la numismatique 123. Dans les communes de Maḍḥàm et de Radmān, en revanche, il est vraisemblable que la date de l’inception corresponde à la création de l’ère. De manière évidemment hypothétique, la création de l’ère ḥimyarite pourrait être contemporaine du premier monnayage, que nous situons vers 70–50 av. è. chr. L’ère de Maḍḥàm serait instituée quelques décennies plus tard, à l’imitation de Ḥimyar, et celle de Radmān, beaucoup plus tard, en 74 è. chr. De manière logique, les premières attestations des ères de Maḍḥàm et Radmān interviennent à une date plus rapprochée de l’inception, 49 ans et 24 respectivement. On ignore quels sont les événements qui correspondent aux inceptions des ères de Ḥimyar, Maḍḥàm et Radmān, mais deux hypothèses peuvent être avancées. Pour Ḥimyar (et de manière moins assurée pour Maḍḥàm), ce pourrait être une étape importante vers l’indépendance. Pour Radmān, ce serait plutôt l’arrivée au pouvoir d’une nouvelle lignée de princes, en relation avec une réforme territoriale. La première preuve de l’existence d’une entité ḥimyarite autonome est le monnayage au nom de Mbhl ḏ-Rydn, qui présente les mêmes caractères que celui du roi de Qatabān Ydʿʾb Ḏbyn bn S²hr (Huth 366, dans Huth 2010, pp. 120–121 ; Robin 2016 c à paraître). 123

LES ÉVOLUTIONS DU CALENDRIER

309

Chacune de ces nouvelles ères est identifiée par le nom d’un personnage qui est probablement à l’origine du nouveau système. Comme nous l’avons vu, deux de ces « computistes », Mabḥūḍ b. Abḥaḍ et Nabaṭum dhu-Kharīf, appartiennent apparemment à des familles qu’on peut qualifier de petite noblesse. On ne sait rien du troisième Abīʿalī b. Ratʿ. Comme je l’ai déjà dit, les Ḥimyarites et les communes voisines de Maḍḥàm et de Radmān sont, en Arabie méridionale, les seules populations qui ont identifié les années par un numéro d’ordre. Cette pratique apparaît vers le début de l’ère chrétienne. Auparavant, il est possible que ces communes aient désigné les années par la mention d’un éponyme 124. Partout ailleurs, de Maʿīn au Ḥaḍramawt, en passant par Samʿī, Sabaʾ et Qatabān, depuis les origines de la civilisation sudarabique jusqu’à l’établissement de la domination ḥimyarite, l’année est identifiée par le nom d’un éponyme. On ignore quel était le rôle exact de ce personnage même si, à Sabaʾ et à Qatabān, il semblerait qu’il ait été en charge de certains rites religieux. Le fonctionnement de l’éponymat est mal connu. On sait seulement que, dans le royaume de Sabaʾ, les éponymes étaient issus de quatre lignages qui assumaient la charge à tour de rôle. Ces lignages, en dehors de Kabīr-Khalīl, n’appartenaient pas à la grande noblesse. Toujours à Sabaʾ, la charge d’éponyme pouvait être exercée à plusieurs reprises, jusqu’à neuf fois 125. Ailleurs, la charge d’éponyme était apparemment exercée une seule année (Samʿī), deux (Maʿīn et Qatabān) 126 ou quatre (Ḥaḍramawt) 127. Les personnages en charge du calendrier en Arabie du Sud ressemblaient peut-être à ceux du Ḥijāz au début du 7e s. è. chr. À la fin du pèlerinage de Makka, le descendant d’un personnage surnommé al-Qalammas avait la responsabilité de proclamer si, dans Pour Ḥimyar voir Kh-Ḥammat al-Ḍabʿ 19 = Ja 3199 = Nājī 1 et Kh-Jarf al-Naʿīmiyya 12, cités ci-dessus, pp. 17–18 ; pour Radmān, YMN 15 (Biʾr al-ʿAyl, Ḥawrān), cité ci-dessus, p. 20 et n. 96. 125 Robin 1994. 126 Beeston 1956, pp. 25–27 pour Maʿīn et pp. 27–28 pour Qatabān. 127 Beeston 1956, pp. 34–35 ; Francuzov 2014, pp. 117–124. 124

310

CHRISTIAN JULIEN ROBIN

l’année qui allait commencer 128, il était prévu ou non d’ajouter un mois intercalaire (afin de maintenir le pèlerinage en phase avec les saisons et donc avec les foires). Le Qalammas et ses descendants étaient eux aussi des personnage assez obscurs, qui n’exerçaient apparemment aucune responsabilité tribale ou religieuse, ni dans leur tribu (Kināna) ni dans celle de Quraysh 129. 4. Le passage du calendrier national au calendrier de Ḥimyar dans les royaumes de Sabaʾ et du Ḥaḍramawt après 270 et 300 Le royaume de Sabaʾ est annexé par celui de Ḥimyar dans les années 270. Quelques textes jettent un peu de lumière sur la manière dont se fait le passage du calendrier sabéen au calendrier ḥimyarite. Avant son annexion, le royaume de Sabaʾ avait ses propres noms de mois et identifiait les années par la mention d’un éponyme. Après l’annexion, la datation par éponyme disparaît rapidement : elle n’est attestée qu’une seule fois dans une inscription sabéenne qui remonte à la première partie du règne de Shammar Yuharʿish (287–296), avant la guerre contre le Ḥaḍramawt 130. On peut donc dater sa disparition vers 300. C’est d’autant plus assuré que l’on dispose de nombreuses inscriptions provenant du Grand Temple de Maʾrib et datant du règne soit du dernier souverain sabéen (Nashaʾkarib Yuhaʾmin Yuharḥib) soit de celui des premiers rois ḥimyarites (Yāsirum Yuhanʿim et son fils Shammar Yu-

Le pèlerinage de Makka avait lieu à la fin de la première décade de dhū ʾl-ḥijja, le dernier mois de l’année. 129 Voir ci-dessous, pp. 42–43 et 48–49. 130 Ja 653. Nous ne retenons dans notre décompte que les textes qui mentionnent un roi ; il n’est donc pas impossible qu’il en existe d’autres, que nous ne savons pas dater précisément (comme Ja 735). 128

LES ÉVOLUTIONS DU CALENDRIER

311

harʿish) 131. En dehors de cette unique datation par éponyme, les textes sabéens utilisent désormais l’ère ḥimyarite 132. Les noms de mois sabéens, en revanche résistent durablement. On ne sera pas étonné d’en trouver un dans l’inscription du règne de Shammar Yuharʿish roi de Sabaʾ et de dhu-Raydān utilisant la datation par éponyme, déjà évoquée 133. Une inscription énonce la date de manière hybride, en associant un mois sabéen avec l’ère ḥimyarite : dhu-Sabaʾ 461 ḥim. 134. Deux autres se servent des mois sabéens pour évoquer divers événements comme le grand pèlerinage annuel de Marib, qui a lieu en dhu-abhī 135, ou de très importants travaux de réfection sur la Digue de Marib, effectués sur l’ordre du roi, en « dhu-Sabaʾ, īlʾīlāt et abhī » 136. La persistance de l’utilisation des mois sabéens s’observe également dans les documents privés en écriture cursive sur bâtonnets de bois. Une correspondance remontant à un roi de date incertaine, mais à situer assurément au 4e s., est datée du seul mois sabéen de

Dans le Grand Temple de Marib, ce sont 12 inscriptions du règne de Nashaʾkarib qui sont datées par éponymes (Ir 22, 25, 26, 70 ; Ja 610, 611, 613, 615 (deux fois), 618, 653, 877 [= CIAS 95.11/o1 n° 2] ; MB 2002 I-28) contre une seule pour les règnes de Yāsirum Yuhanʿim et de Shammar Yuharʿish. 132 Voir MQ Maʾrib 1 et MB 2004 I-123 du Grand Temple de Maʾrib, qui sont examinées ci-après ; YM 1950 / 5 (ḏ-ḫrfn, [47]3 ou [48]3 ḥim.) ; Shibām Kawkabān 1 (ḏ-qyṣn, 487 ḥim.). Voir aussi les textes sur bâtonnets mentionnés ci-après. 133 Ja 653 / 10 et 14, où le dieu Almaqah est remercié pour avoir accordé des pluies abondantes au mois de dhu-mlyt. 134 MQ-Maʾrib 1 : [ḏ-](10)(S¹)b(ʾ). Dans le Grand Temple de Marib, en plus de cette inscription, seules deux autres utilisent l’ère ḥimyarite : ce sont MB 2004 I-147 (489 ḥim., sans nom de mois) et MB 2004 I-123 (dhumabkarān 411 ḥim). 135 Ja 651 / 14-21, également du règne de Shammar Yuharʿish roi de Sabaʾ et de dhu-Raydān. 136 Ja 671 + 788 / 25, du règne de Thaʾrān Yuhanʿim en corégence avec son fils Malkīkarib Yuʾmin, vers 370 è. chr. 131

312

CHRISTIAN JULIEN ROBIN

dhu-Hōbas 137. Quatre autres bâtonnets utilisent une datation hybride, avec des mois sabéens et l’ère ḥimyarite, pratique qui semble durer jusque vers le milieu du 5e s., puisque les années attestées sont : 48[.] 138, 495 139, 522 140 et 549 ḥim. 141. Deux conclusions peuvent être tirées de ces observations. La première est que les noms de mois sabéens sont utilisés plus longtemps dans la sphère privée que dans la sphère publique. Leur dernière attestation dans les inscriptions se situe vers 370, mais sur les bâtonnets vers 440. La seconde conclusion est que, après l’annexion, les noms de mois sabéens et ḥimyarites sont utilisés indifféremment, notamment dans les documents datés en ère ḥimyarite : ils sont donc interchangeables, ce qui implique que les calendriers sabéen et ḥimyarite sont plus ou moins identiques en ce qui concerne la durée de l’année et des mois. Pour le Ḥaḍramawt, la seule donnée réside dans la double datation de ʿAbadān, déjà mentionnée 142. Elle indique que le système local de datation par éponyme pouvait encore être utilisé près de 60 ans après que Ḥimyar ait conquis la capitale, Shabwat. 5. Le calendrier de Ḥimyar après la conversion au judaïsme (c. 380) Après 380, un même calendrier et une même ère sont utilisés dans l’ensemble du royaume de Ḥimyar, avec quelques nuances. La liste des douze mois est désormais dans l’ordre alphabétique ḏ-ʾln, ḏ-ʿln, ḏ-dʾwn (variante ḏ-dʾw), ḏ-ḫrfn (variante, ḏ-ḫrf), ḏ-ḥltn, ḏḥltn ʾḫ(r)[n], ḏ-mʿn (variante ḏ-mʿwn), ḏ-mbkrn, ḏ-mḏrʾn, ḏ-mhltn, qyẓn, ḏX.BSB 139 = Mon.script.sab. 320, dont l’auteur est le roi Thaʾrān, roi de Sabaʾ, de dhu-Raydān, du Ḥaḍramōt et de Yamnat. La titulature implique une date postérieure à la conquête de Shabwat, la capitale du Ḥaḍramawt (Robin 2014 b, pp. 53–54). 138 X.SBS 35 / 6-8, (ḏ)-S¹(b)ʾ 48[0]. 139 L025 / 9-10, ḏ-ns¹(w)[r] (10) 495. 140 X.SBS 46 / 4 (ḏ-ʾln, sans mention de l’année ; mais, à l. 7, ḏ-ʾbhy 522). 141 X.BSB 62 / 6-7 (ḏ-S¹bʾ 549 ḥim.). 142 Voir ci-dessus, p. 16. 137

LES ÉVOLUTIONS DU CALENDRIER

313

qyẓn, ḏ-ṣrbn (variante z-ṣrbn) et ḏ-ṯbtn. Le nom des mois est désormais normalisé, avec systématiquement un ḏ- initial. Si les noms de mois sont d’ordinaire ceux de Ḥimyar, il y a quelques exceptions. Nous avons vu que le calendrier sabéen est utilisé sporadiquement jusque vers le milieu du 5e s. Par ailleurs, deux autres calendriers apparaissent alors. Dans deux inscriptions paléo-arabes des environs de Najrān, les mois ḥimyarites sont remplacés par ceux des Arabes préislamiques, al-muʾtamir et burak 143. Par ailleurs, dans quelques textes, le nom du mois est apparemment remplacé par celui du signe du zodiaque 144. Une inscription provenant de Maḍḥàm est ainsi datée du mois de (ṯ)wrm, « taureau » [à moins qu’il ne faille lire (s²)wrm] 551 145. Ce même nom pourrait se trouver sur deux bâtonnets de lecture et d’interprétation problématique 146. Enfin, un bâtonnet daté de 554 ḥim. mentionne deux fois un mois nommé ʿqrbn, « le scorpion », qui n’est ni ḥimyarite ni sabéen 147. Cette appellation a peut-être un écho dans l’almanach du sultan rasūlide du Yémen, déjà évoqué, qui appelle une saison ʿaqārib al-shitāʾ, « les scorpions de l’hiver », correspondant approximativement à novembre-janvier 148. Pour l’ère utilisée, on observe également une exception: dans les deux inscriptions paléo-arabes des environ de Najrān, l’année n’est pas donnée selon l’ère ḥimyarite, mais très vraisemblablement d’après celle de la province romaine d’Arabie.

Robin et alii 2014. Dans le royaume de Ḥimyar, quelques signes du zodiaque sont représentés sur des plaquettes décoratives et, pour certains, identifiés par un nom : nny (bélier), ṯ[w]rn (taureau), s²r[ṭn] (cancer), ḥẓyn (sagittaire), s¹bltn (vierge) et (ʾ)[r[ poissons) (Laffitte 2003). 145 MQ-al-Qashʿa 1. 146 X.SBS 8 / 4, w-b wrḫn (ṯw)rn ḏ-… ; X.SBS 69 / 4-5, b-wrḫ ḏ(5) (ṯw) [… 147 X.SBS 152 / 9-10 et 10–11, deux fois ḏ-ʿqrbn 554 ḥim. 148 Varisco 1994, p. 119. Cette appellation, qui n’était pas propre au Yémen, appartiendrait à un système liant les années solaires et lunaires. 143 144

314

CHRISTIAN JULIEN ROBIN

B. L’ORDRE DES MOIS, LE DÉBUT DE L’ANNÉE, L’INTERCALATION ET LES ÉQUIVALENCES EN

Sur l’organisation et le fonctionnement du calendrier ḥimyarite, les connaissances se limitent à l’ordre des mois, au commencement de l’année, à l’existence de mois intercalaires et à des équivalences approximatives entre mois ḥimyarites et juliens. Tout le reste est ignoré, notamment la durée précise des mois et de l’année. CALENDRIER JULIEN

1. L’ordre des mois Les inscriptions ne permettent pas de reconstruire l’ordre des mois du calendrier ḥimyarite dans sa totalité 149. C’est grâce à de providentiels textes du Moyen-Âge yéménite (13e et 14e s.) traitant de l’année agricole que cet ordre est connu. Ce type de textes se réfère d’ordinaire aux mois solaires syriens 150. Mais quelques auteurs, sans explication, donnent pour chaque mois syrien un autre nom qui se trouve être la transcription arabe d’un mois ḥimyarite antique 151. Parmi eux, le plus notable est le troisième sultan rasūlide (1295–1297), al-Malik al-Ashraf Mumahhid al-Dīn ʿUmar b. Yūsuf, dans son almanach, qui est l’un des 50 chapitres du traité d’astronomie intitulé al-Tabṣira fī ʿilm al-nujūm, rédigé vers 1271. ʿUmar, avant de succéder brièvement à son père, qui régna 46 ans, fut un savant prolifique dans différents domaines du savoir. La Tabṣira, dont on ne connaît qu’une seule copie, est inédite, à l’exception de l’almanach qui a été publié et étudié par Daniel Varisco (1994). Voir n. 10, l’état des reconstructions dans les années 1950, avant la découverte des listes médiévales. 150 Les mois syriens utilisés dans ces listes sont solaires (et non lunisolaires comme dans l’Antiquité) parce que, en Orient, la christianisation a généralisé l’usage du calendrier de Jules César qui servait à la détermination des fêtes principales (tout particulièrement celle de Pâques) dès le 6e ou le 7e s. 151 Les noms présentent diverses formes : les variantes peu significatives ne sont pas signalées. 149

LES ÉVOLUTIONS DU CALENDRIER

315

L’almanach se fonde sur le calendrier syrien et commence logiquement avec le premier mois de ce calendrier, tishrīn al-awwal (octobre). Le nom du mois ḥimyarite est systématiquement apposé à celui du mois syrien (par exemple tishrīn al-awwal — dhū ṣirāb). Vient ensuite une liste de tous les jours du mois avec, pour chacun, quelques brèves observations dérivées de la tradition des almanachs arabes ou des pratiques et du folklore locaux. Les mois ḥimyarites sont également cités dans un poème agricole médiocrement édité par Muḥammad al-Akwaʿ (1981), et dans trois autres almanach inédits (notamment Bughyat al-fallāḥīn, du sultan rasūlide al-ʿAbbās b. ʿAlī b. Dāʾūd, mort en 1380 152). Ces textes donnent le nom des mois ḥimyarites médiévaux et leur ordre (grâce aux équivalents syriens), mais aussi des indications utiles pour la vocalisation et même la lecture du sudarabique 153. Les noms médiévaux diffèrent quelque peu des noms antiques. Il arrive que l’article ḥimyarite postposé -ān soit supprimé ou remplacé par l’article arabe préposé al-. Par ailleurs, les noms médiévaux peuvent être déformés à la suite de négligences scribales qui ont provoqué des changements de ponctuation. Ces différences ne sont donc pas significatives. Aucun de ces textes médiévaux (pour autant qu’on le sache, puisque deux seulement sont édités et un seul de manière convenable) ne donne la moindre explication ni n’indique la moindre source, ce qui a incité les chercheurs à les utiliser avec prudence. On s’est demandé si les mois ḥimyarites médiévaux étaient encore véritablement utilisés dans les campagnes ou si ce n’était qu’une simple curiosité philologique. En faveur d’un calendrier vivant, on retiendra que les noms de mois présentent des variations qui pourraient être des usages régionaux et que certains d’entre eux sont restés en usage jusqu’à l’époque contemporaine.

A. F. L. Beeston en a utilisé les données dans son article de 1974. Ils ont permis d’établir que le nom du premier mois devait se lire dhu-ḥullatān et non dhu-ḥiggatān (arabe dhū ʾl-ḥijja) comme on le croyait (voir n. 10). 152 153

316

CHRISTIAN JULIEN ROBIN

Voici la liste des mois ḥimyarites donnée par les textes médiévaux, en unifiant la vocalisation 154 et en corrigeant les fautes scribales manifestes 155 : Mois ḥimyarites (listes médiévales)

Mois solaires syriens Mois ḥimyarites (avec son équivalent julien) (textes épigraphiques)

5. dhū (ʾl-)ḥulla, dhū ḥulal

shubāṭ (février)

6. dhū maʿūn 7. dhū ʾl-thāba 8. dhū (ʾl-)mabkar 9. dhū ʾl-qiyāẓ / qayẓ 10. dhū madhrā(n) 11. dhū (ʾl-)khirāf 12. dhū ʿallān 1. dhū (ʾl-)ṣurāb, dhū ʾl-ṣurāb I 2. dhū ʾl-muhla, dhū ʾl-ṣurāb II 3. dhū ʾl-āl, dhū ʾl-awl 4. dhū daʾw, dhū diʾā

ādhār (mars) nīsān (avril) ayyār (mai) ḥuzayrān (juin) tammūz (juillet) āb (août) aylūl (septembre) tishrīn I (octobre) tishrīn II (novembre) kanūn I (décembre) kanūn II (janvier)

I[A]. dhu-ḥullatān [IB. Intercalaire dhu-ḥullatān] II. dhu-maʿūn III. dhu-thābatān IV. dhu-mabkarān V. dhu-qiyāẓān VI. dhu-madhraʾān VII. dhu-khirāfān VIII. dhu-ʿallān IX. dhu-ṣurābān X. dhu-muhlatān XI. dhu-ālān XII. dhu-diʾāwān

On ignore si les équivalences entre le calendrier solaire syrien et le calendrier ḥimyarite sont exactes ou approximatives. Par exemple, pour le mois dhū ʾl-ṣurāb, les listes donnent tishrīn al-awwal (octobre), tandis que Nashwān al-Ḥimyarī, qui écrit un peu plus tôt (il est mort vers 1178), indique : « Ḥimyar appelle aylūl (septembre) “dhū ʾl-ṣirāb” parce que c’est l’époque où l’on coupe ce qui a été planté » 156. Une explication possible serait que dhū ʾl-ṣurāb équivaille non pas à octobre, mais à septembre-octobre ; une autre serait que

Les voyelles /i/ et /u/ se confondent en /ə/ dans un grand nombre de dialectes yéménites, de sorte que les éditeurs vocalisent indifféremment avec un /i/ ou avec un /u/ : voir ṣirāb / ṣurāb ou ḥilla / ḥulla. 155 Le nom du mois ḥimyarite dʾw se trouve ainsi avec les graphies dathaʾ, dibā ou dibāw, qui s’expliquent par de mauvaises ponctuations. Pour le détail de ces variations, voir Robin 2012, pp. 129–130. 156 wa-Ḥimyar tusammī aylūl dhā ʾl-ṣirāb li-anna fī-hi ṣirām al-zarʿ (Nashwān, Shams al-ʿulūm, texte, p. 60, sous ṢRB). 154

LES ÉVOLUTIONS DU CALENDRIER

317

Nashwān s’est trompé 157. Quoi qu’il en soit, le calendrier ḥimyarite médiéval est solaire puisqu’il règle la vie agricole. 2. Le premier mois de l’année Il est possible d’établir que le premier mois de l’année ḥimyarite était, aux 5e et 6e s., dhu-ḥullatān. En 564-565 ḥim. (454-456 è. chr.), le roi Shuriḥbiʾīl Yaʿfur effectue deux séries de grands travaux sur la Digue de Maʾrib. Il fait une première réparation, après une rupture provoquée par l’inondation de l’été 564 158 ; cette réfection est achevée en dhu-diʾāwān (XII, ≈ janvier) 564 159. Au printemps suivant, plus précisément en dhu-thābatān (III, ≈ avril) 565 160, une nouvelle rupture est provoquée par la crue de printemps 161. La réparation des dégâts, qui mobilise 20 000 hommes, est achevée en dhu-diʾāwān (XII, ≈ janvier) 565, manifestement après les pluies de l’été 162. Cette chronologie implique que le millésime change après le mois XII (≈ janvier) et que ce changement est déjà intervenu au mois III (≈ avril). Le nouvel an se situe donc soit au début du mois I (≈ février), soit au début du mois II (≈ mars).

