To the Madbar and Back Again 9004356126, 9789004356122, 2017040759, 9789004357617

Citation preview

To the Madbar and Back Again

Studies in Semitic Languages and Linguistics Editorial Board Aaron D. Rubin and Ahmad Al-Jallad

volume 92

The titles published in this series are listed at brill.com/ssl

Image by Annie Searight

M.C.A. Macdonald in his study

To the Madbar and Back Again Studies in the Languages, Archaeology, and Cultures of Arabia Dedicated to Michael C.A. Macdonald

Edited by

Laïla Nehmé Ahmad Al-Jallad

leiden | boston

The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov LC record available at http://lccn.loc.gov/2017040759

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface. issn 0081-8461 isbn 978-90-04-35612-2 (hardback) isbn 978-90-04-35761-7 (e-book) Copyright 2018 by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill nv incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi, Brill Sense and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, ma 01923, usa. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner.

Contents Introduction xi Publications by M.C.A. Macdonald Tabula Gratulatoria xxvii Tables and Figures xxviii Contributors xxxvii

xvii

part 1 Epigraphy and Philology 1

Les artisans et professions «libérales» dans le domaine nabatéen 3 Laïla Nehmé

2

Anmerkungen zum safaitischen und althebräischen Onomastikon 26 Walter W. Müller

3

Safaitic Prayers, Curses, Grief and More from Wadi Salhub—North-Eastern Jordan 41 Hani Hayajneh

4

Notes on ḥwb in Safaitic Chiara Della Puppa

5

Traditional Music or Religious Ritual? Ancient Rock Art Illumined by Bedouin Custom 81 Ali al-Manaser

6

The Formularies and Their Historical Implications: Two Examples from Ancient South Arabian Epigraphic Documentation 96 Alessandra Avanzini

7

Ancient South Arabian Graffiti from Shabathān (Governorate of al-Bayḍāʾ, Yemen) 116 Alessia Prioletta

69

viii

contents

8

Schreiben, meißeln, Fehler machen. Zur Funktion von Schrift im öffentlichen Raum im antiken Südarabien 154 Peter Stein

9

The Phonemes ẓ and ṭ in the Dadanitic Inscriptions 202 Fokelien Kootstra

10

Dadanitic Inscriptions from Jabal al-Khraymāt (Madāʾin Ṣāliḥ) 218 María del Carmen Hidalgo-Chacón Díez

11

Un sanctuaire de montagne: Mushannaf 238 Maurice Sartre

12

Un pasteur et un soldat? Deux inscriptions grecques d’ époque romaine à l’est du Jabal Ḥawrān 253 François Villeneuve

13

Méharistes et cavaliers romains dans le désert jordanien 270 Pierre-Louis Gatier

14

Goras, sanglier ou jeune lion (ou onagre) ? 298 Jean-Baptiste Yon

15

A Lead Syriac Protective Talisman 309 Sebastian Brock

16

Two New Arabic Inscriptions: Arabian Castles and Christianity in the Umayyad Period 327 Robert Hoyland

17

The Etymology of Ḥattā 338 Ahmad Al-Jallad

18

Are Libyco-Berber Horizontal ṯ and Vertical h the Same Sign? Marijn van Putten

346

contents

part 2 Archaeology, History and Religion 19

The Outer Wall of Taymāʾ and Its Dating to the Bronze Age 361 Arnulf Hausleiter

20

Pottery from the “Midianite Heartland”? On Tell Kheleifeh and Qurayyah Painted Ware. New Evidence from the Harvard Semitic Museum 392 Marta Luciani

21

A Caravan Merchant Family of ‘Antioch on the Chrysorhoas’. A Glimpse of Hellenistic Gerasa as a Caravanserai 439 Ina Kehrberg(-Ostrasz)

22

The Visit of Mālik bin Muʿāwiyah, King of Kindah and Maḏḥiǧ to the Himyarite King Šammar Yuharʿiš in Maʾrib 449 Mohammed Maraqten

23

Der rituelle Umzug des Yadaʿʾil Ḏarīḥ nach Ṣirwāḥ 461 Norbert Nebes

24

Sedentism of Arabs in the 8th–4th Centuries bc 479 Israel Ephʿal

25

Reflections on Arab Leadership in Late Antiquity 489 Greg Fisher

26

A Paradise in the Desert: Iram at the Intersection of One Thousand and One Nights, Quranic Exegesis, and Arabian History 522 Orhan Elmaz

27

Mourning for the Dead and the Beginning of Idolatry in the Kitāb al-Aṣnām and the Spelunca Thesaurorum—an Unknown Parallel to Sūrat at-Takāṯur (q102)? 551 Konstantin M. Klein

28

Temple Inscriptions and “the Death of the God(s)” 567 John F. Healey

ix

x 29

contents

‘The Conception of Jesus’ 581 Hannah M. Cotton Paltiel

part 3 Modern Dialects and Tribes 30

Drink Long and Drink in Peace: Singing to Livestock at Water in Dhofar, Sultanate of Oman 601 Miranda J. Morris and Sālim ʕAwaḏ̣ Ahmad al-Shaḥri

31

South Arabian Sibilants and the Śḥerɛ̄t s̃ ~ š Contrast Alex Bellem and Janet C.E. Watson

32

Was There a “Bedouinisation” of Arabia? Probably Not, at Least in the Way It Has So Far been Portrayed 645 William Lancaster and Fidelity Lancaster Index

713

622

Michael Macdonald Michael C.A. Macdonald’s name is virtually synonymous with Ancient North Arabia. His career, spanning over forty years, pioneered the field of Ancient North Arabian epigraphy, monumentalized, in 2017, by the publication of the Online Corpus of the Inscriptions of Ancient North Arabia (ociana) under his direction. ociana is however the apotheosis of many projects to which Michael contributed or, most often, initiated and directed. The first was the Corpus of the Inscriptions of Jordan Project, between 1980 and 1983, under the auspices of the Centre for Jordanian Studies at Yarmouk University. Aware that, because of the costs of publishing photographs in books, the vast majority of Ancient North Arabian inscriptions had been published from hand-copies, he searched out thousands of unpublished photographs of inscriptions all over the world and created a large photographic archive for the project at Yarmouk University. He combined this with surveys at Jawa, Jabal Qurma, and other sites, in which he initiated a systematic recording methodology which had never before been used in epigraphic expeditions. Instead of travelling from landmark to landmark, Michael systematically marked out specific areas of the basalt desert at Jawa, for instance, and lines of team-members then ‘field-walked’ them, recording and photographing all the material—inscriptions, rock drawings and wusūm—and the archaeological contexts in which they appeared. This systematic approach quickly showed that inscriptions and drawings were not just concentrated at cairns or other landmarks, as had previously been suggested, but could be found on any stone suitable for inscribing among the billions covering the desert floor. In the 1980s he trained his friend and colleague, the late Geraldine King, in all aspects of Ancient North Arabian epigraphy and supported and encouraged her surveys in the Ḥismā desert of southern Jordan, the results of which she used in her doctoral thesis of 1990. This was the first complete description of the Hismaic (formerly ‘Thamudic e’) script, language, and epigraphic formulae, establishing it as a distinct script and dialect rather than simply part of the ‘Thamudic’ Restklassenbildung, or ‘pending-file’, as Michael has called it. A few years later, in 1994, he realised, much earlier than many other scholars, the need to introduce the computer into the field of epigraphic studies and he launched the Safaitic Database Project, obtaining for it a three year grant from the Leverhulme Trust. This database was one of the first of its kind in the field of Semitic epigraphy. With the help of two colleagues to create the structure

xii

michael macdonald

and to enter large numbers of texts, this database eventually contained some 20,000 inscriptions and made it possible to build concordances and indexes which, at that time, was not as easy a task as it is now. The Safaitic Database inspired and formed the basis of ociana. It was enriched by the results of the Safaitic Epigraphic Survey Programme (sesp), which Michael started in 1995 with the aim of locating the sites where Safaitic inscriptions had been found in the nineteenth and early twentieth centuries and for which there were only vague geographical data and handcopies of the texts. The sesp located almost all of these sites, recording their positions and photographing as many of the inscriptions as possible. Between 1996 and 2003, the sesp concentrated on the extremely rich site of Al-ʿĪsāwī, which had been visited in 1904 by the Princeton University Archaeological Expeditions to Syria. In 2012, Michael trained Ahmad Al-Jallad, as he did Geraldine King, in all aspects of Arabian epigraphy. With this knowledge, Ahmad began offering courses in the epigraphy of North Arabia at Leiden University, establishing the first university program with the subject on the regular curriculum. Michael plays an active role in the institutionalization of Ancient North Arabian at Leiden University, advising Ph.D. and ma students and frequently visiting to give lectures and meet with students. Michael’s willingness to undertake fieldwork did not end with the sesp. In the last few years, starting in 2013, he set up the Saudi-British-German ‘Epigraphy and Landscape in the Hinterland of Taymāʾ Project’. This is an interdisciplinary study charting the distribution of the thousands of inscriptions, in many different scripts, in the desert surrounding the oasis of Taymāʾ and comparing this with the hydrology and archaeology of these places in order to try to identify the ancient routes to and from the oasis. It has so far had two seasons, in 2013 and 2015, in which thousands of inscriptions and rock drawings were recorded, and another is being prepared at the end of 2017. Finally, by 2015, the work on ociana showed that many interesting inscriptions from the basalt desert of north-eastern Jordan had been published by different scholars as coming from widely different locations, and often with poor photographs. At the suggestion of his ociana colleague, Ali al-Manaser, and with his extensive help, Michael set up the Badia Epigraphic Survey which systematically searched the areas in four regions in northern Jordan, photographing 10,000 inscriptions with their gps positions. This established the true location of almost all the previously published texts as well as recording many thousand new ones. All this information was entered into ociana in the following year. A second season, led by Ali al-Manaser took place in 2017 with equal success.

michael macdonald

xiii

All these projects, and there are many others, may give the impression that Michael spent most of his time as a research manager, but this is as far from true as it can be and it is high time to talk about his contributions to scholarship. In a short introduction such as this, it is impossible to shine light on all of Michael’s scholarly achievements. His uncanny knowledge of the epigraphy of Syria, Jordan and North Arabia and of more than two thousand years of Near Eastern history, combined with a careful historiographical approach, enabled him to make deep and lasting contributions to the discussion of the cultures of pre-Islamic Arabia, with a focus on the Nabataeans, the nomads and the Arabs in general, the linguistic history of Arabia, as well as writing, literacy, and identity in Ancient Arabia in general. We will highlight some of our favorite works in the following paragraphs, and their significance to the field.

The Interpretation of the Ancient North Arabian inscriptions Michael is a paragon of discretion. He pioneered the methodology of interpreting the terse language of the inscriptions of Ancient North Arabia, going beyond the Arabic dictionaries, contextualizing them within the Ancient Near Eastern world while at the same time bringing them into conversation with modern ethnographic works. This method put an end to the free-for-all approach to interpretation. Michael’s multifaceted approach led to several important milestones in the reconstruction of the lifeways of the ancient Nomads of the Harrah. His 1992 article The Seasons and Transhumance in the Safaitic Inscriptions brilliantly mapped out the seasonal migrations of the ancient nomads through a careful comparison with the migratory patterns of the Rwala Bedouin of the early 20th century, as documented by A. Musil. At other times, ancient sources were the appropriate key to unlocking the meaning of Safaitic texts. In 2014, Michael’s ‘Romans Go Home’ provided a convincing interpretation of a word that had vexed those who studied the inscriptions for decades, ngy. The verb had usually been interpreted as ‘to escape’, which led to many inscriptions dated to the year leaders, kings, and even Caesar himself had “escaped”, a rather unlikely formulation. Comparing the verb to its Ancient South Arabian cognate, ngw ‘to announce’ and understanding the relationship between the nomads and the settled folk based on literary sources, Michael suggested an ingenious solution: ngy meant ‘to announce’, ‘to declare’, and such inscriptions were dated to the year certain men were ‘announced’ commanders of military units. Finally, the phrase snt ngy qṣr h-mdnt could be interpreted meaningfully—it was not the year Caesar escaped the province but rather the year Caesar announced the province, that is, Arabia adquisita (ad 111).

xiv

michael macdonald

Literacy, Languages, and Scripts One of Michael’s most widely known articles is his indispensable Reflections on the linguistic map of pre-Islamic Arabia (2000). In this contribution, he establishes the modern nomenclature for the scripts of Arabia, including Aramaic. Carefully separating language and script, Michael emphasizes that the inscriptions of Ancient North Arabia are not to be understood as Classical Arabic, but rather a sister dialect bundle, probably mutually intelligible with Classical Arabic but distinct in many ways. This work is a constant entry in the bibliography of works on the linguistic situation of pre-Islamic Arabia. One of the hidden gems of this paper is Michael’s suggestion that the orthographic practice of writing the l of the Arabic definite article reflected a phonetic reality, something he labelled the “northern Old Arabic isogloss”. This article was followed by a masterful overview of the grammatical features of the Ancient North Arabian inscriptions (2004), which remains the standard comparative work on the subject. Michael’s attention to the scripts and their phonetics is best represented by his abcs and Letter Order in Ancient North Arabian (1986), where, based on the known Abcedaries, Michael extracted important information about how writing was learned in the desert and to what degree one may come to conclusions regarding the phonetics of a script based on those orders. His views on literacy in the desert and the genre of the Ancient North Arabian inscriptions are beautifully presented in Literacy in an oral Environment (2005). Through close examination of the contents of the North Arabian inscriptions and their place in the landscape, this paper fully elaborates a theory on the purpose and use of writing among the nomads—writing was acquired from contact with settled areas but spread casually among the nomads as a pastime while tending to the herds. In 2015, Michael made an indispensable contribution to the paleographic study of the inscriptions of Arabia in his On the uses of writing in ancient Arabia and the role of palaeography in studying them. This article concretely defines what ‘formal’ and ‘informal’ should refer to in the production of inscriptions and sets out the clear possibilities and limitations of standard paleographic methods when applied to the informal contexts of the North Arabian inscriptions.

michael macdonald

xv

The Nabataeans, the Nabataean Script and the Origins of the Arabic Script This field is probably the one to which Michael devoted the largest amount of time after Ancient North Arabian epigraphy, and out of the many works of his dealing with it, including many editions of new inscriptions, four of them deserve a special mention because they have contributed to real advances in the field. The first chronologically is—surprisingly—a review, but one of these long review articles which go on being useful long after they have been published. This is the review of A. Negev’s Personal Names in the Nabataean Realm (1999), in which Michael not only suggested a number of corrections, as was expected, but also took the opportunity to present his ideas on the use of onomastics in Semitic epigraphy and particularly his leitmotiv that personal names do not provide a suitable material for historical and linguistic studies and are not a distinctive indicator of ethnic identity. The second is the careful and detailed historiographic study of the discovery, identification and decipherment of the Nabataean script, which is included in the publication he and Norman Lewis made, in 2003, of W.J. Bankes’ copies of inscriptions in various scripts, a model of its kind. The same year, Michael published, in the Cincinnati exhibition catalogue, the most detailed and still unrivalled analysis of the Nabataean script, in which he demonstrated, with the support of as many photographs and copies of inscriptions as the editors would allow him to include, that “Nabataean written documents do not form a coherent, homogeneous corpus and [that] it is misleading to assume that a feature in a text from one area is typical of ‘Nabataean’ as a whole”. The regionalisation of the Nabataean script is an important issue which has not been adequately dealt with yet and Michael’s study almost fifteen years ago is a still a landmark in the research landscape. His work on the Nabataean script finds an echo in his most recent approaches of the scripts from the Taymāʾ region and particularly in the definition he is working on of a ‘Taymāʾ Aramaic’ script distinct from both Aramaic and Nabataean. Finally, but this list is in no way exhaustive, the contribution he offered on the occasion of the sudden and untimely death of our colleague A. Sima, he demonstrated, in detail and with the great accuracy he is familiar with, the direct filiation between the Nabataean and Arabic scripts. The palaeographical discussion which accompanies the edition of a previously unpublished text dated to ad275/276, confirms that the Arabic script “in all its forms, simply represents the later phases of the Nabataean script, i.e. in terms of graphic development there are not two scripts—Nabataean and Arabic—but simply one script, developing over centuries and used first to write the Aramaic language, then sometimes Aramaic sometimes Arabic, then Arabic”. This is very

xvi

michael macdonald

true, and this important and in a way provocative statement set the foundation for future research in the field.

The Hellenistic and Roman Ḥawrān Viewed from the Nomads’ Point of View Michael’s contribution to the history of the Ḥawrān of southern Syria and northern Jordan includes many contributions which concern the areas of Jawa, Namāra and the Syrian Ḥarra east of Suwaydāʾ in general. However, his fundamental work is indisputably his Nomads and the Ḥawrān in the Late Hellenistic and Roman Periods which, when published in 1993, was, to use a French expression ‘un pavé dans la mare’, because it called into question many assumptions which he took the time to clarify one after the other: on the existence of socalled ‘Safaitic tribes’, on the alleged sedentary character of the people who wrote the Safaitic inscriptions, on their raiding activities as opposed to banditry, on their subsistence economy, based on pasturing and not on agriculture, on their relations with the Romans, etc. Much of what he wrote in this over 100 pages article now seems obvious, but when placed into the context of the early 1990s and the state of our knowledge on the nomads and their literacy at that time, it was visionary. All who know Michael’s work will recognize that the foregone paragraphs fail to do justice to his invaluable contributions and high scholarly achievements. And all who know the man will attest to the fact that his scholarship is matched, if not even eclipsed, by his personal integrity. Michael is the embodiment of the ideals of scholarship—science for the sake of science. His openness and kindness are responsible for the growth of Ancient North Arabian into a field in its own right, welcoming students into his wonderful house and library, teaching patiently, and always ready to share his vast knowledge with all who wish to learn. Michael is a role-model in all respects. It is a great honor to present this book to a great scholar and a dear friend in celebration of his seventieth year. Ahmad Al-Jallad and Laïla Nehmé October 2017

Publications by M.C.A. Macdonald Books 2010. (ed.) The Development of Arabic as a Written Language. Papers from the Special Session of the Seminar for Arabian Studies held on 24 July, 2009. (Supplement to the Proceedings of the Seminar for Arabian Studies 40). Oxford: Archaeopress. 2009. Literacy and Identity in Pre-Islamic Arabia. (Variorum Collected Studies 906). Farnham: Ashgate. 2005. With C.S. Phillips (ed.). A.F.L. Beeston at the Arabian Seminar, and other Papers. Oxford: Archaeopress.

Articles 2017. How much can we know about language and literacy in Roman Judaea? A Review of M.O. Wise, Language and Literacy in Roman Judaea. A Study of the Bar Kokhba Documents. (The Anchor Yale Bible Reference Library). New Haven, ct: Yale University Press, 2015. Journal of Roman Archaeology 30: 836–846. 2016. “Three Dimensions in Two: Convention and Experiment in the Rock Art of Ancient North Arabia.” In The Archaeology of North Arabia: Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5–8, 2013, edited by M. Luciani. (orea Series 4). Vienna: Austrian Academy of Sciences. 2015 [2016]. With A. Al-Jallad. A few Notes on the Alleged Occurrence of the Group Name ‘Ghassān’ in a Safaitic inscription.” Archiv für Orientforschung 53: 152–157. 2016. With B. Overlaet and P. Stein. “An Aramaic-Hasaitic Bilingual Inscription from a Monumental Tomb at Mleiha, Sharjah, u.a.e.” Arabian Archaeology and Epigraphy 27: 127–142. 2015. “On the Uses of Writing in Ancient Arabia and the Role of Palaeography in Studying them.” Arabian Epigraphic Notes 1: 1–50. 2015. “Was there a ‘Bedouinization of Arabia’?”. In The Arab East and the Bedouin Component: From Late Antiquity to the Ottoman Period, edited by K. Franz, J. Büssow, and S. Leder. Der Islam 92 (1): 42–84. 2015. With contributions by A. Corcella, T. Daryaee, G. Fisher, M. Gibbs, A. Lewin, D. Violante, and C. Whately. “Arabs and Empires before the Sixth Century (Including a new edition of the Ruwāfa Inscriptions).” Chapter 1 in Arabs and Empires Before Islam, edited by G. Fisher, 11–56, 88–89. Oxford: Oxford University Press.

xviii

publications by m.c.a. macdonald

2015. “The Emergence of Arabic as a Written Language.” Chapter 7 in ibidem, 395–417. 2014. “‘Romans go home’? Rome and other ‘Outsiders’ as viewed from the Syro-Arabian Desert.” In Inside and Out. Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, edited by J.H.F. Dijkstra and G. Fisher, 145– 163. (Late Antique History and Religion 8). Louvain: Peeters. 2014. With A.J. Nabulsi, P.-L. Gatier, and S. Timm. “Epigraphic Diversity in the Cemetery at Khirbet es-Samrāʾ.” Palestine Exploration Quarterly 38: 149–161. 2013. With A.J. Drewes, T.F.G. Higham, and C. Bronk Ramsey. “Some Absolute Dates for the Development of the Ancient South Arabian Minuscule Script.” Arabian Archaeology and Epigraphy 24: 196–207. 2012. “Goddesses, Dancing Girls or Cheerleaders? Perceptions of the Divine and the Female Form in the Rock Art of pre-Islamic North Arabia.” In Dieux et déesses d’Arabie. Images et représentations. Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 octobre 2007, edited by I. Sachet and Ch.J. Robin, 261–297. (Orient et Méditerranée 7). Paris: De Boccard. 2012. “Inscriptions, Rock Drawings and wusūm from the Ayl to Ras an-Naqab Archaeological Survey.” In The Ayl to Ras an-Naqab Archaeological Survey, Southern Jordan (2005–2007), by B. MacDonald, L.G. Herr, D.S. Qaintance, G.A. Clark, and M.C.A. Macdonald, 433–465. (American Schools of Oriental Research Archaeological Reports 16). Boston, ma: American Schools of Oriental Research. 2012. “Wheeled Vehicles in the Rock Art of Arabia.” Chapter 12 in The Arabian Horse. Origin, Development and History, edited by M. Khan, 356–395. Riyadh: Layan Cultural Foundation. 2012. With S. Brock, S. Canby, O. al-Ghul, and R.G. Hoyland. “The Semitic Inscriptions.” Chapter c.12 in The Sanctuary of Lot at Deir ʿAin ʿAbata in Jordan. Excavations 1988– 2003, edited by K.D. Politis, 417–419. Amman: Jordan Distribution Agency. 2010 [2011]. “The ‘Abiel’ Coins of Eastern Arabia: A Study of the Aramaic Legends.” In Coinage of the Caravan Kingdoms, edited by M. Huth and P. van Alfen, 403–547. (Numismatic Studies 25). New York: American Numismatic Society. 2010. “Ancient Arabia and the Written Word.” In M.C.A. Macdonald (ed.) 2010 [book section], 5–28. 2010. “The Old Arabic Graffito at Jabal Usays: A New Reading of Line 1.” In M.C.A. Macdonald (ed.) 2010 [book section], 141–143. 2010. “In Memoriam Geraldine King.”Proceedings of the Seminar for Arabian Studies 40: xi–xii. 2009. “Arabs, Arabias, and Arabic before Late Antiquity.” Topoi 16: 277–332. 2009. “arna Nab 17 and the Transition from the Nabataean to the Arabic Script.” In Philologisches und Historisches zwischen Anatolien und Sokotra. Analecta Semitica in Memoriam Alexander Sima, edited by W. Arnold, M. Jursa, W.W. Müller, and S. Procházka, 207–240. Wiesbaden: Harrassowitz.

publications by m.c.a. macdonald

xix

2009. With M. al-Najem. “A New Nabataean Inscription from Taymāʾ.” Arabian Archaeology and Epigraphy 20: 208–217. 2009. “The Decline of the ‘Epigraphic Habit’ in Late Antique Arabia: Some Questions.” In L’Arabie à la veille de l’Islam, edited by J. Schiettecatte and C.J. Robin, 17–27. (Orient & Méditerranée 3). Paris: De Boccard. 2009. “Transformation and Continuity at al-Namārah. Camps, Settlements, Forts, and Tombs.” In Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham. Proceedings of the International Conference held at Damascus, 5–9 November 2006, edited by K. Bartl and ʿA. Moaz, 317–332. (Orient-Archäologie 24). Rahden/Westf.: Leidorf. 2009. “Wheels in a Land of Camels: Another Look at the Chariot in Arabia.” Arabian Archaeology and Epigraphy 20: 156–184. 2009. “A Note on New Readings in Line 1 of the Old Arabic Graffito at Jabal Says.” Semitica et Classica 2: 223. 2009. “A Note on the Inscription below the Drawing of a Horseman.” Pages 372–373 in B. Jacobs and M.C.A. Macdonald, “Felszeichnung eines Reiters aus der Umgebung von Taymāʾ.” Zeitschrift für Orient-Archäologie 2: 364–376. 2008. “The Phoenix of Phoinikēia. Alphabetic Reincarnation in Arabia.” In The Disappearance of Writing Systems: Perspectives on Literacy and Communication, edited by J. Baines, J. Bennet, and S. Houston, 207–229. London: Equinox. 2008. “Old Arabic (Epigraphic).” In Encyclopedia of Arabic Language and Linguistics, iii, edited by K. Versteeg, 464–477. Leiden: Brill. 2008. “Les travaux de René Dussaud dans les régions désertiques de la Syrie du Sud.” In Archéologie, voix de la gloire. Pionniers et protagonistes de l’archéologie Syrienne. Catalogue de l’Exposition, Damas 2008. 2008. With N. Nebes. “Arabien.” In Der Grosse Ploetz. Die Enzyklopädie der Weltgeschichte, 35., völlig neu bearbeitete, Auflage, 117–120, 330–333. Freiburg im Breisigau: Vandenhoeck & Ruprecht. 2006 [2008]. “Death Between the Desert and the Sown. Cave-tombs and Inscriptions near Dayr al-Kahf in Jordan.” Damaszener Mitteilungen 15: 273–301. 2006. “Foreword.” In L’arabo in epoca preislamica: formazione di una lingua, by D. Mascitelli, 11. (Arabia Antica 4). Roma: ‘L’Erma’ di Bretschneider. 2006. “Introductory Remarks [to Papers given on the Beeston Day at the Seminar for Arabian Studies, at the British Museum, on 22nd July 2005].” Proceedings of the Seminar for Arabian Studies 36: 3–4. 2005–2006. “Note on Dadanitic Inscription 12a from Madāʾin Ṣāliḥ.” Page 189 in L. Nehmé, “Inscriptions nabatéennes vues et revues à Madāʾin Ṣāliḥ.” Arabia 3: 179– 225, fig. 117–155 (pp. 345–356). 2005. “Literacy in an Oral Environment.” In Writing and Ancient Near Eastern Society. Papers in Honour of Alan R. Millard, edited by P. Bienkowski, C.B. Mee, and

xx

publications by m.c.a. macdonald

E.A. Slater, 49–118. (Journal for the Study of the Old Testament. Supplement Series 426). Harrisburg, pa: T. & T. Clark. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 2005. “The Safaitic Inscriptions at Dura Europos.” In A Journey to Palmyra. Collected Essays to remember Delbert R. Hillers, edited by E. Cussini, 118–129. (Culture and History of the Ancient Near East 22). Leiden-Boston: Brill. 2005. “Of Rock-Art, ‘desert kites’ and meṣāyid.” In Arabia Vitalis: Arabskiĭ Vostok, islam, drevnyaya Araviya: Sbornik Naychnykh stateĭ, posvyashchennyĭ 60-letiyu V.V. Naumkina, edited by A.V. Sedov and I.M. Smulyanskaya, 332–345. Moscow: Rossiĭskaya Akademiya Nauk. 2005. “Kilroy on the Computer: A Database of Desert Graffiti.” Newsletter of the Council for British Research in the Levant 2005: 65–66. 2005. “A Revised Bibliography of Professor A.F.L. Beeston.” In A.F.L. Beeston at the Arabian Seminar and other papers, including a personal reminiscence by W.W. Müller, edited by M.C.A. Macdonald and C.S. Phillips, 13–36. Oxford: Archaeopress. 2004. “Ancient North Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R.D. Woodard, 488–533. Cambridge: Cambridge University Press. 2004. “Beeston, Alfred Felix Landon (1911–1995).” In Oxford Dictionary of National Biography. Oxford: Oxford University Press. Online edition, http://ezproxy-prd.bodleian .ox.ac.uk:2167/view/article/60459, retrieved September 5th, 2016. 2004. “The Safaitic Inscriptions at Dura-Euopos.” In Doura-Europos: Études v, 1994–1997, edited by P. Leriche, M. Gelin, and A. Dandrau, 119–126. (Publication de la Mission franco-syrienne de Doura-Europos). Paris: Geuthner. 2003. “Languages, Scripts, and the Uses of Writing among the Nabataeans.” In Petra Rediscovered: Lost City of the Nabataeans, edited by G. Markoe, 36–56, 264–266 (endnotes), 274–282 (references). New York-Cincinnati, oh: Abrams and Cincinnati Art Museum. 2003 [2006]. With N.N. Lewis. “W.J. Bankes and the Identification of the Nabataean Script.” With appendices by S. Clackson, R.G. Hoyland, and M. Sartre. Syria 80: 41– 110. 2003. “‘Les Arabes en Syrie’ or ‘La pénétration des Arabes en Syrie’. A question of perceptions.” In La Syrie hellénistique, 303–318 (Topoi Supplément 4). Paris: De Boccard. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 2003. “A Fragmentary Safaitic Inscription from Sīʿ 8.” In Hauran ii: Les installations de Sīʿ 8. Du sanctuaire à l’établissement viticole, edited by J. Dentzer-Feydy, J.-M. Dentzer, and P.-M. Blanc, 277, pl. 182/4–7 (Bibliothèque Archéologique et Historique 164). Beyrouth: Institut Français d’Archéologie du Proche-Orient. 2003. “References to Sīʿ in the Safaitic Inscriptions.” In ibidem, 278–280.

publications by m.c.a. macdonald

xxi

2002. Review of F. Bron, Inventaire des inscriptions sudarabiques, 3: Maʿīn. Paris-Rome, 1998. Journal of Semitic Studies 47: 121–126. 2002. “113. An Inscribed Bronze Figurine of a Camel with an Unusual Saddle.” In Queen of Sheba. Treasures from Ancient Yemen. Catalogue of an exhibition held at the British Museum in 2002, edited by St.J. Simpson, 99. London: British Museum Press. 2001. “Arabi, Arabie e Greci.” Forme di contatto e percezione. In I Greci. Storia Cultura Arte Società. Vol. 3. I Greci oltre la Grecia, edited by S. Settis, 231–266. Torino: Einaudi. A revised English version is published in Macdonald 2009, see book section. 2000. With L. Nehmé. “Al-ʿUzzā.” In Encyclopaedia of Islam, New Edition, vol. x, 967– 968. Leiden: Brill. 2000. “Reflections on the Linguistic Map of Pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 28–79. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 2000. Articles on “Arabia”, “Arabs”, “Camel”, “Dedan”, “Duma”, “Sheba”, “Tayma”. In British Museum Dictionary of the Ancient Near East, edited by P. Bienkowski and A. Millard. London: British Museum Press. 2000. With G.M.H. King. “Thamudic.” In Encyclopaedia of Islam, New Edition, vol. x, 436–438. Leiden: Brill. 1999. “Personal Names in the Nabataean Realm. A Review Article.” Journal of Semitic Studies 44: 251–289. 1998. “Some Reflections on Epigraphy and Ethnicity in the Roman Near East.” In Identities in the Eastern Mediterranean in Antiquity, edited by G. Clarke and D. Harrison. Proceedings of a Conference held at the Humanities Research Centre in Canberra 10–12 November 1997. Mediterranean Archaeology 11: 177–190, pl. 16.1. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 1998. “North Arabian Languages and Scripts.” Version Originale 7: 334–337. 1998. “The Image of the North Arabian Nomad: from Assyrian Annals to Byzantine Chronicles.” [Repeated from the 1997 issue]. Bulletin of the Society for Arabian Studies 3: 8. 1997. “Trade Routes and Trade Goods at the Northern End of the ‘Incense Road’ in the First Millennium b.c.” In Profumi d’Arabia. Atti del Convegno, edited by A. Avanzini, 333–349. (Saggi di Storia Antica 11). Roma: ‘L’Erma’ di Bretschneider. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 1997. “A Bibliography of Professor A.F.L. Beeston.” Proceedings of the Seminar for Arabian Studies 27: 21–48. 1997. “The Namara Rescue Survey 1996.” British Institute at Amman News 4: 6–7. 1997. “The Joint Lecture [The Image of the North Arabian Nomad: from Assyrian Annals to Byzantine Chronicles].” British Institute at Amman News 4: 8–9. 1997. “The Image of the North Arabian Nomad from Assyrian Annals to Byzantine Chronicles.” Bulletin of the Society for Arabian Studies 2: 4.

xxii

publications by m.c.a. macdonald

1996. With M. Muʾazzin, and L. Nehmé. “Les inscriptions safaïtiques de Syrie, 140 ans après leur découverte.” Comptes Rendus de l’Académie des Inscriptions et BellesLettres, 435–492. 1996. “Appendice i. Nabataean Inscriptions copied by W.J. Bankes in the Ḥawrān.” Pages 97–98 in N.N. Lewis, A. Sartre-Fauriat, and M. Sartre, “Travaux en Syrie d’un voyageur oublié.” Syria 73: 57–100. 1996. “Hunting, Fighting and Raiding: The Horse in Pre-Islamic Arabia.” In Furusiya: The Horse in the Art of the Near East, edited by D.G. Alexander, 72–83. Vol. 1. Riyadh: King Abdulaziz Public library. 1996. With A. McQuitty and A.C. Goguel. “In Memoriam Antoni Ostrasz (1929–1996).” Annual of the Department of Antiquities of Jordan 40: 9–10. 1996. “In Memoriam Professor A.F.L. Beeston.” Proceedings of the Seminar for Arabian Studies 26: vi–vii. 1996. “Obituary of Professor A.F.L. Beeston.” Bulletin of the Society for Arabian Studies 1: 8–9. 1996. “Report on the Safaitic Database Project.”Bulletin of the Society for Arabian Studies 1: 10. 1996. “Report on the Safaitic Database Project.” Levant 28: 212–213. 1996. With M. Muʾazzin and L. Nehmé. “Report on the First Season of the Safaitic Epigraphic Survey Programme.” Levant 28: 215. 1996. “The Safaitic Epigraphic Survey.” British Institute at Amman News 3: 4. 1996. “The Safaitic Database Project.” British Institute at Amman News 3: 10–11. 1996. With L. Nehmé. “Une bande dessinée vieille de 2000 ans.” Le Monde de la Bible 98: 43. 1995. “Quelques réflexions sur les Saracènes, l’inscription de Rawwāfa et l’armée romaine.” In 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, edited by H. Lozachmeur, 93–101. Paris: Éditions Recherche sur les Civilisations. A revised English version is published in Macdonald 2009, see book section. 1995. “North Arabia in the First Millennium bce.” In Civilizations of the Ancient Near East, vol. ii, edited by J.M. Sasson, 1355–1369. New York: Scribners. 1995. “Wabar”. Page 1351 in R. Boucharlat, “Archaeology and Artifacts of the Arabian Peninsula.” In ibidem, 1335–1353. New York: Scribners. 1995. “Herodian Echoes in the Syrian Desert.” In Trade, Contact, and the Movement of Peoples in the Eastern Mediterranean: Papers in Honour of J. Basil Hennessy, edited by S. Bourke and J.-P. Descoeudres, 285–290. (Mediterranean Archaeology Supplement 3). Sydney: Mediterranean Archaeology. 1995. “Ṣafaitic.” In Encyclopaedia of Islam, New Edition, vol. viii, 760–762. Leiden: Brill.

publications by m.c.a. macdonald

xxiii

1995. “Report on the Safaitic Database Project.” The Society for Arabian Studies Newsletter 6: 8. 1995. With L. Nehmé. “The Safaitic Database Project.” Orient Express 1995 (1): 16–17. 1995. With M. Muʾazzin and L. Nehmé. “Report on the First Season of the Safaitic Epigraphic Survey Programme.” Orient Express 1995 (3): 88–89. 1995. “Report on the First Season of the Safaitic Epigraphic Survey Programme.” British Institute at Amman News 2: 11. 1994. “A Dated Nabataean Inscription from Southern Arabia.” In Arabia Felix. Beiträge zur Sprache und Kultur des vorislamischen Arabien. Festschrift Walter W. Müller zum 60. Geburtstag, edited by N. Nebes, 132–141. Wiesbaden: Harrassowitz. 1994. With F. Braemer. “Al-Ṣafā.” In Encyclopaedia of Islam, New Edition, vol. viii, 756– 757. Leiden: Brill. 1994. “Iscrizioni thamudee e nabatee.” In Umm al-Rasas Mayfaʿah I. Gli scavi del complesso di Santo Stefano, edited by M. Piccirillo and E. Alliata, 276–278, photo 102 (p. 105) and pl. xxiv, 3, 4 (p. 321). (Studium Biblicum Franciscanum Collectio Maior 28). Jerusalem: Studium Biblicum Franciscanum. 1994. Review of R. Burns, Monuments of Syria: An Historical Guide. London: I.B. Tauris, 1992. Bulletin of the School of Oriental and African Studies 57: 375–376. 1994. “William and Fidelity Lancaster.” British Institute at Amman News 1: 4. 1994. “Extraordinary General Meeting & Reception for Jill Spedding.” British Institute at Amman News 1: 6–7. 1993. “Nomads and the Ḥawrān in the Late Hellenistic and Roman Periods: A Reassessment of the Epigraphic Evidence.” Syria 70: 303–413. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 1992. “The Seasons and Transhumance in the Safaitic Inscriptions.” Journal of the Royal Asiatic Society, 3rd. series, 2: 1–11. 1992. “North Arabian Epigraphic Notes, i.” Arabian Archaeology and Epigraphy 3: 23– 43. 1992. “The Distribution of Safaitic Inscriptions in Northern Jordan.” In Studies in the History and Archaeology of Jordan iv, edited by M. Zaghloul, Kh. ʿAmr, F. Zayadine, R. Nabeel, and N.R. Tawfiq, 303–307. Amman: Department of Antiquities. 1992. “On the Placing of ṣ in the Maghribi abjad and the Khirbet al-Samrāʾ abc.” Journal of Semitic Studies 37: 155–166. 1992. “Inscriptions, Safaitic.” In The Anchor Bible Dictionary, Volume 3, edited by D.N. Freedman, 418–423. New York: Doubleday. 1992. “A Note on the Inscriptions of Fayrān.” Pages 19–20 in P. Grossmann and A. Reichert, “Report on the Season in Fayrān (March 1990).” Göttinger Miszellen 128: 7–20. 1991. “hu 501 and the Use of s³ in Taymanite.” Journal of Semitic Studies 36: 11–36. 1991. “Thamudic and Nabataean Inscriptions from Umm al-Rasas.” Liber Annuus 41: 423–428.

xxiv

publications by m.c.a. macdonald

1991. “Was the Nabataean Kingdom a ‘Bedouin State’?” Zeitschrift des Deutschen Palästina Vereins 107: 102–119. Reprinted with addenda, corrigenda, and index in Macdonald 2009, see book section. 1991. “Epigraphic Gleanings from the Archive of the Palestine Exploration Fund.” Palestine Exploration Quarterly 123: 109–116. 1990. “Inscription en écriture de Teima.” In Aux sources du monde arabe. L’Arabie avant l’Islam, Collections du Musée du Louvre, edited by A. Caubet, 50, no. 11. Paris: Réunion des musées nationaux. 1990. “Camel Hunting or Camel Raiding?” Arabian Archaeology and Epigraphy 1: 24–28. 1989. “Cursive Safaitic Inscriptions? A Preliminary Investigation.” Arabian Studies in Honour of Mahmoud Ghul. Symposium at Yarmouk University December 8–11, 1984, edited by M.M. Ibrahim, 62–81. (Yarmouk University Publications: Institute of Archaeology and Anthropology Series 2). Wiesbaden: Harrassowitz. 1989. “Frederick Victor Winnett [Obituary]”. Proceedings of the Seminar for Arabian Studies 19: 104. 1987. “Nomades et semi-nomades aux époques romaine et byzantine.” Dossiers Histoire et Archéologie 118: 44–46. 1987. “The Nabataeans.” In Treks and Climbs in the Mountains of Rum and Petra Jordan, edited by J. Howard, 17–19. Milnthorpe: Cicerone. 1986. “abcs and Letter Order in Ancient North Arabian.” Proceedings of the Seminar for Arabian Studies 16: 101–168. 1986. “Corpus of the Inscriptions of Jordan Project Field-Work 1982.” Archiv für Orientforschung 33: 252–254. 1984. “Jawa: Corpus des inscriptions de Jordanie (1981–1982).”Revue Biblique 91: 226–230. 1984. “Corpus of the Inscriptions of Jordan Project.” American Journal of Archaeology 88: 227. 1983–1984. “Report on the Work of the Corpus of the Inscriptions of Jordan Project.” Archiv für Orientforschung 29–30: 279–282. 1983. With A. Searight. “Inscriptions and Rock-Art of the Jawa Area, 1982: A Preliminary Report.” Annual of the Department of Antiquities of Jordan 27: 571–576. 1983. “Corpus of the Inscriptions of Jordan Project.” American Journal of Archaeology 87: 195. 1982. With A. Searight. “The Inscriptions and Rock-Drawings of the Jawa Area: A Preliminary Report on the First Season of Field-Work of the Corpus of the Inscriptions of Jordan Project.” Annual of the Department of Antiquities of Jordan 26: 159–172. 1982. “The Corpus of the Inscriptions of Jordan Project.” Liber Annuus 32: 489–493. 1981. “Notes on Some Safaitic Inscriptions.” Appendix g in Jawa, Lost City of the Black Desert, by S.W. Helms, 257–263. London: Methuen. 1980. “A Bibliography of Gerald Lankester Harding.” Annual of the Department of Antiquities of Jordan 24: 8–12.

publications by m.c.a. macdonald

xxv

1980. “Safaitic Inscriptions in the Amman Museum and Other Collections ii.” Annual of the Department of Antiquities of Jordan 24: 185–208. 1979. “Safaitic Inscriptions in the Amman Museum and Other Collections i.” Annual of the Department of Antiquities of Jordan 23: 101–119. 1979. “In Memoriam Gerald Lankester Harding.” Annual of the Department of Antiquities of Jordan 23: 198–200. 1979. Review of F.V. Winnett and G. Lankester Harding, Inscriptions from Fifty Safaitic Cairns. Toronto, 1978. Palestine Exploration Quarterly 111: 137–140. 1976. With G. Lankester Harding. “More Safaitic Texts from Jordan.” Annual of the Department of Antiquities of Jordan 21: 119–133.

In Press With M. al-Najem. Catalogue of the Inscriptions in the Taymāʾ Museum. i. Inscriptions not found in the Saudi-German Excavations. With contributions by F. Imbert and P. Stein. Oxford: Archaeopress. “The Taymāʾ Aramaic, Nabataean, Dadanitic, and Taymanitic inscriptions from the Saudi-German excavations at Taymāʾ 2004–2015.” In Taymāʾ ii. The Inscriptions. The Saudi-German Excavations at Tayma. Final Report, edited by A. Hausleiter and R. Eichmann. Oxford: Archaeopress. Clues as to how a Nabataean may have spoken from a Hismaic inscriptions. In Festschrift for John Healey, edited by G.R. Smith. “Tweets from Antiquity: Literacy, Graffiti, and their Uses in the Towns and Deserts of Ancient Arabia.” In Scribbling through History. Graffiti, Places and People from Ancient Egypt to Modern Turkey, edited by C. Ragazzoli, E. Frood, O. Harmansah, and C. Salvador. London: Bloomsbury. Contributions on “The Desert and its People” and “The Arabs.” In The Blackwells Companion to the Hellenistic and Roman Near East, edited by T. Kaizer. Oxford: Blackwells. “Horses, Asses, and Hybrids, and their Use as Revealed in the Ancient Rock Art of the Syro-Arabian desert.” In Equids in the Ancient Near East, Egypt, and Arabia. Proceedings of a conference in memory of Mary Aiken Littauer, edited by K. Linduff, S. Olsen, and P. Raulwing. Oxford: bar. With C.J. Robin. “The Arabs and Warfare before the 7th Century ce.” In The Cambridge History of War, vol. i, edited by K. Raaflaub and B. Meissner. Cambridge: Cambridge University Press. “Towards a Re-assessment of the Ancient North Arabian Alphabets Used in the Oasis of al-ʿUlā.” Arabian Epigraphic Notes.

xxvi

publications by m.c.a. macdonald

Forthcoming With L. Nehmé. “Bny, ʾl and ʾhl in Nabataean and Safaitic.” In Dûma 2. The 2012 Report of the Saudi-Italian-French Archaeological Project at Dûmat al-Jandal, Saudi Arabia, edited by G. Charloux and R. Loreto. Paris-Riyadh: Saudi Commission for Tourism and National Heritage. “The Ancient South Arabian graffiti in the British Museum.” In Catalogue of the Ancient South Arabian inscriptions in the British Museum, edited by A. Avanzini. “Two fragmentary South Arabian inscriptions.” In the Corpus Inscriptionum Iudaeae Palestinae. “Safaitic inscriptions from Mount Hermon.” In the Corpus Inscriptionum Iudaeae Palestinae. Also published in Macdonald 2009 (see book section), Addenda and corrigenda, 2–3. “Graffiti from the South Hauran Survey 1995.” In The South Hauran Survey Final Report, edited by D.L. Kennedy and P. Freeman. “Notes on Inscriptions and Rock-Drawings from al-Madhi (Hadhramaut) and other Sites.” In Axelle Rougeulle’s final report on her survey in that area. “Note on Mixed Safaitic/Hismaic Inscriptions Carved on an Architectural Fragment from the Zarqa Region of Jordan.” In Report on Landscape Survey of the Zarqa Region of Jordan, edited by M. Wilson and M. Munzi. “The Ancient North and South Arabian inscriptions on the Darb al-Bakra.” In Darb al-Bakra, the Ancient Route Between Ḥegrā and Petra, edited by A.I. Al-Ghabbān. Riyadh.

In Preparation The Origin and Development of the South Semitic scripts. In the series Lehrbücher orientalischer Sprachen, published by Ugarit-Verlag. “What is the language of the Hasaitic inscriptions? The longest Hasaitic text so far discovered.” Atlal. “The Hismaic inscriptions from the Wadi al-Thamad survey, northern Jordan.” In Final Report of the Wadi al-Thamad Survey, edited by M. Daviau.

Tabula Gratulatoria Glen Bowersock, Institute of Advanced Study, Princeton Clive Holes, Oxford University Frédéric Imbert, Université Aix-Marseille Jeremy Johns, Oxford University Ted Kaizer, Durham University Michel Mouton, Centre Français d’Archéologie et de Sciences Sociales Joseph O’Hara, Oxford University Christian Robin, cnrs, Orient & Méditerranée Jérôme Rohmer, cnrs, Orient & Méditerranée Irene Rossi, cnr, Istituto di Studi sul Mediterraneo Antico, Roma Petra Sijpesteijn, Leiden University Phillip Stokes, University of Texas, Austin Mark Whittow, Oxford University

Tables and Figures Tables Luciani (Chapter 20) 1 2 3 4 5

Previously published Qurayyah painted ware fragments from Tell Kheleifeh 406 Previously unpublished qpw pottery from Tell Kheleifeh at the Harvard Semitic Museum 410 Fabric types in the qpw sherds from Tell Kheleifeh at the Harvard Semitic Museum 413 Find-spot of qpw sherds # 1–5 and # 7–10 according to the hand-written Tell Kheleifeh registries in the Harvard Semitic Museum 414 Find-spot of qpw sherds # 12–14 according to the hand-written Tell Kheleifeh registries in the Harvard Semitic Museum 430

Bellem and Watson (Chapter 31) 1

Proto-Semitic and Ancient South Arabian sibilant correspondences and allophones across Modern South Arabian 626

Figures Nehmé (Chapter 1) 1 2 3 4 5 6 7 8 9 10

Pétra, inscription Dalman 1912, no 63 (= mp 446) 8 Umm Jadhāyidh, inscription al-Theeb 2002, no 27 9 Madāʾin Ṣāliḥ, inscription arna.Nab 125 11 Pouzzoles, inscription cis ii 158 13 Pétra, inscription cis ii 372 (= mp 252) 14 Pétra, inscription cis ii 373 + 375 (= mp 253) 15 Madāʾin Ṣāliḥ, inscription MSNab 56/4 16 Madāʾin Ṣāliḥ, inscription MSNab 67/1 17 Madāʾin Ṣāliḥ, inscription MSNab 94/2 18 Umm Jadhāyidh, inscription al-Theeb 2002, no 76 19

xxix

tables and figures

Hayajneh (Chapter 3) 1 2 3 4 5 6 7 8 9 10 11 12 13

Map of Wadi Salhub Salhub 1 43 Salhub 2 44 Salhub 3 46 Salhub 4 48 Salhub 5 50 Salhub 6 51 Salhub 7 52 Salhub 8 56 Salhub 9 57 Salhub 10 59 Salhub 11 61 Salhub 12 63

42

Della Puppa (Chapter 4) 1 2 3 4 5

hnsd 121 = csns 95 78 Tracing of hnsd 121 = csns 95 78 ceds 356 with tracing by the author 79 kwq 118 79 wh 300 with tracing by the author 80

al-Manaser (Chapter 5) 1 2a 2b 2c 2d 3a 3b 3c 3d 4a 4b 5

Map of Jordan showing the location of Wādī Salmā 88 Inscription a 89 Inscription b 89 Inscription c 90 Inscription d 90 Inscription KhBG 17 91 Inscription mssh 12 92 Inscription hch 79 93 Inscription c 1104. 93 Inscription Damascus Museum no. 29067 (Harding 1969) 94 hsd (Harding 1969) 94 Musical instruments. a. Double pipe, yarġul; b. Majwiz or maqrun.

95

xxx

tables and figures

Prioletta (Chapter 7) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Ancient South Arabia 117 asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān asa graffiti from Shabathān

145 145 146 146 147 147 148 148 149 149 150 150 151 151 152 152 153

Stein (Chapter 8) 1

2 3

4

5 6

Sabäische Urkunde auf einem Holzstäbchen (Ausschnitt). Die Vergrößerung (unten) zeigt die vom Schreiber korrigierte Passage am Beginn von Zeile 2; der ursprüngliche fehlerhafte Text ist rot markiert. 156 Mittelsabäische Weihinschrift aus Mārib mit nicht ausgeschöpftem Schriftfeld. 166 Die große Inschriftenstele des ʾAbraha aus Mārib aus dem Jahre 548 n. Chr. Der Ausschnitt zeigt rechts die Schmalseite der Stele mit dem Schluß des Textes, dessen letzte Zeile durch Verbreiterung der Buchstaben komplett ausgefüllt wird. 167 Mittelsabäische Weihinschrift auf einer Bronzetafel. Der untere Teil des Schriftfeldes wurde durch Dehnung der Buchstaben sowie ein ornamentales Motiv aufgefüllt. 168 Mittelsabäische Weihinschrift auf einer Bronzetafel. Der Text reicht unten über das eingerahmte Schriftfeld hinaus. 169 Mittelsabäische Weihinschrift auf einem Steinblock. Die Schrift in den unteren zwei Dritteln ist stark gedrängt, die Buchstaben sind nur noch halb so groß wie in den ersten vier Zeilen. 170

tables and figures 7

8

9

10

11

12

13 14

15 16

17

18

xxxi

Mittelsabäische Weihinschrift auf einem Räucheraltar. Die letzte Zeile der Inschrift befindet sich auf der linken Seitenfläche des Fußes (Teilabb. links). 171 Mittelsabäische Weihinschrift auf einem Steinblock mit gut sichtbaren Linien für die Zeilenführung des Textes. Man beachte, daß der Schreiber die Buchstaben der drittletzten Zeile versehentlich über die Grundlinie hinaus verlängert und damit den Zwischenraum zur nächsten Zeile praktisch beseitigt hat. 172 Der Tatenbericht des Karibʾil Watar aus Ṣirwāḥ. Die Ausschnittvergrößerung läßt nicht nur horizontale, sondern auch vertikale Begrenzungslinien erkennen, welche die Breite sowie die Abstände der Buchstaben voneinander markieren. 173 Mittelsabäische Weihinschrift aus Ṣirwāḥ. Als letzten Buchstaben des letzten Wortes (der Gottesname ʾlmqh = ʾAlmaqah) hat der Steinmetz zunächst ʾ (𐩱) geschrieben und sein Versehen durch Darübersetzen des erforderlichen h (𐩠) korrigiert. 175 Mittelsabäische Weihinschrift aus Ṣirwāḥ. In Zeile 6 hatte der Schreiber zunächst den letzten Buchstaben (𐩬) des Wortes kbtn vergessen, seinen Fehler aber erst nach Fertigstellung des Textes bemerkt. Um das fehlende Zeichen noch unterzubringen, hat der Steinmetz den an der betreffenden Stelle eingemeißelten Worttrenner getilgt und das n an dessen Stelle gesetzt. Für den Trennstrich zum folgenden Wort blieb dann kein Platz mehr. 176 Altsabäische Weihinschrift auf einer aufwendig reliefierten Stele aus Mārib. In Zeile 5 wurde der letzte Buchstabe (𐩥) des Gottesnamens ʾlmqh(w) = ʾAlmaqah(ū) getilgt; die Motivation hierfür ist unklar (es kommen grundsätzlich beide Schreibungen in den Inschriften vor). 177 Ausschnittvergrößerungen der korrigierten Textpassagen aus Abb. 10–12. Die ursprünglichen fehlerhaften Buchstaben sind rot markiert. 178 Mittelsabäische Weihinschrift aus Mārib. Die ersten beiden Zeilen sowie der Beginn von Zeile 3, die den Namen zweier Stifter enthalten, wurden komplett ausgemeißelt und neu beschriftet. 179 Ausschnittvergrößerung des Beginns der Inschrift aus Abb. 14 180 Altsabäische Weihinschrift auf einer verzierten Stele aus dem Wadi al-Ǧawf. Auf dem Rahmen rechts und links der ersten drei Zeilen sind fehlende Buchstaben nachgetragen. 181 Dieselbe Inschrift wie in Abb. 16 in einer anderen Aufnahme. Der Ausschnitt zeigt nicht nur die auf dem Rand nachgetragenen fehlenden Buchstaben der Zeilen 1–3, sondern auch eine großflächige Korrektur, die die gesamte rechte Hälfte der zweiten Zeile umfaßt. 182 Die altsabäische Inschriftenstele aus Abb. 12 an ihrem ursprünglichen Standort

xxxii

19

20 21

22

23

24 25

tables and figures

im Hof des Barʾān-Tempels in Mārib. Das Bild verdeutlicht den repräsentativen Charakter solcher Stelen im öffentlichen Raum. 183 Mittelsabäische Weihinschrift aus Mārib. In Zeile 4 steht das Wort ytʾwln „sie kehren zurück“ mit einem fehlerhaften zusätzlichen ʾ (𐩱) an zweiter Stelle (also y⟨⟨ʾ⟩⟩tʾwln). Der Fehler wurde vom Steinmetz nicht korrigiert. 184 Die Ausschnittvergrößerung aus Abb. 19 zeigt den fehlplazierten Buchstaben in dem Wort y⟨⟨ʾ⟩⟩tʾwln in Zeile 4 (rot markiert). 185 Mittelsabäische Weihinschrift auf einer Bronzetafel. Das letzte Wort der ersten Zeile (bnw „die Söhne von (…)“) wurde zu Beginn der zweiten versehentlich wiederholt (rot eingerahmt). 186 Entwicklung der altsüdarabischen Monumental- (links) und Minuskelschrift (rechts) vom frühen 1. Jahrtausend v. Chr. (altsabäische Periode, aSab) bis in das 6. Jh. n. Chr. (spätsabäisch, spSab). 187 Sabäischer Brief auf einer beidseitig beschrifteten Palmblattrippe. Das Foto zeigt nur deren Vorder- bzw. Oberseite mit den ersten Zeilen des Textes, das Faksimile hingegen den gesamten Text auf Vorder- und Rückseite. 189 Entwurfsskizze eines sabäischen Briefes. Die stickpunktartigen Einträge sind auf drei untereinanderliegende Zeilen verteilt. 190 Entwurfsskizze für eine Weihinschrift auf einem Holzstäbchen. Der dreizeilige Text enthält ein Formular, das sonst nur in Stein gemeißelte Inschriften zeigen. 193

Kootstra (Chapter 9) 1

Photograph of AH 009.1 (courtesy of ociana)

205

del Carmen Hidalgo-Chacón Díez (Chapter 10) 1 2 3 4 5 6 7 8 9 10 11

Map of north-west Saudi Arabia showing the location of Madāʾin Ṣāliḥ and al-ʿUlā 219 Dadanitic inscriptions from Jabal al-Khraymāt at Madāʾin Ṣāliḥ 220 Jabal al-Khraymāt no. 1 in standardised Dadanitic glyphs 220 Jabal al-Khraymāt no. 2 in standardised Dadanitic glyphs 224 Jabal al-Khraymāt no. 3 in standardised Dadanitic glyphs 224 Dadanitic inscription JSLih 375 226 Dadanitic inscription Jabal Ithlib 2 226 Jabal al-Khraymāt no. 4 in standardised Dadanitic glyphs 226 Jabal al-Khraymāt no. 5 in standardised Dadanitic glyphs 228 Jabal al-Khraymāt no. 6 in standardised Dadanitic glyphs 229 Jabal al-Khraymāt no. 7 in standardised Dadanitic glyphs 231

xxxiii

tables and figures 12 13

Jabal al-Khraymāt no. 8 in standardised Dadanitic glyphs 232 Jabal al-Khraymāt no. 9 in standardised Dadanitic glyphs 232

Villeneuve (Chapter 12) 1 2 3 4 5 6

Carte localisant Diyātheh et Tell Freikh 255 Photographie de Tell Freikh vers l’est 256 Photographie de l’inscription grecque no 1 de Tell Freikh 257 Photographie de l’inscription grecque no 2 de Tell Freikh 259 Carte exposant les relations de visibilité entre le fort de Diyātheh, la tour de Marwad et le Tell Freikh 264 Photographie de la tour de surveillance de Marwad, vers le sud-ouest. Noter que seul le limon central de cette tour carrée est resté debout 265

Gatier (Chapter 13) 1 2 3 4

Carte de situation 272 Le bloc inscrit 277 Le sommet du bloc 278 Le bas du bloc 279

Brock (Chapter 15) 1 2 3

Script table 311 Side a reversed image 312 Side b reversed image 313

Hoyland (Chapter 16) 1 2 3 4

View of Khirbat al-Bayḍāʾ 328 Khirbat al-Bayḍāʾ, graffito 329 Kilwa, building text 331 Arabic inscription from Qasr Burquʾ, modern northeast Jordan

334

Hausleiter (Chapter 19) 1 2

Taymāʾ: The Burj Badr Ibn Jawhar, walls dividing the compounds and the outer wall with sand dunes at the inner side seen from North (Compound c) 362 Taymāʾ: The archaeological site of Qrayyah from North (Tuwayil Saʾid) with the Burj Badr Ibn Jawhar, Compounds a and b as well as the outer wall 363

xxxiv

tables and figures

3 4 5 6 7

Sketch plan of Taymāʾ by Charles M. Doughty 364 Sketch plan of the oasis of Taymāʾ 366 Overall plan of the oasis of Taymāʾ in 1979 367 The area of Qasr al-Hamra and Compound w west of it 369 The area of Qrayyah with Compounds (a-e, subcompound a1) and excavation areas 370 8a–b (a) Plan of Qrayyah; (b) Aerial photograph (possibly composed) on which the plan is based 371 9 Northern part of Qrayyah (Compound e): Aerial photograph of 1956 (inset), excavated and surveyed structures and reconstructed part of the northeastern perimeter wall 372 10 Walls of Tayma recorded by the walls-project 376 11 Tayma: the walls of the oasis 377 12 Sampled sites at the outer wall of Taymāʾwith datings from deposits older (green) or younger (red) than the construction of the wall; approximate course of the wall in the eastern part of sabkha 378 13 Eastern part of sabkha: remains of the older eastern wall 379 14 Eastern part of sabkha: Magnetogram and interpretation 380

Luciani (Chapter 20) 1.5, 7–8 Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (with graphic scales). 416 2.9, 10–11 Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (with graphic scales). 417 3 Barbotine pottery from Taymāʾ (a) compared to qpw pottery from Timnaʿ (b). a: Hausleiter 2014: fig. 5a; b: Rothenberg 1988: pl. 18 (colour) and fig. 5, no. 6 (line drawing). 423 4.12–14 Additional Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (at different scales) 430 5 Modified (from Pratico 1993a: pl. 2) general plan of Tell Keleifeh showing the localization of qpw sherds (yellow: single fragment; red: two fragments). For description of the find-spot see above Tables 4 and 5. 431

Kehrberg(-Ostrasz) (Chapter 21) 1

Fig. 1:1a–d, 1:2, 1:3: Toy camels objects nos. 3, 10, and 14, jcw01/109: season 2001, n-w wall trench 109, pre-city wall context, intact Late Hellenistic hypogean tomb (cf. Kehrberg 2004; 2006; Kehrberg and Manley 2002a; 2002c). Fig. 1.4: Photos of camels at Petra, Nov. 2005 (photos i. k–o) 441

xxxv

tables and figures 2

Fig. 2:1a–1c: two handle fragments and a base fragment from jcwp seasons 2000 and 2002: a/ jcw02/507 and c/ jcw02/508: season 2002, west wall trench 500, pre-city wall context (cf. Kehrberg & Manley 2003); b/ jcw00/27: season 2000, s-w wall trench 2000, pre-city wall context (cf. Kehrberg and Manley 2001). Fig. 2.2: clepsydra, object no. 6, jcw01/109: season 2001, n-w wall trench 100, pre-city wall context, intact Late Hellenistic hypogean tomb (cf. Kehrberg 2004; 2006; Kehrberg and Manley 2002a, 2002c) (photos and drawings i. k-o). 442

Maraqten (Chapter 22) 1 2

Location of the inscription mb 2006 i-54 450 The inscription of Mālikum bin Muʿāwiyat (mb 2006 i-54)

451

Nebes (Chapter 23) 1 2 3 4 5 6 7 8 9

Nordwestfassade des ʾAlmaqah-Tempels: c 366 (1) 475 Nordwestfassade des ʾAlmaqah-Tempels: c 366 (2) 475 Nordwestfassade des ʾAlmaqah-Tempels: c 366 (3) 476 Muḥāfaẓat Mārib 11 476 Muḥāfaẓat Mārib 11 477 Muḥāfaẓat Mārib 20/a 477 Muḥāfaẓat Mārib 20/b 477 Muḥāfaẓat Mārib 20/c 478 Muḥāfaẓat Mārib 20/d 478

Ephʿal (Chapter 24) 1

Map showing the location of the sites mentioned in the text

485

Cotton Paltiel (Chapter 29) 1 2

P.Politeuma 4 (recto) 593 P.Politeuma 4 (verso) 594

Bellem and Watson (Chapter 31) 1 2

A Reading palate (l), shown next to one frame of a palatogram (r) generated by ArticulateAssistant during j001’s articulation of s̃ of Śḥerɛ̄t /s̃ehey/ ‘tea’ 631 Palatogram showing contact patterns during j003’s production of s̃es, extracted from /s̃es lo/ ‘not with her’ (left) and šeš, extracted from /okšeš lo/ ‘not fat’ (right) 633

xxxvi 3

4

5

6

tables and figures

Palatogram showing contact patterns during j003’s production of s̃es, extracted from a token of the phrase /s̃es lo/ ‘not with her’ (a different token from fig. 2 above) 633 Individual frames of cumulative epg contact data showing maximal constriction achieved during j001’s production of s̃ in the word /s̃ehey/ ‘tea’ (left), s in the word /derhis/ ‘young female goat’ (centre), and š in the word /šibʕet/ ‘seven’ (right) 634 Three video stills showing lip position during articulation of: (1) s in /sīda/ ‘straight’ (j108); (2) s̃ in /s̃erbaʕ/ ‘cross!’ (j108); (3) s̃ in /s̃aʕgul/ ‘he hurries’ (m026) 636 Three video stills showing lip position of speaker j014 during articulation of: (1) s in /yiʕūrɛ̄s/ ‘they build it f. up’; (2) š in /bšaʕak/ ‘you m.s. have chopped it up’; (3) s̃ in /yis͂ir̍ ek/ ‘he does’ 637

Lancaster and Lancaster (Chapter 32) 1

Map of Arabian Peninsula, showing tribe locations and places mentioned in the paper. 649

Contributors Laïla Nehmé is a senior researcher at the cnrs (umr 8167, Orient et Méditerranée, Paris). She is an expert on Nabataean archaeology (Petra, Hegra, etc.) and Nabataean inscriptions as well as on the development of the Nabataean script into Arabic. She has been co-directing the Madâʾin Sâlih Archaeological Project in Saudi Arabia since 2002. She is presently compiling a corpus of Nabataeo-Arabic and pre-Islamic Arabic texts (late 3rd–6th century ad) and has recently published a two volume book Les tombeaux nabatéens de Hégra (2015, French Academy). Walter W. Müller is Emeritus Professor of Semitic Studies at University of Marburg and a member of the Academy of Sciences and Literature, Mainz, the German Archaeological Institute, and a Corresponding Fellow of the British Academy. His special fields of research and publication are Sabaean studies, i.e. languages, history, religion and culture of pre-Islamic South Arabia as well as Semitic epigraphy and Ethiopian studies. Hani Hayajneh (born 1965 in Deir Al-Siʿneh, Jordan) is a professor of Ancient Near Eastern Languages and civilizations at Yarmouk University, Jordan. He is expert in the field of cultural heritage studies, especially the Intangible cultural heritage. His research focuses on areas including the cultural history and heritage of the Arabian Peninsula from ancient periods to modern times. He represented Jordan on cultural heritage issues in the Intergovernmental Committee of the unesco Convention for Safeguarding Intangible Cultural Heritage. Chiara Della Puppa is a PhD student in Safaitic epigraphy at Leiden University. Her dissertation studies the paleography of Safaitic within the framework of the Jebel Qurma Archaeological Landscape project. Ali al-Manaser gained his PhD Free from the University of Berlin and is currently a researcher at the Khalili Research Centre (krc/University of Oxford). His research focuses on the cultural context of the Ancient North Arabian inscriptions. He is currently working on a Database for Islamic Inscriptions from Jordan and preparing several editions of unpublished Safaitic inscriptions for publication.

xxxviii

contributors

Alessandra Avanzini is Chair Professor of Semitic Philology at the University of Pisa. She has been director of the Department of Historical Sciences of Ancient World. Since 2001 she has been the director of the project “Corpus of South Arabian Inscriptions” for a complete edition of the corpus of Ancient South Arabian. Presently she is the pi of the erc—Advance Grant project “dasi—Digital Archive for the Study of pre-Islamic Arabian Inscriptions” that aims at digitizing and making available online the entire corpus of pre-Islamic epigraphs of Arabia [http:// dasi.humnet.unipi.it/]. She is also responsible for Arabia Antica, a theme portal specific to pre-Islamic Arabia [http://arabiantica.humnet.unipi.it] and director of the series “Arabia Antica”. Alessia Prioletta is a researcher at the cnrs (umr 8167, Orient et Méditerranée, Paris). From 2004 to 2012 she was a postdoctoral fellow at the University of Pisa, and from 2012 to 2015 she held a research position in the same institution within the erc project dasi (Digital Archive for the Study of pre-Islamic Arabian Inscriptions). Her research interests focus on the philology and epigraphy of pre-Islamic Arabia and pre-Islamic southern Arabia in particular. From 2007 to 2010 she worked in many museums of Yemen, which resulted in a monograph publishing the collections of three museums of the Dhamār region (2013). She is the author of several articles on the grammar and writing culture of ancient Yemen. As a member of the French-Saudi Archaeological Mission to Najrān since 2014, she has conducted epigraphic surveys in the areas of Ḥimā (Najrān), Dūmat al-Jandal and the southern Najd. Peter Stein was educated in Assyriology, Semitic Studies and Theology (Old Testament), and is currently working as senior lecturer for Biblical Hebrew and Semitic languages at Jena and Erfurt universities. His main fields of research are the languages and scripts on the Arabian Peninsula in Pre-Islamic times. He has published numerous books and articles on the language and written documentation in Sabaic, particularly on the everyday correspondences written on wooden sticks. Fokelien Kootstra is a PhD candidate at Leiden University at the Leiden Institute for Area Studies (lias). She has an ma in linguistics (2014) and a ba in Arabic studies (2011) from Leiden University. Her current research focuses on the Ancient North Arabian inscriptions from the ancient oasis of Dadan (modern-day al-ʿUlā in Saudi Arabia), with attention to linguistic variation.

contributors

xxxix

María del Carmen Hidalgo-Chacón Díez acquired her doctorate in Semitic languages (with a focus on Ancient North Arabian) from the University of Marburg, Germany. She did her master’s degree in Spanish and Semitic (Arabic) philology at the University of Granada, Spain and spent one year as an Erasmus student in the University François-Rabelais in Tours. She was a lecturer in Yarmouk University (Irbid) and Jordanian University (Amman) in Jordan for five years. She was a research associate in the project dasi (Digital Archive for the Study of pre-Islamic Arabian Inscriptions), University of Pisa, Italy and a researcher in the project ociana Project (The Online Corpus of the Inscriptions of the Ancient North Arabia) based in the Khalili Research Centre, University of Oxford, uk and is a member of davo (Deutsche Arbeitsgemeinschaft Vorderer Orient für gegenwartsbezogene Forschung und Dokumentation e.V.). Currently, she lives in Berlin, Germany. Maurice Sartre is an emeritus professor of Ancient History at the University François-Rabelais, in Tours, France. He is the author of several books about the history of Eastern Mediterranean since the Alexander’s conquest until the beginning of Islam, for instance D’Alexandre à Zénobie. Histoire du Levant antique, (2e éd., Paris, Fayard, 2003; partial English translation The Middle East under Rome, Harvard, The Belknap Press, 2005). He explored the Southern Syria and Southern Jordan to prepare the corpus of Greek and Latin inscriptions from these regions (igls xxi, 4, igls xiii, xiv and xv published, xvi in preparation, the last three in collaboration with Annie Sartre-Fauriat). He is also the author of Histoires grecques (Paris, Le Seuil, 2006; English translation, Snapshots from Antiquiy, Harvard, The Belknap Press, 2009). François Villeneuve born in 1954, he is a Professor at University Paris 1 Pantheon–Sorbonne and an archaeologist active in Syria, Jordan and Saudi Arabia, interested mainly in the Hellenistic and Roman periods. He is interested in relations between peasants, nomads and soldiers, as well as in ancient religious places in the Middle East. He is currently involved in archaeological projects in Dharih, Jordan; Hegra, Saudi Arabia; and Farasan, Saudi Arabia. He was the Director of ifapo, the French Archaeological Institute in the Near East, from 1990 to 1995, in Beirut and Damascus. He is the chairman of the scientific committee of ifpo, the French Institute in the Near East, and the Director of the research center Archeologies & Sciences de l’Antiquité (ArScAn) in Paris and Nanterre.

xl

contributors

Pierre-Louis Gatier is a Senior researcher at the French cnrs in Lyon. He works on the history, archaeology and epigraphy of the Roman and Byzantine Near East. Among his recent publications are books he co-edited on the history of ancient Tyre (Lebanon) and about the survey of the steppe East of Hama in Syria. He is currently preparing a new volume on the Inscriptions de la Jordanie. Jean-Baptiste Yon is a Senior researcher at the French cnrs in Lyon. He has worked extensively on the history and the epigraphy of the Hellenistic and Roman Near East. Among his recent publications are the corpus of the Greek and Latin inscriptions of Palmyra (igls 17), the catalogue of the Greek and Latin inscriptions in the National museum of Beirut (IGLMusBey) and an edited volume on the Ports of the Ancient Indian Ocean. Sebastian Brock is Emeritus Reader in Syriac Studies, Oxford University, and Emeritus Fellow of Wolfson College, Oxford. Robert Hoyland is Professor of Late Antique and Early Islamic Middle East History at nyu’s Institute for Study of the Ancient World. He has published widely on the material and intellectual culture of the region and has conducted fieldwork in Syria, Iraq and Yemen, and now in Azerbaijan. Ahmad Al-Jallad Ph.D. (2012) Harvard University, is an Assistant Professor at Leiden University. He has published on the comparative grammar of the Semitic languages, the history of Arabic, and on the epigraphy of Ancient North Arabia, including An Outline of the Grammar of the Safaitic Inscriptions (Brill, 2015). Marijn van Putten specializes in the historical linguistics of Berber and Arabic, and is currently doing a Post-Doc research on Quranic and Early-Islamic Arabic. Arnulf Hausleiter (ma 1992, PhD 1996 University of Munich, Habilitation 2012 Freie Universität Berlin) is a Near Eastern archaeologist currently based at the Orient Department of Berlin’s German Archaeological Institute (dai). Since 2004 he has been co-directing the Saudi-German multidisciplinary field project at the oasis of Tayma, Northwest Arabia.

contributors

xli

Marta Luciani is Professor of Ancient Near Eastern Archaeology and History at the University of Vienna and is the Director of the Austrian Component of the Joint Archaeological Project at Qurayyah, Saudi Arabia. A leading archaeologist with over 30 years of field experience, she has headed interdisciplinary archaeological projects in multiple lands in the region (Syria, Iraq, Turkey, Saudi Arabia). Recently she has published: Marta Luciani. 2016. (ed.), The Archaeology of North Arabia. Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5th–8th 2013 (orea Archaeological Series 4). Vienna: Austrian Academy of Sciences. Ina Kehrberg(-Ostrasz) completed her graduate and postgraduate studies in Classical and Near Eastern Archaeology at the University of Sydney. She was a permanent member of the International Jarash Archaeological Project from 1982 to 2003, first as archaeologist of the Australian Team. From 1984 to 1996 she worked with her late husband, Antoni Ostrasz, director of the Hippodrome Project, and lastly co-directed the Jarash City Walls Project in 2001–2002. Her main interests are ceramics of the Eastern Mediterranean world, with focus on Jordan, Hellenistic—Byzantine urban sites, and using artefacts assemblages as intimate tools for historical inquiries. Mohammed Maraqten is an archaeologist and specialist of ancient Semitic languages and cultures (Ruprecht-Karls-Universität Heidelberg, Department of Languages and Cultures of the Near East—Semitic Studies, Germany). He conducted archaeological excavations in Jordan, Bahrain, Yemen, Oman, and Morocco and he is epigrapher of the American Foundation for the Study of Man (afsm). He published intensively on Semitic epigraphy and in particular on ancient South Arabian inscriptions and cultural history of pre-Islamic Arabia. He is currently preparing the publication of the newly discovered Sabaic inscriptions from Mahram Bilqis (Maʾrib, Yemen), and the South Arabian minuscule texts written on wooden sticks (the collection of the National Museum of Sanaa). Norbert Nebes is professor of Semitic and Islamic Studies at the Friedrich Schiller University at Jena. He has published mainly on Epigraphic South Arabian and the grammar of Classical Arabic.

xlii

contributors

Israel Ephʿal is professor emeritus of History of the Jewish People and of Ancient Near Eastern Studies, Hebrew University of Jerusalem. Greg Fisher is an historian of the ancient Middle East and is the author and editor of numerous works on the topic, including Between Empires: Arabs, Romans, and Sasanians in Late Antiquity, and Arabs and Empires Before Islam. He lives in Ottawa. Orhan Elmaz studied Computer Science and Arabic Studies in Vienna, where he held a position in Islamic Studies after completing a PhD thesis about hapax legomena in the Qurʾan which was published in 2011. In 2013, he joined the University of St Andrews as a Lecturer in Arabic, where he has been teaching classical and modern Arab culture, classical Arabic language and literature, and media Arabic. His research focuses on Arabic linguistics and lexicography, comparative literature, and Islamic studies with an affinity for digital humanities. He is currently working on a linguistic exploration and description of Hadith Arabic. Konstantin M. Klein is lecturer (akadem. Rat) at the University of Bamberg/Germany. He received his Ph.D. from Brasenose College (Oxford) and works on the Greco-Roman Near East, the religious history of pre-Islamic Arabia, and Palmyrene epigraphy. John F. Healey was Professor of Semitic Studies in the University of Manchester until he retired in 2013. His main interests and publications have been in Nabataean and early Syriac epigraphy. He is a Fellow of the British Academy. Hannah M. Cotton Paltiel (D.Phil. Oxon. 1977) is emeritus Professor of Classics and Ancient History, holder of the Shalom Horowitz Chair in Classics at the Hebrew University of Jerusalem and holds an honorary doctorate in theology from the University of Bern. She edited the Latin and Greek papyri from Masada and the Greek papyri from Nahal Hever published in Desert xxvii 1997. She has written extensively on the legal documents from the Judaean Desert, juxtaposing them with rabbinic literature and papyri from other parts of the Roman World. She is the co-ordinating editor of an ongoing international project, supported by the Deutsche Vorschungsgemeindschaft (dfg) and the Hebrew University, known

contributors

xliii

as the Corpus Inscriptionum Iudaeae-Palaestinae (ciip)—an attempt to create a comprehensive multilingual corpus of all inscriptions, both published and unpublished from the fourth century bce to the seventh century ce, from the territories of Israel and Palestine. Miranda J. Morris is an independent researcher specialising in the Modern South Arabian Languages (msal) and the ethnography of southern Arabia, especially the areas in which the msal are spoken. She first went to live and work in Dhofar in 1975, and has also spent time living and working in the Soqotra archipelago. Sālim ʕAwaḏ̣ Ahmad al-Shaḥri was born in the Hedbīn area of eastern Dhofar. He currently works for the Ministry of Health as the Mushrif Idāri in the Hedbīn Clinic, Dhofar. Alex Bellem gained her ba in Arabic and Turkish from the University of Manchester, followed by an ma and PhD in Linguistics from ucl and soas, respectively. After working for the bbc, she held a post-doctoral fellowship with the Council for British Research in the Levant from 2008 to 2010, becoming their first Research Director (Syria), based in Amman and Damascus. She is currently Assistant Professor in Arabic at the University of Durham and sits on the cbrl’s Committee of Management. Janet C.E. Watson studied at Exeter University and soas, London. She has held academic posts at the universities of Edinburgh, Durham and Salford. She has also held visiting posts at the universities of Heidelberg (2003–2004) and Oslo (2004–2005). She has held the Leadership Chair for Language at Leeds at the University of Leeds since 2013. She was elected Fellow of the British Academy in July 2013. William and Fidelity Lancaster have worked in the Middle East as social anthropologists since 1971, first with the Rwala Bedouin, with whom they and their children lived for eight years, and later in Saudi Arabia, Oman and the United Arab Emirates. They have published a number of important and influential monographs on the Bedouin in different parts of Arabia, Syria and Jordan, including The Rwala Bedouin Today (1981), People, Land and Water in the Arab Middle East, Environments and Landscapes in the Bilād ash-Shām (1999), and Honour is in Contentment. Life before oil in Ras al-Khaimah (uae) and some neighbouring regions (2011), as well as numerous

xliv

contributors

articles. Between 1991 and 1994, William Lancaster was Director of the British Institute at Amman for Archaeology and History (later part of the Council for British Research in the Levant).

part 1 Epigraphy and Philology

∵

chapter 1

Les artisans et professions «libérales » dans le domaine nabatéen Laïla Nehmé

Il y a quelques années, à l’occasion d’une conférence donnée à l’ Institut d’ art et d’archéologie à Paris, je me suis intéressée de près à un thème jusque-là très peu abordé dans les études nabatéennes, à savoir les professions exercées par les auteurs des inscriptions ou attestées dans les sources littéraires relatives aux Nabatéens. Cet intérêt était motivé par le fait que l’ on connaît très mal la société et les institutions nabatéennes, et qu’ une étude systématique des professions mentionnées dans les textes me semblait un excellent moyen d’aborder cette question. De plus, quelques articles avaient été publiés depuis le début des années 1980, traitant à des degrés divers d’ aspects institutionnels ainsi que du personnel militaire nabatéens1. Leur lecture a révélé l’ absence quasi totale d’informations synthétiques sur les professions plus « ordinaires»2. Or, il suffit de considérer le nombre de fois où il nous est demandé, dans un formulaire, d’indiquer notre profession ou de choisir la catégorie socio-professionnelle à laquelle nous appartenons pour se convaincre de l’ importance des métiers dans la compréhension d’ une société donnée. Rassemblées dans des fichiers dont la gestion est confiée à des instituts d’ études statistiques ou économiques, ces données sont triées et analysées et offrent une image de la société qui les a produites: elles fournissent des renseignements sur les sources de sa richesse et attirent l’attention sur ses principaux secteurs d’ activité. C’est à ce même questionnaire que j’ai souhaité soumettre les Nabatéens, en hommage à Michael Macdonald et à ses travaux sur l’ épigraphie nabatéenne. Une petite base de données a servi à rassembler toutes les informations connues sur les professions mentionnées en contexte nabatéen. Les sources 1 Balty 1983, Bowsher 1989, Graf 1994, Briquel-Chatonnet 1995, Teixidor 1995. Dans cette contribution, les signes éditoriaux suivants ont été utilisés : { } encadrent une ou plusieurs lettres incertaines ; [ ] encadrent une ou plusieurs lettres restituées; ---- indiquent une lacune dans le texte ; / indique un saut de ligne ou une variante de lecture entre deux lettres si elles sont placées dans des { }. 2 À l’ exception des métiers liés à la construction et au travail de la pierre.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

4

nehmé

dépouillées sont littéraires3 et épigraphiques. Pour ces dernières, le lexique de J. Cantineau et le dictionnaire de J. Hoftijzer et K. Jongeling constituent la base du dépouillement. Cependant, toutes les publications contenant des éditions d’inscriptions parues depuis une trentaine d’années ont également été passées en revue. Il faut y ajouter les textes appartenant à des corpus d’ inscriptions actuellement en cours de publication: celui des inscriptions de Pétra, dans l’Atlas archéologique et épigraphique de Pétra, dont les fascicules 2 et 3 sont en préparation, celui des inscriptions de Boṣrà, celui des inscriptions du Darb alBakrah (terme qui désigne la section saoudienne de la piste caravanière reliant Hégra et Pétra selon l’itinéraire qui en a été dessiné par A. al-Ghabban4) et enfin celui des inscriptions du Jabal Ithlib, à Madāʾin Ṣāliḥ, qui ont fait l’ objet de relevés systématiques entre 2003 et 2005. Les termes désignant des professions qui apparaissent dans des noms propres n’ont, en revanche, pas été pris en compte, car les personnes qui ont reçu ces noms n’ ont pas forcément exercé les professions contenues dans leur nom5. Il est évident que les données obtenues par ce dépouillement n’offrent pas une image objective de la société nabatéenne. Elles sont déformées par le fait que notre documentation ne concerne qu’une (petite?) partie de la population nabatéenne, à savoir les individus capables d’écrire une inscription et, parmi eux, ceux qui ont décidé de faire connaître leur profession. Il n’y a aucun moyen de savoir a priori si l’échantillon ainsi obtenu est représentatif de l’ ensemble de la population. Malgré tout, ce questionnaire a au moins le mérite de répondre, partiellement au moins, à la question suivante : qui fait quoi et où ? La base de données contient actuellement quatre-vingt dix-huit entrées dans lesquelles on trouve non seulement des professions mais aussi des titres ou des fonctions. Les termes collectés ont été répartis en huit catégories: personnel religieux; personnel politique et institutionnel ; personnel militaire; personnel administratif; personnel lié aux activités agro-pastorales; personnel lié au commerce caravanier; esclaves et serviteurs; artisans et professions libérales. Certaines des catégories mentionnées ci-dessus ont déjà fait l’ objet de commentaires, en particulier celles qui concernent l’ armée6 et les pratiques religieuses7. Pour diverses raisons, et notamment parce que l’ attention qui leur a été portée jusqu’à présent est limitée, les artisans et l’ équivalent de nos profes3 4 5 6 7

Diodore de Sicile, Strabon, Flavius Josèphe, Périple de la mer Érythrée, etc. Voir al-Ghabban 2007 et Nehmé en préparation. A. Negev a dressé la liste de ces noms : Negev 1991 : 83. Bowsher 1989, Graf 1994, Teixidor 1995, Nehmé 2015a. Voir Healey 2001 : 163–165.

artisans et professions libérales

5

sions libérales seront la première cible d’une série d’ articles que je souhaite consacrer aux professions attestées dans la société nabatéenne. Les professions liées au bâtiment, qui ont déjà été étudiées par l’ auteur dans une contribution récente8, ne seront pas abordées ici. Pour mémoire, il s’ agit de ʾmnʾ, «sculpteur» ; bnyʾ, «constructeur» (attesté pour une niche, un autel, un stibadium, un tombeau monumental et une tombe simple) ; kylʾ, « mesureur», ngrʾ, «charpentier»9, pslʾ, «tailleur de pierre» et šydʾ, « plâtrier ». Le présent article portera donc sur les autres sous-catégories d’ artisans et de professions libérales. Les professions sont présentées ci-dessous dans l’ ordre alphabétique araméen.

ʾsyʾ, «médecin» Cette profession est attestée dans deux inscriptions, une de Hégra et une de Pétra. À Pétra, elle apparaît dans une inscription du Wādī Abū ʿUllayqah, au sudouest de Pétra, où se trouve un sanctuaire consacré à Isis. Il s’ agit de mp 869 : dkrwn ṭb w šlm l-hnʾw br zydqwmw sprʾ w l-hnʾt ḥbrh w rḥmh br šʿdʾlhy ʾsyʾ, « Que soient commémorés en bien et que soient sains et saufs Haniʾū fils de Zaydqawmū, le scribe et Hāniʾat son compagnon et son ami fils de Šaʿdʾallāhī, le médecin»10. Ce texte est intéressant car il associe un scribe et un médecin tout en précisant la nature du lien qui existe entre eux. Ces deux personnages faisaient peut-être partie de la même confrérie religieuse qui se réunissait dans des installations rupestres aménagées aux alentours du sanctuaire d’ Isis. À Hégra, ʾsyʾ apparaît dans le texte gravé sur la façade du tombeau ign 44, taillé sur le flanc oriental du Qaṣr al-Bint. L’inscription, JSNab 19, est datée de 26/27 apr. J.-C. et ʾsyʾ est le métier qu’exerçait le propriétaire du tombeau. Ce dernier appartient au type dit Hégra, soit au type le plus élaboré des tombeaux nabatéens, et sa façade mesure 13m de haut11. On peut en déduire que ce médecin était un homme relativement riche, voire très riche. Considérant le rôle dominant des militaires à Hégra, J. Healey12 a proposé de voir dans ce ʾsyʾ un 8 9 10 11 12

L. Nehmé dans Bessac 2007 : 15–21 et Annexe 1, p. 22–26; idem 2015b: 141–143. Mais ngrʾ pouvant également être lu ngdʾ, les attestations de « charpentier» restent incertaines. Lecture J.T. Milik revue dans Roche 2012 : 64–65. Il n’ y a que six tombeaux plus grands à Madāʾin Ṣāliḥ. Healey 1993 : 167.

6

nehmé

«médecin militaire» mais cela reste tout à fait hypothétique. On peut noter que ʾsyʾ figure entre le nom du personnage et son patronyme, ce qui est relativement inhabituel13. Le plus souvent, en effet, la profession est rejetée à la fin de la généalogie de l’auteur du texte, même si elle s’ applique très probablement à ce dernier et non à son père ou à un autre de ses ascendants. On ne sait quasiment rien de la médecine chez les Nabatéens, si ce n’est que ces derniers étaient sans doute réputés pour leurs connaissances dans le domaine des parfums, des médicaments et des poisons, comme le laisse supposer la mention, par Flavius Josèphe, d’une femme d’ Arabie experte dans l’ art des poisons, qui aurait administré à Phéroras (le frère d’ Hérode), à l’ instigation de Syllaios (le ministre d’Obodas iii), un poison mortel14. ʾsyʾ est un participe actif de ʾsy que l’on trouve également en palmyrénien, en judéo-araméen et en syriaque15.

bšmʾ, «parfumeur» D’après J.T. Milik, cette profession est attestée dans deux inscriptions de Pétra, une provenant du Wādī al-Farasah, l’autre du Wādī Abū ʿUllayaqah. La première, mp 130, a fait l’objet d’une réédition par mes soins16, dans laquelle j’ ai montré que la lecture proposée par le savant polonais n’est pas compatible avec les traces visibles sur la pierre. La seconde est une inscription inédite17, une simple signature, mp 1004, dans laquelle il m’a semblé, à l’ examen de la photographie, que la lecture des trois dernières lettres de bšmʾ étaient certaines, tandis que le b initial était peu assuré. La racine bšm signifie « parfum, encens » dans diverses langues sémitiques18. La traduction de bšmʾ par « parfumeur » est approximative (J.T. Milik) car le traitement des parfums, des aromates et des épices se déclinait certainement en plusieurs métiers et spécialisations.

13

14 15 16 17

18

Il en existe un autre exemple dans l’ inscription publiée par A. Negev (1971), selon la relecture qu’ en a proposée J. Teixidor (1986: 263 [= 1973, n° 174]): … whbʾl[hy] [----]ʾ br ḥkmw … On en trouve un troisième exemple dans une des inscriptions contenant le terme ṣyʿʾ, voir ci-dessous. Guerre des Juifs, i, 30, cité par M.-J. Roche (1996 : 96). Références dans Healey 1993 ; Cohen et al. 1970–, s.v. Nehmé 2003 : 220, fig. 11 ; voir désormais Nehmé 2012: 184, mp 130. Les inscriptions du Wādī Abū ʿUllayaqah et la documentation associée ont été transmises par mes soins à M.-J. Roche, qui dirige un projet de prospection épigraphique dans ce secteur et qui a désormais la responsabilité de la publication de ces inscriptions. Cohen et al. 1970–, s.v.

artisans et professions libérales

7

glbʾ, «barbier» Cette profession est attestée dans un seul texte, de Pétra, Dalman 1912: no 63 (= rés 1416, mp 446), gravé sur un rocher en face du ‘sanctuaire’ dit du Qaṭṭār adDayr (sur la rive opposée du ravin où se trouvent les installations cultuelles). Ce ‘sanctuaire’ est en réalité le lieu de réunion de thiases dont les membres ont laissé leurs signatures sur les parois rocheuses environnantes. Le texte se lit wʾlw glbʾ šlm, «Wāʾilū, le barbier, qu’il soit sain et sauf » (fig. 1). Dans le commentaire de l’inscription 1416, le rés fait remarquer que les monnaies nabatéennes représentent les rois avec les cheveux longs et la barbe rasée, ce qui est effectivement le cas dans dans la plupart des portraits royaux19. Il fallait donc bien des barbiers et des coiffeurs. J.T. Milik y a vu un membre du personnel des thiases et des temples20. glb est attesté avec le sens de « couper, tondre», dans diverses langues sémitiques21 et sa traduction ne fait aucun doute.

ḥnṭʾ, «embaumeur» Ce terme est attesté dans une seule inscription, provenant du site de Umm Jadhāyidh, au nord-ouest de Madāʾin Ṣāliḥ, publiée par S. al-Theeb en 2002 (no 27), puis de nouveau en 2010, (no 497)22: tymʾlḥwr br qwpʾ ḥnṭʾ / dkyr bṭb, « Taymʾalẖawwār fils de Qawfā, l’embaumeur» (fig. 2). La racine ḥnṭ signifie, tant en arabe qu’en araméen, «rendre épicé, embaumer, préparer pour l’ inhumation»23. Les analyses réalisées sur des fragments de tissus provenant de plusieurs tombeaux de Madāʾin Ṣāliḥ ont montré que les textiles recouvrant les corps étaient enduits d’éléments gras (probablement une huile végétale) et de composés triterpéniques destinés à en retarder la décomposition. La forte proportion, dans les échantillons étudiés, en α- et β-amyrines, biomarqueurs de la famille des Burséracées a – parmi d’autres arguments – permis de conclure que la substance résineuse étudiée (résine élémi) appartient à cette famille, et

19 20 21 22 23

Les cheveux courts sont une exception. Sur cette question, voir Schwentzel 2005: 151–154; idem 2013 : 189–191. Milik 1972 : 150–151. Cohen et al. 1970–, s.v. Voir aussi Hoftijzer et Jongeling 1995, s.v. Elle fait également partie du corpus à paraître des inscriptions du Darb al-Bakrah, UJadh 97. Brown et al. 1999, s.v. Voir en particulier l’ arabe ḥannāṭ.

8

figure 1

nehmé

Pétra, inscription Dalman 1912, no 63 (= mp 446) photo l. nehmé

artisans et professions libérales

figure 2

9

Umm Jadhāyidh, inscription al-Theeb 2002, no 27 photo l. nehmé

plus précisément au genre Canarium spp24. La préparation des corps, qui avait sans doute lieu au domicile du défunt, pouvait donc être confiée à un ḥnṭʾ.