Si on retenait l’idée que dhū ʾl-ṣurāb correspondait alors à septembre-octobre, il faudrait en déduire que le calendrier ḥimyarite ne s’était pas aligné sur le calendrier syrien ; il serait donc demeuré lunaire avec des mois intercalaires, ce qui semble peu plausible (voir ci-dessous, pp. 48– 49). Nous préférons supposer que Nashwān a donné une équivalence approximative. 158 Sadd Maʾrib 3 = CIH 540 / 54-56, ḏhb ḫrfn. 159 Sadd Maʾrib 3 = CIH 540 / 51-54, wrḫ-hw ḏ-dʾw(52)n ḏ-b-ḫrfn ḏ-l (53) ʾrbʿt w-s¹ṯy w-ḫ(54)ms¹ mʾtm. Pour signaler que l’équivalence donnée par les listes médiévales n’est pas sûre, mais seulement indicative, nous la faisons précéder par le symbole ≈. 160 Sadd Maʾrib 3 = CIH 540 / 60-61, b-wrḫn ḏ-ṯbtn ḏ-l-ḫms¹t (61) [w-s¹]ṯy w-ḫms¹ mʾtm. 161 Sadd Maʾrib 3 = CIH 540 / 58, ḏhb dṯʾn. 162 Sadd Maʾrib 3 = CIH 540 / 97-100, wrḫ-hw (98) ḏ-dʾw ḏ-l-ḫms¹(99)t ws¹ṯy w-ḫms¹ (100) mʾtm. 157

318

CHRISTIAN JULIEN ROBIN

L’inscription du roi Abraha, rédigée en dhu-maʿūn (II, ≈ mars) 658 ḥim. 163, commémore d’importants travaux à la suite d’une rupture survenue en dhu-madhraʾān (VI, ≈ juillet) [65]7 164. La réfection qui avait commencé en dhu-ṣurābān (IX, ≈ octobre) [65]7 165, est suspendue à cause d’une épidémie 166, puis reprend pendant la dernière semaine de dhu-diʾāwān (XII, ≈ janvier) 167 et est achevée en 58 jours 168. Ici, le changement de millésime, qui intervient après le mois IX (≈ octobre), est déjà effectif au mois II (≈ mars). Si on combine les données des inscriptions de Shuriḥbiʾīl Yaʿfur et Abraha (en supposant que le calendrier n’a pas subi de réforme radicale entre 450 et 550), le premier mois de l’année est nécessairement le mois I (≈ février). C’est confirmé par la comparaison des dates de Sadd Maʾrib 4 = DAI GDN 2002/20 (dhu-ḥullatān « le dernier » [I bis] 658) et Sadd Maʾrib 5 = CIH 541 (dhu-maʿūn [II] 658 ḥim.) : il est en effet assuré que le premier texte a été rédigé avant le second parce qu’il en est une ébauche inachevée. Les listes médiévales identifient dhu-ḥullatān avec février. Si cette équivalence était valide pour l’Antiquité, un début d’année en février ne serait pas banal. Dans l’Antiquité tardive, l’année commence en septembre à Byzance et en octobre chez les Syriens. Quant aux juifs, ils ont une année qui a apparemment deux commencements, l’un ancien et tombé en désuétude et l’autre moderne : d’un côté, le premier mois du calendrier est nîsān, qui commence idéalement 169 avec l’équinoxe de printemps, comme à BaSadd Maʾrib 5 = CIH 541 / 134-136, b-wrḫ ḏ-mʿn ḏ-l (135) ṯmnyt wḫms¹y w-s¹(136)ṯ mʾtm. Noter la tournure b-wrḫ ḏ-mʿn, avec annexion (comme en arabe), et non apposition comme c’est la règle en sabaʾique. 164 Sadd Maʾrib 5 = CIH 541 / 45-46, b-wrḫ ḏ-mḏr(46)[ʾ]n ḏ-l-s¹bʿt. 165 Sadd Maʾrib 5 = CIH 541 / 62–63, b-wrḫn ḏ-ṣrbn ḏ-l (63) s¹bʿt. 166 Sadd Maʾrib 5 = CIH 541 / 72–73. 167 Sadd Maʾrib 5 = CIH 541 / 96, b-mdt ḏ-dʾwn ʾḫrtn, « la dernière période de dhu-diʾāwān ». 168 Sadd Maʾrib 5 = CIH 541 / 131-132. 169 Dans la pratique, le début de nîsān et le Jour de l’an se déplacent de quelques semaines dans le cours d’un cycle de 19 ans. C’est seulement 163

LES ÉVOLUTIONS DU CALENDRIER

319

bylone ; mais, depuis une réforme difficile à dater, le Jour de l’an (le premier jour de l’année) est situé idéalement à l’équinoxe d’automne 170. Une année commençant avec l’équinoxe de printemps, qui remontait au Néo-babyloniens, avait été autrefois la norme dans diverses régions du Proche-Orient). On peut donc se poser la question de savoir si l’année ḥimyarite ne commencerait pas, elle aussi, plutôt qu’en février, avec l’équinoxe de printemps, ce que nous examinerons ci-dessous 171. 3. La mise en évidence d’un mois intercalaire prouvant que le calendrier ḥimyarite est lunaire Dans l’Antiquité, la grande diversité des calendriers peut se réduire à deux modèles, le calendrier lunaire babylonien et le calendrier solaire égyptien. Pour mesurer le temps, les références les plus naturelles sont la journée, le mois lunaire et l’année solaire (365,2422 jours) qui règle, avec la succession des saisons, la plupart des activités humaines 172. Cependant, si l’année solaire est bien une référence universelle, sa détermination est une entreprise compliquée parce qu’il n’existe pas d’événements astronomiques facilement observables par le commun des mortels pour fixer le commencement de l’année et sa durée. Les Égyptiens, dès les origines, ont un calendrier solaire fondé sur la crue du Nil. Leur année compte douze mois de 30 jours, avec un ajout de cinq jours appelés « épagomènes » en fin d’année, soit un total de 365 jours. Cette durée, déjà établie au IIIe millénaire avant l’ère chrétienne — ce qui était alors une approximation rela première année du cycle qu’ils tombent à leur date véritable. Voir cidessous, p. 45. 170 Voir aussi ci-dessous, p. 45. 171 Voir ci-dessous, pp. 45–47. 172 Il existe plusieurs définitions de l’année solaire. On peut distinguer, par exemple, l’année « vernale » qui est l’intervalle de temps qui s’écoule entre deux passages successifs du Soleil à l’équinoxe vernal, de l’année « tropique » (qu’on lui préfère aujourd’hui) qui est la moyenne du même intervalle sur l’ensemble du cycle des saisons ; mais ces différences importent peu pour notre propos.

320

CHRISTIAN JULIEN ROBIN

marquable —, est inférieure de 0,2444 jour par an à l’année solaire vernale : il en a résulté une dérive de près de 25 jours par siècle entre le calendrier et le rythme des saisons. Cette dérive, observée très tôt, n’a jamais été corrigée, sans doute par conservatisme, mais aussi parce qu’elle ne provoquait aucun bouleversement à l’échelle d’une vie humaine, avec des fêtes tombant toujours dans les mêmes saisons. La réforme tentée à l’époque hellénistique, après l’établissement d’une durée très précise de l’année solaire, n’a pas été appliquée. Il a fallu attendre la conquête romaine et l’adoption du calendrier réformé par Jules César (46 av. è. chr.) pour que l’ajout d’un jour tous les quatre ans (années bissextiles) apporte une meilleure correction. Le calendrier réformé par Jules César, puis légèrement modifié durant le reste d’Auguste, est le calendrier officiel de l’Empire romain, qui sert progressivement de modèle à tous les calendriers particuliers de l’Empire, notamment en Orient. Il est solaire lui aussi. Même si la durée des années égyptienne et romaine est solaire, il est manifeste que c’est le mouvement de la lune qui a conduit à diviser l’année en douze mois. Le mois lunaire (« synodique »), qui est facilement observable, a en effet une durée moyenne de 29,5306 jours. La plupart des sociétés anciennes, si on excepte l’Égypte, puis Rome qui s’en inspire, fondaient la durée du mois sur le mouvement de la lune. La détermination du début du mois a d’abord été fondée sur l’observation de la lune, puis sur le calcul. Une année ordinaire de 12 mois totalise 354,3670 jours. Sa durée est donc inférieure de 10,8751 jours (soit, approximativement, 11 jours) à l’année solaire. Pour combler cet écart, la solution la plus communément attestée a consisté à ajouter un treizième mois (appelé « intercalaire ») à certaines années : pour cela, on redoublait un mois, avec une préférence assez logique pour le dernier (XII). À Babylone, le mois redoublé était soit le mois XII (addaru) soit le mois VI (ululu) 173. Selon les lieux et les époques, le rythme d’intercalation a considérablement varié. D’abord empirique, il s’est finalement calé sur différents cycles. Grâce à ces cycles, il est devenu possible de 173

Stern 2012, p. 95.

LES ÉVOLUTIONS DU CALENDRIER

321

connaître à l’avance quand un mois serait ajouté, de sorte que le calendrier, de flexible et imprévisible, est devenu fixe et prévisible. Le cycle qui a été le plus pratiqué, et qui est encore celui du calendrier juif contemporain, est appelé couramment « cycle de Méton », d’après un astronome athénien. Ce cycle, mis en œuvre en Babylonie à l’époque achéménide et peut-être même un peu plus tôt, intercale un mois lunaire tous les deux ou trois ans, de manière à totaliser sept mois pour chaque période de 19 ans. Au terme d’un cycle de 19 ans, un petit écart subsiste cependant : la lune a gagné 1h 27’ 33” sur le soleil 174. Il en résulte un décalage d’un jour au bout de 312,5 ans. Dans le royaume de Ḥimyar, l’existence d’un mois intercalaire a été mise en évidence par Norbert Nebes en 2004. Le raisonnement se fonde sur l’inscription Sadd Maʾrib 5 = CIH 541 dont il a déjà été question et sur une ébauche de ce texte récemment découverte, Sadd Maʾrib 4 = DAI GDN 2002/20. Dans la première, Sadd Maʾrib 5 = CIH 541, datée de dhumaʿūn (II, ≈ mars) 658, Abraha indique qu’il est rentré chez lui après une absence de « onze mois » 175. Or il s’est mis en campagne au mois de dhu-qiyāẓān (V, ≈ juin) 657 ḥim. et revenu au mois de dhu-maʿūn (II, ≈ mars) 658, soit dix mois plus tard (sans mois intercalaire), mais effectivement onze si on en ajoute un. Le nom et la place de ce mois intercalaire sont connus, grâce à une inscription inachevée du même roi Abraha qui commémore

Au terme d’un cycle de 19 années solaires, le total des jours écoulés est de 6939,56 (19 x 365,24). Le cycle lunaire de 19 années plus sept mois, soit 235 lunes, a la même durée (235 x 29,53 = 6939,55 jours, avec une petite différence résultant des approximations). L’intercalation des mois supplémentaires peut se faire, par exemple, les années 3, 6, 8, 11, 14, 17 et 19. 175 Sadd Maʾrib 541 = CIH 541 / 130-136 : « Il a (131) a[chevé] son ouvrage en cinquante-(132)huit jours et il est rentré (133) après onze mois.(134) Au mois dhu-maʿūn de l’an (135) six cent cinquante-(136)huit », w-k[ml](131)w mqḥ-hmw b-ṯmny[t w-ḫ](132)ms¹y ymtm w-qf[lw] (133) b-ʾḥd ʿs²r ʾw[rḫ](134)m b-wrḫ ḏmʿn ḏ-l (135) ṯmnyt w-ḫms¹y w-s¹(136)ṯ mʾtm. 174

322

CHRISTIAN JULIEN ROBIN

l’adjonction d’un massif élevé, appelé Qishbānum ou le « Pylône » 176, en avant de l’écluse méridionale de la Digue de Maʾrib (Sadd Maʾrib 4 = DAI GDN 2002/20). Cette seconde inscription est datée de dhu-ḥullatān « le dernier » 658 (ll. 36–41). D’après le texte, l’édification de ce massif a pris la totalité du mois dhu-ḥullatān [65]8 ḥim., sans autre précision (ll. 25–28) : « Avec la puissance, le soutien (2) et l’aide de Raḥmānān,(3) seigneur du Ciel,(4) de son Messie,(5) a édifié le roi (6) Abraha, Zb(7)ymn, roi de Sa(8)baʾ, dhu-Raydān,(9) Ḥaḍramawt et (10) Yamnat, et de leurs Ara(11)bes dans le Haut-Pays (12) et sur la Côte, ʿAw(13)[dān] devant l’ouvrage de (14) [Shuriḥbi]ʾil Yaʿfu[r] 177 (15). (Abraha) l’a édifié de(16)puis son sol ro(17)cheux jusqu’à ses parties hautes,(18) (sur une hauteur de) quarante-et-une (19) coudées, et sur une lon(20)gueur jusqu’au grand canal at(21)teignant quarante-cinq (22) coudées, et qua(23)torze coudées en lar(24)geur. Il en a posé les fondations (25) au début du mois de dhu(26)ḥullatān de (l’an [65])8 (27) et il l’a achevé à la fin (28) de celuici, édifiant (29) avec de la pierre et du mortier (30) une construction éle(31)vée. Voici que l’achèvement (32) complet de son inscription (33) et de son intervention (34) à Ra[ḥbum,] (35) sur la montagne et en avant de (36) Maʿqamān (eut lieu) au mois (37) dhuḥullatān , (38) le dernier, en (l’an) (39) six cent (40) cinquantehuit », b-ḫyl w-n(ṣr) (2) w-rdʾ Rḥmnn (3) mrʾ vac. S¹myn (4) w-Ms¹ vac. ḥ-h(w) (5) (h)qḥ vac. mlk(6)n ʾb(r)h Zb(7)ymn mlk S¹(8)[b](ʾ) w-ḏ-Rydn (9) (w)[Ḥ](ḍ)rmwt w-(10)(Y)m(n)t w-ʾʿr(11) [b-](h)mw Ṭwd(12) [m w](T)ht ʿw(13)[dn b-]qdm mqḥ (14) [S²rḥb]ʾl Yʿf(15)[r] w-(h)qḥ-hw b(16)n mwṯr-hw ẓw(17)[r](n) ʿd(y ʾrʾs¹)-(18)hw ʾḥd w-ʾr(19)bʿy ʾmm wṭw(20)lm s¹-ḏhbn (d) w(21)rdm ḫms¹t w-ʾ(22)rbʿy ʾmm w-ʾr(23)bʿt ʿs²r ʾmm r(24)ḥbm w-hwṯr-hw (25) b-qd(m) wrḫn ḏ-(26)ḥltn ḏ-l-ṯmnt (27) w-hs²qr-hw b-ʿ(28)ḏr-hw mbrʾm (29) b-grbm w-gyrm (30) mbrʾm (ḏ-h)q(31)m w-k-kl tl(32)wn ms³nd-h(33)mw w-mqḥ-hm(34)[w] b-(R)[ḥbm w-](35)b-ʿrn w-qdm (36)

Le nom Qishbānum se trouve dans Sadd Maʾrib 5 = CIH 541 / 104-105. 177 Shuriḥbiʾīl Yaʿfur est alors le dernier roi ḥimyarite à avoir fait des travaux considérables sur la Digue de Marib, près d’un siècle plus tôt. Abraha fonde sa légitimité en se présentant comme son continuateur. 176

LES ÉVOLUTIONS DU CALENDRIER Mʿqm[n] wrḫ-(37)hw ḏ-ḥltn w-s¹ṯ mʾ(t)(41)m

(38)

323

ʾḫ(r)[ n] ḏ-l-(ṯ-)(39)mnt w-ḫ[ms¹=](40)y

Dans la compréhension de ce texte, le point crucial est de savoir si les deux expressions le « mois dhu-ḥullatān [65]8 ḥim. » et le « mois dhu-ḥullatān le dernier 658 » désignent le même mois ou deux mois successifs. L’hypothèse de deux mois successifs paraît raisonnablement assurée si on postule que les travaux qui ont lieu un mois déterminé sont commémorés le mois suivant. C’est ce qu’illustre l’inscription Sadd Maʾrib 5 = CIH 541 / 131–132, datée de dhū maʿūn 658. Les travaux de réfection sur la Digue commencent pendant la dernière semaine du mois de dhū-diʾāwān (XII) 178 et durent 58 jours 179. Ils s’étendent donc sur la totalité du mois de dhu-ḥullatān [I] (≈ février) et sur une grande partie du mois dhu-ḥullatān II (intercalaire), puis sont commémorés le mois suivant, en dhu-maʿūn (≈ mars). Si on applique le même principe à Sadd Maʾrib 4 = DAI GDN 2002/20, les travaux commémorés par ce texte, qui ont lieu pendant toute la durée de dhu-ḥullatān I, sont commémorés en dhuḥullatān II 180. En résumé, il semble assuré que le royaume de Ḥimyar, aux 5e e et 6 s., intercalait occasionnellement un mois supplémentaire, en redoublant le premier mois de l’année. La pratique de l’intercalation avait pour finalité de corriger la différence de durée entre l’année de 12 mois lunaires et l’année solaire. Elle implique donc que les mois ḥimyarites antiques (à la différence de ceux des listes médiévales) étaient lunaires. Dans l’Antiquité tardive, le calendrier solaire romain dominait dans l’ensemble du monde méditerranéen de manière d’autant plus Ligne 96, b-mdt ḏ-dʾwn ʾḫrtn. On peut observer incidemment que cette durée correspond à deux mois lunaires de 29 jours. Mais cela c’est probablement sans importance parce que, dans les calendriers lunaires, un mois de 30 jours suit d’ordinaire un mois de 29 jours. 180 Dans un article rédigé il y a quatre ans, j’avais proposé une autre interprétation de ḏ-ḥltn ʾḫ(r)[n], « le dernier jour du mois dhu-ḥullatān, à laquelle j’ai renoncé (Robin 2012 a, pp. 134–148). 178 179

324

CHRISTIAN JULIEN ROBIN

générale qu’il avait été adopté par tous les chrétiens. En Perse, c’était un autre calendrier solaire qui était utilisé. Le calendrier lunaire, prépondérant au Proche et au Moyen-Orient aux époques néo-babylonienne, achéménide et hellénistique, était devenu résiduel. Il n’était plus conservé que par des populations soit en marge (comme les Arabiques 181) soit entrées en dissidence (comme les juifs ou la secte des Ḥarrāniens 182). Il est donc notable que Ḥimyar utilise encore un calendrier lunaire au 6e s. Dans ce calendrier lunaire, c’est le premier mois de l’année qui était redoublé quand on voulait introduire un intercalaire. Un tel choix était inhabituel. Ailleurs, on préférait redoubler le dernier mois. C’était notamment le parti retenu par le calendrier juif qui dupliquait adār. Le redoublement du premier mois se retrouve peutêtre dans le calendrier de Makka, à moins que ce ne soit le deuxième 183. La mise en évidence d’un calendrier lunaire avec des mois intercalaires dans le royaume de Ḥimyar ne doit pas surprendre puisque, comme nous l’avons vu, plusieurs mois qui sont attestés avec l’adjectif « premier » ou « dernier » semblent avoir été occasionnellement redoublés. Les reconstructions qui précèdent reposent sur l’hypothèse que le calendrier des années 547–548 était identique au calendrier ḥimyarite des siècles précédents. Une autre hypothèse est évidemment envisageable, celle d’une réforme décidée par Abraha à cette date, comportant l’ajout exceptionnel d’un mois intercalaire et l’adoption d’une année de 365 jours (ou de 366 une fois tous les quatre ans), impliquant une redéfinition de la durée des mois afin de les aligner sur ceux d’un autre calendrier, évidemment le calendrier solaire julien. Nous ne la retenons pas parce que, dans l’inscription Sadd Marib 5 = CIH 541, qui présente un bilan apparemment complet des réalisations du roi pendant cette période, une On ne dispose de données concrètes que pour ceux de Ḥimyar et ceux de Makka (concernant ces derniers, voir Robin 2016 a [à paraître]). 182 al-Bīrūnī, al-Āthār al-bāqiya, texte pp. 11, 52–59 et 318–319, traduction pp. 13, 62–69 et 315, pour le 10e s. 183 Robin 2016 a à paraître. 181

LES ÉVOLUTIONS DU CALENDRIER

325

telle réforme n’est pas mentionnée. Nous supposons que la réforme du calendrier intervient plus tard 184. Mais il n’est pas sûr qu’un tel argument soit décisif parce que les textes épigraphiques, ne traitent que de sujets « nobles » qui valorisent leurs auteurs. On observera cependant que, en 547–548, Abraha qui était en campagne dans le Yémen oriental, loin de sa capitale, avait des problèmes à régler bien plus brûlants que la réforme du calendrier. 4. Les équivalents juliens des mois ḥimyarites La question se pose aussi de savoir si les équivalences que les listes yéménites médiévales (13e et 14e s.) établissent entre les mois syriens et les mois ḥimyarites sont exactes. On dispose de deux moyens pour vérifier la précision de ces équivalences. Le premier, qui part du constat que sept des douze mois portent le nom d’une période de l’année agricole, consiste à vérifier si ces mois sont bien positionnés. La référence est évidemment le régime des précipitations au Yémen qui est très spécifique et très régulier. La presque totalité des pluies se concentrent sur deux périodes : une principale en juillet-août, permettant une première récolte, et une seconde, moins abondante, en mars-avril-mai, autorisant d’ordinaire une seconde récolte. À Ṣanʿāʾ, les précipitations atteignent en moyenne 40 mm en juillet et 70 en août ; 20 mm en mars, 50 en avril et 20 en mai. Pendant les autres mois de l’année, la pluviométrie est inférieure à 10 mm. La distribution des précipitations à Ṣanʿāʾ présente un intérêt tout particulier parce que la ville, qui se trouve à la limite du bassin versant qui alimente la Digue de Marib, renseigne sur les périodes où la Digue reçoit de l’eau 185. Les sept mois portant le nom d’une période agricole sont dhumaʿūn (II), dhu-mabkarān (IV), dhu-qiyāẓān (V), dhu-madhraʾān (VI), dhu-khirāfān (VII), dhu-ʿallān (VIII) et dhu-ṣurāb (IX). Pour les autres, dhu-ḥullatān (I), dhu-thābatān (III), dhu-muhlatān (X), dhu-ālān (XI) et dhu-diʾāwān (XII), aucune étymologie convaincante n’a été relevée. 184 185

Voir ci-dessous, pp. 48–50. L’argument est utilisé ci-dessous, pp. 39–40.

326

CHRISTIAN JULIEN ROBIN

Il est remarquable que les noms en relation avec des activités agricoles forment une série allant de mars à octobre. On peut supposer que les autres se référaient à d’autres activités. En particulier, nous avons supposé que le mois XI s’était appelé tout d’abord ṣayd (ce qui veut dire « chasse » 186), avant d’être renommé dhu-ālān. Incidemment, il n’est pas inutile de savoir que les saisons, mentionnées dans quelques inscriptions sabéennes d’époque ḥimyarite, s’appellent : dṯʾ, « printemps » ; ḫrf, « été » ; s¹ʿs¹ʿ, « automne » ; et mly, « hiver » 187. Les deux périodes de pluies sont brq dṯʾn 188 et brq ḫrfn 189. La nomenclature des périodes agricoles se trouve principalement dans les textes yéménites médiévaux déjà évoqués, mais seul l’almanach de la Tabṣira de ʿUmar est édité. On y relève aussi de nombreuses informations sur les pluies 190, notamment leurs noms qui peuvent être intéressants pour notre propos comme kharf (la brume qui enveloppe les pentes du Ẓafār omanais au moment de la Le mot signifie également « pêche », mais cette activité est inconnue dans le Yémen intérieur. 187 Voir Ja 650 / 12–14 « des récoltes de produits agricoles de printemps, d’été, d’automne et d’hi(13)ver abondantes dans tous leurs champs, leurs val(14)lées, leurs champs d’été, leurs domaines et leurs jardins » (frʿ ʾmrt dṯʾ w-ḫrf w-s¹(13)ʿs¹ʿm w-mlym mhrbḍm ʿdy kl ʾrḍ-hmw w-ʾs¹(14)rr-hmw w-mqyẓhmw w-ms²mt-hmw w-tbqlt-hmw), ou Ja 661 / 7–8, « des récoltes de produits agricoles de printemps, d’été, d’automne et d’hiver (8) et des fruits de bonne qualité dans toutes leurs terres et leurs vallées » (w-frʿ ʾmyrt dṯʾ w-ḫrf w-s¹ʿs¹ʿm w-mlym (8) w-ʾṯmr ṣdqm hnʾm bn kl ʾrḍt-hmw w-ʾs¹rr-hmw). 188 Voir Fa 71 / 4-5, « en reconnaissance parce qu’[Il a donné] (5) lors de la saison des pluies du printemps une eau abondante à Marib et dans ses deux vallées jusqu’à saturation » (ḥmdm b-ḏt [s¹qy] (5) b-brq dṯʾn s¹qym mhs²fqm (6) b-Mrb w-s¹ry-hw mhḏrm). 189 Voir Ja 735 / 3-4, « … qu’il accorderait et (4) rendrait pluvieuse la saison des pluies de l’été, au mois de dhu-abhī de l’année de Tubbaʿkarib fils de Wadadʾīl b. Kabīr Kha(5)līl la neuvième, », k-yḫmrn w-(4)s¹qy brq ḫrf b-wrḫ ḏ-ʾbhy ḏ-ḫrf Tbʿkrb bn Wddʾl bn Kbr H̲(5)ll ts¹ʿn. 190 Varisco 1994, p. 106, par exemple, qui situe le début de la seconde saison des pluies début février. 186

LES ÉVOLUTIONS DU CALENDRIER

327

mousson), kharīf, qayẓ 191 ou thawr (pluie du 19 ayyār / mai) 192. Des indications parfois contradictoires rappellent cependant que les données de la Tabṣira ne sont que des approximations. Un passage indique, par exemple, que les précipitations sont particulièrement abondantes entre le 25 tammūz (juillet) et le 30 āb (août) 193, tandis que l’almanach situe leur fin les 29–30 tammūz : « fin des pluies abondantes dans les zone de montagne depuis son premier (?) … » 194. On peut se fonder également sur le lexique relevé dans des sources yéménites très diverses, rassemblé dans quelques dictionnaires spécialisés 195. Voici comment sont définies les sept périodes du calendrier agricole qui ont donné leur nom à un mois ḥimyarite : — ṣurāb (≈ octobre) : c’est le nom de la moisson d’automne qui fait suite aux pluies de juillet-août, en particulier pour l’orge, le raisin (zabīb), les figues, les pois (ʿatar) et le fenugrec (Varisco 1994, pp. 65–66 ; Piamenta 1991, p. 280 ; AlSelwi 1987, pp. 129–130). — maʿūn (≈ mars) : ce terme désigne le début de l’été (Piamenta 1991, p. 469). — mabkar (≈ mai) : le nom est encore utilisé dans diverses régions du Yémen comme synonyme de ayyār, mai (Varisco 1994, p. 69) ; voir aussi bekr el-ṣayf, « début de l’été (saison des pluies, al-Ḥujariyya) » et bikār, « temps de la récolte jusqu’à la mi-août » (Piamenta 1991, p. 37). — qiyāẓ ou qayẓ (≈ juin) : c’est la « fin de l’automne (Ṣaʿda, Ṣanʿāʾ) », le « temps de la moisson jusqu’à la fin avril », etc. (Piamenta 1991, p. 421–422) ; la variante qayẓ renvoie à l’été (Varisco 1994, p. 69). — madhraʾ (≈ juillet) : c’est l’époque des semailles pour le blé et l’orge dans les montagnes du Yémen (Varisco 1994,

Varisco 1994, pp. 108–109. Varisco 1994, pp. 33, 53 (awwal maṭar al-thawr) et 110–111. 193 Varisco 1994, pp. 106–107. 194 nihāyat ghazr al-amṭār fī ʾl-aʿmāl al-jabaliyya min awwali-hi … (Varisco 1994, pp. 36 et 56). 195 Voir notamment Landberg 1920, Al-Selwi 1987 ou Piamenta 1991. 191 192

328

CHRISTIAN JULIEN ROBIN pp. 69–70) ; voir aussi midhraʾ / midhrà, pl. madhārī, « période des semailles », et madhrā(ʾ), madhrāt, « action de semer » (Piamenta 1991, p. 167). — khirāf (≈ août) : le terme est expliqué dans l’almanach par le fait que c’est l’époque de la récolte (ikhtirāf) (Varisco 1994, p. 70) ; il doit être distingué de kharīf, « pluies d’été » (Piamenta 1991, p. 125 ; Al-Selwi 1987, p. 76), même si ces pluies sont précisément celles de juillet-août ; on peut supposer que khirāf est la vocalisation de l’un des termes sabaʾiques écrits ḫrf comme, par exemple, celui qui nomme l’été 196. — ʿallān (≈ septembre) : c’est le début de l’automne, le début de la récolte du sorgho (Varisco 1994, p. 70) ; le mot signifie aussi « pluie d’automne », « saison des pluies ; première pluie, première partie de la saison des pluies ; printemps ; été ; automne, entre la mi-août et la mi-décembre (Jawf) etc. » (Varisco 1994, p. 70 ; Piamenta 1991, p. 339).

Il est manifeste que les mois ḥimyarites portant des noms de périodes agricoles occupent, dans l’année solaire, une place qui s’accorde approximativement avec ces périodes. On peut en déduire que les mois ḥimyarites des 13e et 14e s. n’ont pas dérivé de manière substantielle par rapport à leur place initiale. Pour vérifier la validité des équivalences entre mois ḥimyarites et mois solaires syriens, on peut également se fonder sur les dates des grands travaux de réfection sur la Digue de Maʾrib aux 5e et 6e s. è. chr., qui, de façon évidente, doivent se situer en dehors des deux périodes de pluies, celle de juillet-août et de mars-avril-mai. En 454–456 è. chr. (564–565 ḥim.), le roi Shuriḥbiʾīl Yaʿfur achève une première réfection en dhu-diʾāwān (≈ janvier) et une seconde en dhu-diʾāwān (≈ janvier). Les travaux qu’il a entrepris sont donc terminés un ou deux mois avant les pluies de printemps. En 547–548, sous le règne d’Abraha, les travaux de réfection qui avaient commencé en dhu-ṣurābān (≈ octobre), sont suspendus à cause d’une épidémie, puis reprennent pendant la dernière semaine de dhu-diʾāwān (≈ janvier). Une première phase, qui est commémorée dans Sadd Maʾrib 4 = DAI GDN 2002/20 (daté dhu-ḥullatān II), 196

Voir ci-dessus, p. 37.

LES ÉVOLUTIONS DU CALENDRIER

329

dure du début à la fin de dhu-ḥullatān [I]. On peut supposer que l’absence de pluies a alors incité Abraha à poursuivre les travaux pendant un mois de plus, de sorte que, dans Sadd Maʾrib 5 = CIH 541 (dhu-maʿūn 658 ḥim.), il est indiqué que les travaux ont duré finalement 58 jours, correspondant à la dernière semaine de dhu-diʾāwān (≈ janvier), à dhu-ḥullatān [I] en entier et à la majeure partie de dhu-ḥullatān II. Ils sont commémorés le mois suivant, en dhu-maʿūn (≈ mars). Toujours sous le règne d’Abraha, des contingents communaux commémorent leur participation à des travaux d’entretien sur la Digue en dhu-muhlatān (≈ novembre) 668 ḥim. 197. Ils ont donc été mobilisés après la saison des pluies de juillet-août, sans doute pour une brève période. Quant aux ruptures de la Digue, on peut supposer qu’elles sont la conséquences de pluies exceptionnelles. Dans l’inscription de Shuriḥbiʾīl, les deux ruptures de la Digue sont provoquées par l’inondation de l’été, puis par celle du printemps suivant, en dhuthābatān (≈ avril). À l’époque d’Abraha, la Digue se rompt en dhumadhraʾān (≈ juillet). Dans les deux cas, c’est bien pendant la saison des pluies. Ces correspondances, satisfaisantes en règle générale, prouvent que les mois n’ont pas dérivé de manière significative entre l’Antiquité et le 13e s. Mais une dérive de quelques semaines n’est pas impossible comme nous le verrons 198.

C. UNE RÉFORME DU CALENDRIER LORS DE LA CONVERSION AU JUDÉO-MONOTHÉISME (VERS 380)?

Vers 380, sous le règne du roi Malkīkarib Yuhaʾmin en corégence avec ses fils Abīkarib Asʿad et Dharaʾʾamar Ayman, le royaume de Ḥimyar fait d’un monothéisme inspiré du judaïsme (que nous avons proposé de dénommer « judéo-monothéisme ») la religion officielle, après que nombre de grandes familles se sont converties

197 198

Sadd Maʾrib 6 = Ja 547+546+544+545. Voir ci-dessous, pp. 45–47.

330

CHRISTIAN JULIEN ROBIN

au judaïsme. On peut se demander si cette conversion 199 a eu des conséquences sur le calendrier ou, en d’autres termes, si Ḥimyar a réformé son calendrier pour l’aligner sur l’un des calendriers du judaïsme 200. Les inscriptions n’apportent pas de réponse assurée. On observe que, dans la première moitié du 5e s., l’ancien calendrier du royaume de Sabaʾ disparaît de l’usage. Mais cette disparition peut s’interpréter de diverses manières. Si elle a pu être hâtée par une réforme du calendrier ḥimyarite, elle a pu l’être tout autant par les progrès réalisés dans l’unification du royaume et par l’effacement progressif de Sabaʾ. 1. Quelques innovations peu significatives Pour étayer l’hypothèse que le calendrier ḥimyarite a été réformé au moment de l’adoption du judéo-monothéisme, il faudrait mettre en évidence des innovations significatives par rapport aux périodes antérieures. Un premier changement est le remplacement du nom de mois ṣayd (« chasse »), par un autre, dhu-ālān 201. Comme la chasse a toujours eu, au Yémen, une dimension religieuse 202, il n’est pas exclu que la suppression de toute référence à la chasse dans le calendrier soit la conséquence d’un changement de religion et donc de la conversion au judéo-monothéisme. La signification de ālān qui remplace ṣayd est malheureusement inconnue 203. Un second nom de mois, dṯʾ ou ḏ-dṯʾn, disparaît également, mais comme il s’agit d’une On peut parler de « conversion » parce qu’il y a eu une rupture avec le passé, comme le prouvent le choix d’une divinité inconnue précédemment, la construction de nouveaux lieux de culte et un changement radical de la terminologie. Les familles princières ont été d’autant plus nombreuses à se convertir que le rejet de la religion polythéiste permettait de fermer les anciens temples et de s’emparer de leurs trésors. 200 Sur la question des calendriers juifs de l’Antiquité, voir Stern 2012, pp. 331-353. 201 Voir ci-dessus, p. 11. 202 Serjeant 1976. 203 Varisco 1994, pp. 66–67. 199

LES ÉVOLUTIONS DU CALENDRIER

331

appellation qui n’est attestée que dans quelques graffites, cette disparition signifie sans doute le recul des particularismes régionaux. Une deuxième innovation est un changement dans l’intercalation. Avant 380, deux mois peuvent apparemment être redoublés, dhu-ṣurābān et dhu-qayāẓān, parce qu’ils sont attestés soit sans épithète soit avec l’épithète « le premier » 204. On peut en effet supposer que ces mois existaient aussi avec l’épithète « le dernier » parce que, dans les calendriers de Qatabān et de Sabaʾ, on relève des mois appelés ḏ-Brm, ḏ-Brm « le premier » (qdmn) et ḏ-Brm « le dernier » (ʾḫrn) ou encore ḏ-Ns¹wr, ḏ-Ns¹wr « le premier » (qdmn) et ḏNs¹wr « le dernier » (ʾḫrn) 205. Avant 380, le mois intercalaire ḥimyarite pouvait donc être soit le Ve (dhu-qayāẓān) soit le IXe (dhuṣurābān), avec des places dans le calendrier qui donnent à penser que n’importe quel mois pouvait être redoublé. C’est sans doute un indice que l’intercalation était encore empirique et approximative. Après 380, le mois redoublé est dhu-ḥullatān, le premier de l’année ḥimyarite. Désormais, il est plausible que les règles d’intercalation sont devenues rigides, avec un rythme d’intercalation défini par un cycle et donc connu à l’avance. Un troisième changement est la disparition de toute référence au personnage qui a donné son nom à l’ère ḥimyarite, Mabḥūḍ ibn Abḥaḍ, dont la dernière mention date de 360 è. chr. (470 ḥim.). Or il n’est guère douteux que la charge du calendrier avait une dimension religieuse. Cette disparition pourrait donc résulter d’une réforme calendaire et religieuse ; mais elle peut s’expliquer tout aussi bien par la disparition des ères rivales de l’ère ḥimyarite 206 qui rendait inutile la mention de ce personnage. UAM 327 / 2 (ṣrb qdmn) ; Kh-Jarf al-Naʿīmiyya 18 / 9 (qyẓ qdmn, 308 ḥim.) ; Kh-Umayma 5 / 7 (qyẓ qdmn, 327 ḥim.). 205 Voir ci-dessous, pp. 50–51. 206 L’ère de Radmān, dont la dernière attestation est l’année 198 (alMiʿsāl 5, 272–273 è. chr.), est déjà sortie de l’usage à l’extrême fin du 3e s. puisqu’une inscription de Waʿlān, la capitale de Radmān, utilise l’ère ḥimyarite (voir YMN 13 = Miʿsāl 18, daté de ḏ-mʿn 409 ḥim. soit mars 299 è. chr.). La dernière attestation de l’ère de Maḍḥàm est 388 (355 ± 10 è. chr.) dans Buraʿ al-Aʿlà 1. 204

332

CHRISTIAN JULIEN ROBIN

Concernant d’autres modifications qui auraient pu intervenir dans la répartition des mois de 29 et 30 jours, dans la définition du début du mois et de l’année etc., on ne sait rien. Les inscriptions ne contredisent pas l’hypothèse d’une réforme du calendrier de Ḥimyar après la conversion au judéomonothéisme, mais ne révèlent aucun changement vraiment significatif. 2. La Tradition arabo-islamique L’hypothèse que le calendrier de Ḥimyar pourrait avoir été réformé en prenant modèle sur le calendrier juif s’accorde cependant avec une remarque de deux savants d’époque islamique qui rapportent (ou supposent) que le calendrier des Arabes préislamiques a été réformé de cette manière. Le premier est l’astronome persan Abū Maʿshar al-Balkhī (m. 886, 272 h.), dans Kitāb al-ulūf fī buyūt alʿibādāt. On connaît la teneur de cet ouvrage, qui est perdu, grâce à un texte encore inédit d’un savant du 12e s., heureusement cité par Mahmoud Effendi 207 : « Quant aux Arabes de la Jāhiliyya, ils se servaient des années lunaires en observant l’apparition des croissants comme le font les musulmans. Ils faisaient le pèlerinage le 10 de dhū ʾl-ḥijja. Cette date, cependant ne tombait pas toujours dans la même saison de l’année, mais se déplaçait. Une fois, elle tombait en été et, une autre, en hiver, ou encore dans les deux autres saisons, du fait des divergences entre années solaires et lunaires. Ils voulurent que l’époque du pèlerinage tombât au moment où ils faisaient leur commerce, et que l’air fut tempéré, ni chaud ni froid, à l’époque où les arbres sont en feuille et où on trouve du fourrage, afin que leur voyage à Makka soit aisé, et afin qu’ils y fissent leur commerce tout en s’acquittant de leur acte de dévotion. Ils apprirent alors des juifs la pratique de

Il s’agit de Muntahà ʾl-idrāk fī taqāsīm al-aflāk de Bahāʾ al-Dīn Abū Muḥammad ʿAbd al-Jabbār al-Kharaqī al-Thābitī (1132 [527/ h.] ou 1138 [533/ h.) (ancien fonds Bibl. impériale 1115 = BNF Arabe 2499). Le passage qui concerne le mois intercalaire est heureusement reproduit (pp. 168–172) et traduit (pp. 172–178) par Mahmoud Effendi (1858). 207

LES ÉVOLUTIONS DU CALENDRIER

333

l’intercalation (l’embolisme, al-kabīsa) et la nommèrent al-nasīʾ, c’est-à-dire le “retardement”, sauf qu’ils se distinguèrent des juifs dans certaines pratiques : les juifs intercalaient, dans 19 années lunaires, sept mois lunaires, de sorte qu’elles deviennent 19 solaires alors que les Arabes intercalent, dans 24 années lunaires, 12 mois lunaires » 208.

al-Bīrūnī reproduit cette tradition, en la corrigeant et en la précisant. Il reprend l’idée que les Arabes ont bien imité les juifs, mais suppose qu’ils ont pratiqué l’intercalation non pas de manière rigide (une intercalation tous les deux ans), mais empirique. Quant à l’intercalation que les juifs pratiquaient ce n’était pas 7 mois tous les 19 ans, mais 9 mois tous les 24 ans. Par ailleurs, il complète Abū Maʿshar en se hasardant à préciser l’époque de l’emprunt : « Il (Ḥudhayfa, le premier qalammas) avait emprunté ce procédé aux juifs, environ 200 ans avant l’apparition de l’Islam, sauf que ceux-ci intercalaient 9 mois dans 24 années lunaires » 209.