ṭrq sktʾ, «le fabricant de moule de monnaie » L’ expression ṭrq sktʾ est attestée dans une seule inscription nabatéenne de Umm Jadhāyidh, publiée par S. al-Theeb (2002: no 25 et 2010 : no 495 = UJadh 100). Dans l’editio princeps, elle a cependant été lue comme un seul mot, trqsktʾ, interprété comme dérivant du grec et traduit par « le surveillant, le garde», sans explication. Je préfère, comme l’a suggéré M.C.A. Macdonald (comm. pers.), l’ interpréter comme un nom de métier composé de trq et de sktʾ. Le premier terme pourrait être un participe actif dérivé de la racine ṭrq, qui a plusieurs sens en arabe et en araméen. En arabe, ṭaraqa peut signifier « frapper une pièce de métal», et c’est ce sens qui est retenu ici. Quant à sktʾ, il pourrait s’ agir soit 24

I. Sachet dans Nehmé et al. 2010 : 212. Voir aussi Mathe et al. 2009 (analyses par chromatographie en phase gazeuse couplée à la spectrométrie de masse). Des analyses complémentaires, réalisées en 2010, ont montré la présence de corps gras mais aussi d’un mélange résineux comprenant de l’ élémi et de l’ encens (Boswellia spp). Voir en dernier lieu Bouchaud et al. 2015 : 35.

10

nehmé

de l’araméen sikkā, le «coutre» de la charrue ou « une sorte de bêche », ou une forme aramaïsée, avec s au lieu de š25, de l’ arabe sikkah, « soc de charrue » ou «moule, matrice», une pièce de métal gravée utilisée pour frapper des monnaies, et par extension les monnaies elles-mêmes26. Noter que ṭrq est attesté sous trois formes différentes dans les papyri nabatéens de Naḥal Ḥever, avec le sens de «nouer, attacher» ou de «préparer» le contrat d’ une certaine manière (cf. arabe ṭarīqah, «état, condition, règle, usage »)27.

kḥlyʾ, «maquilleur» Cette profession est attestée dans une seule inscription, inédite, de Pétra, mp 964, gravée dans le Wādī Abū ʿUllayaqah. C’ est la seule inscription relativement claire du point épigraphique auquel elle appartient. Il s’ agirait, d’ après la lecture proposée par J.T. Milik, de la signature avec souhait d’ une personne nommée Kalbū. D’après le commentaire que j’ ai fait de cette inscription avant d’en transmettre la photographie à M.-J. Roche, « la photo est peu claire mais la lecture pas impossible». J.T. Milik propose de voir dans cette profession un «maquilleur» qui jouerait un rôle dans les fêtes organisées autour du sanctuaire d’Isis. kḥl étant attesté en araméen et en arabe avec le sens de « peindre les yeux»28, il n’y a pas de raison qu’il ne soit pas attesté en nabatéen avec un sens lié au traitement appliqué à cette partie du visage.

nwlʾ, «tisserand» Cette profession est attestée en nabatéen dans une seule inscription, arna.Nab 125 (fig. 3), qui provient du Jabal Ithlib à Madāʾin Ṣāliḥ. Elle est gravée, parmi d’autres inscriptions (arna.Nab 124–127), au débouché de la gorge qui marque l’entrée du jabal, sur la gauche lorsque l’on vient du Dīwān, avant de monter sur la terrace le long de laquelle sont gravées de nombreuses autres inscriptions29. Elle a été photographiée par F.V. Winnett et W.L. Reed30 et un nouveau cliché, 25 26 27 28 29 30

L’ un est en effet parfois utilisé à la place de l’ autre en nabatéen: Cantineau 1930–1932, vol. 1, 43, qui cite msgdʾ/mšgdʾ, nps/npš, psl/pšl. Lane 1863–1893 : 1846a et 1387b. Yardeni 2014 : 311. Brown et al. 1999, s.v. C’ est le point épigraphique n° 67 sur la carte publiée dans Nehmé et al. 2010: fig. 2 p. 274. Mais le cliché n’ a pas été publié dans arna.

artisans et professions libérales

figure 3

11

Madāʾin Ṣāliḥ, inscription arna.Nab 125 photo l. nehmé

ainsi qu’une copie, ont été faits par l’auteur en 2004. L’inscription a souffert de l’ érosion durant les cinquante dernières années. La lecture proposée par J.T. Milik et J. Starcky, šly nwlʾ, « Šullay, le tisserand » peut être considérée comme relativement assurée pour plusieurs raisons: 1/ on ne peut pas lire br après šly et on ne s’attend donc pas, à la fin, à la présence d’ un patronyme; 2/ la lettre finale est un ʾ, donc a priori un état emphatique qui s’explique d’autant mieux si l’auteur mentionne sa profession, « Untel le

12

nehmé

…». La traduction de nwlʾ par «tisserand», proposée par les éditeurs, sans commentaire, est probablement fondée sur le judéo-araméen newal, « tisser »31, mais l’arabe connaît nawl et minwal, «métier à tisser ». Il n’est pas surprenant de trouver des tisserands dans la société nabatéenne. La découverte d’ un grand nombre de fragments de textiles dans les tombeaux étudiés et fouillés de Madāʾin Ṣāliḥ32, en lin, en laine et en coton, montre qu’ il existait un marché régional pour ces produits artisanaux. Il n’y a pas de raison de penser qu’ ils ont tous été importés de régions situées en dehors du royaume nabatéen.

nḥšʾ, «forgeron (en cuivre?)» ou «devin » Ce terme est attesté en nabatéen dans une seule inscription, cis ii 158, une dédicace religieuse de Pouzzoles dans laquelle un personnage qualifié de nḥšʾ consacre, en 5 apr. J.-C., les éléments rénovés d’ un sanctuaire33. La pierre est entreposée dans les réserves du musée de Naples et des phographies en ont été faites à ma demande, en lumière rasante, par J.-P. Brun. Le mot nḥšʾ apparaît à la fin de la ligne 1 (fig. 4). Cette inscription est de lecture difficile car elle a beaucoup souffert d’un séjour prolongé à l’air libre, mais l’ estampage du cis montre qu’elle était bien mieux conservée à la fin du xixe siècle. Malgré des traces actuellement ténues, la lecture peut être considérée comme certaine. J. Cantineau traduit nḥšʾ par «forgeron en cuivre» ou « devin » car la racine a d’une part le sens de «pratiquer la divination, observer les signes » et d’ autre part celui de «cuivre, bronze»34. En l’absence d’ autres parallèles épigraphiques, hormis une inscription hatréenne35 dans laquelle on souhaite un «bon augure», nḥšʾ ṭ[bʾ] à une personne, il est difficile de choisir, pour cis ii 158, entre les deux principaux sens de la racine. Le contexte dans lequel se trouve nḥšʾ plaiderait évidemment en faveur d’un «devin ».

31 32 33 34 35

Jastrow 1967, s.v. La racine est également attestée en syriaque pour designer des objets liés au tissage. Nehmé et al. 2010 : fig. 25–30, et surtout Augé et al. 2010: 206 et pl. 7.7–10. Sur ce texte, voir aussi Lacerenza 1988–1989 et Roche 1996: 86–87. Brown et al. 1999, s.v. Aggoula 1991 : no 67.

artisans et professions libérales

figure 4

13

Pouzzoles, inscription cis ii 158 photo j.-p. brun

ṣyʿʾ, «artisan qui façonne les métaux précieux » Cette profession est attestée dans huit inscriptions (toutes des signatures) provenant de trois sites. 1

Pétra (deux textes)

– cis ii 372, copie Lagrange pl. 47 (= mp 252) (fig. 5) : šlm zydw br tymw / ṣyʿʾ, «Que soit sain et sauf Zaydū fils de Taymū, l’artisan qui façonne les métaux précieux». Le texte est parfaitement clair et a été correctement lu par le cis. Ce texte et le suivant font partie d’un groupe de quarante signatures (mp 235– 275 = cis ii 355–384, parmi lesquelles douze sont inédites), gravées dans le secteur d’an-Nmayr, sur une paroi rocheuse verticale surplombant une étroite

14

nehmé

figure 5

Pétra, inscription cis ii 372 (= mp 252) photo l. nehmé

terrasse, au nord du monument connu sous le nom de « Chapelle d’ Obodas »36 et le long du ravin qui permet de remonter de là au Jabal al-Madhbaḥ. Cette paroi rocheuse est recouverte d’une patine foncée et elle est donc particulièrement bien adaptée à la gravure d’inscriptions. Ce groupe est l’ un des cinq groupes de textes gravés dans les environs des aménagements de la « Chapelle d’Obodas»37. – cis 373 + 375, copie Lagrange pl. 47 (= mp 253 et rés 1472) (fig. 6) : šlm tymw br whblhy / ṣyʿʾ, «Que soit sain et sauf Taymū fils de Wahbʾallāhī, l’ artisan qui façonne les métaux précieux». Ce texte est gravé exactement en dessous du précédent. Le cis a bien lu le mot ṣyʿʾ mais a considéré qu’ il faisait partie d’un autre texte (no 375).

36 37

Sur ce monument, voir Tholbecq et al. 2008, Tholbecq 2011 ainsi que Tholbecq et Durand 2013. Voir aussi Nehmé 2012b. Pour la description de ces groupes et leur localisation, voir Nehmé 2002: 251–253 et la carte p. 244.

artisans et professions libérales

figure 6

15

Pétra, inscription cis ii 373 + 375 (= mp 253) photo l. nehmé

Bien que le nom propre Taymū soit très fréquent en nabatéen, le fait que mp 252 et 253 soient gravés l’un au-dessus de l’autre et qu’ ils mentionnent la même profession rend probable l’hypothèse selon laquelle Zaydū est le petitfils de Wahbʾallāhī. Le père et le fils seraient tous les deux artisans du métal et auraient laissé leur signature non loin du lieu où se tenaient les réunions des confréries associées au culte du dieu Obodas. 2

Hégra (cinq textes)

– Deux signatures récemment republiées38, JSNab 58 et 119, contenant le même nom, gravées dans le Jabal Ithlib non loin l’ une de l’ autre. Il est vraisemblable qu’elles mentionnent la même personne, qui aurait gravé son nom deux fois sur le chemin qui l’amenait depuis les installations principales du Jabal Ithlib vers l’endroit où elle avait l’ intention de dresser ou de faire dresser un bétyle en l’honneur du dieu mr bytʾ, le « Seigneur de la maison», comme nous l’apprend JSNab 58. 38

Nehmé 2005–2006 : no 14 p. 200–201, fig. 139 (JSNab 58) et no 24 p. 212 (JSNab 119). JSNab 119 n’ a pas été photographiée en 2004.

16

nehmé

figure 7

Madāʾin Ṣāliḥ, inscription MSNab 56/4 photo l. nehmé

JSNab 58: mnṣb mrʾ bytʾ dy ʿbd whb[ʾ]lhy ṣyʿʾ, « Ceci est le dressoir à stèles du Seigneur de la maison qu’a fait Wahb[ʾ]allāhī, l’ artisan qui façonne les métaux précieux». JSNab 119: ʾḥd whbʾlhy ṣyʿʾ, «Wahbʾallāhī, l’artisan qui façonne les métaux précieux, a pris [possession]». Il est difficile de savoir ce que Wahbʾallāhī a «pris». Peut-être est-ce tout simplement l’endroit où a été gravé le texte. – Trois inscriptions inédites du Jabal Ithlib. Elles ont été photographiées et copiées par l’auteur en 2004. La première est MSNab 56/4 (fig. 7)39, gravée sur la face nord-est d’ une butte rocheuse située à une trentaine de mètres à l’ouest de la gorge du Dīwān: šlmw / šlmw ṣy/ʿʾ / bṭb, «Šalmū Šalmū, l’artisan qui façonne les métaux précieux, en bien». Le nom de l’auteur est répété deux fois et la profession est écrite sur deux lignes, en lettres beaucoup plus petites que le reste du texte. Il est possible que l’auteur ait d’abord voulu écrire šlm, mais comme son nom était šlmw, il a tracé par erreur un w à la fin du premier šlm. 39

Ce numéro est celui qui sera donné dans le corpus des inscriptions de Madāʾin Ṣāliḥ, dans lequel « ms » représente les initiales du site, « 56» le numéro du point épigraphique, tandis que le numéro qui suit le « / » est un numéro d’ ordre au sein de ce point. Ce texte a d’abord été publié par S. al-Theeb (1993: no 89) mais avec une lecture légèrement différente (šlm / šlmy ṣyʿʾ / bṭb).

artisans et professions libérales

figure 8

17

Madāʾin Ṣāliḥ, inscription MSNab 67/1 photo l. nehmé

La seconde est MSNab 67/1 (fig. 8), gravée à gauche de arna.Nab 126, au débouché de la gorge du Dīwān dans le Jabal Ithlib, à 4 m du sol : {d/r}{q}y ṣyʿʾ, « {D/R}{q}ī, l’artisan qui façonne les métaux précieux». La seconde lettre a été lue sur le terrain comme un n mais elle serait très éloignée de la première, et il y a largement la place pour la boucle d’un q dont on distingue encore vaguement le tracé sur la pierre. Le nom est attesté une fois ailleurs en nabatéen, dans une inscription provenant de la région de Tabūk (al-Theeb 2010 : no 912). La troisième est MSNab 94/2 (fig. 9), gravée sur une paroi rocheuse sur le flanc extérieur du Jabal Ithlib, à environ 200m au nord-est de la gorge du Dīwān: šlm {d/r}[q]y ṣyʿʾ br, «Que soit sain et sauf {D/R}{q}ī, l’ artisan qui façonne les métaux précieux, fils de». Il est à peu près certain que les deux textes ont été écrits par la même personne.

18

nehmé

figure 9

Madāʾin Ṣāliḥ, inscription MSNab 94/2 photo l. nehmé

3 Région d’al-Jawf (un texte) Inscription publiée par Kh. al-Muaikil et S. al-Theeb (1996 : no 43). La séquence de lettres ṣyʿʾ y apparaît à la fin du texte, lu dkyr ʾmʿmw br ṣyʿʾ par les éditeurs, qui l’interprètent comme un nom propre. J’ai déjà suggéré ailleurs40 que le premier nom devrait être lu ʾmʿmwbw et que ṣyʿʾ est sa profession. De fait, sur une photographie récemment publiée de ce texte41, il est clair qu’ après ʾmʿmw, il faut lire bw, pas br. Le terme ṣyʿʾ, qui a déjà été commenté42, a été rapproché de la racine arabe ṣwġ, qui signifie «former, façonner, mouler, pour un orfèvre», le ṣāʾiġ étant celui qui réalise ces actions, c’est-à-dire l’artisan qui travaille les ornements en or ou dans un autre métal précieux43. Enfin, il faut rappeler44 que dans l’inscription al-Theeb 2002, no 174, le texte doit être relu dkyryn šbw w šyʿʾlhy w wʾlw. Il n’y a en effet aucun doute sur le fait 40 41 42 43 44

Nehmé 2005–2006 : 201, n. 105. Al-Theeb 2010 : 1400. Ibidem. Lane 1863–1893 : 1747b. Déjà dans Nehmé 2005–2006 : 201.

artisans et professions libérales

figure 10

19

Umm Jadhāyidh, inscription al-Theeb 2002, no 76 photo l. nehmé

qu’ il continue sur la gauche et que ṣyʿʾ correspond en fait au début de šyʿʾlhy. Le terme ṣyʿʾ n’est donc pas mentionné dans ce texte.

qynʾ, «forgeron» Je connais trois attestations de cette profession, une incertaine à Pétra et deux en Arabie du Nord-Ouest. À Pétra, elle apparaît peut-être dans une inscription inédite, mp 529, qui provient de la dernière partie (en amont) de la voie processionnelle menant au Dayr, sur la droite avant d’ atteindre ce monument. Ce texte, dont il ne subsiste que le milieu, a été photographié par un touriste. Seul le début du mot, qy, est visible, la paroi sur laquelle le reste de l’ inscription a été gravée s’étant détachée du substrat rocheux. Avant qy, on distingue les traces de deux lettres, dont la seconde pourrait être un ʾ. La lecture qy[nʾ], proposée par J.T. Milik, reste donc très hypothétique. D’Arabie du Nord-Ouest, on peut citer al-Theeb 1993, no 50 (= 2010 : no 886), provenant de la région de Tabūk: šlm ḥmyn br ʿydw qynʾ bṭb l-ʿlm, « Que soit sain et sauf Ḥumayn fils de ʿAydū, le forgeron, en bien à jamais », ainsi que al-Theeb 2002, no 76 (= 2010 no 547), provenant de Umm Jadhāyidh (= UJadh 307) : ḥzn br g{d/r}t qynʾ, «Ḥazan fils de G{d/r}t [Juddah?], le forgeron » (fig. 10).

20

nehmé

La racine qyn est largement attestée dans les langues sémitiques avec le sens de «travailler le métal»45 et on trouve la profession de qynʾ à Palmyre et à Hatra. À Palmyre, on a notamment la mention, dans une bilingue honorifique, d’ une association de travailleurs du métal (or et argent) qui a fait ériger une statue d’Odaïnat (cis ii 3945).

Autres professions? Plusieurs professions sont citées dans l’index du volume publié en 2010 par S. al-Theeb, qui rassemble toutes les inscriptions nabatéennes de l’ Arabie. Nous commentons ci-dessous celles qui ne nous semblent pas recevables. – gʾyʾ, «couturier» ? : inscription al-Theeb 2010, no 851 (= al-Theeb 1993 no 15), šlm ʿbdʾlgʾ br ʾḥyw gʾyʾ. Le dernier mot de cette signature, gʾyʾ, a été interprété diversement par l’ éditeur. En 1993, il y a vu la seconde partie d’ un nom composé, ʾḥywgʾyʾ, alors qu’en 2010, il l’interprète comme une profession qui serait dérivée de l’arabe gʾy, «rapiécer, raccommoder, réparer». Les deux interprétations sont sujettes à caution, la première parce que le nom propre obtenu serait le premier nom propre nabatéen contenant le toponyme de l’ancienne Wādī Mūsà, couramment utilisé en nabatéen, notamment dans une épithète de Dūšarā46; la seconde car le sens de gʾy est à mettre en relation avec l’idée de «pièce rapportée», la notion de couture arrive donc en second. Une autre possibilité, plus simple, consiste à voir dans gʾyʾ le toponyme Gaia (Wādī Mūsà). L’auteur, qui a écrit son texte sur un rocher de la région de Tabūk, indiquerait son origine, sans toutefois utiliser la préposition mn entre ʾḥyw et gʾyʾ. – dwmʾ, «plâtrier» ? : JSNab 181, selon une des relectures proposées par S. alTheeb (2010: no 275). Le texte se lit ʿylw br šw{d/r}{d/r}wmʾ šlm. J’ ai déjà évoqué ce texte47 car une des lettres, la dixième, est pourvue d’ un point diacritique. Le patronyme a été lu šwdrwmʾ par Jaussen et Savignac mais šwrdwmʾ est plus vraisemblable car le point a sans doute été mis sur le d (le point diacritique est presque toujours mis sur cette lettre en nabatéen). Il n’y a par ailleurs aucune raison de couper ce nom propre en deux, d’ autant que dwmʾ ne signifie pas particulièrement « plâtrier »48. 45 46 47 48

Brown et al. 1999, s.v. Voir des exemples dans Nehmé 2005–2006 : n. 51. Nehmé 2010 : 55, 57, fig. 10. Cohen et al. 1970- : 236–237.

artisans et professions libérales

21

– ktnʾ, «celui qui colle» : dans l’inscription al-Theeb 2002, no 91 (= 2010 : no 563), l’auteur propose de lire šlm rbʾl br / tymw ktnʾ / brh ktbh, « Que soit sain et sauf Rabʾel fils de Taymū al-kattān («celui qui colle »?) son fils l’ a écrit». Au lieu de ktnʾ, nous préférons lire, plus simplement, ktbʾ, « celui qui grave». – šʿʾ, «celui qui applique l’enduit», d’après l’arabe sāʿa: JSNab 371, entre Taymāʾ et Madāʾin Ṣāliḥ, selon une relecture du texte proposée par S. alTheeb (2010: no 426) qui ne semble pas s’imposer, à savoir ʿbd šʿʾ br wʾln au lieu de ʿbyšʿʾ br wʾln de Jaussen et Savignac.

Conclusion Il est tout à fait étonnant de constater que sur les dix professions recensées dans la catégorie définie, quatre concernent le travail des métaux (nḥšʾ ?, ṣyʿʾ, qynʾ, ṭrq sktʾ) et trois autres concernent des métiers en rapport avec l’ encens et les cosmétiques (bšmʾ, ḥnṭʾ, kḥlyʾ). Les trois derniers sont un barbier/coiffeur, un tisserand et un médecin. Cela met en relief le caractère finalement assez restreint de la panoplie des métiers de la catégorie étudiée dans cette contribution – les artisans et professions libérales. Cela est d’ autant plus vrai que la troisième grande catégorie d’artisans concerne les métiers liés au bâtiment au sens large. Les trois grandes familles d’artisans qui mentionnent volontiers leur métier dans les inscriptions nabatéennes travaillaient principalement la pierre, le métal et les parfums. Cela ne surprendra personne pour la pierre, peut-être pour les parfums, mais est plus inattendu pour le métal. Enfin, on peut relever que sur les dix termes présentés dans cette contribution, deux sont des emprunts à l’arabe: ṣyʿʾ et ṭrq sktʾ.

Sigles arna.Nab Inscriptions nabatéennes publiées par J.T. Milik et J. Starcky dans Winnett et Reed 1970. cis ii Corpus Inscriptionum Semiticarum. Pars ii. Inscriptiones Aramaicas continens. Paris, 1889–. JSNab Inscriptions nabatéennes publiées dans Jaussen et Savignac 1909–1922. mp Inscriptions nabatéennes de Pétra, dont le corpus a été préparé par J.T. Milik, vérifié et complété par L. Nehmé et publié en partie dans Nehmé 2012a. P. Yadin Yadin, Y., J.C. Greenfield, A. Yardeni, and B. Levine. 2002. The Documents from the Bar Kokhba Period in the Cave of the Letters. Hebrew, Aramaic

22

rés

nehmé and Nabataean Aramaic Papyri. (Judean Desert Studies). Jerusalem: Israel Exploration Society/The Hebrew University of Jerusalem. Répertoire d’épigraphie sémitique. Paris, 1900–1968.

Bibliographie Aggoula, B. 1991. Inventaire des inscriptions hatréennes. (Bibliothèque archéologique et historique 139). Paris: P. Geuthner. Augé, Chr., V. Bernard, P. Dal-Prà, Y. Gerber, L. Nehmé, I. Sachet. 2010. «Report on the 2004, Fourth Season, of the Saudi-French Archaeological Project at Madâʾin Sâlih». Atlal 20: 197–220 [anglais], pl. 7.1–7.20 et p. 145–158 [arabe]. Balty J.-Ch. 1983. «Architecture et société à Pétra et Hégra. Chronologie et classes sociales; Sculpteurs et commanditaires». Dans Architecture et société. De l’archaïsme grec à la fin de la République romaine, 303–324. Actes du Colloque international organisé par le Centre national de la recherche scientifique et l’École française de Rome (Rome 2–4 décembre 1980). (Collection de l’École française de Rome 66). Rome: École française de Rome. Bessac, J.-Cl. 2007. Le travail de la pierre à Pétra. Technique et économie de la taille rupestre. Paris: Éditions Recherche sur les Civilisations. Bouchaud Ch., I. Sachet, P. Dal-Prà, N. Delhopital, R. Douaud et M. Leguilloux. 2015. «New Discoveries in a Nabataean Tomb. Burial practices and ‘Plant Jewellery’ in Ancient Hegra». Arabian Archaeology and Epigraphy 26: 28–42. Bowsher, J.M.C. 1989: «The Nabataean Army». Dans The Eastern Frontier of the Roman Empire. Proceedings of a colloquium held at Ankara in September 1988, édité par D.H. French et C.S. Lightfoot, 19–30. (bar International Series 553). Oxford: bar. Briquel-Chatonnet, F. 1995. «La pénétration de la culture du Croissant fertile en Arabie: à propos des inscriptions nabatéennes». Dans Présence arabe dans le Croissant fertile avant l’ Hégire. Actes de la Table ronde internationale (Paris, 13 novembre 1993), édité par H. Lozachmeur, 133–141. Paris: Éditions Recherche sur les Civilisations. Brown, F., S.R. Driver, C.A. Briggs. 1999. The Brown-Driver-Briggs Hebrew and English Lexicon. Peobody, ma: Hendrickson. Cantineau, J. 1930–1932. Le Nabatéen (2 volumes). Paris: E. Leroux. Cohen, D., F. Bron et A. Lonnet. 1970–. Dictionnaire des racines sémitiques ou attestées dans les langues sémitiques. Louvain-la-Neuve: Peeters. Dalman, G. 1912. Neue Petra-Forschungen und der heilige Felsen von Jerusalem. Leipzig: J.C. Hinrichs. al-Ghabban, A.B.I. 2007. «Le Darb al-Bakra. Découverte d’une nouvelle branche sur la route commerciale antique, entre al-Ḥigr (Arabie Saʿūdite) et Pétra (Jordanie)». Comptes rendus de l’Académie des inscriptions & belles lettres : 9–24.

artisans et professions libérales

23

Graf, D.F. 1994. «The Nabataean Army and the Cohortes Ulpiae Petraeorum ». Dans The Roman and Byzantine Army in the East. Proceedings of a colloquium held at the Jagiellonian University, Kraków in September 1992, édité par E. Dąbrowa, 265–309. Kraków: Uniwersytet Jagielloński. Healey, J.F. 1993. The Nabataean Tomb Inscriptions of Madaʾin Salih. (Journal of Semitic Studies Supplement 1). Oxford: Oxford University Press. Healey, J.F. 2001. The Religion of the Nabataeans. A Conspectus. (Religions in the GraecoRoman World 136). Leiden: Brill. Hoftijzer, J. et Jongeling, K. 1995. Dictionary of the North-West Semitic Inscriptions. 2 vol. Leiden-New York-Cologne: Brill. Jastrow, M. 1967. A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature. 2 vol. Brooklyn, ny: Shalom. Jaussen, A., and R. Savignac. 1909–1922. Mission archéologique en Arabie. 6 vol. Paris: E. Leroux-P. Geuthner. Lacerenza, G. 1988–1989. «Il dio Dusares a Puteoli». Puteoli 12–13: 119–149. Lane, E.W. 1863–1893. An Arabic-English Lexicon. Derived from the Best and Most Copious Eastern Sources. 8 vol. London: Williams & Norgate. Mathe, C., P. Archier, L. Nehmé et C. Vieillescazes. 2009. «The Study of Nabataean Organic Residues from Madâʾin Sâlih, ancient Hegra, by Gas Chromatography-Mass Spectrometry». Archaeometry 51 (4): 626–636. Milik, J.T. 1972. Dédicaces faites par des dieux (Palmyre, Hatra, Tyr) et des thiases sémitiques à l’époque romaine. (Recherches d’épigraphie proche-orientale 1. Bibliothèque archéologique et historique 92). Paris: P. Geuther. al-Muaikil, Kh.I. et S.A. al-Theeb. 1996. Al-ʾāṯār wa-ʾl-kitābāt al-nabaṭiyyah fī minṭaqat al-jawf. Ar-Riyāḍ: Maktabat al-malik Fahd al-waṭaniyyah. Negev, A. 1971. «A Nabatean Epitaph from Trans-Jordan». Israel Exploration Journal 21: 50–52, pl. 4. Negev, A. 1991. Personal Names in the Nabataean Realm. (Qedem 32). Jerusalem: The Hebrew University. Nehmé, L. 2002. «La Chapelle d’Obodas à Pétra. Rapport préliminaire sur la campagne 2001». Annual of the Department of Antiquities of Jordan 46: 243–256. Nehmé, L. 2003. «Les inscriptions des chambres funéraires nabatéennes et la question de l’anonymat des tombes». Arabian Archaeology and Epigraphy 14: 203–258. Nehmé, L. 2005–2006. «Inscriptions vues et revues à Madāʾin Ṣāliḥ». Arabia 3: 179–225 [texte], 345–356 [figures]. Nehmé, L. 2010. «A Glimpse of the Development of the Nabataean Script into Arabic based on Old and New Epigraphic Material». Dans The development of Arabic as a Written Language, édité par M.C.A. Macdonald, 47–88. (Supplement to the Proceedings of the Seminar for Arabian Studies 40). Oxford: Archaeopress. Nehmé, L. 2012a. Atlas archéologique et épigraphique de Pétra. Fascicule 1. De Bāb as-Sīq

24

nehmé

au Wādī al-Farasah. (Épigraphie & Archéologie 1). Paris: Académie des Inscriptions et Belles-Lettres. Nehmé, L. 2012b. «Le dieu Obodas chez les Nabatéens: hypothèses anciennes et découvertes récentes». Dans Dieux et déesses d’Arabie. Images et représentations. Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 octobre 2007, édité par I. Sachet and C.J. Robin, 181–224. Paris: De Boccard. Nehmé, L. 2015a. « Strategoi in the Nabataean Kingdom: a Reflection of Central Places? ». Arabian Epigraphic Notes 1: 103–122. Nehmé, L. 2015b (éd.). Les tombeaux nabatéens de Hégra. (Épigraphie & Archéologie 2). Paris: Académie des Inscriptions et Belles-Lettres. Nehmé, L. en préparation. Epigraphy on the Edges of the Roman Empire. A Study of the Nabataean inscriptions and Related Material from the Darb al-Bakrah, Saudi Arabia, 1st–5th century ad. Nehmé, L., D. al-Talhi et F. Villeneuve. 2010. Report on the First Excavation Season at Madâʾin Sâlih, Saudi Arabia, 2008. (A Series of Refereed Archaeological Studies 6). Riyadh: Saudi Commission for Tourism and Antiquities. Roche, M.-J. 1996. «Remarques sur les Nabatéens en Méditerranée». Semitica 45: 73– 99. Roche, M.-J. 2012. «A Nabataean Shrine to Isis in Wādī Abū ʿUllayqah, in the SouthWest of Petra». Dans The Nabataeans in Focus: Current Archaeological Research at Petra, édité par L. Nehmé et L. Wadeson, 55–72. (Supplement to the Proceedings of the Seminar for Arabian Studies, 42). Oxford: Archaeopress. Schwentzel, Chr.-G. «Les thèmes du monnayage royal nabatéen et le modèle monarchique hellénistique». Syria 82: 149–166. Schwentzel, Chr.-G. 2013. Juifs et Nabatéens. Les monarchies ethniques du Proche-Orient hellénistique et romain. Rennes: Presses universitaires de Rennes. Teixidor, J. 1986. Bulletin d’épigraphie sémitique (1964–1980). (Bibliothèque archéologique et historique 127). Paris: P. Geuthner. Teixidor, J. 1995. «Le campement: ville des Nabatéens». Semitica 43–44: 111–121. al-Theeb, S. 1993. Aramaic and Nabataean Inscriptions from North-West Saudi Arabia. Riyadh: King Fahd National Library. al-Theeb, S. 2002. Nuqūš jabal umm jaḏāyiḏ al-nabaṭiyyah. Ar-Riyāḍ: Maktabat al-malik Fahd al-waṭaniyyah. al-Theeb, S. 2010. Mudawwanat al-nuqūš al-nabaṭiyyah fī ʾl-mamlakah al-ʿarabiyyah alsaʿūdiyyah. Ar-Riyāḍ: Dārat al-malik ʿAbdulʿazīz. Tholbecq, L. 2011. «Infrastructures et pratiques religieuses nabatéennes. Quelques données provenant du sanctuaire tribal de la “Chapelle d’Obodas” à Pétra». In La pioche et la plume. Autour du Soudan, du Liban et de la Jordanie. Hommages archéologiques à Patrice Lenoble, édité par V. Rondot, Fr. Alpi et F. Villeneuve, 31– 44. Paris: pups.

artisans et professions libérales

25

Tholbecq, L. et C. Durand. 2013. «A Late Second Century bc Nabataean Occupation at Jabal Numayr: the Earliest Phase of the “Obodas Chapel” Sanctuary». Dans Men on the Rocks. The Formation of Nabataean Petra. Proceedings of a conference held in Berlin 2–4 December 2011, édité par M. Mouton et St. Schmid, 205–222. (Supplement to the Bulletin of Nabataean studies 1). Berlin: Logos. Tholbecq, L., C. Durand et Ch. Bouchaud. 2008. «A Nabatean Rock-Cut Sanctuary in Petra: Second Preliminary Report on the “Obodas Chapel” Excavation Project, Jabal Numayr (2005–2007)». Annual of the Department of Antiquities of Jordan 52: 235– 254. Winnett, F.V. et Reed, W.L. 1970. Ancient Records from North Arabia. Toronto: University of Toronto Press. Yardeni, A. 2014. “A List of the Arabic Words Appearing in Nabataean and Aramaic Legal Documents from the Judaean Desert.” Scripta Classica Israelica 33: 301–324.

chapter 2

Anmerkungen zum safaitischen und althebräischen Onomastikon Walter W. Müller

Im Jahre 1963 veröffentlichte der Verfasser dieser Zeilen in der Zeitschrift für die alttestamentliche Wissenschaft, 75. Band, S. 304–316, einen Aufsatz betitelt Altsüdarabische Beiträge zum hebräischen Lexikon, der als Ergänzung zu Wolf Leslaus Ethiopic and South Arabic Contributions to the Hebrew Lexicon (Berkeley and Los Angeles 1958) gedacht war, da sich Leslau fast ausschließlich auf die äthiopischen und neusüdarabischen Sprachen beschränkt hatte. Die etymologischen Ergänzungen bezogen sich damals auf Ludwig Koehlers Lexicon in Veteris Testamenti libros (Leiden 1953). Unmittelbar nach Erscheinen dieses Aufsatzes erhielt der Verfasser ein Schreiben von Rudolf Meyer aus Jena, in welchem er darum bat, für die in Bearbeitung befindliche 18. Auflage von Wilhelm Gesenius’ Hebräischem und Aramäischem Handwörterbuch über das Alte Testament zum etymologischen Teil das Altsüdarabische beizusteuern. Nach Erledigung dieser Aufgabe folgte einige Zeit später die weitere Bitte, zu den hebräischen Personennamen eventuelle Parallelen aus dem Altnordarabischen zusammenzustellen. Wie die beiden Beispiele auf S. 12 zu hebräisch Āgūr ana. (altnordarabisch) pn (Personenname) ʾgr, und auf S. 180 zu hebräisch Barakʾēl saf. (safatenisch/safaitisch) brkʾl, zeigen, erfolgten die Einträge in äußerst knapper Form ohne jeglichen weiteren Kommentar. Der Verfasser hat sich seinerzeit zu den safaitischen Personennamen, welche Entsprechungen im alttestamentlichen Hebräisch haben, eine Reihe Notizen gemacht, die später gelegentlich ergänzt wurden. Aus diesen Aufzeichnungen wurden im Folgenden eine Reihe von Beispielen ausgewählt, die beim Vergleich eventuell neue Erkenntnisse über die safaitischen oder hebräischen bzw. alttestamentlichen Namen erbringen. In den einzelnen Abschnitten wurde namenkundliches Material aus dem Akkadischen und den nordwestsemitischen Sprachen weitgehend ausgespart, da dasselbe einschließlich der relevanten Sekundärliteratur bei Gesenius fast vollständig erfaßt wurde. Dies gilt auch für Varianten der Namensformen in den antiken Versionen des Alten Testaments und für griechische Transkriptionen semitischer Namen. Besonders berücksichtigt wurden auf Hebräisch überlieferte Namen von Nichtisraeliten, etwa von Ismaelitern, Midianitern, Edomitern und Angehörigen ande-

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

anmerkungen zum safaitischen und althebräischen onomastikon 27

rer südlicher Völkerschaften. In den folgenden Untersuchungen wurden einbezogen das umfangreiche Inschriftencorpus des vorwiegend aus Eigennamen bestehenden Safaitischen, das zum späteren Reichsaramäischen gehörige Palmyrenische mit seinen zahlreichen Personennamen arabischen Ursprungs sowie das Nabatäische, dessen Onomastikon ebenfalls arabische Herkunft verrät, ferner als weitere Sprachen des Altnordarabischen das sogenannte Thamudische und das Lihyanische, sodann die altsüdarabischen Sprachen Sabäisch, Minäisch, Qatabanisch und Hadramitisch, und schließlich das Klassische Arabisch und gegebenenfalls Dialekte der Arabischen Halbinsel. Wertvolle Hilfe beim Anfertigen dieses Aufsatzes leisteten dem Verfasser Dissertationen, deren Themen er gestellt und deren Durchführung er betreut hatte; es handelt sich dabei um die Arbeiten von Yusuf Abdallah, Die Personennamen in al-Hamdānī’s al-Iklīl und ihre Parallelen in den altsüdarabischen Inschriften (1975), Rafat Hazim, Die safaitischen theophoren Namen (1986), Fawwaz Al-Khraysheh, Die Personennamen in den nabatäischen Inschriften des Corpus Inscriptionum Semiticarum (1986), Salem Tairan, Die Personennamen in den altsabäischen Inschriften (1992), Said Al-Said, Die Personennamen in den minäischen Inschriften (1995), Hani Hayajneh, Die Personennamen in den qatabānischen Inschriften (1998), Alexander Sima, Die lihyanischen Inschriften von al-ʿUḏayb (1999), und Amida Sholan, Frauennamen in den altsüdarabischen Inschriften (1999). In seinem Aufsatz Das Alte Testament im Lichte der safatenischen Inschriften, in Zeitschrift der Deutschen Morgenländischen Gesellschaft 104 (1954), S. 88–118, weist Otto Eissfeldt zu den gesellschaftlichen Verhältnissen, wie sie besonders die Jakob-Laban-Erzählungen in der Genesis widerspiegeln, auf zahlreiche Parallelen in den besagten Inschriften hin, welche in knapper Form vom Alltagsleben der beduinischen Bevölkerung berichten; er vermerkt dazu, daß, auch wenn Altes Testament und Ṣafā-Inschriften zeitlich nicht zusammengehören, sie doch hinsichtlich des Raumes viel miteinander gemeinsam haben. In Bezug auf das Namenselement Natan hat bereits Julius Wellhausen, Reste arabischen Heidentums. Zweite Auflage, Berlin 1897, S. 145 f., festgestellt, daß das Wort weniger arabisch als hebräisch sei, und man darin Spuren des Einflusses der alten Bevölkerung zu erblicken habe, als deren Erben die Araber angetreten sind. Dem Verfasser dieses bescheidenen onomastischen Beitrags zur Festschrift für Michael Macdonald ist die kritische Einstellung des Geehrten zu Problemen der Personennamen bekannt, etwa daß ein Name lediglich seinen Träger bezeichnet, daß die Etymologie eines Namens kein Beweis für die Sprache oder Stammes- bzw. Volkszugehörigkeit seines Trägers zu sein braucht, daß Namen oft eine Langlebigkeit aufweisen und über ein ursprünglich begrenz-

28

müller

tes Gebiet hinaus auftreten können, oder die Beibehaltung oder Verbreitung von Namen durch das religiöse Bekenntnis ihrer Träger. Darüber kann man sich in seinen neun umfangreichen und gehaltvollen Aufsätzen informieren, welche in dem Sammelband M.C.A. Macdonald, Literacy and Identity in PreIslamic Arabia, zusammengefaßt wurden und die 2009 in der Reihe Variorum Collected Studies Series erschienen sind. Der Beitrag, für den absichtlich ein das Altnordarabische betreffendes Thema ausgewählt wurde, möge als Zeichen der Dankbarkeit für eine jahrzehntelange unverbrüchliche Freundschaft und der Wertschätzung für seine meisterhaften Abhandlungen über die antike Geschichte Nord arabiens gewertet werden. Vielleicht findet Michael Macdonald in den Anmerkungen zum safaitischen und althebräischen Onomastikon doch noch die eine oder andere Notiz, der er zustimmen kann, etwa die Verknüpfung des safaitischen theophoren Personennamens šnrʾl mit der außerhebräischen Bezeichnung des Gottesberges Śěnīr im Alten Testament. Saf. tham. ʾdbʾl (Harding 1971:31); „Gott hat gut erzogen“ (Hazim 1986:2; Shatnawi 2002:644); „Gott lädt ein“ (wwm); hebr. Adběʾēl, Sohn Ismaels und Name eines arab. Stammes (Gesenius 1987–2010:13); „Gott lädt ein“ (hal 1967– 1995:11); asa. (sab.) ʾdbʾl (Tairan 1992:60); „Gott hat sich freigebig erwiesen“ (Hommel 1926:556, Anm. 8); die Übersetzung folgt arab. adaba „(zum Mahl) einladen“. Saf. ʾlʾt, arab. Alāʾa (Harding 1971:63); „Gott ist gekommen“ (wwm); hebr. Ǟlīʾātāh, entweder ēl + ʾth oder wahrscheinlich Kunstname Ēlī attāh „Mein Gott bist du“ (Gesenius 1987–2010:63); „Gott ist gekommen“ (Rechenmacher 2012:131). Saf. ʾṣr, arab. iṣr „Vertrag, Bündnis“ (Harding 1971:51); hebr. Ēṣär, edomitischer Personen- und Sippenname (Gesenius 1987–2010:93); Ēṣär ist eine Nominalform des Typs qitl und würde somit arab. iṣr entsprechen. Saf. ʿḏrʾl (Harding 1971:412); „Gott hilft“ (Grimme 1929:149); hebr. ʿĂzarʾēl „Gott hat geholfen“ (Gesenius 1987–2010:949; Rechenmacher 2012:144); lihy. ʿḏrʾl, ʿzrʾl „Gott hat geholfen“ (Sima 1999:61); die Deutung der saf. und lihy. Personennamen ʿḏrʾl als „Gott hat geholfen“ beruht nicht auf dem Arabischen, sondern auf dem Hebräischen bzw. Nordwestsemitischen (Müller 1980:72; Hazim 1986:84f.). Saf. tham. bdn, arab. Badan (Harding 1971:98); hebr. Bědan, Etymologie umstritten; ? Hypokoristikon < *bad + Gottesname (Gesenius 1987–2010:126); „In der Hand des x“? (Rechenmacher 2012:166); Etymologie strittig; der häufig belegte saf. Personenname bdn spricht für semit. Herkunft; vgl. dazu arab. badan „alte Bergziege“ (statt: „älterer Steinbock“) (A.R. Müller in nbl 1991– 2001:254); arab. Badan (Caskel 1966:218); arab. badan „Wildziege der arabischen Berge“ (Lane 1863–1874:169).

anmerkungen zum safaitischen und althebräischen onomastikon 29

Saf. brk, arab. Bark (Harding 1971:102); arab. Bark, Bārik (Littmann 1943:303); Barīk „Gesegneter“ (wwm); hebr. Bārūk, Baruch (Gesenius 1987–2010:173f.); Partizip Passiv „gesegnet, Gesegneter“ (Noth 1928:183); „x hat gesegnet“, qāttūlForm, d.h. Hypokoristikon des Verbs brk + theophores Element (Rechenmacher 2012:62.150); palm. bryk „gesegnet (von n.n.)“, aram. Partizip Passiv (Stark 1971:79); asa. (spätsab.) bryk, Barīk, Name einer Synagoge (mkrb) (Müller 2010a:149); arab. barīk = mubārak „gesegnet“. Saf. brkt, arab. Baraka (Harding 1971:102); „Segen“; hebr. Běrākāh, Hypokoristikon zu Barakʾēl, (Gesenius 1987–2010:180); „x hat gesegnet“ (Rechenmacher 2010:150); arab. Baraka (Caskel 1966:224), auch Frauenname Baraka „Glück“ (Gratzl 1906:61); vgl. hebr. běrākāh „Segen“ und arab. baraka „Segen, Glück, Fülle“. Saf. bslm, bslmh, zu slm (Harding 1971:106); Bisalāmih „in seinem Frieden“ (Littmann 1943:303); hebr. Bišlām, herkömmlich Personenname Bischlam, persischer Beamter in Palästina, oder aramaisierend bišlām „im Einvernehmen“, lxx ἐν εἰρήνῃ (Gesenius 1987–2010:184); zu den verschiedenen vorgeschlagenen Etymologien des Namens s. abd 1992:i 750; Bišlām wird nur möglicherweise als Name eines persischen Beamten angesehen und als Präposition b + Nomen šlm interpretiert, während Deutungen aus dem Iranischen als unzutreffend angesehen werden (Hutter 2015:41f.); asa. (min.) bslm wird Bāsilum vokalisiert und wie asa. (sab. min. hadr.) bsl mit dem arab. Namen Bāsil zusammengestellt (AlSaid 1995:71). Saf. tham. ḏʾb, arab. Ḏiʾāb (Harding 1971:246); arab. Ḏiʾb (Shatnawi 2002:689); arab. Ḏiʾb „(Schakal)wolf“ (Nöldeke 1904:79); hebr. Zěʾēb „Wolf“, Name eines Midianiterfürsten (Gesenius 1987–2010:291); nabat. dʾbw, Diʾbu „Schakalwolf“ (Al-Khraysheh 1986:58); asa. (sab. min.) ḏʾb, Ḏiʾb (Al-Said 1995:103f.); (sab. qat.) ḏʾbm, Ḏiʾbum (Tairan 1992:115); syr. dyb, Dēb/Dīb, Name eines himjarischen Märtyrers (Moberg 1924:24b); arab. Ḏiʾb, Ḏiʾāb (Caskel 1966:235). Saf. ḏmrt (Harding 1971:287); „beschützt“ (wwm); tham. ḏmr, Ḏamār „tapfer (als Beschützer)“ (Shatnawi 2002:689); hebr. Zěmīrāh, von der Wurzel *zmr < *ḏmr (Gesenius 1987–2010:304); „beschützt“? (Rechenmacher 2012:180); asa. (qat. hadr.) ḏmrm „Beschützer“ oder „beschützt“ (Hayajneh 1998:142); vgl. asa. (sab.) ḏmr „schützen, verteidigen“, wonach der Bestandteil ḏmr in den zusammengesetzten Namen ḏmrʿly, ḏmrkrb, ḏmrydʿ, u. a. als „Beschützer“ gedeutet wird; arab. Ḏamār, (Caskel 1966:235); arab. ḏimr „tapfer“ (als Beschützer). Saf. tham. ḍbʿn, arab. Ḍibʿān (Harding 1971:381), Ḍabuʿān (Shatnawi 2002:716), „männliche Hyäne“ (Nöldeke 1904:79); hebr. Ṣibʿōn, Name eines seïritischen Stammes in Edom (Gesenius 1987–2010:1100); asa. (sab. qat.) ḍbʿn „(männliche) Hyäne“ (Hayajneh 1998:183); arab. Ḍabuʿān (Abdallah 1975:73), Ḍibʿān (Caskel 1966:241).

30

müller

Saf. flṭʾl (Harding 1971:471); „Gott läßt entkommen“ (Grimme 1929:149); hebr. Palṭīʾēl „(meine) Rettung ist Gott“ (Gesenius 1987–2010:1054), „Gott ist Rettung“ (Rechenmacher 2012:119); nabat. plṭʾl (Negev 1991:55); die Übersetzung der saf. und nabat. Personennamen flṭʾl, plṭʾl mit „Gott ließ entkommen, Gott errettete“ ergibt sich nicht aus dem Arabischen, sondern aus dem Hebräischen bzw. Nordwestsemitischen (Müller 1980:72); Gleiches ist für den Namen flṭ anzunehmen. Saf. gdʾl (Harding 1971:154); Gaddʾēl „Gott macht glücklich“ (Littmann 1943:304); Gaddʾil „Gadd ist Gott“ (Hazim 1986:20); hebr. Gaddīʾēl (Gesenius 1987–2010:209); „Gad ist Gott“ (Noth 1928:126); „Gott ist Glück“ (Rechenmacher 2012:116). Der in der Genealogie von Sabaʾ sich findende Personenname Ǧadʾal (al-Hamdānī 1966:106,5) dürfte aus Ǧaddʾil entstellt sein und somit einen Hinweis auf das Vorkommen dieses Namens auch im Sabäischen geben; darauf läßt auch der ebenfalls bei al-Hamdānī belegte Name ʿAbdalǧadd (Wellhausen 1897:2; Caskel 1966:124) „Diener des Ǧadd“ schließen. Der Umstand, daß die Glücksgottheit Gad (griech. Tyche, lat. Fortuna) im Alten Testament (Jes. 65:11) begegnet und daß der Gott gd auch zum Pantheon der Ṣafā-Beduinen zählt, spricht dafür, daß unter dem ersten Bestandteil der Namen Gaddīʾēl bzw. gdʾl auch jene Gottheit zu verstehen sein dürfte. Saf. tham. gšm, arab. Ǧušam (Harding 1971:162); hebr. Gäšäm, Gašmū, Name eines Arabers, Scheich einer arab. Stammesföderation in Südpalästina und im nördlichen Ḥiǧāz (Gesenius 1987–2010:232); nabat. gšmw, Gušamu (Al-Khraysheh 1986:57); lihy. gšm, Gušam (Caskel 1954:146); asa. (sab. qat. hadr.) gšm, Gušam (Hayajneh 1998:113); arab. Ǧušam (Caskel 1966:267–269); der Name Gušam ist im Altnordarabischen und Altarabischen häufig bezeugt, allein bei Caskel, op. cit., werden 118 Personen dieses Namens aufgeführt. Die Etymologie von Ǧušam ist umstritten, und die von den arab. Lexikographen gegebene Erklärung als „Kamelbrust“ ist wohl kaum zutreffend. Da es sich um einen arab. Namen handelt, verbietet sich sowohl die Deutung von Gäšäm/Gašmū nach hebr. gäšäm „Regen“ als „zur Regenzeit geboren“ (hal 1967–1995:197) als auch diejenige von „großer, bedeutender Mann“ nach arab. ǧasuma „groß, stark, gewaltig sein“ (abd 1992:ii 995). Saf. ġfr (Harding 1971:457); hebr. ʿĒpär, vielleicht Bezeichnung eines Capriden-Jungtieres (Gesenius 1987–2010:997); das Junge der Gazelle und des Hirsches (Nöldeke 1904:84); diese Deutungen beruhen auf hebr. ʿopär „Jungtier, Kalb, Kitz von Hirsch, Gazelle oder Reh“ (Gesenius, ebd.) und arab. ġufr „Junges des Steinbocks“ (Nöldeke, ebd.). Da die Form des hebr. Personennamens in der ersten Silbe jedoch ursprünglich nicht den Vokal u, sondern i aufweist, sei zum Vergleich eher arab. ġifr herangezogen, das wie folgt erklärt wird (Lane 1863–1874:2274): „The young of the cow (probably meaning of the bovine ante-

anmerkungen zum safaitischen und althebräischen onomastikon

31

lope called the wild cow)“, d.h. der Oryx(antilope), also „Junges der Oryx“; asa. (sab.) ġfr, (sab. hadr.) ġfrm; arab. Ġufr (Caskel 1966:599). Saf. ġrb (Harding 1971:453); Ġurāb „Rabe“ (Littmann 1943:338); hebr. ʿOrēb = ʿorēb „Rabe“, Name eines Midianiterfürsten (Gesenius 1987–2010:1010; Nöldeke 1904:85; Rechenmacher 2012:171); zu ʿOrēb und Zěʾēb s. G.E. Mendenhall (abd 1992:v 42); asa. (qat.) Sippenname ġrbm, Ġurābum (Ja 191 u. ö.); arab. Ġurāb „Rabe“ (Caskel 1966:275). Saf. tham. ḥbb (Harding 1971:172); Ḥubāb „Schlange“ (Shatnawi 2002:668); hebr. Ḥobāb „Liebling“(?) oder Gottesname (Gesenius 1987–2010: 317); = Ḥubāb = „Schlange“, ursprünglich Gottesname (Kerber 1897:72); Name eines Midianiters; lihy. ḥbb „Geliebter“ (Caskel 1954:146); asa. (sab.) ḥbb, Ḥubāb, ḏḥbb, Ḏū-Ḥubāb „Schlange“ (Tairan 1992:96); arab. Ḥubāb (Caskel 1966:326; Abdallah 1975:42); Th. Nöldeke war der Meinung, hebr. Ḥobāb dürfe man nicht als „Schlange“ auffassen, denn ḥubāb „Liebe“ für Schlange sei ein speziell arab. Euphemismus (Nöldeke 1904:87, Anm. 4); dagegen sprechen allerdings die zahlreichen von der Wurzel ḥbb gebildeten Wörter für Schlange in den äthiosemitischen Sprachen, wie hararī ḥubāb, amharisch ǝbab, u. a.; ob sich in den Personennamen Ḥobāb und Ḥubāb ein Gottesname verbirgt (Wellhausen 1897:146), sei dahingestellt. Saf. ḫṭs, Wurzel unbekannt (Harding 1971:223); ḫṭsʾl (Oxtoby 1968:103, Nr. 403); hebr. Ḥaṭṭūš, Kurzform eines Namens aus einer unbekannten Wurzel ḥṭš + Gottesname (Gesenius 1987–2010:341). Für viele Namen der Nominalform qattūl finden sich im biblischen Hebräisch Entsprechungen der Vollform der Suffixkonjugation + theophorem Element (Rechenmacher 2012:61 f.). Daß auch Ḥaṭṭūš ein Hypokoristikon zu einem nicht belegten Ḥǎṭašʾēl oder Ḥǎṭašyāh sein könnte, wird nahegelegt durch das Vorhandensein von ḫṭsʾl neben häufigem ḫṭs im Safaitischen (Müller 1980:72). Eine Verbalwurzel ḫṭš ist in den Lexika der semit. Sprachen nicht mehr zu belegen und scheint nur noch in Relikten in Personennamen des Hebräischen und Altnordarabischen weiterzuleben; eine Bedeutung läßt sich daher nicht mehr ermitteln. Saf. tham. klb, Kalb (Harding 1971:502); Kalb „Hund“, Kilāb (Shatnawi 2002:735); hebr. Kālēb, a) hypokoristisch für „Hund (im Sinne von Diener) der Gottheit“, b) „hundswütig, toll“ (Gesenius 1987–2010:543); „Hund des x“ (Rechenmacher 2012:35.164); palm. klbʾ „Hund“ (Stark 1971:92); nabat. klbw, Kalbu (Al-Khraysheh 1986:100); lihy. klb (Sima 1999:70); asa. (sab. min. qat.) klb „Hund“ oder „Diener (des Gottes n.n.)“ (Hayajneh 1998:221); asa. klbm (Tairan 1992:187); (sab.) klbnsrm, Kalbnasrim „Hund (d.h. treuer Diener) des (Gottes) Nasrum“, res 4188,2 (statt klbnsšm) und in einer von S. Tairan veröffentlichten Inschrift (Müller 2014:164f.); arab. Kalb, Kilāb (Caskel 1966:369f.); arab. kalb, pl. kilāb „Hund“, kalib „tollwütig“.

32

müller

Saf. mgdʾl (Harding 1971:528); „Lobpreisung bzw. Herrlichkeit Gottes“ (Hazim 1986: 109f.); hebr. Magdīʾēl, < mägäd + ēl, Name eines edomitischen Stammesfürsten (Gesenius 1987–2010:626); „Ruhm Gottes“ (wwm); die Übersetzung von Magdīʾēl durch „Geschenk Gottes, Gottesgabe“ nach hebr. mägäd von U. Hübner (abd 1992:iv 464) scheint weniger angebracht zu sein, da es sich bei dem Namensträger um einen Edomiter handelt; asa. (sab.) Frauenname mgdlt „Ruhm der Lat“ (Sholan 1999: 133); syr. Frauenname mgdl (Drijvers 1972:26). Saf. tham. mlkʾl (Harding 1971:565); Mālikʾil (Hazim 1986:117f.); „Il ist Herrscher“ (Shatnawi 2002:745); hebr. Malkīʾēl, < mäläk + ēl (Gesenius 1987– 2010:687); „König ist Gott“ (Rechenmacher 2012:57); palm. mlkʾl (Stark 1971:95); asa. (altsab.) mlkʾl „Gott ist Herrscher“ (Tairan 1992:204f.). In 1 Makk. 11:39 wird ein Araber namens Ιμαλκουε erwähnt. Im Artikel Imalkue (abd 1992:iii 391) führt A. Kasher diese griech. Namensform auf den arab. Namen ymlkw zurück, und auch Übersetzungen aus neuerer Zeit haben textwidrig und ohne Kommentar Jamliku. Die Vokalisierung Imalkue spricht allerdings gegen die Herleitung des Namens von der Präfixkonjugation des Verbs. Die richtige Deutung des Namens bietet wohl die Vulgata mit der ursprünglichen Form Emalchuel Arabem, was mlkʾl, *Mālikuʾēl, entsprechen dürfte. Saf. tham. mlkt (Harding 1971:565); Malikat (Winnett and Harding 1978:613); hebr. Frauenname Milkāh, < malkāh „Königin“ (Gesenius 1987–2010:687); „Herrscherin“ (abd 1992:iv 824f.); palm. Frauenname mlkt „Königin“ (Stark 1971:95); nabat. Frauenname mlkt (Negev 1991:39); lihy. mlkh, Mālikah (Caskel 1954:149); has. Frauenname mlkt (Sima 2002:172); asa. (min.); unter den ausländischen Ehefrauen, die von Minäern geheiratet wurden, befinden sich vier, welche den Namen mlkt tragen (Al-Said 1995:222); zwei von ihnen stammen aus ġzt, Ghazza, die beiden anderen aus tmlḥ, das mit dem ostarabischen Gerrha zu identifizieren sein dürfte. Die inschriftlichen Belege dieser Frauennamen sind wohl nach arab. Mālika (Aš-Šammarī 1990:661) „Herrscherin, Besitzerin, Herrin“ zu vokalisieren und zu deuten. Saf. mslmʾl (Harding 1971:546; Hazim 1986:59); „Ersetzend ist Gott“ (wwm); hebr. Měšälämyāh „Jahwe ersetzt, ersetzt durch Jahwe“ (Gesenius 1987– 2010:157); „Jahwe ist Entschädigung“ (abd 1992:iv 711 f.); „Ersetzend ist Jahwe“ (Rechenmacher 2012:44.91.124); s. auch slmʾl. Saf. ngh, arab. naǧaha „abweisen“ (Harding 1971:582); „Glanz“ (wwm); hebr. Nogah = nogah „heller Schein, Glanz“ (Gesenius 1987–2010:779); „x ist Glanz“ (Rechenmacher 2012:116); asa. (min.) nght „Glanz“, Name einer fremdländischen, von einem Minäer geheirateten Frau aus ġzt, Ghazza (Al-Said 1995:222). Saf. tham. ntn (Harding 1971:581); Natan (Winnett and Harding 1978:614); „(Gott n.n.) hat gegeben“ (Shatnawi 2002:745); hebr. Nātān, Kurzform von der Wurzel ntn + Gottesname, vgl. Nětanʾēl, Nětanyāhū (Gesenius 1987–2010:863);

anmerkungen zum safaitischen und althebräischen onomastikon 33

„(Gott) hat gegeben“ (Rechenmacher 2012:156); palm. ntny „n.n. hat gegeben“ (Stark 1971:101); nabat. ntn (Negev 1991:45); lihy. ntn, ntnbʿl „Baʿl hat gegeben“ (Caskel 1954:84); qaht. aus Qaryat al-Fāw ntn (Tairan 1992:215); nordmin. ntn „(der Gott n.n.) hat gegeben“ (Al-Said 1995:167); Natnu begegnet als Name eines Araberfürsten in den Annalen des Assurbanipal. Wie eingangs vermerkt, hatte bereits Julius Wellhausen festgestellt, daß das Namenselement Natan weniger arabisch als hebräisch sei und somit den Einfluß der alten Bevölkerung zeige (Müller 1980:72). Saf. qdmʾl (Harding 1971:478); Qadamʾil (Winnett and Harding 1978:603); „Gott ist vorangegangen“ (Hazim 1986:102f.); hebr. Qadmīʾēl, Etymologie unsicher; (?) Wurzel qdm + Endung –ī + ēl, „Gott ist (mir) entgegengekommen“ oder Nomen qädäm + Suffix oder Bindevokal –ī + ēl, „Was vor (mir) ist, ist El“ (Gesenius 1987–2010:1149); „Gott ist vorangegangen“ (Rechenmacher 2012:131); für die Analyse des hebr. Namens spricht asa. yqdmʾl (ebd. 133); asa. (sab. min.) yqdmʾl, Yaqdumʾil „Gott möge vorangehen“ (Al-Said 1995:185). Saf. rfʾl (Harding 1971:283); rfʾl und rfʾʾl, Rafaʾʾil „Gott hat geheilt“ (Hazim 1986: 46f.); hebr. Rěfāʾēl Wurzel rfʾ + ēl, „Gott hat geheilt“ (Gesenius 1987– 2010:1261; Rechenmacher 2012:148); lxx (Tob. 1:1) Ραϕαηλ; palm. rpʾl „Gott hat geheilt“ (Stark 1971:112). Vielleicht sind Namen, die mit dem Verb rfʾ „wiederherstellen, heilen“ gebildet sind, zu den Personennamen zu rechnen, die einen Ersatz zum Ausdruck bringen (Nöldeke 1904:100); eventuell sind auch die Namen mit rpʾ in Bezug auf ein verstorbenes Familienmitglied zu deuten (Rechenmacher 2012:148). Saf. rkb, arab. Rakb (Harding 1971:484); arab. Rākib (Winnett and Harding 1978:578); hebr. Rēkāb (Gesenius 1987–2010:1243f.); der Name ist nach seiner sprachlichen Form und in seiner Bedeutung unerklärt (hal 1963–1995:1152); „(Reit)kamel“ (wwm); asa. (sab.) rkb (Harding 1971:484); sab. rkb „Reitkamel“; arab. Rikāb (maa 1991:980f.); arab. rikāb „Last- oder Reitkamel“. Saf. sbq, arab. Sabbāq (Harding 1971:309); arab. Sābiq, Sabbāq (Winnett and Harding 1978:580); hebr. Šōbēq, Etymologie unsicher […] oder (?) zu arab. sabaka „übertreffen“, sābik „Sieger“ (Gesenius 1987–2010:1331), wo in der (aus Noth 1928:231) übernommenen Passage die arab. Wörter zu sabaqa und sābiq zu korrigieren sind; „Sieger“ (? nach arab. sabaqa „übertreffen“) (Rechenmacher 2012:182); arab. Sābiq (maa 1991:762); Sabbāq (Caskel 1966:492); arab; sābiq „vorangehend, Sieger“; sabbāq „den Vorrang habend“. Saf. sḫrʾl, arab. saḫira „(ver)spotten“ (Harding 1971:312f.); Saḫarʾil (Winnett and Harding 1978:581); Saḫiraʾil „Gott hat (den Feind) verspottet“ (Hazim 1986:55); hebr. Šěḥaryāh, Etymologie unsicher; šaḥar + Jahwe „Die Morgenröte ist Jahwe“ (wozu die Zusammenstellung mit saf. sḫrʾl nicht paßt) (Gesenius 1987–2010:1343); man vergleiche jedoch die Bemerkung, daß Spuren von šaḥar

34

müller

„Morgenröte“ (anders als in asa. sḥr) als Gottesname im Alten Testament nicht nachzuweisen sind (Gesenius, ebd.); „Morgenröte ist Jahwe“ (Rechenmacher 2012:53). Biblische Stellen wie „der Herr verspottet sie“ (Ps. 2:4) oder „du spottest über alle“ (Ps. 59:9), allerdings unter Verwendung des Verbs lāʿag „(ver)spotten“, würden einen Vergleich mit saf. sḫrʾl nicht gänzlich ausschließen. Saf. tham. slmʾl (Harding 1971:325); Sallamʾēl „Gott erhält unversehrt“ (Littmann 1943:332); Salāmʾil „Unversehrtheit (die dem Kind) von Gott (gegeben wird)“ (Hazim 1986:58); „Heil des Il“ oder „Il hat heil bewahrt“ (Shatnawi 2002:705); vielleicht „Ersetzt hat Gott“ (wwm); hebr. Šälämyāhū, Šälämyāh „Jahwe hat ersetzt“ (Gesenius 1987–2010:1370); „Ersetzt hat Jahwe“ (Rechenmacher 2012:134); asa. (hadr.) slmʾl (Francuzov 2009:126); (qat.) slml „Il hat heil bewahrt“ oder „Frieden bzw. Heil des Il“ (Hayajneh 1998:162). Es ist nicht auszuschließen, daß die saf. und tham. Namen slmʾl entsprechend dem Hebräischen mit „Gott hat ersetzt“ zu übersetzen sind; s. auch mslmʾl. Saf. smkʾl (Harding 1971:329); „Gott unterhält“ (Littmann 1943:332); hebr. Sěmakyāhū „Jahwe hat geholfen“ (Gesenius 1987–2010:892); „Jahwe hat unterstützt“ (abd 1992:v 1077); „Jahwe hat gestützt“ (Rechenmacher 2012:144). Da mit dem verbalen Element smk gebildete theophore Namen sonst nur im Nordwestsemitischen begegnen, scheint es angebracht, auch den saf. Personennamen smkʾl nach dem Hebräischen als „Gott hat gestützt, Gott hat unterstützt“ zu deuten (Müller 1980:72). Saf. šnrʾl (al-Manaser 2008:240); Šanīrʾil (Müller 2010b:522f.); hebr. Śěnīr, Bergname, hebraisierende Form einer außerhebräischen Bezeichnung für die Region des Hermon und des Antilibanon oder für einen Einzelgipfel (Gesenius 1987–2010:1293). Da die vom Herausgeber der Inschrift Nr. 376 für die Erklärung des theophoren Namens šnrʾl herangezogenen arab. Wörter der Wurzel šnr, wie šanār „Schande, Unehre“, šannār „ein weißer Vogel“, u. a., sämtlich unbrauchbar sind, wurde erwogen, in šnr den im Alten Testament genannten Śěnīr zu sehen, der ein amoritischer Name für den transjordanischen Berg Hermon ist, zumal der Hermon zu den Gottesbergen zählte, wo die im Alten Testament erwähnte lokale Gottheit Baʿal Ḥärmōn (Gesenius 1987–2010:163) verehrt wurde (Müller 2010b:522f.). Vom Fuße des Berges Hermon stammt angeblich ein Stein mit vier saf. Namensinschriften, die von Michael Macdonald veröffentlicht wurden (Macdonald 2009: Addenda and Corrigenda 2 f.). Saf. šrḥ, arab. Šuraḥ (Harding 1971:345); Šurāḥ, Šurḥ, Šarḥ, Šaraḥ (Winnett and Harding 1978:585); wohl Kurzform von šrḥʾl „Gott hat beschützt“; hebr. Frauenname und Sippenname Śäraḥ, Etymologie unsicher; ? Wurzel śrḥ „aufrichten, kundtun, erklären“ (Gesenius 1987–2010:1299); eine Übersetzung des Namens nach asa. šrḥ, allerdings mit der dem Äthiopischen entnommenen Bedeutung „gedeihen lassen“, war bereits früher vorgenommen worden (Hom-

anmerkungen zum safaitischen und althebräischen onomastikon 35

mel 1897:82); „(Gott) hat beschützt“ (Al-Said 1995:124); lihy. šrḥ, Šarḥ (Caskel 1954:152); asa. šrḥ „(Gott) hat beschützt“ (Al-Said 1995:123f.; Hayajneh 1998:167); vgl. asa. šrḥʾl, Šaraḥʾil „Gott hat beschützt“ (Tairan 1992:136f.) und spätsab. šrḥbʾl, Šuraḥbiʾil „Beschützt von Gott“; arab. Šarāḥ, Šaraḥ, Šuraḥ, Šarḥ, Šurḥ, Šarīḥ (Abdallah 1975:68f.); Šarāḥīl, Šuraḥbīl (Caskel 1966:527.532). Die Deutung der Personennamen ergibt sich aus sab. šrḥ „bewachen, beschützen“ und jemen.arab. šaraḥa „hüten, bewachen“. Saf. tham. tmnʿ, Wurzel mnʿ (Harding 1971:138); „Sie (eine ungenannte Göttin) möge abhalten bzw. fernhalten (von Gefahr)“ (Shatnawi 2002:662); hebr. Timnaʿ, Imperfekt der 3. Person Femininum Singular Qal von der Wurzel mnʿ, „Sie (eine Göttin) schützt/hat geschützt“ (Gesenius 1987–2010:1445f.). Die saf. und tham. Namen von Männern sind sicherlich ebenso zu deuten wie die biblischen, von zwei edomitischen Frauen getragenen Namen, nämlich „Sie möge abhalten“. Saf. trḍy, arab. raḍīy „zufrieden, einwilligend“ (Harding 1971:131); hebr. Tirṣāh, Frauenname, Etymologie unsicher, Wurzel rṣh (?) (Gesenius 1987–2010:1459); nabat. trṣw (Negev 1991:69); lihy. rḍwʾl „Gott hat Wohlgefallen“ bzw. „Wohlgefallen Gottes“ (Sima 1999:75); asa. rḍwʾl „Wohlgefallen bzw. Zufriedenheit Gottes“ (Al-Said 1995:112; Hayajneh 1998:153). Da eine Deutung des von einer Hebräerin getragenen Namens Tirṣāh als „Sie (eine Göttin) hat Wohlgefallen“ kaum möglich ist, sei erwogen, ob es sich bei tirṣāh nicht um ein anderweitig unbezeugtes Nomen mit Präformativ t- der Bildung tiqtal (analog zu tiqwāh „Hoffnung“) handeln könnte, wie dies bereits für den homonymen Ortsnamen Tirṣāh in Cant. 6:4 mit der Bedeutung „Vergnügen, Schönheit“ angenommen wurde (abd 1992:v 573). Somit könnte im Frauennamen Tirṣāh eine Kurzform für „Wohlgefallen bzw. Wohlwollen (Gottes)“ vorliegen. Saf. ṯfn, arab. ṯafana „zurückstoßen“ (Harding 1971:146); Ṯafan „Klippschliefer“ (wwm); hebr. Šāfān = šāfān „Klipp(schliefer)dachs“ (Gesenius 1987– 2010:1405); „Klippschliefer“ (Rechenmacher 2012:171); asa. (hadr.) ṯfnm (M.B. Piotrovskij in Vestnik Drevnej Istorii 1989,2, S. 151), Ṯafanum. Hebr. šāfān „Klippschliefer, Hyrax, Procavia syriacus“ hat eine etymologische Entsprechung im nsa. Mehri ṯōfen und Śḥeri ṯófun; dies erlaubt es, den in einem Graffito aus dem Mahraland vorkommenden hadr. Personennamen ṯfnm als Hyrax zu deuten, und höchstwahrscheinlich auch saf. ṯfn, ehe das Wort durch arab. wabr „Klippschliefer“ verdrängt wurde, das ebenfalls als Name in Wabr (Caskel 1966:581) belegt ist. Saf. yʾsʾl, zu arab. assa „bauen, gründen“ oder aws „Gabe“ (Harding 1971:655); Yaʾūsʾil „Gott gibt Ersatz“ (Hazim 1986:4); tham. yʾwsl, Yaʾūs(ʾ)il „Il gibt Ersatz“ (Shatnawi 2002:755); hebr. Yōʾšiyyāhū, Wurzel ʾšy „heilen“ (vgl. arab. asā „pflegen, behandeln“) + Jahwe (Gesenius 1987–2010:432); Josiah, die Wurzel des

36

müller

Namens ist wahrscheinlich ʾwš „geben“ (abd 1992:iii 1015); „Geheilt hat Jahwe“, zu ʾšy (oder ʾwš „geben“?) (Rechenmacher 2012:84); lihy. yʾwsʾl, Yuʾauwisʾil „Gott gibt Ersatz“ (Caskel 1954:154); asa. (sab. min.) yʾwsʾl, Yuʾāwisʾil „Gott gibt Ersatz oder schenkt“ bzw. „Gott hat Ersatz gegeben (nämlich für ein verstorbenes Kind)“ (Tairan 1992:234). Die tham., lihy. und asa. Schreibweisen des Namens zeigen, daß man es auch im Safaitischen und Hebräischen mit einem Ersatznamen zu tun haben dürfte. Saf. tham. ydʿʾl (Harding 1971:664); Yadaʿʾēl „Gott kennt bzw. weiß“ (Littmann 1943:319); Yadaʿʾil „Gott weiß“ bzw. „Gott kümmert sich (um den Namensträger)“ (Hazim 1986:137f.); „Il weiß bzw. hat erkannt“ oder „Bekannter des Il“ (Shatnawi 2002:756); hebr. Yědaʿyāh, Wurzel ydʿ + Jahwe (Gesenius 1987–2010: 444); „Erkannt hat Jahwe“ (Rechenmacher 2012:80); „Bekanntlich hat ydʿ, zumal wenn es von der Gottheit gesagt wird, im Hebräischen nicht selten die Bedeutung: sich kümmern, sich annehmen“ (Noth 1928:181); hebr. Yědīʿǎʾēl, Partizip Passiv von ydʿ + ēl „der Bekannte Gottes“ (Gesenius 1987–2010:441); „Gekannt von Gott“ (Rechenmacher 2012:52); asa. ydʿʾl, Yadaʿʾil „Gott hat erkannt, weiß“ (Tairan 1992:239f.); „Gott hat erkannt“ bzw. „Bekannter Gottes“ (Hayajneh 1998:273). Da im Arabischen die Wurzel ydʿ im Grundstamm des Verbs nicht vorhanden ist, beruht die Erklärung der saf. und tham. Personennamen ydʿʾl auf dem Hebräischen bzw. Nordwestsemitischen; der im Arabischen überlieferte Name Yadāʿ (Abdallah 1975:98) bezeichnet eine Person in einer himjarischen Genealogie. Vgl. die griech. Personennamen Theognostos und Diognostos gleicher Bedeutung. Saf. zbdʾl (Harding 1971:294); „Geschenk Gottes, Gabe Gottes“ (Hazim 1986:50); hebr. Zabdīʾēl, Wurzel zbd + ēl (Gesenius 1987–2010:292); „Geschenk Gottes“ (Rechenmacher 2012:165); die lxx erwähnen einen Araber namens Ζαβδιηλ (1Makk. 11:17); palm. zbdlh „Geschenk Gottes“ (Stark 1971:85f.). Auch die Deutung der saf., arab. und palm. Personennamen als „Geschenk Gottes“ ergibt sich nicht aus dem Arabischen, sondern aus dem Hebräischen und Nordwestsemitischen (Müller 1980:72). Dies gilt ebenfalls für saf. und tham. zbd (Harding 1971:294; Shatnawi 2002:699), nabat. zbdw (Al-Khraysheh 1986:72), palm. zbd, zbdʾ (Stark 1971:85) und lihy. zbd (Caskel 1954:154), Zabd „Geschenk (Gottes)“; arab. Lexika erklären zabd durch wahb „Gabe, Geschenk“; in griech. Transkriptionen ist der Name belegt als Ζαβδος (Grassi 2012:198). Vgl. zu zbdʾl den griechischen Personennamen Theodoros gleicher Bedeutung. Saf. zbnʾ (Harding 1971:294); „Beschützter“ (wwm); hebr. Zěbīnāʾ, Partizip Passiv des Grundstammes der aram. Wurzel zbn „kaufen“ (Gesenius 1987– 2010:293); „gekauft“ (?) nach aram. zbn „kaufen“ (Rechenmacher 2012:180); onomastisch schwierig (ebd., 180f.); nabat. zbnw, Zabīnu (Negev 1991:25); arab. Zabīna (Caskel 1966:598); in griech. Trankription ist der Name belegt als Ζαβι-

anmerkungen zum safaitischen und althebräischen onomastikon 37

νας (Grassi 2012:195). Auch bei dem nabat. Namen zbynw (wohl Ḏubyānu), der Zabīnu gelesen wurde, wird bei der Deutung auf syr. zbīnā „losgekaufter (Sklave)“ verwiesen (Cantineau ii 1932:91). Es ist jedoch schwierig, die arab. und ana. Namen aus dem Aramäischen erklären zu wollen. In Beduinendialekten der Arabischen Halbinsel begegnen u.a. zentralarab. zabana „Schutz suchen“, zabūn „Beschützer“ (Socin iii 1901:272) und jemen.-arab. zabīn „Zufluchtsuchender“ (Behnstedt 1992–2006:487). Daher sei erwogen, saf. zbnʾ, nabat. zbnw und arab. Zabīna durch „Beschützter“ wiederzugeben, was auch für den singulären hebr. Namen Zěbīnāʾ passend wäre.

Abkürzungen altsab. ana. arab. aram. asa. griech. hadr. has. hebr. jemen.-arab. lihy. min. nabat. nordmin. nsa. palm. qaht. qat. sab. saf. semit. spätsab. syr. tham. zentralarab.

altsabäisch altnordarabisch arabisch aramäisch altsüdarabisch griechisch hadramitisch, ḥaḍramitisch hasaitisch, ḥasāʾitisch hebräisch jemenitisch-arabisch lihyanisch, liḥyānisch minäisch nabatäisch, nabaṭäisch nordminäisch neusüdarabisch palmyrenisch qahtanisch, qaḥṭānisch qatabanisch, qatabānisch sabäisch safaitisch, ṣafāʾitisch, safatenisch semitisch spätsabäisch syrisch thamudisch, thamūdisch zentralarabisch

38

müller

Bibliographie abd. 1992. The Anchor Bible Dictionary, editor-in-chief D.N. Freedman, vol. 1–6. New York: Doubleday. Abdallah, Y. 1975. Die Personennamen in al-Ḥamdānī’s al-Iklīl und ihre Parallelen in den altsüdarabischen Inschriften. Ein Beitrag zur jemenitischen Namensgebung. Dissertation Tübingen. Behnstedt, P. 1992–2006. Die nordjemenitischen Dialekte. Teil 2: Glossar. Wiesbaden: Reichert. Cantineau, J. 1932. Le Nabatéen. ii. Choix de textes – Lexique. Paris: E. Leroux. Caskel, W. 1954.Lihyan und Lihyanisch. (Arbeitsgemeinschaft für Forschung des Landes Nordrhein-Westfalen. Heft 4. Abhandlung). Köln und Opladen: Westdeutscher Verlag. Caskel, W. 1966. Ǧamharat an-nasab. Das genealogische Werk des Hišām ibn Muḥammad al-Kalbī. Band ii. Das Register. Leiden: Brill. Drijvers, H.J.W. 1972. Old Syriac (Edessean) Inscriptions. (Semitic Study Series. No. iii). Leiden: Brill. Francuzov, S.A. 2009. Rajbūn. Nadpisi i ljudi. Sankt-Peterburg: Universitet. Gesenius, W. 1987–2010. Hebräisches und Aramäisches Handwörterbuch über das Alte Testament. 18. Auflage. 1.–6. Lieferung, bearbeitet und herausgegeben von Rudolf Meyer und Herbert Donner. Berlin: Springer. Gratzl, E. 1906. Die altarabischen Frauennamen. Leipzig: Drugulin. Grassi, G.F. 2012. Semitic Onomastics from Dura Europos. The Names in Greek Script and from Latin Epigraphs. (History of the Ancient Near East/Monographs. Vol. xii). Padova: Sargon. Grimme, H. 1929. Texte und Untersuchungen zur safatenisch-arabischen Religion. (Studien zur Geschichte und Kultur des Altertums. 16. Band. 1. Heft). Paderborn: Ferdinand Schöningh. hal. 1967–1995. Hebräisches und Aramäisches Lexikon zum Alten Testament. Von Ludwig Koehler und Walter Baumgartner. Lieferung i–v. Leiden: Brill3. al-Hamdānī, al-Ḥ. 1966. Kitāb al-Iklīl. Al-Ǧuzʾ aṯ-ṯānī, edited by M. al-Akwaʿ. Al-Qāhira: As-Sunna al-Muḥammadīya. Harding, G.L. 1971. An Index and Concordance of Pre-Islamic Arabian Names and Inscriptions. (Near and Middle East Series 8). Toronto: University Press. Hayajneh, H. 1998. Die Personennamen in den qatabānischen Inschriften. Lexikalische und grammatische Analyse im Kontext der semitischen Anthroponomastik. (Texte und Studien zur Orientalistik. Band 10). Hildesheim: Georg Olms. Hazim, R. 1986. Die safaitischen theophoren Namen im Rahmen der gemeinsemitischen Namengebung. Dissertation Marburg. Hommel, F. 1897. Die Altisraelitische Überlieferung in inschriftlicher Beleuchtung. München: G. Franz’sche Hofbuchhandlung.

anmerkungen zum safaitischen und althebräischen onomastikon 39 Hommel, F. 1926. Ethnologie und Geographie des Alten Orients. (Handbuch der Altertums-wissenschaft. Dritte Abteilung. Erster Teil. Erster Band). München: C.H. Beck. Hutter, M. 2015. Iranische Personennamen in der hebräischen Bibel. (Iranisches Personen-namenbuch. Band vii. Iranische Namen in semitischen Nebenüberlieferungen. Faszikel 2. Österreichische Akademie der Wissenschaften. Philosophisch-historische Klasse. Sitzungsberichte, 860. Band). Wien: Verlag der Österreichischen Akademie der Wissenschaften. Kerber, G. 1897. Die religionsgeschichtliche Bedeutung der hebräischen Eigennamen des Alten Testaments. Freiburg i.B.: Herder. Al-Khraysheh, F. 1986. Die Personennamen in den nabatäischen Inschriften des Corpus Inscriptionum Semiticarum. Dissertation Marburg. Lane, E.W. 1863–1874. An Arabic-English Lexicon derived from the best and most copious eastern sources, vol. 1–5. London: Williams and Norgate. Littmann, E. 1943. Safaitic Inscriptions. (Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909. Division iv. Semitic Inscriptions. Section c). Leiden: Brill. maa. 1991. Muʿǧam asmāʾ al-ʿArab. Mawsūʿat as-Sulṭān Qābūs lil-asmāʾ al-ʿArab, vol. 1–2. Bayrūt: Maktabat al-Lubnān. Macdonald, M.C.A. 2009. Literacy and Identity in Pre-Islamic Arabia. (Variorum Collected Studies Series 906). Farnham: Ashgate. al-Manaser, A.Y.Kh. 2008. Ein Korpus neuer safaitischer Inschriften aus Jordanien. (Semitica et Semitohamitica Berolinensia. Band 10). Aachen: Shaker. Moberg, A. 1924. The Book of the Himyarites. Fragments of a hitherto unknown Syriac work. (Acta Reg. Societatis Humaniorum Litterarum Lundensis vii). Lund: C.W.K. Gleerup. Müller, W.W. 1980. „Some remarks on the Safaitic inscriptions.“ Proceedings of the Seminar for Arabian Studies 10: 67–74. Müller, W.W. 2010a. Sabäische Inschriften nach Ären datiert. Bibliographie, Texte und Glossar. (Akademie der Wissenschaften und der Literatur Mainz. Veröffentlichungen der Orientalischen Kommission Band 53). Wiesbaden: Harrassowitz. Müller, W.W. 2010b. Rezension von al-Manaser, A.Y.Kh., Ein Korpus neuer safaitischer Inschriften aus Jordanien. Orientalistische Literaturzeitung 105: 517–524. Müller, W.W. 2014. Südarabien im Altertum. Kommentierte Bibliographie der Jahre 1997 bis 2011. (Epigraphische Forschungen auf der Arabischen Halbinsel Band 6). Tübingen: Ernst Wasmuth. nbl. 1991–2001. Neues Bibel–Lexikon, hrsg. von M. Görg und B. Lang, vol. 1–3. Zürich: Benzinger. Negev, A. 1991. Personal names in the Nabatean Realm. (Qedem. Monographs of the Institute of Archaeology 32). Jerusalem: The Hebrew University. Nöldeke, Th. 1904. „Einige Gruppen semitischer Personennamen.“ Beiträge zur semitischen Sprachwissenschaft, 73–106. Strassburg: Trübner.

40

müller

Noth, M. 1928. Die israelitischen Personennamen im Rahmen der gemeinsemitischen Namengebung. Stuttgart: Kohlhammer. Oxtoby, W.G. 1968. Some Inscriptions of the Safaitic Bedouin. (American Oriental Series Volume 50). New Haven, Connecticut: American Oriental Society. Rechenmacher, H. 2012. Althebräische Personennamen. (Lehrbücher orientalischer Sprachen ii/1). Münster: Ugarit-Verlag. Al-Said, S.F. 1995. Die Personennamen in den minäischen Inschriften. (Akademie der Wissenschaften und der Literatur Mainz. Veröffentlichung der Orientalischen Kommission Band 41). Wiesbaden: Harrassowitz. Aš-Šammarī, H. 1990. Ǧamharat asmāʾ an-nisāʾ wa-aʿlāmihinna. Ar-Riyāḍ: Dār Umayya. Shatnawi, M.A. 2002. „Die Personennamen in den ṯamudischen Inschriften. Eine lexikalisch-grammatische Analyse im Rahmen der gemeinsemitischen Namengebung.“ Ugarit-Forschungen. Internationales Jahrbuch für die Altertumskunde SyrienPalästinas 34: 619–784. Sholan, A. 1999. Frauennamen in den altsüdarabischen Inschriften. (Texte und Studien zur Orientalistik. Band 11). Hildesheim: Georg Olms. Sima, A. 1999. Die lihyanischen Inschriften von al-ʿUḏayb (Saudi-Arabien). (Epigraphische Forschungen auf der Arabischen Halbinsel Band 1). Rahden/Westf.: Leidorf. Sima, A. 2002. „Die hasaitischen Inschriften.“ Neue Beiträge zur Semitistik, hrsg. von N. Nebes, Jenaer Beiträge zum Vorderen Orient Band 5: 167–200. Socin, A. 1900–1901. Diwan aus Centralarabien, hrsg. von H. Stumme, vol. 1–3. Leipzig: G.B. Teubner. Stark, J.K. 1971. Personal Names in Palmyrene Inscriptions. Oxford: Clarenton Press. Tairan, S.A. 1992. Die Personennamen in den altsabäischen Inschriften. Ein Beitrag zur altsüdarabischen Namengebung. (Texte und Studien zur Orientalistik. Band 8). Hildesheim: Georg Olms. Wellhausen, J. 1897. Reste arabischen Heidentums. Berlin: Georg Reimer2. Winnett, F.V. and G.L. Harding. 1978. Inscriptions from Fifty Safaitic Cairns. (Near and Middle East Series 9). Toronto: University Press.

chapter 3

Safaitic Prayers, Curses, Grief and More from Wadi Salhub—North-Eastern Jordan Hani Hayajneh

As a part of the efforts of the Department of Epigraphy at Yarmouk University (Irbid, Jordan) for documentation of the epigraphic heritage of Jordan, an epigraphic survey in Wādī Qaṭṭāfī/Wādī Salḥūb in Northeast Jordan was carried out between 10 and 15 April, 20131 and continued in November 2015 under the direction of the present author. These surveys constitute the first phase of our work in the area, which will be conducted through long term projects, including parts of Wādī Qaṭṭāfī.2 Wādī Salḥūb itself comprises its core rea, as well as numerous branches in different directions (s. Map, Fig. 1). The expedition resulted in documenting and inventorying some hundreds of Ancient North Arabian3 inscriptions (mostly Safaitic4) from 8 sites. In addition, tens of Arabic inscriptions were discovered; most of these are dated to the mid-8th century a.d.5 The aim of the present paper is to present a philological treatment of some newly discovered Safaitic inscriptions from the mentioned survey. The Siglum “Salhub” will be used for this and other future publication of the epigraphical materials from the same area. Personal names will not be treated and common lexemes will not be dealt with unless they are new or rare in Ancient North Arabian.

1 The survey was conducted with participation of some m.a. students at the Department of Epigraphy: Ms. Laura Tashman, Mr. Abdallah Khasawneh (currently PhD candidate in Semitic studies at Philipps University of Marburg), Mr. Muhammad Al-Zuʿbi (Yarmouk University) and in cooperation of Mr. Khaled al-Janaydeh (daj), Mr. Muwaffaq Bataineh (surveyor) and Mr. Hussein Debajah (photographer both at Yarmouk University). 2 Thanks to the Department of Antiquities of Jordan which gave us the permission to conduct future archaeological and epigraphical surveys within ca. 10km2 of the in the region. This first phase was made possible through a financial support by Yarmouk University. 3 On this designation, see. Macdonald (2000, 2004) and Hayajneh (2011 and 2017). 4 See Al-Jallad (2015) for fresh treatment of the contents of the Safaitic inscriptions. 5 See Hayajneh and Weninger (2015).

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

figure 1

Map of Wādī Salḥūb drawn by muwaffa batayneh

42 hayajneh

safaitic prayers, curses, grief and more from wadi salhub

figure 2

43

Salhub 1

l lb bn ʿqrbn bn ḫzr w-rʿy h-nḫl bql ṯlṯ By Lb son of ʿqrbn son of Ḫzr and he pastured (the sheep) in the valley on (fresh) herbage three (times/days?) The reference for the number ṯlṯ “three” is not mentioned in the text.

44

hayajneh

figure 3

a)

Salhub 2

l ys2kr bn ms1k bn mṯʿ bn khl By Ys2kr son of Ms1k son of Mṯʿ son of Khl

safaitic prayers, curses, grief and more from wadi salhub

b)

45

l s2b bn mṯʿ bn khl w-l-h h-ʾs1yt ṣyd ts1 By S2b son of Mṯʿ son of Khl and the/these construction(s) is / are his, (and) he hunted wild-goat(s).

The preposition l in w-l-h is attached to the pronoun of the 3rd person singular which refers to S2b. The reading of the glyph b is certain. The second glyph in the following word is s1, not, b, hence the reading is s1yt. For interpretation, one encounters in Arabic al-āsiyat (pl. ʾawāsī) “the tight construction (building etc.); (ceiling) support, column, pillar; they were named as such because they support the ceiling; buildings and fortified supported by pillars / columns or anything else (of this kind); the rest of the house / dwelling” (Lisan, p. 83). The Arabic form is considered as a loanword from Aramaic (Fränkel 1866: 11) which borrowed it from Akkadian asītu “Turm” (AHv, p. 74). The Hebrew word ‫( ַאְשׁ ִויֶתיָה‬ʾāšəyōwṯe-hā) in Jr 50,15, on the fall of Babylon, is a loanword from Akkadian (halat p. 91). The Hebrew form is encountered, referring to a kind of protective structure or fortification with several translations “foundations; pillars; towers; bulwarks; bastions” (see Utley 2012).6 It is attested in the Hatra inscription, ʾšyt “wall” (dnwsi, p. 123), Syriac ešat “foundation, bottom” (Payne Smith 1903: 31) and Mandaic ašita, ašata, ašaita “wall” (Drower and Maccuch 1963: 40).7 From the presence of this word in Safaitic, we may conclude that it penetrated into the Arabian linguistic realm in a very early stage of the history of the Arabian languages. In the present text, it is impossible to infer whether it is a singular or plural form. The derivatives of the root ṣyd in ana are many, e.g. ṣd “game”, ṣyd “snares; to hunt” (Al-Jallad 2015: 348).8 The following is a collective noun ts1 “buck; he-goat”. The word tays/š belongs to the common Semitic stock, Hebrew: tayiš, Arabic tays “a he-goat; male of the mountain-goat and of the gazelle or that of a complete year, or a yearling hegoat” (Lane, p. 324), Syiac tayšā “he-goat” etc. and in Akkadian daššu (taššu) “buck (said of gazelles and goats)”.9 The word ṣyd can be taken as a form of suffix conjugation, “he hunted (a) wild-goat(s)”

6 In: https://bible.org/assets/pdf/Utley_13AJeremiah.pdf. 7 See drs, p. 35 for more discussion. 8 On the subject of hunting in the Safaitic inscriptions and related matters, e.g. desert kites, etc, see Macdonald (2005) and Maraqten (2015) for a general introduction on hunting traditions in Pre-Islamic Arabia. 9 See Müller (1962: 34) and Sima (2000: 143 f.) for Semitic parallels.

46 c)

hayajneh

l ʿnq bn ws1mt By ʿnq son of Ws1mt

figure 4

Salhub 3

safaitic prayers, curses, grief and more from wadi salhub

a)

47

l s2ʿṯm bn s1r bn ʿwḏ w-qyẓ ʿl mʿzyn w-ndm ʿl ʿhd w-ʿl ḫlh w-ʿl ṭmṯn w-ʿl bʾlt w-ʿl hnʾ w-ʿl ḥnnt w-ʿl hgr w-ʿl bʿtm (?) w-ʿl mḥrb trḥn f-h-lt ʿmy By s2ʿṯm son of S1r bn ʿwḏ and he spent the dry season at Mʿzyn, and he was devastated by grief on account of ʿhd and his uncle and Ṭmṯn and Bʾlt and Hnʾ and Ḥnnt and Hgr and Bʿtm and mḥrb, who had perished, so, O Lt, blind

The verbal qyẓ “spend the dry season” is discussed in different occasions (e.g. Macdonald 1992). The word mʿzyn, which is preceded by the preposition ʿl; could be considered as a locative, cf. mkjs 1: s²ty ʿl ⟨⟨⟩⟩ h- nmrt ‘he spent the winter at the edge of Namārah’. The reading of the name bʿtm is certain; one may notice that the glyph for ʿayn it attached to a short stroke, which is not part of the letter. The verb ndm is translated as “to be devastated by grief”. The lexeme trḥn is an adjective in a form of passive participle in the meaning “perished”.10 The curse by the Goddess Lt is defined by the word ʿmy “blindness”, which is known in Arabic as “to be blind in both eyes” (Lane, 2160). This would recall the meaning of the derivatives of the root ʿwr in Safaitic which has been translated as “blind”. One could argue here that they authors differentiated between ʿwr and ʿmy, whereas the first conveys the meaning “to become one-eyed, damaged” (cf. Arabic ʿawira).11

10 11

For participles in Safaitic, see Al-Jallad (2015: 114 ff.), and (al-) Manaser and Abbadi (2016) for the semantic domain of the word. See also Ababneh (2014).