Plus loin, il adopte une formulation légèrement différente :

Mahmoud Effendi 1858, pp. 168–169 (texte) et 172–173 (traduction) : wa-ammā al-ʿArab fī ʾl-Jāhiliyya fa-kānū yastaʿmilūna sinī ʾl-qamar biruʾyat al-ahilla ka-mā tafʿalu-hu ahl al-islām wa-kānū yaḥujjūna fī ʾl-ʿāshir min dhī ʾl-ḥijja wa-kāna lā yaqaʿu hādhā al-waqt fī faṣl wāḥid min fuṣūl al-sana bal yakhtalifu fa-marratan yaqaʿu fī zamān al-ṣayf wa-marratan fī zamān al-shitā wa-marratan fī ʾl-faṣlayn al-bāqiyayn li-mā yaqaʿu bayna sinī ʾl-shams wa-ʾl-qamar min al-tafāṣil. Fa-arādū an yakūna waqt ḥajjati-him muwāfiqan li-awqāt tijārati-him wa-an yakūna al-hawāʾ muʿtadilan fī ʾl-ḥarr wa-ʾl-bard maʿa tawrīq al-ashjār wa-nabāt al-kalā litashula ʿalay-him al-musāfara ilà Makka wa-yattajirū bi-hā maʿa qaḍā manāsikihim fa-taʿallamū ʿamal al-kabīsa min al-yahūd wa-sammaw-hu al-nasīʾ ayy altaʾkhīr illā anna-hum khālafū al-yahūd fī baʿḍ aʿmāli-him li-anna al-yahūd kānū yakbisūna tisʿa ashara sana qamariyya bi-sabʿat ashhur qamariyya ḥattà taṣīra tisʿa ashara shamsiyya wa-ʾl-ʿArab takbisu arbaʿā (sic) wa-ʿishrīn sana qamariyya biithnay ʿashar shahran qamariyya. 209 al-Bīrūnī, al-Āthār al-bāqiya, texte, p. 12 ; traduction p. 14 (voir la citation complète de ce texte ci-dessous, pp. 57–58). 208

334

CHRISTIAN JULIEN ROBIN « (Les Arabes) apprirent des juifs qui étaient leurs voisins l’intercalation (al-kabs), quelque deux cents ans avant l’hégire. Ils se mirent à s’en servir à la manière des juifs » 210.

Comme cela a déjà été noté, en particulier par François de Blois (« Taʾrīkh »), il est vraisemblable que les affirmations d’al-Bīrūnī se fondent sur son interprétation personnelle des événements et non sur des données transmise par les traditions. Abū Maʿshar al-Balkhī et al-Bīrūnī supposent donc qu’un calendrier juif 211 a servi de modèle quand, dans des temps très anciens, le calendrier des Arabes, incontestablement celui de Makka puisque Abū Maʿshar al-Balkhī traite du pèlerinage à la Kaʿba et qu’al-Bīrūnī mentionne le qalammas, a été réformé. al-Bīrūnī situe cette réforme environ 200 ans avant l’hégire (soit vers 420). Ce nombre s’accorde approximativement avec les sept générations de qalammas mentionnées par al-Bīrūnī entre l’instauration de l’intercalation et les débuts de l’Islam 212. On sait aujourd’hui que, vers 420, c’est le roi Abīkarib Asʿad (c. 375–c. 445, arabe Abū Karib Asʿad le Parfait) qui régnait sur Ḥimyar. Ce roi, selon la Tradition arabo-musulmane, avait introduit le judaïsme au Yémen et revêtu la Kaʿba avec une kiswa pour la première fois. Mais il est fort douteux qu’al-Bīrūnī ait fait la relation parce qu’il ne disposait que d’informations assez confuses sur la chronologie préislamique 213. al-Bīrūnī, al-Āthār al-bāqiya, texte, p. 62 ; traduction p. 73. Concernant les nombreuses variétés du calendrier juif aux époques romaine et byzantine, voir Stern 2001. al-Bīrūnī rapporte que, à son époque, les juifs ne s’accordaient pas sur le rythme des intercalations, mais avaient élaboré trois cycles différents : 2, 5, 7, 10, 13, 16, 18 ; 1, 4, 6, 9, 12, 15, 17 ; 3, 5, 8, 11, 14, 16, 19 (al-Āthār al-bāqiya, texte, p. 55, et traduction, pp. 64–65) 212 al-Bīrūnī, al-Āthār al-bāqiya, texte, p. 12 ; traduction pp. 13–14. 213 Sur l’époque d’Abū Karib, les traditionnistes, Les faisaient l’objet d’avis étonnamment divergents. Abū Karib Asʿad aurait vécu de une à sept générations avant dhū Nuwās (= Yūsuf Asʾar Yathʾar, 522–525 ou l’une des années suivantes) ou même 700 ans avant Muḥammad (Robin 2003, pp. 136–137). Il n’est pas sûr qu’al-Bīrūnī ait eu des idées très pré210 211

LES ÉVOLUTIONS DU CALENDRIER

335

Quant à savoir quel est le calendrier juif qui, selon Abū Maʿshar al-Balkhī et al-Bīrūnī, aurait pu servir de modèle, il est d’autant plus difficile de répondre qu’ils en donnent deux définitions différentes. Les 7 mois intercalaires dans un cycle de 19 ans mentionnés par Abū Maʿshar al-Balkhī font penser aux calendriers juifs, minutieusement décrit par al-Bīrūnī 214, notamment le calendrier rabbinique qui aurait été progressivement fixé entre le 3e et le 5e s. 215. Quant au cycle de 24 ans avec 9 mois intercalaires, mentionné par al-Bīrūnī, il est sans doute identique au cycle de 8 ans avec 3 intercalaires mentionné dans la liste des cinq cycles suivis par les calendriers juifs 216 ; mais l’identité de ceux qui l’utilisent n’est pas précisée. L’estimation de 200 années que donne al-Bīrūnī est reprise par des auteurs postérieurs comme le signale A. P. Caussin de Perceval 217. Elle se trouverait aussi, selon ce dernier, chez un auteur quelque peu antérieur, al-Masʿūdī (m. 956 [345 h.]), qui situerait alors la création du calendrier de Makka, mais je n’ai pas retrouvé cette indication ni dans Murūj al-dhahab ni dans al-Tanbīh.

3. dhu-ḥullatān et nîsān

Aucun des arguments précédents n’est absolument déterminant. Il en reste un dernier à examiner, celui du premier mois de l’année. Dans les calendriers juifs, l’année a en quelque sorte deux commencements. L’un est le Jour de l’an (Roʾsh hash-shana), qui est le premier jour du mois de tishrî, correspondant idéalement à l’équinoxe d’automne. L’autre est le 1er du mois de nîsān (celui de la fête de Pesaḥ), mois effectivement appelé « le premier » dans la cises sur la chronologie des rois de Ḥimyar, dont il n’en traite pas, sinon par allusions (al-Āthār al-bāqiya, texte pp. 40–41, traduction pp. 49–50). 214 al-Bīrūnī, al-Āthār al-bāqiya, texte, p. 52–59 ; traduction pp. 62–69. 215 Stern 2012, p. 334 et n. 115. Selon Sacha Stern, la tradition médiévale qui attribue une redéfinition d’ensemble du calendrier à Hillel II en 358–359 (E.J.W., « Calendar », dans Encyclopaedia Judaica, Vol. 5, Jerusalem, Encyclopaedia Judaica & The Macmillan Company, 1971, col. 43–50, col. 48), n’est pas confirmée par les sources talmudiques. 216 al-Bīrūnī, al-Āthār al-bāqiya, texte, p. 53–54 ; traduction p. 63. 217 Caussin 1843, p. 347.

336

CHRISTIAN JULIEN ROBIN

Bible et de nombreuses sources, dont le commencement correspond idéalement à l’équinoxe de printemps 218. Dans le calendrier ḥimyarite, le premier mois est dhu-ḥullatān qui, dans les listes médiévales, correspond à shubāṭ = février 219. Comme dhu-ḥullatān est assez proche de l’équinoxe de printemps, on peut donc se demander si, initialement, il n’était pas le correspondant ḥimyarite exact du mois juif de nîsān. Il suffit pour cela que l’ajout de un ou deux mois intercalaires aient été omis. Il faut en effet observer que, à l’époque médiévale, le mois solaire syrien appelé nīsān est un peu plus tardif que le mois lunaire juif nîsān. Le nīsān syrien est un autre nom du mois julien « avril » tandis que le 1er nîsān juif rabbinique se déplace dans une fourchette dont le centre se situe un peu plus tôt. Les bornes de cette fourchette sont, de nos jours, le 12 mars et le 13 avril 220 ; dans l’Antiquité tardive, elles étaient sans doute antérieures de quatre ou cinq jours 221. Pour que le mois ḥimyarite antique dhu-ḥullatān ait initialement coïncidé avec le nîsān lunaire du calendrier juif, il faut donc supposer que le calendrier ḥimyarite a dérivé de un ou deux mois du fait de l’omission d’intercalaires, entre l’époque où il a été défini et celle où les équivalences syriennes ont été décidées. Il est malaisé de mettre une telle dérive en évidence. Dans l’almanach d’al-Malik alAshraf, certaines dates s’accordent parfaitement avec les équivalences des listes médiévales : par exemple, la récolte d’automne Stern 2012, pp. 237–238, qui cite une source rabbinique. Il ne faut pas perdre de vue que la date retenue pour l’équinoxe de printemps n’est pas nécessairement très précise : al-Masʿūdī, par exemple, situe cet équinoxe le 15 mars : « le 15 mars, la nuit et le jour ont la même durée », wa-li-khams ʿashara min ādhār yastawī al-layl wa-ʾl-nahār (Murūj al-dhahab, § 1296, texte II, p. 341, traduction II p. 494). 220 Dans le dernier cycle de 19 ans, le 1er nîsān tombait le 28 mars en 1998 (année 1), le 18 mars (année 2), le 6 avril (année 3, après ajout d’un intercalaire), etc. 221 On ne dispose pas de données précises sur le calendrier juif rabbinique avant le 8e s. Mais on sait que les calendriers fondés sur le cycle métonien connaissent une dérive de 1h 27’ 33” tous les 19 ans, ce qui représente plus de 4 jours au bout d’une quinzaine de siècles. 218 219

LES ÉVOLUTIONS DU CALENDRIER

337

(ṣurāb) du sorgho, la céréale la plus commune au Yémen, commence le 17 tishrīn I (octobre) selon al-Malik al-Ashraf ; or l’équivalent ḥimyarite de tishrīn I est précisément dhū ʾl-ṣurāb 222. D’un autre côté, certaines données des listes médiévales pourraient présenter un certain décalage dans le sens requis. La signification de maʿūn, le « début de l’été », ne s’accorde pas avec (≈ mars) qui est trop tôt. Quant à la récolte de khirāf, elle est nécessairement postérieure de deux ou trois mois à la fin des pluies d’été, alors que le mois serait en (≈ août). Dans certains cas au moins, les équivalences données par les listes médiévales semblent antérieures de quelques semaines par rapport aux saisons et aux travaux agricoles évoqués. Mais la valeur de ces arguments est relative. La variabilité des pratiques agricoles est assez grande selon la plante, l’altitude et le régime des pluies, sans même mentionner que le même nom peut désigner des périodes différentes selon les régions. Par exemple, on pourrait penser que les semailles (sens du substantif madhraʾ qu’on retrouve dans le mois dhu-madhraʾān / dhū madhrān ≈ juillet) ont logiquement lieu vers la fin des pluies d’été, dans la seconde moitié du mois août. Mais, dans l’almanach, il est précisé que les semailles du sorgho se font entre le 16 ādhār (mars) et le 28 ayyār (mai), mais surtout à la fin d’avril et au début de mai : c’est donc que le sorgho se plante après les pluies de printemps et non après celles de l’été 223. S’il est malaisé de prouver que les mois ḥimyarites ont dérivé, on peut affirmer, en revanche que l’hypothèse d’une telle dérive entre le milieu du 5e et le milieu du 6e s. s’accorde assez bien avec les dates des interventions sur la Digue de Marib. Au 5e s., les réfections de Shuriḥbiʾīl Yaʿfur sont achevées deux fois en dhu-diʾāwān (≈ janvier), à savoir plus tôt que nécessaire, puisque la saison des pluies commence en mars. Au 6e s., en revanche, la première tranche des travaux d’Abraha est achevée fin dhu-ḥullatān [I] (≈ février) et célébrée en dhu-ḥullatān II (≈ février bis) ; elle est complétée par une seconde tranche en dhu-ḥullatān II (≈ février bis), commémoré en dhu-maʿūn (≈ mars). Toujours au 6e s., une campagne 222 223

Varisco 1994, p. 24, 42 et 166. Varisco 1994, pp. 30, 34, 49, 53 et 165–166.

338

CHRISTIAN JULIEN ROBIN

d’entretien est commémorée en dhu-muhlatān (≈ novembre). Les dates du 6e s. semblent donc parfaitement en phase avec le calendrier des pluies. Une dérive du calendrier ḥimyarite entre le 5e et le 6e s., par omission de un ou deux mois intercalaires, est d’autant moins improbable que le royaume de Ḥimyar a connu une période particulièrement chaotique à partir du début du 6e s., notamment du fait de la dévastatrice conquête aksūmite. Un dernier argument peut être invoqué en faveur d’une correspondance initiale de dhu-ḥullatān avec le mois juif de nîsān. À Makka, à l’époque de Muḥammad, le début de l’année coïncidait approximativement avec l’équinoxe de printemps 224. Or il est vraisemblable que le calendrier de Makka était une variante du calendrier des Arabes préislamiques (dont le premier et le dernier mois étaient al-muʾtamir et burak), et que ce dernier calendrier avait été lui-même façonné sur le modèle du calendrier ḥimyarite 225. C’est d’ailleurs une constante dans l’histoire des calendriers qu’il existe un modèle avec de simples variantes locales 226. Il en résulterait donc que le calendrier ḥimyarite, qui était la référence en Arabie pendant l’Antiquité tardive, avait initialement son commencement à l’équinoxe de printemps. En résumé, il est assuré que, aux 5e et 6e s., l’année ḥimyarite commençait à une date assez proche de l’équinoxe de printemps. De manière plus conjecturale, on peut supposer qu’elle commenEn l’an 10 de l’hégire, Muhammad préside aux cérémonies du pèlerinage, pour la première et la dernière fois. Ce pèlerinage est qualifié de « grand », parce qu’il correspond à la fin d’un cycle et au début d’un nouveau : il donne donc la date idéale du début de l’année. Or c’est aussi le seul pèlerinage de la vie de Muḥammad dont on connaisse la date exacte, les 8, 9 et 10 mars 632, parce que, à cette occasion, Muḥammad a réformé le calendrier. La date des pèlerinages antérieurs est incertaine parce qu’on ne sait pas quand les mois intercalaires étaient ajoutés. L’année qui suit ce grand pèlerinage (1er muḥarram 11 h.) commence le 29 mars 632. C’est elle qui sert de modèle dans le Tableau 1. 225 Sur ces questions compliquées, brièvement évoquées ici, voir Robin 2016 a à paraître. 226 Pour l’Arabie du Sud, voir ci-dessous, pp. 50–55. 224

LES ÉVOLUTIONS DU CALENDRIER

339

çait initialement avec cet équinoxe (hypothèse retenue dans le Tableau 1), mais qu’elle a dérivé entre le 5e et le 6e s. Une telle hypothèse est d’autant plus plausible que le calendrier babylonien commençait avec l’équinoxe de printemps, de même que tous les calendriers qui en dérivaient, de manière sûre (comme le calendrier juif et celui de Makka en 11 h.) ou vraisemblable (comme les calendriers de Sabaʾ, Nabaṭ, Palmyre et, bien sûr, Ḥimyar). Il est donc possible que le premier mois de l’année ḥimyarite, dhu-ḥullatān, ait coïncidait initialement avec le mois lunaire juif de nîsān. Je reviens à la question initiale qui était de savoir si la conversion de Ḥimyar au judéo-monothéisme avait entraîné une réforme du calendrier. Une telle réforme apparaît plausible, pour des raisons politiques et religieuses, sans que les source disponibles permettent de la prouver et de la dater.

D. L’HYPOTHÈSE D’UNE RÉFORME DU CALENDRIER PAR ABRAHA ENTRE 548 ET 560, AVEC ALIGNEMENT SUR LE CALENDRIER JULIEN

Quand Ḥimyar, qui avait un calendrier lunaire avec des mois intercalaires sur le modèle des juifs, est officiellement devenu chrétien vers 530, la question d’un alignement de ce calendrier sur celui des chrétiens, solaire et julien, dans sa version aksūmite ou syrienne, s’est inévitablement posée. Il est inutile de souligner l’importance qu’avait alors le calendrier, notamment pour la fixation de la fête de Pâque(s), aussi bien pour les juifs que pour les chrétiens. Depuis le concile de Nicée, l’orientation générale des églises chrétiennes était de définir la date de Pâques de manière à ce que les chrétiens ne célèbrent pas cette fête au même moment que les juifs 227. La séparation, qui intervient progressivement dans le courant du 4e s., est totale au 6e s. 228. Si une réforme du calendrier semble avoir été indispensable, elle n’est pas encore intervenue en 548 puisque, cette année-là, on observe encore l’ajout d’un mois intercalaire, comme nous l’avons Stern 2012, pp. 326–330, 380–424. La célébration de Pâques « avec les juifs » est désormais confinée à quelques mouvements sectaires (Stern 2012, pp. 412–422. 227 228

340

CHRISTIAN JULIEN ROBIN

indiqué. Ce n’est pas étonnant : jusqu’à cette date, Abraha était entièrement accaparé par la consolidation de son pouvoir, ce que confirme l’absence complète d’inscriptions. Aucune réforme du calendrier n’est prouvée par les textes postérieurs, mais on en observe les conséquences. Comme nous l’avons vu 229, les mois ḥimyarites n’ont pas dérivé de manière significative par rapport aux saisons pendant les six ou sept siècles qui séparent la fin de l’Antiquité et le 13e s. L’absence de dérive est particulièrement significative parce que, après l’effondrement du royaume de Ḥimyar, vers 570–575, il n’y a plus eu d’autorité politique susceptible de gérer le calendrier et de décider des intercalations. S’il n’y a pas eu de dérive c’est sans doute parce que, avant 570, le calendrier ḥimyarite a été réformé et aligné sur le calendrier solaire syrien (lui-même aligné sur le calendrier julien). Dès lors, non seulement le calendrier ḥimyarite était fixé et parfaitement prévisible, mais encore, en cas de doute, il était facile de se tourner vers le modèle syrien. On peut donc faire l’hypothèse que le calendrier ḥimyarite, lunaire au moment de la conquête aksūmite, a été réformé et aligné sur le calendrier solaire julien entre 548 et 570. Il est possible de resserrer encore un peu cette fourchette. La réforme a dû intervenir avant l’inauguration de la Grande Église de Ṣanʿāʾ, appelée en arabe al-Qalīs, vers 560 230. Selon la Tradition arabo-islamique, Abraha voulait faire d’al-Qalīs un centre de pèlerinage chrétien pour toute l’Arabie : une telle ambition impliquait d’avoir le même calendrier liturgique que les chrétiens monophysites et nestoriens qu’on voulait attirer. Une tradition d’al-Azraqī s’accorde également avec l’hypothèse d’une réforme du calendrier ḥimyarite par Abraha. Elle rapporte que les souverains ḥimyarites déléguaient la responsabilité de l’intercalation aux princes kindites d’Arabie centrale. Or, plusieurs générations avant l’Islam (entre trois et sept selon les sources), c’est au personnage, surnommé al-Qalammas qu’échoit la charge de l’intercalation dans le calendrier de Makka :

229 230

Voir ci-dessus, pp. 36–40. Robin 2015 B.