48

hayajneh

figure 5

12

Salhub 412

This inscription was published by Ḥarāḥišah (2007).

safaitic prayers, curses, grief and more from wadi salhub

a)

49

l ʾdʿgt bn s1wd bn ẓlm w wlh ʿl mkyn ʾs2ʿ-h f-h-rḍy ʿyr By ʾdʿgt son of S1wd bn Ẓlm and he was distraught with grief for the places of his companions, so O Rḍy, let be disgrace / blindness (for whosoever would efface).

The grief word wlh “to be distraught”, which occurs often in the Safaitic inscriptions, was treated in different occasion.13 The word mkyn is a broken plural form of mkn, cf. Arabic makān “place, spot, site, location” (see wkas, 468 ff.). It builds with the following word ʾs2ʿh14 a genitive construction “the places of his companions”. Cf. ʾs2yʿ-h “his companions” (e.g. krs 1342 and r 701, nr. 5) which is built as a broken plural form (= ʾafʿāl) from the root s2yʿ, cf. Arabic šayyaʿa-hu “he sent, or sent on, him; he made him, or it, to follow” (Lane, p. 1632), šayʿ (sing.), ʾašyāʿ (pl.) “follower and a friend, or a comrade, or an assistant” (Lane, p. 1632).15 The curse formula ends with the verb ʿyr, which can be interpreted in two ways; either as an imperative form of ʿwr, ʿawwara “to blind”, which is in this case derived from the ii-w root and represents a confusion with iiy root, as attested in ʿyr “blind” (krs 1695) and ryḥ (krs 78),16 or it could be taken as an imperative form of ii-y of the 2nd *ʿayyir “abuse, condemn”, cf. Arabic ʿayyara, ʿāra “blame, rebuke, condemn, insult … etc”, ʿār “shame, disgrace”, taʿāyur “abuse, vituperation” (Lisan, 3188). b)

l gff (?) ….. ʾṯr (?) l ʾ/ṣ mgdy By Gff (?) …..ʾ/ṣ mgdy

The reading of the text is not certain.

13 14 15 16

See Eksell (2005: 167 f.) for this group of words bearing this meaning. For comments on the aberrant spelling, see Al-Jallad (2015: 346). See Hayajneh (2016: 509) for discussion. Al-Jallad (2015: 125).

50

hayajneh

figure 6

Salhub 5

l ṣbḥ bn nql h-ʿrn By Ṣbḥ son of Nql these two asses The rock drawing of two asses makes no doubt that we are dealing with a dual form suffixed by the dual suffix -n.17 The word ʿr “ass” is widely attested in the Safaitic inscriptions.

17

For dual forms in ana, especially in Safaitic, see (Macdoald 2004: 503 and Al-Jallad 2015:61).

safaitic prayers, curses, grief and more from wadi salhub

figure 7

51

Salhub 6

l gl/rdn bn yṭʿn h-nfs1t w-ʿṣr ʿl-h (?) This funerary monument is for Gl/rdn son of Yṭʿn and he (the author of the text) held / concealed (his grief/sadness) in account of him The reading of the last word, especially, the l is certain. Reading it as a z is not probable, as the horizontal stroke which appears on the top of the shaft can create a glyph for l. Al-Jallad (2015: 82) proposes that nfs1t as a combination of nfs1 and a demonstrative -t, (nfs1 +-t), “this funerary monument”.18 He considers it as a fixed phrase used by authors. The word nfs1 has spread from the nws linguistic area to different parts of Arabia. A large number of nfs1-funerary monuments found in pre-Islamic tombs and temples sometimes have short inscriptions including names of the dedicants or deceased. It seems that this word underwent a semantic development from the meaning “soul” and then to “funeral monument”.19

18 19

See Hayajneh (2002 and 2009). See Macdonald (2006: 289) and Hayajneh ( forthcoming).

52

hayajneh

The root ʿ-ṣ-r yields in Arabic ʿaṣara “squeeze, press, etc.” (Lane, 2061). However, derivatives from Biblical Hebrew bear an appropriate meanings that could fit; cf. e.g. ʿāṣar “hold back, detain, restrain strength, …”, and in the from the nif. form, “to be restrained, hindered; to be gathered together” (halat, 823ff.). The semantics of the derivatives denotes the authors’ deep grief on the dead person. The lexeme here is semantically equal to Arabic kaẓama “restrain”, Cf. Quran (3.134) … wa-l-kāẓimīna l-ġayḍaah … “and those who suppress/restrain/ (their) anger/wrath …”

figure 8

Salhub 7

safaitic prayers, curses, grief and more from wadi salhub

a)

53

l nfr bn ws1m h-dr w ṭbb b-ḥrn ḥwbt f- yʾs1 ʾns1

The formulaic structure and vocabulary of this inscription differ greatly from other Safaitic inscriptions. Therefore, the results of the following presentation of the structure, morphology as well as the semantics remains speculative: Dr: it was translated as “place”. It is known with verbs related to camping and spending time.20 Ṭbb: the verb ṭbb refers to the subject Nfr, who could be the author of the inscription. It is attested in a different syntactical position in another Safaitic inscription: ceds 176 l ṣrmt bn qdm w ḥll h- ẓlt w tẓr ṭ{b}b m- ḥrn,21 which will be treated below. The meanings of the verbal and nominal derivatives of the root ṭbb in the Semitic lexica are can be summarized as follow: 1)

2)

20 21 22 23

Syriac ṭab “find out, inquire about”, ṭebbā “news, rumour”, Arabic ṭabba “treat with kindness, be clever, scientist; practice medicine”, ṭabb “clever”; Ancient South Arabian ṭbb “teach, proclaim/judge”,22 Soqotri ṭeb “believe, know”, Geez “to be wise, prudent” (drs, 1048f.). From this semantic domain, we might be able to interpret the verb ṭbb in our inscription as “to know, find out, learn, proclaim” Another semantic field is represented by Syriac ṭabb “fall on”, Egyptian Arabic ṭabb “to reach unexpectedly, to jump on” (drs, 1048f.). Landberg (192–142: 2184ff.) presented the derivatives of this root in the Arabic dialects. He quotes several passages related to the sense of “arriving, reaching a place or situation unexpectedly”, e.g. ṭabbēt li-l-ʿraka “je me suis lance dans le combat”, u-min ṭabbētu ʿala el-faras “dès qu’ il eut sauté en selle sur la jument”, ṭabbēt fi l-bilad … “I arrived / entered in the country …”,23 sārū gebalhom elyāma ṭabbu gurb el-ʿarab “ils marchèrent en avant jusqu’à ce qu’ils fussent arrivés près de leurs contribules”. He quotes also the usage of ṭabb in the Iraqi dialect in the meaning “enter” as well as ṭabban in the meaning of ṭāḥ (= mašēn) “to go, march” (192–142: 2184ff.). Related derivatives in modern Arabic dialects of the Levant, especially For more discussion, see Al-Jallad (2015: 83f.) who relies on the Greek writing version of the word as δαυρ /dawr/ to vocalize the Safaitic version, i.e. not as /dār/. I am indebted to Ms. Chiara Della Puppa (a PhD candidate at Leiden University) for drawing my attention to this inscription. See sd, p. 152. Translation not provided by Landberg.

54

hayajneh

Jordan and Palestine, are attested, e.g. ṭabb / yṭubb ʿalēhum “arriving unexpectedly”. Seeger et. al. (2014: 223) documented some derivatives of this root in the Palestinian dialects, e.g. ṭabb “schlagen, stoßen (jn, etw); dengeln (Sichel, Sense); anfassen, berühren (jemanden, etw fi); umkippen, umdrehen, umwenden, umwerfen (jn, etw); umfallen; da/plötzlich etw tun; (plötzlich) anfangen/beginnen etw zu tun (Subjunktiv); sich (plötzlich) stürzen (auf jn ʿala)” (Seeger et. al. 2014: 223 and see also Militarev 2010:56). Given this semantic range, one could perhaps propose the sense “come upon, descend upon, or pounce upon” for the present verb ṭbb. Ḥrn: According to Macdonald (2009: 7) there are now 23 references to the Ḥawrān (22 to ḥrn and 1 to ḥwrn) in approximately 28,000 Safaitic inscriptions. Ḥwbt: This is a nominal form. The root ḥwb in Semitic languages produced the following derivatives: 1)

2)

3)

24 25

Derivatives of this root are attested in the Safaitic inscription of grieving context and give the meaning of “lamentation” or “supplication” followed by the preposition ʾl +dn or ʿl + a kinship term, beloved, or pn. Ḥwb is attested as a parallel to wgm in two inscriptions incised on the same rock by the same person (krs 1251 and krs 1250).24 Hebrew ḥōb and Aramaic ḥwb “be guilty, indebted”, Syriac ḥawbā, ḥūbā and Arabic ḥawb, ḥūb “debt, sin, crime, offense”, Sabaic ḥwb25 “sin”. This meaning passed to Arabic and Hebrew from Aramaic. The semantic field of the root in Arabic goes into two directions: a) ḥawb “grief, sorrow, trouble, loneliness, solitariness; difficulty, distress” ḥūb “death, destruction, disease, trouble”. ḥawba “anxiety; lack, poverty; want, poverty; state, condition; tenderness of the maternal heart; uterine kinship” (Lane, p. 662 ff. and drs, p. 841 f.). b) ḥawba means also “a person whom one is under an obligation to respect, or honour, or defend, and who may be subjected to loss, or ruin, [if abandoned,] such as a mother, or sister, or daughter, or any other female relation within the prohibited degree of marriage”; alḥaqa Allāhu bi-hi l-ḥawbata “May God bring upon him want, or poverty, or indigence”, ʾilayka ʾarfaʿu ḥawbatī “to thee I make known

See Della Puppa (this volume) for more discussion of the derivatives of the root ḥwb in Safaitic inscriptions. See Beeston (1994: 42).

safaitic prayers, curses, grief and more from wadi salhub

55

my want”, (Lane, 662ff. and drs, p. 841f.). Thus the word designates beings that need to be protected, supported: a wife, a concubine, and even the father and mother (drs, 841 f.). Yʾs1 Derivatives of the root yʾs1 are attested in Safaitic, e.g. yʾs¹ (csns 957; krs 169), as a form of a suffix conjugation, “to despair”; as a 2nd stem yʾs¹ (lp 718) “to upset; to make; despair” etc. (see Al-Jallad 2015: 353). Cf. also Hebrew yāʾēš “abandoner”, nôʾaš “désespérer de” Aramaic yaʾeš “faire désespérer”, Arabic yaʾisa, Ancient South Arabian yʾs1 “désespérer de” (drs, 1141). ʾns1: It could function as a subject of the latter verb, yʾs1, cf. the Arabic collective noun ʾuns,ʾanas “a numerous company of men; a tribe, staying, residing, dwelling, or abiding: the people of a place alighting or abode”; ʾins “the inhabitants of a house” (Lane, p. 114). The unconventional structure of this inscription and given the semantic varieties produced by the root might lead us to the following proposed translations of the individual words used in the text: ṭbb “arrive, come upon / befall; to know, find out, learn; proclaim” ḥwbt “misfortune; lamentation, supplication” yʾs1 “to despair; renounce” ʾns1 “people, inhabitants” Hence, the text can be translated in multiple ways, however, none of which can be preferred of the other: 1) 2)

“By Nfr son of Ws1m at this place, and misfortune came upon / befell Ḥawrān, so the people despaired” “By Nfr son of Ws1m at this place, and he knew / found out / witnessed misfortune in Ḥawrān so the people despaired”

Given the mentioned translation variants, inscription ceds 176 l ṣrmt bn qdm w ḥll h- ẓlt w tẓr ṭ{b}b m- ḥrn, can be translated as follows: By Ṣrmt son of Qdm and he camped in this shelter and awaited a comer / arriver (or: news) from Ḥawrān b)

l ṯlm bn bnḥmʾl By Ṯlm son of Bnḥmʾl

56

hayajneh

figure 9

Salhub 8

l wḥfy bn bzz h-zrb By Wḥfy son of Bzz the lurking-place / enclosure The word zrb occurs in similar formula in Ababneh (2005: 282, Nr. 651). Cf. Zarb and zarība that signify the lurking-place of a hunter, or sportsman, or of an archer, or a shooter; the also signify a well [or pit] which the hunter, or sportsman, digs for himself that he may lie in wait therein for the game or Zarība signifies “enclosure, cattle pen” (Lane, 1224).26

26

For other Safaitic words denoting “enclosures”, see Maraqten (2015: 230).

safaitic prayers, curses, grief and more from wadi salhub

57

figure 10 Salhub 9

a)

l hms1k bn s1krn w-rʿy wgd h-ʾṯbr By Hms1k son of S1krn, and he pastured (and) he found the crops/fruits

58

hayajneh

The last word of the inscription can be read as ʾṯbr. There are traces of an incised concavity inside the rhombus, but it is difficult to consider these traces as part of the glyph. Therefore, we proposed two readings. As for the root ṯbr, we encounter in Safaitic the word ṯbr (c 1758, 4443) “Sagittarius”, ṯbr (lp 679) “destruction”, and ṯbrn “warriors” (nst 3) (Al-Jallad 2015: 349). The word occurs in Dadanitic in the meaning “céréale” (Farès-Drappeau 2005: 206). The proposed meaning here is based on Hebrew šeʿer “grain, corn”27 (e.g. Gen 421 f.-19–26 (see halat, p. 1305)), which fits the context of pasture. b)

l ʿgl bn ʾlht By ʿgl son of ʾlht

c)

ṣnl Ṣnl

27

Including its function as an item of trade.

safaitic prayers, curses, grief and more from wadi salhub

figure 11

Salhub 10

59

60 a)

hayajneh

l s2k bn zkr bn ʿmr w-kmd ḥwlt

While the l is carved as a bent line, its reading as an l is certain is certain, and cannot be confused with b. Derivatives of the root kmd are attested in Arabic, cf. kamida “to be, or become, mournful, melancholy, depressed, downcast; to give oneself to grief, to pine”, kammada “to apply a warm pack round a part of the body; to make warm poultices; compresses”, and ʾakmada “to make mournful, to sadden, to grieve” (wkas, p. 351ff.). The object of the verb kmd is the following word ḥwlt can be interpreted as follows: In the same writing one may suggest the enemy social group ḥwlt, which is known from a number of Safaitic inscriptions that reflect the conflicts between the inhabitants of the Ḥarra and the Ḥwlt tribe. It is associated with the Avalitae, who, according to Pliny, can be identified with the North Arabian oases of Dūma and Ḥegrā;28 the epigraphical evidence proves that the authors of the Safaitic inscriptions were supporters of the Nabataeans against Ḥwlt and no evidence that they support the Ḥwlt (see Abbadi 2015: 72). Given the lexical domains as attested from the Arabic lexicon, one would propose the translation “and he was depressed/ grieved in account of Ḥwlt”, i.e. without the preposition ʿl. This inscription, if the interpretation is correct, forms an opposite argument regarding enmity against the Ḥwlt. Considering the semantic domain under the 2nd stem of the root kmd which indicates the sense of compressing or squeezing (of poultice on a part of the body), one could suggest the meaning “push, pressured” for the verb kmd. Hence, the translation proposed is “… he pushed / pressured Ḥwlt”29 By S2k son of Zkr son of ʿmr and he grieved / lamented (in account of) Ḥwlt By S2k son of Zkr son of ʿmr and he grieved a (misfortunate) change/ state/condition

28 29

Villeneuve (appud Macdonald 1993: 308, note 36). Less probable is the interpretation of ḥwlt as a lexeme, cf. Arabic ḥiwla, ḥūla “a mood, or manner of changing from a state to another; shifting from one thing to another” (Lane, p. 676). Taking the above discussion into account, one may give the meaning “and he grieved a misfortunate (change)” as one would grieve or feel the sorrow for misfortunate conditions, calamity or disasters. The interpretation of kmd leaves the chance to suggest that ḥwlt here can be interpreted as a personal name, however the grief words attested in the Safaitic inscriptions are mostly followed by the preposition ʿl.

safaitic prayers, curses, grief and more from wadi salhub

b)

61

l yʿll b⟨n⟩ bhl By Yʿll son of Bhl

figure 12 Salhub 11

a)

l grm bn rbbʾl bn ʿtrt bn grr (or: ʿrr) bn ġn{m} bn brq (?) […] bn s1r bn bnyt bn dll ḏ-ʾl zhr w-wgm ʿl ʾsmnt ḫl-h w ʿl […]h q ʿl trḥ rġm mny f-h-lt ʿwr w grb wgʿ (or: w gʿ) l-ḏ yʿwr h-s1fr

62

hayajneh

By Grm son of Rbbʾl son of ʿtrt son of Grr (or: ʿrr) son of Ġnm son of Brq […] son of S1rʿ son of Bnyt son of Dll of the lineage Zhr and he grieved for ʾs1mnt, his uncle and for […] who had perished struck down by fate, so, O Lt, let there be blindness and painful scab (or: and starvation) for who would efface the inscription. The grief verb wgm was a subject of several discussions (e.g. Eksell 2002 & 2005). Trḥn: an adjective in a form of passive participle “perished”. The precise meaning of this word is difficult to identify, and has been discussed by many scholars (e.g. Al-Jallad 2015: 114, 348).30 Grb “scab” designates a kind of diseases which is encountered several times in the Safaitic inscriptions and widely known in other Semitic languages, e.g. Arabic ğarab, Akkadian garabu(m). This acquired, infectious and also epidemic spreading disease is a serious fatal disease for both humans and animals (Ababneh 2015: 15). The following cluster of characters, w, g, and ʿ, can be explained in two ways: 1)

2)

b)

either as wgʿ; (1) as a verbal form, which is attested in Safaitic lp 314 w wgʿ ʿl mġny … and he grieved for (see Ababneh 2015: 19 and Fn. 21),31 as a nominal form. Cf. Arabic wağaʿ “pain”. Taking this into account, wgʿ in the present inscription could be interpreted as an adjective in a form of a participle, i.e. *wağīʿ/wāğiʿ “painful,” or divided as w “and” + gʿ (cf. Arabic ğūʿ “starvation”). The latter is attested as an active participle in Safaitic, gyʿ “starving” (c 3811).32 l ʾys1 bn gzm h (?) ṯl (?)m (?) By ʾys1 son of Gzm h-ṯ

The reading of the last three or four signs is not certain. 30 31

32

See also al-Manaser and Abbadi (2016). Al-Jallad (2015: 221) considers the verb ngʿ, which is used frequently in grieving formulae, as a reflection of the N-stem of the root √wgʿ ‘to be in pain’, because the verb wgʿ itself is used in the same grieving context as in wh 946 w wgʿ ʿl Ġyrʾl with wh 960 w ngʿ ʿl Ġyrʾl, both ‘and he grieved in pain for Ġyrʾl’. See Al-Jallad (2015: 121) for discussion on participles.

safaitic prayers, curses, grief and more from wadi salhub

c)

l gls1 bn gr bn ʾlʿz By Gls1 son of Gr son of ʾlʿz

d)

figure 13 Salhub 12

63

64

hayajneh

l zmhr bn yṯʿ bn zmhr bn ʿḏ bn rbʾl bn ʿdnn bn ln ḏ-ʾl gr w-ḥll f-h-lt s1lm s1bʿ ṣlt mʿwr s1fr “By Zmhr son of Yṯʿ son of Zmhr son of ʿḏ son of Rrbʾl son of ʿdnn son of Ln of the lineage Gr, and he encamped, so, O Lt, may he be secure and curse the prayer of whosoever would efface the inscription” After the prayer, one would expect the verb ʿwr followed by mʿwr s1fr. However, instead of that, we encounter a strange expression, which occurs to the best of my knowledge for the first time in the Safaitic inscriptions; it … s1lm s1bʿ ṣlt. The reading is certain; the shape of the glyph for l is bent on the base line, which is similar to the shape of the lam auctoris and in the name Ln in same inscription. Two letters, lt, are visible under the word ṣlt. Their function here is difficult to explain. We believe that the word s1bʿ is part of the invocation, which can be compared with sabaʿa “abuse, revile, vilify; defame; discredit” (Lisān, p. 1927). The verb is in the qal in Biblical Hebrew “swear (make a statement or a promise with an oath), invoking God, pledging something valuable” (halat, p. 1299ff.). The root in Arabic is usually associated with the numerical seven, although it could have been derived from a different root. Kottsieper (2004: 311ff.) discussed this root in the context of Biblical Hebrew evidence. He believes that the Arabic sabaʿa “curse, revile” can be etymologically related to Geʾez sbʿ “bewitch” and cannot be associated with numerical root seven. The nif. forms in Hebrew have negative connotations, as in Arabic, but it develops into a verb of aggression meaning “curse, revile” whence the meaning “swear” developed from the reflexive nature of the nif. Forms of the root šbʿ, while the hif. forms in the meaning “adjure, implore” show that the root didn’t lose this magical-religious nature. The following word ṣlt stands as an object and could be interpreted as “prayer”. To the best of my knowledge, this is the first occurrence of the word ṣlt in any Ancient North Arabian corpus. For the spread of the term ṣlt in Arabia, see Beeston (1994: 42) and Robin (2014: 53) who noticed that many loanwords from Judaeo-Aramaic appear in the Ancient South Arabian monotheistic inscriptions, while there are none in the polytheistic inscriptions, i.e. ‘prayer’ (ṣlt) and ‘favour, (divine) grace’ (zkt) which appeared again 200 years later with the meaning of ‘prayer’ and ‘legal alms’ in Islam.33 There is a scholarly consensus, even among Medieval Arab scholars, that ṣalāt in the Qurʾan is a loanword from Aramaic (Jeffery 1938: 198)

33

For details on the usage of the word ṣlt in the Ancient South Arabian monotheistic period, see Gajda (2009).

safaitic prayers, curses, grief and more from wadi salhub

65

Abbreviations AHw c ceds csns dnwsi drs halat krs Lane

Lisān

lp

sd wh wkas

W. von Soden, Akkadisches Handwörterbuch (Wiesbaden: Harrassowitz, 1965–1981) Safaitic inscriptions in Corpus Inscriptionum Semiticarum, Pars v. Paris. Unpublished Safaitic inscriptions recorded by V.A. Clark in the Eastern Desert Iscriptinos Survey. Clark, V.A. A Study of New Safaitic Inscriptions from Jordan. Unpublished Ph.D. (Melbourne University. Mountains, 1980) J. Hoftijzer and K. Jongeling, Dictionary of the North-West Semitic Inscriptions (Leiden: Brill, 1995) D. Cohen, Dictionnaire de racines sémitiques ou attestées dans les langues sémitiques (Paris: Mouton; Louvain: Peeters, 1970–2012) L. Koehler und W. Baumgartner, Hebräische und aramäische Lexikon zum Alten Testament (Brill: Leiden, 1967–1990) Safaitic inscriptions recorded by the Basalt Desert Rescue Survey in northeastern Jordan in 1990. E.W. Lane, An Arabic-English Lexicon: derived from the best and the most copious eastern sources. (London and Edinburgh Williams and Norgate, 1863–1893). Ibn Manẓūr, Muhammad ibn Mukarram ibn ʿAlī ibn Ahmad ibn Manzūr alAnsārī al-Ifrīqī al-Misrī al-Khazrajī Jamāl al-Dīn Abū al-Fadl. Lisān Al-ʿArab. (Al-Qāhira: Dār Al-Maʿārif, n.d.). E. Littmann, Safaitic Inscriptions. Syria. Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909. Division iv, Section c. (Leiden: Brill, 1943). A.F.L. Beeston; M.A. Ghul; W.W. Müller and J. Ryckmans, Sabaic Dictionary (English-French-Arabic). Louvain-la-Neuve: Peeters (1982). F.V. Winnett—G.L. Harding, Inscriptions from Fifty Safaitic Cairns. Near and Middle East Series 9. (Toronto: University of Toronto Press, 1978). M. Ullmann, Wörterbuch der klassischen arabischen Sprache. (Wiesbaden: Harrassowitz, 1970–1991).

Bibliography Ababneh, M.I. 2005. Neue safaitische Inschriften und deren bildliche Darstellungen. (Semitica et Semitohamitica Berolinensia 6.) Aachen: Shaker. Ababneh, M.I. 2014. “The term ʿwr in the Safaitic inscriptions” [in Arabic]. In A Pioneer of Arabia. Studies in the Archaeology and Epigraphy of the Levant and the Arabian

66

hayajneh

Peninsula in Honor of Moawiah Ibrahim, edited by M. Maraqten and Z. Kafafi, 39–43. (Studies on the Archaeology of Palestine & Transjordan, 10). Rome «La Sapienza». Rome: P. le A. Moro. Ababneh, M.I. 2015. “Krankheit und Heilung in safaïtischen Inschriften.” In Neue Beiträge zur Semitistik, edited by V. Golinets, H. Jenni and H.-P. Mathys, 9–26. Münster: Ugarit-Verlag. Abbadi, S. 2015. “New evidence of a conflict between the Nabataeans and the Ḥwlt in a Safaitic inscription from Wadi Ram.” Arabian Epigraphic Notes 1: 71–76. Al-Jallad, A. 2015. Outline of the Grammar of the Safaitic Inscriptions. (Studies in Semitic Languages and Linguistics). Leiden: Brill. Beeston, A.F.L. 1994. “Foreign Loanwords in Sabaic.” In Arabia Felix. Beiträge zur Sprache und Kultur des vorislamischen Arabien: Festschrift Walter W. Müller zum 60. Geburtstag, edited by N. Nebes, 39–45. Wiesbaden: Harrassowitz. Della Puppa, Chiara (this volume). Drower, E.S. and R. Macuch. 1963. A Mandaic Dictionary. Oxford: Clarendon. Eksell, K. 2002. Meaning in Ancient North Arabian Carvings. (Acta Universitatis Stockholm. Stockholm Oriental Studies 17). Stockholm: Almquiest & Wiksell. Eksell, K. 2005. “The verb wğm in Safaitic inscriptions.” In Alltagsleben und materielle Kultur in der arabischen Sprache und Literatur: Festschrift für Heinz Grotzfeld zum 70. Geburtstag, edited by Th. Bauer and U. Stehli, 163–172. (Social Science 55, 1). Wiesbaden: Harrassowitz. Farès-Drappeau, S. 2005. Dédān et Liḥyān: Histoire des Arabes aus confins des pouvoirs perse et hellénistique (ive–iie avant l’ère chrétienne). (Travaux de la maison de l’orient et de la Méditerranée 42). Lyon: Maison de l’Orient et de la Méditerranée. Fraenkel, S. 1886. Die aramäischen Fremdwörter im Arabischen. Leiden: E.J. Brill. Gajda, I. 2009. Le royaume de Ḥimyar à l’époque monothéiste: l’histoire de l’Arabie du sud ancienne de la fin du ive siècle de l’ère chrétienne jusqu’à l’avènement de l’Islam. (Mémoires de l’Académie des inscriptions et belles-lettres 40). Paris: Académie des inscriptions et belles-lettres. Ḥarāḥišah, R. 2007. “Nuqūš Ṣafawīya Muḫtāra min al-Bādīya al-Urdunīya.” Mağallat alNuqūš wa-r-Rusūm al-Ṣaḫrīya. 1: 29–51. Hayajneh, H. 2002. “Zwei beschriftete Stelen aus dem Museum der Yarmouk-Universität.” Die Welt des Orients 32: 130–137. Hayajneh, H. 2009. “A Fragmentary Ancient North Arabian Inscription from al-Ḥuṣun—Kitim Area near Irbid—Northern Jordan.”Zeitschrift des Deutschen PalästinaVereins 125: 176–178. Hayajneh, H. 2011. “Ancient North Arabian.” In The Semitic Languages: An International Handbook. Edited by S. Weninger, 257–782. (Handbücher zur Sprach- und Kommunikationswissenschaft). Berlin: De Gruyter.

safaitic prayers, curses, grief and more from wadi salhub

67

Hayajneh, H. 2016. “Ancient North Arabian inscriptions, Rock Drawings, and Tribal Brands (Wasms) from Šammāẖ / ʾAyl (ʾĒl) region—Southern Jordan.” In The Shammakh to Ayl Archaeological Survey, Southern Jordan (2010–2012), edited by B. MacDonald, G.A. Clark, L.G. Herr, D.S. Quaintance, H. Hayajneh and J. Eggler, 505–541. (Archaeological Reports Book 24). American Schools of Oriental Research. Hayajneh, H. 2017. “Die vorislamischen frühnordarabischen Sprachformen.” In Nordarabien. Kolloquium 2012 des Deutschen Vereins zur Erforschung Palästinas e.V. Sprachen in Palästina im 2. und 1. Jahrtausend v.Chr. dpv-Tagung “Sprachen und Schriften in Palästina im 2. und 1. Jahrtausend v. Chr.”, edited by Herbert Niehr and Hermann Michael Nieman, 211–230. (Abhandlungen des Deutschen Palästinavereins 43). Harrassowitz: Wiesbaden. Hayajneh, H. Forthcoming. “An Ancient North Arabian inscribed Nefesh Stele from the vicinity of Jerash—North-western Jordan.” Syria. Hayajneh, H. and S. Weninger. 2015. “Ein kleines Gedicht aus der nordjordanischen Bādiya zur Zeit der Mamluken.” Zeitschrift des Deutschen Palästina-Vereins 131,1: 59– 68. Jeffery, A. 1938. Foreign Vocabulary of the Quran. Baroda: Oriental Institute. Kottsieper, I. 2004. “šāḇaʿ.” Theological Dictionary of the Old Testament 14: 311–336. de Landberg, C. 1920–1942. Glossaire Daṯînois. Leiden: Brill Macdonald, M.C.A. 1992a. “The Seasons and Transhumance in the Safaitic Inscriptions.” Journal of the Royal Asiatic Society 2,1: 1–11. Macdonald, M.C.A. 1992b. “Safaitic Inscriptions.” The Anchor Bible Dictionary 3: 418– 423. Macdonald, M.C.A. 2000. “Reflections on the linguistic map of pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 28–79. Macdonald, M.C.A. 2004. “Ancient North Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R.D. Woodard, 488–533. Cambridge: Cambridge University Press. Macdonald, M.C.A. 2005. “Of rock-art, “desert kites” and meṣāyid.” In Arabia Vitalis: Arabskiĭ Vostok, islam, Drevniaia Araviia: sbornik stateĭ, posviashchennyĭ 60-letiiu V. V. Naumkina, edited by A.V. Sedov and I.M. Smilianskaia, 332–345. Moskau. Macdonald, M.C.A. 2006. “Burial between the desert and the sown. Cave-tombs and inscriptions near Dayr al-Kahf in Jordan.” Damaszener Mitteilungen 15, 273–301. Macdonald, M.C.A. 2009. Literacy and Identity in Pre-Islamic Arabia. (Variorum Collected Studies Series cs906). Ashgate Publishing, Farnham/Burlington. al-Manaser, A. and S. Abbadi. 2016. “Remarks on the etymon trḥ in the Safaitic inscriptions.” Arabian Epigraphic Notes 2: 45–54. Maraqten, M. 2015. “Hunting in pre-Islamic Arabia in light of the epigraphic evidence.” Arabian Archaeology and Epigraphy 26, 2: 208–234. Militarev, A. 2010. “A complete etymology-based hundred wordlist of Semitic updated: Items 1–34.” Journal of Language relations 3: 43–78.

68

hayajneh

Müller, W.W. 1962. Die Wurzeln mediae und tertiae y/w im Altsüdarabischen. Eine etymologische und lexikographische Studie. Tübingen: E. Huth. Ratcliffe, R.R. 1998. The “Broken” Plural Problem in Arabic and Comparative Semitic Allomorphy and analogy in non-concatenative morphology. (Current Issues in Linguistic Theory 168). Amsterdam & Philadelphia: John Benjamins. Rezetko, R. and I. Young. 2014. Historical linguistics and Biblical Hebrew: steps toward an integrated approach. (Society of Biblical Literature ancient near East monographs 9). Centro de Estudios de Historia del Antiguo Oriente (uca). Payne Smith, J. 1903. A Compendious Syriac Dictionary. Oxford: Clarendon. [Reprint 1985] Robin, Ch. 2014. “The peoples beyond the Arabian frontier in late antiquity: recent epigraphic discoveries and latest advances.” In Inside and out. Interactions between Rome and the peoples on Arabian and Egyptian frontiers in late antiquity, edited by J.H.F. Dijkstra and G. Fisher, 33–79. Leuven, Paris, Walpole ma: Peeters. Seeger, U. 2014. Wörterbuch Palästinensisch—Deutsch. In cooperation with Laṭīfe Abu l-ʿAsal, Taḥsīn ʿAlāwnih, Rāmi il-ʿArabi Work in Progress Stand: Juni 2014. Unkorrigiertes und unvollständiges Manuskript. Fritz-Thyssen-Stiftung. url: http://www .uni-heidelberg.de/fakultaeten/philosophie/ori/semitistik/seeger_woerterbuch .html (accessed in 15 July 2016). Sima, A. 2000. Tiere, Pflanzen, Steine und Metalle in den altsüdarabischen Inschriften. Eine lexikalische und realienkundliche Untersuchung. (Veröffentlichungen der Orientalischen Kommission der Akademie der Wissenschaften und der Literatur, Mainz, 46). Wiesbaden: Harrasswitz. Tokunaga, R. 1999. “Old South Arabic nfs and its Relation with Stelae.” Bulletin of the Society for Near Eastern Studies in Japan 42, 2: 1–21. Utley, B. 2012. You can understand the Bible: Jeremiah. (Study Guide Commentary Series Old Testament 13). Texas: Bible Lessons International Marshall. (https://bible.org/ assets/pdf/Utley_13AJeremiah.pdf).

chapter 4

Notes on ḥwb in Safaitic Chiara Della Puppa*

1

Introduction

In Safaitic, the root √ḥwb is mainly attested as a verb in the suffix conjugation in the form ḥwb, but there is also sparse evidence for the variant ḥyb1 and for nominal forms.2 It has been documented 90 times3 so far, all instances coming from sites in northeastern Jordan. This paper shall discuss the different contexts in which ḥwb occurs, review previous translations of this word, and propose a new interpretation of its meaning.

2

Previous Interpretations

2.1 The Expression ḥwb ʾl-rḍw In two inscriptions ḥwb is followed by the prepositional phrase ʾl-rḍw. One of the two (KnGQ 5) is from Jebel Qurma and has been published by Knauf (1991),

* I thank Dr. Ahmad Al-Jallad, Prof. Maarten Kossman, and Johan Lundberg for discussing this paper with me and for their valuable comments. All errors remain my own. 1 This form occurs in ceds 438, HaNSB 346, and wh 116, 2442, 3029, 3393, 3838. For a thorough discussion of the different ways in which one could interpret the multiplicity of forms attested for Safaitic medial weak verbs in the suffix conjugation, see Al-Jallad 2015: 119– 120, 125. Both ḥwb and ḥyb could reflect either the G- or the D-stem, and since ḥyb occurs in the same exact contexts as ḥwb, it is impossible to infer which one of the two would represent the G-stem. The two forms could also reflect dialectal variants of the same verbal stem. 2 These are ḥwb (csns 918; wh 300), ḥyb (krs 82), and ḥwbt (ceds 356). 3 This number consists of: 1) the 73 attestations published on the Online Corpus of the Inscriptions of Ancient North Arabia (ociana; http://krcfm.orient.ox.ac.uk/fmi/webd#ociana, accessed on 6th June 2017), 2) one inscription which I have reinterpreted (csns 95, see below), 3) 16 new inscriptions documented within the framework of the Jebel Qurma Archaeological Landscape Project, based at the Faculty of Archaeology of Leiden University and directed by Prof. Peter M.M.G. Akkermans. The inscriptions which have been collected by this project are being studied by the author within the framework of the Landscapes of Survival Project as part of her PhD dissertation.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

70

della puppa

whereas the other one (csns 918) is from Qāʿ al-Ladhāyim, edited by Clark (1979). The texts read as follows: KnGQ 5 = qur 2.482.1 l ngʾ bn qld w ḥwb ʾl-rḍw csns 918 l bʾs¹ bn ḥbb h-ḥ{w}b ʾl-rḍw Knauf interpreted ḥwb as a borrowing from Aramaic4 and translated the expression ḥwb ʾl-rḍw as ‘he sinned against Ruḍā’. He consequently proposed to interpret also csns 918’s phrase h-ḥ{w}b ʾl-rḍw as ‘the (habitual) sinner against Ruḍā’.5 The most obvious problem with this interpretation is that one would have expected the use of the preposition ʿl- instead of ʾl-. The latter never bears the meaning ‘against’ in Safaitic, whereas it appears to be used in petitions and oaths to deities.6 Moreover, although the Classical Arabic G-stem verb ḥāba means ‘to sin’,7 possibly an Aramaic import, there is no compelling reason to prefer this translation of the verb. I will show below (Section 3) that other solutions provided by the Arabic lexica would suit better than ‘to sin’ this and other contexts of ḥwb in Safaitic.8 Clark interpreted the expression as a curse against the deity, and translated csns 918 as: ‘By Bʾs¹ son of Ḥbb, this misfortune of the god Rḍw!’. However, having observed that ʾl as a generic term for ‘god’ is not well attested in Safaitic, he then suggested that the inscription could mean ‘O affliction to Rḍw!’, i.e. a malediction against the deity. But had this been the case, one would have again expected the preposition ʿl- rather than ʾl-. Furthermore, neither curses nor maledictions against deities have been attested in Safaitic inscriptions. Clark cites csns 95 as his only example, proposing the following reading and interpretation: 4 Knauf 1991: 97; cf., e.g., Jewish Palestinian Aramaic ḥwb ‘to be guilty, sin, owe’ (Sokoloff 1990: 189). 5 Knauf argued that the inscription should be interpreted either as a confession or a as boast, but both textual genres are completely unattested in Safaitic inscriptions. 6 See the following inscriptions: c 31 … w s²tky ʾl-lt f ḥnn w s¹lm m-s²nʾ ‘… and he petitioned Lt, so show compassion that he may be secure from enemies’; c 25 … f hy lt w h s²ms¹ ʾṭn ʾl-km ydh l-ṯʾr m-ḏ ʾs¹lf ‘… so, O Lt and S²ms¹, may he cut off his hand for you (in promise) that he will indeed have vengeance against him who has committed this act’ (Readings and translations by Al-Jallad 2015: 227). 7 Cf. Lane 1863–1893: 662a. 8 See for example the grieving contexts in Section 2.2 below. Knauf suggested to translate ḥwb ʿl- in such contexts as ‘to be concerned, worry about pn’, but he did not offer any explanation for this suggestion.

notes on ḥwb in safaitic

71

csns 95 l hʿṣd bn yd w ḥwl { f } bʿd h ʾlt By Hʿṣd son of Yd and he suffered from a squint, so O Lt keep away! Incidentally, a close reading of the inscription (see Fig. 1 and the tracing in Fig. 2) shows that ḥwl, which would represent the first attestation of such a word in Safaitic, can instead be read as another instance of ḥwb. Even though damage on the rock has obscured the lower part of the letter, a closer look shows that what Clark read as an l should be read as a b. Moreover, Clark took as the b of bʿd a letter that is more likely to represent an s¹. This glyph was probably forgotten by the author and added later, since it looks quite pressed between the preceding f and the following ʿ. The shaft of the s¹ is rather visible and rules out the reading of the letter as a b. The new reading of the inscription produces the following text: … w ḥw{b} { f } {s¹}ʿd-h ʾlt The typology of this text is paralleled by a number of inscriptions in which ḥwb is followed by prayers (see Section 3 below for a translation). 2.2 ḥwb in Grieving Contexts In several inscriptions the verb ḥwb is followed by the preposition ʿl- and the name of a relative or beloved of the author. This formula is largely attested in Safaitic grieving inscriptions, and has led Winnett and Harding to translate ḥwb as ‘to grieve’ (Winnett and Harding 1978: 51–52). Their translation is also corroborated by some of the meanings attested in the Arabic lexica for the tDstem of the root: ‘he cried out, expressing pain or grief or sorrow, or lamenting’.9 Al-Jallad adopted the translation ‘to lament’,10 which is also the one used here. See the following examples: wh 2472 l f{ʾ}rn bn ṯʿr w ḥwb ʿl-ḥbb-h By Fʾrn son of Ṯʿr and he lamented over his beloved qur 956.67.1 l bnʾmt bn ʿm w ḥwb ʿl-ʾḫ-h By Bnʾmt son of ʿm and he lamented over his brother krs 1251 l ʿḍl bn s¹ry bn s¹lm w ḥwb ʿl-nn w ḥbb f ḥbb By ʿḍl son of S¹ry son of S¹lm and he lamented over Nn and he was loved and loved indeed 9 10

Lane 1863–1893: 662a–b. Al-Jallad 2015: 323.

72

della puppa

It should be mentioned that the author of the last quoted inscription wrote another text on the very same rock: krs 1250 l ʿḍl bn s¹ry bn s¹lm w wgm ʿl-nn By ʿḍl son of S¹ry son of S¹lm and he grieved for Nn Above krs 1250, inside the same cartouche, there is also the inscription krs 1249 l nn bn s¹ry bn s¹lm. It is written in bigger letters than krs 1250, and it might represent a memorial inscription carved for the same Nn who was grieved upon (the brother of the author of krs 1250 and 1251). Alternatively, it is possible that Nn himself was the author of krs 1249 and that ʿḍl wrote the two other inscriptions afterwards.11 In any case, the very fact that ḥwb appears in a cluster of inscriptions belonging to a grieving setting, lends further circumstantial evidence to the idea that ḥwb in Safaitic had a meaning which could be associated with such contexts.

3

The New Interpretation

While no satisfactory translation for the expression ḥwb ʾl-rḍw has been yet offered, an interpretation of ḥwb as ‘to lament (over someone)’ suits the inscriptions bearing the phrase ḥwb ʿl-[person]. At this point, let us go back to the evidence gathered from the Arabic lexica. In Classical Arabic the G- and the C-stem of ḥwb present meanings related to ‘sin’: ḥāba ‘He sinned; committed a sin, or crime, did what was unlawful’; ʾaḥwaba ‘He pursued a course that led him to sin, or crime’ (Lane 1863–1893: 662a). The D- and the tD-stem, however, exhibit further meanings: ḥawwaba bi-l-ʾibili means ‘he chid the camels by the cry ḥawbi ḥawbi’ and taḥawwaba has, among other meanings,12 ‘He cried out, expressing pain or grief or sorrow, or lamenting, or complaining: he cried aloud, or vehemently … He wept, in impatience, or sorrow, and with loud crying … he cried out, or aloud’ (Lane 1863–1893: 662a–b).13

11

12

13

Some Safaitic inscriptions attest the practice of writing grieving texts upon finding the inscription of a beloved or family member (cf. the ‘Inscription finding’ formulae in AlJallad 2015, § 22.5). taḥawwaba can also mean ‘He abstained from, shunned or avoided, sin, or crime, put it away from himself: he applied himself to acts, or exercises, of devotion’, representing, according to Lane, one of those instances in which the tD-stem is deprived of the meaning of the root (Lane 1863–1893: 662a), in this case the meaning ‘to sin’. Cf. also Yemeni Arabic ḥawb ‘urging crying’ (Piamenta 1990: 112).

notes on ḥwb in safaitic

73

This evidence shows that the lexica provide meanings that can be traced back to two main semantic nuclei: ‘sin’ and ‘cry out’. Unlike the first, the latter would seem to constitute a perfect fit for the phrase ḥwb ʾl-rḍw, which would then translate: KnGQ 5 l ngʾ bn qld w ḥwb ʾl-rḍw By Ngʾ son of Qld and he cried out to Rḍw The expression can therefore be understood as an invocation to the deity, and the nominal form in csns 918 could be accordingly translated as ‘supplication’: csns 918 l bʾs¹ bn ḥbb h-ḥ{w}b ʾl-rḍw By Bʾs¹ son of Ḥbb is this14 supplication to Rḍw Moreover, the translation ‘to cry out’ also suits other contexts, see for example: krs 1377 l nẓl bn dḥmt w ḥwb ʿny m-{g}rb15 f h lt rwḥ [m-]bʾs¹ By Nẓl son of Dḥmt and he cried out, suffering from scabies so, O Lt, deliver from misfortune qur 533.20.1 l gry bn mġyr w tẓr mny f {h} r{ḍ}{w} s¹{ʿ}d ḏ ḥwb By Gry son of Mġyr and Fate lay in wait16 so, O Rḍw, help him who cries out

14

15

16

It is difficult to determine the pragmatic function of the demonstrative article preceding ḥwb. It could refer to the inscription itself as written expression of the prayer, or perhaps to the oral performance of the prayer. My reading of this word differs from the editio princeps, which read the first two glyphs as {q} and b respectively. As rightly pointed out in the ociana commentary to the inscription (http://krc.orient.ox.ac.uk/ociana/corpus/pages/OCIANA_0022006.html, accessed on 1st June 2016), the line crossing the circle of the {q}—and which would form its crossbar— is carved over one side of the circle rather than centrally, and it could be extraneous to the letter. This line in fact would seem to belong to the cartouche of the other inscription on the rock (krs 1376), which is decorated with seven protruding lines. Thus, the letter consists only of a circle and can be interpreted as a {g}. The interpretation of the second glyph, a curving line, is more ambiguous, since it could represent either a b or an r. Considering that an r would produce the attested word grb ‘scabies’, I opted for the latter. The r of rwḥ has a different shape, but note that in this inscription also the bs and the ms exhibit varying shapes. I followed the translation of the Safaitic expression tẓr mny proposed in Al-Jallad 2015.

74

della puppa

Some inscriptions follow the structure l pn w ḥwb + [prayer]. In such cases, ḥwb might also refer to the performance of the actual prayer: csns 95 l hʿṣd bn yd w ḥw{b} { f } {s¹}ʿd-h ʾlt By Hʿṣd son of Yd and he cried out so, help him ʾlt! kwq 118 l nkf bn ʾgd bn lqf bn ryḍ bn rbʿt bn wdnn bn ẖfdt bn ʾbḥ w ḥwb f h rḍy ġnmt17 By Nkf son of ʾgd son of Lqf son of Ryḍ son of Rbʿt son of Wdnn son of Ḫfdt son of ʾbḥ and he cried out so, O Rḍy, grant booty AbaNS 121 l mrʾt bn ykbr bn qnʾl w ḥwb f rḍw ġnyt By Mrʾt son of Ykbr son of Qnʾl and he cried out and so Rḍw grant abundance ceds 356 (see Fig. 4) attests the substantive ḥwbt in the context of a prayer. The -t suffix could reflect a feminine plural ending */āt/, and the noun might be translated either as ‘supplications’ or as ‘lamentations’:18 ceds 356 l ʿbdy bn s¹mk w s¹ʿd-h rḍy m-ḥwbt-h By ʿbdy son of S¹mk and may Rḍy help him on account of his supplications (/lamentations)

4

ḥwb and the Female Figure

One inscription from northern Jordan (wh 300, see Fig. 5) offers some further, although quite puzzling evidence for the interpretation of the word ḥwb and of some recurring anthropomorphic figures in Safaitic rock art. The inscription is associated to the drawing of a she-camel and of a woman with raised hands. Representations of female figures constitute a known motif in Safaitic rock art. Some of the main features shared by such depictions are the long hair and the raised arms, which are sometimes used to draw out the hair. Such is the case for the figure associated to wh 300. This inscription states:

17 18

See Fig. 3. Alternatively, the -t suffix could represent an */at/ ending and one could compare this Safaitic noun to Classical Arabic ḥawbatun ‘Want; poverty; indigence’ (Lane 1863–1893: 662c).

notes on ḥwb in safaitic

75

wh 300 l ʾs¹ bn ḫyft h-bkrt w ḥwb h-mrʾt By ʾs¹ son of Ḫyft is the she-camel and the crying out of the woman Above the inscription there is also a stick-figure man with upraised arms and, according to Winnett and Harding, five circles at the bottom.19 This stick-figure motif mostly accompanies the geometric figure of seven dots, and appears in a number of inscriptions containing the word ḥwb as well (see for example Figs. 3–4). Winnett and Harding translated the phrase ḥwb h-mrʾt as ‘he loved the woman’, interpreting ḥwb as a form derived from ḥabba ‘to love’. Such a development, however, can only be explained through an ad-hoc phonological process unattested in Safaitic. Instead, the expression ḥwb h-mrʾt could be understood as a genitive construction and ḥwb as a verbal noun. The phrase ḥwb h-mrʾt seems to have the same syntactic relationship to the lām auctoris as h-bkrt, since the whole inscription claims the authorship of the two pieces of rock art. ḥwb h-mrʾt offers a unique ‘caption’ of the motif of the female figure in Safaitic rock art, otherwise mostly referred to as dmyt ‘image’, ġlmt ‘slave-girl’, or zmrt ‘musician’.20 In light of this inscription, and following the translations of the word ḥwb discussed above, it would be tempting to interpret some of the female figures of Safaitic rock art as women engaged in a crying activity of some sort: was it a grieving or perhaps a praying custom? Or a yet different practice which involved the participation of women? In his 2012’s paper, Macdonald surveyed closely various occurrences of these female figures in Safaitic and other North Arabian rock art. While rejecting old assumptions that they would represent goddesses, he suggested that these depictions should be subdivided in two main classes: entertainers (engaged in dancing, singing or playing instruments) or, especially in the context of battle scenes, cheerleaders. I would suggest that another class of female figures in Safaitic rock art could be the representation of women in lamentation, perhaps as part of a grieving or of a praying custom. At the same time, it is possible that such a category overlapped to a certain extent with the class of women described by Macdonald as ‘cheerleaders’: it could portray women engaged in wailing or praying activities in war contexts. Indeed, this inscription raises more questions than it provides answers.

19 20

Winnett and Harding 1978: 93; this is unfortunately hard to see from the picture. See Macdonald 2012: 271.

76 5

della puppa

Conclusion

In this paper, I argued that ḥwb had the basic meaning ‘to cry out’ in Safaitic. When it is followed by the prepositional phrase ʾl-rḍw, I proposed the translation ‘to cry out (to the deity)’ (KnGQ 5), and, in the case of the nominal form in csns 918, ‘supplication (to the deity)’. Unlike Knauf’s and Clark’s interpretations, these translations offer the advantage of harmonizing the information from the lexica with the epigraphic evidence. The translation ‘to lament (over someone)’ suggested by Al-Jallad suits better the occurrences of the word in grieving contexts. The use of ḥwb in a unique ‘caption’ next to the representation of a woman with raised hands in wh 300 might suggest that this motif of Safaitic rock art was in some cases meant to portray either lamenting or praying women.

Sigla AbaNS c ceds csns HaNSB hnsd KnGQ krs kwq qur wh

Safaitic inscriptions in Ababneh 2005 Safaitic inscriptions in Ryckmans 1950–1951 Safaitic inscriptions recorded by V.A. Clark on the Eastern Desert Inscriptions Survey in 1980, and published on ociana Safaitic inscriptions in Clark 1979 Safaitic inscriptions in Ḥarāḥišah 2010 Safaitic inscriptions in Al-Ḥāǧǧ 2015, published on ociana Safaitic inscriptions in Knauf 1991 Safaitic inscriptions recorded by G.M.H. King on the Basalt Desert Rescue Survey in north-eastern Jordan in 1989, and published on ociana Safaitic inscriptions from Wadi Qattafi recorded by G.M.H. King, and published on ociana Safaitic inscriptions recorded within the framework of the Jebel Qurma Archaeological Landscape Project, and edited by Chiara Della Puppa Safaitic inscriptions in Winnett and Harding 1978

Bibliography Ababneh, M. (2005). Neue safaitische Inschriften und deren bildliche Darstellungen. (Semitica et Semitohamitica Berolinensia 6). Aachen: Shaker Verlag. Al-Ḥāǧǧ, ʿA.ʿA.ʿA (2015). Dirāsāt nuqūš ṣafawiyyah ǧadīdah min minṭaqat al-ḍuwaylah fī ʾl-bādiyah al-urduniyyah al-šamāliyyah al-šarqiyyah. Unpublished ma Thesis, Hashemite University, Jordan.

notes on ḥwb in safaitic

77

Al-Jallad, A. (2015). An Outline of a Grammar of the Safaitic Inscriptions. (Studies in Semitic Languages and Linguistics 80). Leiden: Brill. Clark, V.A. (1979). A Study of New Safaitic Inscriptions from Jordan. A Thesis presented for the degree of Doctor of Philosophy, Department of Middle Eastern Studies, University of Melbourne. Ann Arbor, mi: University Microfilms International. Ḥarāḥišah, R. (2010). Nuqūš ṣafāʾiyyah mina ʾl-bādiyah al-ʾurduniyyah al-šimāliyyah alšarqiyyah—dirāsah wa-taḥlīl. Amman: Ward. Knauf, E.A. (1991). More Notes on Ǧabal Qurma, Minaeans and Safaites. Zeitschrift des Deutschen Palästina-Vereins. Bd. 107, pp. 92–101. Lane, E.W. (1863–1893). An Arabic-English Lexicon, Derived from the Best and Most Copious Eastern Sources 8 vols. London: Williams and Norgate. Macdonald, M.C.A. (2012). Goddesses, dancing girls or cheerleaders? Perceptions of the divine and the female form in the rock art of pre-Islamic North Arabia. In I. Sachet & Ch.J. Robin (eds), Dieux et déesses d’Arabie. Images et représentations: Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 Octobre 2007. (Orient et Méditerranée, 7), pp. 261–297. Paris: De Boccard, 2012. Piamenta, M. (1990). Dictionary of Post-Classical Yemeni Arabic. Vol. 1 Leiden: Brill. Ryckmans, G. (ed.) (1950–1951). Corpus Inscriptionum Semiticarum. Pars v. Paris: Imprimerie Nationale. Sokoloff, M. (1990). A Dictionary of Jewish Palestinian Aramaic of the Byzantine Period. (Dictionaries of Talmud, Midrash, and Talgum ii). Bar Ilan University Press. Winnett, F.V. and G.L. Harding (1978). Inscriptions from fifty Safaitic cairns. University of Toronto Press.

78

della puppa

figure 1

hnsd 121 = csns 95 photo: ociana

figure 2

Tracing of hnsd 121 = csns 95 drawn by the author

notes on ḥwb in safaitic

figure 3

kwq 118 photo: ociana

figure 4

ceds 356 with tracing by the author photo: ociana

79

80

figure 5

della puppa

wh 300 with tracing by the author photo: ociana

chapter 5

Traditional Music or Religious Ritual? Ancient Rock Art Illumined by Bedouin Custom Ali al-Manaser

Introduction The four new inscriptions published in this article were found in Wadi Salmā in Jordan. The site is 35km north-east of the village of aṣ-Ṣafawī (fig. 1). The wadi itself is approximately 13km long and contains a remarkable concentration of Safaitic, Greek and Islamic Arabic inscriptions on the black, patinated basalt stones that make up the landscape of the ḥarra.

The Inscriptions Inscription No. 1 ( fig. 2a) l s¹ʿd bn gd h-ḫṭṭ By S¹ʿd son of Gd is the carving. This inscription begins as the vast majority of the Safaitic inscriptions do, with the letter l, sometimes known as the lam auctoris and usually translated as “by” (al-Manaser 2008: 75). The two names are well-known from the Safaitic corpus, although the exact combination has not been seen before. Next to the Safaitic inscription are engravings of four figures and an incised line surrounds the inscription and the drawings. Between two of the figures there are seven lines which form an apotropaic sign very common in Safaitic inscriptions. Three of the figures appear to be male and all are looking towards a fourth, the largest, which appears to be a woman playing a double pipe instrument. The artist depicts the female figure’s breasts as two dots in front of a hatched pattern on her upper body, which could be intended to represent her breasts covered by her clothing. If this is correct, the horizontal line at the level of her groin was presumably accidental or added later. The other three figures have no distinguishing sexual characteristics and are presumably male. The

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

82

al-manaser

figure at the top right of the engraving appears to be holding a daf, a type of frame-drum made of hardwood over which is stretched a membrane, usually made of goatskin. The figure below appears to be clapping. M.C.A. Macdonald suggested to me (p.c.) that the central figure is dancing with a swinging belt and wild hair which was later hammered over. She is holding the lyre in an unnatural position, at arm’s length. The wavy lines to the left of the lyre are the most difficult feature to explain. It is possible that this final figure was originally depicted as a woman since previous research on Safaitic drawings of musical instruments has shown that the lyre was typically played by a woman (see hsd fig. 1). It is possible that the hair was hammered over at a later date. Inscription No. 2 ( fig. 2b) l kʾt bn ġṯ h-ḫṭṭ By Kʾt son of Ġṯ is the carving. The figure on the left appears to be playing a double pipe. The style of the hair suggests that it is a woman, since similar hair arrangement has previously been found on a woman (hch 79; wh 442, 568). However, her hair is shorter than that which has been previously found on female figures (see Macdonald 2012: fig. 3). The artist may have tried to depict the subject wearing loose trousers since her calves are considerably thinner than her legs above the knee. This figure’s grip of the double pipe with both hands is similar to the one found in hch 79 and figure 2d here. The tubes of the pipe are separated to show that there are indeed two of them. The position of the three fingers of the left hand shows that both tubes of the double pipe are being held at once. Other depictions (see hsd fig. 3) of a double pipe show the pipes closely bound rather than with a space between them and all these depict the pipe as held with the hands overlapping both segments. The right figure on the stone is probably a male figure—because of the lack of visible hair—who may be holding castanets. The triangle drawn in front of him is difficult to interpret: it may be a mistake or intended to represent part of a musical instrument, such as a drum, but there is not enough detail to decide.

traditional music or religious ritual?

83

Inscription No. 3 ( fig. 2c) l ḥṭm bn s¹mm bn ʿwḏ w h ylt wqy{t} m {b}ʾs¹ h-s¹nt w ġnmt1 By Ḥṭm son of S¹mm son of ʿwḏ and so Lt [grant] protection from distress this year and booty. This stone has been damaged in several places but fortunately the inscription remains unscathed. On the left-hand side of the stone there is a male human figure. This type of figure is common in Safaitic inscriptions (see wh 2502, 2669, 2673; Ababneh 2005: 258). There is another figure playing a lyre in the lower right. It has short, spiky hair, and is in the typical playing position. In the upper left corner, there are some random letters that are not part of the inscription. Inscription No. 4 ( fig. 2d) l ʾs¹lh bn ġṯ h-ḫṭṭ By ʾs¹lh son of Ġṯ is the carving. On the left-hand corner of this stone there is a female figure, as suggested by the cleanly engraved hair, playing a double pipe which the artist has evidently tried to show as tied together at various points along the pipe. The form of the woman shows a belt hanging from either side of her waist. The only other known similar example of a woman wearing such a belt while playing an instrument is the one found in the basalt etching published by Harding (1969: fig. 1; shown as fig. 4b here). These belts are however very common on figures of singing or dancing girls (Macdonald 2012: fig. 1, 6a, b, 7a, and 8a, c). The female figure is standing with the same posture as the others so far examined, half-bent in a stiff stance. This is known to convey that the performer is concentrating and that it accordingly requires great prowess to carry out this activity. The artist has made the female figure much larger than the male figures standing either side of her. It could be suggested that this was to show that her activity was the most important of those being shown (the male figure to her right plays the daf and the one to the left appears to be clapping). The

1 Editorial sigla: { } enclose doubtful letters, [ ] enclose letters which are restored, ---- represents a passage in which one or more letters are completely destroyed, / between two letters indicates an alternative reading. Semitic roots are represented in capital letters, e.g. rwḥ.

84

al-manaser

relatively large size of the female figure is significant as it clearly shows the central role that women played in musical activity in this society. In all known rock drawings where the gender of the figures can be determined, only women are depicted as playing the pipe. Another inscription, kept in the Damascus Museum (Museum no. 29067; fig. 4a), is accompanied by a drawing which depicts a scene in which a horseman carrying a spear is chasing (or possibly raiding) a foal. A dog2 and a scorpion3 are also illustrated in the drawing and between them is a woman with long hair and wearing what appears to be a belt the ends of which hang down from her waist. She is playing what appears to be a single pipe, perhaps a šabbābah. The author has written that the drawing (dmyt) belongs to him (w l-h h-dmyt), but beyond giving us his name he does not tell us anything about it (al-Muʾaḏḏin 1995: 24). It is interesting that the author of the inscription who claims the drawing, s¹krn bn ḥmy bn s¹krn, also carved his name on the stone bearing the inscription and drawing published by Harding (hsd, fig. 4b here). However, on the latter, it is not he but another member of the same lineage group (ḫr bn s²rk bn ḫr) who claims the drawing. The drawing includes two women with lyres who resemble the Yemeni-Ethiopian type, and a male rider (Braun 1999: 163; idem 2002: 221).

Discussion Many scholars have interpreted depictions of musical gatherings in rock drawings as having a religious function, for example rituals such as a rain ceremony. The figures with hands held aloft in KhBG 17 (fig. 3a) have thus been interpreted by al-Khraysheh as taking part in a ceremony to attract rain (2002: 17–18). Depicted in mssh 12 (fig. 3b) is a scene in which a group of men on the right stand in a row clapping. A male figure in the bottom-central portion of the engraving is possibly playing the castanets. In the middle of the scene, another possibly male figure dances. He appears to be wearing trousers, decorated below the knee with tassels that may be made of tails, quills or hair of some kind. On the far left, a woman plays the double pipe, while another claps. In this scene, it is very clear that we are dealing with some kind of dance ceremony which can be compared with the daḥiyyah dance. This dance is known today in the north of Saudi Arabia and across Jordan and Syria, as well as among

2 Compare the dog in the scene on hsd (fig. 4b here). 3 This is one of only a very few rock drawings showing a scorpion.

traditional music or religious ritual?

85

the Bedouin in Sinai and among Iraqis in the western Iraqi desert. However, only Bedouin practice this style of dance at present. The origin of this dance is unclear but the tradition maintains that a group of men travelling in the desert stopped to make camp and discovered a rival group intending to attack them during the night. The first group gathered themselves with their camels and began to clap. The camels made a lot of noise in response to the clapping, frightening the rival group away and averting the attack. This could explain why contemporary performers of the dance make a sound that has been described by modern observers as sounding like the grunts of camels. Another theory about the origins of this dance is that the people in the desert may have devised it as a way of keeping warm during cold desert nights, huddling together to share body heat and clapping to keep moving. However, this explanation does not account for the meaning of the sounds made during the dance, which may just be lost to us. The geographical spread of this dance has led to different names for it, daḥiyyah in Jordan, daḥah in Saudi Arabia and daḥiyyah or sāmr in Iraq but, despite the variety of names, the movements and sounds are exactly the same, pointing to a common origin from one specific locality. The performance of this dance begins with a group of men who separate into two groups. A poem is recited which requires a call and response between the two groups: the first group utters a line of poetry, the second group must repeat it with the difference that the second group must start with the letter that the first group finished on (for example if the final letter in the final word was l, the second group must start by repeatedly saying the sound of that letter, i.e. “lalal”). An inscription published by Harding (hch 79, see fig. 3c) depicts two figures, one dancing and one playing what appears to be a double flute. It is clear that the woman is playing the double flute. Next to the drawing is a Safaitic inscription that includes the word zmrt, which derives from zamara “to play on the flute” (Lane 1251b). In this inscription, zmrt could be interpreted as a verb, the inference being that the woman or slave girl is playing music for the other figure. On the other hand, Al-Jallad has proposed that the h- in h- dmyt zmrt is a presentative particle. He would then translate “behold, a drawing of a flute player”. Alternatively, he suggests that dmyt is a first person singular verb of the suffix conjugation meaning “behold, I have drawn a flute player”. In both cases, the expression would so far be unique (see Al-Jallad commentary of hch 79 in ociana). The word zmr occurs in c 1104 (fig. 3d). Since the drawing there clearly shows a woman, one might have expected zmrt, the feminine form of the verb, as in hch 79, and it is possible that the t was omitted by the copyist. However, if zmr is correct, it could be related to Classical Arabic zawmar “playing”, “a

86

al-manaser

player” (Lane 1251a). In Arabic, the root specifically concerns flute playing, as can be seen in the drawing next to hch 79. However, in c 1104 the woman has no instrument. In Lisān al-ʿArab we find the expression ʾimraʾatun zāmiratun, meaning the woman who plays music “piped through reeds” (Ibn Manẓūr: 327). The nature of pastoral communities required lightweight instruments that were easily transportable, especially the instruments that were used by shepherds. Of the five main instruments depicted in Safaitic drawings, three are wind instruments. The first could be the double pipe, called the yarġul (fig. 5a) which is generally used by Jordanian and Palestinian peoples and consists of two tubes, one shorter than the other. The second could be the majwiz or maqrun (fig. 5b), two pipes of equal length tied together.4 The third wind instrument, called the šabbābah, is a single pipe with seven holes. All three wind instruments are traditionally made from reeds though today some are made of metal, plastic and other materials.

Conclusion Scenes of musical entertainment occur in a number of Safaitic rock drawings. There is nothing to suggest that these scenes represent anything other than entertainment and there is no evidence at all that they had a religious significance (see hch 79; c 2839–2840; hsd 1–5). The majority show the double pipe known today as the majwiz or maqrun which consists of two pipes of equal length bound together (fig. 5b) but in some cases they might be what is known today as the yarġul which consists of two pipes of unequal length bound together (fig. 5a, see Ababneh 2005: 258 and mssh 12). There is no evidence in these drawings for the use of the double aulos in which the pipes diverge forming an inverted v. In those cases where the gender of the player can be discerned, only women seem to be playing the pipes (see Landels 1999: 24–30). The other popular instrument in these scenes is the lyre, which seems to be played mostly, if not entirely, by women (see hsd 1–5). Small drums, like the modern daf, also occur in some scenes, as well as what may be castanets or small bells. These seem to be played by men. There are finally scenes in which those not playing instruments are clapping and/or dancing.

4 Each word comes from a different dialect but they describe the same instrument. The first, majwiz, comes from zawj, meaning “two together”. Maqrun means “to tie” together.

traditional music or religious ritual?

87

Given that so many Safaitic inscriptions describe the unhappiness and insecurity of their authors, it is good to see that they were also able to enjoy themselves and were willing to record it.

Sigla C hch hsd ociana

KhBG mssh wh

Ryckmans 1950–1951. Harding 1953. Harding 1969. The Online Corpus of the Inscriptions of Ancient North Arabia project at the Khalili Research Centre, University of Oxford. Accessed September 5, 2016 http://krc.orient.ox.ac.uk/ociana/index.php Al-Khraysheh 2002. Al-Maʿānī 1999. Winnett and Harding 1978.

References Ababneh, M.I. 2005. Neue safaitische Inschriften und deren bildliche Darstellung. (Semitica et Semitohamitica Berolinensia 6). Aachen: Shaker. Braun, J. 1999. Die Musikkultur Altisraels/Palästinas. Studien zu archäologischen, schriftlichen und vergleichenden Quellen. (Orbis Biblicus et Orientalis 164). Göttingen: Vandenhoeck and Ruprecht. Braun, J. 2002. Music in Ancient Israel/Palestine. Archaeological, Written, and Comparative sources. Grand Rapids, mi: Eerdmans. Harding, G.L. 1953. “The Cairn of Haniʾ.” Annual of the Department of Antiquities of Jordan 2: 8–56, pl. 1–7. Harding, G.L. 1969. “A Safaitic Drawing and Text.” Levant 1: 68–72, pl. 19. Ibn Manẓūr, Muḥammad b. al-Mukarram. 1955–1956. Lisān al-ʿArab, edited by A.F. alWānib. 15 vol. Bayrūt: Dār Ṣāder / Dār Bayrūt. al-Khraysheh, F.H. 2002. al-nuqūš al-ṣafawiyyah min biyār al-ġuṣayn. (Mudawwanat alnuqūš al-ʾurduniyyah, 1). Irbid: Jāmiʿat al-yarmūk. Landels, J. 1999. Music in Ancient Greece and Rome. London: Routledge. Lane, E.W. 1863–1893. An Arabic-English Lexicon. 8 vol. London: Willams and Norgate. al-Maʿānī, S.ʿA. 1999. “Nuqūš ṣafawiyyah jadīdah min wādī salmā / al-ʾurdun.” Majallat al-ʿulūm al-ʾijtimāʿiyyah wa-l-ʾinsāniyyah 4 (2): 29–48. Macdonald, M.C.A. 2012. “Goddesses, Dancing Girls or Cheerleaders? Perceptions of the Divine and the Female Form in the Rock Art of Pre-Islamic North Arabia.” In

88

al-manaser

Dieux et déesses d’Arabie. Images et représentations. Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 octobre 2007, edited by I. Sachet and C.J. Robin, 261–297. (Orient & Méditerranée 7). Paris: De Boccard. al-Manaser, A. 2008. Ein Korpus neuer safaitischer Inschriften aus Jordanien. (Semitica et Semitohamitica Berolinensia 10). Aachen: Shaker Verlag. al-Muʾaḏḏin, M. 1995. “Al-naqāʾiš wa-ʾl-rusūm al-ṣaẖriyyah fī ʾl-ʾāṯār al-ʿarabiyyah.” In Al-naqāʾiš wa-ʾl-rusūm al-ṣaẖriyyah fī ʾl-waṭan al-ʿarabī. Al-muʾtamar al-ṯāliṯ ʿašar lil-āṯār, al-jamāhiriyyah al-ʿuẓmā. Trāblus, 1–7 October 1995, 7–31. Al-munaẓẓamah al-ʿarabiyyah li-l-tarbiyah wa-ʾl-ṯaqāfah wa-ʾl-ʿulūm. Ryckmans, G. 1950–1951. Corpus Inscriptionum Semiticarum. Pars v. Inscriptiones Saracenicas Continens, Tomus 1. Inscriptiones Safaiticae. Paris: Imprimerie nationale. Winnett, F.V., and Harding, G.L. 1978. Inscriptions from Fifty Safaitic Cairns. (Near and Middle East Series 9). Toronto: University of Toronto Press.

figure 1

Map of Jordan showing the location of Wādī Salmā source: google earth

traditional music or religious ritual?

figure 2a Inscription no. 1

figure 2b Inscription no. 2

89

90

figure 2c Inscription no. 3

figure 2s Inscription no. 4

al-manaser

traditional music or religious ritual?

figure 3a Inscription KhBG 17

91

92

figure 3b Inscription mssh 12

al-manaser

traditional music or religious ritual?

figure 3c Inscription hch 79

figure 3d Inscription c 1104.

93

94

figure 4a Inscription Damascus Museum no. 29067 harding 1969

figure 4b hsd harding 1969

al-manaser

traditional music or religious ritual?

figure 5

Musical instruments. a. Double pipe, yarġul; b. Majwiz or maqrun.

95

chapter 6

The Formularies and Their Historical Implications: Two Examples from Ancient South Arabian Epigraphic Documentation Alessandra Avanzini

With Michael I share, I believe, more than a motive of interest and enthusiasm. Something which, without doubt, unites us is our conviction that the creation of it databanks is of great use for the study of the texts of ancient Arabia. Michael is the promoter of the project ociana; while I am the promoter of the project dasi. Both archives are for the moment connected by a link referring one to the other. In the near future common lexical research will be possible in both archives. The advantages of online archives are evident. We will have a databank first and foremost safe from the oblivion of hundreds of published texts in disparate publications not always easy to retrieve. An online archive is easily integrated and updated with new texts and improvements in editions of collected texts. However not only for philological and lexical studies is it of fundamental importance to have consultable on line inscriptions for sub groups, which are articulated chronologically and geographically. This will also help to retrieve historical problems generally. Most of the events which took place over the very long history of Southern Arabia may be reconstructed (a phenomenon which is probably unique in the ancient world) by basing them almost exclusively on the region’s vast epigraphic corpus. We would have certainly reconstructed the thoughts, beliefs and cultural complexity of the ancient South Arabians if we had been aware of their mythology. While the inscriptions are silent on such matters, they are useful for investigating certain general aspects of Ancient South Arabian (asa) history and, in particular, cultural relations among the various areas which had a role in asa culture. asa culture is not definable only on the basis of a geographical area or a historical period, it is also a linguistic family, a writing tradition, shared statecraft, a common identity (based, for the entire long asa history, upon a god, a king and a tribal group), and artistic taste that allow us to recognize and define a culture common to the whole of Western (and Southwestern) Arabia, before Islam. Alongside many common elements there are some aspects which differentiate the various asa cultural areas. For the aspects present in the whole of

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

the formularies and their historical implications

97

the asa culture we may hypothesize a centre of cultural diffusion. The process that led to the creation of asa culture, in many cases, may date back to the proto-historic phase. For those aspects related to a single area, autonomous innovations can be presupposed. However the latter could have taken place in a proto-historic period or during events of the long asa history. asa alphabetic writing originated, without doubt, in one place, which became its centre of diffusion towards other areas of asa culture. The oral teaching of writing, as is always the case, for the teaching of writing in any place and historical period, must be presupposed. At the beginning of asa history, Sabaʾ is the most enterprising commercial kingdom; the only one, and not by chance, mentioned in indirect sources. The centre of diffusion of alphabetic writing could be imagined to have existed in the heart of the Sabaean realm. But, as often is the case in the ancient world, areas of contact are those of the bearers of great cultural innovations. For this reason, and for what is now known, a town in the Jawf region, the city of Nashshān, allied for almost the whole of the eighth century with Sabaʾ, could, as a mere hypothesis, be recognized as being the centre for the spread of asa writing. Various elements give Nashshān the role of being the forefront in the creation and spread of the writing. Most of the private texts in minuscule writing noted up to the present originate from Nashshān, chronologically distributed during the long asa history; so much so as to render perfectly commendable the affirmation of Macdonald: “… it would seem probable that we may have the products of a single scriptorium stretching over some 1500 years”.1 From the most ancient private texts from Nashshān it is possible to posit a common origin of the asa writing, both public and private.2 As well as having an absolute date of an inscritbed stick from Nashshān from the end of the second millennium,3 if confirmed by other dates, this would place the creation of the asa alphabet in the same period as the first attempts of linear alphabetic writings in the Levant. And would force a rethinking on the history of alphabetic writing and would remove credibility from the hypothesis of a direct dependence of the asa alphabet upon the Proto-Canaanite alphabet of the north, as has been recently held by Benjamin Sass.4 Learning writing in the various areas of the Yemen did not regard only the learning of a writing technique but also the learning of textual models into 1 2 3 4

Macdonald 2015: 17. Macdonald 2015: 14 and note 54. Drewes et alii 2013. B. Sass 2005.

98

avanzini

which the epigraphic text must be inserted. In all of the asa documentation the sequence of the various units composing a text is always the same: the subject, the principle verb that characterizes the typology of the text, a possible text which is syntactically and lexically freer and the final invocations. Stereotypical expressions introduce the various units of the text; for example the reason for a dedication or the final invocations to the gods or to the kings. As time went by various aspects specific to a geographical area or a chronological period began to emerge with the creation of writing rules (e.g. the use of matres lectionis for writing vowels) and original formularies. Starting out from inscriptions which have recently been edited, I would like briefly to return to certain historical implications deriving from analysis of the formularies used by a writing school in a certain period and a specific area.

1

The Ancient Jawf

Two writing schools, one in Sabaic (sab) and the other in Minaic (min), are present in the Jawf during the eighth century bc. Relations between Sabaʾ and various city states of the Jawf had long been known. However much more is known of them today thanks to the addition of numerous new texts, which characterize the various phases of war and peace amongst them. Not only is the sab writing school historically predictable in the Jawf, but it is evident in the allied states of the Jawf and those states under its authority. If we content ourselves only with information handed down from the Ṣirwāḥ texts on the great victories of the Sabaean mukarribs Yathaʿamar Watar and Karibil Watar in the Jawf, we will be tempted to identify the contact made between the two schools only in the direction of: Sabaʾ > cities of Jawf, sab > min. But this is not entirely true. A redaction of a text in min follows its own textual rules, and at times this influences redaction of a text in sab. Being in possession of photos of texts reunited in an archive also helps to identify certain palaeographic elements belonging to the two writing schools of the ancient Jawf. M.C.A. Macdonald has written pages of fundamental importance on palaeography.5 I agree wholeheartedly with his severe judgement on the limits of palaeographic studies. However for public texts, for the asa inscriptions, palaeography, if not used incompetently for moving from relative chronological networks to absolute dates, provides elements which are without a doubt historically very interesting. 5 Macdonald 2015: 17–27.

the formularies and their historical implications

99

One of the most evident aspects that characterise inscriptions in min compared to those in sab from the ancient Jawf is linked to the use of the boustrophedon ductus. The boustrophedon ductus is rarely present in the min texts; while it is used regularly in the sab texts in the same region and in the same period. At Haram in the eighth-seventh century bc, writing is in both sab and min. Two contemporary texts (Haram 15 and Haram 12) which refer to the same events are written with one being in sab and the other in min. The sab text is boustrophedon. However the contemporary min text produced in the same city is not. It was partly predictable (but in any case it is interesting to see it confirmed in the documentation) that the few boustrophedon texts in min are attested only in some cities of the Jawf in close relation to the alliance with Sabaʾ (in the city of Maʿīn, in Nashshān before or after the wars with Sabaʾ6). But also an observation needs to be made regarding the palaeographic evolution of the two corpora. The inscriptions of the eighth century bc are paleographically and stylistically easily recognisable. A good example of this is the diagnostic letter s³ in the form of a star, not normalised, for the dating of inscriptions of the eighth century. At the end of the eighth century inscriptions in min lose their own features in the writing of the previous period, while some texts in sab contain conservative features in their palaeographic style. Previously I have given examples of two sab inscriptions from the beginning of the seventh century: Kamma 33 and mafray-Mushji 23.7 The inscriptions of the Jawf in min and in sab are certainly the product of two different schools with their own rules and palaeographic phases of evolution. The use of certain stylistic traits in the formularies often derives from cultural elements and hardly ever from semantic reasons. One of the geographically and chronologically connoted stylistic elements is the deployment, in the inscriptions of the Jawf in the eighth and seventh centuries, of b-rʾẓ for introducing the final invocations and placing the actions recalled into the text, and its own redaction, under the will of god. Finding such a trait is sufficient for suggesting the provenance and date of a text: from the Jawf in the eighth or seventh century. I held that the presence of such a trait, both in min and sab inscriptions, was proof of the existence of a cultural koiné in the Jawf of this period. However,

6 e. g. Maʿīn 38, Maʿīn 99, ym 23250, ym 29827, Moussaïeff 22. 7 Avanzini 2016: 125, figs 38, 39.

100

avanzini

a consultation of the attestations of b-rʿẓ on the online archive allows me to deduce that my affirmation was without doubt a little hurried. b-rʾẓ is attested in min inscriptions from various cities of the Jawf; however, it is only attested in the sab inscriptions of Haram. I have already observed a particularity of the Haram documentation with texts in sab and min. The presence of b-rʿẓ in sab inscriptions from Haram is a loanword from the min formulary. Incidentally, it is interesting to note that the sab inscription Haram 15, where b-rʿẓ appears, has another loanword from min: the min dedication verb s³lʾ and not the sab verb hqny. Also the fact that b-rʿẓ is not used in later inscriptions of the Jawf is, in my opinion, important. It indicates a cultural hiatus in the history of text redaction in the Jawf. I feel it is incorrect to pose the problem of where the Minaeans come from. The Minaeans had always been in the Jawf and in their city. M. Arbach has recently published some inscriptions of the city of Maʿīn to be dated to the end of the eighth century. Here he demonstrates, without any doubt whatsoever, the presence of a city state and of its king up until the beginning of asa history. One evident element of continuity within the social components of the ancient city and the successive kingdom is mention, in inscription Maʿīn 112 (from the end of the eighth century), of one of the principle families (the Gb’n family) of the successive Minaean kingdom. The city of Maʿīn in the eighth-seventh century is an ally of Sabaʾ.8 The inscription of a Minaean king at Marib provides further concrete evidence of an alliance between the ancient city of Maʿīn and Sabaʾ (Schm/Samsara 3). This could be the simple reason to explain why the city of Maʿīn is not cited in the narration of the wars of Karibil Watar in the Jawf. Furthermore, regarding Barāqish, in the res 3946 inscription it is only recalled that it is walled by Karibil Watar; that is, it is part of the Sabaean power; but the city is not remembered in the wars of the king. Nevertheless the hypothesis of an ‘arrival’ of Minaeans, as A. de Maigret and Ch. Robin have recently suggested,9 is not definitely tied only to the lack of Maʿīn’s citation in the Karibil Watar inscriptions. The Italian mission headed by A. de Maigret excavated part of the city of Barāqish and identified a gap in the archaeological records between a more ancient period (the Sabaean period) and a more recent one. As Ch. Robin quite rightly concludes, the date proposed

8 See, for instance, the mention of an alliance with Sabaʾ at the end of the inscription Maʿīn 102 and in ym 2009. 9 Robin and de Maigret 2009: 74 and 79–80.

the formularies and their historical implications

101

by A. de Maigret for the passage between the two periods at Barāqish coincides with the epigraphic dates, at the end of the seventh and the beginning of the sixth century bc. However, for me the interesting archaeological and epigraphic conclusions of the two scholars do not have much to do with the arrival of the Minaeans from areas (more or less) far away from the Jawf. The abandonment of the alliance with Sabaʾ and the pathway to an alliance with the city of Maʿīn, with its consequent immediate abandonment of the documentation in sab in favour of that in min is a clearly proven fact of history; and it re-enters into the perennial oscillations between alliances and wars among the various cities or the various kingdoms of Southern Arabia. Nevertheless there are certain socio-cultural elements that differentiate the kingdom of Maʿīn from other asa kingdoms contemporary to it. But, what interests us here is that from min documentation there emerge important differences, not only with contemporary kingdoms but also with previous periods. The Minaean kingdom was formed thanks to an alliance of the two cities Maʿīn and Barāqish; to which is also added the principal city of the eighth century bc of the Jawf: Nashshān. This kingdom is not considered to be the heir of the glorious past of the cities of the Jawf.10 It marks a hiatus in the history of the Jawf; even though maintaining evident elements of continuity with the past of the city of Maʿīn starting from the divine pantheon and the language. The language of the kingdom does in fact not appear to present any particular elements with respect to the past (or at least these do not clearly emerge in the texts). It keep s³lʾ, the principle verb of the dedicatory texts; but it loses and innovates certain formularies. For example b-rʿẓ is no longer attested. Within the formularies of the Minaean kingdom’s inscriptions (true ‘acid tests’ for any investigation of complex historical problems) there emerge elements of continuity and discontinuity between the past of the city of Maʿīn in the eighth-seventh century and the kingdom. The presence of non-unique elements argues for an internal process and not for a change in the ethnical components within the city of Maʿīn with the arrival of new populations. I am unconvinced about the various historical hypotheses linking changes in asa history to the arrival of populations from places far away, from somewhere else; such as the arrival of the Sabaens from the north, of the Minaeans from Sinai or from Najrān, and of the Arabs from the desert into Haram. Nevertheless my conviction does not exclude the emergence of populations (in asa history

10

A hiatus with the past is, for instance, evident in the artistic taste (Avanzini 2016: 155).

102

avanzini

in general and particularly in the history of the Jawf) that were living there for long, but who, up to a certain period, were socially marginalised.

2

The City of Maryamat

A second example of the possible historical implications of the presence of autonomous formularies in some geographical areas is found in the inscriptions of al-Ādī (the ancient Maryamat) in the Wadi Ḥarīb. Some recently edited inscriptions demonstrate the existence of a writing school, within the strongly centralised Qatabanian writing school, with certain autonomous features. Wadi Ḥarīb, connected to the central area of the kingdom by the pass of Mablaqa, built by the Qatabanian mukarrib Yadaʿab Dhubayān, appeared to have been an outlying area and culturally dependent. Ḥinū az-Zurayr is most likely, as held by J. Schiettecatte, a new-foundation of the Qatabanian kingdom in the Wadi Ḥarīb.11 Not much more than the name of Maryamat was noted until a few years ago. Three beautiful bronze plaques from Maryamat, dedicated to god Ḥawkam, were published in 2009 by F. Bron. With the documentation available up to that time F. Bron was unable to propose their provenance. However today its connection with Maryamat is secure, thanks to the name of the temple and the name of the principle families of Maryamat cited in the texts. F. Bron,12 in my opinion is right when he hypothesised that even a bronze plaque (csai i, 94 = Atlal), published many years previously, could have had the same provenance. Nevertheless a consistent epigraphic corpus in the second half of the first millennium recently published reveals information of historical importance and defines certain cultural features of the city of Maryamat. I also believe it presupposes that in Maryamat there existed an independent tradition of redacting epigraphic texts. As always when we talk of formularies, the problem is primarily a historical one; and only with great caution can it regard the study of language. I do not know if the inhabitants of Maryamat spoke a dialect variant of qat. The texts do not have grammatical differences; they are written in a correct qat. In the texts recognisable elements of linguistic interference, as in the qat inscriptions from the southern plateau in the last centuries bc, are not present. However, for example, as we shall see, some lexical features are definitely

11 12

Mouton and Schiettecatte 2014: 189–190. Bron 2009: 121.

the formularies and their historical implications

103

present. Also regarding the lexicon, a language of fragmentary attestation like qat, requires great caution when it comes to deducing dialectal differences. The interesting historical data present in the new texts from Maryamat are certainly not only reduced, indirectly, to the presence of a possible autonomous writing school, but they emerge directly from the content of the texts. The inscription al-Ādī 22 confirms a translation different from the previous ones suggested by the editor and myself of the csai i, 115 = Arbach-Sayūn 1 inscription. The war, recounted in the text, is not one of the many wars between the king of Qatabān and the king of Ḥaḍramawt, but between the king of Ḥaḍramawt and a few Ḥaḍrami rebel tribes. The Qatabanians go to the rescue of the king and manage to get him to return to Shabwa. The two texts are important for understanding the history of the Ḥaḍramawt kingdom and its difficulties in keeping the vast territory of the Ḥaḍramawt valley, and its tribes, under a central power.13 A recently edited inscription (Collection privée yéménite 2014–3) again from Maryamat, written under the reign of Shahr Yagul Yuhargib, provides interesting information on the relations between Qatabān and Ḥaḍramawt after the help given by the king of Qatabān to the king of Ḥaḍramawt against some rebellious tribes. As it seems, the king of Ḥaḍramawt had to give something in return to the king of Qatabān, and the latter had to give something to the tribes that followed him to war. In the inscription Collection privée yéménite 2014–3 it is clear that members of the tribe Maryamat permanently settled in Shabwa, founded a temple with the same name as that in their original city, and ensured for some time privileged commercial relations with the kingdom of Ḥaḍramawt. The authors of the two texts, who both took part in the same war in helping the king of Ḥaḍramawt, are members of the tribe of Ḥinū az-Zurayr (in the csai i, 115 = Arbach-Sayūn 1 inscription) and of the Maryamat tribe (in the al-Ādī 22 inscription) respectively. The two cities of the Wadi Ḥarīb had close relations with each other, as is evident from the two inscriptions that recall the reconstruction of their city walls. King Yadaʿab Dhubayān Yuhargib builds the walls of Ḥinū az-Zurayr (csai i, 14 = MQ-Ḥinū az-Zurayr 1). The same king is cited in the inscription atm 867 = al-Ādī 2, whose author is a private citizen who was in charge of the construction work of part of the wall surrounding the city under orders from the king.

13

In the third century ad, tribes from the disintegrated kingdom of Qatabān will flood in support of the king of Ḥaḍramawt and once again the king of Ḥaḍramawt can return to his capital (mafray-al-Miʿsāl 4).

104

avanzini

atm 867 = al-Ādī 2, 1–2: Hwfʿṯt bn Ḥmʿṯt ḏ-Yʿd tqdm l-mrʾ-s¹ Ydʿʾb Ḏbyn mlk Qtbn kl mhlk w-mbny ḫms¹ ṣḥfm “Hwfʿṯt, son of Ḥmʿṯt, of the family Yʿd, directed for his lord Yadaʿab Dhubayān, king of Qatabān, the whole realization and construction of the five curtains”.14 It is noteworthy how in this initial part of the text a lexicon (the main verb tqdm, the substantive mhlk) is attested which is well known and often seen in the texts recalling construction work for the community, carried out, or implemented, by the king. Such building work, involving also maintenance of the city walls and the creation of roads that go through mountains (thus uniting various parts of the kingdom) are feats of great importance for the central power. Redaction of texts that recall such construction appear to adopt the same lexicon throughout the whole kingdom. Some regal decrees which grant ownership of buildings to various families are valid for the two cities of the Wadi Ḥarīb. atm 866, 11–12: ʾbytm s¹ḥrm w-ẓrb ʾmlk Qtbn b-hgrnyhn Mrymtm w-Hrbt “the houses to which the kings of Qatabān decreed and gave the propriety in the two cities Maryamat and Ḥinū az-Zurayr (Hrbt)”. Let turn to some particular lexical features of atm 866. The author of this legal text is a king of Qatabān: Hwfʿm Yhnʿm. Paleographically the inscription can be placed into the period which I refer to as b2 and the king could be the son of Shahr Hilal and the father of Shahr Yagyl Yuhargib. The royal genealogy is (pretty much) certain for these kings.

14

At the end of the text the king is invoked as Yadaʿab Dhubayān Yuhargib. Evidently, there is the same king. In the qat documentation there is a mukarrib Yadaʿab Dhubayān, son of Shahr, great constructor of roads and mountain passes. We do not know if this king is the same who led the wars against Ḥaḍramawt and Sabaʾ: Yadaʿab Dhubayān Yuhanʿim, son of Shahr. The inscription atm 867 bears witness to the possibility of referring to the same king both with three and with two elements of his name (Avanzini 2016: 179).

the formularies and their historical implications

105

The editors of the text read and integrate the royal name Hwfʿm Yhnʿ(m) [bn] (ʾ)nmrm mlk Qtbn. In the royal qat onomastics there are some patronymics which are not included in the very limited list of names attributed to the kings (e.g. S²hr Ġyln bn ʾbs²bm, Ydʿʾb Ġyln bn Frʿkrb). In the inscription atm 866 we find ourselves faced with an interesting problem: either King Hawfʿam Yuhanʿim is another king with this name, and not the son of Shahr Hilal; or the patronymic Shahr Hilal generically indicates the king’s belonging to the Qatabanian royal genealogy, and the father of Hawfʿam Yuhanʿim was really such a ʾnmrm. There is also another possibility. The integration of the royal title, recovered also from dasi, is wrong and we should integrate ‘and his son ʾnmrm’. In effect the name of the king’s son (which is not a typically regal name) is, at times, cited after the name of the king. However in atm 866 the king is the author of the decree, and the idea of citing his name without the patronymic is, in my opinion, very improbable. Therefore, there remains the problem of who ʾnmrm was. The tribe of Maryamat is certainly very active as far as trade is concerned. Inscription al-Ādī 5 bears testimony to its presence in Qaryat al-Fāw; the tribe seems implicated in trade by land towards the Gulf. In inscription al-Ādī 21 we find the tribe at Shukaʿ, probably interested in a nascent trade by sea in the Red Sea. A recently published text provides an unexpected and astonishing picture of the trade management ability of the tribe of Maryamat. In the inscription Maraqten-Qatabanic 1, no Qatabanian kings are mentioned, but the palaeography of the text shows stylistic traits ascribable, with enough certainty, to the period I called c. The inhabitants of the city of Maryamat give thanks to the gods for having been able to return safe and sound from a voyage to faraway lands. Maraqten-Qatabanic 1, 7–14: ywm ʿlw w-s¹wfy Ḥwkm w-ʾlhw bytn S²bʿn Ṯwbʾl w-ʿmḏkr bn ʾrḍtw w-ʾhgr s²ʾmt w-Nbṭm w-Ks²d w-Mṣr w-Ywnm w-ywm ʿlw w-s¹wfy Ḥwkm w-ʾlhw bytn S²bʿn Ṯwbʾl bn hgrn Rqmm s¹bʾtm s¹bʾ b-qdmw ḏtn s¹bʾtn b-ybs¹n w-bḥrn “when Ḥawkam and the gods of the temple S²bʿn raised and save Ṯwbʾl e ʿmḏkr from the lands and the towns of the north, Nabatia, Chaldea, Egypt and Ionia. And when Ḥawkam and the gods of the temple S²bʿn raised and save Ṯwbʾl from the town of Petra (Rqmm), during the expedition he did before this expedition, by land and by sea”.