LES ÉVOLUTIONS DU CALENDRIER

341

« Selon (Ibn) al-Kalbī, le premier qui, dans Muḍar, retarda les mois fut Mālik b. Kināna, du fait qu’il était apparenté par mariage à Muʿāwiya b. Thawr le Kindite — il se trouvait alors dans Kinda —. Auparavant, les hommes en charge de l’intercalation appartenaient à Kinda, parce qu’ils étaient avant cela les rois des Arabes de Rabīʿa et de Muḍar et que Kinda était l’un des soutiens des princes ḥimyarites. Alors Thaʿlaba b. Mālik fut en charge de l’intercalation, puis à sa suite al-Ḥārith b. Mālik b. Kināna qui est le Qalammas ; ensuite, après le Qalammas, Surayr b. al-Qalammas fut en charge de l’intercalation. Puis les hommes en charge de l’intercalation appartinrent aux banū Fuqaym issus des banū Thaʿlaba, jusqu’à la venue de 231 l’Islam » .

Une explication plausible de ce transfert de responsabilité des agents de Ḥimyar au Qalammas est que Ḥimyar, entre trois et sept générations avant l’Islam, avait cessé de pratiquer l’intercalation. L’hypothèse que la réforme du calendrier est en relation avec l’inauguration d’al-Qalīs est indirectement confirmée par les traditions qui relatent pourquoi Abraha a voulu détruire la Kaʿba : la toute nouvelle église est profanée par un homme de Kināna, précisément en charge de l’intercalation. Voici comment la Vie exemplaire de Muḥammad le rapporte : « Abraha construisit al-Qalīs à Ṣanʿāʾ. Il édifia une église dont on avait jamais vu la semblable n’importe où sur la terre à cette époque. Il écrivit alors au négus : « Ô roi, j’ai moi-même construit pour toi une église telle que jamais la semblable n’a été bâtie pour un roi avant toi. Et je ne serai pas en reste tant que je

al-Azraqī, Akhbār Makka, texte, pp. 182–183 : wa-qāla al-Kalbī : faawwal man ansaʾa al-shuhūr min Muḍar : Mālik b. Kināna wa-dhālika anna Mālik b. Kināna nakaḥa ilà Muʿawiya b. Thawr al-Kindī — wa-huwa yawmaʾidhin fī Kinda — wa-kānat al-nasāʾa qabla dhālika fī Kinda, li-anna-hum kānū qabla dhālika mulūk al-ʿArab min Rabīʿa wa-Muḍar wa-kānat Kinda min ardāf almaqāwil fa-nasaʾa Thaʿlaba b. Mālik, thumma nasaʾa baʿda-hu al-Ḥārith b. Mālik b. Kināna wa-huwa al-Qalammas, thumma nasaʾa baʿd al-Qalammas Surayr b. alQalammas, thumma kānat al-nasāʾa fī banī Fuqaym min banī Thaʿlaba, ḥattà jāʾa al-Islām. 231

342

CHRISTIAN JULIEN ROBIN n’aurai pas détourné le pèlerinage des Arabes vers elle ». Quand les Arabes parlèrent de cette lettre d’Abraha au négus, l’un des hommes chargé de l’intercalation, issu des banū Fuqaym …, se mit en colère… Le Kinānite se mit en route et parvint à al-Qalīs, où il s’accroupit (pour déféquer), puis sortit et rentra chez lui … » 232.

Si ce texte signifie bien que la réforme du calendrier a rencontré une forte opposition qui se traduit en détestation d’al-Qalīs, l’hypothèse d’une relation entre l’édification d’al-Qalīs et la réforme du calendrier est renforcée. Pour sa réforme du calendrier, Abraha avait le choix entre Aksūm et la Syrie. Les listes médiévales qui donnent les mois syriens et leur équivalent ḥimyarite, suggèrent que c’est le modèle syrien qui s’est imposé. C’est en accord avec la politique ecclésiastique d’Abraha, qui est davantage tournée vers la Syrie que vers l’Afrique 233. Nous aboutissons finalement aux propositions suivantes, qui comportent une part importante d’hypothèse : — 3e et 4e s. : les calendriers sabéens et ḥimyarites sont lunaires, avec des corrections empiriques ; il est probable qu’ils sont identiques sauf pour les noms donnés au mois. — fin du 4e s. : réforme du calendrier ḥimyarite sur le modèle d’un calendrier juif. Le mois intercalaire est désormais obtenu par redoublement du premier mois, dhu-ḥullatān, qui commence idéalement avec l’équinoxe de printemps.

Thumma inna Abraha banà al-Qalīs [ainsi, de préférence à al-Qullays] bi-Ṣanʿāʾ, fa-banà kanīsa lam yura mithlu-hā fī zamāni-hā bi-shayʾ min al-arḍ, thumma kataba ilà al-najāshī : “innī qad banaytu la-ka ayyuhā ʾl-malik kanīsa lam yubna mithlu-hā li-malik kāna qabla-k wa-lastu bi-muntahin ḥattà aṣrifa ilay-hā ḥajj al-ʿArab, fa-lammā taḥaddathat al-ʿArab bi-kitāb Abraha dhālika ilà al-najāshī, ghaḍaba rajul min al-nasaʾa, aḥad banī Fuqaym b. ʿAdī b. ʿĀmir b. Thaʿlaba b. alḤārith b. Mālik b. Kināna b. Khuzayma b. Mudrika b. Ilyās b. Muḍar… Fakharaja al-Kinānī ḥattà atà al-Qalīs fa-qaʿada fī-hā — qāla Ibn Hishām yaʿnī aḥdatha fī-hā — Qāla Ibn Isḥāq : thumma kharaja fa-laḥiqa bi-arḍi-hi (Ibn Hishām, Sīra, texte, vol. I, pp. 43–45 ; traduction anglaise, pp. 21-22). 233 Robin 2014 a, pp. 71-77. 232

LES ÉVOLUTIONS DU CALENDRIER

343

— Entre le milieu du 5e s. et le règne d’Abraha, le mois dhuḥullatān dérive de mars-avril à février. — Entre 548 et 560, nouvelle réforme du calendrier qui devient solaire et julien, sur le modèle du calendrier syrien chrétien. Ce calendrier ḥimyarite solaire est encore en usage aux 13e et 14e s. dans le sud du Yémen, comme l’attestent les textes d’époque rasūlide.

E. DES CALENDRIERS SUDARABIQUES MOINS DIVERS QU’IL N’Y PARAÎT

Incidemment, pour conclure cette étude du calendrier ḥimyarite, je voudrais montrer que la grande diversité des calendriers sudarabiques est plus apparente que réelle. Nous avons déjà vu que, d’une part, les mois des calendriers de Radmān et de Ḥimyar au 3e s. 234 et, d’autre part, ceux des calendriers de Sabaʾ et de Ḥimyar aux 4e et 5e s. 235 sont interchangeables. Cette observation semble pouvoir être généralisée. Parmi les calendriers sudarabiques les plus anciens (ceux des royaumes de Sabaʾ, de Qatabān, de Maʿīn et de Haram), les deux premiers qui sont les mieux connus présentent trois caractères communs qui paraissent significatifs. Le premier est la division du mois en trois décades, appelées frʿ, fqḥy (ou fqḥw) et ʾgby(w) 236. Une telle division du mois est attestée dans le calendrier solaire égyptien, mais aussi dans les calendriers lunaires de Grèce et de Chine. Elle est tout aussi logique que la division du mois en quatre parties (ou semaines), comme le montrent les décades du calendrier révolutionnaire français. Elle n’implique donc pas nécessairement que le modèle des calendriers avec décades soit en Égypte. A. F. L. Beeston, qui a mis en évidence l’existence de ces décades, suppose que les mêmes divisions du mois était utilisée dans

Voir ci-dessus p. 19. Voir ci-dessus p. 26. 236 Beeston 1956, pp. 3–10. 234 235

344

CHRISTIAN JULIEN ROBIN

le royaume de Maʿīn ; mais, cette proposition reste hypothétique puisque un seul nom de décade, ṭʿn, est attesté 237. Un deuxième point commun entre les calendriers de Sabaʾ et de Qatabān est l’existence de mois qui peuvent être mentionnés de trois manières : sans épithète ou avec l’épithète « premier » (qdm) ou encore avec l’épithète « dernier » (ʾḫr). On connaît un mois à Sabaʾ, ḏ-Ns¹wr, et un à Qatabān, ḏ-Brm, qui se présentent sous ces trois formes 238. À Sabaʾ, un second mois, ḏ-ʿṯtr, est également attesté avec l’épithète « premier » 239. L’ajout occasionnel d’une épithète signale probablement la possibilité d’un redoublement 240. Les mois ḏ-Brm « dernier » et ḏ-Ns¹wr « dernier » (plus l’hypothétique ḏ-ʿṯtr « dernier » qui n’est pas encore attesté 241) pourraient donc être des mois intercalaires, d’un type qui reste à définir 242. Il faut noter ici que, dans le calendrier sabéen, deux mois différents pouvaient apparemment être redoublés. L’ajout de mois intercalaires est propre aux calendriers lunaires que l’on veut maintenir en phase avec le rythme des saisons. Beeston 1956, p. 7. Il n’est pas même sûr que ṭʿn soit un nom de décade : c’est seulement quand on aura un nom pour chacune des trois décades du mois que cette hypothèse sera solidement fondée. 238 Robin 2012 a, p. 144. Aux références à ḏ-Ns¹wr qdmn, ajouter L302 et M-SAS 1, et à ḏ-Ns¹wr ʾḫrn, L089, L289 et L319. 239 Ce nom vient d’être découvert dans ATHS 39. 240 S’il est difficile de dire précisément à quelle date remontent les attestations les plus anciennes de ces mois avec épithètes, il est assuré que c’est très tôt, peut-être dès l’époque perse. 241 L’existence d’un mois ḏ-ʿṯtr « premier » implique certainement un ḏ-ʿṯtr « dernier ». 242 Une autre interprétation est théoriquement envisageable : initialement, les calendriers qatabānite et sabéen auraient eu un mois appelé ḏBrm et ḏ-Ns¹wr, qui aurait été dédoublé dans le cours du temps, à la manière du dédoublement des mois syriens tishrīn et kanūn (voir aussi les mois jumeaux du calendrier musulman, rabīʿ et jumādà, dont l’origine n’est pas connue). Cette hypothèse, cependant, paraît peu vraisemblable parce que les deux exemples de dédoublement qui viennent d’être cités datent de l’époque romaine et l’antiquité tardive, alors que les mois qatabānites et sabéens concernés sont beaucoup plus anciens. 237

LES ÉVOLUTIONS DU CALENDRIER

345

Le dernier point commun réside dans un nombre significatif de noms de mois qui sont dérivés d’un nom de divinité, d’un rite ou d’un toponyme. À Qatabān, sur les neuf noms connus — sans compter ceux qualifiés de « premier » (qdmn) et « dernier » (ʾḫrn) —, huit sont dans ce cas : ḏ-ʾbhw (mois du pèlerinage à Sabaʾ), ḏ-ʿm (dieu), ḏ-Brm (Barm, nom antique du wādī Bayḥān), ḏ-Brm qdmn, ḏBrm ʾḫrn, ḏ-Bs²mm (dieu), ḏ-ḏbḥtm (« sacrifice »), ḏ-ms¹lʿt, ḏ-Rbs² (dieu), ḏ-S¹ḥr (dieu), ḏ-Tmnʿ (Tamnaʿ, la résidence royale). Pour Sabaʾ, une petite difficulté tient au fait qu’il n’est pas toujours facile de déterminer quel sont les mois proprement sabéens et ceux qui appartiennent aux calendriers locaux du Jawf. Parmi les mois probablement sabéens qui sont au nombre de quatorze, sept au moins se réfère à la religion : ce sont ḏ-ʾbhy (pèlerinage), ḏ-ʾlʾlt (« divinités »), ḏ-ʿṯtr (dieu), ḏ-ʿṯtr qdmn, ḏ-dnm, ḏ-fls¹m, ḏ-H(w)bs¹ (divinité), ḏ-ḫzfm 243, ḏ-ks²bm, ḏ-mlyt, ḏ-nʿm, ḏ-Ns¹wr (divinité), ḏ-Ns¹wr qdmn, ḏ-Ns¹wr ʾḫrn, ḏ-nylm (rite), ḏ-S¹bʾ, ḏ-S¹ḥr (divinité) 244. À Maʿīn, petit royaume du Jawf, les mois dont le nom s’inspire des pratiques religieuses sont quatre sur dix : ḏ-ʾbhy (mois du pèlerinage à Sabaʾ), ḏ-ʾbrhn, ḏ-ʾṯrt (déesse), ḏ-dṯʾ (et ḏ-bkr dṯʾ), ḏḥḍr (et ḏ-ḥḍr Wd ; « pèlerinage » et « pèlerinage de Wadd » 245), ḏ-nwr, ḏ-S¹mʿ (dieu), ḏ-S²ms¹y (probablement la déesse S²ms¹), ḏ-s³[…], ḏtrdy, ḏ-ṭnft. À Haram enfin, autre petit royaume du Jawf, ils sont

Voir X.SBS 26 / 4-5 (l-ḏ-ḫ(5)zfm b-ḫrf S¹mhkrb bn (6) Kbr H̲ll … … b-(ḏ)- ʾlʾlt ḏ-ḫrf Bʿṯt(8)r bn S²rs¹hmw). 244 On a supposé dans le passé que le calendrier sabéen comptait un mois nommé Ṯwr (voir en dernier lieu Ryckmans Jacques 1973 a, p. 333). Cette opinion se fondait sur Louvre 6 = CIH 671 qui se lit mḏbḥt b-h y(2)ḏbḥn mlkn ṯ(3)wrm b-ywm ts¹ʿm ḏ-Ṯwr, « autel sur lequel sacrifie le roi un taureau le neuvième jour de dhu-Thawr ». L’opinion d’A. F. L. Beeston selon laquelle il vaut mieux comprendre « le neuvième jour de la fête de Thawr » paraît plausible (1956, pp. 9–10). 245 Bar-Int 9 / 3 inédit. (7)

243

346

CHRISTIAN JULIEN ROBIN

deux sur quatre : ḏ-ʿṯtr (dieu), ḏ-mḫẓdm qdmtn 246, ḏ-mwṣbm et ḏ-s¹lʾm (« offrande » 247). Les royaumes de Sabaʾ, Qatabān, Maʿīn et Haram sont centrés sur les oasis de piémont du Yémen intérieur. Ils se distinguent des royaumes de la montagne par de nombreux caractères, tout particulièrement l’organisation sociale. On peut se demander s’il en va de même des calendriers. Une proportion importante des noms de mois à Sabaʾ, Qatabān, Maʿīn et Haram fait référence à des divinités et à des rites. À Ḥimyar, en revanche, comme nous l’avons vu, les mois sont nommés d’après les moments les plus significatifs du calendrier agricole. La même observation est valide pour Samʿī, une vaste confédération tribale des régions au nord de Ṣanʿāʾ, qui fut à haute époque un royaume. Sur huit mois connus, ḏ-ʾbhy, dṯʾ (« printemps »), ḫrf (« saison des pluies »), qyẓ (« chaleur »), s²wrm, ṣrr, ṣydm (« chasse ») et [...]ʾn, quatre dṯʾ, ḫrf, qyẓ et ṣydm, et même cinq si on restitue [mḏr]ʾn, qui appartiennent aussi au calendrier ḥimyarite, se rapportent aux saisons ou aux activités de la campagne 248. Cette parenté des mois de Samʿī et de Ḥimyar conduit à classer les calendriers sudarabiques en deux ensembles, ceux des oasis et ceux de la montagne, les premiers étant « religieux » et les seconds « agricoles ». Concernant la division du mois en décades, on la retrouve peut-être dans Samʿī si on se fonde sur une mention de ḏ-ʾgby dans RES 4176 / 13, quelques siècles avant l’ère chrétienne, dans un décret qui organise l’intégration de Samʿī dans le royaume de Sabaʾ. Pour Ḥimyar, où les informations sont sporadiques jusque vers le milieu du 3e s. è. chr., notre ignorance est totale. Il n’est pas assuré que qdmtn soit un adjectif se rapportant à ḏmḫẓd . Si c’était le cas, on aurait exemple supplémentaire de mois pouvant être redoublé. 247 La langue de Haram amīrite (du 2e s. av. è. chr. au 1er s. ap.), comme celle de Qaryat al-Faʾw, remplace le s³ initial par s¹. 248 Il n’est pas impossible que le mois s²wrm soit également commun aux calendriers de Samʿī et de Ḥimyar si on lit (s²)wrm dans l’inscription ḥimyarite MQ-al-Qashʿa 1 (et non (ṯ)wrm) (voir ci-dessus, p. 27). Pour le mois ḫrf, voir Bron-al-ʿAdan 2 (b-ḫrf (8) ḏ-ḫrf Rbm bn Hwf(9)ʿṯt bn ʿmydʿ). 246

m

LES ÉVOLUTIONS DU CALENDRIER

347

Pour tous ces calendriers anciens (Sabaʾ, Qatabān, Maʿīn, Haram et Samʿī), on ignore quel était l’ordre des mois et le début de l’année. Pour Sabaʾ, on sait néanmoins que les mois ḏ-Hbs¹ et ḏ-ʿṯtr se suivaient (Ja 611 / 7-8), de même que les mois ḏ-S¹bʾ, ḏ-ʾlʾlt et ḏʾbhy (Ja 671 + 788 / 23–25). On sait également que ḏ-S¹bʾ, à la fin du 1er s. av. è. chr., correspondait au mois nabatéen de ṭbt, approximativement décembre-janvier (DAI Ṣirwāḥ 2004-12B+20046+2002-83+1993/3-x, inédit) 249. De ce fait, ḏ-ʾlʾlt, qui suit ḏ-S¹bʾ, tombait en janvier-février 250 et ḏ-ʾbhy (dhu-abhī), le mois du pèlerinage au Grand Temple d’Almaqah à Marib, en février-mars. Vers 370, de gros dégâts sur l’écluse méridionale et le mur de retenue de la Digue de Marib sont réparés pendant trois mois (ḏS¹bʾ, ḏ-ʾlʾlt et ḏ-ʾbhy). Cette intervention, la plus longue que l’on connaisse, n’est pas gênée par la crue qui a été tardive cette annéelà (Ja 671 + 788 / 10–17 et 23–25). La chronologie la plus vraisemblable des événements est une rupture de la Digue pendant la crue de l’été (juillet-août), suivie par une réparation avant la crue du printemps (mars-avril-mai) ; malheureusement, les travaux durent plus de temps que prévu ; ils peuvent néanmoins être menés à leur terme grâce à la faveur divine qui retarde la crue et sont achevés à la fin de dhu-abhī (février-mars) 251. Cette chronologie s’accorde avec celle des campagnes de réparation de Shuriḥbiʾīl et d’Abraha : les quatre qui sont connues sont organisées entre octobre et mars approximativement 252. Par ailleurs, il n’est guère vraisemblable que le roi ḥimyarite ait programmé d’importants travaux entre les pluies du printemps et celles de l’été parce que le période favorable, le mois de juin, est trop courte pour prendre un tel risque. On aurait donc la preuve que le calendrier sabéen est parfaitement stable entre la fin du 1er s. av. è. chr. et la fin du 4e s., soit pendant près de 400 ans. Voir déjà ci-dessus, p. 3 et n. 6. A. G. Lundin situait déjà le mois de ḏ-ʾlʾlt en mars dans sa recension de Beeston 1956 (Lundin 1959, p. 211). 251 Voir Robin 2016 b à paraître. 252 Voir ci-dessus, pp. 39–40. 249 250

348

CHRISTIAN JULIEN ROBIN

Le mois de dhu-abhī était celui du pèlerinage au Grand Temple d’Almaqah de Marib 253. Des sacrifices avaient lieu le septième jour de ce mois, peut-être pour en marquer la fin 254. Le mois de dhuabhī coïncidait aussi avec le début de la saison des pluies. Une inscription 255 date en dhu-abhī une grande procession pour obtenir très rapidement de la pluie 256 et une seconde 257 situe en dhu-abhī des averses torrentielles. En croisant toutes ces données, on arrive à la conclusion que le pèlerinage au Grand Temple d’Almaqah se tenait pendant la première décade du mois de dhu-abhī qui correspondait à février-mars, c’est-à-dire au dernier mois avant l’équinoxe de printemps 258. Il faut rappeler cependant qu’une date plus tardive pour dhuabhī a été défendue par Jacques Ryckmans : « juste avant le début de la saison de pluies d’automne » 259. L’argument se trouve dans l’inscription Ja 735, qui date de la fin du 3e s. ou du début du 4e : les pluies tombant en ḏ-ʾbhy y seraient qualifiées de « pluies d’été » (brq ḫrf) 260 : Ryckmans J. 1973 a, pp. 333–334. Un autel de Marib mentionne le 7 de dhu-abhī (b-s¹bʿm ḏ-ʾbhy, CIH 694) ; un autre, déjà cité (n. 241), le 9 de (la fête de) Thawr (Louvre 6 = CIH 671). 255 Ja 735 / 4-5 et 15–17. 256 Voir Jacques Ryckmans (1973 b) qui interprète cette grande procession comme une cérémonie spécifique pour demander de la pluie (istisqāʾ). Une autre interprétation serait de la considérer comme l’un des rites du pèlerinage annuel. 257 Ja 651 / 14-21. 258 Il semblerait que le pèlerinage de Makka se soit tenu exactement à la même époque. Le seul pèlerinage de la vie de Muḥammad dont on connaisse la date, celui de 10 h., au cours duquel le calendrier fut réformé (avec la suppression des mois intercalaires), a eu lieu les 8–10 mars. Cette date se déduit du fait que la période du pèlerinage (8–10 dhū ʾl-ḥijja) commence trois semaines avant le début de la nouvelle année (le 1er muḥarram 11 h. qui est tombé le 29 mars 632). Voir Robin 2016 b à paraître. 259 Ryckmans Jacques 1973 a, p. 332 et n. 38 ; voir aussi Ryckmans Jacques 1983, p. 17 (automne). 260 « Pluie d’automne » selon Jacques Ryckmans (1973 a, p. 332, n. 38). 253 254

LES ÉVOLUTIONS DU CALENDRIER

349

« … en re(3)connaissance parce qu’Il a accordé et a annoncé à ses serviteurs la commune Sabaʾ Kahlān qu’il accorderait et (4) donnerait de l’eau lors des pluies de ḫrf, au mois de dhu-abhī de l’année de Tubbaʿkarib fils de Wadadʾīl b. Kabīr Kha(5)līl la neuvième », … ḥ(3)mdm b-ḏt ḫmr w-tbs²rn ʾdm-hw s²ʿbn S¹bʾ Khln k-yḫmrn w-(4)s¹qy brq ḫrf b-wrḫ ḏ-ʾbhy ḏ-ḫrf Tbʿkrb bn Wddʾl bn Kbr H̲(5)ll ts¹ʿn

(Ja 735 / 2-5)

Il nous semble que ce texte doive être interprété autrement : Almaqah a promis au mois de dhu-abhī que les pluies du prochain été seraient abondantes. Il est peu vraisemblable que la date se rapporte aux pluies puisque celles-ci sont déjà identifiées par un nom de saison 261. En bref, les calendriers de Sabaʾ et de Qatabān, à époque ancienne, présentent des caractères communs (décades et mois munis d’une épithète), comme ceux de Sabaʾ et de Ḥimyar au 4e s. è. chr. (mois interchangeables et mois munis d’une épithète). Même s’il n’est pas possible de le démontrer véritablement, il est probable que, à une époque donnée, tous les calendriers de l’Arabie méridionale étaient moulés sur le même modèle, et qu’ils ne se distinguaient que par les noms de mois et peut-être quelques dispositions spécifiques. Avec le temps, ces calendriers ont évidemment évolués, mais ils l’ont fait dans le même sens et, apparemment, au même rythme. Une telle hypothèse implique qu’il y a sans doute eu

Il en va sans doute de même dans Fa 71 qui mentionne des « pluies de printemps » (ci-dessus, n. 186) avec une date qui se réfère au mois qui précède dhu-abhī : … et en reconnaissance parce qu’(11)Il a donné de l’eau en abondance à tout Raḥbatān [la région de Ṣanʿāʾ], aux deux H(12)irrān et à tout le domaine d’Almaqah et a donné (13) de l’eau pendant cette saison des pluies de printemps, au mois de dhu-īlʾīl(14)āt de l’année de Maʿdīkarib fils de (15) [Su]mhūkarib b. Faḍḥum … », … w-ḥmdm b-ḏt (11) s¹qy whs²fqn kl Rḥbtn w-H(12)rnhn w-kl mlk ʾlmqh w-s¹q(13)y hwʾ dṯʾn b-wrḫ ḏ-ʾlʾl(14)t ḏ-ḫrf Mʿdkrb bn (15) [S¹]mhkrb bn Fḍḥm … (Fa 71 / 13-14). 261

350

CHRISTIAN JULIEN ROBIN

un modèle, vraisemblablement le calendrier d’un État 262 qui avait acquis une position hégémonique. Ce modèle semble avoir été successivement le calendrier de Sabaʾ, puis celui de Ḥimyar. Pour terminer, je voudrais mentionner un texte publié récemment (mais qui n’a pas été parfaitement compris), prouvant que le calendrier de Sabaʾ est devenu « prédictif » peu avant sa disparition. Comme tous les calendriers lunaires, le calendrier sabéen a probablement été d’abord « empirique », c’est-à-dire fondé sur l’observation de la lune. C’est seulement à partir d’un certain seuil de développement qu’ont été fixés à l’avance la durée de chaque mois et la place des mois intercalaires (qui pouvaient être « prédits ». Ce texte est un certificat en écriture cursive sur bâtonnet : ḏkr Ḥyws²f bnt Tḥyly ʾmt bn ʿṯkln k-ts¹bʾ(2)n w-mḍʾ ʿdy S²bmm l-ḥḍr Qynn b-wrḫ ḏ-Ns¹w(3)r qdmn gyl ḫrf yʿqbn ḫrf Ns²ʾkrb bn S¹(4)mhkrb bn Ḥzfrm ḏ-Ḍmrn … (6) wkwn ḏn ʿlmn b-ḏ-Ns¹wr ḏ-m-ḏn ḫrfn, « Certificat de Ḥaywshāf du lignage Tḥyly servante de Ibn ʿUthkulān, pour qu’elle se rende et parvienne à Shibāmum pour le pèlerinage de Qaynān au mois de dhu-Niswa(3)r le premier, gyl de l’année qui suivra l’année de Nashaʾkarib fils de Su(4)mhūkarib ibn Ḥazfarum dhu-Ḍamrān … (6) A été établi ce document en dhu-Niswar de cette même année » (M-SAS 1) 263.

Les données chronologiques de ce document sont les suivantes : L’hypothèse du calendrier d’un temple est moins probable parce que, si l’on en croit Sacha Stern, le calendrier est une affaire essentiellement politique. 263 Mohamed Maraqten a traduit : « A declaration of Ḥaywšāf, daughter of Taḥylay, female vassal of the clan Banī ʿUṯkulān that she should carry out a journey (2) and proceed into (the city) of Šibāmum to perform the pilgrimage of (the God) Qaynān in the month of Ḏū-Naswar (3) the first. This was in the period of the eponymate year, which succeeds the eponymate year of Našaʾkarib, son (4) Sumukarib of the clan Ḥazfarum, of the family Ḏū-Ḍumrān… (6) And this document was marked in the month Ḏū-Naswar, which is in the same eponymate year. Ḥaywšāf, daughter of Taḥylay. Signature » 262

LES ÉVOLUTIONS DU CALENDRIER

351

— Le document enjoint à une femme de participer au (prochain) pèlerinage de Qaynān qui aura lieu au mois de dhuNiswa(3)r le premier. — Ce sera alors « l’année qui suivra l’année de Nashaʾkarib fils de Su(4)mhūkarib ibn Ḥazfarum dhu-Ḍamrān », c’est-à-dire l’année qui suivra la présente année. Si le document ne nomme pas cette année prochaine, c’est parce que son éponyme n’est pas encore connu. — Le document est rédigé au mois de dhu-Niswar, sans qu’il soit précisé qu’il est le « premier » ou le « dernier ».

Plusieurs observations peuvent être faites : 1. Dans l’année qui suit celle de Nashaʾkarib fils de Su(4)mhūkarib ibn Ḥazfarum, le mois dhu-Niswar est qualifié de « premier ». En revanche, dans l’année précédente (celle de Nashaʾkarib fils de Su(4)mhūkarib ibn Ḥazfarum), il n’a pas d’épithète. On peut en déduire qu’il n’y a pas deux mois appelés « dhuNiswar » dans le calendrier sabéen (comme il y a deux kanūn dans le calendrier syrien et deux rabīʿ dans le calendrier islamique), mais que le mois dhu-Niswar peut être redoublé quand c’est nécessaire. Or, s’il peut être redoublé, c’est que le calendrier est lunaire. 2. L’hypothèse que le mois dhu-Niswar est redoublé paraît confirmée par la terminologie. Alors que « mois » se dit wrḫ, « dhuNiswar le premier » est appelé gyl, hypothétiquement « mois intercalaire » puisque ce mot, n’est attesté qu’en liaison avec le mois dhuNiswar le premier 264. On peut donc supposer que, lors d’un redoublement, l’intercalaire (gyl) est le mois appelé « premier », tandis Voir les trois inscriptions formelles RES 4646 / 19-20 (b-wrḫ ḏNs¹wr qdmn g(20)yl ḫrf Mʿdkrb bn Tbʿkrb bn (21) Ḥzfrm s¹dṯn) ; Gl 1533 = Ja 2855 / 12–13 (b-ḏ Ns¹wr (13) qdmn gyl ḫrf Yṯʿm bn Ḥzfrm ḏ-Rf(14)(d)n ṯkmtn) ; Gl 1572 / (b-[ḏ Ns¹wr] (10) qdmn gyl ḫrf Yqhmlk bn ʾ[b](11)ʿm Kbr Ḫll). Voir aussi les trois textes cursifs X.SBS 31 = Mon.script.sab 161 / 11 (w-kwn ḏn ʿ(10)lmn b-(s²)hrm ḏ-frʿ ḏ-n(11)s¹wr qdmn gyl ḫrf M(12)ʿdkrb bn ʿmʾmr bn (13) Fḍḥm) ; X.SBS 37 A = Mon.script.sab 241 / 3 (ʿdy [xx]xm ḏ-fqḥy [ḏ]-ns¹wr qdmn gyl [ḫr]f T(4)bʿkrb bn Ns²ʾkr[b] bn Fḍḥm) ; ATHS 13 (=YM 11 296) / 4 (wrḫm ḏ-Ns¹(4)wr qdmn gyl ḫrf ʾlrm Kbr H̲ll) ; ainsi que L312 / 9 (…](.) (g)yl ḫrf Mʿdkrb ḏ-Mbḥrt ṯk(10)[mtn …). 264

352

CHRISTIAN JULIEN ROBIN

que le mois régulier (wtḫ, qui conserve ses fêtes) est le « dernier ». C’est aussi le choix du calendrier juif. 3. Concernant l’année prochaine, l’auteur du document ne connaît pas le nom de l’éponyme, mais il sait que le mois dhuNiswar sera redoublé. Il existe donc alors un système d’intercalation qui fixe à l’avance à quels moments les mois intercalaires seront ajoutés.

Malheureusement, le texte est difficile à dater précisément. L’éponyme mentionné n’est pas connu, ce qui suggère une date antérieure au 3e siècle. Quant à la paléographie, elle indique seulement les siècles qui précèdent l’époque ḥimyarite. On peut seulement dire que ce texte est certainement antérieur à la fin du royaume de Sabaʾ (vers 275 de l’ère chrétienne), et peut-être antérieur à 200 de l’ère chrétienne. Pour fixer les idées, nous proposons vers 200 avec une incertitude de l’ordre d’un siècle. La mise en évidence de la prédictivité du calendrier sabéen confirme que, dans les premiers siècles de l’ère chrétienne, le fonctionnement de ce dernier était aussi rigoureux que celui des calendriers proche-orientaux. Il n’est donc pas étonnant qu’il ait servi de modèle aux voisins de Sabaʾ. Si le calendrier sabéen était prédictif dans les premiers siècles de l’ère chrétienne, ce n’est pas parce les Sabéens s’étaient passionnés pour les astres et avaient acquis de grandes connaissances en astronomie et en astrologie, mais plutôt parce que les élites avaient accepté d’utiliser un système étranger dont elle reconnaissaient la perfection. Cet exemple confirme une nouvelle fois que le postulat de l’isolement intellectuel de l’Arabie, si fréquent dans les études sur calendriers d’Arabie et, plus généralement, sur l’Arabie de Muḥamamd, n’est nullement évident. Quand à l’impression d’anarchie que donne la grande diversité des calendriers et des modes de désignation de l’année, elle est sans doute plus apparente que réelle. Comme au Levant, la diversité permettait aux diverses entités politiques d’afficher une autonomie de façade. Sur le fond, tous s’inspiraient des systèmes en vogue à leur époque à Sabaʾ ou à Ḥimyar, systèmes qui eux-mêmes avaient leur origine à Babylone.

BIBLIOGRAPHIE

Al-Azraqī (Abū ʾl-Walīd Muḥammad b. ʿAbd Allāh b. Aḥmad)

LES ÉVOLUTIONS DU CALENDRIER

353

Akhbār Makka, texte : Akhbār Makka wa-mā jāʾa fī-hā min alāthār, éd. Rushdī al-Ṣāliḥ Malḥas, 2 volumes (4e impression), Makka al-Mukarrama (Dār al-Thaqāfa), 1983 (1403 h.). Beaucamp (Joëlle), Briquel-Chatonnet (Françoise) et Robin (Christian) 1999 « La persécution des chrétiens de Nagrān et la chronologie ḥimyarite », Aram, 11–12, 1999–2000, pp. 15–83. Beeston (A. F. L.) 1956 Epigraphic South Arabian Calendars and Dating, London (Luzac), 1956. 1974 « New Light on the Himyaritic Calendar », dans Arabian Studies, I, 1974, pp. 1–6. al-Bīrūnī (Abū ʾl-Rayḥān Muḥammad b. Aḥmad) al-Āthār al-bāqiya, texte éd. Sachau : Chronologie orientalischer Völker von Albêrûnî, herausgegeben von Eduard Sachau, Leipzig (auf Kosten der Deutschen Morgenlandischen Gesellschaft — F. A. Brockhaus), 1878 (Titre arabe : Kitāb al-Āthār albāqiya ʿan al-qurūn al-khāliya, taʾlīf Abī ʾl-Rayḥān Muḥammad b. Aḥmad al-Bīrūnī al-Khawārizmī). al-Āthār al-bāqiya, traduction Sachau : The Chronology of Ancient Nations, an English version of the Arabic text of the Athâr-ulbâqiya of Albîrûnî or “Vestiges of the Past”, collected and reduced to writing by the author in AH. 390–391, A.D. 1000, translated and edited, with notes and index, by Edward Sachau, London (Oriental Translation Fund of Great Britain & Ireland & W. H. Allen), 1879. Blois (François de) « Taʾrīkh I / 1 » : dans Encyclopédie de l’Islam, nouvelle édition (deuxième), X, 2002, pp. 276–283. Bron (François) 2014 « Une nouvelle dédicace sabéenne au dieu Taʾlab provenant du jabal ʿAdan », dans Semitica et Classica, 7, 2014, pp. 203–204. Cantineau (Jean) 1930–1932 Le nabatéen, Paris (Ernest Leroux), 2 vol., 1930 et 1932. Reprise Otto Zeller, Osnabrück, 1978.

354

CHRISTIAN JULIEN ROBIN

Caussin de Perceval (Armand Pierre) 1843 « Mémoire sur le calendrier arabe avant l’islamisme », Journal asiatique, 4/1, 1843, pp. 342–379. Darles (Christian), Robin (Christian Julien) et Schiettecatte (Jérémie), avec une contribution de Ghassan el Masri, 2014 « Contribution à une meilleure compréhension de l’histoire de la Digue de Maʾrib au Yémen », dans F. Baratte, Ch. J. Robin & E. Rocca (édd.), Regards croisés d’Orient et d’Occident. Les barrages dans l’Antiquité tardive, Actes du colloque tenu à Paris, Fondation Simone et Cino del Duca, 7–8 janvier 2011, organisé dans le cadre du programme ANR EauMaghreb (Orient & Méditerranée 14), Paris, De Boccard, 2014, pp. 9–70. Detoraki (Marina) et Beaucamp (Joëlle) 2007 Le martyre de Saint Aréthas et de ses compagnons (BHG 166). Marina Detoraki, éd., Joëlle Beaucamp, trad. Appendice sur les versions orientales par André Binggeli (Collège de France, Centre de Recherche d’histoire et civilisation de Byzance : Monographies 27 ; Le massacre de Najrān 1), Paris (Association des amis du Centre d’histoire et civilisation de Byzance), 2007. Francuzov (S. A) 2014 Istorija Hadramauta b épohu drevnosti (Istorija Hadramauta, c drevnejših vremen do konca britanskogo blad’čestva, Tome 1 = Studia Yemenica), Sankt-Peterburg (Rossijskaja Akademija Nayk, Institut Vostočnih Rukopisej — Peterburgskoe lingvističeskoe obščectvo), 2014. Hillers ( Delbert R.) et Cussini (Eleonora) 1996 Palmyrene Aramaic Texts, Baltimore & London (The Johns Hopkins University Press), 1996. Höfner Maria) 1973 Inschriften aus Ṣirwāḥ H̲aulān (I. Teil) (Sammlung Eduard Glaser VIII = Österreichische Akademie der Wissenschaften, Philosophisch-historische Klasse, Sitzungsberichte, 291. Band, 1. Abhandlung), Wien (Verlag der Österreichischen Akademie der Wissenschaften), 1973. Huth (Martin)

LES ÉVOLUTIONS DU CALENDRIER

355

2010 Coinage of the Caravan Kingdoms. Ancient Arabian Coins from the Collection of Martin Huth (Ancient Coins in North American Collections), New York (The American Numismatic Society), 2010. Ibn Hishām Sīra, texte : al-Sīra al-nabawiyya li-Ibn Hishām, éd. Muṣṭafà ʾlSaqqā, Ibrāhīm al-Abyārī, ʿAbd al-Ḥāfiẓ Shalabī (Turāth alIslām), 2 vol., Bayrūt (Dār al-Maʿrifa), s.d. Sīra, traduction : The Life of Muhammad. A Translation of Ibn Isḥāq’s Sīrat Rasūl Allāh, with Introduction and Notes by A. Guillaume, Oxford University Press, 1955. al-Iryānī (Muṭahhar ʿAlī) & ʿAbd Allāh (Yūsuf Muḥammad) « al-Taqwīm al-ḥimyarī », dans al-Mawsūʿa al-yamaniyya, éd. Aḥmad Jābir ʿAfīf, Ṣanʿāʾ (Muʾassasat al-ʿAfīf al-thaqāfiyya, 2003 (2e édition), 4 vol., I, pp. 722–725. Laffitte (Roland) 2003 « Sur le zodiaque sudarabique », dans Arabia, 1, 2003, pp. 77–87 et pl. 9–11 (pp. 214–216). Landberg (Le comte de) 1920 Glossaire daṯînois, vol. 1 (ʾ-ḏ), 2 (r) et 3 (z-y, publié par K. V. Zetterstéen), Leide(n) (Brill), 1920, 1923 et 1942. Lundin (A. G.) 1959 Recension d’A. F. L. Beeston, Epigraphic South Arabian calendars and dating, London, 1956, dans Problemy vostokovedenija (Problèmes des études orientales), n° 2, 1959, pp. 211–212. Mahmoud Effendi (connu plus tard sous le nom d’al-Falakī) 1858 « Mémoire sur le calendrier arabe avant l’islamisme », dans Journal asiatique, 1858, pp. 109–192. Maraqten (Muḥammad) / Maraqṭin (Muḥammad) 2014 a « Ḥawl al-ʿilāqāt mā bayn bilād al-Shām wa-ʾl-Yaman qabl al-islām », dans A Pioneer of Arabia. Studies in the Archaeology and Epigraphy of the Levant and the Arabian Peninsula, in Honor of Moawiyah Ibrahim, edited by Zeidan Kafafi & Mohammed Maraqten (ROSAPAT = Rome « La Sapienza » Studies on the Archaeology of Palestine & Transjordan, 10), Rome (« La Sapienza » Expedition to Palestine & Jordan), Rāʾid min ruwwād

356

CHRISTIAN JULIEN ROBIN

al-Jazīra al-ʿarabiyya. Dirāsāt fī āthār wa-nuqūsh bilād al-Shām waʾl-Jazīra al-ʿarabiyya, muqaddama takrīman li-l-ustādh al-duktūr Muʿāwiya Ibrāhīm, taḥrīr Zaydān Kafāfī wa-Muḥammad Maraqtin, 2014, pp. 97–114. 2014 b « The pilgrimage of a Sabaean woman », dans A Pioneer of Arabia. Studies in the Archaeology and Epigraphy of the Levant and the Arabian Peninsula, in Honor of Moawiyah Ibrahim, edited by Zeidan Kafafi & Mohammed Maraqten (ROSAPAT = Rome « La Sapienza » Studies on the Archaeology of Palestine & Transjordan, 10), Rome (« La Sapienza » Expedition to Palestine & Jordan), Rāʾid min ruwwād al-Jazīra al-ʿarabiyya. Dirāsāt fī āthār wa-nuqūsh bilād alShām wa-ʾl-Jazīra al-ʿarabiyya, muqaddama takrīman li-l-ustādh alduktūr Muʿāwiya Ibrāhīm, taḥrīr Zaydān Kafāfī wa-Muḥammad Maraqṭin, 2014, pp. 179–197 (de la partie en langues européennes). al-Masʿūdī (Abū l-Ḥasan ʿAlī b. al-Ḥusayn) Murūj al-dhahab, texte : Murūj al-dhahab wa-maʿādin al-jawhar, édition Barbier de Meynard et Pavet de Courteille, révisée par Charles Pellat (Manshūrāt al-Jāmiʿa al-lubnāniyya, Qism alDirāsāt al-taʾrīkhiyya, 11), 7 vol. Bayrūt, 1965–1979. Murūj al-dhahab, traduction : Les prairies d’or, traduction française de Barbier de Meynard et Pavet de Courteille, revue et corrigée par Charles Pellat (Société asiatique, collection d’ouvrages orientaux), tomes I-IV, Paris (Geuthner), 1962, 1965, 1971 et 1989. Martyrion grec : voir Detoraki & Beaucamp 2007 Müller (Walter W.) 2010 Sabäische Inschriften nach Ären datiert. Bibliographie, Texte und Glossar (Akademie der Wissenschaften und der Literatur – Mainz, Veröffentlichungen der Orientalischen Kommission, 53), Wiesbaden (Harrassowitz), 2010. Nashwān b. Saʿīd al-Ḥimyarī Shams al-ʿulūm, texte : Die auf Südarabien bezüglichen Angaben Našwān’s im Šams al-ʿulūm, gesammelt, alphabetisch geordnet und herausgegeben von ʿAẓīmuddīn Aḥmad (“E. J. W. Gibb Memorial” series, XXIV), Leyden (E.J. Brill) - London (Luzac), 1916.