106

avanzini

The whole world in contact with the Qatabanians is mentioned here! Qatabanian caravans also reached Mesopotamia, thus justifying the strategic presence in Qaryat al-Fāw of the tribe of Maryamat. They arrived in Egypt by land and by sea. The Wadi Ḥarīb appear to be for Qatabān what the cities of the Jawf were for Sabaʾ: the main agents of long distance trade managed by the state. Let us now return to the particularities of the epigraphic documentation from Maryamat. Regarding the lexicon, first of all, let us recall various hapax present only in the documentation of Maryamat within asa. The meaning of the verb tn’f ‘to save’ is easily deduced from the context, but the verb is a hapax in asa. In the inscription al-ʿĀdī 28, 8–9: b-ḏtm tnʾf-s¹ b-ḍrm w-s¹lmm “since He saved him in war and peace”; and in lines 11–12: b-ḏtm b-yntʾf-s¹my “because might He save the two of them”. Morphologically speaking, the two verbs follow the rule brought to light by A. Multhoff:15 the prefix t is attested in the suffix verbal form and the infix t in the prefix verbal form (tfʿl vs. yftʿl). And, in this case, the repetition of this morphological phenomenon in all asa languages and over diverse periods allows us to contemplate the presence of a true grammatical rule in all asa languages, as held by A. Multhoff. Incidentally also the stereotypical expression b-ḍrm w-s¹lmm is attested in qat only in this inscription, in Cox 4 and csai i, 94 = Atlal. This also appears to be a formula belonging to the documentation of Maryamat. The presence of “in war and peace” could represent a possible confirmation of Bron’s hypothesis mentioned above, of a provenance of csai i, 94 = Atlal from Maryamat. A substantive which has not been attested elsewhere is ḥlkm (3 times in alʿĀdī 8). My translation suggestion of “as properties” based on the context is totally hypothetical as is also the entire translation of the text proposed in dasi.

15

Multhoff 2010. On the t-stems in Semitic, see Weninger 2011: 157.

the formularies and their historical implications

107

The editor of the text, thinking that the adverbial form of the substantive finds a parallel in the well-known glm (often attested in the qat construction texts) translates ḥlkm as “in their entirety”, which could well be right. What is lacking are etymological comparisons pertinent in the other Semitic languages. The fact of the root in asa languages being essentially used in female onomastics does not help. The context of the text is interesting and merits a brief summary. The authors, members of the principle family of Maryamat, construct buildings, look after fields in a valley with god’s help and the active participation of their tribes. And up to this point the text returns to a textual typology which is familiar and widely attested. The authors are mʿhd qny Ḥwkm “guardians of the property of Ḥawkam”.16 They are put in charge of the temple treasure. If the translation of bn rbḥ ḍʾnm s²ʾmw l-Ḥwkm “with the proceeds of (the sale of) the sheep they purchased for Ḥawkam” was right we could deduce that the temple was in possession of herds of sheep and that the profit from their sale went (obviously) to the temple. Also, rbḥ is a hapax in qat. It is attested in the sab inscription Gl 1572: kl mlʾ w-rbḥ “all the income and the profit”. The context is rather interesting here. It regards, also in this case, the temple and the taxes to pay a god. The priests use their money to build constructions; everything is laid down with the familiar mania of asa classification in all its parts. However, these building works must remain the properties of god, who directly, through His oracle, reiterates that the constructions were completed in His name. The text reaffirms who the owners of the buildings are the priests and the god. The construction work is made possible thanks to the sale of livestock, which evidently belonged to the temple. Fields and buildings belong to the god. This inscription appears to construct a background of a few regal expiatory asa inscriptions where the king must repent for having dared to take possession of the cleric’s goods on behalf of the state and its citizens. Other lexemes present in the corpus of Maryamat are attested in other asa languages, but not in qat. After having affirmed just a short time ago,17 that the pronoun in -h in the qat construction inscriptions is found after different substantives, but only after the verb rṯd, inscription al-ʿĀdī 1 provides an example contrary to my affirmation. Here we, in fact, find a suffix pronoun -hw after the verb fdy.18 16 17 18

Officials covering the same role are attested in Sabaʾ (Sh 17, 3–4: mʿhd ʾlʾltn), and at Timnaʿ (csai i, 37 = mq-hk 7, 2 ʾmnt mbʿl ʿm ḏ-Rymtm). Avanzini 2015: 29. See Prioletta, Arbach 2015: 253 and notes 24, 25 for the meaning of the verb.

108

avanzini

al-ʿĀdī 1, 2: w-fdy-hw ḏt Ẓhrn b-kl ḏtm tkrb-s¹ “and dhat Ẓahrān granted all what he had asked Her” But even more than for rṯd-hw is it evident that the use of hw instead of the regular qat pronoun s¹ is bound to the employment of a verb not attested in qat. The entire expression is a loanword from sab. Despite what the editors maintain,19 in qat the verb, whose subject is a female divinity, appears to me to be normally marked as feminine; while substantives referring to goddesses (mrʾ “Mistress”, bʿl “Lady”) could be masculine. In the case of fdy-hw the verb in a masculine form underlines, I believe, the fixity of a non-productive expression in qat; it is not an example of a masculine verbal form whose subject is a goddess. Also interesting is the sentence that precedes w-fdy-hw: ywm s¹ṭbn ʾlrʾb ḏt Ẓhrn b⟨n⟩20-ʿs³b-s¹ w-fdy-hw “when ʾlrʾbt dedicated to dhat Ẓahrān from his property and She granted …”. A parallel text is to be found in Cox 4, 3–5: ywm s¹(ṭb)n ʿmḏkr Ḥwkm bn ʿs³b-s¹ b-ḏtm s¹wfy Ḥwkm ʿmḏkr b-kl ḏtm tkrbs¹ “when ʿmḏkr dedicated to Ḥawkam from his property, because Ḥawkam granted ʿmḏkr all he had asked Him”. Interesting as the verbs are which mark the textual typology (dedicated, built, decreed) so is the use of the verb s¹ṭbn,21 attested in qat only in these two 19 20 21

Prioletta, Arbach 2015: 253 accept the hypothesis of Multhoff 2010: 32–34. The bronze plaque al-ʿĀdī 1 is broken between b and ʿs³b-s¹. The same text in Cox 4 suggested the integration of n. The editors read: b-ʿs³b-s¹. The verb hṭbn is attested in sab only (three times) in res 3946 “have, give property rights”; s¹ṭbn is attested once in min (mṣm 116, 3: w-mn ys¹s²ʾm w-s¹ṭbn “whoever would sell and transfer the property”). Despite the scarcity of attestations, I find it interesting to come

the formularies and their historical implications

109

inscriptions. The semantic value of the verb is very much parallel with the causative form of qny “hand down possessions to the god, dedicate”. The final element for underlining the value of loanword of fdy-hw could be found in the different conjugation before the verb “to grant”. In Cox 4 bḏtm (a typically qat conjunction) precedes s¹wfy; while before fdy there is the conjunction w-. Also for the redactor of the test it should not appear appropriate to place a qat conjunction before the verb fdy which certainly was not qat. Only at Maryamat within the qat corpus is the substantive fs¹ʾ attested. The substantive is further attested in min; and definitely, both in min and in the inscriptions from Maryamat, it indicates a cardinal point. F. Bron translates it: “east”.22 fs¹ʾ is attested in atm 867= al-ʿĀdī 2, 2–3: bn ḫlfn ḏ-S¹ḥrym l-fs¹ʾ ʿd mḥfdn Yhrʾl “from the gate S¹ḥrym towards fs¹ʾ up until the gate Yhrʾl”. In al-ʿĀdī 4, 5–6: bytn ʾḥrm fs¹ʾytm bn ẓb[r] [… … ʾ](ḥr)m ʿd ẓbr bytn “the temple ʾḥrm towards fs¹ʾ from the corner [… …] ʾḥrm, up to the corner of the temple”. What is morphologically interesting is the form of substantive ( fs¹ʾytm: feminine nisbe); there is a parallel, with the same morphological form used for a cardinal point,23 in sab (Ja 629, 17: ʿdy ḫlf hgrn Ḥlzwm wms²rqytn “up to the gate of the city Ḥlzwm and towards the east”). The most interesting attestation of the substantive is in atm 866 in which it appears evident that in the corpus from Marayamat fs¹ʾ indicates the south.

22 23

across these lexical elements with a connoted semantic meaning and which are present in Early sab and in non-sab languages; they could be hypothetical evidences of a common asa lexicon. Bron 1991: 36. Stein 2003: 63–64.

110

avanzini

atm 866, 6–10: ḏtw s³n Ṣnqn w-byt Whbʾl S¹flyn ms²rqn w-s³n (Ṣnq)n w-byt ḏ-Ḥmym fs¹ʾn w-s³n Ṣnqn w-byt [..]wn bn Bʿn mmqṭn w-s³n ʾbyt bn Ḫs²n w-(b)nw ʾs²s² lʿln “those (the houses given by the king to the recipients of his decree) near Ṣnqn and the house of Whbʾl S¹flyn the eastern one; and near Ṣnqn and the house of ḏ-Ḥmym, the southern one; and near Ṣnqn and the house of [..]wn of Bʿn, the western one; and near the houses of he of Ḫs²n and of the family ʾs²s², the northern one”. The whole text is interesting. Two houses will be the property, through royal volition, of the recipients of the decree. The decree specifies their position in the city by recalling the houses that surround them from all the cardinal points. Also in the inscription Maʿīn 1 ownership of the houses is specified (as in atm 866) by the localisation of the houses. Incidentally, Ṣnqn appears to be a toponym indicating a quarter of Maryamat; but it could also be a common noun “city street”. The meaning of “street” for ṣnq can be found in the sab inscription Ir 13, 22 ʿdy ṣnwq S²bwt “in the streets of Shabwa”. The utility of the on-line archive is evident in this case. If we check the various attestations of ṣnq we see that the concrete meaning of “city street” adapts to other contexts, for which abstract meanings were suggested for the substantive.24 In Maʿīn 13, 3 ʿd ṣnqn (not translated by F. Bron) appears to run perfectly parallel with s³n ṣnqn and could be translated “up to the street”. If we look at the attestations in min of fs¹ʾ (for example in Maʿīn 1, 4–5: wʾs³nnh-s¹ bn ( f )s¹ʾ … w-bn s³nn lʿl … w-bn s³nn ḍrʿ … w-bn s³nn Yṯl) also in min fs¹ʾ could not indicate east, but rather south. Of the four cardinal points attested in Maʿīn 1 only north (lʿl) is certain; and maybe west (rḍʿ) for its etymological comparison with the Arabic raḍʿ sometimes used for the “setting of the sun”. However, Yṯl (evidently the ancient name of the city of Barāqish) and fs¹ʾ could be east and south respectively. 24

In the fragmentary sab inscription res 4337, dealing with the access to a city, the translation given by res for ḏ-ybʾh ṣnq “qui déterminerait le besoin”, seems imaginative. Perhaps, it means “who is entering the street”. Also in cih 81, 8 ṣnq could have a concrete meaning. The god punished the dedicator with the destruction of his houses and w-yhḏlln qny-hmw w-ms²ḥkn ʿdy ṣnqn translated in dasi “so that He ravaged their property, and (He gave) grief in distress”. Despite the difficulty of translating ms²ḥkn (hapax in asa), it seems better to translate “so that He ravaged their property and the buildings (?) in the street”.

the formularies and their historical implications

111

The presence of the preposition s³n in atm 866 merits emphasis. This preposition is absent in the qat inscriptions, but well attested in min and sab. Thus, it is not only the probable loaning of a substantive, but the loaning of a complete phrase. I have already touched upon a textual typology25 present in the Wadi Ḥarīb: these texts are dedicated by members of a tribe that live abroad and not by an individual (e. g. csai i, 59 = res 4329 and csai i, 61 = Ry 497 ([s²ʿ](bn ḏ)(H)rbt ḥwr hgrn S³wʾ and s²ʿbn ʾhrbn ḥwr hgrn Ẓfr respectively). We have already said that some members of the tribe of Maryamat are in Shukaʿ in the south of Jabal al-ʿAwd, an area linked to Sawā and to the Red Sea (al-ʿĀdī 21). The inscription follows the same typology the author, in this case too, is not an individual but the ʾrymn ḥwr S²kʿm. In these three inscriptions and in csai ii, 6 = Ja 2898,26 there is the formulaic expression mṯbm ṯwbw “as an offering they offered” attested only in these inscriptions. To the four inscriptions we can add a recently edited one: Collection privée yéménite 2014–3. We have already quoted this inscription previously. The text contains proof of an established presence of inhabitants from Maryamat at Shabwa. Also in this case, even if the author of the text is an individual person and not a part of a tribe, we are dealing here, therefore, with inhabitants from Maryamat resident abroad. Also in this inscription we find again mṯbm and ṯwb, but the formulary is loose and clarified. The dedication is carried out bn frʿm w-mṯbm frʿ w-ṯwb “from the tribute and the offering he paid and offered”. In any case, and this is the interesting aspect, mṯbm ṯwb is not attested in the rest of the asa corpus. It has to be part of a regional formulary used by the writing school of the Wadi Ḥarīb. Of the inscription Maraqten-Qatabanic 1 we have already cited a passage which ends with b-ybs¹n w-bḥrn “by land and by sea”. I particularly liked this expression because it gives, in my opinion, a positive example of creating formularies, and fixing sentences on the one hand, and on the other hand the dissolving of fixed sentences into elements making up the narrative. Both in min and in sab the expression is known but inversely so: “by sea and by land”. (e. g. m 197, 7: w-kl ʾlʾlt ḏ-bḥrm w-ybs¹m w-ms²rqm w-mʿrbm; Ja 576+577: bn s²ʾmt w-ymnt w-bḥrm w-ybs¹m). This is evidently a formulaic expression;

25 26

Avanzini 2014 a: 63. About this inscription and the reasons of its classification as Marginal Qatabanic, see Avanzini 2014 a: 63.

112

avanzini

the substantive ybs¹ “arid land” is attested both in min and in sab only after bḥr “sea”. The phrases indicating “the north and the south, the west and the east” mean “any place, anywhere”. The author of the qat inscription uses two substantives never before attested in qat and he takes them out of their fixed construction. He does not intend “anywhere” when he uses them in the inscription, but intends their literal sense when relating the tale of his very long journey, first by land and then by sea. Yet another example of a formulary from Maryamat itself appears to be the use of the substantive mlʾt, in absolute state, followed by l + the prefix form of the verb. fb Hawkam 4, 3: w-Ḥwkm mlʾt l-ys¹ʾmn “forever might Ḥawkam protect”; fb Hawkam 5, 2–3: mlʾt l-ys¹ʾmnwn “forever might (the gods) protect”; Maraqten-Qatabanic 1, 14: w-mlʾt l-yʿlw27 “forever might (the gods) raise”.28 Finally, due to its strong valence as a formulary, the use of the verb wzʾ as an auxiliary “do something again” attested in qat only in the inscriptions from Maryamat (in the verbal form yzʾ) is particularly interesting. Its use is widely attested in sab.

27 28

The editor reads: mls¹t surely to be corrected into mlʾt. An example of the same construct in Marginal Qatabanic (csai ii, 14 = res 4336, 4). For a previous wrong interpretation of mlʾt as a verb, see Avanzini 2016: 200.

the formularies and their historical implications

113

For example: Collection privée yémenite 2014–3, 2: w-l-ḏtm b-y(z)ʾ ktrb-s¹ “and for what he should continue to ask Him”; fb Hawkam 2, 4: w- l yzʾ Ḥwkm s¹ʿnm29 Qrs²m bkl ḏtm b-yktrb-s¹ “and might Ḥawkam continue to help Qrs²m in all he will ask Him”; fb Hawkam 6, 6–7: w-l-yzʾ Ḥwkm mtʿ-s¹my bn kl bʾs¹m w-nkytm “and might Ḥawkam continue to save them from all evil and harm”; fb-Hawkam 7, 3: w-l-yzʾ Ḥwkm s¹ʿnm Tbʿʾl b-kl ḏtm b-yktrb-s¹ “and might Ḥawkam continue to help Tbʿʾl in all he will ask Him”; and yet again, in csai i, 94= Atlal, 4: l-yzʾ Ḥwkm s¹ʿn[… “and might Ḥawkam continue to help”. The verb with the meaning of “increase” exists in the rest of the qat documentation, but it is never used as an auxiliary.30 Without doubt at Maryamat epigraphic texts are redacted using stereotypical expressions not yet attested in the rest of the qat. Particular features in the Maryamat formulary are not always due to loans from other writing schools. However, what we can be sure of is that the editing of epigraphic texts at Maryamat have autonomous features which, in a few cases are affected by contacts with formularies present in the documentation of other asa languages. The inhabitants of Maryamat were open to cultural contacts beyond the kingdom of Qatabān. As always we can see the result of such a process, but if my hypothesis of the existence of a partially autonomous writing school at Maryamat were true, we could deduce the historical importance of the city and its long history prior to the hitherto known documentation. In conclusion, in my opinion, having at one’s disposition texts redacted within the norms of a writing school there is, on the one hand, a limit to understanding the grammatical norms of the languages actually spoken by the ancient South Arabians.31 On the other hand, such texts are an extremely useful tool for investigating the history of the region and its peoples.

29

30

31

Interesting to note is m suffixed to the infinitive. Before the publication of the epigraphic corpus from Maryamat the suffix -m after the infinitive of verbs at derived stem (also in this example the verb is at causative: s¹ʿnm < ʿwn) was attested only in legal texts. yzʾ as an auxiliary is also found in two qat inscriptions from the second century ad (cias 47.82/o 2, 10–11 and cias 95.11/o 2, 9). But this is a completely different matter. The two late inscriptions show some traits of linguistic interference with the sab. Avanzini 2014 b: 1–2.

114

avanzini

Bibliographical References For the bibliography related to the epigraphs mentioned in this contribution, see Corpus of South Arabian Inscriptions (http://dasi.humnet.unipi.it/). Avanzini, A. 2014 a. “Appendix to M. Arbach «Queques remarques sur la chronologie de l’Arabie du Sud aux iie–ier siècle avant l’ère chretienne».” In Arabian and Islamic studies. A collection of papers in honour of Mikhail Borishovich Piotrovskij on the occasion of his 70th birthday, edited by A.V. Sedov, 58–65. Moskow. Avanzini, A. 2014 b. “From inscriptions to grammar: notes on the grammar of nonSabaic languages.” In Languages of Southern Arabia. Papers from the special session of the seminar for Arabian studies held on 27 July 2013. (Supplement to the Proceedings of the Seminar for Arabian Studies, 44), edited by O. Elmaz and J.C.E. Watson, 1–8. Oxford: Archaeopress. Avanzini, A. 2015. Ancient South Arabian within Semitic and Sabaic within ancient South Arabian. (Quaderni di Arabia Antica, 2). Rome: «L’Erma» di Bretschneider. Avanzini, A. 2016. By land and by sea. A history of South Arabia before Islam recounted from inscriptions. (Arabia Antica, 10). Rome: «L’Erma» di Bretschneider. Bron, Fr. 1991. “Deux inscriptions de la porte ouest de Maʿīn.”Proceedings of the Seminar for Arabian Studies 21: 35–40. Bron, Fr. 2009. “Trois nouvelles dédicaces qatabanites à Ḥawkam.” Orientalia 78/2: 121– 126. Drewes, A.J., Higham, T.F.G., Macdonald, M.C.A. and Bronk Ramsey, C. 2013. “Some absolute dates for the development of the ancient South Arabian minuscule script.” Arabian Archaeology and Epigraphy 24/2: 196–207. Macdonald, M.C.A. 2015. “On the uses of writing in ancient Arabia and the role of palaeography in studying them.” Arabian Epigraphic Notes 1: 1–50. Mouton, M. and Schiettecatte, J. 2014. In the desert margins. The settlement process in the ancient South and East Arabia. (Arabia Antica, 9). Rome: «L’Erma» di Bretschneider. Multhoff, A. 2010. “tfʿl/ftʿl—Die verbalen T-Stämme im Altsüdarabischen.” Folia Orientalia 47: 20–69. Prioletta, A. and Arbach, M. 2015. “Inscriptions qatabānites inédites d’une collection privée yéménite.” Semitica 57: 243–271. Robin, Ch.J. and de Maigret, A. 2009. “Le royaume sudarabique de Maīʿn: nouvelles données grâce aux fouilles italiennes de Barāqish (l’antique Yathill).” Comptes rendus de l’Académie des Inscriptions et Belles-Lettres: 57–96. Sass, B. 2005. The alphabet at the turn of the millennium. The west Semitic alphabet ca. 1150–850bce. The antiquity of the Arabian, Greek and Phrygian alphabets. Tel Aviv: Emery and Claire Yass publications in archaeology.

the formularies and their historical implications

115

Stein, P. 2003. Untersuchungen zur Phonologie und Morphologie des Sabäischen. (Epigraphische Forschungen auf der Arabischen Halbinsel, 3). Rahden/Westf.: Marie Leidorf GmbH. Weninger, S. 2011. “Reconstructive Morphology.” In pages The Semitic languages: an International Handbook, edited by S. Weninger, G. Khan, M.P. Streck and J.C.E. Watson, 151–178. (Handbücher zur Sprach- und Kommunikationswissenschaft, 36). Berlin: De Gruyter Mouton.

chapter 7

Ancient South Arabian Graffiti from Shabathān (Governorate of al-Bayḍāʾ, Yemen) Alessia Prioletta

1

Introduction

The graffiti published in this study were discovered in December 2003 by the French archaeological mission “Mission Qatabān” (mq), directed by Christian Robin.1 The team returned to the site during the 2004 first archaeological campaign at Ḥaṣī.2 Shabathān is located about 24 Km north-east of al-Bayḍāʾ, the administrative centre of the same governorate (map). The site is located on a rocky ridge overlooking the surrounding region, about 2030m above sea level. The graffiti are especially found on the south side of the rock wall, over a length of about 150m. On the plateau, the remains of an ancient construction are recognized, resting on a natural platform. The archaeologists assume that this building might be related to the rock sanctuary and the worship of ʿAthtar testified to in the texts but they admit, however, that its layout rather resembles the typical single-celled structures of the Yemeni Bronze Age.3

1 The researchers who participated in the 2003 campaign were: Mounir Arbach, Hédi Dridi, Jean-Baptiste Rigot, Christian Julien Robin, and Jérémie Schiettecatte. I would like to thank the “Mission Qatabān” for granting me permission to publish this material. 2 See the unpublished report in Mounir Arbach et al., “Comptes rendus de trois campagnes de fouilles 2004–2006 par la Mission Qatabān” (https://halshs.archives-ouvertes.fr/halshs -00581418). The results have been published in the review Arabia. See Hédi Dridi, “Rapport préliminaire sur les deux premières campagnes de fouilles sur le site de Ḥaṣī (al-Bayḍāʾ, Yémen),” Arabia 3 (2005–2006): 11–16; Jean-Baptiste Rigot, “Premier aperçu de l’environnement et de l’ aménagement hydro-agricole dans la région de Ḥaṣī (province de al-Bayḍāʾ), Yémen,” Arabia 3 (2005–2006): 17–25. 3 It is a rectangular building of about 10 × 8 m with north-south exposure, built with locally sourced stone slabs, of which a door with it orthostat is recognized. Fragments of pottery were found in the building, whose dating, however, is difficult to specify. This information is contained in the unpublished report of the 2003 campaign, prepared by Jérémie Schiettecatte, whom I thank for having provided me with the text.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

Ancient South Arabia

117

figure 1

ancient south arabian graffiti from shabathān

118

prioletta

The major site of the region during pre-Islamic times was Ḥaṣī, located a few km south of Shabathān. Ḥaṣī was home to the powerful clan of the banū Ḥaṣbaḥ, a lineage whose princes (qayls) were at the head of the tribe of Maḍḥām (Mḍḥym)4 and controlled a large territory, which at the time of their greatest expansion (from about the 2nd–3rd c. ad), comprised two other surroundings tribes: Datīna (Dtnt)5 to the east, and Sufārum (S¹frm) to the west. So far, it was unclear whether, before that time, Shabathān was already under control of Maḍḥām. The data derivable from these graffiti prove that in earlier times this site was perhaps administered by the Datīna tribe. Dozens of graffiti have been found everywhere in this area, many of them discovered by the French Mission (mafray and mq) during its surveys and still unpublished.6 Only a small part of these graffiti has been examined analytically and with good photos.7 I found it appropriate that this contribution should appear in a volume in honour of Michael Macdonald, in the hope that it will stimulate further research on this subject in Ancient South Arabian (asa) epigraphy.

2

The Texts

The graffiti carved at the ridge of Shabathān have not been systematically documented. From the available photos, roughly one hundred texts could be deciphered. Of these, only about 37 texts are presented here, as a result 4 Christian Robin summed up the historical data concerning this clan coming from the epigraphic and Arab-Islamic sources in “Les banū Ḥaṣbaḥ, princes de la commune de Maḍḥām,” Arabia 3 (2005–2006): 31–110. I accept here the vocalization of the tribe’s name suggested by him and based on the transcription transmitted in the Arabic sources. I am aware, however, that other vocalizations could also be possible, because the y in Mḍḥym could mark ī, ē or even ay. 5 In the Arabic sources, the name has become Dathīna. The history and geographical borders of ancient Datīna have been studied in Hermann v. Wissmann, Zur Archäologie und antiken Geographie von Südarabien. Ḥaḍramaut, Qatabān und das ʿAden-Gebiet in der Antike. (Uitgaven van het, Nederlands Historisch-Archaeologisch Instituut te İstanbul, 24) (Istanbul: Nederlands Instituut voor het Nabije Oosten, 1968), 90–93. The Maḍḥām tribe is also treated at pp. 67–89. 6 Among them, especially the graffiti of the Van Lessen collection should be mentioned, published in Albert Jamme, Miscellanées d’ancient arabe ii (Washington, 1971) and Albert Jamme, Miscellanées d’ancient arabe iii (Washington, 1972). The more important texts are listed in Robin, “Les banū Ḥaṣbaḥ”. See also footnotes 41–43 in this study. 7 Mounir Arbach and Iwona Gajda, “La plus ancienne inscription sudarabique datée d’après

ancient south arabian graffiti from shabathān

119

of a selection that took into account their graphic features as well as the linguistic and historical data considered as more relevant for the objectives of this paper.8 Since the texts are short and repetitive, and predominantly contain proper names, no translation is given, and the few lexical items are analysed in part 4. The most relevant historical and cultural data are also treated in parts 5 and 6. The graphic (and linguistic) features of these texts allow us to deduce that they are the product of private individuals rather than of professional scribes, who left their prayers to the deity engraved on the rock of an open-air sanctuary. Following a recent definition by Michael Macdonald, the Shabathān graffiti should belong to the category of the “graffiti of a literate society”, as they are not produced by the nomads and semi-nomads of the desert, but by members of a settled literate society such as that of ancient South Arabia, in which “reading and writing have become essential to its functioning”.9 It seemed useful to divide the texts into two groups on the basis of the type of script. In the first one, the authors visibly attempted to reproduce the formal— or monumental—asa script while in the second, the style is much closer to the informal—minuscule or cursive—variety of this script.10 In practice, this division remains somewhat arbitrary and overly rigid, as both groups show how in the carving of a graffito by a non-professional writer other dynamics occur other than the strict stylistic formalization of a scribal school. Group i Rock Face 1—Right (Fig. 2) This big block of smooth basaltic rock bears 13 main texts and several faint scratches, incised with different techniques and likely during different times.11

8 9 10

11

une ère et autres inscriptions rupestres de la région d’ al-Bayḍāʾ,” Syria 79 (2002): 293–306; Christian J. Robin, “Les inscriptions de Ḥaṣī,” Raydān 7 (2001): 179–223. The rest of this collection, containing minor texts or texts of difficult reading, will be the subject of a subsequent publication. Michael Macdonald, “On the uses of writing in ancient Arabia and the role of palaeography in studying them,” Arabian Epigraphic Notes 1 (2015): 9–10. Recently, Macdonald has suggested that the definition of “monumental” should be avoided, since it blurs the distinction between purpose and register of a script. He equally recommended that “cursive” be used in the most restricted sense, namely when the letters are joined together. See Macdonald, “On the uses,” 4–5. Numeration begins from the upper right angle. The transcription criteria adopted in the edition of the texts are the following: ( ) indicates uncertain or partial reading, [ ] a restored lacuna, [.] an unrestored lacuna; word dividers are transcribed with a space, except when

120

prioletta

1 (fig. 2) 1 2

Nbṭʿm bn ʿmkrb | ḏ-Yhyll b-ʾḏn ʿ(s³)tr

This text is finely incised with the double line technique. It is read from right to left, though the n’s are back-to-front. The letter ṭ has the central vertical stroke drawn diagonally. Almost the whole s³ is obscured by an abrasion, and only the final part of the lower right stroke is visible. Both the name and the patronymic of the author are known in asa, though Nbṭʿm is more a Qatabanian name. An extraneous thinly incised Arabic graffito is visible below the text. 2 (fig. 2) 1 2

b-ʾḏn ʿṯtr w-ʿm Rʾbm ḏ-Lqḥm

Inscribed below no. 1, this text is enclosed by a zigzag border and is read from right to left. The letters are also chiselled with a zigzag line; the m’s are back-tofront. A symbol made up of three parallel verticals intersected by a horizontal stroke closes the text. Rʾbm is very common in all the asa languages. 3 (fig. 2) 1 2

Nbṭʿm ḏ-Lqḥm b-ʾḏn-ʿṯtr |

Text chiselled in a zigzag line below no. 2 and read from right to left. The presence of a word divider after the god’s name likely suggests that the inscription begins on the lower line, although one cannot exclude that the invocation formula opens the text, as in nos. 2 and 4. In line 2, the word divider is however missing between ʾḏn and ʿṯtr. The form of the ṭ is again with diagonal central stroke as in no. 1; the n’s are back-to-front; the ḥ is upside down and has only the three prongs extended downwards. The author’s name has also been found in no. 1. they appear at the beginning or at the end of a line, in which case they are indicated with the symbol |; the symbol - is used with clitic particles and pronouns, and also replaces spaces in texts that have no word dividers; the symbol – indicates a word that begins on one line and ends on the following one.

ancient south arabian graffiti from shabathān

121

4 (fig. 2) 1 2

b-ʾḏn ʿ[… …] S¹wṭʿm

Text inscribed below and to the left of no. 1 with a zigzag line. If the interpretation offered is correct, the text would have been incised upside down in relation to the other inscriptions on the rock. It would read from left to right with no word dividers and be closed by a symbol in the shape of the letter ḥ. The m and n would be back-to-front. The ṭ has again a diagonal central stroke and the b and ḏ have no horizontal strokes. The name, composed by the theonym ʿm, is new. 5 (fig. 2) 1 2

| Ynhb ḏ-Lqḥm | b

Text inscribed below and to the left of no. 3; it is read from right to left and the n is back-to-front. It has been left uncompleted in line 2 and is followed by extraneous thinly scratched letters. Ynhb is known from Sabaic. Rock Face 1—Left (Fig. 3) 6 (fig. 3) 1

S²hdn |

A graffito carelessly inscribed with a zigzag line below and to the left of no. 1, read from right to left. The letter d is back-to-front. S²hdn is a new name. 7 (fig. 3) 1

Qs²mm-ḏ-Lqḥm

This text is inscribed to the left of no. 6; it is read from right to left and has no word dividers. The characters are not homogeneously chiselled and the last ones are more faintly drawn. Worth noting is the shape of the ḏ with only one horizontal stroke; the ḥ is upside down. The name Qs²mm occurs in Sabaic though only as a tribe name.

122

prioletta

8 (fig. 3) 1

S²hrn

A graffito carelessly inscribed upside down and to the left of no. 7. 9 (fig. 3) [… …] Hwkn [… …] Text carved with a zigzag line below no. 7. The beginning and the end are not clear on the stone. 10 (fig. 3) 1 2 3

Nbṭʿm ḏ-S¹ʾrn ʿṣwyn b-ʾḏn-ʿs³– tr |

Text inscribed below no. 9 and enclosed by a cartouche. It is chiselled with a zigzag line and is read from right to left. The n’s are back-to-front; the ṭ is open at the bottom; the ḏ has only one horizontal stroke; the w is squared. The nisbe ʿṣwyn is new; however, the similar name ʿṣyt is known as a component of the lineage ʿṯkln, originating from Maʾrib. 11 (fig. 3) 1 2

Bʿw[..]– bn-b-ʾḏn

Text thickly carved below no. 10 and with no word dividers. The reading is not clear on the left side and the interpretation remains altogether uncertain. The w is oversized. There is no trace on the stone of the divine name. 12 (fig. 3) 1

ʾlḏḫr-ḏ-Yhyll

The text is thinly chiselled with a zigzag line scratched below no. 11. It is read from right to left and has no word dividers. The letter ḫ is drawn in informal

ancient south arabian graffiti from shabathān

123

style; the second ḏ has one horizontal stroke. ʾlḏḫr has been frequently found in various asa languages. 13 (fig. 3) 1 2 3 4

Lḥym ḏ-ʾb– ʾns¹ b-ʾḏn ʿztr ḏ-ʾbnṣ– lm

Text inscribed to the left of no. 10 with a zigzag line and enclosed by a cartouche. It is read from right to left. The ḏ’s have one horizontal stroke; the letters ʾ and m are back-to-front. Worth noting is the spelling of the divine name with the second radical z. Lḥym is much attested in the asa languages; ʾbʾns¹ is more usual as a name of a man rather than as a lineage name. 14 (fig. 4) 1 2

ʿmydʿ bn Mfʿ(m)

Text thickly chiselled with a zigzag line and read from right to left. The d is oversized; the b lacks the horizontal stroke. The end of the patronymic is unclear on the stone. ʿmydʿ is a well-known name while the patronymic is new. 15 (fig. 4) 1 2 3

Ġfrm ḏ-Lq– ḥm b-ʾḏn ʿṯ– (tr)

This text is inscribed with the double line technique to the right of no. 14 and is separated from it by a thin border. It is read from right to left and several letters are upside down (ġ, b and ʾ); the ḏ has one horizontal stroke. Ġfrm is known from Qatabanic and Sabaic. Casual scratches are incised on the upper right corner. 16 (fig. 4) 1 2

ʾbm [.]ʾdn b-ʾḏn ʿs³tr

124

prioletta

Graffito inscribed below no. 15 and enclosed by a thin border. The text is faint and carelessly carved compared to the others near it. It is read from left to right but has back-to-front ʾ, m and n; the b has no horizontal stroke; the s³ is in the shape of an asterisk. The initial letter of the second name could be a s² (backto-front), n or m (implying the lack of a word divider and a non-back-to-front stance). The name ʾbm is rare; names such as Nʾdn or Mʾdn are, however, known in the asa languages. 17 (fig. 5) 1

Nmrm ḏ-Yhyll

Text read from left to right. Nmrm is a common asa name. 18 (fig. 5) 1 2

ʿmkrb ḏ-Yhy– yll |

Inscribed below no. 17 in bigger size and with the same style. The text is also read from left to right. ʿmkrb has also been frequently found in the various asa languages. 19 (fig. 6) 1 2 3

Hmʿl ḏ-Lqḥm b-ʾḏn-ʿ(s³)tr

Text inscribed on three lines and with a thin vertical border on the sides. It is read from left to right and has no word dividers. The first letters have been more carefully chiselled compared to those of line 3. The h (l. 1) and ḏ (l. 2) are ligatured; the ḥ is upside-down. 20 (fig. 7) 1 2 3

ʿs³tr ḏ-ʾbnbṣ– lm

ancient south arabian graffiti from shabathān

125

Text chiselled with a zigzag line and read from right to left. The ʿ is in the shape of a triangle; the ʾ and b are written as monograms; the n and m are ligatured. 21 (fig. 8) 1 2

ʿmḏḫr ḏ-ʾḏylm b-ʾḏn | (ʿm) ḏ-Ṣrm w-ʾḫ-s¹ ʿs³bm |

This text, incised in big dimensions, stands out from the others for its graphic features. It seems that the style is somehow archaizing having right angles and big circles. Round r and closed m are privileged. It is read from right to left but has back-to-front m, ʾ and ḫ. The peculiar form of the y and ṯ, whose verticals are not centred with regard to the circles, should also be noted. The names borne by the author and his brother are typically Qatabanian; the family name is new though the root ʾḏl is known in onomastics. At the bottom lower left there are some signs joined to form sorts of monograms. 22 (fig. 9) 1 2 3

Nbṭʿm ḏḪrf b-ʾḏn ʿs³(tr)

Text thickly carved and enclosed by a cartouche. It is read from left to right and has back-to-front m ḫ and n (in l. 2); the b, ṭ and ḏ lack the horizontals. The last two letters of the god’s name are not clear on the stone. 23 (fig. 10) 1 2

Ġfyrm Mʿhr | b-ʾḏn [… …]

Text thinly chiselled with a zigzag line and drawn in a regular formal style; it is read from right to left. The divine name is abraded on the stone. To the left, extraneous scratches are visible on the upper corner while thinly carved signs can be read, from right to left, at the bottom: ʾlḏ. 24 (fig. 11) 1

ʾbs²bm-ḏ-Qynm |

126

prioletta

This text is carefully chiselled with the zigzag technique; it is read from top to bottom and has no word dividers. The letters, although they are drawn with a quite regular formal style, lack a consistent stance: ʾ, b and n are kept frontal while the rest is turned 45° clockwise or counterclockwise (the m is drawn in both stances). 25 (fig. 12) 1 2

R[.]ʾm bn | Brʾ(n)

This text belongs to a group of four texts carved on a flat rock wall, all of them carved with the zigzag technique and having a quite homogeneous style reminiscent of the formal inscriptions of around the 1st c. bc–1st/2nd c. ad. Despite the poor quality of the picture, the reading is generally clear, except for no. 25 at the top and no. 28 at the bottom of the rock. In this inscription, the m is back-to-front. 26 (fig. 12) 1

Ṯbʾl Lqḥm (S²r)ḥnʿt

Written below no. 25. The family name, Lqḥm, is here juxtaposed to the author’s name. Ṯbʾl has been found in Sabaic. The name S²rḥnʿt (“may ʿAthtar protect me” or “ʿAthtar has protected me”), made up of verb and apochoristic theonym, is interesting. 27 (fig. 12) 1

ʾlwhb ḏ-Rḥbtm |

Below no. 26 and enclosed by a cartouche. It is read from right to left though the m is back-to-front. The h is likely upside down, since ʾlwhb is a very well attested name while ʾlws¹b appears to be new. 28 (fig. 12) 1 2 3

Ḍbʿn-ḏ-Ṣdqm | w-ḏ Lqḥm ʿmr [.]ʾ[.] ḏ-[.]lmtm

ancient south arabian graffiti from shabathān

127

Word dividers are used irregularly; the m’s are back-to-front. The end of line 2 is not clear on the stone, but a possible reading could be: Ḏʾb. In line 3, the first letter of the lineage name is partially abraded and only the two horizontals of what might be a ʾ, k etc. are visible. All of the names have been found in asa onomastics. 29 (fig. 13) 1

S²rḥbʾl

Text bearing only one name carefully incised in a double-line rectangular cartouche and read from right to left. The style reproduces perfectly that of the formal inscriptions dating back to about the 4th–6th c. ad: letters with apexes at their ends, the r with tilted ends and the b with triangular top. S²rḥbʾl has been typically found in Sabaic. Group ii 1 (fig. 14) 1 2 3

Ṣbḥkrb ḏ-S²bs³n b-ʾḏn-ʿs³tr ḏ-ʾ– bn Ṣlm w-ʿ(m)

This text is enclosed in a zigzag border and is read from right to left. It makes use of word dividers in the shape of short vertical strokes. The letters are also chiselled with a zigzag line and several of them are back-to-front (k, n and m). In addition to the informal style of the letters, the curious shape of the letter ḏ must be noted, having the horizontals inclined to form a rightward triangle. The author’s name is known from Qatabanic. S²bs³n is the clan’s name after which the toponym Shabathān was named. 2 (fig. 15) 1 2 3 4 5

ʾbrtʿ ḏ-Hwfʿm b-ʾḏn-ʿtr ḏ-ʾbnṣl– m

128

prioletta

This text, which seems to have been carved with the same technique as no. 1, is read from left to right and has no word dividers. A vertical line on the left side separates it from what may be a drawing or signs. Some letters are upside-down (b and l), and the n’s are back-to-front. The letter h in line 2 lacks its vertical stroke. In this line, the half-rhombus shape of the f is particularly notable. The shape of b and ḏ vary in line 3, being the former with no horizontal stroke and the latter with only one horizontal stroke. In line 4, the shape of ṣ, slightly tilted leftward, is not clear because of a stray scratch on the upper right corner. An extraneous scratch seems also to follow the l at the end of the line. The author’s name is known from Qatabanic and Sabaic while his family name is new. 3a (fig. 16) 1

| (R)ṣf[..] mʿhd ʿs³tr ḏ-ʾb(ṣl)m |

3b (fig. 16) 1

| S²rḥʾl ʿbd ʿmḏm w-ʾw(s¹)b[.]m |

At least four different main texts are invisible on the photograph, incised on this big rock face. However, the poor quality of the photograph as well as the irregular undulated surface of the rock makes the reading difficult and their interpretation doubtful. Texts 3a and 3b, thinly incised on the upper right edge, might have been the product of one hand. They are read from right to left and have word dividers in the shape of vertical strokes, which also begin and end each of the texts. The shape of the informal m, open at the bottom, is particularly remarkable. The ḏ has one horizontal stroke. Lightly scratched signs difficult to read are seen to the right of these texts. 4 (fig. 16) 1 2 3 4 5

ʿmʾns¹ bn Frʿkrb Mrtʿtm mʿhd ʿtr | [.]w[.] (ḏ)w ʾbṣlm w-(ʿ)– m (ḏ)-Ṣrm w-ʿ(m) ḏ-(M)ʿrtm |

This text is read from right to left and makes use of word dividers in the shape of vertical strokes. The n’s are back-to-front and the s¹ lacks its vertical stroke.

ancient south arabian graffiti from shabathān

129

The f has the form of a half circle. A chip on the stone obscures the reading at the beginning of line 3, and the interpretation remains thus unclear. Above this text, there is a two-line lightly scratched graffito, of which only few letters can be recognised in line 1: [….]qn-bn-Lqḥ[…]. 5 (fig. 16) 1 2

Ḥyʿm ḏ-S²y[..]

This short text is incised below no. 4. It reads from left to right but has the m back-to-front and the s² oriented 45° counterclockwise. The ḥ has only its three prongs. The author’s name is known; the end of the family name is unclear. 6 (fig. 17) 1 2

Nmrm ḏ-Yhyll b-ʾḏn ʿs³tr ḏ-ʾbṣlm-(w)-ʿm

This text is read from right to left, begins on the lower line and has no word dividers. The m’s have an informal style compared to other signs that are more formal (e.g. ʾ); the ḏ’s have one horizontal stroke. Nmrm has been commonly found in all the asa languages. Some extraneous signs are thinly scratched at the end of line 1, and below this line there is a sign that can possibly be interpreted as a false start. 7 (fig. 18) 1 2 3

Ṣdn Ḏbḥ– m w-S²wfth w-Ḫ( f )lm Hdr

This text is incised on the flat face of a long rock block together with no. 8. It is read downwards having the letters oriented 45° counterclockwise. The signs are carved with the zigzag technique and word dividers, in the shape of short shafts, are used. Compared with the stance of the other signs, the ṣ and ḥ are upside-down and the n is back-to-front. It is worth noting again the combination between formal (e.g. ḏ) and informal shapes (m, ḫ) of the letters. Ṣdn, Ḏbḥm and Ḫflm are all known from asa though they are not very usual; the other names are new.

130

prioletta

8 (fig. 18) 1 2

ʿmḏkr ḏ-S²rgn mʿhd ʿtr

This text is incised below no. 7 with the same orientation and technique. It also has word dividers in the shape of short strokes. The k is back-to-front and the d is turned 45° counterclockwise compared to the other signs. The m’s are drawn in informal style and the h lacks its vertical stroke. ʿmḏkr ḏ-S²rgn has been found in a Qatabanic inscription (see chapter 5).

3

Graphic Features

The graffiti from Shabathān have many interesting graphic and epigraphic peculiarities. These texts are engraved on boulders and rocks of basalt origin having an irregular and undulated surface. Just as in the case of the thousands of Safaitic graffiti, this is probably the reason why there is such a great extent of freedom in carving. The many carving techniques (incision, scratching, chiselling with zigzag line or double line etc.) possibly suggest that different implements were used. There is no fixed direction of writing and the texts can run right-to-left, leftto-right, downwards, upwards. The letter orientation is also absolutely free: virtually—even though some are more so than others—all the glyphs can be written upside-down or back-to-front. What is most astonishing is also the letter shape: not only is a combination between formal and informal shapes found in group ii, but a great inconsistency in the shape of the glyphs is also found in group i, where many letters lack their horizontals or verticals. The use of word dividers is not only rare, but also inconsistent within a single text. Word dividers in the shape of short vertical strokes, such as those used in Taymanitic inscriptions, are attested in the second group. There are also some examples of letter joining and ligaturing (i. 19, 20, ii. 7) similar to those found in the Safaitic graffiti.12 If a graffito—even that produced by settled people who must have had a certain degree of writing or reading skills—is indeed the result of “a personal, 12

Examples of such experiments of letter joining and ligaturing in Safaitic have been described in Michael Macdonald, “Cursive Safaitic inscriptions? A preliminary investigation”, in Arabian Studies in Honour of Mahmoud al-Ghul: Symposium at Yarmouk University December 8–11, ed. Moawiyah M. Ibrahim (Harrassowitz: Wiesbaden, 1984), 62–81.

ancient south arabian graffiti from shabathān

131

individual act” in which “any divergences from the formal letter shapes will tend to be individual rather than generic, i.e. they will differ from text to text”,13 I would not venture to attempt a palaeographic approach, least of all a diachronic arrangement of each single text. I will instead limit myself to some basic remarks.14 Within the texts of the first group, which reproduce a version, albeit much aberrant, of the formal script, the following observations can be made. Nos. 3, 21 and 23 (and possibly 24) appear to be the most archaic (or archaizing), having closed triangle m’s, right-angled n’s and big circles. Some texts (nos. 1, 5, 7, 14 and 15) show thickening of the letter extremities, a feature of which the earliest occurrences are dated to about the beginning of the 3rd c. bc. The texts nos. 25– 28 have graphic features that can be comparable to those of the period between the 1st c. bc and the 2nd c. ad. No. 29 is the most recent one, possibly with a dating between the 4th–6th c. ad. In the second group, the form of some letters is identical or very close to the correspondent glyphs in the earliest stages of the informal alphabet (especially the ʾ, h, l, m, n, s³, ṣ). I would be inclined to compare these forms particularly with the latest stages of phase ii and with phase iiia, which would give a chronological range of more that 600 years up the 1st c. bc.15 Specialists of Ancient North Arabian inscriptions will also inevitably notice some resemblance in the glyphs of this group with the northern alphabets of the south Semitic family, especially those of the Oasis North Arabian (ona), such as Dadanitic and Taymanitic. There doesn’t seem to be any sociolinguistic implication in choosing the formal or informal style, and the only relevant difference is that many of the people who wrote in the informal style were priests.

13 14

15

Macdonald, “On the uses”, 10. For the latest studies on Qatabanic palaeography, see Alessandra Avanzini, Corpus of South Arabian Inscriptions i–iii. Qatabanic, Marginal Qatabanic, Awsanite Inscriptions (Arabia Antica 2) (Pisa: Edizioni Plus-Università di Pisa, 2004), 26–33. For the evolution of the two systems in use in ancient South Arabia, see Peter Stein, “Palaeography of the Ancient South Arabian script. New evidence for an absolute chronology”, Arabian Archaeology and Epigraphy 24 (2013): 186–195. Cf. the stages and their approximate dating in Stein, “Palaeography”, 6–7 and fig. 2.

132 4

prioletta

Language, Formularies and Lexicon

The texts are short and contain almost exclusively only proper names followed by a religious formula that repeats the same. They are, therefore, almost totally devoid of linguistic data. Consequently, the language can only be defined indirectly either by analysing the historical and cultural clues in the texts themselves or considering the political history of Shabathān over time. Thus, it can be inferred that the area of Shabathān and Datīna underwent the same linguistic process observed in the epigraphic sources from the neighbouring tribes who settled in the southern regions (such as Maḍḥām and RadmānKhawlān) between Qatabān and the Indian Ocean; that is: there was a shift from Qatabanic to Sabaic as a consequence of the changed political situation and of the independence gained from Qatabān rule.16 The single text (i. 21) of which the language can be defined on reliable grounds because it mentions the author and his brother (ʾḫ-s¹), is in a nonSabaic language—most probably Qatabanic—, as evidenced by the use of the suffix pronoun (-s¹).17 Despite these limits, however, some individual data emerge the same, especially from a phonological point of view. Following a recurring phenomenon in the informal setting of graffiti, it may be argued that these texts reflect to a greater extent the language spoken by their authors.18 Thus, there are indications that the interdentals are replaced by dentals in the language underlying these texts. In fact, the strong fluctuations in the spelling of the divine name ʿAthtar is a hint of a sound change in the realization of the voiceless fricative interdental ṯ, which was probably pronounced with its dental counterpart s³.19 16

17 18

19

On this subject, see Alessia Prioletta, “The town of Ḥalzaw (Ḥlzwm) between Qatabān, Radmān and Ḥimyar: an essay on political, religious and linguistic history,” Semitica et Classica 6 (2013): 126–128. Of course, this process is observed in the written language. The issue of which language was spoken by the population of this area of the southern highlands of Yemen is still an open question. The text would be, however, in Sabaic if it were assumed that the s¹ is in fact an upsidedown h. This has been admirably described by Macdonald in the case of the nomads of the north Arabia non-literate societies who produced the Safaitic inscriptions: “A lack of formal teaching would suggest that there were also no conventional or historical spellings and that the authors of these texts wrote phonetically, as they spoke”. See Michael Macdonald, “Literacy in an oral environment”, in Writing and Ancient Near Eastern Society: Papers in Honour of Alan R. Millard, ed. P. Bienkowski, C. Mee and E. Slater (New York/London: t&t Clark International, 2005), 87. The voiced series (ḏ/z) remains stable instead.

ancient south arabian graffiti from shabathān

133

The spelling ʿs³tr is in fact the most used one, both in the first and in the second group. Only in the first group 3 examples of the historical spelling ʿṯtr are found. It is difficult to argue whether the spelling with the etymological consonant reflect a more ancient epigraphic phase. In fact, the graphic variations between the inscriptions are not decisive enough to suggest a substantial chronological difference. Besides, if, on the one hand, it could be noticed that all of the four occurrences of ʿṯtr are found in texts written by members of the dhū-Lqḥm clan, on the other hand, a few occurrences of ʿs³tr are also found in texts of the same clan. The spelling with s³ ( ṭ in the spoken language at the oasis of Dadan on their own. However, in combination with the other instances of *ẓ being represented as ṭ, especially the misspelling of ẓll, the spelling of the personal names may lend additional support to the idea that this sound change took place in the language of the inhabitants of the oasis. If *ẓ and *ṭ had indeed merged one would expect to see this reflected in both the spelling of the personal names and the rest of the inscriptions. Therefore their absence could have been a counter argument against the merging of *ẓ and *ṭ in the spoken language at the oasis.

4

Discussion

The evidence from the inscriptions seems to point to the merging of ẓ and ṭ to ṭ at some point during the history of the writing of the Dadanitic inscriptions. It is unclear whether this happened in the spoken language while the inscriptions were already being made, or whether the inscriptions reflect a written register different from the spoken language from the start of their production. A third option would be that there was synchronic linguistic variation at Dadan and that the merging of ẓ and ṭ was only a feature of a part of the population of the oasis whose language did not end up as the main written language of the oasis. If ẓ merged with ṭ this seems to indicate that the reflex of ẓ was voiceless in Dadanitic, similar to its realization in Old Arabic and probably Pre-Hilalian Maghrebian dialects (Al-Jallad 2015b: 90, 92).

17

Even though the name Michael comes from Hebrew originally, the English pronunciation of it can tell us for example that [i] came to be pronounced as [ɑɪ] in modern English or that English lacked a velar fricative.

the phonemes ẓ and ṭ in the dadanitic inscriptions

211

In order to better understand the variation attested in the representation of *ẓ, it is helpful to consider other linguistic variation in the Dadanitic corpus as well. Below a brief discussion of this will follow. 4.1 Other Variation It is interesting to note that variation can be found in other areas of the grammar of the Dadanitic inscriptions as well: there are attestations of both ʾafʿal and hafʿal causative forms (e.g. hẓll in ah021 but ʾẓll in ah 097) and there is variation in number congruence with the dual.18 Most of the causative forms occur in the same formulaic context: in the dedication part of the inscription and in the blessing formula. There are two inscriptions in which both an h- and an ʾ-causative occur (see examples 15and 16).19 ah 120 13. ʿbd/bn//rmh/s1//lḥ/ḏġb//t/ʾẓll/b//ʿd/ml/kn[/l-]//h/b-bdr/l-ḏġbt//f-rḍ-hmy ʿbd son of Rmh priest of Ḏū-ġābat performed the ẓll on behalf of the property that belongs to him at Bdr for Ḏū-ġābat so favor him ah 021 = u 041 14. mgd/bn/ḥrb//ẖṯmh/hẓll//{l-}ḏġbt/b-khl//b{ʿ}d/ẖrf-h/b-b{dr}//{ f-rḍ-}h/ w-ʾẖrt-h Mgd son of Ḥrbẖṯmh performed the ẓll {for} Ḏū-ġābat at Khl on behalf of his autumn crops at Bdr so favor him and his offspring u 079bis 15. w----t/bldʾ//----l/hẓll/h- ẓ//ll//b-bṯ//r/bʿd/n{ḫ}l-h w-//dṯʾ-h/b-ḏʿmn//l-ḏġbt f-rḍ//hm//w-ʾṯb-hm 18 19

For an overview of the variation in number agreement in the inscriptions from al-ʿUḏayb see Sima (1999, 117). There is one object with a verb in each variant, but they seem to belong to two different inscriptions (Al-Saʿīd 1419/1999: 4–24, no. 1, side 1–2 = Sima 2000: 252–254). Even though the inscription containing the h- causative was the first to be written on the object, it is impossible to tell how much time passed between the production of each inscription.

212

kootstra

… he performed the ẓll ceremony at Bṯr on behalf of his palm trees and his crops of the season of the later rains at Ḏʿmn for Ḏū-ġābat so favor them and reward them ah197 16. ḥggw/h-nq/w-hġnyw/b-bt-hm/l- ----//tn/l-ḫrg/w-ʾẓlw/b-h-mṣd/ẓll/h-[nq]/ /l- ḏġbt they performed the feast and dedicated (lit. made increase wealth?) at their temple for … tn for ḫrg and they performed the ẓll of the [nq] for Ḏū-ġābat The co-occurrence of both variants clearly shows that at least at some point during the production of the inscriptions both forms were known and used, making it impossible to make any claims about chronology based on the occurrence of either. Inscriptions with two dedicants can be formulated in two different ways:20 ‘standard dual agreement’ in Dadanitic uses a plural verb with a dual suffixed pronoun; in other inscriptions the category of the dual seems to have been lost and verb and pronouns are both in the plural form.21

20

21

It has also been suggested that Dadanitic has a third person dual suffix –h in the suffix conjugation (ʾẓllh ah 199; ʾẓlh u 019; ʾfyh u 026) (Stiehl 1971, 18; Drewes 1985, 168), but these are certainly marginal forms, occuring only a few times in the entire corpus. There is a third group of inscriptions in which number agreement seems to be confused. ah 120 for example, which mentions a single dedicant, a singular verb, a singular pronoun on the property that is mentioned but a dual pronoun in the blessing formula; u 075 in which two dedicants are mentioned with a plural verb, a singular pronoun on the property that is mentioned and a plural pronoun in the blessing formula. It may be possible to interpret the ‘mistaken’ pronouns in these examples as referring to different people outside the text rather than as actual mistakes. The dedicant in ah120 may have been asking for favor from the deity for himself and someone else even though he made the inscription/dedication alone. Even if these are indeed mistakes, they would speak more to the formulaic nature of the texts and the strong division between the separate parts of the texts than to a lack of knowledge of how to use proper number agreement by the inscribers.

the phonemes ẓ and ṭ in the dadanitic inscriptions

213

ah 058 17. …]bn/bn[…//wmd/bnt/ẓbyh//ʾẓllw/h-ẓll/b-khl//bʿd/ṯbrt-hmy/b-ḏ//ṯʿʿl/frḍ-hmy/w-s1//ʿd-hmy … son of bn … and Md daughter of ẓbyh performed (pl) the ẓll ritual on behalf of their (du) grain at Ḏṯʿʿl so favor them (du) and aid them (du) u 064 18. s²rd/w-ʾḫ-h/ʿbd//s¹mh/bnw/ʿyḏ//ḥrn/ʾẓlw/h-ẓll/l-ḏġbt//ll/bʿd/nḫl-h//m/bḏʿmn/wḏ//kn/l-hm/b-bdr//f-rḍ-hm S²rd and his brother S¹mh sons (pl)22 of ʿyḏḥrn performed (pl) the ẓll ritual for Ḏū-ġābat ll23 on behalf of their (pl) palm trees at Ḏʿmn and what belongs to them (pl) at Bdr so favor them (pl) If we include evidence from personal names one finds that while the shift –at > -ah had not taken place in the language of the inscriptions, the number of Dadanitic personal names written with final –h (most likely representing –ah: 189 unique names with very little repetition of names) is only slightly higher than those with final -t (most likely representing –at: 151 unique names with rare repetition). 19. Final –t personal names e.g. ġlbt, bnklbt, ḏbʾt, fhdt, ġrbt, ḥmmt, nʿmt 20. Final -h personal names e.g. bqlh, brgh, ḏkrh, ḥs1nh, mgyh, nḍrh, s2grh

22 23

Note that there is one inscription (u 011) in which the dual seems to be preserved on the noun (bny ‘two sons of …’) but the verbs and pronouns take the plural form. On the photograph the two l’s and the word are clearly visible at the beginning of line 4 (ociana accessed 4/6/2017). An attempt seems to have been made however to obliterate the l’s. It is unclear whether this was done by the author of the inscription who realized his mistake or by a later passer-by.

214

kootstra

While especially the variation in h- and ʾ- causatives and number agreement need further study before anything can be concluded about them with any certainty, the fact that variation occurs in the phonology, morphology and syntax of the inscriptions, supports the idea that the variation in representation of ẓ represents actual linguistic variation and are not simply scribal errors. The presence of a scribal school at Dadan, producing formal monumental inscriptions, makes it difficult to quantify the variation that we see. Such a scribal tradition would promote consistency and uniformity in the language and content of the inscriptions that they produced, which means they likely did not reflect the spoken language of the oasis directly. This makes it difficult to say how representative the single ṭll slip-up is for the language(s) spoken at the oasis. The presence of such a scribal school sets the Dadanitic corpus apart from the other ana corpora, which had clear writing traditions (Al-Jallad 2015a: 6–7), but do not seem to have had the industry connected to them that would come with the need for skilled labor that was needed to carve the Dadanitic inscriptions in relief.24 The co-occurrence of the different causative forms within two inscriptions shows that, while chronology may have played some part in how the variation in the inscriptions came about, it can definitely not explain all the variation that we see. While the loss of dual agreement may point to diachronic variation, the distribution of the variation will need to be investigated further to be able to conclude this with any certainty, as other causes such as substrate influence or influence from register cannot be excluded at the moment.25

24

25

The level of skill required to carve inscriptions in relief can be compared to that of the Sabaic inscriptions, which most likely meant that these inscriptions were commissioned (Macdonald 2010: 7). A preliminary survey of the distribution of ‘standard’ dual agreement and the loss of dual agreement suggests that register played an important role in the choice of either variant, based on manner of inscribing. A chronological explanation seems less likely since the linguistically more archaic ‘standard’ agreement co-occurs with the more progressive ʾcausatives, but never with the linguistically more archaic h-causative. Only inscriptions with the linguistically more progressive loss of the dual co-occur with the h-causative. This will be addressed in a future publication.

the phonemes ẓ and ṭ in the dadanitic inscriptions

215

Conclusions As discussed above, there are several inscriptions that seem to contain forms that point to the merging of ẓ and ṭ in Dadanitic. The clearest example of this is ah009.1 in which a ṭll form occurs. Due to the formulaic nature of the texts however, it is difficult to say based on this example how widespread this sound change was. Other occurrences of ṭ for *ẓ could support this idea, but are also open to other explanations. The nṭr forms could be cultural loans from Aramaic, but the fact that it occurs in a formulaic context (albeit not in one of the more common formulae) makes it less likely that nẓr and nṭr reflect two different meanings. The personal names might also be drawn from another language which merged *ẓ and *ṭ, therefore they cannot be taken as clear evidence for the sound change in Dadanitic, if no names would have occurred in which ẓ and ṭ merged however, this would have cast serious doubt on the reality of *ẓ > ṭ. The other examples of linguistic variation attested in the Dadanitic corpus seem to further support the idea that there was a difference between the written and spoken languages at Dadan. Even though some of the other linguistic variation attested in the inscriptions may be due to diachronic change, the existence of variation across several linguistic features, and especially the cooccurrence of the ʾ- and h-causatives in two inscriptions, suggest that variant forms were available along-side each other at the oasis, and people were using forms and grammatical features in their inscriptions that were not, or no longer, part of their everyday language.

Sigla ah atm c csns csai dasi HaNSB hch JaL JSLih

Dadanitic inscriptions in Abu ʾl-Ḥasan (1997) Arbach, Bāṭāyiʿ and al-Zubaydī (2013) Ryckmans (1950) Clark (1979) Corpus of South Arabian Inscriptions, part of dasi Digital Archive for the Study of pre-Islamic Arabian Inscriptions, directed by Alessandra Avanzini (University of Pisa) http://dasi.humnet.unipi.it/ Ḥarāḥišah (2010) Harding (1953) Dadanitic (formerly Lihyanite) inscriptions published by Jamme Dadanitic (former Liḥyanite) inscriptions in Jaussen, A. and Savignac, R. (1909)

216 krs lp ociana u res wh

kootstra Safaitic inscriptions recorded by King (1999) Littmann (1943) Online corpus of Inscriptions from Ancient North Arabia http://krcfm .orient.ox.ac.uk/fmi/webd#ociana Inscriptions from al-Uḏayb in Sima, A. (1999) Sabaic inscriptions in Ryckmans (1952) Winnett and Harding (1978)

References Abu ʾl-Ḥasan, H.A.D. 1997. Qirāʾa Li-Kitābāt Liḥyaniyyah Min Jabal ʿikma Bi-Mintaqat Alʿulā. Riyād: maktabat al-malik fahd al-waṭaniyyah. Al-Jallad, A. 2015a. An Outline of the Grammar of the Safaitic Inscriptions. Leiden/Boston: Brill. Al-Jallad, A. 2015b. “On the Voiceless Reflex of *́ṣ and *̱ṭ in Pre-Hilalian Maghrebian Arabic.” Journal of Arabic Linguistics 62: 88–95. Arbach, M., A. Bāṭāyìʿ, and Kh. al-Zubaydī. 2013. “Nuqūš Qatbāniyyah Jadīdah.” Raydān 8: 49–103. Caskel, W. 1954. Liḥyan Und Liḥyanisch. Köln: Westdeutscher Verlag. Clark, V.A. 1979. “A Study of New Safaitic Inscriptions from Jordan.” A thesis presented for the degree of Doctor of Philosophy, Deparment of Middle Eastern Studies, Melbourne: Ann Arbor. Drewes, A.J. 1985. “The Phonemes of Lihyanite.” In Mélanges Linguistiques Offerts À Maxime Rodinson per Ses Élèves, Ses Collègues et Ses Amis, edited by Ch. Robin, 165– 173. Paris: Librairie orientaliste Paul Geuthner, S.A. Farès-Drappeau, S. 2005. Dédan et Liḥyān: Histoire Des Arabes Aux Confins Des Pouvoirs Perse et Hellénestique (ive–iie S. Avant L’ère Chrétienne). Lyon: Maison de l’Orient et de la Méditerranée. Ḥarāḥišah, R. 2010. Nuqūš Ṣafāʾiyyah Mina ʾl-Bādijah Al-ʾUrdunniyyah. Amman: Ward books. Harding, G. Lankester. 1953. “The Cairn of Hani.” Annual of the Department of Antiquities of Jordan 2: 8–56. Harding, G. Lankester. 1971. An Index Concorcdance of Pre-Islamic Arabian Names and Inscriptions. Toronto and Buffalo: University of Toronto Press. Littmann, E. 1943. Safaïtic Inscriptions. Vol. Section d. Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909, Division iv. Leiden: Brill. Macdonald, M.C.A. 1997. “Trade Routes and Trade Goods at the Northern End of the ‘incense Road’ in the First Millennium b.c.” In Profumi d’Arabia: Atti Del Convegno, edited by A. Avanzini, 333–349. Saggi Di Storia Antica 11. Rome.

the phonemes ẓ and ṭ in the dadanitic inscriptions

217

Macdonald, M.C.A. 1999. “Personal Names in the Nabataean Realm; a Review Article.” Journal of Semitic Studies 44 (2): 251–189. Macdonald, M.C.A. 2008. “Ancient North Arabian.” In The Ancient Languages of SyriaPalestine and Arabia, edited by R.D. Woodard, 197–224. Macdonald, M.C.A. 2010. “Ancient Arabia and the Written Word.” Supplement to the Proceedings of the Seminar for Arabian Studies; The Development of Arabic as a Written Language 40: 5–27. Macdonald, M.C.A. 2015. “On the Uses of Writing in Ancient Arabia and the Role of Paleography in Studying Them.” Arabian Epigraphic Notes: An Open Access Online Journal on Arabian Epigraphy 1: 1–49. Noth, M. 2010. Die Israelischen Personennamen Im Rahmen Der Gemeinsemitischen Namengebung. Beiträge Zur Wissenschaft Com Alten Und Neuen Testament 46. Rohmer, J., and G. Charloux. 2015. “From Liḥyān to the Nabataeans: Dating the End of the Iron Age in North-West Arabia.” Proceedings of the Seminar for Arabian Studies 45: 297–320. Ryckmans, G. 1950. Corpus Semiticarum Pars v. Paris: Imprimerie Nationale. Ryckmans, G. 1952. “Epigraphical Texts.” In An Archaeological Journey to Yemen (March– May 1947). Vol. ii. Cairo: Government Press. Scagliarini, F. 2002. “The Origin of the Qanāt System.” aram 14: 567–577. Sima, Alexander. 1999. Die Lihyanischen Inschriften von Al-ʿUḏayb (Saudi Arabien). Epigraphische Forschungen Auf Der Arabischen Halbinsel 1. Rahden/Westfahlen: Verlag Maire Leidorf GmbH. Stein, P. 2010. Die Altsüdarabische Minuskelinschriften Auf Holzstäbchen Aus Der Bayerischen Staatsbibliothek in München. Vol. 1. 2 vols. Epigrapische Forschungen Auf Der Arabischen Halbinsel 5. Tübingen: Ernst Wasmuth Verlag Tübingen. Stiehl, Ruth. 1971. “Neue Liḥyānische Inschriften Aus Al-ʿUḏaib.” In Christentum Am Roten Meer, edited by F. Altheim and R. Stiehl. Vol. 1. Berlin. Winnett, F.V. 1939. “The Place of the Minaeans in the History of Pre-Islamic Arabia.” Bulletin of the American Schools of Oriental Research 73: 3–9. Winnett, F.V., and G. Lankester Harding. 1978. Inscriptions from 50 Safaitic Cairns. Near and Middle Easten Series 9. Toronto: University of Toronto Press. Winnett, F.V., and W.L. Reed. 1970. Ancient Records from North Arabia. Toronto: University of Toronto Press.

chapter 10

Dadanitic Inscriptions from Jabal al-Khraymāt (Madāʾin Ṣāliḥ) María del Carmen Hidalgo-Chacón Díez*

a

Introduction

The inscriptions published here were found at Jabal al-Khraymāt, west of Madāʾin Ṣāliḥ (ancient Ḥegrā), in northwest Saudi Arabia (fig. 1). Madāʾin Ṣāliḥ lies about 22km north of the oasis of al-ʿUlā, ancient Dadan. On the upper part of a rock face, nine Dadanitic inscriptions and one Nabataean have been incised (fig. 2). In addition to these inscriptions there are several drawings of animals. The most striking elements of these drawings are the figures of two large lions facing left.1 Jabal al-Khraymāt no. 5 cannot be read clearly since the tail of the right lion is carved over it, which means that no. 5 was carved before the lion was drawn, while the other inscriptions were probably carved later, especially no. 2 and 6, which carefully avoid the left lion’s tail and front legs respectively. Three drawings of ibexes have been partially covered by one or other of the lions, one by the back legs of the right lion and the others by the body of the left lion, confirming that some of the drawings were also completed before the lions were drawn. The Dadanitic texts found at Madāʾin Ṣāliḥ represent a small proportion of the Dadanitic corpus, the vast majority of which has been found in the oasis of al-ʿUlā. Besides the nine texts which are analysed here, at least sixty-five other Dadanitic texts have been recorded at Madāʾin Ṣāliḥ, most of which are graffiti and consist only of a genealogy. Others, which contain the expression nṭr ddn, “(he) guarded Ddn”, occur in the area of Jabal Ithlib2 which is the only place where this expression appears. The reason may be that this area was used as a * I am very grateful to Michael Macdonald for his helpful comments on an early draft of this article. He is, of course, in no way responsible for its shortcomings. 1 A clearly incorrect analysis of the drawings was done by Nayeem (2000: 98), who describes this petroglyph as being located at the site of “Hassu Aba Mafir at Jabal al-Brar”, south-east of Taymāʾ. This is incorrect and the inscription is located at Jabal al-Khraymāt, as Michael Macdonald (p.c.) explained to me. See also Wood 1975: 51. 2 Abū ʾl-Ḥasan 2002: 255–297.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

dadanitic inscriptions from jabal al-khraymāt

figure 1

219

Map of north-west Saudi Arabia showing the location of Madāʾin Ṣāliḥ and al-ʿUlā l. nehmé & j. schiettecatte

watching-post (Abū ʾl-Ḥasan 2002: 257–258) north of the oasis of al-ʿUlā. Those who carved the texts may have been involved in the work of guarding this oasis from the dangers which could befall it (Hidalgo-Chacón Díez 2015: 142–143). Unfortunately, the inscriptions studied in this contribution do not give any indication as to the date or even the time frame during which they were written.

b

The Inscriptions

No study of these inscriptions has been published before, but photographs of them are shown in Nayeem (2000: 101, fig. 103), in Wood (1975: 51),3 and recently on Wikipedia.4

3 In both books, the inscriptions are considered erroneously as Thamudic. 4 https://en.wikipedia.org/wiki/Mada%27in_Saleh#/media/File:Madain_Saleh_(6720062703) .jpg (Retrieved 06/04/2016).

220

hidalgo-chacón díez

figure 2

Dadanitic inscriptions from Jabal al-Khraymāt at Madāʾin Ṣāliḥ source: https://en.wikipedia.org/wiki/mada%27in_saleh#/media/ file:madain_saleh_(6720062703).jpg (last retrieved september 8, 2016)

Jabal al-Khraymāt No. 1 ( fig. 3)5

figure 3 Jabal al-Khraymāt no. 1 in standardised Dadanitic glyphs

ks¹ṭ w ṭbʿ {ṯ}mny {q}d w grs¹ rḥw f ʾn ṣ{b}w b-{k}l {s¹}th---5 Editorial sigla: { } enclose doubtful letters, [ ] enclose letters which are restored, ---- represents a passage in which one or more letters are completely destroyed, / between two letters indicates an alternative reading. Semitic roots are represented in capital letters, e.g. rwḥ. Note that these inscriptions will bear numbers MS115Dad 1–9 in ociana and in the corpus of inscriptions from Madāʾin Ṣāliḥ, in which 115 is the number of the epigraphic point to which they belong.

dadanitic inscriptions from jabal al-khraymāt

221

(1) Ks¹ṭ (2) and Ṭbʿ {the (3) two appreciated by Qd} and Grs¹ relieved (?) (4) for they desired {in all} {s¹}th----. The interpretation of this text is extremely problematical. Firstly, several letters are not clear; secondly, the text shows linguistic features that have not been found previously in Dadanitic. Considering the lack of comparative material, the suggested interpretation remains tentative. Line 1 Ks¹ṭ. This personal name is very well-known in Safaitic (Harding 1971: 500) but it appears here for the first time in Dadanitic. The vocalization of the name may be either *kāseṭ or *kusayṭ (a diminutive form of *kāseṭ), as can be inferred from the Greek transcription of the name, Χασετος and Χοσετος (see Littmann et al. 1907–1921: 72, no. 94 and 432–433, no. 799; Wuthnow 1930: 147). Line 2 Ṭbʿ. The vertical stroke below the ʿ belongs to the q of the third line. A reading of this name as Ṭby is therefore not possible, given that if the vertical stroke did not belong to the q, the y below would have a form quite unlike that of the y immediately before it. The personal name Ṭbʿ has not been found previously in Ancient North Arabian (ana) but is known in Qatabanian where it is composed of the element ṭyb and an abbreviation of the theophoric element ʿṯtr.6 On the other hand, the name Ṭabʿa is attested in the Arabic tradition (Caskel 1966: 554). Lines 2–3 {Ṯ}mny. The middle stroke of the ṯ rises from the centre of the “x”, rather than descending from the centre, as is normal. This may be because there is a small hole in the stone at the place where the middle stroke of the ṯ should have been incised. {ṯ}mny could be considered to derive from ṯmn, Arabic ṯammana, “to appreciate” (Ibn Manẓūr, s.v. ṯmn) and the adjective ṯamīn, “of high value” (Lane 1968: 355a). It would then be considered as a substantive used as an adjective in the sense of “something appreciated, of high value” referring to the individuals mentioned before it. It is in the dual nominative case expressed by -y (compare the Dadanitic inscriptions JSLih 72 l. 3–4: kbry s²ʿt h-n{ṣ}, “the two kabirs of the company of H-n{ṣ}”), contrary to Arabic where the dual in the nominative case ends with long vowel -ā after the drop of the –n in the construct.

6 Hayajneh 1998: 184.

222

hidalgo-chacón díez

{Q}d. This personal name has not been found before in Dadanitic, though compare the names Qdd and Qddh (see The Online Corpus of the Inscriptions of Ancient North Arabia, ociana, s.v. qd) which are well attested in Safaitic, as well as Qdt in Hismaic (King 1990: 536). In Arabic, the names Qadd, Qudād and Qudayd are known (Caskel 1966: 454, 469). Grs¹. To the best of my knowledge, this name occurs here for the first time in Dadanitic. For its etymology, one may refer to the Classical Arabic nouns jars and jaras which mean, respectively, “a low, faint, gentle, slight or soft sound” and “a bell” (Lane 1968: 409b–c). This name is also attested in Safaitic (see ociana, s.v. grs¹) and in Minaic (al-Said 1995: 80). Rḥw could be a verb in the third person plural of the suffix conjugation derived from rwḥ, which is cognate to Arabic rāḥa, “to relieve”, “to be large, spacious” (Ibn Manẓūr, s.v. rwḥ). In Punic and in Palmyrene Aramaic texts, rwḥ means “to widen”, “to enlarge”, “to set at large”, “to relieve” (Hoftijzer and Jongeling 1995: 1062) and in Sabaic rwḥ means “to extend, to enlarge boundaries” (Beeston et al. 1982: 119). The root is widely attested in Safaitic (see ociana, s.v. rwḥ) with the sense of “to grant relief (from adversity and uncertainty)”, see Al-Jallad (2015: 339), and it occurs almost entirely in prayers. In this text, it probably means “they relieved”. Line 4 f ʾn. The combination f ʾn could be compared to the Classical Arabic phrase fa-ʾinna, which coordinates clauses that justify, with the meaning “for” in an assertive phrase.7 This is also attested in the Dadanitic inscription Al-Ḫuraybah 17 l. 4:8 ----{m}n s¹rq f ʾn yṣbr b-mh s¹r[q]----, “---- {who} stole, he should be arrested for what {he has stolen} ----” (al-Saʿīd 2013–2014: 299–300, no. 9). Note that this text suggests the existence of a conditional system, probably with the deductive aspect of f ʾn in the apodosis. Ṣ{b}w. The reading of the second letter is doubtful but it is probably a b, despite the fact that it differs slightly from the b in b-kl in the same line. If the reading is correct, ṣbw should be interpreted as the third person plural of the suffix conjugation of ṣbw. Ṣabā is attested in Classical Arabic with the meaning “to desire”, “to long for” (Lane 1968: 1649c), as well as in Assyrian

7 For the different functions of fa-ʾinna in Classical Arabic see Larcher 2009: 51–55. 8 We use here the siglum given in the ociana project.

dadanitic inscriptions from jabal al-khraymāt

223

ṣebû (ṣabû), “to wish, desire, need” (cad, 119–120, vol. 16), Syriac ṣebā, “to be willing, to wish, prefer, to seek to” (Payne Smith 1903: 472), and Official Aramaic, Palmyrene, Nabataean and Jewish Aramaic ṣby, “to want, to long for” (Hoftijzer and Jongeling 1995: 957f.). The orthographic rules of Dadanitic show the plene spelling of the verb iii = w or y. Thus, in Dadanitic, the third person plural of the suffix conjugation of bny is bnyw, “they built” (see cll 26 l. 1) and of ʾgw is ʾgww, “they organized” (see ah243 l. 4; Nasif9 1988: 99, pl. clvii line 3 and u 88 l. 2). It might be expected, therefore, that this text would show the third person plural of the suffix conjugation as *ṣbww rather than ṣbw, if it derives from ṣbw (see above). The explanation for that could be that in bnyw the /y/ clearly has a full consonantal value, *banayū, whereas in ʾgw versus ʾgww and in ṣbw there was uncertainty in the pronunciation, some people saying *ʾagawū (ʾgww) and others *ʾagaw (ʾgw) or *ṣabaw (ṣbw). In Classical Arabic, the third person plural of the suffix conjugation of the verb daʿā, derived from dʿw, is daʿaw, “they called”. The similarity of this verb in both languages could be interpreted as an areal feature, as is also the case of the Arabic article ʾl found in some Dadanitic texts, see below Jabal al-Khraymāt no. 5 and 6 (l. 3 and 5). b-kl. Note that kl is written here with a single consonant, contrary to the usual spelling with double consonants, kll. {s¹}th----. The right stroke of the s¹ extends above the horizontal line at which it should have ended. It could be interpreted as a word divider joined to the s¹ at this point. Very tentatively, it could also be suggested that it is the letter ʾ. The rock-face seems to be broken on the left and the inscription may originally have continued. The text is certainly difficult to interpret. Its language shows features that have not been attested previously in any Dadanitic text, such as the verbs, or that have been attested only once, such as f ʾn. Moreover, it contains none of the formulas typically found in the majority of the Dadanitic texts.

9

This reference only contains the photograph. The text is analysed in ociana.

224

hidalgo-chacón díez

Jabal al-Khraymāt No. 2 ( fig. 4)

figure 4

Jabal al-Khraymāt no. 2 in standardised Dadanitic glyphs

This is located above the back of the left lion. Unfortunately, the inscription is crudely carved and the surface is covered with extraneous scratches so that it is impossible to provide a firm reading of some of the letters. At the beginning of the inscription, below the ʿ and joined to the bottom of the ṭ, there is a curved line which has a lighter patina than the letters and which does not belong to the text. ʿṭy ḥrd/{ʿ/w}f----g{}/b{/..}gr{q} ʿṭy {desired} (?) ----. ʿṭy. This is a very common personal name in Safaitic (see ociana, s.v. ʿṭy) and it has been found once before in Dadanitic (Nasif 1988: 85, plate cxiv). It occurs in Arabic as ʿAṭīya (Caskel 1966: 205). The name can be interpreted as coming from Arabic ʿaṭāʾ in the sense of “gift” (Lane 1968: 2085b–c). ḥr{d}. Because what follows this word is damaged, I cannot at present offer a satisfactory reading. But if we consider the last letter as {d}, thus ḥr{d}, it could be interpreted as a verb derived from ḥrd which in Arabic means “to pursue, to desire, to intend”, “to retire from his people” (Lane 1968: 543c–544a). In Hebrew, ḥrd produces words with the meanings “to tremble”: ḥārēd meaning “anxious, frightened at”, ḥărādāh, “trembling, fear” (Köhler and Baumgartner 2001: 350– 351). In Sabaic, the verb ḥrd means “land reserved by a deity, land watered by a deity” (Beeston et al. 1982: 70). As a personal name, Ḥrd is found several times in Safaitic (see ociana, s.v. ḥrd). In conclusion, the reading is not certain but it is possible, and the suggested meaning is the one the root ḥrd has in Arabic. Jabal al-Khraymāt No. 3 ( fig. 5) figure 5 Jabal al-Khraymāt no. 3 in standardised Dadanitic glyphs

This is carved in large letters at the top of the panel.

dadanitic inscriptions from jabal al-khraymāt

225

mlḥ bn brʾh Mlḥ son of Brʾh. The same genealogy has been found in two other Dadanitic texts carved by the same author. One is JSLih 375, located north of the oasis of al-ʿUlā and the other is Jabal Ithlib 2, in Madāʾin Ṣāliḥ (fig. 6–7).10 From the epigraphic point of view, it is interesting to note that the letters of the word bn are joined at the bottom. The occasional joining of letters, repeated in three separate inscriptions by the same author, may point to the fact that the Dadanitic script developed in documents written in ink.11 Mlḥ. This is a very common personal name.12 In Classical Arabic, maluḥa means “to be, to become goodly, beautiful, or pretty”; “beautiful of colour”, “beautiful and bright” (Lane 1968: 2731b). Brʾh. The personal name Brʾh could be connected to Punic bryʾh, “security”, “relief”, Aramaic brʾ, “healthy”, “healed”, Syriac berē, “free” and Arabic bariʾa, “to be healed”, “to be innocent, freed” (Cohen et al. 1970–: 80), as well as to Sabaic bry, “keep someone healthy” and the nouns bry and bryt, “health, soundness, prosperity” (Beeston et al. 1982: 32).

10 11

12

Inscription Jabal Ithlib 2 corresponds to MS036Dad 1, and this epigraphic point contains two other Dadanitic texts (p.c. from L. Nehmé). In his brilliant forthcoming article “Towards a Re-assessment of the Ancient North Arabian Alphabets Used in the Oasis of al-ʿUlà”, Michael Macdonald has analysed the ana scripts used in the oasis of al-ʿUlā. He has suggested that there are many features in the Dadanitic texts, such as their nearly universal right-to-left direction, and the aforementioned joining of letters in some graffiti, which imply that the Dadanitic script “had long been used to write documents in ink on soft or portable materials such as papyrus, parchment or potsherds, or that texts were cut with a sharp point on sticks or palm-leaf stalks, etc., and that this practice continued in parallel with the carving of inscriptions”. He has compared the Dadanitic inscriptions to Safatic, Hismaic and the so-called “Thamudic” graffiti all of which were carved in many directions, a characteristic of scripts which were used only for carving on stone (pp. 7–10). The name could also be read glḥ.

226

hidalgo-chacón díez

figure 6 Dadanitic inscription JSLih 375 jaussen and savignac 1914, atlas, pl. 139

figure 7

Dadanitic inscription Jabal Ithlib 2 photo l. nehmé

Jabal al-Khraymāt No. 4 ( fig. 8)

figure 8 Jabal al-Khraymāt no. 4 in standardised Dadanitic glyphs

This is located above the right lion. ġṯṯt fʿl ʾlʾs¹d (1) Ġṯṯt made the (2) lion. The glyph of the ġ in the name Ġṯṯt is carved differently from the usual Dadanitic letter-shape ġ: the two parallel strokes are joined at the bottom, forming a triangle. As for the f in fʿl, it has a slightly squarer form than usual. Line 1 The personal name Ġṯṯt has not appeared previously in Dadanitic. It has also not been found previously in this particular form in any other language. However, the personal name Ġṯṯ is known in Qatabanic (Hayajneh 1998: 206; see also for more parallels, Sholan 1999: 116) as a feminine name. In Safaitic (Harding 1971: 452) and Hismaic (King 1990: 532), the names Ġṯ and Ġṯt appear, but

dadanitic inscriptions from jabal al-khraymāt

227

it is not certain if these personal names derive from ġṯṯ or from the very common Arabic root ġwṯ, “to help, to save”. Ġṯṯt seems to reflect the Arabic form ġaṯāṯah, “thinness, leanness” (Ibn Manẓūr, s.v. ġṯṯ). Dadanitic orthography does not show geminated consonants, so the repetition of ṯ here means that the two consonants were separated by a short vowel, as they may be in kll,13 in the expression ʾẓll h-ẓll, and in the participle ʿrr (see, for more details Macdonald 2004: 496–497, 510). Ġṯṯt is a male personal name, as is confirmed by no. 7 below, where it is followed by bn, “son of”. Lines 1–2 The occurrence of ʾl-ʾs¹d (also in Jabal al-Khraymāt no. 6 below), formed with ʾl- rather than with h- or hn- (the latter before ʿ and ʾ) which are common in Dadanitic, can be erroneously interpreted as an argument for considering these texts as Arabic or Dadano-Arabic.14 The article ʾl- is however an areal feature shared also by Classical Arabic, it is not a characteristic which requires all the texts in which it appears to be considered as Arabic. Indeed, as Al-Jallad (2014: 6, see pp. 14–15 for more details) has pointed out: The ʾl article is not restricted to Classical Arabic, but is also attested across the varieties written in the ana scripts. Given the great diversity of article forms and assimilation patterns within Arabic itself, the fact that ʾl is not restricted to Arabic, and that the entire category is the result of areal diffusion, one is not on particularly strong grounds when using the phonological shape of the article as a diagnostic feature. It should also be noticed that the ancient Arabs are said to have worshipped the god Yaġuṯ in the form of a lion (Khoury 2001: 243). Although the veneration of this god is not attested in either ana or Nabataean, we may find it in the Dadanitic compound personal name yġṯ-ʾl in JaL 65a (Hidalgo-Chacón Díez 2010: 274 and the commentary there), as well as in the Nabataean personal name ʾmrʾyʿwt (al-Khraysheh 1986: 39). Even in these names, however, yġṯ need not be interpreted as a theophoric element: Dadanitic yġṯ-ʾl could be translated as “Yaġūṯ is god” or as “ʾIl aids”, and in ʾmrʾyʿwt, yʿwt, which is the Aramaic equivalent of Yaġūṯ, could be the name of a god or that of a partic-

13 14

Exceptionally this is found as kl in Jabal al-Khraymāt no. 1 l. 4 (see above). The term Dadano-Arabic refers to an inscription written in the Dadanitic script but which contains elements from both Dadanitic and Arabic, see Macdonald 2000: 52–53.

228

hidalgo-chacón díez

ularly venerated individual. In asa inscriptions, it is completely unknown as a divine name and is only found in a Sabaic inscription (rés 5002) as an attributive personal name (Robin 2007: 831–832), even though Yaġuṯ was venerated by the tribes of Maḏḥij and Murād. It was also known at the beginning of the Islamic period, and venerated not only by the Maḏḥij, but also by other tribes like Qurayš and Hawāzin (Fahd 1968: 191–193). Jabal al-Khraymāt No. 5 ( fig. 9) figure 9 Jabal al-Khraymāt no. 5 in standardised Dadanitic glyphs

Between no. 4 and the back of the lion. s¹lmn f---- h-tmʿl S¹lmn f---- the clever (?). As mentioned in the introduction, the drawing of the right lion was made after this inscription and partly hides it. S¹lmn. The text begins with the personal name s¹lmn, which also appears in the Dadanitic inscriptions JSLih 77, u 16, u 77.1, Al-ʿUḏayb 44 and Qaṣr al-Ṣāniʿ 7. f----. The second word is difficult to interpret with certainty because the tail of the lion was carved over it. tmʿl. This substantive, determined by the article h-, has not been found before in any Dadanitic inscription. It is possible that it is related to the Arabic adjective maʿl which means “an agile, acute, clever (man)” (Lane 1968: 3022a). A personal name with the same form may occur in the Safaitic inscription bpws 195, {t}m{ʿ}l.

dadanitic inscriptions from jabal al-khraymāt

229

Jabal al-Khraymāt No. 6 ( fig. 10)

figure 10 Jabal al-Khraymāt no. 6 in standardised Dadanitic glyphs

To the left of the front legs of the left lion. ʿs²bt fʿl ʾl-ʾs¹d b-s²hr ʾl-gb{y} b hnʾmnt (1) ʿs²bt (2) made (3) the lion (4) in the month of (5) ʾl-{Gby} at (6) Hnʾmnt. Line 1 ʿs²bt. This is the only occurrence of this personal name in Dadanitic. It is known, however, in Safaitic (Winnett and Harding 1978: 343 = wh 2245) and in Arabic (al-)ʿUšba (Caskel 1966: 576; Hess 1912: 40). The name may be related to the Arabic epithet ʿašabat, “short”, applied to a man and ʿašīb, “short, and ugly, or despicable”, said of a woman (Lane 1968: 2050b). In Sabaic, the name Yaʿšib exists, deriving from ʿšb, “to be dry” (Ryckmans 1934: 171). Lines 2–3 fʿl ʾl-ʾs¹d see inscription no. 4 above. Line 4 b-s²hr, “in the month”. Šahr is the normal Arabic word for expressing “month”. S²hr is also found in Safaitic (Al-Jallad 2015: 344). In Taymanitic, “month” is expressed by yrḫ (see WTay 20 in Winnett and Reed 1970: 102–103)15 and in asa 15

Winnett and Reed read yrh rather than yrḫ.

230

hidalgo-chacón díez

by wrḫ (dasi, s.v. wrḫ). Note that s²hr, yrḫ and wrḫ have not been previously found in Dadanitic. Line 5 Gby. I would suggest that this is the name of the month, determined by the Arabic article ʾl. I would suggest interpreting gby here on the basis of Arabic jabā, “to return, to go back” (Lane 1968: 378b), Sabaic gbʾ, “to return” (Beeston et al. 1982: 48) and Gəʿəz gabʾa (Leslau 1987: 176). However, the Semitic root gbʾ/w/y also has the sense of “to collect water”, and “to collect money or tribute” (Cohen et al. 1970–: 93–94). There are therefore two possible interpretations for the month’s name: 1/ it refers to the month during which water or money was collected; or 2/ it refers to the month of return (perhaps from a religious feast or from a trade journey). Our knowledge of month names and their functions in Dadanitic is non-existent at present, and so these two interpretations can be no more than hypotheses. As a personal name, Gby occurs several times in Safaitic (see ociana, s.v. gby). Line 6 From an etymological point of view, ʾmnt is a word with the feminine ending -t derived from ʾmn: Gəʿəz ʾamna, “to believe, to trust, to have faith in, to have confidence” (Leslau 1987: 24); Hebrew ʾmn, “to prove to be firm, religious, faithful” (nifʿal) and “to have trust in, to believe in” (hifʿil) (Köhler and Baumgartner 2001: 63–64); Syriac ʾamīn, “true, lasting” (Payne Smith 1903: 19); Arabic ʾamina, “to trust a person with respect” (causative-stem, “to give security”) and ʾamīn, “safety”, ʾamānat, “trust, deposit” (Nöldeke 1952–1954: 41); Sabaic ʾmn, “be secure, to do something with impunity” and ʾmnt, “person/thing under protection of someone” (Beeston et al. 1982: 6); and finally Safaitic ʾmn, “a guard” (Al-Jallad 2015: 300). As a personal name, ʾmnt is found both as a feminine (Qatabanian, see Hayajneh 1998: 87) and masculine name (Safaitic, see ociana, s.v. ʾmnt). See also the Arabic personal name ʾĀminat (Caskel 1966: 156). The etymology of the place name suggests that this was likely a place of safety, probably a vantage point for defending the area from potential external attack (as also could be the case of the aforementioned Dadanitic graffiti from Jabal Ithlib). It is worth recalling that the text was found high up on a rock-face, in an area west of the Wādī al-ʿUlā which is on the way to oasis of al-ʿUlā. There are no other Dadanitic inscriptions in the immediate surroundings. Since the place name hnʾmnt is not attested in any other Dadanitic inscription, it suggests that it could have been a private property. One can compare the place names attested only in the Dadanitic inscriptions from al-ʿUḏayb (Hidalgo-Chacón Díez 2014). These are votive inscriptions ded-

dadanitic inscriptions from jabal al-khraymāt

231

icated to the gods who are asked, by means of a particular ceremony or rite, to protect the agricultural assets and private properties located in specified places. Jabal al-Khraymāt No. 7 ( fig. 11)

figure 11 Jabal al-Khraymāt no. 7 in standardised Dadanitic glyphs

ġṯṯt bn ʾs¹d dʿw ʾs¹fr (1) Ġṯṯt son of ʾs¹d (2) read [the] inscription. Line 1 Ġṯṯt. This personal name also appears in no. 5, for which see above. ʾs¹d. The patronym ʾs¹d, meaning “lion” is very well-known in Arabian onomastics (see for instance Safaitic and Dadanitic in ociana, s.v. ʾs¹d). Line 2 dʿw. Despite a slight doubt about the reading of the first letter, it seems likely that the word is dʿw, a verb derived from dʿw, a cognate of Arabic daʿā, “to claim”, “to make come”, “to pray to God” (Cohen et al. 1970–: 290). However, the translation of dʿw in the epigraphical context is more likely “to read (the inscription)” (see Al-Jallad 2015: 309). ʾs¹fr. The substantive s¹fr, “script” is well attested in Dadanitic, with the sense of “inscription”, and it is usually preceded by the Dadanitic article h- (see for instance ah213 l. 3: h-s¹fr ḏh, “this inscription”). The form ʾs¹fr in this text may be interpreted as the broken plural of s¹fr, to be compared with Dadanitic ʾnḫl-h (Al-ʿUḏayb 73 l. 4) and ʾdṯʾ-h (Al-ʿUḏayb 71 l. 5), the plurals of nḫl, “palm garden” and dṯʾ, “crop of the season of the later rains” respectively, both determined by the enclitic personal pronoun -h. However, given the fact that ʾs¹fr is determined neither by the article hn- nor by a personal pronoun, I very tentatively suggest that the noun is singular, determined by the Arabic article ʾl in which the l has been assimilated to the following letter: ʾl + s¹fr > ʾ-s¹fr. Inscription JSLih 276, which has been classified as Dadano-Arabic (Macdonald 2000: 53), shows the same structure: zdlh bn {kl}b ḏ-ʿmrtʿ f ʿrr ḏġbt ʿrr ʾ-s¹fr ḏh, “Zdlh son of {Kl}bd

232

hidalgo-chacón díez

of the lineage of ʿmrtʿ. And so, may Ḏġbt dishonour he who dishonours this inscription”. Jabal al-Khraymāt No. 8 ( fig. 12) figure 12 Jabal al-Khraymāt no. 8 in standardised Dadanitic glyphs

Below the left lion. This graffito contains only a personal name. ʿmr This name is well-known in ana (see ociana s.v. ʿmr). Jabal al-Khraymāt No. 9 ( fig. 13) figure 13 Jabal al-Khraymāt no. 9 in standardised Dadanitic glyphs

At the top to the right. This graffito contains only a personal name. fr This personal name is also found in Safaitic (see ociana, s.v. fr) and in the Hismaic graffito kjc 513 (King 1990: 412).

c

Conclusion

The texts studied here show some features that are unusual in Dadanitic inscriptions: 1.