LES ÉVOLUTIONS DU CALENDRIER

357

Nebes (Norbert) 2004 « A new ʾAbraha inscription from the Great Dam of Mārib », dans Proceedings of the Seminar for Arabian Studies, 34, 2004, pp. 221–230. Nuʿmān (Khaldūn Hazzāʿ ʿAbduh) / Noman (Khaldon) [2012] « A Study of South Arabian Inscriptions from the Region of Dhamār (Yemen) », Dottorato in Orientalistica, Università di Pisa (manuscrit), [2012]. Piamenta (Moshe) 1991 Dictionary of Post-Classical Yemeni Arabic, Leiden (Brill), 2 parts, 1991. Pirenne (Jacqueline) 1974 « A Palaeographical Chronology of the Sabaean-Dated Inscriptions, with Reference to Several Eras », dans Proceedings of the Seminar for Arabian Studies, 4, 1974, pp. 118–130. Prioletta (Alessia) 2013 Inscriptions from the Southern Highlands of Yemen. The Epigraphic Collections of the Museums of Baynūn and Dhamār (Arabia Antica, 8), Roma (« L’Erma » di Bretschneider), 2013. Robin (Christian Julien) 1981 « Le calendrier himyarite : nouvelles suggestions », dans Proceedings of the Seminar for Arabian Studies, 11, 1981, pp. 43– 53. 1994 « À propos d’une nouvelle inscription du règne de Sha’rum Awtar, un réexamen de l’éponymat sabéen à l’époque des rois de Sabaʾ et de dhū-Raydān », dans Arabia Felix. Beiträge zur Sprache und Kultur des vorislamischen Arabien, Festschrift Walter W. Müller zum 60. Geburtstag, herausgegeben von Norbert Nebes, Wiesbaden (Harrassowitz), 1994, pp. 230–249. 1998 « Décompte du temps et souveraineté politique en Arabie méridionale », dans Proche-Orient ancien : temps vécu, temps pensé, Actes de la Table-Ronde du 15 novembre 1997 organisée par l’URA 1062 “Études sémitiques”, édités par Françoise Briquel-Chatonnet et Hélène Lozachmeur (Antiquités sémitiques, III), Paris (Jean Maisonneuve), 1998, pp. 121–151 et 9 fig. entre les pp. 144 et 145.

358

CHRISTIAN JULIEN ROBIN 2003 « Le judaïsme de Ḥimyar », dans Arabia, 1, 2003, pp. 97– 172. 2012 a « Nouvelles observations sur le calendrier de Ḥimyar », dans Rassegna di Studi Etiopici, IV (Nuova Serie), 2012, pp. 119–151. 2012 b « South Arabia, Ethiopia and Socotra », dans Ingo Strauch (éd.), Foreign Sailors on Socotra. The inscriptions and drawings from the cave Hoq (Vergleichende Studien zu Antike und Orient, 3), Bremen (Hempen Verlag), 2012, pp. 437–446. Notices relatives aux inscriptions sudarabiques et aksūmites dans « Catalogue of inscriptions ans drawings in the cave Hoq », pp. 25–218 ; « Sudarabiques et Aksūmites à Suquṭra d’après les inscriptions de la grotte de Ḥōq », pp. 438–442 ; « Suquṭra dans les inscriptions de l’Arabie du Sud », pp. 443– 446 ; « Les Palmyréniens en Arabie du Sud », pp. 488–492. 2013 « Matériaux pour une prosopographie de l’Arabie antique : les noblesses sabéenne et ḥimyarite avant et après l’Islam », dans Christian Julien Robin et Jérémie Schiettecatte, Les préludes de l’Islam. Ruptures et continuités dans les civilisations du Proche-Orient, de l’Afrique orientale, de l’Arabie et de l’Inde à la veille de l’Islam (Orient et Méditerranée, 11), Paris (de Boccard), 2013, pp. 127–270. 2014 a « The peoples beyond the Arabian frontier in Late Antiquity : recent epigraphic discoveries and latest advances », dans Jitse H. F. Dijkstra & Greg Fisher, Inside and Out. Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity (Late Antique History and Religion, 8), Leuven (Peeters), 2014, pp. 33–79. 2014 b « Le roi ḥimyarite Thaʾrān Yuhanʿim (avant 325– v. 375). Stabilisation politique et réforme religieuse », dans Jerusalem Studies on Arabic and Islam, 41, 2014, pp. 1–95. 2015 a « L’Arabie dans le Coran. Réexamen de quelques termes à la lumière des inscriptions préislamiques », dans François Déroche, Christian Robin et Michel Zinc, Les origines du Coran, le Coran des origines (Académie des Inscriptions et Belles-Lettres, Actes de colloque), Paris (Diffusion De Boccard), 2015, pp. 27–74.

LES ÉVOLUTIONS DU CALENDRIER

359

2015 b « La Grande Église d’Abraha à Ṣanʿāʾ. Quelques remarques sur son emplacement ses dimensions et sa date », dans Interrelations between the Peoples of the Near East and Byzantium in Pre-Islamic Times, ed. Vassilios Christides (Semitica Antiqva 3), Córdoba: Oriens Academic, 2015, pp. 105–129. 2015 c « Quel judaïsme en Arabie ? », à paraître dans Christian Julien Robin (éd.), Le judaïsme de l’Arabie antique (Judaïsme antique et Origines du christianisme, 1), Turnhout (Brepols), à paraître. 2016 a « Die Kalender der Araber vor dem Islam », dans Nora Schmidt, Nora Katharina Schmid & Angelika Neuwirth (édd.), Denkraum Spätantike. Reflexionen von Antiken im Umfeld des Koran, Wiesbaden (Harrassowitz), 2016, pp. 299–386 (à paraître). 2016 b « Marib et Makka : deux pèlerinages de l’Arabie préislamique qui se tenaient à la veille de l’équinoxe de printemps », à paraître dans Graeco-Arabica (éd. V. Christides), 12, 2015. 2016 c « Tamnaʿ et Qatabān, l’état des lieux », dans Alessandro de Maigret & Christian Julien Robin (éd.), Gli scavi italofrancesi di Tamnaʿ (Repubblica dello Yemen). Rapporto finale (Orient et Méditerranée, 20), Paris (de Boccard), 2016, pp. 17– 97. Robin (Christian Julien), al-Ghabbān (ʿAlī I.) et al-Saʿīd (Saʿīd F.) 2014 « Inscriptions antiques récemment découvertes à Najrān (Arabie séoudite méridionale) : nouveaux jalons pour l’histoire de l’oasis et celle de l’écriture et de la langue arabes », dans Académie des Inscriptions et Belles-Lettres, Comptes rendus, 2014, pp. 1033–1127. Robin (Christian Julien) et Ṭayrān (Sālim) 2012 « Soixante-dix ans avant l’Islam, l’Arabie toute entière dominée par un roi chrétien », dans Académie des Inscriptions et Belles-Lettres, Comptes rendus de l’année 2012, pp. 525–553. Ryckmans (Jacques) 1973 a « Le repas rituel dans la religion sud-arabe », dans Symbolae biblicae et mesopotamicae Francisco Mario Theodoro de Liagre

360

CHRISTIAN JULIEN ROBIN

Böhl dedicatae, ediderunt M. A. Beek, A. A. Kampman, C. Nijland, J. Ryckmans, Leiden (Brill), 1973, pp. 327–334. 1973 b « Un rite d’istisqâ’ au temple sabéen de Mârib », dans Annuaire de l’Institut de Philologie et d’Histoire orientales et slaves, XX, 1968–1972, pp. 379–388. 1983 « Biblical and Old South Arabian Institutions : Some Parallels », dans Robin Bidwell and G. Rex Smith, Arabian and Islamic Studies, Articles presented to R. B. Serjeant on the occasion of his retirement from the Sir Thomas Adam’s Chair of Arabic at the University of Cambridge, London and New York (Longman), 1983, pp. 14–25. Al-Selwi (Ibrahim) 1987 Jemenitische Wörter in den Werken von al-Hamdānī und Našwān und ihre Parallelen in den semitischen Sprachen (Marburger Studien zur Afrika- und Asienkunde, Serie B : Asien, Band 10), Berlin (Verlag von Dietrich Reimer), 1987. Serjeant (R. B.) 1976 South Arabian Hunt, London (Luzac), 1976. Starcky (Jean) 1966 « Pétra et la Nabatène », dans Supplément au Dictionnaire de la Bible, VII, Pastorales—Pirot, 1966, col. 886–1017. Stern (Sacha) 2001 Calendar and Community. A History of the Jewish Calendar, 2nd Century BCE — 10th Century CE, Oxford University Press, 2001. 2012 Calendars in Antiquity. Empires, States, and Societies, Oxford University Press, 2012. Tabṣira : voir Varisco 1994 Varisco (Daniel Martin) 1994 Medieval Agriculture and Islamic Science. The Almanac of a Yemeni Sultan 5publications on the Near East, 6), Seattle and London (University of Washington Press), 1994. Yule (Paul) 2013 Late Antique Arabia. Ẓafār, Capital of Ḥimyar, edited by ... (Abhandlungen der Deutschen Orient-Gesellschaft, 29), Wiesbaden (Harrassowitz), 2013.

LES ÉVOLUTIONS DU CALENDRIER

361

Zaid (Zaydoon) & Maraqten (Mohammed) 2008 « The Peristyle Hall : remarks on the history of construction based on recent archaeological and epigraphic evidence of the AFSM expedition to the Awām temple in Mārib, Yemen », dans Proceedings of the Seminar for Arabian Studies, 38, 2008, pp. 327–340. Sigles La plupart des textes peuvent être consultés sur la Toile, à l’adresse : http://csai.humnet.unipi.it Nous donnons ici la référence qui donne la bibliographie la plus récente. ʿAbadān 1 : Müller 2010, pp. 50–54. ATHS 13 (=YM 11 296) ; 39 (YM 10 278) : Maraqten (Mohammed), Altsüdarabische Texte auf Holzstäbchen. Epigraphische und kulturhistorische Untersuchungen (Beiruter Texte und Studien, 103), Würzburg (Ergon Verlag in Kommission), 2014. Av Būsān 4 : Avanzini (Alessandra), « Problemi storici della regione di al-Ḥadāʾ nel periodo preislamico e nuove iscrizioni », dans P. Fronzaroli (éd.), Studi Yemeniti (Quaderni di semitistica, 14), Florence (Istituto di linguistica e di lingue orientali, Università di Firenze), 1985, pp. 53–115, pp. 99–101, pl. 25. Le site de Būsān (l’antique Buʾsān) se trouve à mi-chemin entre Baynūn et Ḍāf. Bar[āqish]-Int[érieur] 9 : inédit (Barāqish). B[āfaqīh]R[obin]-Yanbuq 47 : Müller 2010, p. 93. Yanbuq est une ravine qui se jette dans le wādī ʿAmaqīn. Bron-al-ʿAdan 2 : Bron 2014, pp. 203–204 et Fig. 1. al-ʿAdan est une montagne du Nihm au nord-est de Ṣanʿāʾ. Buraʿ al-Aʿlà 1 et 2 : al-Aghbarī (Fahmī ʿAlī b. ʿAlī), « Nuqūsh sabaʾiyya jadīda taḥtawī ʿalà aqdam naqsh tawḥīdī muʾarrakh », dans Raydān, 8, 2013, pp. 167–183 (en version électronique uniquement). Il s’agit ici du Buraʿ au sud de Bayḍāʾ-Ḥaṣī ( à distinguer du jabal Buraʿ, proche d’al-Ḥudayda).

362

CHRISTIAN JULIEN ROBIN

CIAS : Corpus des Inscriptions et des Antiquités Sud-arabes (Académie des Inscriptions et Belles-Lettres), Louvain (Éditions Peeters), tome I, sections 1 et 2, 1977 ; tome II, fascicules 1 et 2, 1986. — CIAS 95.11/o1 n° 2 = Ja 877, pp. 335–336. — CIAS 95.11/o2 n° 2 = CSAI II, 12 = Ja 878. CIH 46, 448, 540, 541, 671, 694 : Corpus Inscriptionum Semiticarum, IV, Inscriptiones Ḥimyariticas et Sabaeas continens, Paris (Imprimerie nationale), tomus I-III, 1899–1930. — CIH 46 = Gl 799 : Müller 2010, p. 33. — CIH 448 + Hakir 1 : Müller 2010, p. 35. — CIH 540 = Sadd Maʾrib 3 : Müller 2010, pp. 68–73 ; Darles et alii 2014, passim. — CIH 541 = Sadd Maʾrib 5 : Müller 2010, pp. 110–117 ; Darles et alii 2014, passim. — CIH 671 = Louvre 6. CSAI : Avanzini (Alessandra) (éd.), Corpus of South Arabian Inscriptions I-III. Qatabanic, Marginal Qatabanic, Awsanite Inscriptions (Philological Studies, Arabia Antica 2), Pisa (Università di Pisa, Edizioni Plus), 2004. — CSAI I, 594 = Ja 358. — CSAI II, 12 = CIAS 95.11/o2 n° 2 = Ja 878. DAI GDN 2002/20 = Sadd Maʾrib 4 : Nebes 2004 ; Darles et alii 2014, passim. DAI Ṣirwāḥ 2004-12B+2004-6+2002-83+1993/3-x, inédit : voir Nebes (Norbert), « Eine datierte nabatäisch-sabäische Bilingue aus Sirwah », dans Jemen-Report, 37 / 1, 2006, p. 10 ; le même, « Die Nabatäer in Südarabien », dans Antike Welt, 40. Jg., Heft 1 / 2009, pp. 52 et suiv. DhM 204 : Robin 2014 b, p. 42 et fig. 9 ; Prioletta 2013, n° 94, pp. 187–188 et 200 (photographie). DJE (Deutsche jemenitische Expedition) 25 : Müller 2010, pp. 44– 45. Fa[khry] 71 : Ahmed Fakhry, An Archaeological Journey to Yemen (March-May, 1947), Part II, Epigraphical Texts, by G. Ryckmans (Service des Antiquités de l’Égypte), Cairo (Government Press), 1952.

LES ÉVOLUTIONS DU CALENDRIER

363

Garb[ini]-Bayt al-Ashwal 2 : Müller 2010, p. 59. Bayt al-Ashwal est un village proche de Ẓafār, où de nombreuses pierres antiques sont remployées. Garb-Hakir 2 : Müller 2010, p. 22. Ghūl YU 100 : Hayajneh (Hani 2004), « Eine Sammlung von fragmentarischen altsüdarabischen Inschriften aus dem Jemen », dans Arabian Archaeology and Epigraphy, 15 / 1, 2004, pp. 120– 148, pp. 140–143 et Fig. 50 (p. 142). Gl[aser] 799 = CIH 46 : Müller 2010, p. 33. Gl 1533 = Ja 2855 : Höfner 1973, pp. 29–35 et Pl. XI. Gl 1541 : Müller 2010, p. 38. Gl 1572 : Höfner 1973, pp. 35–39 et Pl. XII. Gl 1593-1596 = Kh-Nūna 1 A : Nuʿmān 2012, pp. 150–151 et Fig. 106. Nūna est un village à 15 km à l’est de Maʿbar. Gr[jaznevič] 11 : Južnaja Aravija. Pamjiatniki Drevney Istorii i Kultury, 1, Moscow (Glavnaja redakzija vostočnoj literatury), p. 21, fig. 11. Ḥaṣī 5 : Müller 2010, p. 48. Ḥaṣī 8 : Müller 2010, p. 2. Höfner AF 3 = Munich 91-315336 : Müller 2010, p. 21. Ir 22, 25, 26, 70 : al-Iryānī (Muṭahhar ʿAlī), Fī ta’rīkh al-Yaman. Sharḥ wa-taʿlīq ʿalà nuqūsh lam tunshar, 34 naqshan min majmūʿat al-qāḍī ʿAlī ʿAbd Allāh al-Kuhālī, [Ṣanʿā’] (Markaz al-dirāsāt al-yamaniyya), 1973. Ist 7608 bis + Wellcome A 103664 : Ryckmans (Jacques), « L’inscription sabéenne chrétienne Istanbul 7608 bis », dans Journal of the Royal Asiatic Society, 1976, pp. 96–99 et 1 pl. ; Beeston (A. F. L.), « The South Arabian collection of the Wellcome Museum in London », dans Raydān, 3, 1980, pp. 11– 16 et fig. 1–3 ; en dernier lieu, Robin (Christian-Julien), « Joseph, dernier roi de Ḥimyar (de 522 à 525, ou une des années suivantes) », dans Jerusalem Studies in Arabic and Islam, 34, pp. 1– 124, pp. 96–100. Ja 358 = CSAI I, 594 : Jamme (Albert), Pièces épigraphiques de Ḥeid Bin ʿAqîl, la nécropole de Timnaʿ (Hagr Koḥlân) (Bibliothèque du Muséon, 30), Louvain (Institut orientaliste), 1952.

364

CHRISTIAN JULIEN ROBIN

Ja 547+546+544+545 = Sadd Maʾrib 6 : Darles et alii 2014, pp. 30–39. Ja 610, 611, 613, 615, 618, 650, 651, 653, 661, 669, 670, 671 + 788, 735, 877, 878 : Jamme (Albert), Sabaean Inscriptions from Maḥram Bilqîs (Mârib) (Publications of the American Foundation for the Study of Man, III), Baltimore (The Johns Hopkins Press), 1962. — Ja 671 + 788 : Robin 2014 b, pp. 40–41. — Ja 877 = CIAS 95.11/o1 n° 2, pp. 335–336. — Ja 878 = CSAI II, 12 = CIAS 95.11/o2 n° 2, pp. 346–347. Ja 889 : Müller 2010, p. 3 Ja 1819 = Shirjān 27 : Jamme (Albert), Miscellanées d’ancient (sic) arabe, II, Washington D.C., 1971 (brochure ronéotée), pp. 84– 85. Ja 2356 a = Shirjān 12 : Müller 2010, pp. 46–47 ; Robin 2014 b, p. 48 et Fig. 14 (p. 88). Ja 2855 : voir Gl 1533 Ja 3199 = Kh-Ḥammat al-Ḍabʿ 19 = Nājī 1 : Nuʿmān 2012, pp. 99–100 et Fig. 71. Kh[aldūn]-al-Harūj 1 : Nuʿmān 2012, p. 145 et Fig. 20. Le village d’al-Harūj se trouve à 10 km au nord de Hakir. Kh-al-Harūj 2 : Nuʿmān 2012, p. 176 et Fig. 114. Kh-Ḍāf 1 : Nuʿmān 2012, pp. 143–144 et Fig. 97. Kh-Ghawl al-ʿAjmāʾa 1 : Nuʿmān 2012, pp. 48–49 et Fig. 20. Ghawl al-ʿAjmāʾa est un lieu-dit au nord de Sanabān (bourg à 20 km à l’ouest de Radāʿ). Kh-Hakir 1 : Nuʿmān 2012, p. 89 et Fig. 53 A-B. Kh-Ḥammat al-Ḍabʿ 1 : Nuʿmān 2012, p. 52 et Fig. 25. Ḥammat al-Ḍabʿ est un lieu-dit à 3 km de Sanabān (bourg à 20 km à l’ouest de Radāʿ). Kh-Ḥammat al-Ḍabʿ 5 : Nuʿmān 2012, p. 91 et Fig. 57. Kh-Ḥammat al-Ḍabʿ 14 : Nuʿmān 2012, pp. 96–97 et Fig. 66. Kh-Ḥammat al-Ḍabʿ 15 : Nuʿmān 2012, p. 97 et Fig. 67. Kh-Ḥammat al-Ḍabʿ 16 : Nuʿmān 2012, pp. 97–98 et Fig. 68. Kh-Ḥammat al-Ḍabʿ 17 : Nuʿmān 2012, p. 98 et Fig. 69.