2.

The occurrence of the Arabic article ʾl instead of the common Dadanitic h- (hn- before ʾ and ʿ) in texts no. 4 and 6. A similar form of the article appears also in JSLih 71 l. 5: b-l-ḥgr, “in al-Ḥigr”. A case of assimilation of the Arabic article may occur in no. 7: ʾl + s¹fr > ʾ-s¹fr (cf. JSLih 276 ʾ-s¹fr ḏh, “this inscription”). The particle f ʾn cognate with Arabic fa-ʾinna in no. 1 l. 4. This occurs also in the text Al-Ḫuraybah 17 l. 4 from the oasis of al-ʿUlā.

dadanitic inscriptions from jabal al-khraymāt

3.

4.

233

Some words have not been encountered in other Dadanitic texts, as is the case of the verbs rḥw, “(they) relieved(?)” and ṣ{b}w, “(they) desired”, both in no. 1 l. 3–4 and tmʿl, “the clever (?)” in no. 5. Some of the personal names are attested for the first time in Dadanitic, such as Ks¹ṭ, Ṭbʿ and {Q}d (no. 1), Ġṯṯt (no. 4 and 7), ʿs²bt (no. 6), and Fr (no. 9), or have appeared only once before, as ʿṭy (no. 2), but have already been found elsewhere in ana and/or in later Arabic sources.

These features are by no means decisive reasons for identifying the language as Arabic (or even another variant of ana) for two reasons: Firstly, as we have seen above, the common Arabic article ʾl is an areal feature resulting from contact with languages spoken by adjacent populations. At this stage of our knowledge, it is even impossible to know whether ʾl- and ʾ- were two different forms (or dialectical variants) for h-/hn- within the Dadanitic language, or whether they occur in texts carved by speakers of another language, such as Arabic. Secondly, the fact that a personal name appears in a text does not mean that the language of that text is the same as the etymological language of the name. The misuse of onomastics to identify ethnicity by some scholars has been challenged by Michael Macdonald who rejects the assumption that “names contain linguistic information and that this represents the language use of their bearers”, and that the names found only or principally in some inscriptions “were exclusive to a particular community” and therefore “can be used to identify its members wherever they are found” and finally the misunderstanding that “adjacent communities were in some way in watertight compartments and that fashions in name-giving in one group would have no influence on another” (Macdonald 1998: 187–188). It is difficult to say in which language these texts are written, and they might actually be written in a ‘mixed’ language since they contain linguistic features of Arabic and of other varieties of ana than Dadanitic, this being the result of the existence of several languages in contact in this area.

Sigla ah bpws cad

Dadanitic Inscriptions published in Abū ʾl-Ḥasan 1997 (no. 1–196) and Abū ʾl-Ḥasan 2002 (no. 197–347). Unpublished Safaitic inscription in the digital archive ociana, University of Oxford. Roth, M.T., R.D. Biggs, J.A. Brinkman, M. Civil, W. Farber, E. Reiner,

234

hidalgo-chacón díez

M.W. Stolper (ed.). 1956–2010. The Assyrian Dictionary of the Oriental Institute, the University of Chicago. Chicago: University of Chicago. cll Dadanitic inscriptions published in Caskel 1954. dasi Digital Archive for the Study of pre-Islamic Arabian Inscriptions. Accessed June 4, 2016. http://www.dasiproject.eu Jabal Ithlib, Qaṣr al-Ṣāniʿ, and Al-ʿUḏayb Dadanitic inscriptions in the digital archive ociana, University of Oxford. JaL Dadanitic inscriptions published in Jamme 1974. JSLih Dadanitic inscriptions published in Jaussen and Savignac 1909–1922. ociana The Online Corpus of the Inscriptions of Ancient North Arabia. http:// krcfm.orient.ox.ac.uk/fmi/webd#ociana (accessed September 14, 2017). Qaṣr al-Ṣāniʿ See Jabal Ithlib. rés Répertoire d’épigraphie sémitique. Paris, Imprimerie Nationale, 1900– 1905. U. Dadanitic inscriptions published in Sima 1999. Al-ʿUḏayb See Jabal Ithlib.

References Abū ʾl-Ḥasan, Ḥ. 1997. Qirāʾah li-kitābāt liḥyāniyyah min jabal ʿikmah bi-minṭaqat al-ʿulā. ar-Riyāḍ: Maktabat al-malik Fahd al-waṭaniyyah. Abū ʾl-Ḥasan, Ḥ. 2002. Nuqūš liḥyāniyyah min minṭaqat al-ʿulā. “Dirāsah taḥlīliyyah muqāranah”. ar-Riyāḍ: Wizārat al-maʿārif. Beeston, A.F.L., M.A. Ghul, W.W. Müller, and J. Ryckmans. 1982. Sabaic Dictionary (English-French-Arabic). Louvain-la-Neuve: Peeters; Beyrouth: Librairie du Liban. Caskel, W. 1954. Lihyan und Lihyanisch. (Arbeitsgemeinschaft für Forschung des Landes Nordrhein-Westfalen, Geisteswissenschaften 4). Köln: Westdeutscher Verlag. Caskel, W. 1966. Ǧamharat an-Nasab. Das genealogische Werk des Hišām Ibn Muḥammad al-Kalbī. Leiden: Brill. Cohen, D., F. Bron, and A. Lonnet. 1970–. Dictionnaire des racines sémitiques ou attestées dans les langues sémitiques. Louvain-la-Neuve: Peeters. Fahd, T. 1968. Le Panthéon de l’Arabie centrale à la veille de l’Hégire. (Bibliothèque archéologique et historique 88). Paris: P. Geuthner. Harding, G.L. 1971. An Index and Concordance of Pre-Islamic Arabian Names and Inscriptions. (Near and Middle East Series 8). Toronto: University of Toronto Press. Hayajneh, H. 1998. Die Personennamen in den qatabānischen Inschriften: lexikalische und grammatische Analyse im Kontext der semitischen Anthroponomastik. (Texte und Studien zur Orientalistik 10). Hildesheim: G. Olms. Hess, J.J. 1912. Beduinennamen aus Zentralarabien. (Sitzungsberichte der Heidelber-

dadanitic inscriptions from jabal al-khraymāt

235

ger Akademie der Wissenschaften. Philosophisch-historische Klasse 3). Heidelberg: C. Winter. Hidalgo-Chacón Díez, M. 2010. Die theophoren Personennamen in den dadanischen Inschriften. Marburg/Lahn: Philipps-Universität. Hidalgo-Chacón Díez, M. 2014. “Place Names in the Dadanitic Inscriptions of alʿUḏayb.” Adumatu 30:15–30. Hidalgo-Chacón Díez, M. 2015. “The Distribution of the Dadanitic Inscriptions according to their Content and Palaeographical Features.” Proceedings of the Seminar for Arabian Studies 45: 139–148. Hoftijzer, J., and K. Jongeling. 1995. Dictionary of the North-West Semitic Inscriptions. 2 vol. Leiden-New York-Köln: Brill. Ibn Manẓūr, Muḥammad b. al-Mukarram. 1955–1956. Lisān al-ʿArab, edited by A.F. alWānib. 15 vol. Bayrūt: Dār Ṣāder / Dār Bayrūt. Al-Jallad, A. 2014. “On the Genetic Background of the Rbbl Bn Hfʿm Grave Inscription at Qaryat al-Fāw.” Bulletin of the School of Oriental and African Studies 77 (3): 445– 465. Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. Leiden-Boston: Brill. Jamme, A. 1974. Miscellanées d’ancient arabe vii. Washington, dc: [privately produced]. Jaussen, A., and R. Savignac. 1909–1922. Mission archéologique en Arabie. 6 vol. Paris: E. Leroux-P. Geuthner. Khoury, A.T. 2001. Der Koran, Arabisch-Deutsch. Übersetzung und wissenschaftlicher Kommentar von Adel Theodor Khoury. Gütersloh: Mohn. al-Khraysheh, F. 1986. Die Personennamen in den nabatäischen Inschriften des Corpus Inscriptionum Semiticarum. Marburg/Lahn: Philipps-Universität. King, G.M.H. 1990. “Early North Arabian Hismaic. A Preliminary Description Based on a New Corpus of Inscriptions from the Ḥismā Desert of Southern Jordan and Published Material.” (PhD diss., School of Oriental and African Studies, University of London). Köhler, L., and W. Baumgartner. 2001. The Hebrew and Aramaic Lexicon of the Old Testament. Leiden: Brill. Lane, W. 1968. An Arabic-English Lexicon. 8 vol. Reprint of the edition published in London, 1863–1893. Beirut: Librairie du Liban. Larcher, P. 2009. “Les systèmes conditionnels en ʾin de l’arabe classique.” Bulletin d’études orientales 58: 205–232. Leslau, W. 1987. Comparative Dictionary of Geʿez (Classical Ethiopic): Geʿez-English / English-Geʿez with an Index of the Semitic Roots. Wiesbaden: Harrassowitz. Littmann, E., D. Magie, and D.R. Stuart 1907–1921. Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909, Division iii. Greek and Latin Inscriptions in Syria, Section a. Southern Syria. Leiden: Brill.

236

hidalgo-chacón díez

Macdonald, M.C.A. 1998. “Some Reflections on Epigraphy and Ethnicity in the Roman Near East.”Mediterranean Archaeology 11: 177–190. Repr. in Macdonald, M.C.A. 2009: chapter iv. Macdonald, M.C.A. 2000. “Reflections on the Linguistic Map of Pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 28–79. Repr. in Macdonald, M.C.A. 2009: chapter iii. Macdonald, M.C.A. 2004. “Ancient North Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R.D. Woodard, 488–533. Cambridge: Cambridge University Press. Macdonald, M.C.A. 2009: Literacy and Identity in Pre-Islamic Arabia. (Variorum Collected Studies Series cs906). Farnham-Burlington, vt: Ashgate-Variorum. Macdonald, M.C.A. forthcoming. “Towards a Re-Assessment of the Ancient North Arabian Alphabets Used in the Oasis of al-ʿUlà.” Arabian Epigraphic Notes. Nasif, A.A. 1988. Al-ʿUlā: An Historical and Archaeological Survey with Special Reference to its Irrigation System. Riyadh: King Saud University. Nayeem, M.A. 2000. The Rock Art of Arabia: Saudi Arabia, Oman, Qatar, the Emirates & Yemen. Hyderabad: Hyderabad Publishers. Nöldeke, T. 1952–1954. Belegwörterbuch zur klassischen arabischen Sprache, edited and reviewed by J. Kraemer. Berlin: De Gruyter. Payne Smith, R. 1903. A Compendious Syriac Dictionary: Founded upon the Thesaurus Syriacus of Robert Payne Smith, edited by J. Payne Smith (Mrs Margoliouth). Oxford: Clarendon Press. Robin, C.J. 2007. “Yaghūth.” In Encyclopaedia of Islam, vol. xii, Supplement, edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, and W.P. Heinrichs, 831–832. Leiden: Brill (Second edition). Ryckmans, G. 1934. Les noms propres sud-sémitiques. (Bibliothèque du Muséon 2). Louvain: Bureaux du Muséon. al-Said, S. 1995. Die Personennamen in den minäischen Inschriften. Eine etymologische und lexikalische Studie im Bereich der semitischen Sprachen. (Veröffentlichungen der orientalischen Kommission der Akademie der Wissenschaften und der Literatur— Mainz 41). Wiesbaden: Harrassowitz. al-Saʿīd, S.F. 1434–1435/2013–2014. “Nuqūš dādān: ad-dalālah wa-l-maḍmūn.” In Kunūz ʾaṯariyyah min dādān: natāʾij tanqībāt al-mawāsim as-sabʿah al-ʾūlā, edited by S.F. asSaʿīd, and A.S. al-Ġazzī, vol. 2, 281–302. (Dirāsāt ʾāṯāriyyah mīdāniyyah 1). Ar-Riyāḍ: al-jamʿiyyah as-saʿūdiyyah li-l-dirāsāt al-ʾaṯariyyah. Sholan, A. 1999. Frauennamen in den altsüdarabischen Inschriften. (Texte und Studien zur Orientalistik 11). Hildesheim: G. Olms. Sima, A. 1999. Die lihyanischen Inschriften von al-ʿUḏayb (Saudi-Arabien). (Epigraphische Forschungen auf der Arabischen Halbinsel 1). Rahden/Westf.: Leidorf. Winnett, F.V., and G.L. Harding. 1978. Inscriptions from Fifty Safaitic Cairns. (Near and Middle East Series 9). Toronto: University of Toronto Press.

dadanitic inscriptions from jabal al-khraymāt

237

Winnett, F.V., and W.L. Reed, with contributions by J.T. Milik and J. Starcky. 1970. Ancient Records from North Arabia. (Near and Middle East Series 6). Toronto: University of Toronto Press. Wood, R. (ed.). 1975. An Introduction to Saudi Arabian Antiquities. Riyadh: Department of Antiquities and Museums. Wuthnow, H. 1930. Die semitischen Menschennamen in griechischen Inschriften und Papyri des vorderen Orients. (Studien zur Epigraphik und Papyruskunde 1.4). Leipzig: Dieterich.

chapter 11

Un sanctuaire de montagne: Mushannaf Maurice Sartre

Sur les hautes terres du Jabal al-ʿArab, à 1500m d’ altitude, dans la partie nord du plateau central de la montagne, se situent à la tête du wādī homonyme, sur la pente nord-ouest de la vallée, sur une sorte d’ éperon tourné vers le sud, les vestiges d’un village antique aujourd’hui largement délaissé au profit de constructions plus modernes, Mushannaf (Dussaud 1927: 359). Souvent visité et décrit, notamment en raison de la présence d’ un temple assez bien conservé et de nombreuses inscriptions, il avait été identifié par W.H. Waddington avec Neila et Nelkômia, en s’appuyant sur le texte d’ une inscription (838)1. Cette identification a été très largement acceptée sans contrôle, malgré les réserves de R. Dussaud, et on la retrouve encore mentionnée ici et là2, bien que j’aie montré depuis près de 20 ans que l’ antique Neeila se trouvait bien loin de Mushannaf, à Inkhil (igls xiv, no 461), en Batanée. De plus, il n’y a pas davantage de raison d’admettre que Neila/Neeila désigne le même lieu que Nèlkômia. Il faut donc reconnaître que nous ignorons le nom antique de cette petite capitale des hauts plateaux. Cependant, il n’ est pas à exclure que l’inscription 830 en suggère le nom. En effet, après l’ empereur, le gouverneur et le centurion chargé de surveiller des travaux, est cité le « koinon des Manenoi », suivi du nom de deux magistrats chargés d’ exécuter l’ ouvrage. Cette succession de responsables irait bien dans le sens de la présence des villageois eux-mêmes en quatrième position, même si le nom est philologiquement peu compatible avec celui que porte Mushannaf aujourd’ hui. On connaît d’ailleurs d’autres exemples de noms antiques sans aucun rapport avec la toponymie actuelle: Orela-Jnaynah, Lebea-Bthaynah, Airè-Ṣanamein, Phaina-Mismiyyah pour ne citer que quelques exemples proches puisés dans la région.

1 Pour ne pas alourdir l’ appareil de notes de cet article, nous ne donnons ici que les références au futur corpus (tome xvi des igls), mais on trouvera à la fin un tableau de concordance indiquant l’ équivalence entre notre référence et les principales publications antérieures. 2 L’ identification de Mushannaf à Nela/Neila est encore affirmée sans aucune réserve par Weber 2006 : 117, renvoyant à Abu Assaf 1998.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

un sanctuaire de montagne : mushannaf

239

Le village venait d’être réoccupé par des Druzes depuis deux mois quand Cyril C. Graham, premier voyageur européen, y passa en 18573. Nombre de voyageurs s’y rendirent depuis cette date, et tous en décrivirent le vestige principal, celui qui attire l’attention de tous: le temple. La première vraie description de ce dernier est due aux archéologues de l’expédition américaine de Princeton en 1899 (aaes, ii, p. 346–351, fig. 122–123 et quatre photographies). Il se dresse sur le côté est d’un bassin en partie naturel et en partie artificiel, et est entouré sur trois côtés d’une vaste cour pavée délimitée par un mur où s’ ouvre une simple porte au nord, porte surmontée d’un linteau avec une inscription datant du règne d’Agrippa ier (815). Une autre inscription trouvée dans la cour mentionne Marc Aurèle (827), mais sans précision sur les travaux entrepris. H.C. Butler, en raison des détails de l’architecture du temple, rattachait cette inscription à la construction de celui-ci qu’il datait donc à la fin du iie siècle. Selon Butler, les murs ouest et nord du temple étaient d’origine, tandis que la façade avait dû être abattue et refaite grossièrement «quand le temple fut fortifié ». Un simple examen de l’état actuel confirme cette analyse, même si l’ on voit encore en façade l’emplacement des angles du bâtiment. Le mur de défense construit à l’ est avait utilisé en remploi, parmi d’autres, des blocs du temple permettant de reconstituer le décor des jambages de la porte principale fait de grappes de raisin en haut relief et de placer des consoles ainsi que des niches de chaque côté du portail. Ce temple se distingue de ceux de ʿAtīl, selon Butler, par son podium bas, son décor (frise de grecques et entablement de rinceaux d’ acanthes et de fleurs, surmonté d’oves et de fers de lance) tout autour de la cella, et par l’ ordre composite de ses colonnes. Bien qu’il n’ait vu aucun élément prouvant que le fronton de façade du pronaos soit arqué, Butler restituait néanmoins un fronton de ce type sous prétexte que «dans le Ḥawrān beaucoup de temples étaient ainsi». Grâce aux relevés de Butler, Clarence Ward fut en mesure de proposer une restitution et une datation du temple dès 1907, utilisant sans originalité les textes datés pour établir les différentes phases de construction du temple et de son temenos (Ward 1907: 1–6). Les Américains revinrent à Mushannaf en 1904 et explorèrent à cette occasion d’autres bâtiments au sud et à l’ est du temple et sur les pentes au sud-est, dans lesquels devaient se trouver des inscriptions (paes ii: 340). Le village conserve encore aujourd’hui des vestiges importants du grand temple. Celui-ci a fait récemment l’objet d’une thèse de Renate Barksay-Regner

3 On trouvera une liste plus complète des voyageurs dans la thèse de Barcsay-Regner 2007: 11– 16, mais nous ne signalons ici que ceux qui ont copié des inscriptions.

240

sartre

qui constitue désormais l’étude de référence4. La nôtre ne portera donc pas sur l’architecture du sanctuaire5, mais on voudrait, après avoir rapidement traité des divinités auxquelles il est consacré, s’ intéresser aux fidèles, grâce à un corpus d’une exceptionnelle importance (soixante-douze inscriptions grecques)6, et montrer son audience étendue à l’ échelle de la Syrie du Sud, voire de la province. Il ne semble pas qu’en dehors de Sīʿ un seul sanctuaire de la région ait joui d’une telle réputation. On tentera de comparer les quelques sanctuaires connus dans le Jabal al-ʿArab à partir des consécrations effectuées. Deux divinités sont nommées dans le corpus du village. D’ après l’ inscription la plus ancienne du corpus (815), le sanctuaire, dénommé simplement « maison », aurait été édifié par les villageois à l’occasion du retour d’ un roi Agrippa : Ὑπὲρ σωτηρίας κυρίου βασι|λέως Ἀγρίππα καὶ ἐπανόδου, κα|τ’ εὐχὴν Διὸς καὶ πατρ⟨ῴ⟩ου Ἀθ̣η̣ν̣α̣̃[ς] | σ̣ υ̣⟨ν⟩οδ̣ο⟨ς⟩ ὁμονο[ί]ας τὸν οἶκον ᾠ κοδόμ[ησεν]. Pour le salut de notre maître le roi Agrippa et son retour, conformément au vœu fait à Zeus et à Athéna ancestrale, l’ association de la concorde a fait construire le temple. Waddington tenait pour Agrippa ier et l’allusion à son retour concernerait son voyage à Rome en 41, lorsqu’il quitta la ville après l’ assassinat de Caligula et l’avènement de Claude. Cette opinion a été reprise par Schürer (1909: 553, mais Schürer 1973: 445, n. 19 laisse place au doute, Agrippa ier ou Agrippa ii), puis par Dittenberger (ogis 418), Sourdel (1952: 71, n. 6) et Chaniotis (151 no i, voir n. 6).

4 Barcsay-Regner 2007 ; ce travail très fouillé, riche de nombreux plans et illustrations, avait été précédé de quelques travaux préparatoires: Barcsay-Regner 1996: 22–26. Cf. pour le remploi moderne des matériaux antiques, Barksay-Regner 1991: 42–47, avec un plan de l’ensemble du temple et des maisons associées. 5 Freyberger 1998: 59, propose une date précise pour la fondation, 41 ap. J.-C., en se fondant sur l’ inscription 815. Barcsay-Regner 2007 : 96–99, suit d’assez près les inscriptions pour dater construction et embellissements successifs. 6 Renate Barcsay-Regner a confié à notre collègue Angelos Chaniotis le dossier d’inscriptions qu’ elle avait rassemblé et celui-ci a eu la courtoisie de nous communiquer au printemps 2011 ce dossier issu de la thèse publiée, puis la thèse elle-même. Qu’il en soit ici remercié. Les textes inédits ont été recensés dans le seg 57, no 1885–1901. seg 57, no 1884, donne une table de concordance entre la thèse de Barcsay-Regner et le volume igls à paraître. La mention « Chaniotis » renvoie à la publication des inscriptions dans la thèse de R. Barcsay-Regner.

un sanctuaire de montagne : mushannaf

241

J.H. Mordtmann (1884: 122) tenait pour Agrippa ii sans plus de justification. On rappellera qu’il est impossible, épigraphiquement, de distinguer le père du fils, et que le «retour royal» peut concerner n’importe quel voyage, y compris une visite royale à Mushannaf. Deux divinités sont associées dans l’inscription. D’ une part, Athéna Patrôos, dont l’épithète est curieusement placée avant le nom ; d’ autre part Zeus, à qui conviendrait aussi bien la même épithète. D’autres inscriptions du même village mentionnent les deux divinités, mais séparément. Ainsi Athéna apparaît seule dans deux inscriptions (812 et 814) avec l’épithète κυρία: (812) Πρόκλος ὁ καὶ Μασπ̣ [---]δου Κα|ναθηνὸς βουλευτὴς καὶ Θαι(μ)ος ὁ | καὶ Τειμόθεος καὶ Ἀντίοχος ὁ καὶ | Σαμεθος ἀδελφοὶ τῇ κυρίᾳ Ἀθη|νᾷ τὸ πρόπυλον σὺν παντὶ κοζ|μῷ ἐκ τῶν ἰδίων ᾠκοδόμησαν. Proclos dit aussi Mas …, fils de … dos, Canathénien, bouleute, Thaimos dit aussi Timothéos et Antiochos dit aussi Saméthos, ses frères, à notre maîtresse Athèna ont consacré le propylée avec tout son décor, à leurs frais. (814) [---]οτε ἐπιδημήσαντα |[---]α τοῦ ὑπερθύρου |[---]νων τῇ κυρίᾳ Α̣ θ̣η[νᾷ]. … à son retour … le linteau … à la maîtresse Athéna. Elle est dépourvue d’épithète sur une autre consécration (813) : Οναιφ|ος Ραγε|λου ἐπο|ίησεν τ|ῇ Ἀθηνᾷ. Onaiphos fils de Ragélos a fait pour Athèna. Zeus Kéraunios est honoré, seul, par un vétéran (816) : Διὶ Κερονείῳ | Ἰούλιος Οὐάλη[ς] | δω Κασσίου οὐ⟨ε⟩[τρ]|ανὸς ἠπύησ[εν ---]|[--]. À Zeus Kéraunios, Iulius Valens … fils de Cassius, vétéran, a fait faire …. On peut s’interroger sur la nature réelle de Zeus Tonnant, sans doute assez proche du Baʿalšamīn, «maître des cieux», et donc de l’ orage et de la pluie, qui est le grand dieu masculin de la Syrie du Sud (Sourdel 1952 : 71–72).

242

sartre

Peut-être Zeus et Athéna sont-ils réunis dans une inscription où ils ne sont pas nommés par leur nom mais seulement comme « les dieux ancestraux» (823): [Ὑπὲρ σωτ]|η̣ ρίας [καὶ] νε[ίκης κυρίου ---] |[---]|σαθος Οτασ[ου ---]|[---ταῖς ἑο]|ρταῖς πατρῴ[ων θεῶν ---]|[--- εὐξ]ά̣μ|ε̣νος τρίων τι[---] |[---]ὅλον τὸ | ἔθνος καὶ Ε[---]. Pour le salut et la victoire de notre maître … sathos fils d’ Otasos … pour les fêtes des dieux ancestraux … en ex-voto trois … l’ ethnos tout entier ( ?) et e …. Les autres textes ne mentionnent pas les dieux, même lorsqu’ il s’ agit d’ offrandes en leur honneur: ils étaient bien connus de tous. Y eut-il d’ autres divinités honorées? Th. Weber signale une tête et un relief du musée de Suwaydāʾ provenant de Mushannaf représentant un dieu barbu avec des cornes de cerf qu’il identifie au Zeus Hammon, patron de la iiie légion Cyrenaica stationnée à Bostra (Weber 2006: 117, pl. 80 a–c et e) ainsi qu’ une statue d’ homme en tunique retrouvée dans une cour à l’opposé du sanctuaire, qui pourrait être celle d’un sacrificateur (Bolelli 1986: 322 no 45 et 89 et 348, no 7, pl. ii, 7 ; Weber 2006: 117, pl. 80g). Les textes sont rarement datés, mais beaucoup fournissent néanmoins des indices. On a vu que la plus ancienne inscription remonte au temps des Hérodiens, sans que l’on puisse préciser s’il s’agit du règne d’ Agrippa ier (41–44) ou de celui de son fils Agrippa ii (v. 50–v. 90). On trouve ensuite des inscriptions datées du temps de Marc-Aurèle (826, 827), de Commode (830), de Sévère Alexandre (831); une autre est datée précisément par les consuls de 235 (828). Il faut sans doute placer au ive siècle au plus tôt l’ offrande d’ un officialis nommé Flavius Malchos fils de Iustus, originaire de Nèlkômia: le gentilice impérial porté de la sorte par les agents impériaux renvoie en effet à une date assez tardive (838). Mais d’autres inscriptions non datées appartiennent sans doute aux iie-iiie siècles si l’on en juge par l’écriture, comme une inscription très mutilée (814); d’autres sont peut-être plus anciennes, comme l’ inscription en relief 817 (peut-être dès le ier siècle, quoique Chaniotis préfère le iiie siècle), le iie siècle ou le début du iiie (820, 821 selon Chaniotis). Une seule inscription ouvertement chrétienne est datée avec précision de l’année 387 de l’ère provinciale, soit 492–493 (825) : il devait y avoir longtemps que les cultes païens avaient cessé, même dans ce sanctuaire de montagne. Une autre inscription chrétienne (une croix est gravée en tête), très fragmentaire, concerne la construction d’un pigeonnier, sans doute au-dessus d’ un tombeau,

un sanctuaire de montagne : mushannaf

243

ce qui m’incite à en placer la date vers le milieu du ive siècle, comme toutes les inscriptions semblables (848). La provenance diverse des fidèles qui ont laissé leur marque permet d’ affirmer le caractère non strictement villageois du sanctuaire. Certes, nombre de fidèles portent des noms indigènes sans indiquer d’ origine, ce qui nous prive de renseignements sur leur compte, mais les noms appartiennent clairement à l’onomastique propre au Ḥawrān: Θαιμος (812, 819), Σαμεθος (812), Οναιφος (813), Μαλεχου (817, 817a), Οβαισαθου (819), etc. Pour ceux qui indiquent une origine ou le nom d’un groupe, on peut relever les indications suivantes. Figurent des individus originaires de la région, mais non de Mushannaf même. Ainsi, un bouleute de Canatha, Proclos, fait une consécration avec ses frères à Athéna (812); on notera que deux frères portent un nom grec et un surnom indigène, et le troisième l’inverse. Un autre Canathénien apparaît dans un texte qui doit être une inscription honorifique (829) plutôt qu’ une épitaphe (mais ce n’est pas une consécration); il me semble devoir être identifié à un personnage connu par ailleurs, originaire du village de Slaym, qui fut ambassadeur à Rome sans doute en 149, lors du règlement d’ un différend entre Canatha et Soada après l’érection de ce dernier village en cité (igls xvi, no 82 et no 321)7. Un troisième Canathénien – mais qui pourrait être identique au premier puisqu’il est bouleute et anonyme – est mentionné dans un texte très mutilé où il est question d’une dépense faite «à ses frais» (829a). On trouve également mention de deux entrepreneurs ([ἐ]ργολά̣β̣οι) originaires de Bostra (839), qui ont financé quelque chose dont le nom a disparu. Une femme de Bostra au nom indigène bien attesté, Ονημαθη, se trouve sur une inscription en compagnie de son mari et de son fils sans que l’ on sache s’ il s’ agit d’une consécration, d’une construction publique ou sacrée, ou de celle d’ un tombeau (857). Deux autres textes sont d’interprétation plus difficile. Une dédicace en l’ honneur d’Athéna a été attribuée par J.T. Milik (Milik 1972: 91) à un homme originaire de Palmyre ou de Syrie du Sud, Onaiphos fils de Ragelos, pour la raison que son nom se trouve à Palmyre (813); on restera prudent à cet égard, de même que pour les autres suggestions du même savant en faveur d’ une possible origine palmyrénienne des dédicants du 817 (ibidem), du 821 (ibidem), des ergolaboi (839), et des individus mentionnés au 842 et 855, sur des bases uniquement onomastiques. En revanche, le rapprochement a été fait avec de

7 La réalisation d’ une ambassade apporte une gloire particulière à ceux qui l’accomplissent et à leurs descendants : à Mushannaf même, un ancien centurion rappelle que son grand-père avait accompli une ambassade (849b).

244

sartre

meilleures raisons entre un T. Iulius Oraros mentionné à Palmyre (igls xvii, no 381) et un Iulius Oréros qui célèbre l’accomplissement de sa liturgie dans un texte très difficile de Mushannaf (833). Ce peut être un Palmyrénien (Milik 1972: 92, Yon 2007: 396–397). À côté de ces mentions d’habitants des cités de la région, un lot de textes relève des offrandes accomplies pour le salut des empereurs. Deux ont été dressées au temps du gouverneur Avidius Cassius, vers 171, pour le salut de Marc Aurèle par Titus Aurelius Quirinalis Gemellus (826 et 827), centurion de la iiie légion Gallica, que l’on trouve également à l’ œuvre à Phaina du Trachôn (igls xv, no 4) où il devait être en garnison. Une dédicace de même type pour Sévère Alexandre est érigée par un certain Aurelius Marcus Chaammôn, fils d’Alexandre, dont les titres sont perdus; son cognomen le désigne comme d’origine locale (831), ce qui n’empêche pas qu’ il ait été soldat, sans doute officier. Au titre des agents romains, il faut encore signaler un syndic des nomades (818), un agent de l’ officium du gouverneur (sans doute celui de Bostra), Flavius Malchus, qui indique son origine, Nèlkômia, village non identifié mais qui ne peut pas être Mushannaf (838). On trouve également des soldats (dont un strator) (828), et un vétéran (816). Bien qu’il ne s’ agisse pas de consécrations ou d’offrandes aux dieux, il faut néanmoins signaler une sur-représentation des soldats dans l’épigraphie funéraire du village, alors qu’ il n’ est pas attesté comme lieu de garnison: un vétéran cavalier légionnaire (840), un tesserarius de la iiie légion Cyrenaica (841), un centurion (844a), un ancien beneficiarius devenu centurion (847) et des vétérans (844, 845a, 849b – ancien centurion –, 849c, 859c). Il est vrai qu’il existe des garnisons dans plusieurs villages peu éloignés, mais on a l’impression que se faire inhumer à Mushannaf recèle une valeur particulière, peut-être simplement parce que le tombeau avait une chance d’être remarqué par davantage de passants. Enfin, plusieurs offrandes sont le fait de groupes. Une consécration (823) semble effectuée au nom de l’ ethnos entier, sans que l’ on puisse préciser davantage en raison de l’état de la pierre. Une dédicace à Athéna est dédiée par le synodos de la Concorde (815), une association locale dont on ne sait rien d’autre. Au 830, l’offrande – un étage supplémentaire d’ un édifice non désigné – provient du koinon des Manènoi, en 189–190. On ne sait ce qu’ est ce koinon, terme vague qui peut aussi bien désigner un village entier comme on en a des exemples à Ṣūr al-Lajā (igls xv, no 96), à Agraina (igls xv, no 275), à Ṣanamayn (igls xiv, no 555), une communauté particulière au sein du village, voire une tribu. En revanche, la mention de la tribu des Chauchabènoi (827a) ne renvoie qu’à un seul dédicant dont rien n’indique qu’ il agisse au nom de sa tribu; mais le texte est mutilé. Il est probable que cette tribu soit

un sanctuaire de montagne : mushannaf

245

localisée dans le village voisin de Rāmah, que nous avons identifié avec Sodala, mentionné dans une inscription de Namarāh (1394). Les Aouarènoi, désignés comme une tribu (835), offrent une kamara, c’est-à-dire un īwān, une salle de réunion. S’agit-il d’une tribu du village? d’un groupe extérieur au village? Le nom est inconnu par ailleurs. Enfin, une épitaphe concerne un homme qui se déclare membre de la tribu des Rhaseiaènoi (845 : φυλῆς Ρασειαιη[νῶν -]), probablement un groupe extérieur au village si l’ adjectif qui suit son nom constitue une indication d’origine villageoise (Γαυτος Νατιμαθου Φωσμανειος) ; il aurait choisi Mushannaf comme lieu de sépulture en vue. On ne peut malheureusement rien dire des fêtes en l’ honneur des dieux, bien qu’une inscription y fasse allusion: [---ταῖς ἑο]|ρταῖς πατρῴ[ων] θεῶν ---] (823). On est mieux renseigné sur les offrandes laissées par les fidèles. Elles portent la plupart du temps sur des réfections ou des parties ajoutées ou rénovées d’édifice. Ainsi, un bouleute de Canatha finance avec ses frères les propylées du sanctuaire (812). Un autre fidèle a financé l’ érection de deux colonnes, soit qu’elles aient formé seules un monument commémoratif, soit qu’ il s’agisse de deux colonnes d’un portique ou d’ un édifice plus important (817a); on privilégiera néanmoins la première solution, car le même fidèle a offert aussi deux petites Victoires (817), auxquelles les deux colonnes pouvaient servir de support ou de cadre. Il est parfois difficile de faire le partage entre les offrandes des fidèles et les constructions entreprises par la communauté ou les responsables du sanctuaire. Ainsi, il me semble que les inscriptions où apparaissent des magistrats villageois comme un syndic (818b) ou des pistoi (834, avec aussi un syndic), des trésoriers sacrés (818a, 832), relèvent difficilement de l’offrande, si ce n’est de la communauté entière. Ainsi est construit un édifice nommé «le petit sigma» (τὸ σιγμ[ά]τειον, 818b), un arc (834), une niche avec un reposoir pour le dieu (832). Relève en revanche de l’offrande individuelle une partie (ou la totalité?) d’ un entablement (821, 831), une autre partie du sanctuaire du dieu dont le nom n’est pas identifiable (833), la rénovation d’un portique (838), un édifice ou édicule avec tout son décor construit en «acte de piété» (les deux dédicants ont agi [εὐ]|σεβοῦντες) (837). En revanche, les statues font l’objet de plus rares consécrations. Les textes mentionnent explicitement deux petites Victoires (817), terme que l’ on retrouve à Ṣanamayn en Batanée (igls xiv, no 559) ; la pierre qui porte le texte montre que les deux statuettes étaient dressées sur une sorte d’ autel sur le dessus duquel était creusée une cupule pour les libations. Une autre inscription (819) évoque l’offrande d’une Nikè, sans doute l’ une de ces statues dont le Ḥawrān a fourni de nombreux exemples, peut-être en suivant le modèle fourni par le monument commémoratif de la Victoire érigé à Saḥr (Weber 2009). Bien que Nikè ne soit jamais mentionnée à Mushannaf comme divinité honorée,

246

sartre

l’offrande d’une victoire convient parfaitement à Athéna/Allat, divinité guerrière. Peut-être faut-il ajouter une autre offrande de Nikè avec le texte 859a : trois lettres (αλε qui peuvent être le début d’Alexandros) figurent sur le socle d’une statue de Nikè dont ne subsiste que le bas de la robe. Le texte 824 est gravé sur un bloc orné d’un buste masculin malheureusement mutilé dont ne subsiste que le bas de la tête que l’ on devine barbue. Le personnage est vêtu d’un manteau étroitement serré de l’ échancrure duquel sort une main tenant un objet que l’on ne peut identifier. S’ agit-il du dieu ou du dédicant? Le texte ne porte plus que […] | εὐσεβ|ῶν ἀν|έθη|κεν, ce qui interdit toute conclusion. Cela peut fournir une première piste pour éclairer les raisons de la popularité du sanctuaire de Mushannaf chez les soldats et les vétérans: c’ est la déesse qui apporte la victoire. On la célèbre donc aussi bien chez les soldats romains que chez les auxiliaires indigènes, que représente ici le syndic des nomades (818). Michael Macdonald a en effet souvent rappelé qu’ une telle appellation, « les nomades», ne pouvait émaner que de l’administration romaine, et que jamais un nomade ne se désignerait de la sorte. La Victoire est particulièrement invoquée pour le salut de l’Empereur (826, 827, 827a, 830, 831) comme pour celui du roi Agrippa avant lui (815), y compris en dehors des milieux militaires (les Manènoi, 830, le synodos de la concorde 815). Il est difficile d’ aller plus loin dans cette voie, et l’épithète «ancestrale» n’éclaire pas davantage la personnalité de la déesse, si ce n’est pour insister sur son caractère local : on est en présence d’une interprétation grecque d’Allat, ou Lât comme elle est mentionnée à maintes reprises dans les inscriptions des nomades de la Ḥarrah8. Il me semble en revanche que le sanctuaire est aussi fréquenté pour une autre raison, qui diffère quelque peu de cette première fonction, même si elle ne lui est pas complètement étrangère. On a vu que la plus ancienne inscription datée (815) mentionne la construction du sanctuaire pour remercier les dieux ancestraux, Zeus et Athéna, du retour du roi Agrippa. Quel que soit le roi en question, quel que soit le voyage qu’il a effectué, on loue les dieux d’ avoir favorisé son retour sain et sauf. Or, cette mention n’est pas isolée. Elle est explicite dans un texte malheureusement très mutilé (814) mais où la mention ἐπιδημήσαντα ne fait aucun doute et renvoie à un retour heureux au pays. Je crois qu’on peut rajouter un troisième texte, où cette mention reste implicite

8 Seize inscriptions safaïtiques proviennent des abords de ce village ainsi que des gravures anépigraphes dont une scène de combat entre un cavalier armé d’une lance et un fantassin avec un bouclier et une épée : Zeinaddin 2000 : 285, no 2 pl. 66a–b.

un sanctuaire de montagne : mushannaf

247

mais s’impose. En effet, j’ai indiqué que le Canathénien Αλειφος Αβγαρου était sans doute le même homme qui est mentionné à la fois dans une inscription sur marbre de Suwaydāʾ datée de 149, portant une lettre impériale au sujet du litige entre Canatha et Soada (321) (Sartre 2013: 130–131) et dans une brève inscription de Slaym, village tout proche à la fois de Canatha et de Soada, qui rappelle qu’ il a accompli une ambassade à Rome. Événement sans aucun doute considérable dans la vie de ce villageois: je ne doute pas que l’ inscription qui l’ honore à Mushannaf (peut-être la dédicace d’une statue) ne soit la conséquence de son retour en bonne santé. Les dieux de Mushannaf, non contents d’accorder la victoire – du moins pour la guerrière Athéna –, seraient donc aussi des dieux protecteurs, des dieux du bon retour, des dieux du salut des voyageurs en quelque sorte. On aimerait savoir si Iulius Oreros (833), qui fait une consécration à l’ occasion de l’accomplissement de sa liturgie, se trouve dans la même situation : les ambassades comptent au nombre des liturgies, mais il en est bien d’ autres possibles. La célébrité de Mushannaf, au moins à l’échelon régional, pousse certains à venir y faire édifier leur tombeau. Le fait est d’autant plus frappant qu’ il s’ agit pour une part d’agents impériaux en poste auprès du gouverneur d’ Arabie siégeant à Bostra: son tesserarius (c’est-à-dire son porte-parole et transmetteur de mots de passe), un bénéficiaire/corniculaire, par exemple. Faut-il penser que l’inhumation près d’un sanctuaire réputé passait pour avoir des vertus particulières? Les dieux de Mushannaf protègent le voyageur, qui leur rend volontiers grâce après son bon retour; faut-il en déduire qu’ on attend aussi leur secours pour le dernier voyage? Rien n’étaye une telle interprétation, faute de documents explicites, mais la question mérite peut-être d’ être posée.

Équivalences entre igls xvi (à paraître) et les principales publications antérieures9 321 812 813*

W.H. Waddington, ISyrie, no 2307. W.H. Waddington, ISyrie, no 2216; A. Chaniotis, p. 158, no vii et pl. 150a et 150b. M. Dunand, RBi 41, 1932, p. 575, no 121; (seg vii, no 1075); A. Chaniotis, p. 168, no xxii et pl. 152b.

9 On n’a pas reproduit ici la totalité de la bibliographie mais seulement indiqué une ou deux publications accessibles, et ajouté le renvoi à la publication confidentielle d’Angelos

248 814* 815 816 817* 817a 818 818a* 818b 819 820 821 822 823* 824* 824a* 825* 826*

827 827a 828* 829

sartre R.F. Burton et C.F.T. Drake, Unexplored Syria, ii, pl. 6, no 126 et p. 386; (A. Chaniotis, p. 178, no xliv). W.H. Waddington, I.Syrie, no 2211; (ogis 418; igrr iii, no 1260); W.K. Prentice, aaes, no 380; (A. Chaniotis, p. 151, no i). R.F. Burton et C.F.T. Drake, Unexplored Syria, ii, pl. 6, no 119 et p. 386; (A. Chaniotis, p. 167, no xx et pl. 152a). M. Dunand, RBi 41, 1932, p. 577–578, no 128 (seg vii, no 1076; A. Chaniotis, p. 168, no xxiii et pl. 147c et 152 c et e). A. Chaniotis, p. 183, no lii et pl. 155a (seg 57, no 1890). W.K. Prentice, aaes, no 383; (A. Chaniotis, p. 167, no xxi). A. Chaniotis, p. 178, no xlviii et pl. 148b (dessin) et 154b (photo) (seg 57, no 1886). A. Chaniotis, p. 181, no xlix et pl. 138a et 154b (seg 57, no 1887). M. Dunand, Musée, p. 42, no 55 (A. Chaniotis, p. 169, no xxv). R.F. Burton et C.F.T. Drake, Unexplored Syria, ii, pl. 6, no 117 et p. 386; (A. Chaniotis, p. 164, no xiv). M. Dunand, RBi 41, 1932, p. 575, no 120; (seg vii, no 1074; A. Chaniotis, p. 169, no xxiv). W.H. Waddington, I.Syrie, no 2223; (A. Chaniotis, p. 167, no xix). A. Chaniotis, p. 187, no lix et pl. 157a et 157b (seg 57, no 1885). A. Chaniotis, p. 182, no l et pl. 154c (seg 57, no 1888). A. Chaniotis, p. 182, no li et pl. 148a, 155a (seg 57, no 1889). W.H. Waddington, I.Syrie, 2235; A. Chaniotis, p. 166, no xviii et pl. 151c et d. W.H. Waddington, I.Syrie, no 2214 (fr. c) (aaes, no 381a); W.K. Prentice, aaes no 381 (fr. a); (A. Chaniotis, p. 154, no iii et pl. 149a, fr. a et c) et p. 179, no xlvii, fr. b (sur photo de R. Barcsay-Regner, pl. 147b, 149b). W.H. Waddington, I.Syrie, no 2212; (W.K. Prentice, aaes, no 380a; A. Chaniotis, p. 153, no ii). R.F. Burton et C.F.T. Drake, Unexplored Syria, ii, p. 6, no 127 et p. 387, no 127; (A. Chaniotis, p. 178, no xlv). W.H. Waddington, I.Syrie, no 2215; (igrr iii, no 1263); A. Chaniotis, p. 157, no vi et pl. 149d. M. Dunand, RBi 41, 1932, p. 577, no 126 (seg vii, no 1080; A. Chaniotis, p. 177, no xli).

Chaniotis dans la thèse de Renate Barcsay-Regner (auteur des photos); cf. ci-dessus n. 6. Les inscriptions marquées d’ une astérisque ont été revues par nous. Toutes proviennent de Mushannaf, sauf la première (Suwaydāʾ) et la dernière (Namārah).

un sanctuaire de montagne : mushannaf 829a* 830* 831* 832 833* 834 835* 836 837 838* 839* 839a* 840* 841 842* 843* 844* 844a* 845 845a* 846* 847 848* 848a* 849 849a 849b* 849c*

249

Inédit. W.H. Waddington, I.Syrie, no 2213 (igrr iii, no 1262); A. Chaniotis, p. 155, no iv et pl. 149c. W.K. Prentice, aaes, no 382; A. Chaniotis, p. 156, no v et pl. 69a et 148c. W.H. Waddington, I.Syrie, no 2218; (A. Chaniotis, p. 160, no ix). W.H. Waddington, I.Syrie, no 2221; M. Dunand, RBi 41, 1932, p. 578, no 129 (seg vii, no 1082); A. Chaniotis, p. 163, no xii et pl. 150d. W.H. Waddington, I.Syrie, no 2219; (A. Chaniotis, p. 161, no x). W.H. Waddington, I.Syrie, no 2220; A. Chaniotis, p. 162, no xi et pl. 150c. W.H. Waddington, I.Syrie, no 2222; M. Dunand, RBi 41, 1932, p. 576, no 124 (seg vii, no 1078; A. Chaniotis, p. 164, no xiii et pl. 151a). M. Dunand, RBi 41, 1932, p. 577, no 127 (seg vii, no 1081; A. Chaniotis, p. 165, no xvi). W.H. Waddington, I.Syrie, no 2217; (A. Chaniotis, p. 159, no viii). M. Dunand, RBi 41, 1932, p. 576, no 122 (seg vii, no 1077; A. Chaniotis, p. 165, no xv et pl. 147a et 151b). Inédit. W.H. Waddington, I.Syrie, no 2228; (igrr iii, no 1265); A. Chaniotis, p. 172, no xxx et pl. 153a. W.K. Prentice, aaes, no 384; (A. Chaniotis, p. 176, no xxxix). W.H. Waddington, I.Syrie, no 2226; A. Chaniotis, p. 171; no xxviii et pl. 152d. M. Dunand, RBi 41, 1932, p. 576–577, no 125 (seg vii, no 1079); A. Chaniotis, p. 176, no xl et pl. 153c et 153d. W.H. Waddington, I.Syrie, no 2227; (igrr iii, no 1266; A. Chaniotis, p. 172, no xxix). Inédit. W.H. Waddington, I.Syrie, no 2224; (A. Chaniotis, p. 170, no xxvi). Inédit. W.H. Waddington, I.Syrie, no 2228 a; (A. Chaniotis, p. 173, no xxxi et pl. 153b). W.H. Waddington, I.Syrie, no 2225; (igrr iii, no 1264); (A. Chaniotis, p. 170, no xxvii). Inédit. A. Chaniotis, p. 184, no liii et pl. 155c (seg 57, no 1891). W.H. Waddington, I.Syrie, no 2231a; (A. Chaniotis, p. 174, no xxxv). A. Chaniotis, p. 185, no lv et pl. 156a (seg 57, no 1893). A. Chaniotis, p. 184, no liv et pl. 155d (seg 57, no 1892). A. Chaniotis, p. 185, no lvi et pl. 156b (seg 57, no 1894, qui indique par erreur no xlvi et donc igls xvi, no 859).

250

sartre

850*

M. Dunand, Musée, p. 48–49, no 74 et pl. xxi, no 74; A. Chaniotis, p. 177, no xlii et pl. 154d. 851 W.H. Waddington, I.Syrie, no 2232; (A. Chaniotis, p. 175, no xxxvi). 852* M. Dunand, Musée, p. 58, no 99 et pl. xix, no 99; A. Chaniotis, p. 177, no xliii et pl. 154e; (seg 56, no 1928). 853 W.H. Waddington, I.Syrie, no 2233; (A. Chaniotis, p. 175, no xxxvii). 854* A. Chaniotis, p. 187, no lx et pl. 157c (seg 57, no 1895). 855 W.H. Waddington, I.Syrie, no 2234; (A. Chaniotis, p. 175, no xxxviii). 856 W.H. Waddington, I.Syrie, no 2231; (A. Chaniotis, p. 174, no xxxiv). 857 W.H. Waddington, I.Syrie, no 2229; W.K. Prentice, aaes, no 385; (A. Chaniotis, p. 173, no xxxii). 858 W.H. Waddington, I.Syrie, no 2230; (A. Chaniotis, p. 174, no xxxiii). 858a* A. Chaniotis, p. 186, no lvii et pl. 156c (seg 57, no 1899, avec renvoi erroné à notre 854g). 858b* A. Chaniotis, p. 186, no lviii et pl. 156d (seg 57, no 1900, sans renvoi à igls xvi). 859 M. Dunand, RBi 41, 1932, p. 576, no 123; (A. Chaniotis, p. 178, no xlvi). 859a A. Chaniotis p. 188, no lxi et pl. 139c (seg 57, no 1897). 859b A. Chaniotis, p. 188, no lxii et pl. 157d (seg 57, no 1898). 859c* A. Chaniotis, p. 189, no lxiii et pl. 158a (seg 57, no 1896). 859d A. Chaniotis, p. 189, no lxiv et pl. 158b. 859e A. Chaniotis, p. 190, no lxv et pl. 158c. 859f * A. Chaniotis, p. 190, no lxvi (seg 57, no 1901). 859g* A. Chaniotis, p. 190, no lxvii et pl. 12c–d. 859h*. Inédit 1394* W.H. Waddington, I.Syrie, no 2265.

Sigles et abréviations aaes igls igrr ogis paes

American Archaeological Expedition to Syria in 1899–1900, tome iii, W.K. Prentice, Greek and Latin Inscriptions. London: W. Heinemann, 1908. Inscriptions grecques et latines de la Syrie. Cagnat, R. 1908–1911. Inscriptiones graecae ad res romanas pertinentes. Paris: E. Leroux. Dittenberger, W. 1905–1907. Orientis graeci inscriptiones selectae. Leipzig: S. Hirzel. Littmann, E., D. Magie, D.R. Stuart et W.K. Prentice. 1904–1921. Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909. Division iii Greek and Latin Inscriptions in Syria. Section a, Southern Syria. Leyde: Brill.

un sanctuaire de montagne : mushannaf Rbi seg

251

Revue Biblique internationale. Supplementum Epigraphicum Graecum. Leiden.

Bibliographie Abu Assaf, A. 1998. Gabal Hauran und seine Denkmäler. Mohafaza von Souweida. Damaskus: Ministère du Tourisme. Barcsay-Regner, R. 1991. «Zur Wiederverwendung und Rezeption antiker Baugliede in modernen Bauten im Hauran». Damaszener Mitteilungen 5: 39–48. Barcsay-Regner, R. 1996. «Einige Beobachtungen zu Planung und Ausführung am Tempel von Muschennef». Dans Bericht über die 38. Tagung für Ausgrabungswissenschaft und Bauforschung: vom 11. bis 15. Mai 1994 in Brandenburg (Koldewey Gesellschaft), 22–26. Bonn: R. Habelt. Barcsay-Regner, R. 2007. Der Tempel von Muschennef. Ein frühkaiserliches Heiligtum im Hauran/Syrien. Freiberg: Universitätsbibliothek «Georgius Agricola». Bolleli, G. 1986. «La ronde-bosse de caractère indigène en Syrie du Sud». Dans Hauran i. Recherches archéologiques sur la Syrie du Sud à l’époque hellénistique et romaine, vol. 2, édité par J.-M. Dentzer, 311–372. Paris: P. Geuthner. Dussaud, R. 1927. Topographie historique de la Syrie antique et médiévale. (Bibliothèque archéologique et historique 4). Paris: P. Geuthner. Freyberger, K.S. 1998. «Die frükkaiserzeitlichen Heiligtümer». Damaszener Forschungen 6: 59. Milik, J.T. 1972. Dédicaces faites par des dieux (Palmyre, Hatra, Tyr) et des thiases sémitiques à l’époque romaine. (Bibliothèque archéologique et historique 92). Paris: P. Geuthner. Mordtmann, J.H. 1884. «Griechische Inschriften aus dem Hauran». Archäologisch-epigraphische Mitteilungen aus Österreich-Ungarn 8: 180–192. Sartre, M. 2013. «Dionysias d’Arabie». Dans Le voyage des légendes. Hommages à Pierre Chuvin, édité par D. Lauritzen et M. Tardieu, 123–138. Paris: cnrs. Schürer, E. 1909. Geschichte des jüdischen Volkes im Zeitalter Jesu Christi, i. 4e édition. Leipzig: Hinrichs. Schürer, E. 1973. The History of the Jewish People in the Age of Jesus Christ (175b.c.– a.d.135), vol. i. Edinburgh: t & t Clark. Sourdel, D. 1952. Les cultes du Hauran à l’époque romaine. (Bibliothèque archéologique et historique 53). Paris: Imprimerie nationale. Waddington, W.H. 1870. Inscriptions grecques et latines de la Syrie, recueillies et expliquées. Paris: Didot. Ward, Cl. 1907. «The Temple at Mushennef, Ḥaurân, Syria». American Journal of Archaeology 11: 1–6.

252

sartre

Weber, Th. 2006. Sculptures from Roman Syria in the Syrian National Museum at Damascus. Vol. 1. From Cities and Villages in Central and Southern Syria. Worms: Wernersche Verlagsgesellschaft. Weber, Th. 2009. Hauran iv : Sahr al-Ledja. Vol. ii. Die Skulpturen aus Sahr und die Statuendenkmäler der römischen Kaiserzeit in südsyrischen Heiligtümern. (Bibliothèque archéologique et historique 184). Beyrouth: Ifpo. Yon, J.-B. 2007. «De l’araméen en grec». Mélanges de l’Université Saint-Joseph 60: 381– 429. Zeinaddin, H. 2000. «Safaitische Inschriften aus dem Ǧabal al-ʿArab». Damaszener Mitteilungen 12: 276–289.

chapter 12

Un pasteur et un soldat? Deux inscriptions grecques d’époque romaine à l’est du Jabal Ḥawrān François Villeneuve

Notre ami Michael Macdonald, éminent spécialiste des inscriptions nord-arabiques, celles entre autres qu’on dit conventionnellement et improprement « safaïtiques», s’est plusieurs fois interrogé sur leurs relations avec les inscriptions grecques, ou plus exactement l’usage occasionnel – rare – du grec par les auteurs, pasteurs ou principalement pasteurs de ces textes de la steppe. Il a d’autre part, non sans une salutaire vigueur, dans un article paru dans Syria et qui a fait date1, «remis les pendules à l’heure» sur les groupes nomades et leur rôle, dans le Ḥawrān et à la périphérie orientale de cette région. Ayant longtemps scruté le Ḥawrān des villageois sédentaires de l’ Antiquité, et longuement fréquenté l’une des zones de contact privilégiées – aujourd’hui comme dans l’Antiquité – des sédentaires et des nomades à l’ est du Jabal Ḥawrān, je ne peux trouver meilleur moyen de lui rendre hommage et lui témoigner mon amitié que de publier deux brèves inscriptions grecques « égarées», seules, au milieu de rochers portant de très nombreuses inscriptions safaïtiques, sur un site épigraphique de cette région, Tell Freikh. Certes, le cas de textes grecs « perdus» dans une masse d’inscriptions rupestres sémitiques, notamment nord-arabiques, n’est pas unique: on peut ainsi en trouver sur les itinéraires caravaniers du nord du Ḥijāz – mais il s’agit généralement alors de gens de passage ou de caravaniers, et ces documents sont au total fort rares. Beaucoup plus rares encore sont les textes grecs en milieu pastoral2, et rarissimes ceux

1 Macdonald 1993. 2 Pour la seule ḥarrah à l’ est du Ḥawrān, et marginalement dans le Ḥawrān même, toujours hors des villages et villes antiques, le nombre des inscriptions nord-arabiques, ici majoritairement safaïtiques, plus ou moins recensées, laissées par des pasteurs, semble dépasser les 20 000 (Thuillier 2010 : 37) ; celui des grecques en milieu équivalent et sur support équivalent – rochers – me semble difficilement dépasser la trentaine; en 1993, M. Macdonald en mentionnait sept (Macdonald 1993: 348, n. 289–290) ; Macdonald et al. 1996: 480–487 (site i, inscr. 11 et 12 ; site j, inscr. 1 et 2) y ajoutait deux bilingues gréco-safaïtique du Wādī Rushaydah et d’un affluent du Wādī Shām, publiées, en signalant la présence de quelques autres, bilingues ou grecques (dont une grecque publiée p. 464, e2, graffite en deux noms indiscutablement nord-

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

254

villeneuve

qui dépassent un peu le niveau du simple graffito portant un nom et parfois un patronyme pour donner une indication d’activité, sinon de métier. Les deux documents de Freikh appartiennent à cette catégorie et pourront à ce titre intéresser le public. L’Institut français d’archéologie du Proche-Orient a conduit, de 1985 à 1988, avec l’aide de la Direction Générale des Antiquités et Musées de Syrie, l’ étude monographique détaillée du site, presque exclusivement romain (en date), de Diyātheh (fig. 1), en bordure est du Jabal Ḥawrān, très précisément au contact du Ḥawrān et de la ḥarrah. À la différence de tous les villages antiques de cette bordure orientale, disposés en chapelet vers 1500 m d’ altitude, Diyātheh est à 1050m d’altitude et à plusieurs kilomètres plus à l’ est, en bordure du Wādī Shām, un important torrent intermittent. Ce site antique, que les paysans druzes n’ont réoccupé que durant quelques années du milieu du xxe siècle, est remarquable par l’association d’un village d’une petite centaine de maisons avec un grand fort militaire romain, des puits aménagés dans le lit du wādī, et un grand terroir irrigué à l’aval du site, là où le Wādī Shām achève de sortir des gorges pour s’étaler sur son piémont. Diyātheh marque la fin du monde habité par des sédentaires, toutes périodes confondues : au-delà, il n’ y a plus rien de construit jusqu’à l’Euphrate3. L’étude de Diyātheh, sans fouille mais détaillée, a naturellement conduit à l’exploration méticuleuse de son environnement. Cette exploration a mené à la découverte d’une éminence rocheuse appelée Tell Freikh, au milieu de la plaine formée par le cône de déjection du Wādī Shām, qu’ elle domine d’ une arabiques). Bader 1997 en a publié trois autres du Wādī Salmah en Jordanie du Nord-Est, et Bader 2009 encore trois autres. Zeinaddin 2008 : 336, sur la base du Safaitic Epigraphic Survey Project engagé par M. Macdonald en 1995 dans la région de Namārah, signale (sans donner les textes) douze nouveaux textes grecs (sur un total de 376 inscriptions), que H. Zeinaddin attribue pour partie à des soldats du poste militaire romain qu’il considère dater du courant du iie siècle et à des «Safaiten» (horresco referens !) « die sich in der griechischen Schrift übten» ; Macdonald et al. 1996, Al-Jallad et al-Manaser 2016 ont publié trois bilingues ‘partielles’ grécosafaïtiques, s’ ajoutant au texte grec « thematically close to the Safaitic inscriptions» qui est le no 2 de Al-Jallad et al-Manaser 2015 ; le no 1 de cette même publication, en revanche, passionnant texte préislamique de sept lignes en arabe écrit en lettres grecques, s’écarte un peu de la série des grecques thématiquement et géographiquement ‘safaïtiques’ et des bilingues gréco-safaïtiques, et pourrait être éventuellement plus tardif. Nous ajoutons ici deux textes grecs qui, heureusement, contribuent à infirmer un peu, progressivement, l’assertion d’AlJallad et al-Manaser 2015 : 51, que « most of the Greek inscriptions of this area […] contain only names ». 3 Villeneuve et Sadler 2001. Travaux conduits par Serge Sadler, architecte, avec F. Villeneuve et Hassan Hatoum, archéologues, et Philippe Thévenin, technicien coopérant.

inscriptions grecques d’ époque romaine

figure 1

255

Carte localisant Diyātheh et Tell Freikh s. sadler, f. villeneuve, s. vatteoni, 2003

quinzaine de mètres (carte fig. 1; photographie fig. 2). Nous sommes à 3,5 km à l’ est-sud-est de Diyātheh et à 970m d’altitude. La colline a un peu plus de cent mètres de longueur nord-sud, et une cinquantaine de mètres d’ est en ouest. La surface est très rocheuse et chaotique au sommet. Le toponyme, dont nous conservons par commodité la transcription française d’ époque mandataire, s’ écrirait plus proprement Furaykh, diminutif de farkh, donc « oisillon » – mais il peut s’agir d’un sobriquet: le tell d’Oisillon. Nous l’ avons visité à plusieurs reprises, en particulier de 1985 à 1988. Le tell n’ est que rarement habité par des nomades: nous n’y avons vu qu’une fois une tente, au printemps 2002. Il ne comporte pratiquement pas de vestiges ‘bâtis’ (hormis ceux indiqués cidessous), mais il est très riche en inscriptions rupestres: plusieurs centaines de documents, presque tous safaïtiques, et quelques rares arabes.

256

figure 2

villeneuve

Photographie de Tell Freikh vers l’ est f. villeneuve, mars 1988

Dans cette série, nos prospections de la fin des années 1980 ont permis de documenter (photographie, positionnement en plan) soixante-huit inscriptions safaïtiques4, disposées (comme la grande majorité des autres) en deux groupes d’une vingtaine de mètres de diamètre, séparés par un vide d’ une trentaine de mètres de long, toutes sur la face est de l’ éminence. Dans la zone sudest se trouvent deux grosses tombes-tumulus de plan ovoïde d’ allure ancienne, sensiblement sur la crête du tell et orientées est-ouest, et un peu plus bas à l’ est quatre petites tombes de plan rectangulaire, récentes, également orientées estouest. Un sinueux muret de pierres sèches de 15 m de longueur sépare tombes anciennes et récentes et passe au-dessus et à l’ ouest de toutes les inscriptions enregistrées, au nombre de vingt-trois. Dans la zone nord-est, quarante et une inscriptions safaïtiques et une arabe sont enregistrées; à l’ est de celles-ci, près de la crête, deux pierriers de 3 et 5m de diamètre sont séparés par un espace de 8m où se trouvent deux tombes récentes du type décrit ci-dessus, contiguës, tandis qu’une troisième se trouve juste à l’ouest du pierrier nord. Un muret longe ce pierrier nord sur 5m, le séparant de facto des inscriptions. Deux inscriptions safaïtiques enregistrées sont à l’écart, encore plus au nord. En résumé,

4 M. Macdonald a procédé à une première étude très détaillée de vingt-quatre de ces inscriptions (communication écrite du 23/08/2016 sur la base des photographies et informations que lui a remises notre mission).

inscriptions grecques d’ époque romaine

figure 3

257

Photographie de l’ inscription grecque no 1 de Tell Freikh b. villeneuve 17/04/2002

il y a une répartition de l’espace de fait, abstraction faite de toute chronologie: les inscriptions sont toutes plus à l’est et plus bas que l’ ensemble pierrierstombes-murets. Lors d’une visite de vérification le 17 avril 20025, deux inscriptions grecques ont été découvertes dans le grand groupe de rochers nord-est, mêlées aux inscriptions safaïtiques mais toutes proches l’une de l’ autre. Elles sont réalisées selon la même technique que le plus grand nombre des inscriptions safaïtiques de Freikh: martèlement sur rocher non retaillé, avec un gros percuteur, donnant des lettres épaisses, assez dépourvues d’angle aigu. Les lettres ont une hauteur assez régulière, de 6–7cm.

Inscription no 1 (fig. 3) Quatre lignes. Sur le haut de la face verticale d’un rocher. Le texte est complet. ζηνοδωροσ αυρηλιουυπι κασατηρησα τηνσκηνησ

5 F. Villeneuve, accompagné de Béatrice Villeneuve.

258

villeneuve

Les sigmas sont lunaires, mais assez anguleux. l. 1: avant le ζ se trouve un signe de forme intermédiaire entre un u en fer à cheval et un o non fermé au sommet (avec une sorte de petite flèche partant en bas vers la gauche), plus épais que les lettres de l’ inscription. Ce signe était là avant l’inscription, car pour l’éviter le graveur a écrit un ζ plus petit que les autres lettres et l’a décalé vers le haut. La 6e lettre est un oméga de type ω. l. 4: après le sigma, petit signe supplémentaire en forme d’ apostrophe. Au dessous de la l. 4, décentré vers le droite, petit dessin, de la taille d’ une lettre: deux signes superposés? Ou un φ? Ζηνόδωρος / Αὐρηλίου ὑπ(ε)ῖ/κας ⟨ἐ⟩τήρησα / τὴν σκήνη⟨ν⟩ υπικας ne peut guère être compris que comme ὑπεῖκας, participe aoriste de ὑπείκω, «obéir». Il s’agit là de la seule difficulté, ou plutôt de la seule surprise, le recours à un participe aoriste étant plutôt inattendu dans un texte fruste. Traduction mot à mot: Zénodoros fils d’Aurélios, obéissant, j’ai gardé la tente. Il semble permis de proposer une traduction moins littérale mais plus réaliste: Zénodoros fils d’Aurélios. J’ai obéi: je garde le campement. Zénodoros est un nom grec théophore (littéralement « don de Zeus ») d’ une grande banalité, très largement répandu dans tout l’ Orient et autour de la Méditerranée. Il peut s’agir, mais ce n’est nullement nécessaire, de ‘l’ interprétation’, en grec, d’un théophore sémitique de même signification, construit sur un grand dieu masculin. Aurélios est le latin Aurelius grécisé, très commun également, et plus encore à partir de 212 apr. J.-C., date où, l’ édit de l’ empereur romain M. Aurelius Antoninus (dit Caracalla) connu comme Constitutio antoniniana ayant accordé la citoyenneté romaine à tout homme libre de l’ Empire, ceux accédant à cette citoyenneté le firent massivement en prenant le nom de l’empereur régnant: Aurelius.

Inscription no 2 (fig. 4) Trois lignes. Sur la face horizontale d’un rocher, tout près de l’ inscription no 1. Le texte est complet.

inscriptions grecques d’ époque romaine

figure 4

259

Photographie de l’ inscription grecque no 2 de Tell Freikh b. villeneuve 17/04/2002

τανοσζαβου δουποιμενει θρεματαρ Les sigmas sont lunaires, mais assez anguleux. Les epsilons sont lunaires. Le rho final de la l. 3 ne fait pas sens; il peut être gravé par erreur, ou ce peut être comme le possible phi final de l’inscription no 1 un sigle que je ne comprends pas. Τανος Ζαβου/δου ποιμαίνει / θρέμ(μ)ατα⟨ ⟩ La restitution ne fait pas de difficulté, une fois le ε de ποιμενει corrigé en αι et le second μ réinséré dans θρέμματα, un pluriel du collectif θρεμμα, « animaux domestiques, brebis, chèvres», pluriel qu’on traduira donc par « les troupeaux (sous-entendu: d’ovins et caprins)». Traduction: Tan fils de Zaboud fait paître les troupeaux. Le nom «Tan» (ẓn) est «très commun dans les inscriptions safaïtiques »6. On le retrouve même probablement, en grec, sous la forme Ταενο(ς) comme 6 Courrier électronique de M. Macdonald à F. Villeneuve, 30/11/2003.

260

villeneuve

patronyme, dans un graffite rupestre encore inédit du nord du Ḥijāz7. Zaboud, voire Zabboud, est une variante familière des très répandus anthroponymes sémitiques construits sur la racine zbd, présents dans les inscriptions nordarabiques. Je n’hésite pas à traiter les deux textes comme contemporains: leur voisinage, parmi des centaines d’inscriptions d’une autre langue, et l’ identité de leur graphie y convient. On ne peut guère, en revanche, s’ aventurer à les dater. La graphie n’est pour cela d’aucun secours. Le seul élément susceptible de donner une indication chronologique, mais de valeur uniquement statistique, est le nom Aurélios: comme indiqué supra, il est beaucoup plus répandu à partir de 212 apr. J.-C. qu’avant. On peut donc privilégier, sans certitude, une date au iiie ou au ive siècle apr. J.-C.8. Nos deux inscriptions sont donc en grec, seules à Tell Freikh, mais elles ont en commun avec les inscriptions safaïtiques au milieu desquelles elles se trouvent: – un support: rocher brut; – un mode de gravure: au percuteur lithique; – et un contenu: en termes d’activité: garder le campement, faire paître le troupeau – celui par excellence des inscriptions safaïtiques. En revanche, elles ne contiennent pas de formule de prière (formules qui ne sont certes pas la règle mais qui sont fréquentes dans les textes nord-arabiques). En forçant le trait: ce sont donc des inscriptions nord-arabiques … en grec. Pourquoi sont-elles en grec? Rappelons d’abord qu’elles sont tout à fait voisines l’ une de l’ autre, parmi un très grand nombre d’inscriptions safaïtiques : ce voisinage n’a aucune chance d’être dû au hasard et ne peut pas s’expliquer par la simple contemporanéité qui en découle.

7 Il s’ agit de l’ inscription grecque no 3 de Umm Jadhāyidh, découverte lors de la prospection épigraphique et archéologique dirigée par le Dr. Ali al-Ghabbân pour le compte du Service saoudien du tourisme et des antiquités, en collaboration avec Laïla Nehmé. Les inscriptions grecques me sont confiées pour étude dans le cadre de la publication de cette prospection. 8 L’ argument d’Aurelius et de la constitution antoniana ne donne certes qu’un terminus post quem. Cependant, dépasser le ive siècle apr. J.-C. paraît hasardeux: les autres inscriptions de Freikh, sauf les quelques arabes – non datées – sont safaïtiques, donc généralement considérées antérieures au ve siècle ; un nom théophore païen comme Zénodoros ne disparaît certes pas avec la christianisation, mais tend à devenir nettement moins fréquent.

inscriptions grecques d’ époque romaine

261

Sont-elles pour autant identiques au regard de leurs auteurs et de leurs contenus? Non. Une première différence réside dans l’origine linguistique des noms. L’auteur de l’inscription no 2, Tanos Zaboudou, a un nom et un patronyme parfaitement sémitiques, son nom propre étant même bien nord-arabique; ẓn bn zbd ne surprendrait personne dans une inscription safaïtique. Celui de l’ inscription no 1 a un nom parfaitement grec et un patronyme parfaitement latin. De toute évidence, cela n’exclut en rien qu’il soit un sémite oriental dont le nom Zénodoros peut ‘traduire’, on l’a vu, un nom sémitique théophore, le père, Aurelius pouvant être lui aussi un habitant de la région devenu citoyen romain en 212 ou descendant de quelqu’un qui l’est devenu à cette date; on sait aussi qu’ on pourrait fort bien trouver en safaïtique, donc chez les pasteurs de cette steppe, un zndrs ʾrls (Zénodoros Aurélios) – encore que la conjonction d’ un nom propre et d’un patronyme grecs transcrits relèverait d’une forte rareté. Certes, mais l’ origine régionale d’un Zénodoros fils d’Aurelius peut se trouver n’importe où dans l’empire romain, partie orientale en tout cas. Le nom de l’ auteur de l’ inscription no 1 n’indique pas, à la différence de celui de l’ inscription no 2, une origine locale (la steppe) ni même régionale (le Ḥawrān ou le Proche-Orient). Cette différence va de pair avec deux autres. – Une différence de disposition. L’inscription no 1, sur un support vertical, est composée en quatre lignes relativement parallèles et de même longueur, dont la ligne supérieure est judicieusement calée sur le bord supérieur du rocher, et elle est cadrée latéralement par de larges marges laissées vacantes sur le rocher (cadrage soigneux, voulu: il a obligé le graveur à composer avec la marque plus ancienne, en forme de fer à cheval, que nous avons signalée): nous ne sommes pas très loin de la disposition que pourrait avoir une inscription gravée par un lapicide dans un cartouche rectangulaire. L’inscription no 2, à la façon d’une très grande majorité d’ inscriptions safaïtiques, s’adapte à son support, horizontal et irrégulier, sans souci de géométrie: les lignes ne sont pas rectilignes; elles sont de longueur irrégulière. – Une différence de contenu qui est perceptible. « Tanos », dans l’ inscription no 2, a l’activité pure du berger des inscriptions safaïtiques : il fait paître le troupeau. Ce qu’il écrit en grec est une traduction précise de la formule classique: w rʿy h-ḍʾn, «et il a fait pâturer les moutons »9.

9 Texte partiel d’ une inscription safaïtique découverte par Geraldine King dans le cadre du Black Desert Rescue Survey, en Jordanie du Nord-Est, et mentionnée par Macdonald 1993: 341 sous la référence krs 216–218.

262

villeneuve

Zénodoros quant à lui garde «la tente», ou le « campement ». Est-il un pasteur nomade qui garde une tente ou un groupe de tentes de sa famille, de son clan? C’est possible. Mais il précise qu’ il a « obéi » à une instruction de garder ce campement. L’instruction ne vient pas forcément d’ une autorité familiale ou clanique. Le campement n’est pas nécessairement un campement de nomades. Il faut rendre compte de l’ensemble de ces observations, et notamment de ces différences. L’un de nos deux ‘auteurs’ au moins sait le grec – sans doute pas les deux, pour des raisons statistiques (deux inscriptions grecques sur plusieurs centaines ici; quelques unités ou quelques dizaines d’ inscriptions rupestres grecques dans la ḥarrah sur plus de 20 000): donc l’ un des deux a appris ‘le grec’ à l’autre, ou a écrit en grec à sa place. Lectio facilior : celui qui sait le grec est Zénodoros. Il le sait plutôt bien : il est allé cherché le participe aoriste pour indiquer son obéissance puis l’ indicatif aoriste pour indiquer ce qu’il fait – garder –, maniant donc fort bien la ‘valeur d’aspect’ de l’aoriste grec, qui indique des événements bien délimités dans le temps: j’ai obéi, je me suis mis à garder. Zénodoros et Tanos se comprennent, soit que Tanos sache lui-même un peu de grec, soit que Zénodoros sache en plus du grec parler un dialecte nord-arabique, ou que sa langue maternelle soit proche du nord-arabique: Zénodoros est donc en mesure de traduire en grec ce que Tanos dit qu’il fait – faire paître le troupeau. Il le traduit fort bien aussi, en utilisant le présent de l’indicatif (valeur de durée: le pâturage sur lequel veille Tanos est moins limité dans le temps que la garde qu’ exerce Zénodoros); détail intéressant, il choisit la 3e personne du singulier pour le verbe: il fait paître. Moi je me suis mis à garder, lui fait paître. On voit donc qui est le rédacteur: c’est Zénodoros. Qui est le graveur? Nous n’en savons rien, mais de façon symétrique, dans un échange de services, Tanos, qui a de fortes chances de savoir graver une inscription safaïtique, a probablement montré à Zénodoros comment écrire avec un percuteur lithique. Or Zénodoros choisit de graver, ou de faire graver par Tanos, sur un rocher vertical et en respectant un cartouche rectangulaire théorique, comme on écrirait une inscription en ville, dans un village ou à l’ armée. Ce que ne fait pas Tanos pour graver (ou faire graver par Zénodoros) le constat de sa propre activité. Lectio facilior toujours: Tanos est un pasteur nomade, qui fait paître le troupeau de sa famille, ou de son propriétaire, dans les pâturages du bas-Wādī Shām, et qui s’est perché sur le tell pour bien voir. Zénodoros n’en est pas un. C’est un soldat, d’un corps auxiliaire, ou même d’ un corps de supplétifs comme il en existait parfois. Il a obéi a un ordre: il garde un campement, seul probablement dans cette tâche. Soit c’est une mission de renseignement: il y

inscriptions grecques d’ époque romaine

263

a là ou à proximité un campement de nomades et Zénodoros doit le surveiller. Soit c’est une mission d’exploration: le campement est celui d’ une patrouille militaire, qui se déplace dans la steppe et dont la ou les tentes doivent être gardées en attendant son retour. Cette zone est en effet, dès le ier siècle apr. J.-C. et plus encore à partir du tournant des iiie-ive siècles, une zone de surveillance militaire intensive10 – et qui d’autre surveiller que les nomades, ici? Diyātheh est le point majeur du dispositif, en raison des bons puits d’eau dans le lit du Wādī Shām en cet endroit. Mais de Diyātheh (fig. 5), la vue ne porte pas dans toutes les directions: un relais visuel, la haute tour de surveillance de Marwad (fig. 6), a donc été construit, à une demi-heure de marche au sud (et une demi-heure aussi à l’ ouest de Freikh), pour bien contrôler le bas des pentes du Jabal Ḥawrān au sud – en relation visuelle cette fois avec le fort romain d’ aṣ-Ṣaʿnah. Il est d’ ailleurs tout à fait possible qu’avant la construction de la tour de Marwad, le tell Freikh ait assuré cette fonction de relais visuel pour la surveillance. Dans ce cadre quadri-séculaire (ier-ive siècles) de surveillance des pasteurs par les troupes de princes locaux puis de l’armée romaine, et bien entendu de coexistence et d’échanges, je propose donc, sur la base de nos deux inscriptions, de restituer – évidemment sans certitude, et sans date bien que le iiie siècle ou plutôt le début du ive, une fois le fort de Diyātheh construit, me paraissent les plus probables – ce fragment de micro-histoire. Le soldat Zénodoros et le pasteur Ẓan, peut-être seuls à Freikh avec les bêtes, s’ennuient, se rencontrent, font connaissance et passent le temps11. Tanos (Ẓan) montre à Zénodoros comment graver une inscription à la façon ‘safaïtique’, et Zénodoros montre à Tanos comment on écrit le grec. Ils font des travaux pratiques: les inscriptions expliquent ce que fait chacun des deux hommes. L’absence de généalogie de Tanos, de détails sur son activité, de prière, sans être absolument significatives, me semblent confirmer ce contexte de ‘conversation appliquée’. Ces rencontres dans la steppe ne sont pas toujours aussi interactives: la fameuse inscription grecque du lieu-dit Jathūm, dans le nord-est jordanien, est de nature différente: plus élaborée, gravée et non martelée, elle fait connaître le citharède et le barbier d’ un détachement militaire qui se plaignent d’être dans le désert12. 10

11 12

Villeneuve et Sadler 2001. À Diyātheh, une forte tour de surveillance de plan carré est construite sur le rebord du plateau dominant le wādī, dès le ier siècle apr. J.-C. Vers 300 apr. J.-C., elle est intégrée, comme tour d’ angle sud-ouest, dans un vaste fort à quatre tours d’ angle, plus tours à mi-longueur des côtés, et deux portes. Macdonald 2006. Mowry 1953 ; Schwabe 1954.

264

figure 5

villeneuve

Carte exposant les relations de visibilité entre le fort de Diyātheh, la tour de Marwad et le Tell Freikh f. villeneuve, sophie vatteoni, 2003

inscriptions grecques d’ époque romaine

figure 6

265

Photographie de la tour de surveillance de Marwad, vers le sud-ouest. Noter que seul le limon central de cette tour carrée est resté debout f. villeneuve, mars 1988

Nos deux documents n’infirment pas, tellement ils restent, avec quelques autres13, exceptionnels, la règle d’exclusion réciproque entre l’ usage épigraphique du grec, entièrement hégémonique dans les villages (et villes) du Ḥawrān à l’époque romaine après la disparition du nabatéen et avant l’ apparition de l’arabe, et l’usage épigraphique – toujours rupestre – du « safaïtique », exclusif dans la steppe et présent parfois sur certains terrains de pacage entre les villages. J’ai proposé depuis longtemps ce constat d’ exclusion, et M. Macdonald, en dépit de réticences peut-être marginales, ne me semble pas l’ avoir récusé14. La règle ainsi énoncée, les exceptions, par principe intéressantes, sont de deux types: inscriptions grecques (ou bilingues) dans la steppe, ou inscriptions safaïtiques dans les villages. Le premier type, comme déjà indiqué ci-dessus, est rarissime, surtout quand on en retranche les purs et simples militaires, comme ceux de Jathūm ou ceux du secteur du fort de Namārah, accompagnés de quelques villageois en affaire avec les soldats du fort15. Les publications (voir 13 14

15

Voir ci-dessus, n. 2. Macdonald 1993: passim, en particulier p. 309 : « … the discovery of a handful of Safaitic inscriptions at Umm al-Jimâl or Bosra does not mean that nomads were necessarily present in these places … ». Villeneuve 1989 : 135–138. Waddington 1870 : no 2265–2284.

266

villeneuve

n. 2) qui font croître, très lentement, ce petit lot, donnent des explications diverses, quand il y en a, à la connaissance, ou à la présence du ‘grec’ dans la steppe. Loin d’être quelque chose d’assez normal dont il n’ y aurait pas à s’étonner outre mesure, comme M. Macdonald a semblé plusieurs fois dans son œuvre le penser, cette connaissance est à chaque fois extraordinaire et doit correspondre à des histoires individuelles sortant de l’ ordinaire. Ce que je viens de présenter peut correspondre à un schéma qui aurait fonctionné dans plusieurs cas, et non uniquement cette fois-ci à Freikh – expliquant alors une partie des bilingues ‘usuelles’ composées de simples noms. Quant aux éventuelles inscriptions safaïtiques dans les villages, la contribution de H. Zeinaddin16 sur une petite sélection de ses découvertes d’ inscriptions safaïtiques dans la montagne du Ḥawrān, donc dans la zone des villages – contribution très importante et malheureusement déparée par de très nombreuses erreurs à l’impression des translitérations –, a pu un temps faire hésiter M. Macdonald17 sur l’exclusion réciproque. Une carte, les listes topographiques et certaines expressions de cet article laissaient penser que nombre de ces inscriptions safaïtiques avaient été trouvées dans les villages, « im Areal der Dörfer» (expression en elle-même ambiguë). Une lecture attentive montre que pour quatorze des dix-huit inscriptions publiées il n’en est rien : tantôt trouvées le long des wadis, tantôt à plusieurs centaines de mètres ou à des kilomètres à l’extérieur des villages. En revanche, les inscriptions no 4 à 7 ont été trouvées sur des éminences rocheuses auxquelles sont accrochés deux villages druzes voisins, respectivement Ḥarīsah (no 4) et Shaʿaf (no 5–7). Mais il est tout à fait possible que ces textes safaïtiques soient antérieurs à l’ installation villageoise antique qui, en limite orientale de la montagne du Ḥawrān, est souvent tardive dans l’époque romaine. De plus, ces inscriptions des tells de Shaʿaf et de Ḥarīsah sont rupestres, comme le sont toutes les autres; elles n’ont donc rien de commun avec les inscriptions villageoises – en grec – qui sont sur des blocs d’architecture. Elles doivent par hypothèse logique être jugées l’ expression de pasteurs ou chasseurs présents de façon temporaire sur des terroirs villageois, ou en rase campagne avant la construction des villages. C’ est bien ce qu’indique le contenu de tous ces textes, avec la possible exception du no 6, un graffite trouvé sur un rocher de la colline volcanique où est accroché le village druze de Shaʿaf18. L’éditeur le publie ainsi : l ṭʿʾl ḏ ʾl s2ʿf, et le traduit : « Von Ṭāʿil, der von den Einwohnern von Šaʿaf ist».

16 17 18

Zeinaddin 2000. M. Macdonald, communications orales et écrites, 2002–2005. Zeinaddin 2000 : 276, no 6.

inscriptions grecques d’ époque romaine

267

La signification de ʾl comme «habitants» (d’un village) me semble cependant problématique. H. Zeinaddin traduit ailleurs ʾl de façon constante19 et classique par «Stamm», donc «lignage». Un lignage de nomades s2ʿf seraitil à exclure? Au demeurant, l’existence d’un village antique à Shaʿaf n’ est pas documentée avec pleine certitude20. Il pourrait donc s’ agir d’ une fausse exception. De la même façon, à Diyātheh, origine de notre enquête ici présentée, notre découverte, sur une pierre d’une maison du village d’ époque romaine, d’ un graffite safaïtique – anthroponyme plus patronyme21 – pourrait faire figure d’ exception. Il n’en est pourtant rien, au contraire: le graffite figure sur le lit d’ attente d’un corbeau, donc à un emplacement situé en hauteur et recouvert par une poutre en basalte du plafond. Il ne pouvait donc pas être lu, ni même vu. Le corbeau est une pierre en partie brute et pour le reste à peine dégrossie: le graffite a été gravé sur le roc dont il a été tiré, ou sur pierre errante, comme il arrive souvent aussi, avant qu’elle soit extraite, ou ramassée, pour la construction de cette maison. J’ai daté l’édification de la centaine de maisons du village de Diyātheh des iie et iiie siècles apr. J.-C.22: le graffite safaïtique est fort pro19 20

21

22

Dès l’ inscription suivante, notamment, pourtant trouvée sur la même éminence rocheuse : Zeinaddin 2000 : 277, no 7. M. Sartre (communication écrite, 13/07/2016) m’indique que seules quatre inscriptions grecques ont été retrouvées à Shaʿaf – par lui-même uniquement, à partir de 1988 seulement. Il s’ agit des documents qui seront publiés dans les Inscriptions grecques et latines de la Syrie, xvi, no 932–935. Sur ces quatre textes, deux proviennent de façon certaine d’un autre lieu que Shaʿaf, village vers où ils ont été déplacés. On peut hasarder l’hypothèse qu’il en soit de même pour les deux autres, d’ autant que les vestiges archéologiques semblent faire défaut à Shaʿaf, qui pourrait n’avoir jamais été un village antique (la seule indication contraire proviendrait de Burckhardt 1822, p. 94, mentionnant en 1810 un «Tel Shaaf» avec une “ruined city” (sic !)). Au contraire, M. Sartre a eu connaissance de la découverte par H. Zeinaddin de soixante-douze ( !) inscriptions safaïtiques sur la colline rocheuse de Shaʿaf. On peut donc faire soit l’ hypothèse d’ un lieu non bâti dans l’Antiquité, et largement fréquenté par les pasteurs, comme souvent pour les éminences; soit celle, que propose M. Sartre, d’ un village ‘tardif’ dans la période romaine, s’installant en un lieu précédemment fréquenté par les pasteurs. Rien ne peut être tranché, d’autant que H. Zeinaddin aurait également découvert là une inscription nabatéenne, dont ni la date, ni le contenu, ni la nature du support ne me sont connus. Ce graffite encore inédit a été étudié par M. Macdonald: communication écrite personnelle du 28/05/2004. Le savant y rappelle que Dussaud et Macler 1903 avaient publié (no 821) une copie « chaotic » d’ un autre graffite safaïtique de Diyātheh, présentant exactement les mêmes caractéristiques de texte écrit « on the non-dressed face of a corbel in a ruined house ». Villeneuve et Sadler 2001. La fréquentation du site de Diyātheh par des pasteurs, laissant

268

villeneuve

bablement antérieur (ier siècle av. ou apr. J.-C.) à l’ apparition d’ un village à Diyātheh, à l’époque où les pasteurs de langue nord-arabique fréquentaient cette zone de puits le long du Wādī Shām. Je ne suis pas certain, en revanche, que les deux documents publiés ici contribuent à étayer la séduisante thèse, souvent exposée23 puis peaufinée, par M. Macdonald, d’une écriture passe-temps, qui ne serait presque que dessin, dans une société de l’oralité qui, si je suis bien notre savant ami, ne serait pas véritablement alphabétisée. Les essais d’apprentissages linguistiques et graphiques, croisés selon moi, dont témoignent nos deux petits textes, pourraient montrer le contraire.

Bibliographie Bader, N. 1997. «Inscriptions grecques de Wadi Salma dans la région désertique du nord-est de la Jordanie». Le Muséon 110: 459–466. Bader, N. 2009. Inscriptions grecques et latines de la Syrie, t. xxi, Inscriptions de la Jordanie, t. 5/1, La Jordanie du Nord-Est. Beyrouth: Institut français du Proche-Orient. Burckhardt, J.L. 1882. Travels in Syria and the Holy Land. London: John Murray. Dussaud, R. et Fr. Macler. 1903. Mission dans les régions désertiques de la Syrie moyenne. Paris: Imprimerie nationale-Leroux. Al-Jallad, A. et A. al-Manaser. 2015. «New Epigraphica from Jordan i: A pre-Islamic Arabic Inscription in Greek Letters and a Greek Inscription from North-Eastern Jordan». Arabian Epigraphic Notes 1: 51–70. Al-Jallad, A. et A. al-Manaser. 2016. «New Epigraphica from Jordan ii: Three SafaiticGreek Partial Bilingual Inscriptions». Arabian Epigraphic Notes 2: 55–66. Macdonald, M.C.A. 1993. «Nomads and the Ḥawrān in the Late Hellenistic and Roman Periods: A Reassessment of the Epigraphic Evidence». Syria 70: 303–413. Macdonald, M.C.A. 2006. «Literacy in an Oral Environment». Dans Writing and Ancient Near Eastern Society, édité par P. Bienkowski, C. Mee et E. Slater, 49–118. New YorkLondon: t & t Clark. Macdonald, M., M. al-Muʾazzin et L. Nehmé. 1996. «Les inscriptions safaïtiques de Syrie, cent quarante ans après leur découverte». Comptes rendus de l’Académie des inscriptions et belles-lettres : 435–494.

23

quelques graffites dont celui présent sur la pierre utilisée ensuite en corbeau, peut avoir coexisté avec la forte tour de surveillance isolée, construite dès le ier siècle apr. J.-C. selon notre proposition, devenue vers 300 apr. J.-C. la tour sud-ouest du fort romain d’époque tétrarchique. Déjà en conclusion de Macdonald 1993 : 388.

inscriptions grecques d’ époque romaine

269

Mowry, L. 1953. «A Greek Inscription at Jathum in Transjordan». Bulletin of the American Schools of Oriental Research 132: 34–41. Schwabe, M. 1954. «Note on the Jathum Inscription». Bulletin of the American Schools of Oriental Research 135: 38. Thuillier, M. 2010. «Les inscriptions rupestres grecques et latines au Proche-Orient aux époques hellénistique et romaine». Mémoire de master 1 (inédit), Université Paris 1. Villeneuve, F. 1989. «Citadins, villageois et nomades. Le cas de la provincia Arabia (iieive siècles ap. J.-C.)». Dialogues d’histoire ancienne 15 (1): 119–140. Villeneuve, F. et S. Sadler 2001. «Occupation du sol et vestiges architecturaux sur les marges arides de Syrie du Sud. L’exemple de Diyatheh». Dans Conquête de la steppe et appropriation des terres sur les marges arides du Croissant fertile, édité par B. Geyer, 159–187. (Travaux de la maison de l’Orient 36). Lyon: Maison de l’Orient. Waddington, W.H. 1870. Inscriptions grecques et latines de la Syrie recueillies et expliquées. Paris: Firmin Didot. Zeinaddin, H. 2000. «Safaitische Inschriften aus dem Ǧabal al-ʿArab». Damaszener Mitteilungen 12: 265–289, pl. 58–69. Zeinaddin, H. 2008. «Die Inschriften von al-Namara». Dans Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, edité par K. Bartl et A. Moaz, 333–337. (Orient-Archäologie 24). Rahden: Leidorf.

chapter 13

Méharistes et cavaliers romains dans le désert jordanien Pierre-Louis Gatier*

Les innombrables graffites, souvent accompagnés de dessins, qui ont été tracés sur les rochers dans les sites des zones arides du Proche-Orient – particulièrement dans la ḥarrah couverte de roches basaltiques qui s’ étend depuis le sud de la Syrie jusqu’au nord de l’Arabie Séoudite, mais aussi dans les régions voisines – sont reconnus désormais comme la principale source d’ information sur le monde des nomades de l’Antiquité et ils participent à l’ ensemble de la documentation sur l’histoire du Proche-Orient antique. Nous le devons largement à Michael Macdonald. Il a su donner une impulsion décisive à leur étude, bien entendu avec les solides bases linguistiques et philologiques héritées de ses prédécesseurs, et aussi avec le regard critique nécessaire pour en assurer la valeur de documents historiques. Les différentes langues et écritures utilisées pour ces graffites, sémitiques dans leur écrasante majorité, reflètent les pratiques et les connaissances des populations nomades. Néanmoins, le grec et le latin, langues du pouvoir et langues de communication, réussissent parfois à se glisser très minoritairement parmi ces textes gravés sur des galets mobiles ou des parois rocheuses. Il faudra un jour faire l’ inventaire de ces rares graffites grecs ou latins du désert, dont Michael Macdonald a parfois rencontré et publié quelques exemplaires1, mais ici, en hommage d’ amitié, je lui offrirai un de ces textes inhabituels, qui a l’intérêt d’être inédit et donc d’ apporter quelque nouveauté, en dépit des difficultés de lecture inhérentes à ce type de document. Il s’agit d’un galet ou plutôt d’un bloc qui provient de Bāyir, dans le désert jordanien.