LES ÉVOLUTIONS DU CALENDRIER

365

Kh-Ḥammat al-Ḍabʿ 19 = Ja 3199 = Nājī 1 : Nuʿmān 2012, pp. 99–100 et Fig. 71. Kh-Ḥawarwar 1 : mention dans Nuʿmān 2012, p. 64. Kh-al-Jarasha 1 : Robin 2014 b, pp. 42–43 et fig. 10 ; Nuʿmān 2012, pp. 176–177 et fig. 115. al-Jarasha est un village à 5 km au nord du jabal al-Lasī. Kh-Jarf al-Naʿīmiyya 4 : Nuʿmān 2012, pp. 79–80 et Fig. 38. Jarf al-Naʿīmiyya est un lieu-dit à 1 km au nord de Hakir. Kh-Jarf al-Naʿīmiyya 5 : Nuʿmān 2012, pp. 80–81 et Fig. 39. Kh-Jarf al-Naʿīmiyya 6 : Nuʿmān 2012, p. 81 et Fig. 40. Kh-Jarf al-Naʿīmiyya 7 : Nuʿmān 2012, pp. 81–82 et Fig. 41. Kh-Jarf al-Naʿīmiyya 10 : Nuʿmān 2012, pp. 83–84 et Fig. 44. Kh-Jarf al-Naʿīmiyya 12 : Nuʿmān 2012, p. 85 et Fig. 46. Kh-Jarf al-Naʿīmiyya 13 : Nuʿmān 2012, pp. 85–86 et Fig. 47. Kh-Jarf al-Naʿīmiyya 14 : Nuʿmān 2012, pp. 86–87 et Fig. 48. Kh-Jarf al-Naʿīmiyya 15 : Nuʿmān 2012, p. 87 et Fig. 49. Kh-Jarf al-Naʿīmiyya 16 : Nuʿmān 2012, p. 88 et Fig. 50. Kh-Jarf al-Naʿīmiyya 18 : Nuʿmān 2012, pp. 144–145 et Fig. 98. Kh-Nūna 1 A = Gl 1593-1596 : Nuʿmān 2012, pp. 150–151 et Fig. 106. Kh-Sayya 3 : Nuʿmān 2012, p. 103 et Fig. 77. Sayya est (semble-t-il) un lieu-dit à l’est de Sanabān. Kh-Umayma 1 : Nuʿmān 2012, p. 74 et Fig. 28. Umayma est un lieu-dit à 1 km à l’est de Hakir. Kh-Umayma 2 : Nuʿmān 2012, pp. 43–44 et Fig. 11. Kh-Umayma 3 : Nuʿmān 2012, pp. 74–75 et Fig. 29. Kh-Umayma 4 : Nuʿmān 2012, pp. 132–133 et Fig. 81. Kh-Umayma 5 : Nuʿmān 2012, p. 133–134 et Fig. 82. Kh-Umayma 6 : Nuʿmān 2012, p. 134 et Fig. 83. Kh-Umayma 13 : Nuʿmān 2012, pp. 75–76 et Fig. 30. Kh-Umayma 16 : Nuʿmān 2012, p. 138 et Fig. 89. Kh-Umayma 25 : Nuʿmān 2012, p. 32 et Fig. 4 A-B. Khaldūn-ʿAbd Allāh, zabūr = Mus. Univ. Dhamār 34 : Müller 2010, p. 36.

366

CHRISTIAN JULIEN ROBIN

L026, L089, L289, L302, L312 et L319 : Drewes (Abraham J.) & Ryckmans (Jacques), Les inscriptions sudarabes sur bois dans la collection de l’Oosters Instituut conservée dans la bibliothèque universitaire de Leiden, Texte révisé et adapté par Peter Stein, Édité par Peter Stein et Harry Stroomer, Wiesbaden (Harrassowitz), 2016. Louvre 6 : Calvet (Yves) et Robin (Christian), Arabie heureuse, Arabie déserte. Les antiquités arabiques du Musée du Louvre (Notes et documents des Musées de France, 31), Paris (Réunion des Musées nationaux), 1997. Louvre 6 = CIH 671. MAFYS (Mission archéologique française au Yémen-Sud)-Ḍuraʾ 3 = RES 4069 : Müller 2010, pp. 87–88. MAFRAY (Mission archéologique française en République arabe du Yémen)-Abū Thawr 4 : Müller 2010, p. 85. Le site de Abū Thawr (l’antique Manhiyatum) se trouve à une vingtaine de kilomètres à l’ouest d’al-Bayḍāʾ (l’antique Nashqum). MAFRAY-Hajar Ṣabāḥ 1 : Robin (Christian) et Dridi (Hédi), « Deux barrages du Yémen antique », dans Académie des Inscriptions et Belles-Lettres, Comptes rendus de l’année 2004, pp. 67–121, pp. 113–114 et fig. 41 (p. 95) ; ligne 4, lire mmnn-hmy Ynʿm (d’après Kh- Ḥawarwar 1 / 2–3). Le bourg de Hajar Ṣabāḥ se trouve à une douzaine de kilomètres au sud-est de Radāʿ. MAFRAY-al-Ḥijla 1 : Müller 2010, p. 12. al-Ḥijla est un bourg de Qaran, au nord de Qāniya. MAFRAY-Sāriʿ 5 : inédit (al-Miʿsāl). Sāriʿ est le nom du village le plus proche d’al-Miʿsāl. MAFRAY-Sāriʿ 6 : Müller 2010, p. 8. al-Miʿsāl 2 : Müller 2010, pp. 25–26. al-Miʿsāl 4 : Müller 2010, pp. 19–20. al-Miʿsāl 5 : Müller 2010, pp. 28–31. al-Miʿsāl 6 : Bāfaqīh (Muḥammad ʿAbd al-Qādir), « al-Miʿsāl 6 », dans Raydān, 6, 1994, pp. 78–88 de la section arabe. al-Miʿsāl 16 = YMN 9 : Müller 2010, p. 17. al-Miʿsāl 18 = YMN 13 : Müller 2010, p. 40. M[aḥram]B[ilqīs] 2002 I-28 : Maraqten (Mohammed), « Legal documents recently discovered by AFSM at Maḥram Bilqīs, near

LES ÉVOLUTIONS DU CALENDRIER

367

Mārib, Yemen », dans Proceedings of the Seminar for Arabian Studies, 36, 2006, pp. 53–67. MB 2004 I-123 : texte inédit daté de dhu-mabkarān 411 ; il y est fait allusion dans Maraqten 2014 a, p. 100 (où une première interprétation de la date était 404). MB 2004 I-147 : texte inédit (la date est citée dans Zaid et Maraqten 2008, p. 338). M[ission]Q[atabān]-al-Ḥāṭ 1 : Müller 2010, p. 32. Le village d’alḤāṭ se trouve en amont du wādī Shirjān. MQ-Minkath 1 : Robin 2014 b, p. 43 et fig. 11. Minkath est un village voisin de Ẓafār, construit en partie avec des pierres antiques. MQ-al-Jifjif 1 : Müller 2010, p. 23. al-Jifjif est un lieu-dit dans le wādī Shirjān, en amont de Qaryat Ḥusayn. MQ-al-Qashʿa 1 : inédit. Il a été fait mention de ce texte dans Robin 2012, a pp. 126– 127. al-Qashʿa est un lieu-dit de ʿAlma, à l’est de Shirjān. MQ-Ḍarāwiʿ 2 et 3 : inédits. Il a été fait mention de ce texte dans Robin (Christian Julien), « Les banū Haṣbaḥ, princes de la commune de Maḍḥām », dans Arabia, 3, 2005–2006, pp. 31– 110 et fig. 59–67 (pp. 324–327), notamment pp. 91–92. La ravine de Ḍarāwiʿ se trouve sur les hauteurs, à l’ouest du wādī Shurjān. MQ-dhū-Wayn 1 : Müller 2010, p. 4. Le village de dhū Wayn se trouve à l’est de Bayḍāʾ-Ḥaṣī. MQ-dhū-Wayn 2 : Müller 2010, p. 5. MQ-dhū-Wayn 4 : Müller 2010, p. 6. MQ-dhū-Wayn 9 : Arbach (Mounir) & Gajda (Iwona), « La plus ancienne inscription sudarabique datée d’après une ère et autres inscriptions rupestres de la région d’al-Bayḍāʾ », dans Syria, 79, 2002, pp. 293–306, pp. 302–304, fig. 8. MQ-Ḥayd Mūsà 1 = RES 4196 : Müller 2010, p. 37. Le village de Ḥayd Mūsà se trouve dans le wādī Shirjān, en amont de Qaryat Ḥusayn. MQ-Maʾrib 1 : Robin 2014 b, pp. 44–46 et Fig. 12 (p. 86). M-SAS 1 : Maraqten 2014 b.

368

CHRISTIAN JULIEN ROBIN

Nājī 1 = Kh-Ḥammat al-Ḍabʿ 19 = Ja 3199 : Nuʿmān 2012, pp. 99–100 et Fig. 71. Le lieu-dit Ḥammat al-Ḍabʿ se trouve à 3 km à l’est de Sababān (soit à près de 20 km à l’ouest de Radāʿ). RES 3383, 4176, 4196, 4197 bis, 3958, 4069, 4646, 5085 : Répertoire d’épigraphie sémitique, publié par la Commission du Corpus Inscriptionum Semiticarum (Académie des Inscriptions et BellesLettres), Paris (Imprimerie nationale), tomes I-VIII, 1900– 1967. — RES 3383 : Müller 2010, p. 58. — RES 4197 bis : Müller 2010, p. 9. — RES 3958 : Müller 2010, pp. 14–15. — RES 4069 = MAFYS (Mission archéologique française au Yémen-Sud)-Ḍuraʾ 3 : Müller 2010, pp. 87–88. — RES 4196 = MQ-Ḥayd Mūsà 1 : Müller 2010, p. 37. — RES 5085 : Müller 2010, pp. 66–67. Sadd Maʾrib 3 = CIH 540 : Müller 2010, pp. 68–73 ; Darles et alii 2014, passim. Sadd Maʾrib 4 = DAI GDN 2002/20 : Nebes 2004 ; Darles et alii 2014, passim. Sadd Maʾrib 5 = CIH 541 : Müller 2010, pp. 110–117 ; Darles et alii 2014, passim. Sadd Maʾrib 6 = Ja 547+546+544+545 : Darles et alii 2014, pp. 30–39. Shibām Kawkabān 1 : Sholan (Amida) & Gajda (Iwona), « A new Ḥimyaritic inscription from Šibām Kawkabān dated to the year 487 of the Ḥimyaritic era », dans South Arabia and its Neighbourgs. Phenomena of Intercultural Contacts. 14. Rencontres sabéennes, éd. Iris Gerlach = Archäologische Berichte aus dem Yemen, XIV, 2015, pp. 161–169. Shirjān 12 = Ja 2356 a : Müller 2010, pp. 46–47 ; Robin 2014 b, p. 48 et Fig. 14 (p. 88). Shirjān 27 = Ja 1819 : Jamme (Albert), Miscellanées d’ancient (sic) arabe, II, Washington D.C., 1971 (brochure ronéotée). ThUM 34 : Prioletta (Alessia), Inscriptions from the Southern Highlands of Yemen. The Epigraphic Collections of the Museums of Baynūn and

LES ÉVOLUTIONS DU CALENDRIER

369

Dhamār (Arabia Antica, 8), Roma (« L’Erma » di Bretschneider), 2013, pp. 322–323 et photographie p. 328. UAM 327 : Müller 2010, p. 7. X.SBS 8, 10 A, 11, 26, 31, 35, 37 A, 46, 62, 69, 139, 152 : Stein (Peter), Die altsüdarabischen Minuskulinschriften auf Holzstäbchen aus der Bayerischen Staatsbibliothek in München. I : Die Inschriften der mittel- und spätsabäischen Periode, 2 vol., Tübingen (Ernst Wasmuth Verlag), 2010. — X.BSB 139 = Mon.script.sab. 320 : Robin 2014 b, pp. 53–54. YM 1695 : Müller 2010, p. 42–43. YM 1950 : Müller 2010, p. 57. YMN 9 = al-Miʿsāl 16 : Müller 2010, p. 17. YMN 10 : Müller 2010, p. 16. YMN 13 = al-Miʿsāl 18 : Müller 2010, p. 40. YMN 15 : Yūsuf Muḥammad ʿAbd Allāh, « Mudawwanat alnuqūsh al-yamaniyya al-qadīma. Naqsh Biʾr al-ʿAyl », dans alIklīl (Ṣanʿāʾ), vol. [6], fasc. 3–4, automne 1988, pp. 250–260. Ẓafār 08~235 : texte inédit mentionné et reproduit dans Yule 2013, Table 10.1, pp. 170–171 (« from al-Gusr »), et Fig. 10.9, 1 = [p. 180]. Ẓafār 10~002 : Robin 2014 b, pp. 41–42 et fig. 8 ; Müller (Walter W.), « Chapter 10. Three Late Sabaic Inscriptions with Royal Names from Ẓafār », dans Yule (Paul), Late Antique Arabia. Ẓafār, Capital of Ḥimyar, edited by ... (Abhandlungen der Deutschen Orient-Gesellschaft, 29), Wiesbaden (Harrassowitz), 2013, pp. 167–170 et fig. 10.14.1, 10.15.2 et 10.15.1. ẒM 1 : Müller 2010, pp. 75–76. ẒM 704 : Müller 2010, p. 41. ẒM 2263+2262+2264 : Müller 2010, p. 24. Zubayrī al-ʿAwd 1 : Müller 2010, pp. 10–11 ; Robin (Ch.), « Ḥimyarites Kings on Coinage », dans Coinage of the Caravan Kingdoms, Studies in Ancient Arabian Monetization, edited by Martin Huth and Peter G. van Alfen (Numismatics Studies, 25), New York (The American Numismatic Society), 2010, pp. 357–381, pp. 371–373. Le jabal al-ʿAwd est une haute

370

CHRISTIAN JULIEN ROBIN montagne au sud-est de Ẓafār ; on y trouvait dans l’antiquité la commune Dhubḥān. Carte 1, Le Yémen antique

LES ÉVOLUTIONS DU CALENDRIER Tableau 1

Le calendrier ḥimyarite dans les inscriptions commémorant les réfections de la Digue de Maʾrib aux 5e et 6e s. è. chr.

1. Réfection du roi Shuriḥbiʾīl Yaʿfur en 564-565 ḥim. : Sadd Maʾrib 3 = CIH 540 — [Saison des pluies d’été 564] [Rupture de la Digue de Maʾrib] — dhu-diʾāwān (XII, ≈ janvier) 564 Fin de la réfection — dhu-thābatān (III, ≈ avril) 565 Nouvelle rupture provoquée par la crue du printemps — dhu-diʾāwān (XII, ≈ janvier) 565 Fin de la réfection ; commémoration de ces travaux dans une inscription →Changement de millésime en dhu-ḥullatān (I, ≈ février) ou dhumaʿūn (II, ≈ mars) 2. Réfection du roi Abraha en 567-568 ḥim. Texte A, construction du pylône : Sadd Maʾrib 4 = DAI GDN 2002/20 — dhu-ḥullatān (I, ≈ février) [65]8 Travaux du début à la fin du mois — dhu-ḥullatān « le dernier » (I bis, ≈ février bis) 658 Commémoration des travaux le mois suivant →Changement de millésime en dhu-ḥullatān (I, ≈ février) ou auparavant Texte B, totalité des travaux : Sadd Maʾrib 5 = CIH 541 — dhu-qiyāẓān (V, ≈ juin) 657 Départ d’Abraha en campagne — dhu-madhraʾān (VI, ≈ juillet) [65]7 Rupture de la Digue de Maʾrib — dhu-ṣurābān (IX, ≈ octobre) [65]7 Début des travaux, bientôt suspendus à cause d’une épidémie — dernière semaine de dhu-diʾāwān (XII, ≈ janvier) Reprise des travaux, achevés en 58 jours — onze mois Durée de l’absence d’Araha — dhu-maʿūn (II, ≈ mars) 658 Commémoration de ces travaux dans une inscription →Changement de millésime : assurément après dhu-ṣurābān, au plus tard en dhuma ʿūn (II, ≈ mars) ; probablement en dhu-ḥullatān (I, ≈ février) ou dhu-maʿūn (II, ≈ mars) parce que le mois dhu-diʾāwān (XII, ≈ janvier) qui précède, mentionné sans année, appartient encore à 657. De dhu-qiyāẓān (V, ≈ juin) 657 à dhu-maʿūn (II, ≈ mars) 658, on ne compte que 10 mois : comme Abraha en mentionne 11, un mois intercalaire a été ajouté. Hypothèse de reconstruction chronologique de l’expédition et des travaux d’Abraha — dhu-qiyāẓān (V, ≈ juin) 657 : départ d’Abraha en campagne — dhu-madhraʾān (VI, ≈ juillet) [65]7 : rupture de la Digue de Maʾrib — dhu-ṣurābān (IX, ≈ octobre) [65]7 : début des travaux, bientôt suspendus à cause d’une épidémie

371

372

CHRISTIAN JULIEN ROBIN — dernière semaine de dhu-diʾāwān (XII, ≈ janvier) [657] : reprise des travaux — dernière semaine de dhu-diʾāwān (XII, ≈ janvier) [657], totalité du mois de dhu-ḥullatān [I] [65]8 et majeure partie de dhu-ḥullatān II 658 : travaux qui durent 58 jours au total — dhu-ḥullatān II 658 : Abraha pense arrêter les travaux après la constructions du pylône en dhu-ḥullatān [I] et commande la gravure d’une inscription ; mais il change d’avis, fait poursuivre les travaux et abandonne l’inscription inachevée dans le sable — onze mois d’absence d’Abraha de dhu-qiyāẓān (V) à dhu-maʿūn (II) ; dhu-qiyāẓān (V), dhu-madhraʾān (VI), dhu-khirāfān (VII), dhu-ʿallān (VIII), dhu-ṣurābān (IX), dhu-muhlatān (X), dhu-ālān (XI), dhu-diʾāwān (XII), dhuḥullatān (I), dhu-ḥullatān (I bis), dhu-maʿūn (II) — dhu-maʿūn (II, ≈ mars) 658 : commémoration de ces travaux dans une inscription

LES ÉVOLUTIONS DU CALENDRIER

373

Tableau 2

Le calendrier babylonien et quelques calendriers du ProcheOrient qui en dérivent (début d’année à l’équinoxe de printemps) 1. (mars-avril) {Add.} 2. (avril-mai) 3. (mai-juin) 4. (juin-juillet) 5. (juillet-août) 6. (août-sept.) 7. (sept.-oct) 8. (oct.-nov) 9. (nov.-déc.) 10. (déc.-janv.) 11. (janv-févr.) 12. (févr.-mars) {Add.}

Babylone 265 nisannu

Nabaṭ Judaïsme (Palmyre) 266 nysn nîsān (30 jours)

aiaru

ʾyr

simanu duzu

sywn tmwz (qynn)

abu

ʾb

ululu

ʾlwl

tashritu araḥsamnu

tšry — (knwn)

kislimu ṭebetu

kslw, šmrʾ (kslw, kslwl) ṭbt

shabaṭu

šbṭ

addaru {addaru I}

ʾdr

{addaru II}

iyyār (29 j.)

sîwān (30 j.) tammûz (29 j.) āv (30 j.) elûl (29 j.)

tishrî (30 j.) (mar)ḥeshwān (29 ou 30 j.) kislêw (29 ou 30 j.) ṭevet (29 j.) shvāṭ (30 j.)

{adār I, mois intercal., 30 j.) adār {adār II ou we-adār} (29 j.)

Syrie

7. nîsān 8. iyyār

9. ḥzîrān 10. tāmûz 11. āb

12. êlûl

1. teshrî(n) 2. teshrî(n) II 3. kānûn I

4. kānûn II 5. shbāṭ 6. ādār

Ḥimyar (hypoth.) dhu-ḥillatān {dhu-ḥillatān I} {dhu-ḥillatān II} dhu-maʿūn

Makka (11 h.) al-muḥarram

dhu-thābatān dhu-mabkarān

{ṣafar I} ṣafar {ṣafar II} rabīʿ al-awwal rabīʿ al-thānī

dhu-qiyāẓān

jumādà al-ūlà

dhu-madhraʾān I dhu-khirāfān dhū ʿallān

jumādà alākhira rajab shaʿbān

dhu-ṣurābān

ramaḍān

dhu-muhlatān

shawwāl

dhu-ālān

dhū ʾl-qaʿda

dhu-diʾāwān

dhū ʾl-ḥijja

Stern 2012, p. 75. Dans ce calendrier, les mois intercalaires ont d’abord été ajouté de façon empirique (en redoublant notamment ululu ou addaru, Stern 2012, p. 95), puis de manière systématique (et donc prédictive), en ne redoublant plus que addaru selon un cycle rigide. 266 Cantineau 1930–1932, 2, pp. 170–171 ; Starcky 1966, col. 940– 941. À Palmyre, ce sont les mêmes noms avec une différence (qynn à la place de tmwz) et une variante (kslw et kslwl). Dans le tableau, ces spécificités palmyréniennes (Hillers & Cussini 1996, p. 443) sont indiquées dans la colonne « Nabaṭ, Palmyre » entre parenthèses. 265