* J’ exprime ma reconnaissance aux savants confrères qui m’ont apporté généreusement leur aide : Julien Aliquot pour la lecture du graffite et pour la carte, Denis Genequand pour les questions concernant les sites de Maʿān et de Bāyir, Burton MacDonald et David Kennedy pour celui de Qaṣr al-Abyaḍ, Jean-Baptiste Yon pour l’onomastique et Maurice Sartre pour m’ avoir donné accès à son article sur Namārah avant sa publication. 1 Entre autres, Macdonald et Searight 1982 : 167 ; Macdonald et al. 1996: 464; 480–487.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

méharistes et cavaliers romains dans le désert jordanien

i

271

Le site de Bāyir

Le site de Bāyir, dont le toponyme évoque les puits, se trouve un peu à l’ est de la route actuelle qui, du nord-est au sud-ouest, joint al-Azraq à Maʿān, à michemin entre ces deux villes, à environ 120km de chacune d’ entre elles. À vol d’ oiseau, une centaine de kilomètres séparent Bāyir du gros bourg de Ṭafīlah vers l’ouest et du grand axe Nord/Sud que constitue, à l’ est, le Wādī Sirḥān, ce long corridor qui relie l’oasis d’al-Jawf à Azraq (fig. 1). Le lieu est situé sur la bordure orientale calcaire du plateau jordanien parcourue par des vallées sèches qui filent vers le Wādī Sirḥān, dont celle du Wādī Bāyir, orientée vers le nord-est, et celle du Wādī al-Siʾr. Bāyir, dans une zone de croupes arides, est connu du monde savant par les brefs séjours de quelques voyageurs et par les prospections menées par plusieurs équipes archéologiques. Le site attend toujours d’être fouillé. Avant 1918, quelques voyageurs intrépides et autres aventuriers y font étape. Burckhardt connaît la bonne source de «Byr Bair » où il passe en venant de « Kalaat el Hassa», très probablement le site de l’actuel al-Ḥasā, pour se rendre dans le Wādī Sirḥān2. Carruthers semble le plus ancien voyageur européen à avoir signalé les restes du château, qu’il a vu en 1909 et qu’ il considère comme un caravansérail routier3. En effet, ce qu’en écrit Moritz, un peu avant, n’est pas le résultat d’une observation personnelle, mais semble plutôt dépendre d’ informations inédites de Georg August Wallin4. Alois Musil a beaucoup parcouru la région entre 1896 et 1902, puis dans les années 1906–1915, et il visite en 1909 le site qu’il nomme «Bâjer», qu’il considère lui aussi comme un khān5. Quelques années plus tard, Gertrude Bell, partie à dos de chameau de Zīzā au sud d’Amman le 7 janvier 1914, en route vers le sud-est avec une petite escorte, arrive le 20 janvier en début d’après-midi à Bāyir, qu’ elle nomme Bir Baʿir et Qasr Baʿir6. Elle écrit que tous les rezzous (« ghazzus ») passent par le lieu, du fait des puits qui ont de l’eau jusqu’en été, et qu’ elle n’a jamais autant entendu de récits de ces raids. Elle photographie les restes du château, qu’ elle juge omeyyade, et en dessine le plan, avant de repartir le 21 janvier dans la journée. Ses photos, inédites pour la plupart, sont disponibles dans ses

2 3 4 5 6

Burckhardt 1822 : 665. Carruthers 1910 : 243–244. Moritz 1908 : 427–428. Musil 1926 : 25 ; idem 1927 : 324–325 ; voir aussi Musil 1928. Bell 1927 : 327–331 ; eadem 2000 : 48–51.

272

figure 1

gatier

Carte de situation j. aliquot

méharistes et cavaliers romains dans le désert jordanien

273

archives, à l’université de Newcastle7. Le fameux T.E. Lawrence, dit « d’ Arabie », mentionne «Bair», où il se rend en juin 1917 en provenance du Wādī Sirḥān avec une troupe de cinq cents hommes montés sur des chameaux, au cours des opérations de la «Révolte arabe»8. Les installations des quatre puits y ont été dynamitées par les Turcs, mais la remise en service de deux d’ entre eux permet d’abreuver les animaux pendant une semaine. Par la suite, Bāyir sert aux insurgés arabes de base de départ et de repli pour plusieurs raids. Les observations de Lawrence restent celles d’un visiteur de passage; elles le conduisent néanmoins à attribuer le château aux « Ghassanides » (désormais « Jafnides»). À l’époque du Mandat britannique, une première prospection systématique s’ organise, sous la direction d’Henry Field, qui passe avec son équipe à Bāyir en novembre 1927 et 1928, particulièrement à la recherche de silex taillés préhistoriques. Les photos et le plan, qui seront publiés longtemps après, sont les derniers réalisés avant la destruction du château9. Une grande citerne double, avec ses canaux d’approvisionnement, est alors reconnue près du Wādī Bāyir. Par la suite, en décembre 1932, Nelson Glueck fait une brève visite sur le site en compagnie de George et Agnes Horsfield. Il ne reste alors plus grand chose du château. En effet, un fort a désormais été construit pour la Légion Arabe sur une colline, à 500m du château ancien, dont les vestiges ont été presque entièrement détruits pour en récupérer les blocs. Glueck et les époux Horsfield examinent néanmoins en détail le site, dessinent un plan de ce qui reste du château et opèrent le premier ramassage de céramique10. Enfin, en mars 1939, le pionnier de l’archéologie aérienne, Sir Aurel Stein, fait – par la route de terre – un court passage à Bāyir, en cherchant à se rendre à Kilwah, et il en profite surtout pour examiner les installations hydrauliques. Ses observations et les photos prises par la raf à cette occasion n’ont été publiées qu’ assez récemment par David Kennedy11.

7

8

9 10 11

Deux des photos de Gertrude Bell ont été publiées, l’une par David Hogarth (1927, entre les p. 4 et 5) ; l’ autre par Denis Genequand (2012 : 405). Voir l’album y, de y-183 à y-195, sur le site web de l’ Université de Newcastle (http://www.gerty.ncl.ac.uk/). Lawrence 1967: 288–291 ; voir aussi p. 412–413, 460, 574. De même, le 15 août 1918, deux des puits abreuvent à la fois un millier de réfugiés et une troupe montée sur six cents chameaux ; voir p. 578–580 et 592–593. Field 1960. Voir aussi, dans le même ouvrage, les contributions de Florence Day et d’Eric Schroeder. Glueck 1933 : 381 ; idem 1934 (Explorations i) : 72–75 ; idem 1940: 42–43. Kennedy 1982: 255–258 ; voir p. 206, 209–210 et 294. Même texte, Stein 1985: 287–290, voir p. 279, 422–423.

274

gatier

À partir des années 1980, des missions plus longues permettent d’ approfondir les recherches. La prospection conduite en 1981 par Scott Rolston et Gary Rollefson, bien qu’orientée vers la préhistoire, est accompagnée par une collecte de céramique sur le site du château et par une étude de graffites nordarabiques par Vincent Clark12. La même année, l’ équipe réunie autour de Geoffrey King pour prospecter les sites byzantins et islamiques de Jordanie s’ est rendue à Bāyir et y a collecté de la céramique13. L’Université du Yarmouk à Irbid (Département d’ épigraphie de l’ Institut d’archéologie et d’anthropologie) mène en 1988, 1989, 1990, puis 1993, avec pour les deux dernières campagnes la collaboration de l’ ismeo de Rome, une prospection épigraphique et archéologique dirigée par Fawwaz al-Khraysheh et Jacqueline Gysens Calzini14. Ces recherches élargissent la perception de l’ensemble du site de Bāyir en étendant largement la zone étudiée autour du château et des trois puits voisins (site 13) et en découvrant une série de sites secondaires. Certains (sites 1 à 12) livrent seulement des graffites, mais deux autres (sites 14 et 15) révèlent des restes de bâtiments ou de carrières d’extraction des pierres. Le site 14, à environ 5 km au nord du château, comporte trois constructions rectangulaires inconnues jusqu’ alors et placées sur une éminence, à proximité de la grande citerne repérée déjà par Henry Field et Aurel Stein. Les trois constructions sont des podiums rectangulaires (26×6,80m; 20×6,85m; 16,80×11,30m), parallèles entre eux. Jacqueline Gysens Calzini les décrit comme des sortes de restes de chapelles ou de temples, ce que la découverte de fragments de tuiles de terre cuite et celle d’ un graffite nord-arabique invoquant une déesse sembleraient conforter. Le site 15 est une carrière déjà repérée par Henry Field. Une collecte de tessons entreprise par l’équipe n’a pas été suivie de la publication annoncée. On peut ajouter à cette description générale du site la présence de tombes anciennes ou récentes. Plusieurs auteurs mentionnent le discret cimetière des bédouins modernes. Mais certaines sépultures, réelles ou légendaires, sortent du commun. Burckhardt signale la tombe du fils du Sultan Ḥassan15. On montre à Gertrude Bell un grand tombeau de 4m sur 4, qui est attribué soit à un héros de la geste hillalienne, un certain Maraʾi « qui sortit du Nejd avec les Bani Hallal […] et conquit la Tunisie», soit à un adversaire arabe des Turcs en 1852 à Ḥudaydah16. L’expédition Field décrit dans le lit du Wādī Bāyir 12 13 14 15 16

Rolston et Rollefson 1982 : voir p. 213 ; Clark 1987 : 185–191. King et al. 1983 : 398–399 et 420 ; King 1987 : 91–105. Calzini 1993 ; Gysens et Ruffo 1995 ; Gysens Calzini et al-Khraysheh 1995. Burckhardt 1822 : 665. Bell 2000 : 49–51.

méharistes et cavaliers romains dans le désert jordanien

275

le même tombeau monumental construit en pierres, sans se prononcer sur son occupant, construction dont il n’est plus question dans les écrits des visiteurs suivants17. En revanche, Lawrence est conduit par le chef des bédouins Ḥuwaytāt sur la tombe de son fils ʿAnnad, tué dans un conflit tribal, et cet autre monument funéraire paraît être resté un objet de vénération jusqu’ à nos jours18. Pour revenir aux vestiges les plus importants, les ruines du château sont datées par les différents auteurs en fonction de trois critères, les sources textuelles, l’architecture ou la céramique: – La mention dans les sources arabes19, en préférant soit les textes qui se rapportent au jafnide («ghassanide») al-Ḥārith ibn Jabalah, soit ceux qui sont relatifs à l’omeyyade al-Walīd ii, conduit à choisir la fin de la période byzantine d’un côté – avec Musil, Moritz, Lawrence et Stein – ou le début de l’époque islamique de l’autre, avec Lammens20. – L’observation de l’architecture des vestiges conservés jadis, ou des photos antérieures à 1928, et l’examen des plans du bâtiment21 mènent à dater le bâtiment de la période omeyyade et à le ranger dans la catégorie des «châteaux du désert», qui sont vus de nos jours comme les résidences des élites de l’époque. Bell, Schroeder et Day ont pris ce parti, suivis par Creswell, Sauvaget et Genequand22. Ce dernier considère que le château est resté inachevé, comme Qaṣr aṭ-Ṭūbah ou al-Mushattā, et que cela explique l’absence de céramique omeyyade. À une époque où l’ on reconnaissait moins bien qu’aujourd’hui le caractère tardo-antique des constructions omeyyades, certains auteurs, comme Musil, Lawrence et Stein, attribuaient aux «Ghassanides» nombre des «châteaux du désert », dont Bāyir.

17 18 19

20 21

22

Field 1960 : 78. Lawrence 1967 : 289–290 ; Gysens Calzini et al-Khraysheh: 355. Musil 1927: 324–325, n. 76, a fait le bilan de ces sources et en propose des équivalences entre des toponymes qu’ il écrit Bâjer ou Ḳaṣr Bâjer (Bāyir), Ubajr, Ubâjer, Abâʾir et Wubajr; voir aussi Musil 1928: 286–287. Les auteurs modernes renvoient principalement à Abū ʾlFaraǧ al-Iṣfahānī, Kitāb al-Aġānī, ii, 108, et xi, 87, non vidi, où le site est nommé Abāiʾr. Lammens 1910 : 102, n. 7, et 109, n. 2. Trois plans ont été publiés : par Glueck (1934, Explorations i: 73, pl. 19), dessiné par G. Horsfield, et qualifié de « ridiculous » par Creswell; par un membre de l’équipe de Field, E. Schroeder (Schroeder 1960: 98, fig. 29) ; par Gysens Calzini et al-Khraysheh 1995: 32. Seul celui de Schroeder est antérieur à la destruction de l’essentiel des restes du château. On retrouvera peut-être un jour le plan de G. Bell. Creswell 1932–1940 : vol. 1, 642–643, plan p. 642 ; Sauvaget 1939: 39–40.

276

gatier

– L’étude des quatre séries de tessons, collectés sans être publiés en détail (Glueck23, Rolston et Rollefson, King, Gysens Calzini) et décrits comme appartenant massivement à la céramique peinte nabatéenne, pousse à y voir un fort nabatéen; c’est le parti de Glueck (qui pense cependant que le fort nabatéen aurait été repris par le château omeyyade), puis de Rolston et Rollefson (avec deux collectes où la céramique nabatéenne constitue 95 % et 100% des tessons), dont les conclusions sont suivies par Parker24. Là encore, les connaissances ont évolué et l’on sait que la céramique peinte fine de tradition nabatéenne se poursuit jusqu’au iiie siècle et même, sous des apparences plus grossières, plus tard. King, qui a effectué des ramassages de tessons, souligne d’ailleurs la présence importante de la céramique « Late Roman / Early Byzantine». Il propose une occupation continue depuis les Nabatéens jusqu’au ive siècle – qui lui semble la date la plus récente possible pour le château – en récusant toute idée d’occupation omeyyade25. Jacqueline Gysens Calzini présente brièvement la céramique qui a été collectée en deux points, près du château et près des trois podiums, et elle la considère comme essentiellement nabatéenne. Elle compare les tours semicirculaires du long côté le moins détruit du château avec leurs fondations rectangulaires et en conclut que le bâtiment avait connu deux périodes de constructions et plusieurs niveaux d’occupation26. Pour ma part, je n’ai pas pu me rendre à Bāyir et mes informations proviennent de la documentation publiée.

ii

Un graffite grec

Le regretté Fawwaz al-Khraysheh, épigraphiste spécialiste des graffites nordarabiques, dont je salue ici la mémoire, avait conduit dans la région de Bāyir les missions épigraphiques évoquées ci-dessus. Il s’ était appliqué à protéger certaines des pierres inscrites provenant de ce secteur en les déposant dans les musées ou dans les entrepôts des Antiquités. Devenu directeur du Département des Antiquités de Jordanie, il m’avait fait l’ honneur de me proposer d’étudier le bloc dont il sera question ici (fig. 2) et de le publier dans le rap-

23 24 25 26

Néanmoins, Glueck a publié des photos de céramiques. Parker 1986 : 116. King et al. 1983 : 398–399 et 420 (liste de la céramique par période). Gysens Calzini et Ruffo 1995 : 33.

méharistes et cavaliers romains dans le désert jordanien

figure 2

277

Le bloc inscrit

port final qu’il préparait, mais qu’il n’a pas pu achever. Le 4 novembre 2004, il avait fait transporter la pierre à l’Ifpo d’Amman, où je résidais, pour me faciliter généreusement son étude. Je m’acquitte donc, avec quelque retard, de cette dette scientifique, mais l’emplacement actuel du bloc inscrit, qui avait été retourné au Département, m’est inconnu. Ce graffite n’est pas le premier texte grec de Bāyir. Les recherches épigraphiques conduites avec Jacqueline Gysens Calzini avaient d’ abord permis de retrouver quatre graffites grecs et un latin, parmi d’ autres textes dans des langues et écritures nord-arabiques ou en nabatéen. Par la suite, J. Gysens Cal-

278

gatier

figure 3

Le sommet du bloc

zini avait publié ces premiers textes grecs et latins27. Les graffites grecs, de simples noms, mal établis hormis Λούκιος, n’apportent guère d’ informations, mais il est évident que le texte latin nomme des militaires. En 1993, divers graffites sémitiques nouveaux s’étaient ajoutés, ainsi que cinq courts textes grecs – tous restés inédits à ce jour – et un sixième texte grec décrit comme long de treize lignes28. Ce dernier est manifestement celui qui figure sur le bloc que Fawwaz al-Khraysheh m’a confié, bien qu’ il me semble comporter quinze lignes plutôt que treize. Il n’est pas possible de clairement l’ attribuer à l’ un ou l’autre des sites de graffites cartographiés dans les rapports de Jacqueline Gysens Calzini. Il s’agit d’un gros galet plat en calcaire gréseux jaunâtre, ou plutôt d’ un fragment de la dalle de rocher naturelle, de forme irrégulière, dont la partie supérieure, en forme de cornes, est légèrement bombée en surface (fig. 3–4). La seule cassure récente se trouve sur l’ensemble de la bordure de droite, sans dommages pour le texte. Un trait horizontal sépare la première ligne du reste du 27

28

Gysens Calzini 1993 : 107–111, publie quatre blocs, mais le bloc no 3 porte deux mots en grec, considérés comme deux graffites. Le bloc 4, en latin, mal déchiffré, n’est illustré que par un dessin ; on y reconnaît les noms L. Attili[us] Iulianus, Sabinus et peut-être [Fl]auin[us], mais il y a d’ autres possibilités (Gauinus, Octauinus, etc.). Gysens Calzini et Ruffo 1995 : 360.

méharistes et cavaliers romains dans le désert jordanien

figure 4

279

Le bas du bloc

texte. Dimensions: 46×40×10cm. Lettres relativement effacées, de dimensions variables: 1–3,5cm. Ἄλε Γετούλων̣. Μ̣νησθῇ Φλάις Βηρύλλιο̣[ς] ἐκκύης ἄλα δρομεδαρίων Ἀντ(ωνινιανή/άνη/ιάνη?) 4 κ̣ [αὶ] Φλάις Μάξιμος τόρμα Δημητρί[ου καὶ] μνησθῇ κον̣τβ̣ ερνάλιος καὶ Γερμανὸς καὶ Ασμαλεθ̣η̣[ς] καὶ Αβειβα καὶ Μα̣λχο̣ς ̣ 8 καὶ [ca 5 l.] Φλ[άις] καὶ [ca 7 l.] 9–11 [--- 3 lignes ---] 12 καὶ [---] καὶ [---] καὶ μνησ̣ θ̣ῇ [---] [---]. Aile des Gétules. Qu’on se souvienne de Flavius Bèrullios, eques de l’ ala Ant(oniniana/ana/iana?) des méharistes, et de Flavius Maximos de la turme de Dèmètrios, et qu’on se souvienne de son camarade (= contuber-

280

gatier

nalis), et de Germanos, et d’Asmaléthès ( ?), et d’ Abeiba, et de Malchos et … Flavius et … et … et … et qu’on se souvienne … Notes critiques: epsilon et sigma lunaires; lettres triangulaires dont la diagonale de droite se prolonge vers le haut et dont la barre est penchée vers la gauche; phi en arbalète. Malheureusement, je n’ai pas pu déchiffrer les sept dernières lignes, très finement gravées. Je ne signale pas les quelques lettres isolées que je crois distinguer à droite, aux lignes 9 à 15, très peu lisibles. À la ligne 2, je pointe un omicron qui pourrait également être lu comme un sigma devant une épaufrure. Les deux dernières lettres de la ligne 3 sont en ligature. On a gravé une liste de noms précédés de μνησθῇ, « qu’ on se souvienne d’ untel», «qu’il soit fait mémoire d’untel». La formule est habituelle dans des lieux de passage ou parfois de séjour, comme des sanctuaires ou des points particuliers du désert, et dans les cas les plus fréquents, comme ici, elle n’a aucun caractère funéraire. Le bloc n’a d’ailleurs rien d’ une stèle ou d’ un quelconque monument funéraire. La liste de Bāyir entre dans une petite série d’ autres graffites de militaires romains, comme ceux de Namārah en Syrie29, de Qaṣr alAbyaḍ en Jordanie30, et de deux sites d’Arabie Séoudite de la région de Hégra: le Jabal Ithlib et Qubūr al-Jundī, distants de 7km l’ un de l’ autre31. Les cinq graffites grecs du Jabal Ithlib paraissent tous concerner des soldats de l’ aile des Gétules, alors que les douze graffites grecs et le graffite latin de Qubūr al-Jundī ne paraissent nommer que des membres d’une aile de dromedarii. Pour la bonne compréhension du texte de Bāyir, il me semble qu’ il ne faut pas ponctuer avant ἄλα et qu’il vaut mieux considérer que ce mot n’est pas décliné mais fonctionne comme un génitif en rapport avec le mot précédent. Il ne s’agirait donc pas de deux listes, l’une de l’aile des Gétules avec un seul nom et l’autre de l’aile des dromedarii avec les noms suivants, mais, à mes yeux, de la simple mention de l’aile des Gétules au-dessus de la ligne horizontale, suivie au-dessous par les formules de commémoration qui concernent en premier

29

30 31

Maurice Sartre (2016) fait le bilan de l’ épigraphie grecque et latine de ce site, dont plusieurs textes avaient été publiés auparavant. Pour l’épigraphie sémitique, voir Macdonald 2008 et Zeinaddin 2008. Sartre 1993 : 120–121, no 94, pl. 40. Ci-dessous Sartre 1993: no 94. Textes publiés à partir de relevés de Ch. Huber, J. Euting, A. Jaussen et R. Savignac, par Henri Seyrig en 1941 (Speidel 1977: 703–705; Sartre 1982b: 29–35, la meilleure publication à ce jour). Graf 1988: 192–203, a revu certains des textes de Qubūr al-Jundī (d’où seg 38, no 1663–1668), et il améliore la lecture du seul graffite latin (ae 2009, no 1619). Voir, sur les deux sites, Nehmé 2009 : 44–48.

méharistes et cavaliers romains dans le désert jordanien

281

lieu un membre de l’aile des dromedarii, puis d’autres soldats dont l’ unité n’est pas précisée, mais qui devraient appartenir à ce même régiment de méharistes. Traditionnellement, le mot latin ala est transcrit εἴλη en grec, mais ici on a procédé au plus simple, en translittérant directement le latin32, sous la forme ἄλα, sans la décliner, à la l. 3. À la première ligne, ἄλε, pour ἄλη, est une sorte de forme intermédiaire de nominatif (plutôt qu’ un génitif pour ἄλαε). De même, comme on le rencontre parfois33, le terme latin eques, « cavalier», est directement translittéré, ἐκκύης, plutôt que traduit, ἱπ(π)εύς. Ces traits de langue se retrouvent dans les graffites du Jabal Ithlib dont il sera question cidessous: ἄλε et ἄλα, non déclinées derrière ἐκκύης ou ἐκύης34. En revanche, les autres graffites militaires de cette région, à Qubūr al-Jundī, 7 km plus au sud, mentionnent plus classiquement le mot εἴλη, en le déclinant deux fois, et ils signalent à deux reprises également un ἱπ(π)εύς. La subdivision d’ une aile est le peloton nommé turme, turma, ce qui est transcrit en grec généralement τύρμα, mais τόρμα dans notre graffite de Bāyir, forme peu courante qui se rencontre cependant à Qubūr al-Jundī35. Maurice Sartre a publié, il y a quelques années, dans son corpus épigraphique du sud de la Jordanie, un bloc inscrit qui avait été rupestre, « fragment de calcaire blanchâtre arraché au rocher pour être transporté », et qu’ il avait vu au Musée de Mādabā. La pierre, inscrite en grec, provenait de « Qasr al-Abyad, petit site ruiné près de la naissance du wadi al-Hasa, non loin de l’ endroit où la route du désert traverse ce wadi»36. Le texte était lu : Μνησθῇ Φλάις βειλ λιος δρομεδ[ά]ρις [.]δλλαι[---] Notes critiques: grâce aux deux photos de la planche 40 du corpus, il est possible de corriger la fin de la l. 1 et le début de la l. 2 : Βερ[ύλ]λιος. On peut voir les traces de l’upsilon et du premier lambda à la fin de la première ligne, sous les égratignures de la pierre. En revanche, après le bèta, il semble s’ agir d’ un epsilon, comme Sartre l’a lu, plutôt que d’un èta. Toutefois, la confusion 32

33 34 35 36

Il y a d’ assez nombreux autres exemples d’ usage d’ἄλη, mais à ma connaissance le mot est alors décliné, voir ae 2007, no 1478, en Galatie, ἄλης Φλαό̣[ι]ας ; seg 43, no 956, à Sagalassos, ἄλης Γαιτουλῶν ; seg 45, no 2119, en Égypte, ἄλης Αὐγούστης, etc. Gatier 2007 : 180–182. Sartre 1982b : 30–31, no 1, 3, et 4 ; 1 et 3. Ibidem : 31–32, no 9 et 13 ; 10 et 13 ; 6. Sartre 1993 : no 94.

282

gatier

entre les deux sons et les deux lettres est très courante dans l’ Antiquité tardive, ce qui ne crée pas de difficulté ici. Dans les lettres triangulaires de la l. 3, faut-il rechercher ἄλα Ἀν̣[τ ---] ? Je n’ai pas pu identifier précisément le site de Qaṣr al-Abyaḍ dans la région du Wādī al-Ḥasā. Il se peut qu’il soit désigné sous deux ou plusieurs noms différents. Il pourrait être reconnu dans le site de Rujm al-Abyaḍ, à 20 km au nord de la gare d’al-Ḥasā, ou bien de préférence dans celui de ʿAynah à environ 2km à l’est de l’agglomération d’al-Ḥasā37. Quoi qu’ il en soit, il me semble possible d’identifier ce Flavius Bérullios, δρομεδάριος à Qaṣr alAbyaḍ, au Flavius Bèrullios, ἐκκύης à Bāyir. Si cette proposition convenait, elle montrerait soit que le terme ἐκκύης, eques, peut désigner non seulement un cavalier mais aussi un méhariste, ce qui me semble improbable, soit que δρομεδάριος, dromedarius, s’applique à tous les soldats de l’ aile des dromedarii, quelle que soit leur monture, cheval ou méhari. C’ est cette dernière solution qui me paraît préférable et, par ailleurs, il semble difficile d’ imaginer que le même militaire dans un régiment régulier ait été une fois cavalier, une autre fois méhariste. Ces considérations sont confortées par les graffites grecs de Qubūr al-Jundī, dont il sera question ci-dessous et qui montrent la présence de cavaliers dans une aile de méharistes. Je ferai donc l’hypothèse que le graffite de Bāyir mentionne d’ abord un cavalier spécifiquement désigné comme tel, Flavius Bèrullios, et ensuite des méharistes appartenant à la même unité: Germanos, Asmaléthès, Abeiba, et Malchos. Entre ces deux groupes, le cas de Flavius Maximos n’est pas clair, mais on pourrait privilégier la proximité avec le texte qui précède: ce soldat serait également un cavalier, membre de la turme de Dèmètrios, et donc le camarade de Flavius Bèrullios dans le même peloton du régiment. Les lignes 4 et 5 pourraient donc se comprendre: «Qu’on se souvienne aussi de Flavius Maximos, son camarade (= contubernalis) dans la turme de Dèmètrios ». L’onomastique distingue les deux Flauii, qui seraient des cavaliers si l’ on me suit, des autres soldats, qui seraient des méharistes, mais il est peut-être possible de reconnaître un autre Flavius, ligne 8. Le gentilice Flavius, après avoir été celui des empereurs de la dynastie flavienne dans la seconde moitié du ier siècle (69–96 apr. J.-C.), est – entre autres – celui de Constantin et de ses premiers successeurs. De ce fait, Flavius a un large succès au ive siècle, particulièrement chez les fonctionnaires et les militaires, ce qui fournit un élément

37

Communications personnelles de Burton MacDonald et de David Kennedy.

méharistes et cavaliers romains dans le désert jordanien

283

de datation assez sûr38, même si, évidemment, le nom ne reste pas confiné aux périodes de règne des empereurs qui le portent. Les noms uniques ou cognomina, Maximos, Dèmètrios et Germanos, d’origine latine ou grecque, sont banals et se rencontrent très fréquemment en Arabie, et plus généralement au Proche-Orient39. Asmaléthès pose problème; Ασγαλοθη, à Khirbat as-Samrāʾ, lui ressemble lointainement40; on connaît un Ἀσμαλῆς, à Soknopaiou Nesos en Égypte41. Abeiba, Αβειβα, – bien qu’avec une désinence inhabituelle, puisqu’ on attendrait Αβειβας – et Malchos, Μα̣λχο̣ς,̣ ont chacun une racine sémitique qui a servi à différents anthroponymes. Les Habibas, Habibos et Habiba (fém.) – avec le sens de «cher», «chéri» – sont fréquents dans le Ḥawrān, rares à Palmyre, mais relativement bien attestés dans la cohorte des Palmyréniens à Doura42 ; les noms de type Malchos et Malichos sont très répandus au Proche-Orient en général. Bérullios/Bèrullios est un nom grec rare. Globalement, l’ onomastique a une apparence plus hauranaise, disons sud-syrienne, que palmyrénienne.

iii

Deux régiments de l’armée romaine

L’ intérêt du graffite de Bāyir est de réunir dans un même texte deux corps de troupes qui se trouvent en un même lieu, à une même date, quoique les soldats mentionnés ne semblent appartenir qu’au deuxième régiment nommé, une aile de méharistes. Il est en effet clair qu’il n’y a qu’ une seule graphie du début à la fin de l’inscription. Les soldats n’ont pas écrit leurs noms à la suite, mais un seul homme s’est chargé de graver l’ensemble du texte. Les ailes de méharistes. Puisqu’il ne s’agit pas ici d’ écrire l’ histoire de l’ usage guerrier du chameau dans l’Antiquité, contentons-nous ici de nous intéresser aux méharistes, les dromedarii43. À l’époque impériale, on connaît dans l’armée romaine un seul véritable corps de méharistes, qui est caserné au Proche-Orient, rangé dans les troupes auxiliaires et modelé à l’ image des ailes de cavalerie. Cinq diplômes militaires montrent la présence de cette troupe au deuxième siècle en Arabie et en Syrie. Depuis peu, on possède trois diplômes militaires qui concernent l’Arabie, province qui jusqu’ à récem38 39 40 41 42 43

Par exemple, Gatier 1998 : 406, no 131 (Φλάεις). Sartre 1985 : 214 ; 198 ; 193. Maximos est particulièrement répandu dans l’armée. Gatier 1998 : 369, no 9. P.Lond. 2, p. 159–160, no 322, 12 ; cpr 1, 33, 12. Bader 2009 : 120–121, 284, 289 et 324, no 154–156, 628, 637 et 707. Yon 2012: 398, no 532. P.Dura. 100, 11, 27 ; 101, 27, 15 ; 101, 30, 10 ; 101, 43, 5–6. Dąbrowa 1991 ; Rance 2015 : 125–126.

284

gatier

ment n’apparaissait pas dans ces documents. Datés, l’ un probablement de 126, l’autre de 142, et le troisième de 145, ils témoignent tous de la présence de l’ ala I Ulpia droma(dariorum) Palmyr(enorum) (milliaria) dans cette province44. Le régiment est une aile milliaire, catégorie plus rare et troupe en principe de meilleure qualité que les ailes quingénaires. Ses méharistes ont été recrutés à l’origine à Palmyre, sous Trajan, probablement pour organiser l’ annexion de l’Arabie en 106 et pour servir dans des contrées relativement similaires à leur région d’origine. Deux autres diplômes, qui concernent, cette fois, les troupes de la province de Syrie en 153 et 156–157 apr. J.-C., mentionnent une ala I Ulpia droma(dariorum) (milliaria), qui n’est pas dite « des Palmyréniens», mais qui est probablement le même régiment, transféré d’ une province à l’ autre entre 145 et 15345. On n’a pas assez remarqué que ces diplômes militaires, dont le premier est maintenant celui de 126, sont les plus anciens témoignages de l’ existence du mot latin dromedarius/dromadarius, fabriqué dans l’ armée romaine à partir du grec dromas, «coursier», qui nomme exclusivement le chameau de monte (ou de selle), et du suffixe latin –arius, fréquent pour désigner métiers et fonctions. Le dromedarius est un militaire légèrement armé, monté sur un chameau rapide – à une bosse – pour circuler dans des zones où les cavaliers sont moins à l’aise. Certains auteurs modernes, confondant les motards et les camionneurs, attribuent à tort aux dromedarii la fonction de transporter les fournitures de l’armée sur des chameaux de charge. Le terme de « méhariste», que j’ utilise ici pour traduire dromedarius, ou son équivalent grec δρομεδάρις / δρομεδάριος, est lui-même une fabrication, par les troupes coloniales françaises de jadis en Afrique du Nord, à partir du mot arabe francisé « méhari » qui sert à désigner le chameau de monte. Après 157, on perd les traces de l’ ala I Ulpia droma(dariorum) (milliaria), dont on ne connaît qu’un seul officier, un préfet mentionné dans une inscription de Palmyre, elle-même datée des environs de 15746. L’inscription de Rīmat al-Luḥf, site du Lajā dans la province d’Arabie, où apparaît parfois un officier de méharistes, a paru concerner une troupe stationnée en Égypte, à tort me

44 45

46

Spaul 1994 : 27–28, 104–105, très insuffisant. Weiss et Speidel 2004 (ae 2004, no 1925); Eck et Pangerl 2016. Diplôme de 153, Weiss 2006 : 265–271; ae 2006, no 1841. Diplôme de 156–157, ils 9057. La proposition de Weiss, p. 283, de ne voir dans les mouvements de l’aile d’une province à l’ autre que le déplacement de vexillations ne me semble pas pertinente. Yon 2012 : 196–197, no 202 (= igls, 17/1, no 392), inscription bilingue, grecque et araméenne palmyrénienne où la désignation de l’ aile, complète en grec, est raccourcie et simplement translittérée en araméen : ʾlʾ drmdryʾ.

méharistes et cavaliers romains dans le désert jordanien

285

semble-t-il, et elle ne doit pas être retirée du débat47. Par ailleurs, dans une inscription de Bostra, la capitale de l’Arabie, les equites singulares ou gardes du corps du gouverneur Marcus Caecilius Fuscianus Crepereianus Floranus, associés à des dromedarii, honorent ce personnage vers 19848. Le texte laisse envisager la présence dans la province d’une troupe de méharistes, même s’ il ne permet pas de montrer sa nature à la fin du iie siècle : aile I Ulpia des dromedarii de retour de Syrie? Nouvelle aile? Ou simples méharistes détachés d’ autres troupes? En effet, si l’on ne connaît pas de dromedarii dans les légions, certaines cohortes auxiliaires peuvent comporter des méharistes, sans que le nom et la titulature de ces unités reflètent cette composition. Il s’ agit toujours de cohortes, quingénaires ou milliaires, equitatae et donc composées à la fois de fantassins et de cavaliers, dont certaines comptent également quelques dromedarii. Même si le site précis de leur casernement n’est parfois pas connu, il semble que ces régiments ne se rencontrent qu’ en Égypte et au ProcheOrient49. En tout cas, la documentation disponible à leur sujet repose presque uniquement sur la papyrologie, comme l’illustre le cas de la seule cohorte proche-orientale de cet ensemble, la cohors xx Palmyrenorum, stationnée à

47

48 49

Sartre-Fauriat et Sartre 2014 : 445–446, no 392 (= igls, 15, no 392), fournissent une bonne photo de l’ inscription qu’ ils restituent: Iulius Can|didus uetra|nus ex dup|l(icario) Val(eriae) drum(edariorum), en suivant, parmi leurs prédécesseurs, Waddington 1870: 556, n° 2424. Speidel 1977: 703–704, comprenait la fin du texte: ex dup|l(icario) v al(ae) drum(edariorum), en considérant le v comme un chiffre, ce qui serait une manière inusitée de désigner une troupe. Je lis et restitue: ex dup|lu(cario) al(ae) drumm(edariorum), en remarquant les deux m en ligature à la fin du texte et en notant que l’emploi du u à la place du i est relativement fréquent en épigraphie latine (par exemple optumus pour optimus). Si l’ on me suit, le texte n’a plus rien à voir avec l’ Égypte où était stationnée l’ ala Valeria, mais il montre que la seule désignation d’ « aile des dromedarii » sans précision suffit à faire reconnaître la troupe, parce qu’ il n’y avait alors qu’ un seul régiment de méharistes dans la province (d’ Arabie). Le milieu et la deuxième moitié du iiie siècle, avec le retour en vogue du gentilice Iulius sous l’ empereur Philippe l’ Arabe, conviendraient bien au style de l’ écriture. Sartre 1982a: 143–145, no 9071 (= igls, 13/1, no 9071): eq(uites) sing(ulares) exerc(itus) Araḅ(ici) item droṃ(edarii). Dąbrowa 1991: 365, n. 7, énumère ces cohortes: I Augusta praetoria Lusitanorum equitata, ii Ituraeorum equitata, I Numidarum equitata. La présence de la cohors ii Hemesenorum dans cet ensemble paraît très douteuse, fondée sur deux inscriptions grecques peu déchiffrables trouvées en Palmyrène, qui contiennent la formule μνησθῇ, suivie par un nom de personne, par ce qui est compris comme le terme abrégé δρομε(δάριος) et par des signes qui sont interprétés comme l’ abréviation du nom de la cohorte, voir ae 1952, no 239.

286

gatier

Doura-Europos à la fin du iie siècle et au début du iiie, dont une bonne partie des papyrus retrouvés dans la ville constituent les archives. Deux turmes, ou pelotons, réunissaient un total de trente-six à trente-neuf méharistes dans cette cohorte milliaire xx Palmyrenorum, alors que les cohortes quingénaires se contentaient d’une seule turme de quatorze dromedarii50. Il faut attendre l’Antiquité tardive pour qu’il soit de nouveau question, dans notre documentation, de régiments expressément définis comme de méharistes. Ce sont encore des ailes, mais elles sont cette fois au nombre de quatre. La Notitia dignitatum, dont la section concernant l’ Orient est datée des environs de 400 apr. J.-C., mentionne quatre troupes de méharistes en tout et pour tout: d’une part en Égypte, trois ailes sous l’autorité du dux de Thébaïde, l’ ala tertia dromedariorum à Maximianopolis, l’ ala secunda Herculia dromedariorum à Psinaula (Psinabla) et l’ ala prima Valeria dromedariorum à Precteos (Prectis); d’autre part au Proche-Orient, à Admatha, une seule aile sous le dux de Palestine, l’ ala Antana dromedariorum51. Aucune de ces quatre ailes n’a le qualificatif de milliaire, ce qui montre qu’elles ne sont pas issues de l’ ala I Ulpia droma(dariorum) (milliaria) du Haut-Empire. Thébaïde et Palestine sont les deux seuls duchés à disposer de troupes expressément dites de méharistes. En revanche, les autres duchés, ceux d’Arabie, de Phénicie, d’ Euphratésie, de Mésopotamie et d’Osrhoène, pour s’en tenir au Proche-Orient, et tout le reste de l’Afrique du Nord y compris la Libye en semblent alors dépourvus. Il n’est pas impossible que des méharistes faisaient partie de régiments de cavalerie dans certaines de ces régions, mais, pour l’heure, ils restent invisibles dans la documentation publiée. On doit soigneusement distinguer le site nommé Admatha/Ammatha par la Notitia dignitatum qui, dans l’Antiquité tardive, se trouve dans la province de Palestine Troisième, de celui quasi homonyme d’ Amatha/Amata/Hamatha, dans la province d’Arabie, l’actuel Azraq52. Admatha est considérée comme l’équivalent d’Ammatha, toponyme qui se rencontre dans l’ édit de Bersabée53 – un texte fiscal connu par une inscription grecque du vie siècle apr. J.-C. – où il figure entre Zodocatha, la moderne Ṣadaqah et Aridela, actuelle Gharandal. Il est aussi question dans les papyrus de Pétra, également du vie siècle, du kastron d’Ammatha54. On place le site, dont le nom désigne en araméen des bains chauds, au lieu-dit al-Ḥammām, « le bain », au nord de Maʿān 50 51 52 53 54

Dąbrowa 1991 : 364. Notitia dignitatum, Or. (éd. Seeck 1876) : 31, 48 ; 31, 54; 31, 57; 34, 33. Sartre 1982a : no 9101 ; ae 2001, no 1975 et no 1976 (ae 1987, no 963 et no 964). Di Segni 2004 : 151–152, frag. 2 (Bulletin épigraphique 2005, no 538; seg 54, no 1643). P.Petra 23, 1 ; 25, 4.

méharistes et cavaliers romains dans le désert jordanien

287

en Jordanie, bien que la base archéologique de cette identification ne soit pas très assurée, avec une confusion possible entre un fortin romain et une résidence omeyyade55. Denis Genequand, qui a prospecté le site, montre que ce qui avait été souvent considéré comme la forteresse romaine est un établissement d’ époque omeyyade comprenant trois grands bâtiments carrés résidentiels56. L’ éventuel fort romain aurait pu se trouver sur le tell d’ al-Ḥammām, situé très près d’une des trois habitations omeyyades, mais les vestiges visibles autrefois ont été entièrement détruits à l’époque récente. La qualification de l’aile casernée en Palestine, Antana, que fournit la tradition manuscrite de la Notitia Dignitatum, est inconnue par ailleurs. Elle a semblé corrompue et deux solutions alternatives ont été proposées, Antiana ou bien Antoniniana. Le graffite de Bāyir ne permet pas de trancher et mes restitutions laissent la question ouverte, bien que ma préférence aille à Antoniniana, épithète dynastique qui remonterait aux Sévères. L’adjectif Antiana, qui renverrait à un certain C. Antus, légat de Germanicus, se rencontre dans la titulature de l’ ala Gallorum et Thracum Antiana, connue en Syrie en 54 et 88 apr. J.-C., puis en Syrie-Palestine entre 139 et 186, grâce à des diplômes militaires et à une inscription latine de Scythopolis57. Plusieurs savants ont pensé que l’ ala Antana dromedariorum de la Notitia était l’ ala Gallorum et Thracum Antiana des diplômes58, ce qui me semble peu vraisemblable. Quoi qu’ il en soit, il est question de cette ala Ant(-) dans le graffite de Bāyir, et non pas de l’ ala I Ulpia droma(dariorum) Palmyr(enorum) (milliaria). Il convient désormais de dater ce graffite ainsi que les autres mentions de méharistes que l’ on rencontre en Arabie et en Palestine. Les douze graffites grecs de Qubūr al-Jundī près de Hégra dont il a été question ci-dessus sont de brèves acclamations qui commencent par la formule μνησθῇ, «qu’il soit fait mémoire»59. Ils peuvent tous être rapportés, de même que le graffite latin qui leur est associé, à des méharistes ou à des cavaliers d’ une aile de méharistes, qui n’est pas nommée. Ils comprennent les termes εἴλη et ἱππεύς de préférence à ἄλα et ἐκκύης, comme l’illustre le graffite 13 de la numérotation de Maurice Sartre: Μνησθῇ Οὔλπις Μάγνος ἱππαὶς εἴλης δρομεδάριος. La présence dans cette dernière inscription du gentilice Ulpius, celui de l’ empereur Tra55

56 57 58 59

Parker 1986: 101–102, après d’ autres savants, dont Jaussen, Savignac et Musil; voir le commentaire de Maurice Sartre sur les différentes autres propositions, en annexe à Sartre 1993 : no 94. Genequand 2003 et idem 2012 : 291–295, 304–305, 348–350. Last et Stein 1990 : 224–228 ; ae 1990, no 1013 ; voir Eck 2016: 233–235. Voir Last et Stein 1990 : 226 et n. 15. Sartre 1982b : 30–33.

288

gatier

jan, n’est pas probante en elle-même, mais elle oriente plutôt vers une période ancienne, le iie siècle apr. J.-C., ce que pourrait confirmer le voisinage avec les deux graffites nabatéens 18 et 20 qui désignent des cavaliers et avec le graffite latin 21 dont le sens est proche de celui des formules grecques60. Par ailleurs, le nom unique Severus du graffite grec 9 ne fournit pas d’ élément de datation particulier, parce qu’il est assez répandu, avant comme après la dynastie sévérienne, particulièrement en Arabie où il peut dissimuler un « nom indigène»61. Ces graffites, qui présentent de forts caractères communs, concernent très probablement l’ ala I Ulpia droma(dariorum) Palmyr(enorum) (milliaria), avant son départ en Syrie, même si on ne peut pas totalement exclure qu’ il s’agisse d’une autre aile, qui l’aurait remplacée dans la seconde moitié du iie siècle. D’ autres graffites, à Namārah62, site du Ṣafā, environ 50 km à l’ est du Jabal Druze, ne se trouvent pas sur des parois, comme ceux de Qubūr al-Jundī et du Jabal Ithlib, mais sur des rochers. Ils sont particulièrement nombreux en diverses langues et écritures et voisinent avec le célèbre texte bilingue du tombeau de Imruʾlqays/Maraʾ al-Qays, mort en 332 (ou 328). Les textes grecs et latins se datent approximativement depuis l’ époque sévérienne, au plus tard, jusque dans le courant du ive siècle. Plusieurs particularités rapprochent Namārah de Bāyir: la situation dans le désert près d’ un point d’ eau alimenté toute l’année, la présence de tombeaux anciens et plus récents, les vestiges d’installations construites où il est difficile de reconnaître un fortin et, malgré cela, le rôle militaire. Tous les graffites de Namārah ne relèvent pas de l’épigraphie militaire, mais un grand nombre d’ entre eux appartiennent à cette catégorie, en particulier ceux qui sont en latin, au nombre d’ une dizaine, dont plusieurs mentionnent des légions ou des légionnaires. Quatre graffites grecs peuvent être rapprochés du texte de Bāyir. Deux concernent des méharistes et deux autres nomment des Flauii63. Les troupes des deux dromedarii ne sont pas nommées et rien ne prouve ni n’exclut que ces derniers appartiennent à une aile spécialisée, même si le mot ala figure dans deux textes latins, sans précision. L’aile des Gétules. L’ala Gaetulorum ueterana est connue uniquement au Haut-Empire et elle doit être soigneusement distinguée d’ une autre troupe, 60

61 62 63

Texte corrigé par David Graf (1988: 195–196); voir les photos, p. 208–210, où la datation au iie siècle est confirmée par la forme des lettres et la graphie très régulière des textes grecs, qui sont regroupés dans un même secteur et semblent gravés par la même main. Sartre 1985 : 237 ; Gatier 1998 : 367–368, no 3. Nom conventionnel que je conserverai, voir Sartre 2016. Sartre 2016 : no 11 et 13, 16 et 17.

méharistes et cavaliers romains dans le désert jordanien

289

l’ ala Flauia Gaetulorum casernée en Mésie inférieure. La présence de cette ala Gaetulorum ueterana est attestée en Arabie par les trois diplômes militaires où elle figure parmi les troupes auxiliaires de la province, aux côtés de l’ ala I Ulpia dromadariorum Palmyrenorum milliaria, en 126, 142 et 14564. Il est probable que ce régiment de cavalerie a participé, en provenance de la Judée où il était caserné auparavant, à l’annexion du royaume nabatéen et à sa réduction en province d’Arabie en 106. Le dernier témoignage daté qui le concerne est le diplôme de 145 et il est absent de la Notitia dignitatum. Mais on connaît, dans une inscription de Tomi (Mésie inférieure) datée entre le milieu du iie siècle et les années 230 du iiie siècle, un certain Sedatius Apollônios, « préfet de l’ aile des Gétules en Arabie»65. Les cinq graffites grecs du Jabal Ithlib paraissent tous concerner des soldats de l’aile des Gétules et on peut supposer qu’ils sont contemporains. On a pensé parfois que l’aile des dromedarii, mentionnée à Qubūr al-Jundī, à 7 km de là, était un détachement de cette aile des Gétules66, mais cette proposition n’est pas soutenable, comme le démontrent désormais les diplômes militaires. Le graffite 1 désigne un cavalier, Μνησθῇ [---] ἐκκύης ἄλα Γετού(λ)ων ; le graffite 3 est un texte de même type; dans le graffite 4, peu clair, il est question d’ une aile sans qu’on puisse l’identifier. Rien ne montre, en l’ absence de bonnes photos publiées qui permettraient de se faire une opinion sur leur écriture, que les graffites du Jabal Ithlib qui concernent l’aile des Gétules sont strictement contemporains de ceux de Qubūr al-Jundī. Nous avons vu que le vocabulaire dont ils faisaient usage est plus une translittération du latin en grec qu’ une traduction, un peu à la façon du graffite de Bāyir. Cela me semble donner des indications de chronologie relative qui montreraient que les graffites du Jabal Ithlib ne sont pas contemporains de ceux de Qubūr al-Jundī, mais postérieurs. En effet, ce phénomène est assez commun dans l’ Antiquité tardive. Revenons au graffite de Bāyir, pour lequel on ne dispose que de trois éléments de datation. Le premier est l’écriture, ce qui reste imprécis, surtout dans le cas de graffites. On se contentera de remarquer qu’ elle peut appartenir au iiie siècle comme au ive. Le second élément est l’épithète d’Ant(-) qui qualifie l’ aile des méharistes. S’il s’agit d’Antoniniana, cette titulature a été donnée par Caracalla ou Élagabal, entre 211 et 222, ce qui constituerait un terminus post quem ;

64 65 66

Spaul 1994: 124–125, à corriger; Weiss et Speidel 2004 (ae 2004, no 1925); Eck et Pangerl 2016. ae 1974, no 579 ; seg 24, no 1064, avec une datation du iie ou iiie siècle; seg 27, no 401, avec une hypothèse de datation des environs de 232–233. Sartre 1982b : 33–34, présente l’ état ancien du débat.

290

gatier

s’il s’agit d’Antana ou d’Antiana, cela ne fournit aucune information chronologique utile. Le troisième est l’onomastique, et plus précisément la présence de deux ou trois Flauii dans la liste de Bāyir, ce qui donne un terminus post quem plus tardif, dans le règne de Constantin (306–337). Il ne paraît pas possible de descendre trop bas dans le ive siècle, vu l’absence de toute trace de christianisme. Je pense que l’on peut donc placer le graffite de Bāyir dans les années médianes du siècle, entre 325 et 375. Le site n’appartenait alors plus à la province d’Arabie, mais à la Palestine, depuis les redécoupages de la Tétrarchie67.

iv

Fonction de Bāyir

Tous les graffites grecs et latins de Bāyir ne datent pas nécessairement de la même période et rien n’oblige à considérer tous les textes grecs comme militaires. Le graffite latin 4, publié jadis, est certainement ancien, surtout si le praenomen du premier personnage figure au début de son nom, usage qui disparaît dans la première moitié du iie siècle68. Le texte est clairement un document militaire et on soupçonne que le nom Lucius, gravé en grec, du graffite 1 pourrait être aussi celui d’un soldat. Mais l’ exemple de Namārah montre que certains des graffites grecs peuvent être l’ œuvre de villageois des régions voisines. Rappelons aussi que plusieurs autres textes grecs de Bāyir sont restés inédits. Les raisons qui poussent des individus à graver leur nom, et un peu plus parfois, sur des rochers dans le désert, ont souvent été discutées. Dans le cas présent, même si le désœuvrement et l’ennui inhérents à la vie militaire ne sont pas à négliger, il est également possible que le lieu, malheureusement peu précis, où a été gravé le texte qui nous retient ici, ait eu une fonction religieuse qui expliquerait le formulaire, μνησθῇ. Il faut comprendre quel était le rôle de Bāyir qui y justifiait la présence de soldats romains, mais il n’est pas possible de savoir si cette présence a été permanente, prolongée ou intermittente. Sans oublier les trois bâtiments sur podium repérés par Jacqueline Gysens Calzini, les trouvailles de céramique considérée comme nabatéenne puis romaine tardive ou byzantine ancienne me semblent dépasser ce qu’aurait entraîné le simple usage des puits. Elles montrent que le lieu a connu des périodes soit de forte fréquentation soit d’ installation d’ un établissement permanent. On a vu que le bâtiment décrit comme un château n’était ni une forteresse ni un caravansérail, mais qu’ il devait être considéré, en

67 68

Sartre 1982b : 71–73. Gysens Calzini 1993 : 107–111.

méharistes et cavaliers romains dans le désert jordanien

291

suivant notamment le raisonnement de Denis Genequand, comme une résidence aristocratique omeyyade inachevée, même si, paradoxalement, pas ou peu de céramique correspond à cette phase de construction. Cependant, il n’est pas possible, dans l’état actuel des connaissances, de déterminer s’ il y avait un fort construit en dur pour les soldats qui ont laissé des graffites ou pour ceux d’autres périodes. La question se pose de manière voisine pour le site de Namārah, en Syrie du Sud. L’existence à Bāyir d’ un fortin nabatéen, romain ou byzantin précédant le château omeyyade reste donc une hypothèse à vérifier. Quoi qu’il en soit, seuls des travaux archéologiques renouvelés ainsi que la publication des graffites inédits permettront un jour de répondre à ces incertitudes. L’eau que fournissent deux des quatre puits creusés près du château, en bordure du lit du Wādī Bāyir, a permis à Lawrence et ses accompagnateurs d’ abreuver 500 à 600 chameaux. Henry Field décrit une consommation d’ eau de la même ampleur lors du passage de son groupe en novembre 1927: « With a bucket, we drew up extremely dirty water. A terrible odor permeated the area. This was caused by the refuse of a large tribe with hundreds of camels, who had passed northwards only a day or two before»69. Le grand avantage du site de Bāyir est donc sa possibilité d’offrir, même en été70, une grande quantité d’ eau dans une région qui en est dépourvue. Beaucoup de savants et de commentateurs, pour qui le chameau est un animal synonyme de caravane, ont voulu – à tort – reconnaître un khān ou caravansérail dans le château71. Nombre d’entre eux ont donc placé Bāyir sur des itinéraires qui illustreraient son rôle de carrefour essentiel des communications aussi bien Est/Ouest que Nord/Sud, voire de pivot entre l’ Égypte et la Babylonie72. Le géographe arabe du xe siècle, al-Muqaddasī, place en effet le site de Wubayr – que Musil identifie à Bāyir – sur la route d’ Amman à Taymāʾ73. Des routes vers Qaṭrānah à l’ouest ou vers le Wādī Sirḥān à l’ est pas-

69 70 71 72

73

Field 1960 : 43. Bell 2000 : 48–49, avec la précison que les sheikhs des Banī Ṣakhr (Sukhur) y campent avec 500 tentes. Musil 1927: 508, identifie Obaira, Ὄβαιρα, de la Géographie de Ptolémée (iie siècle apr. J.-C.) à Bāyir. Pour Glueck (1940: 42–43 ; 1945–1949, Explorations iv: 47–49), Bāyir est un «important traffic center» au croisement de routes vers Pétra et ʿAqabah, Kilwah, Taymāʾ et Hégra, le Wādī Sirḥān et al-Jawf. Parker 1986: 529 : route depuis la «route du roi» ou via nova vers l’ est, par le Wādī al-Ḥafira vers Bāyir, puis vers le Wādī Sirḥān. Voir Potts 1988: 132, 135. Musil 1927 : 517 ; repris par King 1987 : 25.

292

gatier

seraient par Bāyir74. Il me semble que beaucoup de ces itinéraires sont des lignes tirées par les Modernes entre des points, hors de toute hiérarchie, en mélangeant les sentiers de montagne, les pistes malaisées et les grands axes routiers. La carte des déplacements de Lawrence vers et à partir de Bāyir en 1917 et 1918 montre, par exemple, les multiples possibilités de circulation autour d’un point bien alimenté en eau, mais elle ne trace pas de routes à proprement parler75. L’idée d’une grande route de transport traversant Bāyir, dans l’ Antiquité comme aux périodes plus récentes, me semble devoir être récusée. Elle est contredite par les descriptions du site. La plupart des voyageurs font état des difficultés considérables qu’ils ont éprouvées à se rendre à Bāyir, à cause du relief, de la médiocrité des pistes et de la distance entre d’ autres puits et le site. Cependant, il ne s’agit pas de nier toutes les possibilités de circulation qui justifient une étape à Bāyir pour abreuver sa monture. Il est intéressant de noter, par exemple, que la route de Burckhardt depuis le Wādī al-Ḥasā jusqu’ à al-Jawf paraît un itinéraire bien établi, qui passe par les deux différents sites où nous connaissons des graffites du soldat Flavius Bèrullios76. En réalité, les itinéraires régionaux sont multiples, mais les principaux ne concernent pas Bāyir. Si des puits bien alimentés sont utiles pour des déplacements de voyageurs et de marchandises à moyenne distance, ils sont surtout indispensables aux mouvements des troupeaux des éleveurs nomades. Les histoires de rezzous de Gertrude Bell, comme les récits de Carruthers77, de Musil78 et de Field, montrent que le ravitaillement en eau des troupeaux de chameaux des tribus constitue la fonction principale du site en toute saison, en attirant les nomades et en fournissant l’occasion de rencontres ou de conflits. Dans l’ Antiquité, la surveillance de ces nomades justifiait la présence, permanente ou intermittente, des soldats romains aux abords des puits de Bāyir pour exercer la police du désert. Les

74

75 76 77

78

Carruthers 1910: 243, envisageait même une route de l’Égypte à la Babylonie, par Maʿān, Bāyir et al-Jawf; voir Potts 1988: 148–149, sur la route de Qaṭrānah à Taymāʾ et celle de Pétra au Jawf. Gysens Calzini et Ruffo 1995 : 36–37. Lawrence 1967 : map ii (entre les p. 64 et 65). Burckhardt 1822 : 665. Alexander Carruthers (1910: 243), est fait prisonnier à Bāyir avec son escorte par «a raiding party of Beduin, they were encamped at the well, on the look-out for Arabs coming in to get water ». A. Musil (1926: 25) explique qu’ en été il est extrêmement dangereux de se trouver près de Bāyir, parce que tous les groupes de pillards du désert sont contraints de venir à ces puits qui offrent les seules possibilités d’ abreuvement en cette saison, mais que le danger est encore plus grand en période de conflits inter-tribaux.

méharistes et cavaliers romains dans le désert jordanien

293

Modernes n’ont pas manqué de comparer l’action de la Légion Arabe de Glubb Pasha construisant un fort près des puits avec celle des soldats romains.

v

Conclusion: un document d’histoire militaire

En attendant la publication ou une republication détaillée d’ autres ensembles de graffites rupestres grecs ou latins, comme ceux de la région de Hégra, et en attendant aussi l’étude des autres trouvailles de Bāyir, notamment des textes grecs, le graffite présenté ici enrichit la documentation sur l’ histoire militaire du Proche-Orient tardo-antique. Il témoigne de la présence à Bāyir, très à l’ est du plateau jordanien, vers le milieu du ive siècle, de soldats romains appartenant à deux unités différentes, l’une de cavaliers, l’ ala Gaetulorum ueterana, et l’ autre de cavaliers et de méharistes, l’ ala Ant(-) dromedariorum. Pour la première aile, il s’agit du témoignage le plus récent la concernant ; pour la seconde, au contraire, c’est le plus ancien. Le texte montre, ou plutôt confirme, que les rares alae dromedariorum n’étaient pas composées seulement de méharistes, mais qu’elles comprenaient aussi des cavaliers. En revanche, rien ne prouve que les ailes ordinaires de cavalerie, comme l’aile des Gétules, aient comporté des méharistes. Depuis les réformes territoriales de Dioclétien, Bāyir dépendait vraisemblablement de la Palestine et, même si la situation du Wādī Sirḥān plus à l’ est n’est pas bien connue au ive siècle, le site se trouvait à l’ extrémité désertique du monde romain, un peu comme Namārah. Bāyir aussi, grâce à de précieuses ressources en eau, servait pour l’armée romaine de point de contrôle des nomades. Le site permettait de surveiller ceux qui venaient y abreuver leurs troupeaux et peut-être aussi de contrecarrer les raids et rezzous.

Sigles ae Année épigraphique, Paris. Bulletin épigraphique «Bulletin épigraphique», Revue des études grecques. igls Inscriptions grecques et latines de la Syrie. ils Inscriptiones Latinae Selectae. Berlin: Weidmann, 1892–1916. seg Supplementum Epigraphicum Graecum. Leiden.

Les sigles papyrologiques correspondent aux conventions exposées dans Checklist of Editions of Greek, Latin, Demotic and Coptic Papyri, Ostraca and Tablets, en ligne: http://library.duke.edu/rubenstein/scriptorium/papyrus/texts/clist .html, consulté le 27/07/2016.

294

gatier

Bibliographie Bader, N. 2009. Inscriptions grecques et latines de la Syrie, t. xxi, Inscriptions de la Jordanie, t. 5/1, La Jordanie du Nord-Est. Beyrouth: Institut français du Proche-Orient. Bell, G.L. 1927. The Letters of Gertrude Bell, édité par Lady Bell. London: E. Benn. Bell, G.L. 2000. The Arabian Diaries 1913–1914, édité par R. O’Brien. Syracuse, ny: Syracuse University Press. Burckhardt, J.L. 1822. Travels in Syria and the Holy Land. London: J. Murray. Calzini Gysens, voir Gysens Calzini. Carruthers, A.D. 1910. «A Journey in North-Western Arabia, 1909». The Geographical Journal 35: 225–248. Clark, V.A. 1987. «Safaitic and Thamudic Inscriptions from Wādī Bāyir, Jordan». Zeitschrift des Deutschen Palästina-Vereins 103: 185–191. Creswell, K.A.C. 1932–1940. Early Muslim Architecture (Umayyads a.d.622–750), vol. 1– 2. Oxford: Clarendon Press. Dąbrowa, E. 1991. « Dromedarii in the Roman Army: A Note». Dans Roman Frontier Studies 1989: Proceedings of the 15th International Congress of Roman Frontier Studies, Held at Canterbury, 2–10 September 1989, édité par V.A. Maxfield et M.J. Dobson, 364– 366. Exeter: University of Exeter. Day, F. 1960. «Historical Notes on Burquʿ, Bayir and Dauqara». Dans North Arabian Desert Archaeological Survey, 1925–1950, édité par H. Field, 150–160. Cambridge, ma: The Peabody Museum, Harvard University. Di Segni, L. 2004. «The Beersheba Tax Edict Reconsidered in the Light of a Newly Discovered Fragment». Scripta Classica Israelitica 23: 131–158. Eck, W. 2016. «Zu Militärinschriften im römischen Iudaea. Epigraphische Vorarbeiten zum ciip». Zeitschrift für Papyrologie und Epigraphik 197: 231–239. Eck, W. et A. Pangerl. 2016. «Ein Diplom für die Hilfstruppen der Provinz Arabia, ausgestellt unter Hadrian, wohl im Jahr 126». Zeitschrift für Papyrologie und Epigraphik 197: 227–230. Field, H. 1960. North Arabian Desert Archaeological Survey, 1925–1950. Cambridge, ma: The Peabody Museum, Harvard University. Gatier, P.-L. 1998. «Les inscriptions grecques et latines de Samra et de Rihab». Dans Fouilles de Khirbet es-Samra en Jordanie, 1. La voie romaine, le cimetière, les documents épigraphiques, sous la direction de J.-B. Humbert et A. Desreumaux, 359–431. Turnhout: Brepols. Gatier, P.-L. 2007. «Decapolitana». Syria 84: 169–184. Genequand, D. 2003. «Maʿān, an Early Islamic Settlement in Southern Jordan. Preliminary Report on a Survey in 2002». Annual of the Department of Antiquities of Jordan 47: 25–35. Genequand, D. 2012. Les établissements des élites omeyyades en Palmyrène et au ProcheOrient. Beyrouth: Institut français du Proche-Orient.

méharistes et cavaliers romains dans le désert jordanien

295

Glueck, N. 1933. «Prehistoric Rock-Drawings in Transjordan». American Journal of Archaeology 37: 381–386. Glueck, N. 1934–1951. Explorations in Eastern Palestine, i–iv (Annual of the American Schools of Oriental Research, vol. 14, 15, 18–19, 25–28). New Haven: American Schools of Oriental Research. Glueck, N. 1940. The Other Side of the Jordan. Cambridge, ma: American Schools of Oriental Research. Graf, D.F. 1988. «Qurā ʿArabiyya and Provincia Arabia». Dans Géographie historique au Proche-Orient (Syrie, Phénicie, Arabie, grecques, romaines, byzantines), sous la direction de P.-L. Gatier, B. Helly et J.-P. Rey-Coquais, 171–211. Paris: Éditions du cnrs. Gysens Calzini, J. 1993. «Greek and Latin Graffiti from Bayir, South-East Jordan». Dans Arabia Antiqua, Hellenistic Centres around Arabia, édité par A. Invernizzi et J.-F. Salles, 105–112. Roma: Istituto italiano per il Medio ed Estremo Oriente. Gysens Calzini, J. et F. al-Khraysheh. 1995. «Preliminary Report of Reconnaissance Survey in the Region of Wādī Bāyir». Annual of the Department of Antiquities of Jordan 39: 355–364. Gysens Calzini, J. et G. Ruffo. 1995. «A Preliminary Report on a Reconnaissance Survey in Southeast Jordan (Region of Wādī Bāyir)». East and West 45: 23–44. Hogarth, D.G. 1927. «Gertrude Bell’s Journey to Hayil». The Geographical Journal 70 (1): 1–17. Jaussen, A. et R. Savignac. 1914. Mission archéologique en Arabie, ii. Paris: E. Leroux. Kennedy, D.L. 1982. Archaeological Explorations on the Roman Frontier in North-East Jordan. The Roman and Byzantine Military Installations and Road Network on the Ground and from the Air. (bar International Series 134). Oxford: bar. King, G.R.D. 1987. «The Distribution of Sites and Routes in the Jordanian and Syrian Deserts in the Early Islamic Period». Proceedings of the Seminar for Arabian Studies 17: 91–105. King, G.R.D., C.J. Lenzen et G.O. Rollefson. 1983. «Survey of Byzantine and Islamic Sites in Jordan. Second Season Report, 1981». Annual of the Department of Antiquities of Jordan 27: 385–436. Lammens, H. 1910. «La Bādia et la Ḥīra sous les Omaiyades. Un mot à propos de Mshattā ». Mélanges de la Faculté Orientale 4: 91–112. Last, R. et A. Stein. 1990. «Ala Antiana in Scythopolis: A New Inscription from BethShean». Zeitschrift für Papyrologie und Epigraphik 81: 224–228. Lawrence, T.E. 1967. Seven Pillars of Wisdom, a Triumph. London: Jonathan Cape. Macdonald, M.C.A. 2008. «Transformation and Continuity at al-Namāra: Camps, Settlements, Forts and Tombs». Dans Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, édité par K. Bartl et A. Moaz, 317–332. Rahden: Leidorf.

296

gatier

Macdonald, M.C.A. et A. Searight. 1982. «The Inscriptions and Rock-drawings of the Jawa Area: A Preliminary Report of the First Season of Field-work of the Corpus of Inscriptions of Jordan Project». Annual of the Department of Antiquities of Jordan 26: 159–172. Macdonald, M.C.A., M. al-Muʾazzin et L. Nehmé. 1996. «Les inscriptions safaïtiques de Syrie, cent quarante ans après leur découverte». Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres : 435–494. Moritz, B. 1908. «Ausflüge in der Arabia Petraea». Mélanges de la Faculté Orientale 3: 387–436. Musil, A. 1926. The Northern Ḥeǧâz: A Topographical Itinerary. New York: American Geographical Society. Musil, A. 1927. Arabia Deserta: A Topographical Itinerary. New York: American Geographical Society. Musil, A. 1928. Palmyrena: A Topographical Itinerary. New York: American Geographical Society. Nehmé, L. 2009. «Quelques éléments de réflexion sur Hégra et sa région à partir du iie siècle après J.-C.». Dans L’Arabie à la veille de l’Islam. Bilan clinique, édité par J. Schiettecatte et Chr. J. Robin, 37–58. Paris: De Boccard. Parker, S.T. 1986. Romans and Saracens. A History of the Arabian Frontier. Winona Lake, in: Eisenbrauns. Potts, D.T. 1988. «Trans-Arabian Routes of the Pre-Islamic Period». Dans L’Arabie et ses mers bordières, i, Itinéraires et voisinages, sous la direction de J.-F. Salles, 127–168. Lyon: Maison de l’Orient. Rance, P. 2015. «Camels: Late Empire». Dans The Encyclopedia of the Roman Army, édité par Y. Le Bohec, G. Brizzi, E. Deschler-Erb, G. Greatrex, B. Rankov, M. Reddé, vol. 1, 125–126. Chichester: Wiley-Blackwell. Rolston, S.L. et G.O. Rollefson. 1982. «The Wadi Bayir Paleoanthropological Survey: A First Season Report». Annual of the Department of Antiquities of Jordan 26: 211–219. Sartre, M. 1982a. Inscriptions grecques et latines de la Syrie, t. xiii/1, Bostra. Paris: P. Geuthner. Sartre, M. 1982b. Trois études sur l’Arabie romaine et byzantine. Bruxelles: Latomus. Sartre, M. 1985. Bostra, des origines à l’Islam. Paris: P. Geuthner. Sartre, M. 1993. Inscriptions grecques et latines de la Syrie, t. xxi, Inscriptions de la Jordanie, t. 4, Pétra et la Nabatène méridionale. Paris: P. Geuthner. Sartre, M. 2016. «Namāra du Ṣafā». Syria 93: 45–66. Sartre-Fauriat, A. et M. Sartre. 2014. Inscriptions grecques et latines de la Syrie, t. xv/2, Le plateau du Trachôn et ses bordures. Beyrouth: Institut français du Proche-Orient. Sauvaget, J. 1939. «Remarques sur les monuments omeyyades, i, Châteaux de Syrie». Journal Asiatique 231: 1–59. Schroeder, E. 1960. «Architectural Report». Dans North Arabian Desert Archaeological

méharistes et cavaliers romains dans le désert jordanien

297

Survey, 1925–1950, édité par H. Field, 94–104. Cambridge, ma: The Peabody Museum, Harvard University. Seeck, O. (éd.). 1876. Notitia Dignitatum. Berlin: Weidmann. Seyrig, H. 1941. «Antiquités syriennes, 37. Postes romains sur la route de Médine». Syria 22: 218–223 (= Antiquités Syriennes, iii, 162–167, et p. vii. Paris: P. Geuthner, 1946). Spaul, J.E.H. 1994. Ala2: The Auxiliary Cavalry Units of the Pre-Diocletianic Imperial Roman Army. Andover: Nectoreca Press. Speidel, M.P. 1977. «The Roman Army in Arabia». Dans Aufstieg und Niedergang der römischen Welt, ii Principat, 8, édité par H. Temporini et W. Haase, 687–730. Berlin: De Gruyter (= Roman Army Studies, i, 229–272. Amsterdam: Gieben, 1984). Stein, A. 1985. Sir Aurel Stein’s Limes Report. The Full Text of M.A. Stein’s Unpublished Limes Report (his Aerial and Ground Reconnaissances in Iraq and Transjordan in 1938–39), édité par S. Gregory et D. Kennedy, vol. 1–2. (bar International series 272). Oxford: bar. Waddington, W.H. 1870. Inscriptions grecques et latines de la Syrie, recueillies et expliquées. Paris: Didot. Weiss, P. 2006. «Die Auxilien des syrischen Heeres von Domitian bis Antoninus Pius. Eine Zwischenbilanz nach den neuen Militärdiplomen». Chiron 36: 249–298. Weiss, P. et M.P. Speidel. 2004. «Das erste Militärdiplom für Arabia». Zeitschrift für Papyrologie und Epigraphik 150: 253–264. Yon, J.-B. 2012. Inscriptions grecques et latines de la Syrie, t. xvii/1, Palmyre. Beyrouth: Institut français du Proche-Orient. Zeinaddin, H. 2008. «Die Inschriften von al-Namāra». Dans Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, édité par K. Bartl et A. Moaz, 333–337. Rahden: Leidorf.

chapter 14

Goras, sanglier ou jeune lion (ou onagre) ? Jean-Baptiste Yon

Après avoir il y a peu traité de certains noms d’animaux utilisés comme anthroponymes dans les langues sémitiques et éventuellement traduits en grec1, il m’a semblé intéressant de revenir sur le sujet, et en particulier sur un cas limite de l’interprétation, avec le nom Goras, attesté dans diverses langues et sur divers sites du Proche-Orient. Le sujet de cette étude d’ onomastique paraissait approprié dans un recueil en l’honneur d’un savant qui a bien souvent mis en lumière les dangers qu’il y avait à assimiler l’origine d’un nom à l’ origine ethnique de ceux qui le portaient. Michael Macdonald a aussi fréquemment souligné les difficultés que l’on peut rencontrer à interpréter les formes linguistiques des langues sémitiques quand elles sont transcrites dans des alphabets élaborés pour les langues indo-européennes. Dans le cas présenté ici, un seul nom, Goras, transcrit en grec et en latin, peut désigner selon les cas deux ou même trois animaux différents. Les difficultés sont augmentées par la présence de l’iranien, un groupe linguistique indo-européen mais assez éloigné du grec et du latin utilisés habituellement au Proche-Orient. L’interprétation est encore compliquée par le fait que des anthroponymes homophones sont également attestés dans d’ autres régions du monde grec et romain (en Macédoine et en Thrace ou en Afrique du Nord), sans aucun rapport étymologique. Pour cette raison, on présentera d’ abord les différentes formes du nom Goras avant de terminer avec des noms évoquant Goras par l’aspect, mais dont il est difficile d’établir de manière certaine s’ ils dérivent d’une des formes précédentes ou s’il existe réellement un lien autre que phonétique.

Goras Dans le recueil récent des Inscriptions grecques d’Iran et d’Asie centrale (igiac), au no 12, le patronyme du graveur est Γόρου (Ἀρίστων Γόρου)2. Il s’ agit d’ une ins1 Voir Yon 2013: deux anthroponymes, l’ un grec, l’ autre araméen, peuvent désigner le même animal (ʿgylw / Μόσχος). 2 Voir déjà seg 7, no 13.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

goras, sanglier ou jeune lion (ou onagre) ?

299

cription de Suse, en l’honneur d’un certain Zamaspès (au nom iranien), datée de 1/2 apr. J.-C. Dans une note de cette édition, P. Bernard donne une explication iranienne («à Suse en tout cas») à ce nom, pour lequel on a parfois préféré une étymologie sémitique3. Comme cela a été reconnu depuis longtemps, il peut s’ agir d’une adaptation grecque du vieil iranien *varāza, « sanglier ». Dans le même ouvrage, l’avis de Ph. Huyse est plus mesuré (« peut-être »)4. Il est vrai que, sans être impossible, une origine sémitique est peut-être moins probable à Suse, où les noms grecs dominent dans l’ épigraphie grecque, majoritairement antérieure au ier siècle apr. J.-C.5. De la même manière, il est possible de faire le rapprochement avec l’iranien gōr, « onagre», explication proposée pour le nom attesté dans la littérature syriaque sous la forme gwryʾ et également surnom porté par l’un des rois sassanides du ve siècle, Varhām6. Pourtant, il existe bien un nom semblable en araméen, gwrʾ, qui est traduit « jeune lion»7, mais qui désigne à proprement parler le petit d’ un animal (chien ou lion). Dans l’onomastique juive, le même nom apparaît sous diverses formes, dont Γωρίων / gwrywn, mais aussi gwryʾ8, dans tous les cas en lien avec le nom sémitique du petit d’un lion. Il peut donc être utile de reprendre l’ensemble de la documentation pour voir ce qu’il en est des exemples assez nombreux identifiés en Syrie orientale et en Mésopotamie septentrionale. Dans les Res Gestae Diui Saporis apparaissent les anthroponymes Γουραζ, Γοραζης/ος et Γοραζδουκτ, qui désignent de grands personnages du royaume sassanide, évidemment iraniens d’ origine ou

3 P. Bernard dans igiac, 52, n. 136. 4 Ph. Huyse, igiac, 307, dans l’ index commenté des anthroponymes iraniens. Le nom est absent de Huyse 1988, qui ne tient pas compte des anthroponymes de Doura Europos publiés dans Cumont 1926. 5 igiac, 34, n. 72. La situation est différente à une époque plus tardive, en particulier en raison des déportations de populations pratiquées par les grands rois sassanides: par exemple pour le nom des évêques à partir du ve siècle, voir Chabot, Synodicon orientale: 683. Toutefois, même dans l’ épigraphie grecque, les noms sémitiques (no 18, Βηλτιβανατις; no 40, Αβδο- (?); no 42, Ραχιμναιος) sont au moins aussi nombreux que les noms iraniens (11–12, Ζαμάσπης). J’ai exclu les noms des souverains. 6 Gignoux et al. 2009 : 74, no 198a–c, Gōriy « onagre» (voir no 200, Gōr-Šahryār), avec renvoi à Gignoux 1986: no 395 (Gōr), et idem 2003 : no 146 (Gōr, où une étymologie alternative est proposée pour ce même nom). Sur gwryʾ voir infra. Pour Vahrām, dernièrement Traina 2009: 165–166 : « l’ épithète devait signifier “grand” ou “puissant”». 7 Stark 1971: 81. Voir aussi Nöldeke 1904 : 78, avec les exemples palmyréniens et dans d’autres lieux du Proche-Orient (dont en arabe). 8 Ilan 2002: 367–368; pour le grec, Flavius Josèphe, bj iv, 159 (entre autres). C’est le nom Gurion de l’ hébreu moderne. Le nom apparaît sous la forme gwryʾ: Ilan 2011: 342 (voir infra).

300

yon

de culture9. En arménien (Varaz10), et en grec Οὐαράζης (Procope, De Bellis vii, 27, 3 et 10–11; viii, 13, 10) et Βαράζης (Agathias iv, 13, 3–4)11, désignent des nobles arméniens. Chez Plutarque, le nom Γούρας est celui d’ un frère de Tigrane d’Arménie qui défend Nisibe contre Lucullus (Lucullus 32, 5–6 : ἀδελφὸς Τιγράνου Γούρας). Là encore, une étymologie iranienne s’ impose. Dans la région de l’Euphrate, un nom Goras est bien attesté, à Tell Aḥmar (au sud de Karkemish / Djerablous), pour Marcus Gora f(ilius) sig(nifer) c(o)h(ortis) I Ascalonit(orum) domo Hemesa12. Le personnage est originaire d’ Émèse, ce qui rend l’étymologie iranienne moins immédiatement tentante. Lors de la publication de ce texte, P.-L. Gatier avait rassemblé les exemples du nom dans l’ensemble du Proche-Orient, ce qui permet de rester succinct ici, en attirant néanmoins l’attention sur le nombre particulièrement élevé d’ exemples douréens13, ce qui n’empêche pas que le nom soit répandu à travers l’ ensemble de la Syrie (Massif calcaire, steppe et Syrie du Sud)14.

9

10

11 12 13

14

Respectivement lignes 62, 57 et 50 du grec. Huyse 1999: vol. 1, 125–126, pour les équivalents moyen-perse et parthe. À la ligne 57, il s’ agit clairement d’un nom de famille: voir le commentaire de Huyse 1999: vol. 2, 134–135 (§ 42.4); ibidem à propos du passage de l’ iranien /Wa-/ au grec Γο(υ)-. Voir Garsoian 1989: 422, pour des exemples nombreux dans l’histoire arménienne. Sur l’ étymologie et les exemples iraniens, Justi 1895 : 348–349; Gignoux 1986: 173–174, no 940– 944. Voir sans doute aussi IosPE ii, no 451 = cirb no 1282, l. 14 (Ουαράζακον) dans une liste de fidèles du Théos Hypsistos à Tanaïs (l’ onomastique iranienne est très présente). Gatier 1994 : 151–157 (ae 1994, no 1764), repris Choix igls no 57. Gatier 1994: 153–154, à propos de l’ étymologie du terme avec liste d’attestations en Syrie. À Doura Europos, liste des attestations épigraphiques dans Grassi 2012: 47, avec commentaire p. 182: Cumont 1926: 405, no 50 (Γορας), et 443, no 121a (Γορου au gén.); Dura Prel. Rep. ii, 123, no d36 (Γορας ; seg 7, no 565) ; Dura Prel. Rep. v, 18, no 378 (Γορας); Dura Prel. Rep. vii/viii, 168, no 871 (Γορου au gén.). Dans les papyrus de la cohorte xx Palmyrenorum, P. Dura 97, 3 (Malchus Gora [dans ce cas, comme les suivants, génitif patronymique; cf. ChLA vii, p. 352]) ; 98, ii, 17 (Bassus Gora) ; 100, xlii, 28 (Aurel(ius) Zabdibolus Gora); 101, x, 26 (Bassus Gora) ; 101, xlii, 19 (Zabdibolus Gora); 115a, i, 3 (Themarsa Gora); 115a, i, 15 (Zabdibolus Gora). On a donc au moins quatre soldats différents; un autre est Aurel(ius) Nisamsus Guris (100, xli, 28 ; 101, xli, 16). Pour les formes peut-être apparentées, Goremis et Gorippus, voir infra. À Émèse, Γορ[- -], sans doute à restituer Γορ[ου] au génitif (igls v, no 2391); en Syrie du Sud (Trachôn), igls xv, no 105 et 105a (Γορας), no 397 (Γορo[υ], au génitif, nom de clan). En igls xv, no 312, la lecture proposée par les premiers éditeurs (⟨Γ⟩ου⟨ρ⟩α, paes iiia, no 8016) a été corrigée et le texte est donc à écarter de la discussion. On le retrouve également sous la forme Γουρας dans la région du Massif calcaire, ainsi igls iv, no 1425 (Dana, Γουρου) et 1480 (al-Bārah, Γουρα, au gén.). Un soldat originaire d’Antioche est Caius Iulius Gurati

goras, sanglier ou jeune lion (ou onagre) ?

301

Je me contenterai donc d’ajouter quelques attestations, soit épigraphiques, soit littéraires, en particulier syriaques. Le même nom est aussi attesté à Karab près de Balanée15, et sans doute également comme Γοροῦς à Apamée16. Plus près de l’Euphrate, il apparaît à Palmyre, sous les deux formes gwrʾ / gwry ; dans deux inscriptions de la diaspora, une avec la forme latine Guras (Flauius Guras Iiddei / gwrʾ ydy)17 et une avec la forme Goras (Maximus Gora, idr iii/2, no 20, à Sarmizegetusa). Enfin, sur l’Euphrate (P. Euphr. 6–7), une femme s’ appelle Μαθ̣θαβεινη Αββα τοῦ Γ[ο]ρα18. L’onomastique régionale, et dans ce dernier dossier en particulier, est surtout sémitique, ce qui n’empêche pas que les noms iraniens soient également bien attestés (Worod19). La même indétermination règne pour un des derniers exemples, dans le Khuzestan (Iran sud-occidental, à une soixantaine de kilomètres au sud-est de Suse), puisque gwrʾ est le nom d’ un évêque de Šoušterê (dont le nom actuel est Šuštar), au synode de 42020. Comme toujours, on peut se demander s’il s’agit du nom sémitique d’un chrétien21 – éventuellement déporté par les rois sassanides depuis la Haute-Mésopotamie ou la Syrie –, ou bien d’un nom iranien. Il est possible que l’absence de la finale z/s permette de pencher pour la première solution (à moins qu’ il ne s’ agisse encore de l’onagre), mais la réponse appartient plus sûrement à un spécialiste des langues iraniennes. Dans le cas des exemples égyptiens, assez nombreux au Mons Claudianus, le contexte est nettement sémitique : O. Claud. iii, no 455, 471, 472, 474. Il y a là deux individus, Γορας Ἀμπ[λ]ιάτου (1er et 3e textes) et Γορας Βαραθη, dont le second au moins est d’ascendance proche-orientale, Βαραθης renvoyant sans doute à Palmyre, ou à la Syrie du Nord et à la région de l’Euphrate. Avant d’en passer à des anthroponymes peut-être apparentés, il convient aussi de faire la liste de noms pratiquement homophones, qui n’ont apparemment aucun rapport avec le Proche-Orient. En effet, je ne crois pas que les deux

15 16 17 18 19 20 21

filius Domitianus Antiochia ex Syria Coele et Proculus filius eius : ae 1993, no 1789 = rmd 3, no 189. Voir aussi seg 20, no 345, à Bāṣūfān (Γορη- -). L’ inscription est conservée au Musée national de Beyrouth (inv. 26350), IGLMusBey 384. Balty 1981 : 194, no 6. cis 3906 (bilingue araméen-latin, idr iii/1, 154 ; pat 0251, à Tibiscum), 4062, 4348–4349 pour le premier (pat 0408, 0706–0707). cis 4281 (= pat 0638) pour gwry. Feissel et al. : 23, n. 68. P. Euphr. Syr. a et b, sous la forme wrwd ; P. Euphr. 16, sous la forme Ουροδος. Chabot, Synodicon orientale, 42, l. 18. On retrouve d’ autres noms sémitiques parmi les évêques du lieu: Chabot, Synodicon orientale, 683.

302

yon

inscriptions d’Afrique proconsulaire (Théveste, cil viii, no 1843 [16500]) et de Numidie (Markouna, cil viii, no 4279), qui portent les noms de Gurai et de Gura (féminin)22 se rattachent à un milieu d’origine sémitique. Il est clair que des noms semblables, en particulier Γούρας, sont assez bien attestés ailleurs, ainsi en Macédoine. Dans sa récente synthèse, D. Dana lui donne une origine thrace occidentale23. Pour cette raison, le rapprochement fait par l’ éditeur du Recueil des inscriptions chrétiennes de Macédoine, au no 236, entre le nom d’ un prêtre à Philippes (Γουράσιος) et le nom sémitique attesté par exemple en igls iv, no 1480, n’est pas absolument certain, au regard des exemples macédoniens plus proches géographiquement. Comme c’est aussi le cas à Doura Europos ou à Zeugma, il est fort possible que certains noms macédoniens typiques ont été conservés longtemps dans l’onomastique des élites locales de fondations hellénistiques, peuplées dès l’origine par des Gréco-Macédoniens24. La question pourrait éventuellement se poser pour Γουρᾶς / Γορας, en raison des nombreux exemples macédoniens: toutefois, l’étude du contexte des différentes attestations montre clairement que Γορας se trouve principalement parmi des noms typiquement sémitiques, et pas nécessairement en relation avec les notables de la ville de Doura. Certes, un Aurelius Goras est gazzophylax (Αὐρήλιος Γορας γαζζοφ(ύλαξ)), ce qui indique peut-être une position de premier plan dans l’ administration du sanctuaire d’Artémis, mais rien ne montre des liens avec des personnages à l’onomastique macédonienne. En effet, il s’agit sûrement du même que le père (ou le grand-père selon Fr. Cumont) d’un certain Ορθονοβαζος (Ορθονοβαζος Γορου)25 dans une autre inscription du même sanctuaire. On se trouve renvoyé à une onomastique de type plutôt iranien, ce qui correspond d’ ailleurs bien à la fonction de gazzophylax (voir les remarques de Fr. Cumont, ad loc., ou cirb no 45 et 49). De même, il y a peu d’ éléments pour conclure à un milieu spécifiquement proche-oriental dans le cas d’ un affranchi enterré dans la nécropole de la via Ostiense à Rome26. Les personnes de même condition sont nombreuses dans le reste de ce corpus romain, et donc les noms grecs, mais l’onomastique sémitique ou iranienne y apparaît peu importante.

22

23 24 25 26

Diehl, ilcv no 2050b pour un autre exemple (Gureus). Un terme apparenté semble attesté dans la toponymie moderne (Qasr Gurai), ce qui révèle éventuellement une origine locale du nom. Dana 2014 : 191 (liste des exemples). Yon 2014. Cumont 1926 : 405, no 50, et 443, no 121a. Solin 1996 : vol. 3, 603 : Q(uinti) Agillei Q(uinti) Kani / l(iberti) Gorae (ISOstiense, no 107).

goras, sanglier ou jeune lion (ou onagre) ?

303

Noms apparentés Un autre groupe de noms est constitué de formes que l’ on peut penser apparentées, tel l’anthroponyme Gortha / Gorthas (Γορθα / Γορθας), connu en Syrie du Nord ou à Zeugma27. Dans ces deux cas, il s’agit de noms féminins (à en juger par les corbeilles qui les accompagnent sur ces stèles funéraires) qui peuvent être considérés comme des formes féminines de Γουρας / Γορας. On pense aussi au nom d’un hérétique, cité par Eusèbe (he iv, 22, 5 : Γορθαῖος, ὅθεν Γοραθηνοί), Théodoret (dans une liste d’hérétiques, Haereticarum fabularum compendium, pg 83, col. 388: καὶ Γορθέου), et encore au 3e Concile de Constantinople (Γόρθεος)28. Comme le rappelle Épiphane, Gorthaios donna son nom à une secte judaïsante29, issue (ou proche?) des Samaritains. Il n’y a pas d’ indications géographiques, mais il faut plutôt supposer une origine palestinienne. Le nom est dans ce cas bien évidemment masculin, ce qui fait que le rapprochement avec les noms féminins Γορθα / Γορθας est incertain. À Doura, d’autres noms proches, Γοριαια et Γοραιος, sont des transcriptions du nom gwry, connu à Palmyre, ou de gwryʾ syriaque, donc très clairement des formes dérivées de Goras / Guras30. On retrouve le même nom pour un martyr fameux à Édesse (Γουρίας / gwryʾ)31. Dans ce cas, l’ interprétation par l’ iranien

27 28 29

30 31

Γορθας à Zeugma : seg 46, no 1754 = Ergeç et Yon 2012: 185, no 41. Γορθα à Hiérapolis: ReyCoquais 1998 : 198, no 2 (seg 48, no 1859). aco ii, 2 (Concilium universale Constantinopolitanum tertium [680–681]), Document 11, 476. Épiphane, Haeres. 12, édit. Holl, i, 205 : Κατὰ Γοροθηνῶν, ἀπὸ Σαμαρειτῶν αἱρέσεως γ, τῆς δὲ ἀκολουθίας ιβ. Γοροθηνοί τε καὶ οἱ ἄλλοι. Ἐσσηνοὶ δὲ ἐγγὺς τῶν ἄλλων γινόμενοι τὰ ἴσα ἐκείνοις πράττουσι· μόνοι δὲ Γοροθηνοὶ καὶ Δοσίθεοι τὴν φιλονεικίαν εἰσὶ πρὸς Σεβουαίους κεκτημένοι. καὶ ποιοῦσιν αὐτοὶ τὰς ἑορτάς, φημὶ δὲ Γοροθηνοὶ καὶ Δοσίθεοι, ὅταν οἱ Ἰουδαῖοι ἐπιτελῶσι, τῶν τε Ἀζύμων καὶ Πάσχων, Πεντηκοστῆς τε καὶ Σκηνοπηγιῶν καὶ νηστείας τῆς παρ’ αὐτοῖς νομιτευομένης ἡμέραν μίαν. οἱ δὲ ἄλλοι οὐχ οὕτως ποιοῦσιν, ἀλλὰ ἰδιαζόντως ἐν μησὶν οἷς προείπαμεν, « Part 12. Epiphanius Against the Gorothenes: 1 : 1 But the Gorothenes and the others were not convinced by the Sebuaeans. When Essenes are in the neighbourhood of the others they do the same as they only the Gorothenes and Dositheans have the quarrel with the Sebuaeans. 1 : 2 And they, I mean the Gorothenes and Dositheans, keep the Festivals of Unleavened Bread, Passover, Pentecost and Tabernacles, and their one set fast day, when the Jews observe them. The others though (i.e., the Sebuaeans) do not keep them then, but in their own way in the months I have mentioned. » (p. 35 de la trad. de Williams 1987, i). Cumont 1926, 413, no 58 ; Dura Prel. Rep. v, 18, no 379. Voir Grassi 2012: 47, 181–182. Fiey 2004: 88–89, no 181 ; Thes. Syr. i, col. 691 ; Gebhardt 1911 pour les versions grecques. Syriaque dans Burkitt 1913, p. g et suivantes. Synaxaire de Constantinople au 15 novembre (col. 225, éd. Delehaye): Ἄθλησις τῶν ἁγίων ὁμολογητῶν Γουρία, Σαμωνᾶ καὶ Ἀβίβου μαρτυ-

304

yon

pourrait être en concurrence avec l’étymologie araméenne, bien qu’ elle puisse apparaître moins probable. De manière curieuse, ce nom n’a, semble-t-il, pas eu de postérité, puisqu’on ne connaît pas d’autres attestations, alors que l’ histoire du martyr a eu un certain succès, au moins littéraire, si l’ on en croit les traductions. À Doura encore, au iiie siècle, les soldats Goremis et Gorippus portent des noms sans étymologie claire32. Le premier nom est peut-être à rapprocher des noms grmy / grymy connus à Palmyre (grmʾ dans les textes safaïtiques), tandis que le second nom doit se rattacher à un nom arabe bien attesté en nabatéen (paes iiia, Nab. no 59: grpw) et en safaïtique (grf, Harding 1971: 159), avec des parallèles en grec en Syrie du Sud (Γορεπου, gén., à Ḥarrān, et Γορπος, à Malaḥ aṣ-Ṣarrār, paes iiia, no 7941 [= igls xv, no 264] et 714) et peut-être en Cilicie à Korykos (mama iii, no 683: Γορίππου, gén.)33. De ce dernier, on peut sans doute rapprocher un nom assez semblable d’ aspect, celui d’ un moine syriaque, Šleîmûn bar Gorop (šlymwn br grp), qui aurait vécu à la fin du viie siècle (Assemani, bo iii/1, 459). Enfin, on a proposé récemment de corriger en Gorippus le nom de l’auteur tardo-antique plus connu comme Corippus, en le rattachant au nom connu à Doura et à Korykos34. Γοραιθ dans une inscription inédite découverte au sud de Tartous35 est peutêtre apparenté à l’un de ces groupes onomastiques, mais – comme pour le nom d’hérétique Γορθαῖος / Γόρθεος et comme pour Goremis et Gorippus –, il n’est pas certain qu’il faille nécessairement le rapprocher de Goras. Si ces anthroponymes ne sont peut-être pas tous tirés de noms d’ animaux, ils ont l’avantage d’attirer l’attention sur des zones d’ ombre de l’ onomastique. Il

32

33 34 35

ρησάντων ἐν Ἐδέσῃ τῇ πόλει. Οὗτοι ὑπῆρχον ἐπὶ Διοκλητιανοῦ βασιλέως καὶ Ἀντωνίνου δουκός· ᾧτινι καὶ διαβληθέντες ὡς τοὺς ἐν τῇ φυλακῇ χριστιανοὺς θεραπεύοντες παρέστησαν, Γουρίας μὲν ὁρμώμενος ἀπὸ Σαρκιγειτνᾶς κώμης, Σαμωνᾶς δὲ ἀπὸ Γάναδος, τρόφιμοι ἄμφω Ἐδέσης καὶ ἐν τάξει ἱερέων ὑπάρχοντες. Les villages d’ origine des deux martyrs ne sont pas connus par ailleurs. P. Dura 100, xxxvii, 5 (Aurel(ius) Goremis Iadaei) ; 101, xxxvii, 7 (Goremis Iadei); 103b, 5 ( ?). P. Dura 100, xxxvii, 11 (Aurel(ius) Iulius Gorippus), et 101, xxxvii, 13 (Iulius Gorippus); 100, xl, 9 ([Aurel(ius) Gorip]pus Valenti), et 101, xl, 16 (Gorippus Valentini). Il y a donc deux Gorippus différents. Les éditeurs des P. Dura, p. 62, font de Gorippus un nom mixte («apparently as a hybrid»: une sorte d’ hippogriffe ?), sans justifier ni expliciter leur proposition. Riedlberger 2015. Dans la plaine au sud de Tartous: Rey-Coquais 1974: 226–227 et 246. Mouterde (dans ReyCoquais 1974: 227, n. 3) écrivait que « Γοραιθ semble être un diminutif arabe de même sens que le nom palmyrénien gwrʾ, “petit” de lion ou de chien».

goras, sanglier ou jeune lion (ou onagre) ?

305

faudrait pouvoir, dans chaque cas, étudier le contexte, le milieu où le nom est attesté, pour en tirer des conclusions sur l’étymologie possible. Qui plus est, dans certaines des régions concernées ici, à l’articulation des mondes sémitiques et iraniens (Suse, Osrhoène, et même Euphrate), il serait illusoire de vouloir distinguer ce qui appartient à chaque sphère culturelle, et évidemment encore plus dangereux d’attribuer au porteur d’un tel nom une origine ethnique précise, danger contre lequel Michael Macdonald a plusieurs fois mis en garde. On peut même imaginer que dans certains cas, l’ ambiguïté pouvait être volontaire, et qu’on choisissait à dessein des noms qui pouvaient trouver une interprétation dans plusieurs des langues utilisées sur place. Des phénomènes de ce type ont déjà été signalés au Proche-Orient pour des anthroponymes latins et sémitiques36. La différence est peut-être que dans le cas de Goras, si les animaux sont bien différents, on peut leur trouver des similitudes du point de vue des connotations («sauvagerie, caractère indomptable »), ce qui serait un autre élément d’explication.

Sigles et abréviations ae Année épigraphique. Paris: puf. Assemani, bo Assemani 1725. Chabot, Synodicon orientale Chabot 1902. ChLA Chartae Latinae Antiquiores. Lausanne-Olten: Urs Graf Verlag. Choix igls Yon et Gatier 2009. cil Corpus Inscriptionum Latinarum. Berlin: Reimer puis De Gruyter. cirb Corpus Inscriptionum Regni Bosporani. Moscou-Léningrad: Académie des sciences. cis Corpus Inscriptionum Semiticarum. Paris: Imprimerie nationale. Diehl, ilcv Diehl, E. 1961–1970. Inscriptiones latinae christianae veteres, Berlin: Weidmann. Dura Prel. Rep. Excavations at Dura-Europos, Preliminary Reports i–ix. New Haven: Yale University Press, 1929–1952. idr Inscriptiones Daciae Romanae. Bucarest: Academiei rs Romania. igiac Rougemont 2012. IGLMusBey Yon et Aliquot 2016. igls Inscriptions grecques et latines de la Syrie.

36

Sartre 2007.

306

yon

IosPE

Inscriptiones antiquae Orae Septentrionalis Ponti Euxini graecae et latinae. Saint-Pétersbourg: Societatis Archaeologicae Imperii Russici. ISOstiense Inscriptiones Latii veteris Latinae supplementum ostiense. Berlin: De Gruyter. mama Monumenta Asiae Minoris Antiqua. Manchester: Manchester University Press. O. Claud. Mons Claudianus. Ostraca graeca et latina. Le Caire: Institut Français d’Archéologie Orientale. P. Dura. Welles et al. 1959. P. Euphr. Feissel, Gascou et Teixidor 1997. paes Littmann et al. 1904–1921. pat Hillers, D.R., and E. Cussini. 1996. Palmyrene Aramaic Texts. Baltimore-London: Johns Hopkins University Press. Res Gestae Diui Saporis Huyse 1999. rmd Roxan, M.M. et P. Holder. Roman Military Diplomas. London: Institute of Archaeology. seg Supplementum Epigraphicum Graecum. Leiden. Thes. Syr. Thesaurus Syriacus. 2 vol. Oxford, Clarendon Press, 1879–1901.

Bibliographie Assemani, G.S. 1725, Bibliotheca Orientalis Clementino-Vaticana iii/1, De Scriptoribus Syris Nestorianis. Rome: Typis Sacrae Congregationis de Propaganda Fide. Balty, J.-Ch. 1981. Guide d’Apamée. Bruxelles: Centre belge de recherches archéologiques à Apamée de Syrie. Burkitt, F.C. 1913. Euphemia and the Goth, with the Acts of Martyrdom of the Confessors of Edessa. London-Oxford: Williams and Norgate. Chabot, J.-B. 1902. Synodicon Orientale ou Recueil de synodes nestoriens publié, traduit et annoté par J.-B. Chabot d’après le ms. syriaque 332 de la Bibliothèque Nationale et ms. k. vi, 4 du Musée Borgia, à Rome. Paris: Libraire C. Klincksieck. Cumont, F. 1926, Fouilles de Doura-Europos (1922–1923). (Bibliothèque archéologique et historique 9). Paris: P. Geuthner. Dana, D. 2014. Onomasticon Thracicum (OnomThrac): répertoire des noms indigènes de Thrace, Macédoine orientale, Mésies, Dacie et Bithynie. (Meletemata 70). Paris: De Boccard. Ergeç, R. et J.-B. Yon. 2012, « Nouvelles inscriptions». Dans Zeugma iii. Fouilles de l’habitat. Vol. 2. La maison des synaristôsai. Nouvelles inscriptions, édité par C. Abadie-Reynal, 151–200. Lyon: Publications de la Maison de l’Orient.

goras, sanglier ou jeune lion (ou onagre) ?

307

Feissel, D., J. Gascou et J. Teixidor. 1997. «Documents d’archives romains inédits du Moyen Euphrate (iiie s. après J.-C.). ii. Les actes de vente-achat (P. Euphr. 6 à 10)». Journal des savants: 3–57. Fiey, J.-M. 2004. Saints syriaques. Édité par L.I. Conrad. Princeton: Darwin Press. Garsoian, N.G. (éd. et trad.) 1989. The Epic Histories Attributed to P‘awstos Buzand (Buzandaran Patmut‘iwnk‘). Cambridge, ma: Harvard University Press. Gatier, P.-L. 1994. «Une inscription latine du Moyen Euphrate». Syria 71: 151–157. Gebhardt, O. von. 1911. Die Akten der edessenischen Bekenner Gurjas, Samonas und Abibos. Leipzig: J.C. Hinrichs. Gignoux, Ph. 1986. Iranisches Personennamenbuch. ii. Mitteliranische Personennamen. 2. Noms propres en moyen-perse épigraphique. Wien: Österreichischen Akademie der Wissenschaften. Gignoux, Ph. 2003. Iranisches Personennamenbuch. ii. Mitteliranische Personennamen. 3. Noms propres en moyen-perse épigraphique, Supplément [1986–2001]. Wien: Österreichische Akademie der Wissenschaften. Gignoux, Ph., Chr. Jullien et Fl. Jullien. 2009. Iranisches Personennamenbuch. vii. Iranische Namen in semitischen Nebenüberlieferungen. 5. Noms propres syriaques d’origine iranienne. Wien: Österreichischen Akademie der Wissenschaften. Grassi, G.F. 2012. Semitic Onomastics from Dura Europos. The Names in Greek Script and from Latin Epigraphs. Padoue: s.a.r.g.o.n. Harding, G.L. 1971. An Index and Concordance of Pre-Islamic Arabian Names and Inscriptions. (Near and Middle East Series 8). Toronto: University of Toronto Press. Huyse, Ph. 1988. «Zum iranischen Namengut in Dura-Europos». Anzeiger der Österreichischen Akademie der Wissenschaften 125: 27–40. Huyse, Ph. 1999. Corpus Inscriptionum Iranicarum iii.i.1. Die dreisprachige Inschrift Šābuhrs i. an der Kaʿba-i Zardušt (škz). London: School of Oriental and African Studies. Ilan, T. 2002. Lexicon of Jewish Names in Late Antiquity. Part i: Palestine 330bce–200ce. (Texts and Studies in Ancient Judaism 91). Tübingen: Mohr Siebeck. Ilan, T. 2011. Lexicon of Jewish Names in Late Antiquity. Part iv: The Eastern Diaspora 330bce–650ce. (Texts and Studies in Ancient Judaism 141). Tübingen: Mohr Siebeck. Justi, F. 1895. Iranisches Namenbuch. Marburg: N.G. Elwert. Littmann, E., D. Magie, D.R. Stuart et W.K. Prentice. 1904–1921. Publications of the Princeton University Archaeological Expeditions to Syria in 1904–1905 and 1909. Division iii Greek and Latin Inscriptions in Syria. Section a, Southern Syria. Leyde: Brill. Nöldeke, Th. 1904. Beiträge zur semitischen Sprachwissenschaft. Strassburg: K.J. Trübner. Rey-Coquais, J.-P. 1974. Arados et sa pérée aux époques grecque, romaine et byzantine. Paris: P. Geuthner. Rey-Coquais, J.-P. 1998. «Deux stèles inscrites de Syrie du Nord». Syria 75: 193–200.

308

yon

Riedlberger, P. 2015. «Again on the Name ‘Gorippus’ – State of the Question – New Evidence – Rebuttal of Counterarguments – The Case of the Suda». Dans Corippe, un poète latin entre deux mondes, édité par B. Goldlust, 243–269. Paris-Lyon: De Boccard. Rougemont, G. 2012. Inscriptions grecques d’Iran et d’Asie centrale. Part ii Vol. 1. Inscriptions of the Seleucid and Parthian Periods and of Eastern Iran and Central Asia Inscriptions in non-Iranian Languages. London: School of Oriental and African Studies. Sartre, M. 2007. «The Ambiguous Name: The Limitations of Cultural Identity in Graeco-Roman Syrian Onomastics». Dans Old and New Worlds in Greek Onomastics, édité par E. Matthews, 199–232. (Proceedings of the British Academy 148). Oxford: Oxford University Press. Solin, H. 1996. Die Stadtrömischen Sklavennamen: ein Namenbuch. 3 vol. Stuttgart: F. Steiner. Stark, J.K. 1971. Personal Names in Palmyrene Inscriptions. Oxford: Clarendon Press. Traina, G. 2009. 428. Une année ordinaire à la fin de l’empire romain. Paris: Les Belles Lettres. Welles, C.B., R.O. Fink et J.F. Gilliam. 1959. Excavations at Dura-Europos, Final Reports v-1, The Parchments and Papyri. New Haven: Yale University Press. Williams, F. (éd. et trad.) 1987. The Panarion of Epiphanius of Salamis: Book i (Sects 1– 46). (Nag Hammadi Studies 35). Leyde: Brill. Yon, J.-B. 2013. «Weasels and Calves. Animals and Onomastics from Qaryatain to the Euphrates». Dans Animals, Men and Gods in the Ancient Near East. Papers on Archaeology and History in Honour of Roberta Venco Ricciardi, édité par A. Perruzzeto, L. Dirven et Fr. Dorna Metzger, 99–102. Oxford: Archaeopress. Yon, J.-B. 2014. «L’onomastique de Doura à l’époque parthe». Ktèma 39, 199–210. Yon, J.-B. et P.-L. Gatier. 2009. Choix d’inscriptions grecques et latines de la Syrie. (Guides archéologiques 6). Amman, Beyrouth, Damas, Alep: Presses de l’Ifpo. Yon, J.-B. et J. Aliquot. 2016. Inscriptions grecques et latines du Musée national de Beyrouth. (Bulletin d’archéologie et d’architecture libanaises, Hors Série xii). Beyrouth: Ministère de la Culture. Direction générale des antiquités.

chapter 15

A Lead Syriac Protective Talisman Sebastian Brock

I, Apollonius of the city of Tyana, by the gift bestowed upon me, made talismans, … and every evil creeping thing I have subdued with the yoke of subjugation. I have brought them all low, and enslaved them by (my) talismans.1

∵ A number of Syriac texts make mention of telesmata made by Apollonius of Tyana, and the lead inscription published here,2 described as a telesma, would appear to be an excellent example of the talismans mentioned in this text. Lead tablets and rolls containing texts of a protective or magic nature are known from the Near East in Late Antiquity in both Greek and Mandaic,3 as well as amulets on other metals in Jewish Aramaic.4 Although texts in Syriac on leather5 and on bowls6 of a generally related nature are known, the lead inscription published here in honour of Michael Macdonald, which was intended to provide protection for agricultural land against destruction by insects and wild animals, would seem to be unique.

1 From “A Discourse of Apollonius” (of Tyana), Hoffmann 1880: xvi–xviii. English translation, based on Mingana Syr. 566, in Brock 1969: 217–219. 2 With the kind permission of the owner, Elias Assad, to whom I am also most grateful for the photographs. The tablet is illustrated and briefly described in the Catalogue of his collection, Assad 2015: 28–29. 3 Müller-Kessler 1999: 197–209, and eadem 2005: 136–137. Her edition of the 35 Mandaic lead rolls in the British Museum is forthcoming. The Greek defixiones are almost all curse tablets or spells; a representative selection is to be found in Gager 1992. 4 See especially Naveh and Shaked 1985; idem 1993. 5 Gignoux 1987. Cf. also the text edited by J. Naveh (1997: 33–38). A further text of a similar character has turned up in the British Library among some Coptic documents (p.c. from V. Nersessian, former Curator of the Eastern Christian Collections). 6 Moriggi 2014.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

310

brock

The rectangular lead tablet measures c. 8× 25 cm and is inscribed on both sides, with twenty-six lines on what is taken here to be the recto, and twentyone lines on the verso. The script appears in a mirror reversed form, the inscription evidently having been in a properly readable form in the mould from which the lead plate was cast. The top has been damaged and there is further damage about eight lines down, where the plate has been slashed by a sharp object (possibly a plough, if the tablet had been buried). Apart from the damaged area, the letters are for the most part reasonably legible. The provenance of the tablet is unfortunately completely unknown; it is, however, known that the prior owner was P.H.D.S. Wikramaratna (1916–2010), a collector of oriental ceramics and benefactor of the Victoria and Albert Museum and British Museum in London, and the Ashmolean Museum in Oxford.7 The style of the script corresponds to what has been termed ‘medial estrangelo’, a later form of estrangelo that is found in literary manuscripts from the eighth/ninth century onwards, but becoming rare by the twelfth/thirteenth century apart from continuing to be found in large format liturgical manuscripts (see further below, on Script). The mention of St George indicates that the object has a Christian background, which is also indicated by the Greek name Theophilus. The Christian background is further indicated by the presence of a number of terms which are also found in Syriac liturgical texts; this suggests that the person who composed the text was himself a priest or deacon.

Script The script is a later form of estrangelo, sometimes entitled ‘medial estrangelo’, which is characterised by closed he, waw and usually (but not here) medial mim; dalat and rish are angular, and not rounded (apart from a 24). The greatest variety of forms occurs (as often is the case) with alaph and taw. Most alaphs are tipped c. 45 degrees so that the lower right arm has become horizontal (e.g. a 2, b 16); there are also a few serto forms (e.g. a 7). Tau is characterised by a left-slanting ascender, as well as a more vertical one, and the right arm descending below the base line and sometimes ending in a flourish (notably a 23); the hollow space, typical of the early estrangelo, has disappeared. Other distinctive features include: he, whose right descender sometimes curves round to the left (e.g. a 17); final kaph is very pronounced (e.g. a 22); final nun, when

7 On him, see the obituary in the Transactions of the Oriental Ceramic Society 74 (2009/10): xx– xxi.

a lead syriac protective talisman

figure 1

311

Script table

independent, begins at the top, rather than lower part of the line of writing (e.g. b 15); the left half of semkath is raised higher than the right (which is sometimes rather flattened), and is never joined to the left. The script as a whole is not dissimilar to that of Gignoux’s Syriaque i and iii (his ii would seem to have a higher proportion of serto alaph),8 and among the Syriac bowls Moriggi’s no. 10 and 24 would seem the closest parallels. Gignoux tentatively assigned his three documents to the sixth or seventh century, while Klugkist allocated the Syriac bowls with an estrangelo type of script to the seventh to ninth centuries.9 If one turns to literary manuscripts, this type of medial estrangelo does not come into common use before the ninth century, and in Hatch’s Album of Dated Syriac Manuscripts, the nearest comparable hand would seem to be that of the famous East Syriac so-called Masora, British Library Add. 12,138, written in Ḥarrān in 899,10 although there medial mim is closed and semkat is joined to the left. The matter is complicated by the fact that the more cursive hands, found in some early colophons, provide a link between the documentary hands of the three legal documents on leather from the early 8 9 10

Gignoux 1987: 3, considered ii to be by a different hand. Klugkist 1982: 214. Cf. also Moriggi 2014: 16. Hatch 1946: pl. clxvi. There is now a photographic edition of the whole manuscript by Loopstra (2014).

312

figure 2

brock

Side a reversed image

a lead syriac protective talisman

figure 3

Side b reversed image

313

314

brock

240s and the distinguishing features of medial estrangelo, as well as with early forms of serto script (first found in manuscripts of the eighth century).11 All this makes any attempted dating of the lead tablet very uncertain, but perhaps one might tentatively suggest approximately the ninth century, with the lee-way either side of at least a century or more.

Text It is not clear which side should be treated as the recto and which the verso. In the edition below, the side concluding with the reference to St George is taken as the verso. Red denotes uncertain readings. a (recto?)

‫ ]ܕ [ ܐ‬焏‫ܘܩܦ‬狏‫ܒ‬ ‫ܐ ܬܘܪܐ‬犯‫ ܐܡ‬爯‫ܫܝ‬熏‫ܡܠ‬ 焏‫ܘܩܦ‬狏‫ ܒ‬爯‫ ܨܠ̈ܡܝ‬爯‫ܘܬܪܝ‬ 焏‫]ܘܩ[ܦ‬狏‫[ ܒ‬.…]‫ܘܒܡ‬ 牟‫ ܫܒ‬爯‫ܬܐ ܕܗܠܝ‬熏‫ܕܚܣܝܟ‬ 犯‫ ܕܐܡ‬焏‫ܢ‬狏‫ ܚܝܠ‬焏‫ܓ̈ܡ‬狏‫ܦ‬ ‫ܐ‬犯‫ ܒ‬煟‫ܗ ܟ‬犯‫ܐ ܠܒ‬煿‫ܐܠ‬ 爏‫ܐܝ‬.[ ]‫ܠ‬ ‫[ ܒ‬.] ‫ ܬܘ‬爏‫ܐܝ‬犯‫ܓܒ‬ 爏‫ܝ‬焏‫[ ܡܝܟ‬.…]‫ܚ‬ 焏‫ ܒܚܝܠ‬焏‫ܓ̈ܡ‬狏‫ ܦ‬爯‫ܗܠܝ‬ 焏‫ܢ‬犯‫ ܐܡ‬燿‫ ܠ‬焏‫ܣ‬熏‫ܕܢܡ‬ ‫ܬܐ‬熏‫ ܘܚܝ‬焏‫ ܒܝܫ‬焏‫ܪܚܢ‬ ‫ܢ‬熏‫ܢ ܘܬܫܢ‬熏‫ܐܙܠ‬狏‫ܐ ܕ‬狏‫ܒܝܫ‬ 犯‫ ܕܒܣܦ‬焏‫ ܕܐܪܥ‬焏‫ܡ‬熏‫ܬܚ‬ 焏‫ ܟܡ‬爟‫ ܠܥܠ‬煿‫ܢ ܠ‬熏‫ ܬܥܠ‬焏‫ܘܠ‬ 煿‫ܪܥ‬焏‫ ܒ‬焏‫ܕܗܕܐ ܛܠܣܡ‬ 爟‫ ܬܘܒ ܠܥܠ‬煿‫ܐ ܠ‬熟‫ܚ‬狏‫ ܢ‬焏‫ܘܠ‬ 狏‫ܐ ܘܦ‬煿‫ ̈ܩܡ‬爯‫ ܕܗܠܝ‬焏‫ܟܡ‬ ‫ܕܐ‬煿‫[ ܒ‬爯‫ ܪܫܝܡ]ܝ‬爯‫ ܗܠܝ‬焏‫ܓܡ‬ 營‫ܐ ܨܠ‬煿‫ ܐܠ‬焏‫ܛܠܣܡ‬ 營‫ ܘܫܡܠ‬燿‫]ܠ[ܘܬܝ ܐܕܢ‬ 11

Healey 2000 and (independently) Briquel-Chatonnet 2001.

1

5

10

15

20

315

a lead syriac protective talisman

犯‫ܗ ܕܒܣܦ‬狏‫ܟ ܠܥܠ‬犯‫ܘܒ‬ 煿‫ܡ‬熏‫ܘܡ ܬܚ‬狏‫ ܘܚ‬犯‫ܘܛ‬ 燿‫ ܒܚܝܠ‬焏‫ ܒܝܫ‬焏‫ ܢܟܝܢ‬爯‫ܡ‬ 焏‫ܕܟܝܢ‬熟‫ ܡ‬焏‫ܠ‬

25

b (verso?)

熏‫ܓ‬犯‫ܗܡ‬ ‫ܪ‬熏‫ [ܝܣܝܢ‬.. ]‫ ܕ‬煿‫ܘܩܦ‬狏‫ܒ‬ ‫ ܐܬܪܒ‬焏‫ ܕܝܡ̈ܡ‬煿‫ܘܩ‬狏‫ܒ‬ 狏‫ ܬܠ‬爯‫ ܕܗܠܝ‬焏‫ܘܩܦ‬狏‫ ܒ‬犯‫ܒܦܝ‬ ‫ܐ ܘܬܘܪܐ‬犯‫ ܐܡ‬爯‫ܫܝ‬熏‫ܡܠ‬ 煿‫[ܘܩܦ‬狏‫ ܒ‬爯‫ ܨܠܡ]ܝ‬爯‫ܘܬܪܝ‬ 熏‫ ] [ܐ ܕܗ‬焏‫ܕܩܝܦ‬ 焏‫ܢ‬煟‫ ܘܥ‬焏[‫ ]ܙ̈ܒܢ‬牯‫ܡܫܚܠ‬ ݀ ‫ ܘ‬焏‫[ܢ‬.…] ‫ܘܪ‬煿‫ ܘܒ‬焏‫ܡܝ‬ 焏‫ܡ‬犯‫ ܘܟ‬焏狏‫ ܙܝ‬焏‫ܙܪܥ‬ ̈ ‫ܘ‬ 爏‫ ܟ‬犯‫ܗ ܕܒܣܦ‬狏‫ܥܠܠ‬ ̈ ̈ 焏‫ܗܝ ܕܬܘܦܝܠ‬熏‫ ܚ‬營‫ܡ‬熏‫ܝ‬ ‫ܐ‬犯‫ ܘܬܫ‬熏‫ ܣܝܣܝܢ‬犯‫ܒ‬ ‫ܐ‬狏‫ܪܟ‬熏‫ܗܝ ܒ‬熏‫ܒܡܪܗܛ‬ ‫ܗܝ‬熏‫̈ܒܒ‬煟‫ܘܢ ܒܥܠ‬犯‫ܕܓ‬熟‫ܘܢ‬ ‫ܐ‬狏‫ ܡ̈ܝ‬燿‫ܘܘܢ ܐܝ‬煿‫ܘܢ‬ 焏‫ ܣܡܝ‬燿‫ܩܒܝܪܐ ܘܐܝ‬ ‫ܗ‬煿‫ ܐܠ‬爏‫ܕܐ ܬܟ‬熏‫ ܢܓ‬焏‫ܕܠ‬ ‫ ܕܗܘ‬爿‫ܪܓܝ‬熏‫ܝ ܓܝ‬犯‫ܕܡ‬ 焏‫ ܡܫܝܚ‬爯‫ܟ ܠ‬犯‫ܡܒ‬ ‫ܬܐ‬熏‫ܝܡܢ‬煿‫ ܒ‬爯‫ܪ ܠ‬犯‫ܫ‬

5

10

15

20

Below the writing on side b there is a circle circumscribing a smaller circle; to the left and right of this are further smaller letters, written at right angles: To the left: probably eight letters, beginning with ‫ܐ‬, the remainder being very uncertain. To the right: [‫ܗ]؟‬. ‫[ܘܡـ‬.]

Comments on Readings a 4: the initial word bears some similarities with the small letters at the top of b; the second letter could be quph rather than mim. No sense can be made of it.

316

brock

a 5: ‫ܬܐ‬熏‫“ ܕܚܣܝܟ‬abstinence” is probably an error for ‫ܬܐ‬熏‫“ ܕܚܣܝܢ‬strengthening”. The flourish at the end of the word evidently should be taken as tau, with a barely visible serto alaph following it. a 8: the final letters are probably the end of a name of an angel; unfortunately the letter before the theophoric ending is very uncertain. a 9: the first letter seems to be gamal rather than lamad. a 10: the first character seems to be ḥeth, but could also be ʿe + yod. a 11: although not always clear, the readings seem certain. a 13: 焏‫ ܪܚܢ‬will certainly be an error for 焏‫ܪܚܫ‬.

̈ ‫ ܩ‬is an error for 焏‫ܩ̈ܡܥ‬, “amulets”. a 19: the restoration presupposes that ‫ܐ‬煿‫ܡ‬ The word break (pt/gmʾ) is a feature almost entirely absent in manuscripts before about the ninth century (and then remains rare). a 25: the kaph of the last word is written beneath the lamad (compare, probably, b 9). b 1: the mysterious combination of letters has some similarities with the initial word of a 4; another possible reading might be 熏‫ܡܩܢܓ‬. b 2: the five letters at the end of the line could also be read as ‫ܪ‬熏‫[ܣܚ‬. b 3: it is tempting to take 煿‫ܘܩ‬狏‫ ܒ‬as an error for 煿‫ܘܩܦ‬狏‫ ܒ‬but the singular suffix will not fit with the following word, for which seyame seem to be present, ̈ ‫“ ܝܡ‬seas”, is the most probable reading. The final four letters and the and 焏‫ܡ‬ first word of line 4 are baffling, and no interpretation comes to mind. b 4: the first two letters are absolutely clear. b 7: the first letter seems best taken as quph (it cannot be semkat, providing 焏‫“ ܣܝܦ‬sword”, which could have suited the context), but the name Caiaphas here would be very surprising; the interpretation thus remains puzzling. b 9: beneath the he of the last word there is a beth, perhaps intending bhur “shine”, though the verbal form is otherwise unattested.

a lead syriac protective talisman

317

b12: ‫ܗܝ‬熏‫ ̈ܚ‬will be an error for ‫ܗܝ‬熏‫ܚ̈ܝ‬. b 18: 爏‫ ܬܟ‬would seem best taken as an error for 爏‫ܬܟܝ‬, “trustworthy”.

Translation a. (1) By the might [of these three] (2) zodiacal signs, Lamb, Bull, (3) and Two Images; by the might (4) [of …;] by the m[i]ght (5) of the strength (?) of these seven (6) powerful words which God (7) spoke to his Son when he created (8) the […]-iel (9) Gabriel, come […] (10) […] Michael (11) these seven words with the power (12) of law: To you I speak, (13) evil creeping creature and evil (14) (wild) animals, that you should go and leave (15) the confines of the land of bspr, (16) and do not enter it for ever, as long as (17) this talisman (is) in his land, (18) and do not be seen by him again, for ever, (19) as long as these charms (?) and these (20) words are inscri[bed] on this (21) talisman. O God, incline (22) to me your ear,12 and bring to completion (23) and bless the produce of bspr; (24) preserve and seal13 his confines (25) from evil harm through your invincible (26) power. b. (1) ? ? ? ? (2) By the might of […] (3) by the might(?) ? ? ? (4) ? ?; by the might of these three (5) zodiacal signs, Lamb, Bull, (6) and Two Ima[ges]; by the power (7) of the ? [.] him who (8) changes [time]s and seasons,14 (9) [e.g. may there be blessed? ?] and water, […] (10) and the seed(s), olive(s), and vineyard, (11) and crops of bspr all (12) the days of the life of Theophilos, (13) the son of Sisino(s). And may you cause to reside (14) on his moneys blessings, (15) and may his enemies be kept away, (16) and become like dead (and) (17) buried people, and like a blind man (18) without a guide. Trust⟨ed⟩ (is) the God (19) of Mar Giwargis who (20) blesses us. O Christ, (21) confirm us in the faith.

12 13 14

Ps. 17:8b. Apart from the transposition of “to me” and “your ear”, the wording is that of the Peshitta. A standard term of protection also found in liturgical texts (where it has the sense of “make the sign of the cross” over something). Dan. 2:21; the Peshitta has a different form of the verb, paʿʿel instead of shaphʿel.

318

brock

Annotation a. 2: by the might of (b-tuqpa): though no exact parallel to this phraseology is found in either the bowls15 or the exorcisms in the baptismal rites, perhaps compare the adjective taqipa which features in (e.g.) Moriggi no. 6 line 2 and no. 8 line 7 (in other bowls it is the demons who are described as taqipe), and ba-šma taqipa d-alaha, “by the mighty name of God” at the beginning of the series of exorcisms in the archaic Syria baptismal rite attributed to Timothy of Alexandria.16 2–3: zodiacal signs, Lamb, Bull, Two Images: in the magic bowls the term malwaša features on one of those published by C. Müller-Kessler;17 it is the standard term for the zodiacal signs in Syriac astronomical texts. The names of the signs here are likewise the standard Syriac terms for Aries, Taurus and Gemini;18 at some point, however, tʾome, “Twins”, came to replace “Two Images”: thus “Two Images” is still used by Job of Edessa (9th century),19 but no longer in the anonymous Causa Causarum,20 perhaps of the 11th century. According to a Greek text under the name of Apollonius Aries, Taurus, and Gemini are the signs that constitute Spring, running from mid-March to midJune;21 other texts, both Greek and Syriac, simply equate these three signs with April, May and June.22 Their mention here is presumably because agricultural crops were especially vulnerable to insect infestation and damage by wild animals.

15 16 17

18 19

20 21 22

Note, however, Naveh and Shaked 1985: Amulet 4 (Hebrew), line 33 twqpw. For this see note 23, below. 2005: 100 (no. 33b), with her note, 102. She kindly draws my attention to the mention of the signs of the zodiac in a Mandaic lead roll, published by Greenfield and Naveh 1985: 98–99, pl. 21–22. For these, see the Appendix; also Gignoux 1988. Ed. A. Mingana. 1935. Encyclopædia of Philosophical and Natural Sciences as Taught in Baghdad about a.d. 817, or Book of Treasures, by Job of Edessa. Cambridge: Heffer & Sons: 235, 435 (text). Ed. C. Kayser. 1889. Das Buch von der Erkenntnis der Wahrheit oder der Ursache alle Ursachen. Leipzig: J.C. Hinrichs: 193. Ed. F. Nau. 1907. Apotelesmata Apollonii Tyanensis, in Patrologia Syriaca i.2. Paris: col. 1375. Thus the Greek Zodiologia e and y in Hübner 1983: 150, 191–192. For Syriac, e.g. Mingana Syr. 266, f. 1v, and Budge 1913: 483.

a lead syriac protective talisman

319

5–6: seven words: these will be the seven utterances of God, “Let us …”, in Genesis 1:3–24, prior to the creation of man. 9–11: evidently three angels are involved; the name of the first angel remains uncertain, possibly Raphael (cf. Moriggi 2014: no. 6, line 7 and 48, line 2). 12: law: it is unclear whether “law” is in a general sense, or refers to the Law (perhaps less suitable in the present context). to you I speak: compare the West Syriac baptismal rite attributed to Timothy of Alexandria where, in the pre-baptismal exorcism the priest addresses any demon present with the words “It is to you I speak, unclean demon”.23 13: 焏‫ܪܚܢ‬, which is meaningless, will definitely be an error for 焏‫“ ܪܚܫ‬creeping thing(s)”, which happens to fit very well with the excerpt from Apollonius, quoted at the beginning. For the combination of 焏‫ ܪܚܫ‬with ‫ܬܐ‬熏‫ ܚܝ‬compare Ezekiel 38:20. 14: go and leave: compare the baptismal rite attributed to Timothy, where the demon is ordered to “depart and leave” (焏‫ܩ ܘܢܫܢ‬熏‫)ܢܦ‬.24 15: bspr: the vocalization is uncertain (Bispar, Bospor[os]?),25 and the name does not seem to be known elsewhere. Although it looks as if it might be an Iranian name, there seems to be no suitable etymology.26 17: talisman (ṭelesma, fem.!): talismans are associated with Apollonius in a number of Syriac,27 as well as Greek texts; in some cases the term refers to pagan statues which were regarded as possessing certain magic powers.28 23 24 25

26 27 28

Section 13.xi, edited in Brock 1970: 376, 408–409 (for parallels). Section 13.xxxi (see note 23); the reverse order of verbs is found in Moriggi 2014: no. 24. The Lexicon of Greek Personal Names, edited by P.M. Fraser and E. Matthews, provides isolated examples of the name Bosporos. Vol. i, The Aegean Islands, Cyprus, Cyrenaica. Oxford: University Press, 1987: 103 (Delos?, Rhodes); Vol. iv, Macedonia, Thrace, Northern Regions of the Black Sea. Oxford: University Press, 2005: 73 (Thrace); while Vol. va, Coastal Asia Minor: Pontos to Ionia. Oxford: University Press, 2010: 101, has an example of Bisoporis (Bithynia). None of these localities, of course, would have had Syriac-speakers among their population. It does not feature in Gignoux et al. 2009. E.g. Chronicon ad annum 1234, i, edited by J.B. Chabot (csco Scr. Syri 36, 1920): 132, “this Apollonius from the city of Tyana made many talismans (ṭlysmʾ sagiʾe, masc.)”. This will be what the Uighur monk Rabban Sauma saw in Constantinople in 1287: Borbone 2009: 71 (tr.), 171, 26* (text); the Greek plural ṭlysmṭʾ is used here.

320

brock

19: charms (qamhe): the reading is not entirely clear and the translation assumes that qmhʾ is an error for qmʿʾ, which occurs frequently on the bowls. 20: inscribed (ršimin): another term found in baptismal texts, where the verb is used in connection with the baptismal candidates who are “inscribed” or “marked” with oil.29 22–23: bring to completion and bless (šamli w-barrek): both verbs are common in liturgical texts, including in the Institution Narrative of a number of Anaphoras (though there is no case of the pair occurring together).30 24: preserve and seal: both verbs are again common in liturgical texts; for “seal” (ḥtum), compare for example the baptismal rite attributed to Timothy.31 b. 7: see comments on the readings; although 焏‫ ܩܝܦ‬seems fairly certain, it is hardly likely that Caiaphas is intended here; the meaning remains unclear. 12: Theophilos (twpylʾ): the name is rarely found in Syriac texts. It is not clear what relationship he has with bspr: one possibility might be that bspr was his tenant farmer. 13: Sisino(s): again, the name (normally Sisinios) is rare. There is no connection here with the legend of St Sisinios, a version of which turns up on two magic bowls and an amulet, discussed by Naveh and Shaked.32 cause to reside (tašre): another verb commonly found in liturgical texts (where its origin can be traced back to Jewish Aramaic ašri with škinta as object).33 15: may his enemies be kept away (nezdagrun): active forms of the verb zgar feature in baptismal texts, e.g. the rite attributed to Timothy.34 29

30 31 32 33 34

E.g. section 17 of the rite attributed to Timothy (see note 23): metršem plan bašma d- … “n is marked in the name of …”; in some other rites “they are marked as sheep in the flock of Christ”. See Brock 2000: 186. Sections 11 and 15 (the imperative occurs in the latter). 1985: 111–122. See Brock 2006: 14. Sections 11, 35 (see note 23).

321

a lead syriac protective talisman

18: for ‫ܕܐ‬熏‫“ ܢܓ‬guide” in connection with the blind, see Matthew 23:16. 19: the presence of St George is not so surprising given his huge popularity over the centuries in the Middle East and elsewhere. 20: confirm us in the faith: probably a reminiscence of i Peter 5:9.

Index of Proper Names a 15, 23, b 11 犯‫ܒܣܦ‬ a 9 爏‫ܐܝ‬犯‫ܓܒ‬ b 19 爿‫ܪܓܝ‬熏‫ܓܝ‬ a 10 爏‫ܝ‬焏‫ܡܝܟ‬ b 20 焏‫ܡܫܝܚ‬ b 13 熏‫ܣܝܣܝܢ‬ b 12 焏‫ܬܐܘܦܝܠ‬

Index of Words Occurring a 22 焏‫ܐܕܢ‬ a 14 ‫ܐܙܠ‬ b 16, 17 燿‫ܐܝ‬ a 7, 12, 21, b 18 ‫ܐ‬煿‫ܐܠ‬ a 6 犯‫ܐܡ‬ a 2, b 5 ‫ܐ‬犯‫ܐܡ‬ a 15, 17 焏‫ܐܪܥ‬ b 9 ‫ܪ؟‬煿‫ܒ‬ b 14 ‫ܐ‬狏‫ܪܟ‬熏‫ܒ‬ a 3, 14, 25 焏‫ܒܝܫ‬ b 15 焏‫ܒܒ‬煟‫ܒܥܠ‬ a 7, 9, b 13 ‫ܐ‬犯‫ܒ‬ a 13, 23, b 20 ‫ܟ‬犯‫ܒ‬ a 17, 20 ‫ܗܕܐ‬ b 16 ‫ܗܘܐ‬ b 21 ‫ܬܐ‬熏‫ܗܝܡܢ‬ a 5, 11, 19, 20 爯‫ܗܠܝ‬ a 5 ‫ܬܐ‬熏‫ܚܣܝܢ‬ b 8 [焏‫]ܙܒܢ‬

‫‪brock‬‬

‫ܙܓ‪b 15 (pass.) 犯‬‬ ‫ܙܝ‪狏‬ܐ ‪b 10‬‬ ‫ܚ‪熟‬ܐ )‪a 18 (pass.‬‬ ‫ܚܝ‪熏‬ܬܐ ‪a13‬‬ ‫ܚ݀ܝ‪b 12 焏‬‬ ‫ܚܝܠ‪a 25 焏‬‬ ‫ܚܝܠ‪狏‬ܢ‪a 6 焏‬‬ ‫ܚ‪狏‬ܡ ‪a 24‬‬ ‫ܛܠܣܡ‪a 17, 21 焏‬‬ ‫ܝ‪熏‬ܡ‪b 12 焏‬‬ ‫ܟܡ‪a 19 焏‬‬ ‫ܟ‪犯‬ܡ‪b 10 焏‬‬ ‫ܠ‪熏‬ܬ ‪a 22‬‬ ‫ܡ‪熟‬ܕܟܝܢ‪a 26 焏‬‬ ‫ܡ̈ܝ‪b 9 焏‬‬ ‫ܡܝ‪狏‬ܐ ‪b 16‬‬ ‫ܡܠ‪熏‬ܫ‪a 2, b 5 焏‬‬ ‫ܡ‪̈犯‬ܗܛ‪b 14 焏‬‬ ‫ܡ‪犯‬ܝ ‪b 19‬‬ ‫ܢܓ‪熏‬ܕܐ ‪b 18‬‬ ‫ܢܛ‪a 24 犯‬‬ ‫ܢܟܝܢ‪a 25 焏‬‬ ‫ܢܡ‪熏‬ܣ‪a 12 焏‬‬ ‫ܣܡܝ‪b 17 焏‬‬ ‫ܥ‪煟‬ܢ‪b 8 焏‬‬ ‫ܥ‪a 16 爏‬‬ ‫ܥܠܡ‪) 焏‬ܠܥܠ‪a 16, 18 (爟‬‬ ‫ܥܠ‪狏‬ܐ ‪a 23, b 11‬‬ ‫ܦ‪狏‬ܓܡ‪a 6, 11, 19–20 焏‬‬ ‫ܨܠ‪a 21 焏‬‬ ‫ܨܠܡ‪a 3, b 6 焏‬‬ ‫ܩܒܝ‪犯‬ܐ ‪b 17‬‬ ‫ܩܝܦ‪焏‬؟ ‪b 7‬‬ ‫ܩܡ‪煿‬ܐ ‪a 19‬‬ ‫ܪܚܫ‪a 13 焏‬‬ ‫ܪܫ‪a 20 爟‬‬ ‫ܫܒ‪a 5 牟‬‬ ‫ܫܚܠ‪b 8 牯‬‬ ‫ܫܡܠ‪a 22 營‬‬ ‫ܫܢ‪a 14 營‬‬

‫‪322‬‬

323

a lead syriac protective talisman

b 13 (af ʿel) ‫ܐ‬犯‫ܫ‬ b 21 ‫ܪ‬犯‫ܫ‬ a 18 ‫ܬܘܒ‬ a 2, b 5 ‫ܬܘܪܐ‬ a 15, 24 焏‫ܡ‬熏‫ܬܚ‬ b 4 狏‫ܬܠ‬ a 3, b 6 爯‫ܬܪܝ‬

Appendix: Names of the Signs of the Zodiac in Syriac Names of the signs of the zodiac are found in several different Syriac texts, and the list below is based on the following selection (the letters are used in the table below to refer to these works): a b

c

d Ea/b

f Ga/b h i

35 36

Disciples of Bardaisan (3rd century?), in British Library, Add. 14,658, f. 149v (cf. F. Nau, Patrologia Syriaca i.2 [1907]: col. 513). Sergius of Reshʿaina (d. 536). On the course of the moon. Edited by E. Sachau. 1870. Inedita Syriaca. Halle, 113–118; repr. Hildersheim: Olms, 1968. Severus Sebokht (7th century). On the Constellations. In F. Nau. 1929– 1930. “Le traité sur les constellations écrit en 660 par Sévère Sébokht.” Revue de l’Orient chrétien 27: 375. Job of Edessa (9th century). Cf. note 19. Causa Causarum (11th century?). Edited by C. Kayser. 1889. Das Buch von der Erkenntniss der Wahrheit oder Ursache aller Ursachen. Leipzig: J.C. Hinrichs: 193, 217. Magic texts on Leather:35 see Gignoux 1987. Book of Medicines. Cf. Budge 1913: i, 443–444, 464–465. Treatise of Shem.36 Edited by A. Mingana. 1917. Bulletin of the John Rylands Library 4: 59–118. Bar ʿEbroyo (Barhebraeus, d. 1286). 1899. Le livre de l’ ascension de l’ esprit. Edited by F. Nau, i, 114–117. (Bibliothèque de l’ École pratique des Hautes Études 121). Paris: E. Bouillon.

Gignoux tentatively suggests a date of 6th/7th century, but the script (which is more developed than the present lead tablet) points to a date several centuries later. Charlesworth 1978 wanted to date the work to the 1st century bc, but this is most implausible (see my review, Brock 1979: 125–126).

324

brock

Earlier

Later

Aries Taurus Gemini Cancer Leo Virgo Libra Scorpio Sagitarius

emra rel. tawra omnes tren ṣalme ABDGbH sarṭana omnes arya omnes šebelta abcefghi qenšalma abef ʿeqarba a–e ghi ṣalma rabba ab41

dekra c (margin)37

Capricorn Aquarius Pisces

gadya omnes dawla omnes nune omnes

tʾome CEGabI

btulta d38 masʾata cdeg39hi gzdwm f40 kashaṭa CDEGbHI qešta Ga42

References Assad, E.M. 2015. Berule men shmayo: Pearls from Heaven. The Assad Collection of Syriac Manuscripts and Works of Art. London: the author. Borbone, P.G. 2009. Storia di Mar Yahballaha e di Rabban Sauma. Cronaca siriaca del xiv secolo. Moncalieri: Lulu. Briquel-Chatonnet, F. 2001. “De l’écriture édessénienne à l’estrangela et au serto.” Semitica 50: 81–90. Brock, S. 1969. “Notes on Some Texts in the Mingana Collection.” Journal of Semitic Studies 14 (2): 205–226. Brock, S. 1970. “A New Syriac Baptismal Ordo Attributed to Timothy of Alexandria.” Le Muséon 83: 367–431.

37 38 39 40 41 42

Thus in Jewish Aramaic (already in a text from Qumran, Cave 4 [4q318]); in Akkadian imru is both “lamb” and “ram”. Mandaic mbrʾ concurs with Syriac emra. Thus Jewish Aramaic (and Hebrew). Elsewhere in g (p. 483) masʾata is equated with qnyšlmʾ. The Middle Persian term. Ishoʿdad of Merv (9th cent.), Comm. on Matthew (ed. M. Gibson): 28–29, has both tren ṣalme and ṣalma rabba. Thus Jewish Aramaic (and Hebrew). Later in g (p. 483) kašaṭa is equated with qešta.

a lead syriac protective talisman

325

Brock, S. 1979. Society for Old Testament Study, Book List 1979. Sheffield: Society for Old Testament Study. Brock, S. 2000. “Towards a Typology of the Epicleses in the West Syrian Anaphoras.” In Crossroad of Cultures. Studies in Liturgy and Patristics in Honour of Gabriele Winkler, edited by G. Winkler, H.J. Feulner, E. Velkovska, and R.F. Taft, 173–192. (Orientalia Christiana Analecta 260). Rome: Pontificio istituto orientale. Repr. in Fire from Heaven: Studies in Syriac Theology an Liturgy. Aldershot: Ashgate, 2006, chapter viii. Brock, S. 2006. “The Use of the Syriac Versions in the Liturgy.” In The Peshitta: its Use in Literature and Liturgy, edited by B. ter Haar Romeny, 3–25. (Monographs of the Peshitta Institute Leiden 15). Leiden: Brill. Budge, E.A.W. 1913. Syrian Anatomy Pathology and Therapeutics or “The Book of Medicines”. The Syriac Text, Edited from a Rare Manuscript, with an English Translation, etc. i. Introduction. Syriac Text. London, etc.: Oxford University Press. Repr. Amsterdam: Philo Press, 1976. Charlesworth, J.H. 1978. “Rylands Syriac ms. 44 and a New Addition to the Pseudepigrapha: the Treatise of Shem, Discussed and Translated.” Bulletin of the John Rylands Library 60: 376–403. Gager, J.G. 1992. Curse Tablets and Binding Spells from the Ancient World. Oxford: Oxford University Press. Gignoux, Ph. 1987. Incantations magiques syriaques. (Collection de la Revue des Études Juives 4). Louvain-Paris: Peeters. Gignoux, Ph. 1988. “Les noms des signes du zodiaque en syriaque et leurs correspondants en moyen-perse et en mandéen.” In Mélanges Antoine Guillaumont. Contribution à l’étude des christianismes orientaux, edited by R.G. Coquin, 299–304. (Cahiers d’Orientalisme 20). Genève: Patrick Cramer. Gignoux, Ph., C. Jullien, and F. Jullien. 2009. Iranisches Personennamenbuch vii, Iranische Namen in semitischer Nebenüberlieferungen 5, Noms propres syriaques d’origine iranienne. Wien: Österreichische Akademie der Wissenschaften. Greenfield, J.C., and J. Naveh. 1985. “A Mandaic Lead Amulet with Four Incantations.” Eretz Israel 18: 97–106. [In Hebrew] Hatch, W.H.P. 1946. An Album of Dated Syriac Manuscripts. Boston: American Academy of Arts and Sciences, 1946. Repr. with a foreword by L. Van Rompay, Piscataway, nj: Gorgias, 2002. Healey, J.F. 2000. “The Early History of the Syriac Script: A Reassessment.” Journal of Semitic Studies 45 (1): 55–67. Hoffmann, G. 1880. “‘A Discourse of Apollonius’ (of Tyana).” In Iulianos der Abtrünnige. Leiden: Brill. Hübner, W. 1983. Zodiacus Christianus. Jüdisch-christliche Adaptionen des Tierkreises von der Antike bis zur Gegenwart. (Beiträge zur Klassischen Philologie 144). Königstein/Ts.: Hain.

326

brock

Klugkist, A.C. 1982. “Midden-Aramese Schriften in Syrië, Mesopotamië, Perzië en aangrenzende gebieden.” PhD diss., University of Groningen. Loopstra, J. 2014. An East Syrian Manuscript of the Syriac ‘Masora’ Dated to 899 ce, i–ii. Piscataway, nj: Gorgias. Moriggi, M. 2014. A Corpus of Syriac Incantation Bowls. Syriac Magical Texts from LateAntique Mesopotamia. (Magical and Religious Literature of Late Antiquity 3). Leiden-Boston: Brill. Müller-Kessler, C. 1999. “Interrelations Between Mandaic Lead Rolls and Incantation Bowls.” In Mesopotamian Magic: Textual, Historical, and Interpretative Perspectives, edited by T. Abusch and K. van der Toorn, 197–209. Groningen: Styx. Müller-Kessler, C. 2005. Die Zauberschalentexte in der Hilprecht-Sammlung, Jena. Wiesbaden: Harrassowitz. Naveh, J. 1997. “A Syriac Amulet on Leather.”Journal of Semitic Studies 42 (1): 33–38. Naveh, J., and Sh. Shaked. 1985. Amulets and Magic Bowls. Aramaic Incantations of Late Antiquity. Jerusalem: Magnes. Naveh, J., and Sh. Shaked. 1993. Magic Spells and Amulets. Aramaic Incantations of Late Antiquity Jerusalem: Magnes.

chapter 16

Two New Arabic Inscriptions: Arabian Castles and Christianity in the Umayyad Period Robert Hoyland

Michael Macdonald is justly famous for his studies on the inscriptions of Arabia and its adjoining regions and on the wider socio-political significance of these lapidary texts. I have benefitted enormously from his expertise over the years and as a small tribute to him and in recognition of his substantial contribution to the field I present here two little known Arabic texts from the early Islamic period.

1

A Graffito from Khirbat al-Bayḍāʾ, Modern South Syria

The rather fine castle of Khirbat al-Bayḍāʾ (fig. 1) is located in the basalt desert approximately 100km southeast of Damascus. It lies between the sites of Namara and Jabal Says, both well known for their pre-Islamic Arabic inscriptions. In a thorough study of the building Heinz Gaube (1974) argued that it too should be assigned to the pre-Islamic period, most likely constructed for one of the chiefs of the tribe of Ghassan, the most important Arab allies of the Byzantines in the sixth century ad. It was thought to be bereft of epigraphic remains, but in the course of an expedition to Jabal Says in 1999 I discovered an inscription there (fig. 2). It is a graffito rather than a building text, but it is interesting for the fact that it is precisely dated. It is etched onto one of a very large number of broken blocks that are strewn about the lava field on which the castle is built. It was located only about ten metres from the main entrance. Its surface is very rough and this must be borne in mind when approaching the text. Reading The second line of the text clearly bears the date: sanat sitt wa-ʿishrīn wa-miʾa, “year one hundred and twenty-six”, which in the era of the Hijra corresponds to ad743–744. Unfortunately the first line is not so easy to read. It would appear to have an initial bismillāh (“in the name of God”). It is possible that the next word is al-raḥmān, the first part of the standard accompaniment to bismillāh, i.e. alraḥmān al-raḥīm (“the merciful the compassionate”), but it seems impossible to

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

328

figure 1

hoyland

View of Khirbat al-Bayḍāʾ

read the second part, al-raḥīm, into the following letters. An alternative would be to read the word after bismillāh as irḥam (“have mercy on”). This would be a very common way to begin a graffito, the most usual aim of such texts being to solicit God’s indulgence. There are numerous texts on the lip of the crater above the castle of Jabal Says which have exactly this purpose and were inscribed by those spending time there (‘Ushsh 1964). However, this section of the first line after the bismillāh eludes sure decipherment. After a short gap, which ends in what looks like an isolated rāʾ though could just be a mark on the rock, there appear the letters l-ʾ-q-l-y-m. If one posits a preceding aliph the most obvious reading would be al-iqlīm (“the district”). One could link this with the suggested reading of the previous word as irḥam and interpret it as a request for God’s mercy upon the district, though I know of no other occurrence of such a sentiment in an inscription. One would rather expect the inscriber to be saying that he was stationed in the district or protected it; it is tempting to read ḥamā al-iqlīm (“he defended the district”), which would be appropriate, but would ignore the two letters before the ḥāʾ and would steal the initial aliph from al-iqlīm (assuming ḥamā to be written with aliph rather than aliph maqṣūra). Another solution would be to assume that we have the usual call for mercy on a person, and here that person has a Greek name beginning with the letter kappa or phi, which in Arabic usually

two new arabic inscriptions

figure 2

329

Khirbat al-Bayḍāʾ, graffito

gains an initial aliph (thus Polemon becomes Aflīmūn, Plato Aflatūn, Clypea Iqlībiyya, and so on), but this is perhaps clutching at straws. The next part of the first line is partially effaced, making any sure reading impossible. Then at the very end is the word fī (“in”), which is part of the dating section (“in the year …”). After the date, which takes up the majority of the second line, there occur the letters ʿayn and aliph, which may relate to the partially effaced word above it; perhaps the inscriber made a mistake in writing his name, tried to erase it and then inserted the correct letters below; or else he may have intended to write more, but realised he did not have the space, time or energy. Significance Heinz Gaube published a monograph on Khirbat al-Bayḍāʾ in 1974, presenting it as a residence of a member of the chiefly family of the tribe of Ghassan, which had enjoyed the favour of the Roman Empire in the sixth century. In the absence of any dated artefacts or historical sources, he relied on stylistic arguments and the fact that the Ghassanid family had influence in the region where Khirbat al-Bayḍāʾ was located. In an article written three decades later he reviewed his hypothesis and diluted his earlier arguments somewhat, concluding that the building could also have been commissioned by an Umayyad ruler like ‘Abd al-Malik (685–705), even though he still considered a Ghassanid chief to be a more likely candidate (Gaube 2004). It is difficult to make a convincing case for a Ghassanid owner, since we have no residential buildings that can be confidently ascribed to a Ghassanid or any Arab tribal leader from the pre-Islamic period, whereas we have numerous examples of Umayyad palatial residences (Genequand 2003). Of the seven inscriptions that mention a Ghassanid leader five relate to church buildings

330

hoyland

that had received the patronage of a Ghassanid (Qasr al-Hayr al-Gharbi and Dumayr, both north of Damascus, Samma’, south of Damascus, Tell al-ʿUmayri near ‘Amman, and Rusafa in eastern Syria1), one is a graffito of a soldier posted by a Ghassanid leader as a guard (Jabal Says, southeast of Damascus), and one is on a private house (al-Hayyat, south of Damascus) erected during the phylarchate of a Ghassanid (Genequand 2015). These inscriptions do reveal the influence of the Ghassanids in Syria, especially in the region of Damascus, and it is by no means impossible that they commissioned the construction of secular buildings in that region, but these texts do not in themselves provide confirmation of that. The inscription I discovered does not offer incontrovertible support for an Umayyad patron, since it may have been etched by a traveller rather than a resident, but it certainly adds some weight to the argument that Khirbat al-Bayḍā’ is an Umayyad foundation.

2

Kilwa, Modern North Saudi Arabia

In 1936 Nelson Glueck published a report about a site named Kilwa, which lies near Jabal Tubayq, at a crossroads of routes going from Amman to Tayma (via the Bayir Wells) and Aqaba to al-Jawf.2 It is a few miles south of the modern Jordan-Saudi border and some 100 miles northeast of Tabuk. Glueck describes two buildings placed either side of an open courtyard, another building to the south housing a cistern, a shallow reservoir, and to the northeast “numerous small circular graves and small square and rectangular cells” (p. 14). Over one of the latter Glueck noted a stone lintel with an Arabic inscription (fig. 3). Almost certainly this site is to be identified with that visited by Gertrude Bell in 1914 when she was encamped by the Jabal Tubayq: A sort of little khan on a khabra which had been walled round so as to form a birkeh—I make no doubt that it was a menzil on the road from Baʾir to Taimah. There was a Cufic inscription, dated, but alas I could not make it out with certainty. Anyhow there it is, the only ruin in these parts, good squared stones and mortar brought from far.3 1 The short inscription, “the fortune of Mundhir is victorious”, is on the apse of a basilical building just outside the walls of Rusafa. Without the inscription one would have taken it to be a church, but the inscription led scholars to think of it as a praetorium or audience hall. Most recently see fowden 1999, 149–173, who suggested that it served both functions. 2 Gleuck 1936; there was also an earlier very brief report (Gleuck and Horsfield 1933). 3 http://www.gerty.ncl.ac.uk/diary_details.php?diary_id=1095 (diary entry for 2/2/1914).

two new arabic inscriptions

figure 3

331

Kilwa, building text

Unfortunately she did not provide for us a photograph or drawing of the inscription. She says it was dated whereas that given in Glueck’s report is not dated. Either she was mistaken, or the inscription was more complete in her day, or she saw a different text. Reading Glueck consulted L.A. Mayer of the Hebrew University about the inscription and was told that it “belongs to the end of the first millennium ad” and that it reads: bismillāh ʿamala hādh … qaliyya (“In the name of God, x made this cell”). On both counts this assessment had long seemed to me to be wrong, but without good photographs nothing certain could be said. In 2010 Christian Robin told me he had visited the site and kindly allowed me to see his photographs. After I had finished working on the text, but not yet submitted it for publication, Saba Farès, who has participated in a Saudi-French prospection of Kilwa and its environs, published a new reading, though without providing much commentary and ignoring some obvious problems (Farès 2010). Since this text is very significant, it seems worthwhile according it its due and proper treatment. I will first give my own suggested reading and then discuss the various possible alternative readings and highlight difficulties. …‫ حفه اهل كلا \ تكلا من اقل‬/ ‫† بسم الل ّٰه حفره‬ † In the name of God, the people of Kallā / Taklā, from (the church / province of …?), engraved / encircled it.

332

hoyland

† Bismillāh The initial Maltese cross suggests that this inscription was made by Christians. The basmalah (“in the name of God”) is the Arabic form of the Greek phrase en onomati tou theou, which in the late sixth century became required by law to be placed at the beginning of all written texts. ḥafarahu / ḥaffahu The first letter is a straight downward stroke at 45 degrees to the horizontal and passes slightly below the horizontal; it cannot, therefore, be an ʿayn, as Mayer had thought, but must be a j/ḥ/kh. Next comes a circular letter, so m/f/q. The last letter is a very clear hāʾ, the same shape as the hāʾ of the previous Allāh. It is very difficult to be sure if there is anything between the m/f/q and the hāʾ. Farès decided there was nothing between them and read ḥmt (‫)حمة‬, which she translates as “sanctuary”, the Arabic word for which is ḥiman (‫ى‬ ً ‫)حم‬. She offers no justification at all for why she departs from the usual writing of this word and I can find no instance where such a departure occurs.4 Two further considerations need to be taken into account here. Firstly, based on the vast corpus of Arabic inscriptions at our disposal, one would expect our inscription to start, after the basmalah, with either a verb or a demonstrative pronoun (hādhā / ‘this’), and not simply with a noun. Secondly, the letter m in the initial word bi-sm and in the later min sits in the middle of the base line, whereas the circular letter in the middle of the word we are discussing sits on top of the baseline; it is therefore more likely to be f/q than m. If one decides that there is no letter between the f/q and the final hāʾ then one could read ḥaffahu, “encircled/enclosed it”, presumably then an allusion to the making of the surrounding wall of the complex. This seems to me the best reading, though there is possibly a small downward stroke after the f/q, which one might take to be a rāʾ, and then a break in the base line before the final hāʾ. This would then give the reading ḥafarahu, “engraved it” (referring to the inscription and/or the accompanying cross) or “excavated it” (referring to some nearby construction work). Ahl Kallā / Taklā The reading ahl seems clear. The following word must then be the name of a place or a group/tribe, the sense evidently being “the people of x”. It would

4 She cites the Arabic dictionary, the Lisān al-ʿarab, for her translation (Farès 2010, p. 244), but for ‫ حمة‬the Lisān just gives the options of (from the root ḥmy) ḥuma/“venom” and (from the root ḥmm) ḥamma/“hot spring” and ḥumma/ “vehemence” and “blackness”.

two new arabic inscriptions

333

seem to end in a joined lām and aliph. The preceding letter is either a kāf or a dāl/dhāl. The question is does anything come before that? There is a short horizontal stroke, which is in alignment with the base line of the word, but there is no clear vertical stroke attached to it; perhaps it was just a false start for the k/d/dh? In this case one might read kallā and assume a connection with Greek kella / Latin cella, used to refer to the cells inhabited by monks, for it is plausibly on such a structure that this inscription is placed. Thus the inhabitants were the people of “cells”, and this perhaps lies behind the modern place-name of Kilwa. If one prefers the other option, that some letter precedes the k/d/dh, the most likely is a single tooth of the letter b/t/th/n/y. One possibility would be Thekla,5 which would be rather apt, as Thekla was for most of her life an ascetic living in a cell in the mountains of Syria and was revered as a saint across the Levant by the fifth century ad (Festugiere 1968). Min …? The next word, min (“from”), is very well and deeply carved. The first letters of the following word are also clear: aliph–m/f/q–lām–b/t/th/n/y. There then ensues a series of short vertical strokes, which could be combinations of b/t/th/n/y, or could include a sīn (and one might then read iqlīsiya, for ekklésia, “church”). Over the first two strokes there is a hint of a horizontal stroke over them, forming the circle of the letter m, so that one could read iqlīm (“district”). The latter reading is Farès’ choice, identifying it with “the name of a region near Damascus”, citing the thirteenth-century Muslim geographer Yāqūt al-Ḥamawī. Since the stone is badly chipped at this point, it is also possible that another word originally followed and we could then understand “from the district of x” (or “from the church of x” if we read iqlīsiya). Finally, one could think of a place name that would fit the postulated letters, such as iqlībiyya on the Tunisian coast (ancient Clypea), but this would be somewhat arbitrary. Date Turning now to the second problem of Mayer’s assessment of this inscription, the date, “the end of the first millennium ad” seems to me much too late as regards both the script and the context (a Christian building in Arabia). Glueck also showed the inscription to Fathers Vincent and Savignac, who judged that it was “no later than the ninth century”. As an absolute terminus ante quem this is perhaps reasonable, but it seems to me that a date around the late

5 This reading is favoured by Farès 2010, p. 244, though again giving no discussion (“la lecture de ce nom est claire, elle ne présente aucune difficulté”).

334

figure 4

hoyland

Arabic inscription from Qasr Burquʾ, modern northeast Jordan

seventh/early eighth century would suit better the script. The style of the letters is quite similar, for example, to the Burquʾ inscription of ah 81/ad 700 from northeast Jordan, especially the lām-aliph (fig. 4).6 And this receives some support from Glueck’s own observation that the basalt buildings resemble the buildings of the sixth–eighth centuries ad that are found in southern Hauran (p. 9). As for the function of the building complex, some scholars had taken it to be a hostel or staging post of some sort, an idea suggested by its strategic location on a crossroads between an east-west road (al-JawfʿAqaba) and a north-south one (ʿAmman-Tayma), whereas Glueck favoured a “monastic settlement” (p. 16). The latter option is supported by the SaudiFrench team, which uncovered a church at Kilwa during excavations in 2010. It could of course serve both functions; indeed, many monasteries offered accommodation. Further clarification, as with the reading of the inscription, will have to await the publication of the archaeological prospection currently being undertaken by the Saudi-French team. Significance This text has some relevance to both Christian Arabic and Arabian Christianity. Christian Arabic inscriptions (in the sense that they are inscriptions made by and for Christians) provide us with the earliest examples known so far of the use of the Arabic script. We have earlier Arabic inscriptions, but they are written in other scripts, most commonly Nabataean Aramaic. These Christian Arabic texts have been found in two main areas: southwest Arabia and Syria. The former group are very simple texts (robin 2014), mostly comprising names accompanied by crosses, which gives them the appearance of grave markers, and indeed one specifies the death date of the person. A few bear dates and, if we assume that the era used is that of the province of Arabia,7 they belong to the 6 Gruendler 1993, pp. 18–19. This is noted also by Farès 2010 245, who adds that “les céramiques trouvées à Kilwa, dans la fouille et en context, appuient cette datation” and notes that a report on the ceramics is being prepared by Marie-Odile Rousset. 7 It seems likely since it yields a plausible date for these texts, namely the late fifth/early sixth

two new arabic inscriptions

335

late fifth and early sixth century. The latter group, from Syria, are more varied: names appended to a Greek-Syriac inscription (dated to ad 512) on a church lintel from Zebed, southeast of Aleppo, a graffito (dated to 527) of an official recording his dispatch by an Arab Christian “king” allied to Rome on a rock from Jabal Says, southeast of Damascus, and an Arabic-Greek inscription (dated to 567) on a church lintel at Harran, south of Damascus, commissioned by a local Arab Christian “phylarch” (Fisher 2015, 410–415, by Michael Macdonald). The fact that all the pre-Islamic Arabic inscriptions in Arabic script are from a Christian context suggests that there is some connection between Christianity and the emergence of the Arabic script. As I have pointed out elsewhere, two particular developments in the region of Syro-Arabia in the late Roman period are likely to have contributed to this emergence, namely the activity of Christian missionaries among Arabophone communities and the rise of an Arabophone Christian provincial elite (Hoyland 2008). At the very least these two groups gave an impetus to the use of the Arabic script, which might well have accelerated its evolution from the Nabataean Aramaic script. One could possibly go further and argue that the fact that these two groups were accustomed to writing in Syriac played a part in this evolution.8 The two groups of pre-Islamic Christian Arabic texts discovered so far are located at a distance of at least 1800 kilometres from each other. The Kilwa text helps to bridge this spatial gap; it is from a later date, but it is reasonable to assume that Arabic had been used for writing there—and very probably in nearby communities of northwest Arabia—for a while. We may, therefore, need to revise the commonly held view that the Christian Arabic religious texts that survive from the eighth-century Levant were written by Christians who had only recently switched to writing in Arabic as a result of increasing conversion to Islam and the dominance of Arabic in the bureaucracy of the Umayyad Empire.9 It is perhaps more plausible that Arabic was already one of

century, which fits with what we know of the emergence of the Arabic script. This means that by this time the era of the Roman province of Arabia was recognized and employed far beyond its core area (northwest Arabia, southern Jordan/Palestine). 8 The Arabic letters are clearly derived from Nabataean Aramaic (Nehme 2010), but the question is whether the evolution towards what we call the Arabic script was an entirely internal development or involved any external influence, such as the conventions/aesthetics of the Syriac script(s). 9 Griffith 1988, e.g. p. 3: “it was at this time [750–850] that conversion to Islam must have become more appealing to upwardly mobile Christians. So it is not surprising that it is also to this era, the first Abbasid century, that one must date the appearance of the earliest Christian Arabic literature”.

336

hoyland

the written languages of Christianity in the Syro-Arabia region before Islam. Evidence for this is slim, but there are some hints: for example, the famous Christian scholar Jerome (d. 420), who spent a substantial portion of his life in Palestine, speaks of “the Arabic language” in the context of his translations of the books of Daniel and Job, and an Arabic gloss on the Psalms surviving on a papyrus in the mosque of Damascus may be of pre-Islamic date.10 It also seems reasonable to argue that the Qurʾan’s deep acquaintance with Biblical material and the sophistication of its discourse on monotheism presumes the existence of a corpus of monotheist literature in Arabic in west Arabia. It is usually argued that such a corpus could only have circulated orally in the pre-Islamic period (Griffith 2013, 53, and El-Badawy 2013, 8), but if we have Christians in Arabia writing Arabic inscriptions in Arabic script is it not also plausible that they were using it to write other sorts of texts, including religious ones? As regards Arabian Christianity, the Kilwa text suggests that it was not extinguished by the caliph ʿUmar i (634–644), as the Muslim tradition would have us believe, and did not suffer an immediate decline after the Arab conquests, as modern scholars had tended to think. Rather, as Rob Carter has recently and persuasively shown with regard to East Arabia, Christianity flourished in Arabia until at least the ninth century. This he demonstrates on the basis of the archaeological record, since textual evidence is very patchy and sporadic. One expects that this is the case also for northwest Arabia, and the inscription and site of Kilwa provide some evidence for that.

Bibliography Carter, R. 2008. “Christianity in the Gulf during the First Centuries of Islam”, aae 19: 71–108. El-Badawy, E. 2013. The Qurʾan and the Aramaic Gospel Traditions. London: Routledge. Farès, S. 2010. “L’inscription arabe de Kilwa: nouvelle lecture”, Semitica et Classica 3: 241–248. Festugière, A.-J. 1968. “Les énigmes de sainte Thècle”, Comptes rendus de l’Académie des Inscriptions et des Belles-Lettres 112: 52–63. Gaube, H. 1974. Ein arabischer Palast in Südsyrien, Ḫirbet el-Baida, Beirut. Gaube, H. 2004. “Wie ist Hirbat al-Baydaʾ chronologisch einzuordnen?”, Oriente Moderno n.s. 23: 449–467.

10

Millar 2010, 76 (Jerome); Al-Jallad forthcoming (Psalms) and see also Al-Jallad forthcoming for discussion of pre-Islamic Levantine Arabic.

two new arabic inscriptions

337

Genequand, D. 2006. “Umayyad Castles: the shift from late antique military architecture to early Islamic palatial building.” In Muslim Military Architecture in Greater Syria, edited by Hugh Kennedy, 3–25. Leiden: Brill. Genequand, D. 2015. “The Archaeological Evidence for the Jafnids and the Nasrids.” In Arabs and Empires before Islam, edited by G. Fisher, 172–213. Oxford: Oxford University Press. Glueck, N. 1936. “Christian Kilwa”, jpos 16: 9–16. Glueck, N. and Horsefield, A. and G., 1933. “Pre-historic rock-drawings in Transjordan”, American Journal of Archaeology 37: 381–386. Griffith, S. 1988. “The Monks of Palestine and the Growth of Christian Literature in Arabic”, Muslim World 78: 1–28. Griffith, S. 2013: The Bible in Arabic, Princeton: Princeton University Press. Gruendler, B. 1993. The Development of The Arabic Scripts: From The Nabatean Era to the First Islamic Century according To the Dated Texts, Atlanta ga. Hoyland, R. 2008. “Epigraphy and the Linguistic Background to the Qurʾan” In The Qurʾan in its Historical Context, edited by G. Reynolds, 51–69. Abingdon: Routledge. Al-Jallad, A. forthcoming. “The earliest stages of Arabic and its linguistic classification” In The Routledge Handbook of Arabic Linguistics, edited by R. Bassiouney and M. Benmamoun. New York: Routledge. Al-Jallad, A. forthcoming. The Damascus Psalm Fragment: Middle Arabic and the Legacy of Old Higāzī. Chicago. Millar, F. 2010. “Jerome and Palestine”, Scripta Classica Israelica 29: 59–79. Nehmé, L. 2010. “A glimpse of the development of the Nabataean script into Arabic based on old and new epigraphic material.” In The development of Arabic as a written language, edited by M.C.A. Macdonald, 47–88. Oxford: Oxford University Press. Robin, Ch. 2014. “Inscriptions antique de la region de Najrān”, Académie des Inscriptions et Belles-Lettres: Comptes Rendus de l’année 2014: 1033–1128. ʿUshsh, M., 1964. “Kitābāt ʿarabiyya ghayr manshūra fī Jabal Usays.” al-Abḥāth 17: 227– 316. Yāqūt al-Ḥamawī, Muʿjam al-buldān, edited by F. Wüstenfeld, Göttingen, 1866–1873.

chapter 17

The Etymology of Ḥattā Ahmad Al-Jallad

a

Introduction

The particular isoglosses that define Arabic against the other ancient languages of the Arabian Peninsula have attracted the interest of Michael Macdonald for decades and he has made important contributions to this subject.1 Always cautioning against treating all of the non-Ancient South Arabian languages of Arabia as examples of Classical (or near classical) Arabic, Macdonald emphasized the need for establishing a linguistic definition of Arabic, rather than appealing to geography or ethnicity to identify the language. It is in honour of his pioneering contributions to the establishment of a linguistic definition of Arabic that I wish to dedicate the following study of an important Arabic isogloss: the subordinating conjunction *ḥattay. Arabic is the only Semitic language that employs the subordinating conjunction *ḥattay,2 “until” (also introducing purposes clauses and concession): Qurʾanic Consonantal Text (qct) ḥty; Classical Arabic ḥattā; early JudaeoArabic ḥattē, and Modern Arabic ḥattā, but still in some Yemenī varieties, ḥattē. The Arabic conjunction is roughly equivalent to ʿad or ʾad, “until, as long as” attested in Northwest Semitic, Ancient South Arabian, and Akkadian.3 ʿad is often accompanied by a clause nominalizer: Official Aramaic ʿd zy, Hebrew ʿad kî, Dadanitic ʿdky, and Qaryat al-Fāw4 ʿdky.5 While a reflex of *kī is attested in Arabic,6 *ʿad seems to have disappeared entirely as an independent preposi-

1 For example, the excellent discussions in Macdonald 2000; 2008; 2009. 2 This reconstruction will be justified in 3.1 below. 3 On its distribution in the Semitic languages, see Lipiński (2001: §48.17). Lipiński goes on to connect ʿad with the verbal root ʿdw/y, which in asa means “to march” and in Arabic “to speed”. Sima (1999) contents that the preposition derives from an original ʾad, which would rule out a connection to the aforementioned verbal root. 4 By this, I mean the grave inscription of Rbbl bn Hfʿm at Qaryat al-Fāw. I hesitate to assign this inscription to a specific language for the reasons explained in my 2014 article on its classification. 5 On the syntax of ʿad and various compounds connected to it, see Pat-El (2008). 6 Interestingly, it is only attested as kay, ky in the qct. The diphthongization of the final vowel

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

the etymology of ḥattā

339

tion.7 Thus, it would seem that at the Proto-Arabic stage, a new subordinating conjunction *ḥattay was innovated and replaced the common Semitic ʿad and constructions based on it—this much is not controversial based on the comparative evidence. What remains to be explained it the conjunction’s etymology and that is the goal of this paper.

b

Traditional Etymologies

The Classical Arabic conjunction ḥattā has been the subject of great attention by the Arabic grammarians and has enjoyed some discussion among modern scholars of Arabic as well.8 Sadan (2012: 199–201) presents an excellent summary of the traditional opinions regarding its etymology. Most lexicons naturally list this term under the root ḥtt, which has to do with scraping or rubbing something off. Sadan cites the example ḥatta-hū ʿani š-šayʾi, “he repelled him, drove him back”, or “turned him back, from the thing” (Lane 1863–1893: 508c) as a synonym of radda-hū, and the meaning of the verbal noun ḥattun as “quickly”. Al-Azharī states that it is a common opinion among a number of grammarians that ḥattā derives from the faʿlā noun pattern of the word ḥatt, meaning “ending something”, although he disagrees.9 Without problematizing this etymology, both classical and modern scholars have attempted to explain the meaning of “until” through appealing to different shades of the meaning of the verb ḥatta. Fleischer (1885: i.2, 403–404), for example, suggests from the meaning “to scrape off” one can arrive at the sense of to knock against or bump into something, “an etwas stossen”, which then changed to an arrival at something, “reichen bis zu etwas”, and finally to seems ad hoc. All of the other languages attest a form that goes back to *kī, as *kay in Hebrew should have yielded *kê, just as *ʿalay yields ʿalê, “on, upon”. 7 It may survive in the preposition bʿd, “on account of”, */beʿad/, attested in Safaitic (Al-Jallad 2015: 147). Once the independent ʿad was lost, the compound was reanalysed as belonging to the root bʿd, and perhaps confounded with the pre-existing preposition baʿd, “after”. 8 For an excellent overview of these studies, see Sadan 2012. 9 Sadan gives the following quotation from the Tahḏīb of al-Azharī: wa qāla baʿḍuhum ḥattā faʿlā mina l-ḥatti wa huwa l-farāġu mina l-šayʾi, trans. “and some of them (grammarians) say that ḥattā is a faʿlā (noun pattern) from ḥatt, which means to be free from a thing”. Sadan translates al-farāġu mina l-šayʾi as “ending something”, but being free of something, in line with the “scraping off” meaning of ḥatta seems more appropriate. Al-Azharī disagrees because according to him the final vowel of ḥattā cannot be raised (experience imāla), but this is demonstrably false, as both Yemenī dialects and early Judaeo-Arabic have a form ḥattē. See below under c.1.

340

al-jallad

“Erreichung eines Endpunktes” arriving at an end point.10 There is absolutely no evidence to suggest these intermediate steps for the verb ḥatta, and the entire chain of development is without parallel. Al-Sāmarrāʾī takes the meaning of *ḥattay directly from the meaning “to scrape off”, which he takes to signify uprooting and “arriv[ing] to the end, that is to the end of the matter” (apud Sadan 2012: 201). Both al-Sāmarrāʾī and Fleisher take the conjunction as a nominal pattern faʿlā, derived from the root ḥtt. It is clear that the traditional etymologies are rather farfetched and find no parallel in the world’s languages, as far as I know. While typology cannot dictate etymology, there remains no internal evidence for the grammaticalization of this conjunction from the verbal root ḥtt through any of the proposed paths of semantic change.

c

New Proposal

In 2014,11 I suggested in passing that *ḥattay was an Arabic innovation, derived from a compound of the noun ḥadd, “limit, edge, border” and an original subordinating conjunction *tay/*tī. I would like to flesh out this proposal in full here. There are three reasons to prefer this etymology over the ones proposed above. 1.

2.

10 11 12

While Classical Arabic only knows the form ḥattā, an etymologically transparent by-form is attested in some Levantine dialects, ḥaddī-ta, “until” (Abu Haider 1979: 116). The cluster is broken up by an ī, which is found on prepositions before pronominal suffixes in other Modern Arabic dialects, for example, in Egyptian Arabic: ga ʾablī-ha b-yōm, “he arrived a day before her”.12 This suffix may in turn be related to the ē vowel added to prepositions of nominal origin in other Semitic languages, e.g. Aramaic qŏdāmēhon, “before them”; Gəʿəz tāḥteka, “under you” (Lipiński 2001: §48.23). Compounds with the suffix *tay are attested, such as kīf-ta, “how”, lēš-ta, “why” (Abu Haider 1979: 116), and perhaps even at the Proto-Semitic stage in *matay, “when”. These constructions support the parsing of ḥattā as a more primitive element + *tay.

Translations by Sadan 2012: 200. Al-Jallad 2014: 6 and note 28. Quotation adapted from Hinds and Badawi (1986: 684a).

the etymology of ḥattā

3.

341

While the grammaticalization of the word for “until” from a word meaning “boundary, limit, edge” is logical, it is rather uncommon in the world’s languages. However, unlike the “scraping off” examples, it does occur in several languages of East Africa, most notably in Swahili mpaka, “until” < “boundary”.13 In the Indo-European languages, words for “until” usually consist of compounds signifying an “end point”, e.g. English “until” consists of un, “as far as, up to” and till, “to”, from Proto-Germanic *tila, cf. German Ziel, “limit, end, goal”!

Point one proves a connection with the word ḥadd, “border”, rather than derivatives of the root ḥtt, and the second two points provide further support for this derivation. However, two outstanding questions regarding the form ḥattā remain. The first is its phonological reconstruction, namely, the sound changes that take an underlying ḥadd+tay to ḥattā everywhere. The second issue is the identification of the second element. c.1 Reconstruction The spelling with the alif-maqṣūrah in the qct indicates that the final vowel was not a pure ā but rather originally a diphthong *ay or possibly *ī. I have argued extensively elsewhere that the alif-maqṣūrah represented only original diphthongs and triphthongs and did not collapse to /ā/ until after Arabic orthography was fixed.14 The vast majority of spoken dialects exhibit a form with a final ā, ḥattā, as in Classical Arabic, and this can be explained through the collapse of the word-final diphthong *ay to ā. A similar sound change operates in the reflexes of the preposition “on, upon” *ʿalay, which gives ʿalā in Classical Arabic, but ʿalay- in forms with suffixes and *ʾilay, “to, towards”, Classical Arabic ʾilā.15 Reflexes of *ḥattay that continue to exhibit a non-ā reflex of the final diphthong can be found in Yemen, for example in the dialect of an-Naḏ̣ īr: ḥattē (Behnstedt 1987: 243).16 In the vocalized Middle Arabic documents from the Cairo Genizah, dated to the high Middle Ages, we find an /ē/ reflex in the

13 14 15 16

I thank my colleague Professor Maarten Mous for this example. An article on the problem of mpaka in East Africa is currently under preparation by Professor Mous. For example, in Al-Jallad 2014: n. 32, and especially in Al-Jallad 2017: §5.1; also relevant to this question is the Graeco-Arabic inscription a1 (Al-Jallad and al-Manaser 2015). The exception seems to be the construct dual, but this could be the result of paradigmatic levelling from the unbound form, -ayni. The final ē here must come from a diphthong rather than a long ā, which does not

342

al-jallad

final syllable as well: ḥattē yiftaḥ ʿaynu, “until he opens his eye” (Khan 2011: 827). Thus, the combined evidence of Arabic orthography and modern dialectal forms that continue to exhibit an ē-reflex that cannot be explained through raising require a reconstruction of the conjunction at the Proto-Arabic stage as *ḥattay. The second issue regards the transformation of the underlying /ḥaddtay/ to /ḥattay/. The uncontracted form is not known; the earliest document attesting this conjunction with certainty is the Qurʾan, where it is always spelled ḥty. One possible Safaitic example, in an uncertain context, is attested as ḥt.17 Thus, it would seem that in the earliest studiable stages, the contraction had already occurred. While this sound change is to some extent ad hoc, a comparable shift is attested in the word for “six”. This word should be reconstructed for Proto-Semitic as *sidṯu,18 and its original consonantism is preserved in early Sabaic, Minaic, and Qatabanic as s1dṯ. Almost all Semitic languages experience phonetic changes in this word: Ugaritic attests ṯṯ, which seems to imply the following regressive assimilations: *sidṯ > *siṯṯ > ṯiṯṯ. Regressive assimilation gave rise to Gəʿəz səssu from earlier *siṯṯ < *sidṯ. In Arabic and Aramaic, the form is sitt and šettā, respectively. The Arabic suggests that the final interdental shifted to a stop under the influence of the preceding /d/, but that the voicing features spread regressively. Thus, in Arabic the following changes took place: *sidṯ > *sidt > sitt. While the same development can explain the Aramaic form, Benjamin Suchard (p.c.) explains to me that it would be easier to account for the Aramaic with a Proto-Northwest Semitic *siṯṯ, as this form can explain the reflex of this numeral in all the Northwest Semitic languages. Now, what this shows is that a cluster of dt, in an isolated lexeme, can experience regressive assimilation of voice in Arabic. An original *ḥaddtay fits this category and so the transformation to ḥatt-tay is not unexpected. So then, we are left to explain the reduction of the sequence /ttt/ to /tt/. The simplification of a cluster of three consonants of an identical quality to a geminate of course seems natural and hardly requires a lengthy justification. Nevertheless, my colleague and friend Bruno Herin informs me (p.c.) that in his corpus of Salti (Jordan) Arabic, the dental geminate of pseudo-verb “to want”, badd is reduced when followed by a consonant, so badhāš [bat-ha:-

17 18

experience this change. Behnstedt (1987: 133–134) illustrates this clearly with examples such as matē, “when” (< Proto-Semitic *matay; Hebrew māṯay) versus ʾillā, “except”. The relevant part of the inscription reads: c 4384 ṣyd ḥt wqf f l-{ʾ}kl w {q}t-h, “he hunted until exhaustion, so may he eat and be protected” (Al-Jallad 2015: 164). See Fox (2003: 80) for this reconstruction (actually *šidθ for him) and cognates, and Lipiński (2001: 287) on some possible Afro-Asiatic cognates and parallels.

the etymology of ḥattā

343

ʃ], “she doesn’t want” rather than **baddhāš.19 A similar phenomenon would then have operated in the grammaticalization of the compound *ḥadd+tay into a subordinating conjunction, *ḥattay, perhaps even preceding the regressive voice assimilation. A final minor issue comes to mind, namely, whether such a construction would have required a case ending. Most prepositions of nominal origin in Classical Arabic take the accusative /a/. In this case, the contraction of ḥadd and tay would be challenged by the presence of a vowel between the two elements. It is possible that this vowel was simply lost in the grammaticalization of this conjunction, as the entire term came to be conceived of as a single element, or that it was never there to begin with. Prepositions such as *ʿalay (Classical Arabic ʿalā) and *ʾilay (Classical Arabic ʾilā), for example, exhibit caseless forms when taking pronominal suffixes. So, we find ʿalay-ka, “upon you” or ʾilay-ka, “to you”.20 If these forms took the accusative originally, then the triphthong aya would have collapsed to ā in Classical Arabic, yielding ʿalāka and ʾilāka.21 Another good example is fī, which is frozen in the construct genitive rather than inflecting based on its syntactic role. Other prepositions do not take case endings: min-ka, “from you”, rather than **mina-ka, ʿan-ka, “about you”, rather than **ʿanaka. This could suggest that ḥadd was fit into such a model, and the following *tay was treated as a suffix. Whichever explanation we choose, it is clear that the first element must have lacked a case ending when the phonological contraction occurred. c.2 The Identity of *tay Beside reflexes of *ḥattay, many Arabic dialects possess a simple conjunction ta, which has all of the same functions as *ḥattay and has been considered by most a contraction of the latter.22 If this is the case, then it would require us to explain the Modern South Arabian teh/te/ta etc., “until, till, when” (SimeoneSenelle 2011: 1100) as borrowings from Arabic. While this may very well be

19

20 21

22

I thank Dr. Herin for sharing his expertise with me. Note especially the regressive voice assimilation following the simplification of the cluster! This example provides an exact parallel to the case of ḥattā, where *dd > t. Some scholars have suggested that these endings are construct duals, at least formally, in which case the absence of case vowels would be expected. Rabin (1951: 65) points out that the “Ancient West Arabian” of Northern Yemen had forms such as ʿalāka and ʿalāha, but attributes these to the contraction of the diphthong [ai] to [æ:]. For example, Damascene Arabic (Cowell 1964): ʾaddēš baddo ta-yəḫlaṣ, “how long until it’s done” (353); mā bbaṭṭel ʾəṭlob ta-mūt, “I won’t stop pleading till I die” (358).

344

al-jallad

true, the grammaticalization scenario proposed above suggests that the second element was productive at some point in the history of Arabic. Assuming this, I will treat the Modern Arabic ta and msa tV as original and attempt to explain the derivation of *ḥattay from this starting point. Outside of msa and Arabic, a related conjunction seems to appear in the interrogative pronoun “when”, reconstructed by Weninger as either *matī or *matay (2011: 170), based on Akkadian mati, Hebrew māṯay, qct mty, and Mehri mayt. This pronoun may be broken down further to the general interrogative *ma/mā, “what” and the element *tay/*tī, which could be one and the same as Arabic and msa *tay, “until”. One can compare this to Latin quando, “when”, which is a compound deriving from quam, “how” and dō, “until”. If this reconstruction is correct, then it would indicate that the latter element was frozen in the interrogative *matay, and supplanted almost everywhere by the more transparent nominal *ʿad, with the exception of Arabic and msa, where *ʿad did not enjoy as much success. This brings us back to *ḥattay. I would argue that the original conjunction *tay, “until” was reinforced by the word for border, limit, “ḥadd”, yielding a compound strikingly similar to English “until”, with the original preposition un < Proto-Germanic *unda, “until” (Kroonen 2013: 559) and til < Proto-Germanic *tila, “planned point in time”, but Middle Low German til, “border, goal” and Old High German zil, “border, aim” (ibidem: 517).

References Abu Haider, F. 1979. A Study of the Spoken Arabic of Baskinta. Leiden: Brill. Behnstedt, P. 1987. Die Dialekte Der Gegend von Ṣaʿdah (Nord-Jemen). (Semitica Viva 1). Wiesbaden: Harrassowitz. Cowell, M.W. 1964. A Reference Grammar of Syrian Arabic. Washington: Georgetown University Press. Fleischer, H.L. 1885. Kleinere Schriften. 3 vol. Leipzig: Hirzel. Fox, J. 2003. Semitic Noun Patterns. Winona Lake, in: Eisenbrauns. Hinds, M., and S.M. Badawi. 1986. A Dictionary of Egyptian Arabic: Arabic-English. Beirut: Librairie du Liban. Al-Jallad, A. 2014. “On the Genetic Background of the Rbbl Bn Hfʿm Grave Inscription at Qaryat al-Fāw.” Bulletin of the School of Oriental and African Studies 77 (3): 445–465. doi:10.1017/S0041977X14000524. Al-Jallad, A. 2015. An Outline of the Grammar of the Safaitic Inscriptions. (Studies in Semitic Languages and Linguistics 80). Leiden: Brill. Al-Jallad, A. 2017. “Graeco-Arabica i: the Southern Levant.” In Arabic in Context, edited by A. Al-Jallad, 99–186. Leiden: Brill.

the etymology of ḥattā

345

Al-Jallad, A., and A. al-Manaser. 2015. “New Epigraphica from Jordan i: A Pre-Islamic Arabic Inscription in Greek Letters and a Greek Inscription from North-Eastern Jordan.” Arabian Epigraphic Notes 1: 51–70. Khan, G. 2011. “Middle Arabic.” In Weninger et al. 2011: 817–835. Kroonen, G. 2013. Etymological Dictionary of Proto-Germanic. Leiden: Brill. Lane, E.W. 1863–1893. An Arabic-English Lexicon. London: Taylor. Lipiński, E. 2001. Semitic Languages: Outline of a Comparative Grammar. (Orientalia Lovaniensia Analecta 8). Second edition. Leuven: Peeters. Macdonald, M.C.A. 2000. “Reflections on the Linguistic Map of Pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 28–79. Macdonald, M.C.A. 2008. “Old Arabic.” In Encyclopedia of Arabic Language and Linguistics, edited by K. Versteegh. Leiden: Brill. Macdonald, M.C.A. 2009. “Arabs, Arabias, and Arabic before Late Antiquity.” Topoi 16: 277–332. Pat-El, N. 2008. “Historical Syntax of Aramaic: A Note on Subordination.” In Aramaic in its Historical and Linguistic Context, edited by H. Gzella and M. Folmer, 55–76. Wiesbaden: Harrassowitz. Rabin, C. 1951. Ancient West-Arabian. London: Taylor’s Foreign Press. Sadan, A. 2012. The Subjunctive Mood in Arabic Grammar. (Studies in Semitic Languages and Linguistics 66). Leiden: Brill. Sima, A. 1999. “Etymologisches Zu Akkadisch Adi ‘bis, Bis zu’ (Präp. Loci et Temporis).” Archiv für Orientforschung 46: 213a–215b. Simeone-Senelle, M.-C. 2011. “Modern South Arabian.” In Weninger et al. 2011: 1073–1113. Weninger, S. 2011. “Reconstructive Morphology.” In Weninger et al. 2011: 151–176. Weninger, S., G. Khan, M. Streck, and J. Watson (ed.). 2011. The Semitic Languages: an International Handbook. (Handbücher Zur Sprach- Und Kommunikationswissenschaft 36). Boston-Berlin: De Gruyter Mouton.

chapter 18

Are Libyco-Berber Horizontal ṯ and Vertical h the Same Sign? Marijn van Putten*

In Michael Macdonald’s essay “Literacy in an Oral Environment” (2005), he explores the use of writing practices of Safaitic, and draws a comparison to the informal literacy of the Tuaregs in the Tifinagh script. As he points out, the Tifinagh letters “are thought to be descended from the letters of the LibycoBerber script, which was used in North Africa from before the Christian era. However, the latter is still not entirely deciphered and the exact process of this descent is far from clear.” (Macdonald 2005: 56). Inspired by Macdonald’s invaluable contributions to the understanding of the Ancient North Arabian epigraphy, this article presents a humble contribution to the analysis of the grammar and orthography of the Libyco-Berber inscriptions. Libyco-Berber (lb) inscriptions can be classified into two general types, which have slightly different orthographic practices.1 The first type is the horizontal script found in Dougga, in the form of longer, monumental texts (some of which are Punic bilinguals), written from right to left. The second type is the vertical script, which is mostly found in shorter, funerary inscriptions written from bottom to top. There are some orthographic differences between the two script types. The sign known as ṯ exclusively occurs in the horizontal script, and almost exclusively in word-final position.2 The horizontal script, moreover has a different sign for ṭ than the vertical script. In the horizontal script, the sign looks like a leftward facing trident, while in the vertical script it consists of 4 vertical lines of equal length next to each other. Besides this, the horizontal script quite consistently marks word boundaries.

* I would like to thank Benjamin Suchard, Lameen Souag and Maarten Kossmann for providing useful comments on a draft of this paper. 1 The transcription used for Libyco-Berber is different from the one used by Chabot (1940). The following signs are different: U = w, Ẓ = z², I = y, Ç = ṣ, T₁ = ṯ. 2 The two exceptions to this are tgyṯh (ril 3) and trbṯn (ril 6).

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

are libyco-berber horizontal ṯ and vertical h the same sign?

347

That ṯ occurs exclusively in the horizontal script and almost exclusively in word-final position is, of course, rather unusual. In this paper I wish to examine the possible phonetic value of this sign and its grammatical function in the Libyco-Berber texts collected by Chabot (1940).

1

The Phonetic Value of ṯ

Chabot transcribes ṯ as T₁ because it seems to correspond to t in Punic bilinguals. The evidence for this identification is rather limited. It is exclusively based on the lb ypmṭṯ :: Punic ypmṭt (ril 1: 1, 3). This would appear to be quite strong evidence in favour of the identification of ṯ as word-final *t. However, in other environments Punic word-final t corresponds to nothing in Libyco-Berber: lb wdštr :: Punic ʕbdštrt (ril 1: 2). Moreover, the Libyco-Berber name transcribed in Punic as mqlʔ (ril 1: 4) which does not having a direct equivalent in ril 1, because damage has made the name unreadable, should probably be equated to the Libyco-Berber name mqlṯ (ril 3: 10), and mqlh (ril 3: 5), lending further credence to the idea that the final ṭ may represent a word-final vowel.3 This irregularity of the representation of final t might be a result of a Punicinternal development. Final t appears to have been lost in Punic, while it remained written in the Punic orthography, see Kerr (2010: 125) and Jongeling (1994: 58, 67). The phonetic identification of ṯ as t is therefore highly uncertain, and it is therefore likely that the sign ṯ represents a word-final vowel.

2

The Syntactic Distribution of ṯ

Outside of the name ypmṭṯ, ṯ mostly appears on titles. Many of these titles are also attested without this ṯ, which seems to suggest that ṯ is some kind of suffix.4 The table below sums up an overview of the titles.

3 I thank Dr. Maarten Kossmann for providing me with this important observation. 4 The idea that the ṯ is a suffix is not new, e.g. Prasse (1972: 158) suggests it is a deictic and Chaker (1984: 257) suggests it’s a 3sg.m. possessive suffix. Galand (2002: 39f.) shows that neither explanation is very convincing in the context of the inscriptions.

348

gld mṣṣkw gldmṣk mwsn gz²b

van putten

Punic

ril

ṯ-suffix

ril

h-mmlkt mṣṣkwy h-ʔd[r] ḥmšm h-ʔš rbt mʔt gzby

2 2 2 2 2, 10

gldṯ mṣṣkwṯ gldmṣkṯ mwsnṯ gz²bṯ

2,11 11 11 11 11

‘king/royalty’5 ‘architect(s)?’6 ‘lord of 50 men’7 ‘centurion’8 ‘?’

On the basis of the different translation of gld and gldṯ as Punic mlk and hmmlkt respectively, Rössler (1958: 108) concludes that the ṯ-suffix, in this place must function as derivational suffix that creates abstract nouns. While this analysis is plausible for this pair, the other pairs are not easily explained in the same way. An abstract noun of ‘architect’ or ‘centurion’ does not seem particularly plausible. Moreover, to my knowledge and examination of the plates, mlk in the transcription of ril1 is completely reconstructed, and therefore there is no true opposition between mlk and h-mmlkt.9 We have to examine in which contexts the non-suffixed and suffixed titles appear. Examining this, it becomes clear that the context in which the ṯsuffix appears is conditioned by the syntax, rather than derivational semantics (Galand 2002: 38). Non-suffixed forms appear whenever the title comes first followed by a name (+ lineage), while ṯ-suffixed forms appear whenever the title follows a name:

5 Long recognised as the Berber word *a-ǵăllid ‘king’ (e.g. Rössler 1942: 299). 6 Suggested to be an agentive noun (prefix m-) + causative s- of the root ṣk ‘to build’, e.g. Marcy (1936: 30, 37). 7 Suggested by Prasse (1972: 158) to consist of *a-ǵăllid ‘king’ + m- agentive of ṣk ‘to build’, literally “head mason”. An alternative solution might be *a-ǵăllid măẓẓăḱ ‘be small’ 3sg.n., literally ‘small king/chief’, cf. Kb. məẓẓi (Dallet 1982: s.v. mzy), Awj. məššək (Van Putten 2014: s.v. mšk). 8 Suggested by many authors as an agentive noun (prefix m-) based on the verb stem of ‘to know’ *√wsn. A perfect cognate is found in Tuareg əmusăn ‘sage, someone who knows’ (Heath 2006: s.v. sn). Due to the rather conspicuous mismatch in semantics with ‘centurion’ there is room to doubt this etymology. 9 Hoftijzer & Jongeling (1995: 646f.) suggest that Punic mmlkt has lost its abstract meaning completely and just means ‘prince, king’, thus being basically identical in meaning to mlk ‘king’ (ibid.: 640 ff.).

are libyco-berber horizontal ṯ and vertical h the same sign?

349

gld ril10:2

gld fš[-------] ‘The king is Fš-’

msnsn gld-ṯ w-gyy gld-ṯ ril 2:6 ril 2:1 (punic) l-msnsn h-mmlkt bn gʕyy h-mmlkt ‘for Massinissa the King son of Gyy the King’ mwsn ril10:3

mwsn dd[---] ‘The centurion is Dd-’

ril10:3

mwsn zm[r ] w[---] ‘The centurion is Zmr son of—’

ril 3:4

sdyln mwsn-ṯ w-ṭbn mwsn-ṯ w--‘Sdyln the centurion son of Ṭbn the centurion son of—’ mṣṣkw

ril 2:9 ril2:3 (punic)

mṣṣkw mgn w-yrštn w-sdyln mṣṣkwy mgn bn yrštn bn sdyln ‘The architect is Mgn son of Yrštn son of Ysdyln’

ril10:5

mṣṣkw šmn [w--‘The architectis Šmn son of—’

ril 3:6

šmn mṣṣkw-ṯ w-mgn mwsn-ṯ w-[k]nsln ‘Šmn the architect son of Mgn the centurion son of Knsln’ gldmṣk

ril2:10–11 gldmṣk m[…] // wšyn gld-ṯ wmgn gld-ṯ ril2: 4 h-ʔd[n] ḥmšm h-ʔš mqlʔ bn ʔšyn h-mmlkt bn mgn h-mml[kt] ‘The Lord of 50 men is: Mqlʔ son of Šyn the king, son of Mgn the king’

350

van putten

ril 10: 6

gldmṣk brh [w--‘The lord of 50 men is Brh son of—’

ril 3:7–8

msbl gldmṣk-ṯ w-šmn gld-ṯ w-šyn gld-ṯ ‘Msbl the lord of 50 men, son of Šmn the king son of Šyn the king’

gz²b ril2:9 ril2:3 (punic)

gz²b mgn w-šfṭ gzby mgn bn šfṭ ‘The gzb is Mgn son of Šfṭ’

ril 10:7

gz²b šfṭ [w--‘The gzb is šfṭ son of—’

ril 3:9

zmr gz²b-ṯ w-rš w-tnkw ‘Zmr the gzb son of Rš son of Tnkw’

An Exception: šfṭ The title šfṭ, which is a Punic loanword (Hoftijzer & Jongeling 1995: 1182), does not receive a ṯ-suffix when it follows a name. Being a loanword, it may have not been integrated enough to receive the Libyco-Berber ṯ-suffix. There are no examples where the title precedes the name. ril 2:6 ril 2:1

msnsn gld-ṯ w-gyy gld-ṯ w-zllsn šfṭ l-msnsn h-mmlkt bn gʕyy h-mmlkt bn zllsn h-šfṭ ‘For Massinissa the king son of gyy the king, son of zllsan the magistrate’

An n-Suffix The title gldgyml does not receive a ṯ-suffix, but rather a suffix n in the same environment where other titles get a ṯ-suffix (Galand 2002: 44 f.). An explanation for this variation is not forthcoming. ril 2:10 ril 2:4 (punic)

gldgmyl (sic!) zmr w-msnf w-šmn gldgyml zmr bn msnf bn ʕbdʔšmn ‘The gldgyml is Zmr son of Msnf son of Šmn.’

ril 3:10–11

mqlṯ gldgyml-n w-mzdkh mwsnṯ w-yrštn gld-ṯ ‘Mqlṯ the gldgyml son of Mzdkh the centurion son of Yrštn the king’

are libyco-berber horizontal ṯ and vertical h the same sign?

ril 5: 5

knsln gldgyml-n w-knzdṯ ‘Knsln the gldgyml son of knzdṯ’

ril 11: 8

ṭbn gldgyml-n w-šnk mwsn-ṯ w-bl[l] ‘Ṭbn the gldgyml son of Šnk the centurion son of Bll.’

351

An h-Suffix The word mwsn ‘centurion’ occurs once in the form mwsnh where it is followed by two names. ril 2:8 ril 2:2

mwsn-h šnk w-bny d-šfṭ w-m[--- // w-tnkw rbt mʔt šnk bn bny w-šfṭ bn ngm bn tnkw ‘The centurions are Šnk son of Bny and Šfṭ son of M-/Ngm son of Tnkw’

Rössler (1958: 107) suggests that this h-suffix is the suffix of the status indeterminatus masculine plural. There is rather little evidence for Rössler’s tripartite system of status indeterminatus/status determinatus/status demonstrativus, and therefore it does not seem a particularly useful avenue of investigation, especially as there is no corresponding system attested in any of the modern Berber languages. Rössler’s suggestion that the h-suffix could be a plural suffix seems probable. It however does not seem necessary to assign this suffix a separate grammatical function from the plural suffix -n that is presumably present in ril 1:6 nbb-n ‘artisans’ and ril 1:7 nb[-]-n ‘smiths’, cf. Tuareg e-nhăḍ pl. i-nhăḍ-ăn ‘smith’. The modern Berber languages have a highly irregular system of plural formations, therefore this h-suffix could simply be another plural formation without a separate grammatical function. The obvious candidate for a parallel Berber plural formation is Prasse’s Plural 7 (Prasse 1974: 59f.) which suffixes *-aʔ and turns preceding non-high vowels into high vowels, e.g. pb *ta-ʔmar-t pl. *ti-ʔmir-aʔ, cf. Tu. tămart pl. timir ‘beard’, Ghd. tomărt pl. tmiro; Zng. taʔmmart pl. təʔṃṃuraʔ; Fig. tmart pl. timira.10

10

Note however that if the identification of mwsn as əmusăn ‘knower’ is correct, the modern Tuareg plural form just has -ăn: əmusăn pl. imusăn-ăn.

352 3

van putten

Understanding the ṯ-Suffix

Now that we have examined the syntactic behaviour of the ṯ-suffix we may examine what its exact grammatical function may be. The ṯ-suffix exclusively shows up whenever the title directly modifies a name. It therefore appears to be some kind of attributive marker, or perhaps a relative marker ‘name who is title’.11 Such a marker for purely nominal attributes is absent in modern Berber languages, and it is therefore difficult to decide what its origin, phonetic value or exact form would have been.

4

The Absence of the ṯ-Suffix in Vertical Script

The function of the ṯ-suffix in horizontal script is is clearly some form of attributive marker. The sign ṯ, however, is completely absent in vertical script. Considering its rather common occurrence and clear grammatical function in the horizontal script, this is surprising. Vertical inscriptions often have a structure ‘name word son of name word’ or, ‘name son of name word (word, word)’. Considering that these are funerary inscriptions, it seems likely that the words that follow names should be considered titles of some kind. If we examine some of the most commonly returning words that follow names (Chabot 1940: xv), it is striking that they generally end in the sign h: rṣh, mdyth, mnkdh, mskrh, mswh, nm·nh nmgnwh, nbybh, nfzyh, nnbyh, nndrmh, nšfh, nz²dbh Most of these are indeed likely to be titles, as all but rṣh appear to be agentive nouns, a formation that lends itself well for creating titles. In Berber, the m/n- prefix marks agentive nouns; the alternation between the two allomorphs is phonetically conditioned by whether there is a labial *f, *b, *β or *m later in the word (Prasse 1974: 67). As we can see, the titles that have an initial m/n perfectly match this distribution, which strongly suggests these can be identified as being related to the Berber agentives. The presence and function of the h in vertical script therefore seems to be identical to the ṯ in horizontal script. Below follow some examples:

11

A similar analysis was already suggested by Galand (2002), who identifies the ṯ-suffix as

are libyco-berber horizontal ṯ and vertical h the same sign?

ril 12 Punic

mstrt[-] rṣ-h w-tkmls msw-h mʕstrt ‘Mstrt the rṣ son of Tkmls the msw.’

ril 20

zmrywn w-msryr[n] nz²db-h ‘Zmrywn son of Msryrn the nz²db.’

ril 31 Punic

bhnh msw-[h] w-yfdt mwl-h mdyt-h mnkd-h … l-bʕlḥnʔ bn yfdʕt h-mdytʔ … ‘Bhnh the msw son of Yfdt the mwl the mdyt the mnkd’

ril 72 Punic

ygwknh w-knrdt nnby-h l-ygwʕkn bn knrdʕt bn msyʕln … ‘Ygwknh son of Knrdt the Nnby’

ril 85 Latin

fwsth w-srnh n[b]yb-h msw-h mnkd-h [L. F]austus Asprenatis f. N. tr. ..t.ici Vix(it) annis lxxv ‘Faustus son of Asprenas the nbyb, the msw, the mnkd.’

ril 97

mrgm w-swrh mskr-h ‘Mrgm son of Swrh the mskr’

ril 121

tmy w-byn nm·n-h ‘Tmy son of Byn the nm·n’

ril 252 Latin

kndyl w-mšgt nšf-h chindial wisicit f tribu misicir vix an xxxx ‘Chindial son of Mišigit the nšf ’

ril 280

kndyhl w-šdh nfzy-h ‘Chindial (?) son of Šdh the nfzy’

ril 418

brtrzrn w-krdd nmgnw-h ‘Brtrzn son of Krdd, the nmgnw’

353

the Berber direct object clitic -t. This relies on its identification as as the phoneme /t/. As already pointed out, there are reasons to doubt such an analysis.

354

van putten

ril 204

yftn w-ms[y]h nndrm-h ‘Yftn son of Msyh the nndrm’

There are two cases where these titles have the Berber feminine singular ta-…-t circumfix, which is subsequently followed by the h-suffix. The feminine counterpart to rṣh (also ril 811, 968): ril 720

[-----]n[--]gh trṣt-h wlt-s ‘-n-gh, the trṣt, his/her daughter’

The feminine counterpart to mswh: ril 80

bn-s bnt[-] tmswt-h ‘his/her tomb/wife,12 bnt the tmswt’

tglt-h is probably the feminine counterpart to gld-ṯ ‘king’, with voice assimilation of final d to the feminine t, as found in Berber today, e.g. Tashlhiyt tagllitt ‘queen’ (Sabir 2010: s.v. agellid). ril 81:

bn-s wrtfrh tglt-h ‘his/her tomb/wife Wrtfrh the queen’

h in Horizontal Script If the horizontal ṯ sign is equivalent to the vertical h, this leads us to the question what the horizontal h sign stands for. The horizontal and vertical h signs look identical, and therefore their initial identification as the same sign is attractive. None of the items with a final h in the horizontal script occur in the vertical script with h. The function of this final h in horizontal script is unclear. There are three words with a final h, which might reliably be read as Berber: ril 1:6 ril 1:7 ril 2:8

12

n-šqrh Punic šyr ‘wood’, cf. Berber asɣar ‘wood’ n-zlh Punic brzl ‘iron’, cf. Berber uzzal ‘id.’ mwsn-h ‘centurions’ has been analyzed as having an -h as a plural suffix.

Chabot (1940: xvi) argues that bns stands for ‘his/her tomb’, while Rössler (1958: 109) argues that bn-s in some contexts, including those of ril 80 and 81 must be read as ‘his wife’ rather than ‘his stone’. The Proto-Berber *βăn(n) ‘wife’ attested in Tuareg hănni ‘my wife’, hănnis ‘his wife’ and Chaouia in (ʕəmmi) ‘the wife (of the paternal uncle)’ (Kossmann 1999: 99) is a convincing cognate.

are libyco-berber horizontal ṯ and vertical h the same sign?

355

Three other words have a final h which is not well understood: ril 3:1

bnmzbkh ‘stone of the altar’?, Rössler (1958: 109) suggests bn ‘stone’ + a Libyco-Berber rendition of the Punic mzbḥ ‘alter’. ril 3:12 tgyṯh ‘the work’, translated by Rössler (1958: 111) as such, with no further explanation for this translation (a similar translation in Chabot 1940: 5, also without comment). ril 7 bz³n twnth occurs as a phrase throughout this inscription. The meaning is unclear.

Three instances of h are found in broken contexts: ril 4:1 [-]rfth, ril 6:3 [-]bšwšh, ril 6:5 [-]mhy. The sign occurs twice in two words that are otherwise unanalysable, where the parallel Punic text is understood, but the Libyco-Berber text is unclear. It does not seem to contain the word *măraw ‘ten’ or *asəwwas ‘year’.13 sbsndh gldṯ sysh gld mkwsn ril 2:7 ril 2:1 (punic) b-št ʕsr š[mlk] mkwsn ‘In year 10 of the king Micipsa’ (translation of Punic) h occurs in a word final position for three names. ril 3:5 mqlh, ril 4:3, 7, 10: brh and ril 11: kkh. If mqlh is indeed equivalent to the name mqlṭ (ril 3:10) as suggested above, then this would constitute evidence that the horizontal script ṯ is equivalent to the horizontal script h.

5

Conclusion

In this article I have argued that the sign ṯ, unique to the horizontal libycoberber script, is functionally identical to the sign h of the vertical script. The horizontal ṯ sign and the vertical h sign are both used as suffixes that modify attributive titles that follow a name. These suffixes therefore have the same function, and if the language of horizontal and vertical script represent the exact same language, we should consider horizontal ṯ the phonetic equivalent of vertical h. 13

Zyhlarz (1932–1933) suggests sbs is to be analysed as a reflex of *asəwwas with a shift *ww > bb, similar to the development we find in the Zenaga cognate äššäbbäš ‘year’ (TaineCheikh 2008: 476). The rest of the phrase remains difficult to analyse, so this interpretation is far from certain.

356

van putten

The actual phonetic value of the horizontal ṯ and vertical h remains uncertain. The only cases where we have h in bilinguals with Punic, it is unclear what h represents: ril 31 Punic

bhnh msw-h w-yfdt mwl-h mdyt-h mnkd-h l-bʕlḥnʔ bn yfdʕt h-mdytʔ … ‘For Baʕlḥannō the msw, son of yfdt the mwl, the mdyt the mnkd’

lb h corresponds to Punic ʕlḥ, and Punic ʔ. Presumably, this spelling renders the hypocoristic variant bannō (Jongeling 1994: s.v. banno). The other bilingual inscription that has lb h, does not have a reflex of h in the Punic at all: ril 72 Punic

ygwknh w-kndrt nnby-h l-ygwʕkn bn kndrʕt … ‘For Ygwkn[h] son of Kndrt the nnby’

The sign identified as h in the horizontal script, cannot readily be equated to the vertical h, and might have an altogether different phonetic value. Only if we accept the names mqlh and mqlṯ attested in ril 3 to be the same name, we may conclude that ṯ and h were allographs of the same sign in the horizontal script.

Bibliography Chabot, J.-B. 1940. Recueil des Inscriptions Libyques. Paris: Imprimerie Nationale. Chaker, S. 1984. Textes en Linguistique Berbère. Introduction au domaine berbère. Paris: Editions du centre national de la recherche scientifique. Galand, L. 2002. L’indication des titres et des fonctions en libyque. In: L. Galand. Études de linguistique berbère. Leuven & Paris: Peeters, pp. 38–47. Heath, J. 2006. Dictionnaire Touareg du Mali: Tamachek-Anglais-Français. Paris: Karthala. Hoftijzer, J. & K. Jongeling. 1995. Dictionary of the North-West Semitic Inscriptions. Leiden, New York & Boston: Brill. Jongeling, K. 1994. North-African Names from Latin Sources. Leiden: Research School cnws. Kerr, R.M. 2010. Latino-Punic Epigraphy. A Descriptive Study of the Inscriptions. Tübingen: Mohr Siebeck. Kossmann, K. 1999. Essai sur la phonologie du proto-berbère. Köln: Rüdiger Köppe Verlag.

are libyco-berber horizontal ṯ and vertical h the same sign?

357

Macdonald, M.C.A. 2005. “Literacy in an Oral Environment”. In: P. Bienkowski, C. Mee & E. Slater (eds.) Writing and Ancient Near Eastern Society. Papers in Honour of Alan R. Millard. New York & London, t&t clark. Marcy, G. 1936. Les Inscriptions libyque bilingues de l’Afrique du Nord. Paris: Imprimerie Nationale. Prasse, K.-G. 1972. Manuel de Grammaire Touaregue (tăhăggart) i–iii. Phonétique— Ecriture—Pronom. Copenhague: Akademisk Forlag. Prasse, K.-G. 1974. Manuel de Grammaire Touaregue (tăhăggart) iv–v. Nom. Copenhague: Akademisk Forlag. Putten, M. van. 2014. A Grammar of Awjila Berber (Libya). Based on Paradisi’s Work. Köln: Rüdiger Köppe Verlag. ril = Chabot 1940. Rössler, O. 1942. Libyca. Wiener Zeitschrift für die Kunde des Morgenlandes 49: 282–311. Rössler, O. 1958. Die Sprache Numidiens. In: Sybaris—Festschrift für Hans Krahe zum 60. Geburtstag am 7. Februar 1958, Wiesbaden, pp. 96–120. Sabir, A. 2010. Taknarit. Diccionario Español—Amazigh. Amazigh—Español. Taine-Cheikh, C. 2008. Dictionnaire zénaga-français. Köln: Rüdiger Köppe Verlag. Zyhlarz, E. 1932–1933. “Das Wort für “Jahr” im Altlibyschen”Zeitschrift für EingeborenenSprachen 23, 75–77.

part 2 Archaeology, History and Religion

∵

chapter 19

The Outer Wall of Taymāʾ and Its Dating to the Bronze Age* Arnulf Hausleiter**

1

Introduction

One of the most intriguing remains of the ancient oasis settlement of Taymāʾ, Northwest Arabia, is a large system of walls, surrounding the oasis and subdividing it into different spatial and most probably functional units within its 9.2 sq.km large surface. Long before Western travelers visited the oasis in the 19th century ce, their impressive remains have been noted by early Arab geographers and historians, such as 11th century ce geographer al-Bakri reporting on a wall which was a farsakh in length running along a brook (cf. Buhl— Bosworth 2000; cf. Bawden et al. 1980: 73, for a similar description by 10th century ce writer al-Istakhri; see also Jaussen—Savignac 1997: 147). Still today, the walls impact the visual appearance of the ruins of Taymāʾ within the constantly growing modern settlement (Figs. 1–2). Apart from being a physical border between different physical and spatial zones, usually considered in the context of warfare and fortification, walls embody a wider range of concepts, not only of practical function, such as

* I wish to thank Laïla Nehmé and Ahmad Al-Jallad for inviting me to join this Festschrift. Since 2010 Michael has been working with the Saudi-German expedition at Taymāʾ substantially contributing to the systematic study of epigraphic remains from the excavations (Macdonald forthcoming), the Taymāʾ Museum (Macdonald and al-Najem forthcoming) as well as from sporadic findings in the settlement (al-Najem and Macdonald 2009) or in the areas surrounding the oasis (Macdonald and Jacobs 2009). With this contribution I would like to express my sincere thanks to Michael for sharing his immense knowledge with all of us as well as for his constant interest in the epigraphy, history and archaeology of Taymāʾ, from which we greatly benefited. This includes his numerous seminal publications referring to the history and culture of Northwest Arabia which for constraints of space, cannot be mentioned here (cf., however, at least Macdonald 1995; 1997; 2004). Michael unites patience, perseverance, humor and scholarship in special way. ** The article was completed during my stay as 2015–2016 Visiting Research Scholar at the Institute for the Study of the Ancient World (isaw) at New York University.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

362

figure 1

hausleiter

Taymāʾ: The Burj Badr Ibn Jawhar, walls dividing the compounds and the outer wall with sand dunes at the inner side seen from North (Compound c) dai orient-abteilung, a. hausleiter

preventing from sand dunes, probably even flood events, but also social and economic meaning, such as manifesting segregation within the population, delimitation of different economic, religious or legal spheres as well as symbolic demarcations. Walls embody also two very opposite principles, those of impermeability and permeability. The existing walls of Taymāʾ, according to recent investigations, preserved over a total length of approx. 18.2km, constitute a highly complex system of different building units, combining different building materials (siltstone and mud bricks) and techniques, impacting their physical properties and thus functional characteristics. Taken together (and also regarding the individual sections) they constitute an example of architecture, which can be without exaggerating labeled monumental. Walls factually dividing oasis settlements in different districts, though most probably offering a certain degree of permeability, have been observed at other places in Northern and Western Arabia, such as Dumat al-Jandal (Charloux et al. 2014), Qurayyah (Parr et al. 1970: 220–225), al-Hayit (cf. Hausleiter and Schaudig 2016; Hausleiter—Schaudig, in press), to mention only those examples where excavations, surveys or other types of investigations have been carried out recently; the situation at al-Khuraybah/Dadan seems to be characterized by walls of smaller dimensions, though the building of the Hijaz railway

the outer wall of taymāʾ and its dating to the bronze age

figure 2

363

Taymāʾ: The archaeological site of Qrayyah from North (Tuwayyil Saʾid) with the Burj Badr Ibn Jawhar, Compounds a and b as well as the outer wall dai orient-abteilung, a. hausleiter

as well as agricultural cultivation may have contributed to damaging the walls (cf. Parr et al. 1970: 207). The aim of the present contribution is not so much a detailed discussion of function and meaning of the wall system for the life of the ancient oasis or other more cognitive aspects; nor are construction process, technical details, architectonic features, building materials, access points and settlement patterns etc. object of our considerations; some of these aspects have been recently dealt with elsewhere (Schneider 2010a, 2010b; 2011; 2012; 2016; forthcoming; cf. Bawden et al. 1980: 77–78; 91–92 as well as Abu Duruk 1988: 16–26, the latter summarizing previous investigations, which will not be discussed in the present contribution in detail). It is rather focusing on the construction date of the outer part of the core area of the settlement and this for three major reasons: 1)

In the context of the perception of Taymāʾ as a 1st millennium bc site, the construction of the large walls has always been dated respectively— mainly based on a mention of the building of a palace and the reinforcement of the walls in the verse account on King Nabonidus (556–539bc),

364

hausleiter

figure 3

2)

3)

Sketch plan of Taymāʾ by Charles M. Doughty doughty 1921: 287

who took residence at Taymāʾ between 552 and 542 bc (cf. Schaudig 2001); it had been constantly overlooked, however, that this text was not authored by the king himself and may include a semantic different from royal accounts of that king;1 This ‘global’ dating of the wall(s), narrowing the long-lasting history of occupation of the oasis to a short-time historical event, though of fundamental significance for the understanding of the Babylonian-Arabian relations, overshadowed the chronologically highly differentiated history of construction of the walls as outcome of the oasis’ development and the pertinent subdivision of the settlement in different parts through time. One aspect of a radical change of perspective with regard to these longterm processes is connected to the evidence for the construction of the inner wall considerably after that of the outer wall, as it has been proved by stratigraphy, associated artifacts, written evidence as well as 14C-dating (see below); Recent archaeological salvage excavations by the Saudi-German expedition and a Saudi team revealed dated evidence for late 3rd / early 2nd

1 According to H. Schaudig (personal communication 2005) these building operations at the far flung site of Taymāʾ may have been deliberately mentioned in this text in order to blame the king for the neglect of his hometown Babylon. If developing this hypothesis further, there may not have been a palace of this king at Taymāʾ either (cf. Eichmann et al. 2006b; Hausleiter 2011; 2012).

the outer wall of taymāʾ and its dating to the bronze age

365

millennium bc burials some of which containing ceremonial bronze weapons, such as fenestrated axes and several types of thrusting weapons. These new findings not only placed the dating of the beginning of a pottery attested in 17th to 14th century bc contexts, i.e. Red Burnished Ware (rbw) to the turn of the 3rd to 2nd millennium bc (Hausleiter— Zur 2016), at the same time matching the end of preceding, late 3rd millennium bc Gritty Ware (gw; Tourtet et al., forthcoming); they also changed the view on the settlement history of the oasis which emerges now as a major and significant Bronze Age settlement of little less than the maximum size of the walled oasis which, at the latest, had been reached in the Early Iron Age (12th to 9th century bc). At Taymāʾ, at this time, elsewhere know as Early-to-Middle Bronze Age transition, local burial customs are attested, which were influenced by regional communal practices known from warrior graves in Syria and the Levant, suggesting close cultural relations between these two areas (Hausleiter—Zur 2016). In this context it was of eminent interest to investigate the dating of the outer wall as result of public investment into a place which may have been of urban character.2

2

Early Investigations

In 1888 it was Charles M. Doughty, who published for the first time a sketch plan of the area of Taymāʾ (Fig. 3) nowadays known as Qrayyah (already Philby 1957: 86). Whereas his short textual description refers to the length of the walls and the relatively elevated position in respect to the modern settlement (Doughty 1921: 287), his map is more detailed, including not only the central tower, the Burj Badr Ibn Jawhar, the highest built structure preserved at the site (cf. Philby 1957: 85–86; Bawden et al. 1980: 75), but also the then two major wells of the oasis, the Haddaj and Waddaj.3 Slightly later Ch. Huber and J. Euting visited the site. The former only once mentions “l’ ancien mur d’ enceinte” (Huber 1891: 326), whereas the latter published a sketch drawing of the southern parts of the oasis (Fig. 4); he observed that Taymāʾ is, on three 2 Cf. Edens and Bawden 1989 for the discussion of oasis urbanism at the Bronze to Iron Age tradition. 3 The Waddaj has not been identified with any existing structure in one of the compounds. Whether it is the well east of Compound e (Bawden et al. 1980: 75) or the large well in Compound b (Wellbrock and Grottker in Hausleiter et al., forthcoming a) cannot been stated with certainty, even though the latter may be a more suitable candidate; see also Philby 1957: 86.

366

figure 4

hausleiter

Sketch plan of the oasis of Taymāʾ euting 1914: 148

the outer wall of taymāʾ and its dating to the bronze age

figure 5

Overall plan of the oasis of Taymāʾ in 1979 bawden et al. 1980, pl. 60

367

368

hausleiter

sides, enclosed by artificial “Erd- und Steinwälle”, which he interpreted as fortificatons (ibid.: 177); at the western branch of the wall, he identified remains of a 1m wide stone wall and a tower with rectangular ground plan (ibid.: 198). On his map he furthermore mentions an “alter Hügelwall mit kleinen Turmresten” and a “Haupt-Tor”—all west of the oasis and the site—referring to the southwestern sector of the outer wall as well as to the western branch, which he explicitly mentions as one of the place visited.4 A. Jaussen and R. Savignac in 1909 made some observations on the walls of Taymāʾ (1997: 151–153) and published the first photographs, i.e. of the southern walls of Compound a (ibid., Pl. lxiii.2; 3; Pl. lxiv.1). H.StJ. Philby (1957: 81) mentioning the ruins and the walls, apparently visited only the Burj Badr Ibn Jawhar (see above). F.V. Winnett and W.L. Reed (1970) briefly commented upon the walls, mentioning the sand dunes (ibid.: 176) and reproduced photographs of the walls separating Compounds c and d (ibid.: 24, Fig. 22), the centrally located tower known as Burj Badr Ibn Jawhar (ibid.: Fig. 23), Compounds a and b (ibid.: 25, Fig. 24), the western branch (ibid.: Fig. 25–26), making again reference to the ‘gateway’ which had been observed by Euting earlier (see above), and the southern wall of Compound a (ibid: 26, Fig. 27). P.J. Parr in surveying the area of Taymāʾ in the 1960s with G.L. Harding and J.E. Dayton closely investigated a tower at the site of Mintar bin ʿAttīya, a plateau northwest of the oasis, identifying the former as part of the oasis’ fortification system (Parr et al. 1972: 23–26; 40–46; see already Philby 1957: 70); he dated this structure to the Neo-Babylonian period (Parr 1979: 41). This location has already been visited by Ch. Huber (1891; without further description). The 1979 expedition to Taymāʾ resulted in a general description of the “city walls” (Bawden et al. 1980: 74; 77–78)5 and the first publication of a plan of the entire oasis, which represented actual built features in the right proportions (Fig. 5). In the context of overall considerations of topography, spatial differentiation and use of the ancient settlement, the different parts of the oasis were labeled “Compounds”: Compounds a–e (with subcompound a1 in the northwestern corner of Compound a; Fig. 7) are located to the South of the “Haddaj district” a term introduced by Bawden and Edens (1988: 198, Fig. 1),6 whereas

4 The Haupt-Tor (main gate) may refer to the opening of the western branch near Qasr al-Radm. 5 Cf. Schneider 2010b: 21–22 relativizing the term “city wall” in view of the results of his investigations. 6 This scale of subdividing the oasis in large units along socio-economic functions is similar to the division of the settlement by the hydrological team (cf. Wellbrock and Grottker, forthcoming): Sabkha, Oasis, Qrayyah, Modern City.

the outer wall of taymāʾ and its dating to the bronze age

figure 6

369

The area of Qasr al-Hamra and Compound w west of it bawden et al. 1980: pl. 62b

Compound w lies to the Northwest of Qasr al-Hamra (Fig. 6). Bawden et al. (1980: 75), observing the marked in-turning of the inner wall at the eastern flank of Compound e hypothesized the existence of a further compound which would have been destroyed by modern constructions. As to the functional subdivision of the oasis, Compounds e and a1 were considered of public in nature; Compounds a and b, as manufacture-related and commercial areas; the (city) walls were attributed a defensive function (ibid. 91, though with “?”, whereas near Qasr al-Hamra a gate complex was located).7 Both the general plan and the one depicting the central part of Taymāʾ (Qrayyah) seem to be based on an aerial photograph which was apparently shot in the 1960s the visible remains on it serving as blueprint for the entries of architectural remains in the schematic plan (compare Figs. 8a and 8b).8 As a

7 For the location of the other gateways, cf. Bawden et al. 1980: 77. This opening may coincide with the Haupt-Tor observed by Euting 1914: 148 (see above). 8 The photograph has been published by Abu Duruk 1988: 138 (Arabic pagination).

370

figure 7

hausleiter

The area of Qrayyah with Compounds (a-e, subcompound a1) and excavation areas dai orient-abteilung, s. lora

the outer wall of taymāʾ and its dating to the bronze age

figure 8a–b

371

(a) Plan of Qrayyah; (b) Aerial photograph (possibly composed) on which the plan is based (a) bawden et al. 1980: pl. 62a; (b) dai orient-abteilung m. cusin, courtesy municipality of taymāʾ

result, any topographic features are lacking on this plan. As stated by the same authors, the northern delimitation of Compound e and therefore its dimension are difficult to determine (ibid.). However, a recent analysis of an aerial photograph from 1956 shows traces of the former course of at least the eastern part of the northern wall (Fig. 9 inset box). If this part continued to the Northwest, one may reconstruct Compound e as a “citadel” fully surrounded by a substantial wall (Fig. 9), even more emphasizing the significance of this location within the topography of the oasis.9 In contrast to these walls, the enclosure surrounding Area e with its Nabataean temple Building e-b1 as well as a large 2nd to 4th century ad building complex known as Building e-b6 located northwest of it (Fig. 7), is considered different in character since exclusively dealing with defined architectural units rather than with entire compounds.10 9

10

For the presence of a much more articulated “citadel” at the site of Qurayyah, cf. Parr et al. 1970: 220–224 with Fig. 11 on p. 222; cf., instead the completely different spatial and topographical arrangement of the oasis of Dadan in al-Khuraybah (ibid.: 206–207). The same is valid for the enclosures of Early Iron Age Building o-b1 (Hausleiter 2014: 417, Fig. 22) and probably also for the enclosure of a well in Compound a (the enclosure itself currently labeled Building h-b1). Qasr al-Radm, enclosing a well of yet unknown date has always been considered a building rather than an enclosure (cf. the recognition of its

Northern part of Qrayyah (Compound e): Aerial photograph of 1956 (inset), excavated and surveyed structures and reconstructed part of the northeastern perimeter wall dai orient-abteilung, s. lora

hausleiter

figure 9

372

“defensive or control function” by Bawden et al. 1980: 82; more recently, Northedge 2009) in 2013 the structure was for the first time systematically recorded by an architect (Knop 2014); in this context is was suggested that the demolition of parts of the building into a camel draw should be dated to the period after the visit of J. Euting to Taymāʾ (1884), since he found the building intact (Euting 1914: 157; cf. Knop-Hausleiter 2016).

the outer wall of taymāʾ and its dating to the bronze age

373

Abu Duruk (1988) offers a brief description of the walls and its features (ibid.: 12–13) before going into details of construction, building materials and techniques and dating in the context of a thorough description in combination with his excavation report (ibid.: 16–26); he lists the Mintar bani Attiya and Tuwayyil Saʾid as “square towers” recognizing that “neither of them is attached to the city wall”;11 further rectangular structures on the “northern escarpment” he addresses as “scattered towers” aimed at probably “command[ing] the northern approaches to the city” (ibid. 12).

3

Dating the Walls of Taymāʾ

As to the dating of the walls no reliable data directly associated with the investigations of the walls had been provided. R. Jaussen and A. Savignac (1997), conscious of the lack of comparable evidence, considered the walls as preIslamic (“antérieures à l’invasion arabe”) attributing their construction to the ancient Aramaic population (ibid. 153). Philby (1957: 85–86) attributed the circular upper part of the Burj Badr Ibn Jawhar probably rightly to the medieval period; beneath that, there is a rectangular structure connected to the masonry crown of the outer wall, thus being of much earlier date (see below). Winnett and Reed (1970: 175–176), in discussing the archaeological evidence, in particular the pottery, stress the need of excavations for resolving the chronological dating of the remains of Taymāʾ. P.J. Parr et al. (1970: 223) pointed to the close resemblance of the walls of the oases of Qurayyah and Taymāʾ, but this observation did not let them express an explicit dating of the Taymāʾ wall to the 2nd millennium bc, as concisely summarized by Abu Duruk (1988: 23). Only subsequently, P.J. Parr (1979: 41) seemed to slightly prefer a NeoBabylonian date for the walls of Taymāʾ, not without referring to his earlier assertion regarding the similarity between the Taymāʾ and Qurayyah walls. Shortly afterwards, Bawden et al. (1980: 91), based on excavations at Qasr alHamra, considered a dating to the ‘Neo-Babylonian years’ (supposedly the ten years between 552–542bc) for “much of the wall, if not all of it and Compound w” (the latter directly adjacent to the al-Hamra area); however, they concluded at the same time that “the attribution of such large numbers of architectonic units to Nabonidus must remain somewhat conjectural” (ibid.) Abu Duruk (1988: 13) with reference to P.J. Parr (1979) also suggests a NeoBabylonian date of the walls, “since at no other time would the very consider-

11

Not much of these remains is visible any more today.

374

hausleiter

able resources for an enterprise of such magnitude have been available locally”. At last, A.R. Al-Ansary and H. Abu al-Hasan in a small publication on Taymāʾ published within a series of books for the wider public focusing on sites on the caravan route (2005: 42) repeat the dating of the walls to the time of the Babylonian king.12 In view of this situation and in the context of the reasons described further above, it was one of the aims of the Saudi-German project to closely investigate the wall system by means of archaeological excavations. Therefore, in the second year, after surveying various parts of the wall system, an area west of the central part of the ancient settlement characterized by comparatively narrow deposits (Area c; see Fig. 7), was selected for a transect including both outer and inner wall (nos. 2 and 1, respectively, after Schneider 2010b: 3, Fig. 1). Altogether six trenches of Area c (Squares c1; c4–5; c7; c9; c10) were directly related to the exploration of the wall system (see below). Since it was clear, that only selected parts of this complex could be investigated by this exploration method within a project whose aim was to establish the history of occupation of this important oasis within the environmental framework,13 a research project within the Saudi-German expedition, exclusively dedicated to the exploration of the city wall was launched. This first systematic investigation of the entire wall system of Taymāʾ respectively its visible remains, was carried out in the years 2006–2008 by a collaborative enterprise of the (then) Brandenburgisch-Technische Universität (btu) Cottbus, funded by the Fritz Thyssen-Stiftung (Schneider 2010a; 2010b; 2011; 2012; 2016; forthcoming), in collaboration with the Division of Building Archaeology at the dai’s head office. The project resulted in the complete recording of the surface remains by means of surveying, 3D-modelling as well as of altogether 44 archaeological soundings (w1–w44) aimed at closer investigating areas of particular interest revealing new data on building material and construction techniques; elements enabling for the reconstruction of the chaîne opératoire of the building process, functional and semantic aspects as well as relative and absolute dating of the system and its different units. In the context of these investigations, the previous estimates of length, width and height of the walls were replaced by accurate measurements (Schneider 2010b; 2016), whereas the heterogeneity of the system and its parts was confirmed (ibid.: 9; for a summary of the evidence of the outer wall cf. ibid.: 6–15). The preserved walls, accord-

12 13

The regal years are given as 555–539 bc. Cf., e.g. for a different approach, i.e. the full excavation of the wall system of the Assyrian capital Ashur, Andrae 1913.

the outer wall of taymāʾ and its dating to the bronze age

375

ing to the classification criteria were divided into different sectors which were numbered accordingly (cf. Fig. 10–11). Whereas the recording of the entire wall system of Taymāʾ according to the standards of building archaeology is owed to the above mentioned wallsproject, it were archaeological excavations in combination with geoarchaeological research methods, providing those results which were significant for the first reliable dating of the outer and inner walls of Taymāʾ. The following paragraphs offer a brief summary of the sequence of exploration of archaeological contexts providing data impacting the dating of the outer walls of the oasis (for the location of the sampling sites, cf. Fig. 12). Area a: Outer Wall Built before the Early Iron Age Excavations (2004) in a small building which had been attached to the outer wall14 apparently in the context of rebuilding the upper part of the mud brick wall with a construction of silt stones (Eichmann et al. 2006a: 103–107; Schneider 2010b: 10) revealed two occupation phases of the building which have been attributed to the Early Iron Age (12th to 9th centuries bc) by means of two 14Cdates. The earlier 14C-date (ta 142: 1189–933 calBC15) originates from a sand layer outside of Building a-b1 from its older occupation level (ol a:2) and beneath a sequence of deposits on which a wall of the younger occupation level (ol a:1) has been built; the second one is from an ash layer within a sequence of floors in the edifice belonging to the younger ol a:1 ta 190: 970–813 calBC.16 A third sample (ta 4484), aimed at verifying the previous dates, was recovered from a sand deposit beneath the building (a:3) and provided a 1016–858 calBC date,17 providing a terminus post quem for the foundation of Building a-b1 and at the same time implying that ta 142 may not have been in its original context, when discovered. In any case, the two datings obtained from secure contexts fully coincided with the occurrence of well known and well dated Early Iron

14

15 16 17

Observation of concentrations of surface pottery led to the decision to excavate in this area. The surface pottery seems to have already been identified by Bawden et al. (1980: 89). ta 142, su 154: Charcoal; analyzed by the dai-Laboratory, Berlin; sample no. Bln 5669; Radiocarbon date: 2873±31; 2σ range, probability 95.4 %; calibrated with OxCal v4.2.4. ta 190, su 133: Charcoal; analyzed by the dai-Laboratory, Berlin; sample no. Bln 5670; Radiocarbon date: 2735±31; 2σ range, probability 95.4 %; calibrated with OxCal v4.2.4. ta 4484, su 179: Charcoal (tamarix, chenopodiaceae, identified by R. Neef, Scientific Division at the Head Office of the dai); analyzed by the Leibniz-Labor für Altersbestimmung, Christian-Albrechts-Universität, Kiel; sample no. kia 30895; Radiocarbon date: 2800±25; 2σ range, probability 95.4 %; calibrated with OxCal v4.2.4.

figure 10 Walls of Tayma recorded by the walls-project courtesy p.i. schneider

376 hausleiter

the outer wall of taymāʾ and its dating to the bronze age

figure 11

377

Tayma: the walls of the oasis dai orient department, s. lora; numbers and surveying data according to the walls-project

Age Painted and Plain Wares (Hausleiter 2014: 414 with Figs. 17–22 on pp. 415– 417; Maritan et al. in press), predominantly defined by the building complex in Area o and to that of Sanaʾiye Ware (Hausleiter 2014: 415–418; Maritan et al. in press), which, in the context of this excavation area has been dated for the first time by means of a stratigraphic sequence (Hausleiter—Machel forthcoming). Based on the evidence from Area a, it could be concluded that the construction of the outer wall pre-dates the period of the Early Iron Age with a probable construction date of the mid-2nd millennium bc and subsequent additions and extensions (Schneider 2010b: 3).

378

hausleiter

figure 12 Sampled sites at the outer wall of Taymāʾwith datings from deposits older (green) or younger (red) than the construction of the wall; approximate course of the wall in the eastern part of sabkha dai orient-abteilung, s. lora

Area c, Square c1: Outer Wall Built before the Late 3rd Millennium bc (Inner Wall of Late 1st Millennium and Nabataean Date) Aimed at providing information on the areas immediately outside of the outer wall, at the same time at exploring its foundation, Square c1, a 15 m wide and 35 m long trench was excavated west of the outer wall (2005) simultaneously reaching it (Intilia 2010: 109–111; 113–116; for the other trenches cf. Intilia 2011: 71–77). In the context of the attempt of dating the time of when massive aeolian sand dunes accumulated to the wall, a series of sequenced samples for dating by means of Optically Stimulated Luminescene (osl) combined with 14C-dating was taken from the alluvial deposits beneath the wall (Tay 5) respectively from the inactive dunes accumulated to the standing wall (Tay 8; Klasen et al. 2011; for the location of the sampling sites ibid.: 1819, Fig. 1). The

the outer wall of taymāʾ and its dating to the bronze age

379

figure 13 Eastern part of sabkha: remains of the older eastern wall dai orient-abteilung, m. haibt

dating of the first series (Tay 5 / Tay 8) provided the following ages (listing from bottom to top; ibid.: 1820, Fig. 2): deposits beneath the wall (igs = inactive gravel sheets): 6600±300 a (= years before present; Tay 8a); 6000±400 a (Tay 5a);18 4900±400 a (Tay 5b); sand deposit in a small ditch cut into the former deposits, considered as trace of first human activity in the area: 4900±300 a (Tay 8b); the inactive dune (id) generated ages of 5100±400 a (Tay 8c), 4400±300 a (Tay 8d), 3900±200 a (Tay 8e) and 4000±200 a (Tay 8f; ibid., Table 1; 1824: Fig. 11). 14Cdates from samples Tay 8d and 8e cover the periods between 4420–4190 cal bp (Tay 8 d2 hk) and 4080–3870 cal bp (Tay e2 hk).19 A second series (Tay 176) was sampled in 2009 from deposits above supposedly collapsed mud bricks from the outer wall. The slightly younger ages of 5100–3900 a (Klasen et al., forthcoming) confirmed the results of the first series. As overall outcome of these investigations, a construction date of the outer wall west of Compound(s) c (and e) after the beginning of what is known as

18 19

The text of Klasen et al. 2011: 1824 (right column) exchanges sample nos. Tay 5b and Tay 5a; vgl. ibid., Fig. 2 and ibid.: 1820, Table 1. For the 14C-dates cf. Klasen et al. 2011; 1822, Table 2.

380

hausleiter

figure 14 Eastern part of sabkha: Magnetogram and interpretation komp 2017, fig. 21

‘Early Bronze Age’ elsewhere could be adopted, whereas the dunes would have accumulated as early as the late 3rd millennium (and later), implying that the construction of the wall had been completed before that. These results were of ground-breaking character not only radically changing the above mentioned datings but opening the perspective of an Early to Middle Bronze settlement within the walled area, possibly of ‘urban’ character. The consistently high dating of these samples and the derived construction date of this part of the wall as well as contexts uncovered in the frame of the walls-project laid the foundation for further investigations of deposits pre-dating the Late Bronze or Early Iron Age at the oasis. Before excavation in Area c started, a charcoal fragment from a mud brick of the outer wall was 14C-dated (1961–1773 calBC20), although the context of the

20

ta 3949, su 100: Charcoal; analyzed by the Leibniz-Labor für Altersbestimmung und Isotopenforschung, Christian-Albrechts-Universität Kiel; sample no. kia 24634; Radiocarbon date: 3548±26; 2σ range, probability 95.4 %; calibrated with OxCal v4.2.4.

the outer wall of taymāʾ and its dating to the bronze age

381

mud brick could only tentatively be reconstructed (i.e. approx. 60 m north of Area a). Subsequent excavations in Square c7, more than 100 m north of Area a, revealed two supports of the outer wall, each built parallel to the previous wall and resting onto the sand dune, which had accumulated against the existing outer wall (Intilia 2011: 71–72); it is probable that the mud brick belonged once to one of these annexes. It should be added that the inner wall was also reached by the Area c excavations (Squares c4 and c5); based on archaeological excavation, associated artifacts, inscriptions and 14C-dating it was established, that an older phase may have been built before the 5th / 4th century bc,21 whereas a substantial rebuilding, probably caused by a major flood event, took place towards the end of the Nabataean period (cf., Schneider 2010b: 15–17). A monumental ditch, detected on a length of 0.5km of 12m width and approx. 6m depth (Götzelt 2010; Intilia 2010: 113–114; Schneider 2010b: 15) belonged to this second phase of the inner wall, using the excavated material from the ditch for filling the two shells of the wall. Square w41: Outer Wall Built before the Mid-2nd Millennium bc In the context of the walls-project, a tower-like building at the inside of the Western Branch (no. 5 at Schneider 2010b: 3, Fig. 1) was identified, located opposite of Qasr al-Radm. Of this building, which was excavated in two seasons (2008 and 2009), an upper storey with a lateral doorway was excavated (Sperveslage forthcoming a; b). The ceiling of the rectangular room, supported by a central pillar carrying large triangular shaped building elements which were in turn covered by flat stones of smaller dimensions showed two using stages represented by two floor levels. Since there was no opening of the ceiling the upper storey may have been accessed by an external stairway, probably the same which allowed arriving at the 2nd (?) floor i.e. that with the central pillar; but no remains of such a stairway survived. The existence of a 1st and possibly also a ground floor can currently only be hypothesized based on the general height of the current surface. However, excavation did not reach below the floor level of the 2nd floor. The two stages of occupation (w41:1a and b) revealed a combination of both Red Burnished Ware (rbw) and of rbw with 21

This hypothesis is based on the content of Aramaic inscription ta 964 (Stein in press; differently al-Said 2010: 141–142), mentioning the making of a breach into an—existing— wall; this operation was carried out by Naṭir-Ēl, governor of Lawdhan, king of Lihyan. A 5th / 4th century bc funerary stele with a very fragmentary Aramaic inscription (Stein, in press) and a carved representation of an armed male figure was re-used in the foundation of the latest inner wall (Intilia 2010: 114–115).

382

hausleiter

Barbotine decoration, a pottery which was subsequently identified as of earlyto-mid 2nd millennium bc date. At the time of the investigation the associated 14C-dates from a fireplace and one bowl which was installed in a conglomerate of gypsum provided date ranging from the 17th to 15th centuries calBC for this apparently last using stage of the upper storey (Hausleiter 2014: 402, note 8). The construction of the wall, to which the building had been attached, would have happened relatively earlier, i.e. at the latest before the 17th century bc. Area l: Belonging to the Outer Wall but with No Dating Evidence Three squares of Area l, aimed at investigating the transition between Compounds b and c, revealed remains of the possible foundation of the outer wall in this part and possible remains of a passageway (Eichmann et al. 2010: 116–117). Location, building materials and techniques suggest its belonging to the outer wall (as similar to the remains uncovered in Squares w9 and w10) but from the stratigraphic observations and the pottery record no independent absolute dating could be drawn. Square q3: Outer Wall Built before the Mid-2nd Millennium bc As in Square w41, excavation trenches initiated by the project investigating the walls of Taymāʾ, were continued due to archaeological discoveries associated with those of the walls. Square q3, directly adjacent to and east of Square w10 (Schneider 2012: 107–108) is located at the outer (and southern) side of the outer wall in the southern part of Qrayyah (at the same time located within the northernmost area of Compound a, thought to be a subsequent addition to the outer wall, probably of Early Iron Age date). This part of the outer wall, contrary to the previously discussed trenches, lies east of the Burj Badr Ibn Jawhar, the central tower of the system mentioned above (cf. Schneider 2011: 105; 109; 2012: 107). The trench was aimed at identifying the nature of a square building apparently attached to the badly preserved outer wall which leads towards the Eastern Branch (no. 3 after Schneider 2010b: 3, Fig. 1) embracing the eastern parts of the oasis. The occurrence of an assemblage of Qurayyah Painted Ware (qpw) in a narrow strip south of that building in Square w10, in a homogenous deposit promised a stratified sequence of deposits of Late Bronze Age date, a period, which, at that time, had hardly been represented in the archaeological records of the excavation project.22 In adjacent Square q3, the suspected sequence of floors, probably external walking horizons, was sub-

22

Hausleiter 2014: 406–408; for previously discovered qpw sherds at Taymāʾ, cf., Intilia 2016 (forthcoming).

the outer wall of taymāʾ and its dating to the bronze age

383

sequently excavated (briefly: Hausleiter 2014: 404; Luciani—Machel, forthcoming), resulting in a ceramic sequence including rbw, qpw and plain specimens, possibly of Early Iron Age date (but dissimilar to the known types of Early Iron Age Painted and Plain Wares from Area o) followed by mid-to-late 1st millennium bc ceramic sherds. Although samples of organic material resulted in early 2nd millennium bc 14C-dates, the stratigraphically oldest dates in the sequence are of the mid-2nd millennium bc (1605–1425 calBC).23 Dating Provided by the Walls-Project The results of the investigations of the walls-project with regard to the chronology of the walls can be summarized as follows (Schneider 2010b: 3; 6–9): Outer walls of the oasis, southeastern parts of the western branch; southwestern part of the eastern branch; walls of Compounds a1 and b: from the mid-2nd millennium bc onwards. The wall of Compound a, based on considerations of materials, techniques and conceptual aspects, was understood as chronologically subsequent to the outer wall; furthermore, the identification of remains of towers at the southern side of the outer wall (= the “northern” wall of Compound a) supported this interpretation (ibid.: 19–20). Square w43: Outer Wall (Western Branch) Built after the Late 3rd Millennium bc osl-dating of deposits from beneath the western branch of the outer wall near Qasr al-Radm were obtained from Square w43. This square lies at the inner façade of the western branch of the wall, south of the supposed ‘gate’ (Euting 1914: 148; Winnett—Reed 1970: 176). The age of sample Tay 171–173 ranging from 6200–4000 a before today suggests a construction date for this part of the wall in the second half of the 3rd millennium bc at the earliest (Klasen et al., forthcoming). Square w9: Outer Wall Built after the Late-4th-to-Early-3rd Millennium bc This square, excavated in the context of the walls-project, already in 2006 (Schneider 2011: 109; 2012: 107), had revealed interesting evidence, since, in addition to burial remains probably connected to other reported burials in this area, in particular near the (old) water tower of Taymāʾ (cf. Petiti et al. 23

ta 12790, su 7740 (originating from the northern sequence within the Square): Charcoal; analyzed by the Center for Applied Isotope Studies, University of Georgia, Athens, ga; sample no. ugams 17145; Radiocarbon date: 3220±30; 2σ range, probability 95.4%; calibrated with OxCal v4.2.4.

384

hausleiter

2014: 373; cf. ibid. 372), a possible passageway between Compounds a and d as well as levels below the foundation of the outer wall, had been reached. The latter evidence was of particular interest, since, in connection with the results from Square q3, and especially in view of the results from Square c1 (see above) it appeared possible to consolidate the hypothesis of the construction date of the outer wall in an area, as Squares w41 and w43, fairly distant to those trenches which had been providing the dating evidence so far. Fortunately, in the meantime, ceramic horizons pre-dating Red Burnished Ware and rbw with Barbotine decoration had been identified in the central part of the settlement, in particular in Square e11 and Area E-East leading to the identification of earlier Gritty Ware (gw), preceded by Reddish Coarse Ware (rcw) featuring a ‘sandwich’-like coloring at the fresh break, its fabric at the same time characterized by very coarse mineral inclusions (Hausleiter—Zur 2016; Tourtet et al. forthcoming). Sherds of the latter ware had already been identified in Square q3, but at low quantity, and in the context of overall chronostratigraphic considerations, the sounding in Square w9 seemed a most suitable place for validating this early ceramic chronology in the context with archaeological remains predating the construction of this part of the outer wall. In spring 2015 the western section of Square w9 was cleaned, and pottery as well as organic samples for 14C-dating were collected. The dating of the sample resulted in a late 4th / early 3rd millennium bc date (3323–2928 calBC)24 of the charcoal remains beneath the outer wall, constituting a terminus post quem for the construction of the outer wall in this area.

4

Discussion and Perspectives

Scientific dating of a series of samples in combination with stratigraphic archaeological investigations and the pottery chronology of the site points to a dating of the construction of the outer wall of Taymāʾ to the 3rd millennium bc. At three sites (Squares c1, w9 and w43) data from deposits beneath the earliest construction phase of the outer wall has been provided adding to those data which originate from excavations of deposits subsequent to the construction and, as in case of Area a, repair of the wall. If considering only the youngest dates, an approximate mid-3rd millennium bc dating for the construction of 24

ta 17356, su 5908: Charcoal (ficus carica identified by R. Neef, Scientific Division at the Head Office of the dai, Berlin); analyzed by the Center for Applied Isotope Studies, University of Georgia, Athens, ga; sample no. ugams 24656; Radiocarbon date: 4430±25; 2σ range, probability 95.4 %; calibrated with OxCal v4.2.4.

the outer wall of taymāʾ and its dating to the bronze age

385

the outer wall emerges for the wall west of the central part of the settlement (2) in Area c; a similar date may be adopted for western branch in Square w43 (5) and hypothetically also for the eastern branch (3) in Square q3, east of the Burj. The late 4th / early 3rd millennium bc 14C-date with associated pottery coincides with several 14C-dates from the core area of the settlement (Area e), i.e. in deposits beneath Building e-b1 (Hausleiter—Zur 2016), where a pottery sequence prior to early 2nd millennium bc (and probably slightly older) Red Burnished Ware has now been established (Tourtet et al., forthcoming), i.e. of Reddish Coarse Ware (rcw), Gritty Ware (gw), and Red Burnished Ware (rbw). In spite of this admittedly attractive scenario, it is, however, necessary to consolidate this hypothesis with further stratified dating evidence. At the same time it is clear that more information on the planning and implementation of the building process of only this part of the wall system, including its organizational and administrative aspects and the sequence of individual building activities (see already Schneider 2010b), would lead to a more differentiated reconstruction of the chronological scheme developed in the present contribution, at the same time allowing for discussing the wider socio-economic aspects of this building project. Elementary part of this discussion is the presence of contemporary, possibly elite-related burials with a local tradition of grave buildings and a set of grave goods very similar to the type of warrior graves known in the Syro-Levantine area of the Early to Middle Bronze Ages, which, in recent years, have been excavated south of the oasis; details of these burial contexts have been discussed elsewhere (Hausleiter and Zur 2016). Together with the above mentioned evidence for a pottery producing oasis with building remains in an area where still (?) in the Nabataean period a temple is located (thus potentially indicating a long lasting building tradition), this suggests at least cultural contacts, most probably with an economic background with the Levant, if not generally pointing to Taymāʾ as large 3rd millennium bc oasis, competing with the urban centers all over the region. Most recent soundings in the eastern part of the sabkha, approx. 80 m west of the eastern wall, at the same time south of the sabkha wall (4b), aimed at obtaining stratigraphic information regarding the remains of a bead-producing silex industry (Haibt 2013), revealed traces of a further wall made of stone rubble, approximately n-s oriented. This wall clearly seems to pre-date the eastern wall (3c)—due to its lower foundation level—as well as the deposition of flint tools at of the production site of these beads (Haibt 2015a; 2015b), which has been 14C-dated to approx. 4,000 bc (Square se 4; . (s. Hausleiter—Zur 2016) generally corresponding to the attestation of bead producing communities in the Middle East from the Chalcolithic to Early Bronze Age (Frenez—Kenoyer

386

hausleiter

2015). In addition to this, the presence of anomalies interpreted as channels in the geomagnetic records (Komp 2017; Fig. 14), which, according to their extension and orientation may be contemporaneous to the older n-s wall are most intriguing, if considering a possible use for agricultural irrigation. Combining this chronological indicator and the evidence for the beginning of oasis cultivation slightly earlier, i.e. around 6600 calBP based on dated pollen records from the sabkha deposits (M. Dinies, personal communication; Dinies et al. 2016), a cultural and economic horizon arises, implying a walled oasis settlement right from the beginnings of its existence onwards, probably with contemporary irrigation facilities providing the context for specialized craftsmanship at the end of the 5th millennium bc.

Acknowledgments Since 2004, archaeological investigations at Taymāʾ have been carried out by the Saudi Commission for Tourism and National Heritage (scth) in collaboration with the German Archaeological Institute’s (dai) Orient Department, Berlin. The main funding of the German component has been provided by the German Research Foundation (dfg), Bonn. My sincere thanks go to R. Eichmann for constant support; to M. Dinies, A. Zur and M. Haibt for discussion and valuable comments. S. Lora produced the plans for this contribution. A. Wang (New York) read and corrected the English text. Needless to say, that the responsibility for the content of this article lies fully with the author.

References Abu Duruk, H.I. 1988. Introduction to the Archaeology of Tayma. Riyadh: Department of Antiquities and Museums. Al-Ansary, A.R. and H. Abu al-Hassan. 2005. Tayma—Multaka alkhadarat. Riyadh: Dār al-Qawāfil lil-nashr wal-tawzīʿ. Andrae, W. 1913. Die Festungswerke von Assur. (Wissenschaftliche Veröffentlichunge der Deutschen Orient-Gesellschaft, vol. 23). Leipzig: J.C. Hinrichs. Bawden, G. and Ch. Edens. 1988: “Tayma Painted Ware and the Hejaz Iron Age Ceramic Tradition.” Levant 20: 197–213. Bawden, G., Ch. Edens and R. Miller. 1980. “Preliminary archaeological investigations at Tayma.” Atlal 4: 69–106. Buhl, F. and C.E. Bosworth. 2000. “Taymāʾ.” In Encyclopedia of Islam, vol. 10: 430–431. Leiden: Brill2.

the outer wall of taymāʾ and its dating to the bronze age

387

Charloux, G., A.K. al-Tirad, A.A. al-Qaʾid, A.I. al-Dayil, T.A. al-Maliki, A. Marcolongo, Q. Morel, D.M. al-Qahtani, M.A. al-Qahtani, J. Schiettecatte and P. Siméon. 2014. “The Western Enclosure Wall Sector c.” In Duma 1—2010 Report on the Saudi-ItalianFrench Archaeological Project at Dûmat al-Jandal Saudi Arabia, edited by G. Charloux and R. Loreto, 147–257. Riyadh. Dinies, M., R. Neef, B. Plessen and H. Kürschner. 2016. “Holocene Vegetation, Climate, Land Use and Plant Cultivation in the Tayma Region, Northwestern Arabia.” In The Archaeology of North Arabia: Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5th–8th 2013, edited by M. Luciani, (Oriental and European Archaeology Band 4). Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Doughty, C.M. 1921. Travels in Arabia Deserta. New York:3. Edens, Ch. and G. Bawden. 1989. “History of Tayma and Hejazi trade during the first millennium bc.” Journal of the Economy and Social History of the Orient 32: 48– 103. Eichmann, R., A. Hausleiter, M.H. al-Najem and S.F. al-Said. 2006a. “Tayma—Spring 2004, Report on the Joint Saudi Arabian-German archaeological project.” Atlal 19: 91–116. Eichmann, R., H. Schaudig and A. Hausleiter. 2006b. “Archaeology and Epigraphy at Tayma, Saudi Arabia.” Arabian Archaeology and Epigraphy 17, 163–176. Eichmann, R., A. Hausleiter, M.H. al-Najem and S.F. al-Said. 2010. “Tayma—Autumn 2004 and Spring 2005, 2nd Report on the Joint Saudi Arabian-German archaeological project.” Atlal 20: 101–147. Eichmann, R., A. Hausleiter, M.H. al-Najem and S.F. al-Said. 2011. “Tayma—Autumn 2005 and 2006 (Spring and Autumn), 3rd Report on the Joint Saudi Arabian-German archaeological project.” Atlal 21: 64–118. Euting, J. 1914. Tagebuch einer Reise in Inner-Arabien—Zweiter Theil, herausgegeben von E. Littmann. Leiden: Brill. Frenez, D. and J.M. Kenoyer. 2015. “Stone Beads of Prehistoric Oman: Continuities in Technology and Style.” Paper presented at the Annual Meeting of the American Schools of Oriental Research 2015, Atlanta, November 18th–22nd, 2015. Götzelt, Th. 2010. “Area g.” In Eichmann et al. 2010: 107–109. Haibt, M. 2013. The Chalcolithic beadmakers of Tayma. Unpublished Bachelor thesis. Berlin: Freie Universität Berlin. Haibt, M. 2015a. “Investigations in the eastern sabkha”. In Saudi-German Joint Archaeological Project at Tayma—Spring Season 2015 (9th February–26th March): 5th Weekly Report, edited by A. Hausleiter and M.H. al-Najem, 16–18. Unpublished. Haibt, M. 2015b. “Investigations in the eastern sabkha”. In Saudi-German Joint Archaeological Project at Tayma—Spring Season 2015 (9th February–26th March): 6th Weekly Report, edited by A. Hausleiter and M.H. al-Najem, 20–21. Unpublished.

388

hausleiter

Hausleiter, A. 2011. “Ancient Taymaʾ: an Oasis at the Interface Between Cultures— New research at a key location on the caravan road.” In Roads of Arabia—The Archaeological Treasures of Saudi Arabia, edited by U. Franke and J. Gierlichs in collaboration with S. Vassilopoulou and L. Wagner, 102–123. Berlin: Wasmuth. Hausleiter, A. 2012. “North Arabian Kingdoms.” In A Companion to the Archaeology of the Ancient Near East, edited by D.T. Potts, 816–832. Chichester: Blackwell-Wiley. Hausleiter, A. 2014. “Pottery Groups of the Late 2nd / Early 1st Millennia bc in Northwest Arabia and New Evidence from the Excavations at Tayma, with a contribution by M. Daszkiewicz.” In Recent Trends in the Study of Late Bronze Age Ceramics in SyroMesopotamia and Neighbouring Regions, International Workshop, Berlin, 2–5 November 2006, edited by M. Luciani and A. Hausleiter, 399–434. (Orient-Archäologie vol. 32). Rahden/Westfalen: Verlag Marie Leidorf. Hausleiter, A. and H. Schaudig. 2016. “Rock Relief and Cuneiform Inscription of King Nabonidus at al-Haʾit (Province of Haʾil, Saudi Arabia), Ancient Padakku.”Zeitschrift für Orient-Archäologie 9: 224–240. Hausleiter, A. and H. Schaudig in collaboration with S. Baier, A. al-Dayel, Kh. al-Hāʾiti, M. al-Hāʾiti, N. al-Rashīdi and S. al-Rowaisān. In press. “A New Relief with Cuneiform Inscription of King Nabonidus of Babylon (556–539bc) from al-Hāʾit, Saudi Arabia.” Atlal. Hausleiter, A. and L. Machel. Forthcoming. “Excavations in Area a—Stratigraphy and Pottery.” Hausleiter, A. and A. Zur. 2016. “Taymāʾ in the Bronze Age (c. 2,000bce): Settlement and Funerary Landscapes.” In The Archaeology of North Arabia: Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5th–8th 2013, edited by M. Luciani, 135–174 (Oriental and European Archaeology Band 4). Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Hausleiter, A., R. Eichmann, M.H. al-Najem and S.F. al-Said. Forthcoming a. “Tayma 2008—5th Report on the Joint Saudi Arabian-German Archaeological Project.”Atlal. Hausleiter, A., R. Eichmann, M.H. al-Najem and S.F. al-Said. Forthcoming b. “Tayma 2009—6th Report on the Joint Saudi Arabian-German Archaeological Project.” Atlal. Hausleiter, A., R. Eichmann, M.H. al-Najem and S.F. al-Said. Forthcoming c. “Tayma 2010—7th Report on the Joint Saudi Arabian-German Archaeological Project.” Atlal. Huber, Ch. 1891. Journal d’un voyage en Arabie (1883–1884). Paris: Imprimerie nationale. Intilia, A. 2010. “Area c.” In Eichmann et al. 2010: 109–116. Intilia, A. 2011. “Area c.” In Eichmann et al. 2011: 71–79. Intilia, A. 2016 (forthcoming). “Qurayyah Painted Ware: A Reassessment of 40 Years of Research on Its Origins, Chronology and Distribution.” In The Archaeology of North Arabia: Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5th–8th 2013, edited by M. Luciani, 175–256. (Oriental

the outer wall of taymāʾ and its dating to the bronze age

389

and European Archaeology Band 4). Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Intilia, A. and F. Tourtet. 2015. “Square w9 / Sounding.” In Saudi-German Joint Archaeological Project at Tayma—Spring Season 2015 (9th February–26th March): 4th Weekly Report, edited by A. Hausleiter and M.H. al-Najem, 2–3. Unpublished. Jaussen, A. and R. Savignac. 1997. Mission archéologique en Arabie. (Publication de la Société des fouilles archéologiques). Vol. 1–3 (Reprint of the 1909–1922 edition). Cairo: ifao. Klasen, N., M. Engel, H. Brückner, A. Hausleiter, A. Intilia, R. Eichmann, M.H. al-Najem and S.F. al-Said. 2011. “Optically stimulated luminescence dating of the city wall system of ancient Tayma (nw Saudi Arabia).” Journal of Archaeological Science 38: 1818–1826. Klasen, N., M. Engel and H. Brückner. Forthcoming. “Geoarchaeological research: Luminescence dating at the city wall.” In Hausleiter et al. forthcoming b. Knop, C. 2014. Taymaʾ, Qasr al-Radm, Saudi-Arabien: Bauforschung an einem historischen Bauwerk in einer Oasenstadt. Unpublished Master Thesis, Technische Universität Berlin. Knop, C. and A. Hausleiter. 2016. “Tayma, Saudi-Arabien. Bauforschung am Qasr alRadm: Die Arbeiten der Jahre 2013–2015”, e-Forschungsberichte des Deutschen Archäologischen Instituts 2016–3: 155–160. Komp, R. 2017. Preliminary Field Report on a Geomagnetic Survey in the Oasis of Tayma, Saudi Arabia, 15–26 February 2015, Unpublished Report. Luciani, M. and L. Machel. Forthcoming. “Excavations in Square q3.” In Hausleiter, A., Eichmann, R., al-Najem, M.H. and al-Said, S.F. Forthcoming. “Tayma 2014–2015— 11th and 12th Reports on the Joint Saudi Arabian-German Archaeological Project.” Atlal. Macdonald, M.C.A. 1995. “North Arabia in the First Millennium bce.” In Civilizations of the Ancient Near East, edited by J.M. Sasson, vol. 2, 1355–1369. New York: Scribner. Macdonald, M.C.A. 1997. “Trade routes and trade goods at the northern end of the ‘Incense Road’ in the first millennium b.c.” In Profumi d’Arabia—Atti del convegno, edited by A. Avanzini, 333–349. Rome: Bretschneider. Macdonald, M.C.A. 2004. “Ancient North Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R. Woodard, 488–533. Cambridge: Cambridge Universtiy Press. Macdonald, M.C.A. Forthcoming. “Taymāʾ Aramaic, Nabataean, and Taymanitic Inscriptions from the Saudi-German Excavations at Taymāʾ, 2004–2015.” In Macdonald. Tayma ii—Catalogue of the inscriptions from the Saudi-German Excavations, edited by M.C.A. Macdonald. Oxford: Archaeopress. Macdonald, M.C.A. and B. Jacobs. 2009. “Felszeichnung eines Reiters aus der Umgebung von Taymāʾ.” Zeitschrift für Orient-Archäologie 2: 364–376.

390

hausleiter

Macdonald, M.C.A. and M.H. al-Najem. 2009. “A new Nabataean inscription from Taymāʾ.” Arabian Archaeology and Epigraphy 20: 208–217. Macdonald, M.C.A. and M.H. al-Najem, with contributions by F. Imbert and P. Stein. Forthcoming. Tayma iii—Catalogue of the Inscriptions in the Taymāʾ Museum, Oxford: Archaeopress. Maritan, L., Mazzoli, C., Tourtet, F. and Hausleiter, A. Forthcoming. “Pottery analysis.” In Hausleiter et al. forthcoming b. Northedge, A. 2009. The Umayyad Desert Castles and Pre-Islamic Arabia, In Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad al-Sham, Proceedings of the International Conference held at Damascus, 5–9 November 2006, edited by K. Bartl and A. Moaz, 243–259 (Orient-Archäologie vol. 24). Rahden/Westfalen: Verlag Marie Leidorf. Parr, P.J. 1979. “Archaeological Sources for the early history of n.w. Arabia.” In Sources for the history of Arabia: Proceedings of the First International Symposium on Studies in the History of Arabia, 23rd–28th of April, 1977, Riyadh, edited by A.R. al-Ansary, 37–44. Parr, P.J., Harding, G.L. and Dayton, J.E. 1970. “Preliminary Survey in n.w. Arabia, 1968.” Institute of Archaeology Bulletin 8/9: 193–242. Parr, P.J., Harding, G.L. and Dayton, J.E. 1972. “Preliminary Survey of n.w. Arabia 1968.” Institute of Archaeology Bulletin 10: 23–61. Petiti, E., A. Intilia and A. Hausleiter. 2014. “Bioarchaeological Investigations at a 4th– 3rd century bc Cemetery at Tayma, North-West Arabia.” In Proceedings of the 8th International Congress on the Archaeology of the Ancient Near East, 30 April–2 May 2012, University of Warsaw, edited by P. Bielinski, M. Gawlikowski, R. Koliński, D. Ławecka, A. Sołtysiak, and Z. Wygnańska, vol. 3, 371–390. Wiesbaden: Harrassowitz. Philby, H.StJ. 1957. The Land of Midian. London: Ernest Benn. al-Said, S.F. 2010. “Aramaic Inscriptions from Tayma, 2005 Season.” In Eichmann et al. 2010: 138–143. Schaudig, H. 2001. Die Inschriften Nabonids von Babylon und Kyros’ des Großen, samt der in ihrem Umfeld entstandenen Tendenzschriften: Textausgabe und Grammatik. (Alter Orient und Altes Testament vol. 256). Münster: Ugarit-Verlag. Schneider, P.I. 2010a. “On the organization and building techniques of the large walls of Tayma.” In Eichmann et al. 2010: 117–119. Schneider, P.I. 2010b. “Die Mauern von Tayma.” In Aktuelle Forschungen zur Konstruktion, Funktion und Semantik antiker Stadtbefestigungen, Kolloquium 09./10. Februar 2007 in Istanbul, edited by J. Lorentzen, F. Pirson, P.I. Schneider, P.I. and U. WulfRheidt, 1–25. Istanbul: Ege Yaynilari. Schneider, P.I. 2011. “Investigations on the wall-system of Tayma.” In Eichmann et al. 2011: 102–112. Schneider, P.I. 2012. “The City Wall System of Tayma 2007.” In Eichmann et al. 2012: 106– 113.

the outer wall of taymāʾ and its dating to the bronze age

391

Schneider, P.I. 2016. “Tayma.” In Ancient Fortifications—A Compendium of Theory and Practice, edited by S. Müth, P. Schneider, M. Schnelle, P. De Staebler, 343–352. Oxford: Oxbow. Schneider, P.I. Forthcoming. “Investigations of the wall-system of Tayma.” In Tayma iv—Reports 2, edited by A. Hausleiter, R. Eichmann and M.H. al-Najem. Oxford: Archaeopress. Sperveslage, G. Forthcoming a. “Excavations in Square w41, 2008–2009.” In Hausleiter et al. forthcoming c. Sperveslage, G. Forthcoming b. “Archaeological Research on the Wall System at Tayma—Part i: The Western Branch and Compound w.” In Tayma iv—Archaeological Reports 2, edited by A. Hausleiter, R. Eichmann and M.H. al-Najem. Oxford: Archaeopress. Stein, P. Forthcoming. “Die reichsaramäischen Inschriften aus Taymāʾ, Kampagnen 2005–2009.” In Tayma ii—Catalogue of the inscriptions from the Saudi-German Excavations, edited by M.C.A. Macdonald. Oxford: Archaeopress. Tourtet, F., M. Daszkiewicz and A. Hausleiter. Forthcoming. “Tayma Pottery: Chronostratigraphy, archaeometric studies, cultural interaction.” In Proceedings of the Workshop “The Archaeology of the Arabian Peninsula: Connecting the Evidence”, 8th International Congress on the Archaeology of the Ancient Near East, edited by M. Luciani. Vienna. Wellbrock, K. and M. Grottker. Forthcoming. “Hydrological investigations.” In Hausleiter et al. forthcoming b. Winnett, F.V. and W.L. Reed with contributions by J.T. Milik and J. Starcky. 1970. Ancient Records from North Arabia. Toronto—Buffalo: Toronto University Press.

chapter 20

Pottery from the “Midianite Heartland”?1 On Tell Kheleifeh and Qurayyah Painted Ware. New Evidence from the Harvard Semitic Museum2 Marta Luciani

“Scherben bringen Glück”3 says the old adage for nuptials … so even if I am well aware that compared to lavish inscriptions or to the rock graffiti Michael is a master in deciphering, sherds do not seem very important, I will try my luck by discussing and contextualising a dozen ceramic fragments. One may question whether this pottery, though painted, but excavated long before the jubilee himself was born, can still be considered relevant nowadays. Nonetheless, I am confident that Michael more than anybody can appreciate how new research—be it in the field, in the library or in the boxes of a museum—is critical in helping us gain a better understanding of the ample inconnus in the archaeology of the Arabian Peninsula.

1 Parr 1982: 129. 2 It is with great pleasure that I take the opportunity to thank Peter Der Manuelian, Director of the Harvard Semitic Museum, Joseph A. Greene, Deputy Director, and Adam J. Aja, Curator of Collections for their constant support and encouragement as well as for permission to publish the material discussed in the present article. To Joe Greene go my particularly heartfelt thanks for his repeated counsel and generous sharing of his knowledge in all things Near Eastern and for reading a previous version of this article. The charm and friendship of the Cantley Sato family has made my stays in Cambridge a researcher’s dream. I also wish to acknowledge Laura Machel for digitising my pencil drawings and Andrea Intilia for participating unrestrictedly his prior knowledge on Tell Kheleifeh’s most ancient pottery and his sustained interest in Qurayyah Painted Ware even after writing an 83-pages long article about it (Intilia 2016). Erez Ben-Yosef and Uzi Avner were both speedy and very helpful in answering my electronic queries about Wādī ʿArabah sites. Of course, I remain the sole responsible for the interpretations expressed in this article. This contribution was written during my tenure as Visiting Scholar and Adjunct Professor at New York University’s isaw (Institute for the Study of the Ancient World). I am indebted to its Director, Roger Bagnall and the library staff (David Ratzan and Gabriel McKee) for their consistent backing during this research. 3 Moreover, these sherds have been brought by Glueck [Nelson] so they cannot but bring luck!

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

pottery from the “midianite heartland”?

393

By investigating eleven pottery sherds stemming from Tell Kheleifeh,4 Southern Jordan, six5 of which have not been published before, I would like to show how they can positively be identified as Qurayyah Painted Ware6 (hence abbreviated as qpw), against their original definition as ‘Edomite’ and notwithstanding the fact that the site’s main occupational period is consistent with the Iron Age/‘Edomite’ date.7 On the basis of novel interpretations of old publications and new data both from Qurayyah itself and partially also from Taymāʾ, I shall propose a reliable chronological frame for this material. Moreover, I would like to consider the role of Tell Kheleifeh as find-spot located in the central part of the distributional area of qpw pottery as it has grown to be defined through research in the last years.8

The Site Still well recognisable on satellite images, the low mound of Tell Kheleifeh, located only half a kilometre from the present-day coastal line at the northern tip of the eastern branch of the Red Sea, most likely represented a significant harbour settlement. After its almost fortuitous discovery by F. Frank during his ʿArabah survey,9 because of its location “immediately north of the present shore-line of the Gulf of ʿAqabah, and just west of the boundary line between Palestine and Transjordan”,10 Tell Kheleifeh was visited in 1934 and twice in 1936 by 4 5 6

7 8 9 10

Tell Khulayfah would be a more exact transliteration of the toponym, but Tell Kheleifeh is kept in this contribution because that is how the site is known among archaeologists. But see Addendum at the end of this article which brings the total count to fourteen, nine of which were previously unpublished. Also known as ‘Midianite’ pottery. For an overview of this ‘perhaps too unwise’ definition and the choice of the new one, see Parr 1988: 73–74. Of course, by calling it now ‘Qurayyah Painted Ware’ I by no means imply it necessarily must all have been produced at that Northern Ḥijāzī site. This can only be established through archaeometric analyses. Furthermore, even if in this article I only discuss painted specimens, the same ware (actually multiple fabric types, see below Table 3 and Parr 1982: 128) is used for both painted and unpainted vessels. Glueck 1967: 10: “no earlier than the latter part of the eighth century b.c., and no later than the sixth century b.c.”. For an overview on qpw see Intilia 2016, focused exclusively on published evidence, and Singer-Avitz 2014 discussing qpw’s chronology. Frank 1934. Glueck 1935: 48. Frank (1934: 244) had written: tell el-chlēfi “liegt etwas westlich der durch

394

luciani

N. Glueck’s team in the course of the Explorations in Eastern Palestine.11 In a two-day period during the second of these visits, the first soundings were carried out. Its discoverer, already in 1934, noticing on the surface of the site pre-Roman pottery sherds, had immediately proposed the identification of Tell Kheleifeh with biblical Ezion-Geber, Solomon’s port on the Red Sea.12 For these pioneers of field research, the biblical narrative was the driving interest and force beyond investigations in the Near East. But the seminal work carried out by W.M.F. Petrie13 had long shown that other evidence, such as pottery, represented significant data for establishing the chronology of sites. Glueck seemed to have fully embraced this methodology: “As has been pointed out repeatedly, it is now possible from surface finds of pottery or of sherds alone to determine with a considerable degree of accuracy the age to which a particular site belongs, even when all other indications are missing”.14 In view of these promising methodological premises, it is still now slightly disheartening to see the many ways in which the unbalanced relationship between texts and archaeology ended unduly influencing the correct interpretation of the function of the discovered structures, the chronology of the site and by extension also of the date of all pottery assemblages found therein. In the years 1938–1940 Glueck decided to investigate the site during three seasons of excavation. While different preliminary reports by the excavator appeared in the following years,15 neither final publication nor a comprehensive study of all remains was ever put into print. A selective reappraisal of the

11 12 13

14

15

eine hohe Kokospalme bezeichneten jetzigen Grenze zwischen Palästina und Transjordanien, also noch auf palästinischem Boden”. Glueck 1935 and 1939a: 3. Frank 1934: 244. “Our special notice must be given to the potsherds lying strewn all over the surface. Pottery is the very key to digging; to know the varieties of it, and the age of each, is the alphabet of work. A place may be dated in a minute by its pottery on the surface, which would require a month’s digging in the inside of it to discover as much from inscriptions or sculptures” (Petrie 1892: 158). Glueck 1935: 4. Glueck had also taken part in the famous excavations headed by W.F. Alrbight at Tell Beit Mirsim, where scientific archaeology took shape for the first time (Wright 1970: 27). There he “became Albright’s first serious student of ceramic chronology” (ibidem: 29). I owe this reference to Joe Greene. However, Glueck’s perspective remained that beyond scanty Egyptian sources, “the Bible remains the chief source of information aside from the archaeological evidence of the country itself” (ibidem: 1). Glueck and Albright 1938, Glueck 1938, 1939b, 1940, 1965, 1967, 1993.

pottery from the “midianite heartland”?

395

results of all field campaigns supported by a limited re-survey of the site was carried out by G.D. Pratico,16 almost fifty years after the initial investigations. In his publication, Pratico stated quite clearly that, because the excavation was carried out with inadequate methodological standards and field recording systems, he would be able to discuss only certain aspects of the site. He thereby proposed to reconstruct a sequence for the architecture uncovered by Glueck that could be divided into five different consecutive periods. Period i—which has been recognised as having three sub-phases—was characterised by the presence of a so-called ‘Four-Room Building’. Contrary to earlier readings, this building was interpreted as to have been fenced in by a casemate wall from the onset of its life. A single rectangular, mud-brick mastabah tomb was identified aligned outside the north-eastern perimeter wall of this phase. In Period ii, when a square fortification wall-cum-doublechambered entrance gate in the western part of the southern side of the enclosure was added, the settlement was significantly enlarged. Several constructions filled the area between the casemate building and the outer wall (Period iii). Eventually, In Period iv a–c a number of buildings overlapped this intermediate phase and covered large parts of the previous constructions. After initial proposals of a Solomonic date for Tell Kheleifeh to suit the interpretation of the site as biblical Ezion-Geber, the study of the pottery proved that Periods i through iv should range in time between the late 8th and the 6th centuries bce, without, however, it being possible to attribute or anchor any specific assemblage to any single corresponding layer of phase.17 To a Period v have been ascribed those remains (one Phoenician inscribed sherd, eight different ostraca, and one jar with South Arabian/Minean inscription) that could not be associated with visible architectural remains and that range in dating from the mid-5th and 4th century bce and down to a date close to the very end of the 3rd century bce in the instance of one inscribed Rhodian jar handle.18 From the beginning of the excavation, the interpretations of both detailed features and of the Period i building as a whole and its chronology have been strongly influenced by the a priori identification of the site with the biblical narrative on one hand, and, on the other, by the investigations’ main research question: was the ʿArabah valley the location of “Solomon’s (copper) mines”?19

16 17 18 19

Pratico 1985, 1992, 1993a, 1993b, 1997. Pratico 1985 and 1993a. DiVito 1993: esp. 62. For a critical review see Muhly 1984 and Finkelstein 2005; idem 2014 for the identification with Ezion-Geber in the Assyrian period.

396

luciani

So it is not surprising that even if the original interpretation of the ‘FourRoom Building’ as a smelter has been revised by the excavator himself—by proposing to interpret the structure as a granary first, and a citadel-fortress later—, renewed investigations in 1999 by a Canadian team20 have proven that earlier excavations had missed crucial information. The new project has shown that the settlement was apparently not confined solely within the perimeter of the walls and was therefore more extended than research in the 1930s and 1980s had thought.21 Furthermore, Mussell’s excavations, while confirming the early ‘Four-Room Building’s relations with the fortifications, have shown that the traces of greenish-glazed or dark red surface on fired-bricks composing the walls should be attributed to a fire conflagration that affected the building,22 not to smelting activities. Most remarkable is the presence of 1m-deep wall constructions below ground, reaching down to the virgin soil and enveloped by a “red clay-like soil with heavy calcite inclusions”.23 It is under this layer that a second mastabah tomb was uncovered and partially excavated. Because of the detailed observations made in the field, it was possible to establish that this construction—and by comparison also the mastabah uncovered by Glueck?24—must be seen as the earliest of the architectural phases as yet archaeologically documented on the site of Tell Kheleifeh. However, no precise date has been given for this structure by its excavator, who does not mention any additional chronological periods beyond the Iron Age ones already established during the 1980s reappraisal.25

Pre-First Millennium bce Evidence from Tell Kheleifeh All revisions of the chronology of the architecture and of the finds from Kheleifeh verged on shifting the date from the early (‘Solomonic’ 10th century bce) to the full or mature/late Iron Age (‘Edomite and Assyrian’ late 8th to 6th century bce). It was only with the discovery and excavation of the so-called

20 21 22 23 24

25

Mussell 1999 and 2000. After the premature death of M.-L. Mussell no work has been resumed on the site. Mussell 2000. Mussell 2000: 577. Mussell 1999: 6. In his report Glueck (1940: 9–10) identified the size and type of mud-brick masonry of the tomb with the ones of the casemate building. Therefore, he interpreted this construction as the “grave of the important man who built” the fortress-settlement on Tell Kheleifeh. Mussell 1999 and 2000.

pottery from the “midianite heartland”?

397

Egyptian Temple at Timnaʿ (Wādī Manaʿiyyah, Site 200),26 where a pottery similar to a handful of painted sherds found in Tell Kheleifeh was identified and dated to the 13th to 12th century bce, that a new perspective emerged. B. Rothenberg and J. Glass were the first to lay the basis for recognising and proving that the Tell Kheleifeh sherds published in 196727 as Edomite and dated to the Iron Age period must be considered related and contemporaneous to one of the four different categories of pottery found in the Timnaʿ Temple, the painted pottery they called ‘Midianite’.28 As they suggested, the correct attribution to an earlier period of the painted pottery production from Kheleifeh (and other sites) was possible only once the sherds were looked at in direct autopsy and not only from drawings or photographs. From its onset, the petrographic analysis of this material revealed that the ‘Midianite’ sherds “differ in clay and temper from any pottery known from Palestine and Syria”.29 The find-spot of the painted sherds in Tell Kheleifeh had been indicated as being Stratum iv, i.e. the most superficial and recent one.30 Understandably, once they were recognised to be similar to the Timnaʿ ones, and therefore older than the levels attested on the site, also their context of provenance was put into question.31 Y. Aharoni proposed that they must have come from an area outside Glueck’s excavation proper.32 Mussell, in her renewed work at Tell Kheleifeh, pointed out that the earliest evidence on the site was a mastabah tomb found close to the one identified by Glueck during his third campaign of excavation,33 north of the main build-

26 27 28

29 30 31

32 33

Rothenberg 1988 with previous references. Now see also Avner 2014 and below. Glueck 1967. Before him, jug Reg. No. 248, i.e. our # 1 had been depicted by Wright in his 1959 article. Rothenberg and Glass 1983: 76 “belong to the Midianite pottery group”. However not all Rothenberg’s and Glass’ identifications are to be considered certain (for Tell el-ʿAjjūl see Intilia 2016: 202–203 and for Bir el-ʿAbd see Luciani forthcoming a). Parr et al. 1970: 238 footnote 50 had already noted the similarity with the painted pottery found in Qurayyah and assigned it a Late Bronze Age date. Rothenberg and Glass 1983: 68 relaying A. Slatkin’s analytical work. Glueck 1967: 10. But see below Tables 4 and 5 with details of the find situation from the Tell Kheleifeh registries. Meshel (1975: 49–50) recognising that the similarities between the ‘Midianite’ painted pottery from Timnaʿ and the sherds published by Glueck meant that the latter should also be identified as ‘Midianite’, suggested that their attribution to Tell Kheleifeh Stratum iv must be erroneous (I wish to thank Z. Weiss for translating the Hebrew text of Z. Meshel’s article for me). Cited in Rothenberg and Glass 1983: 120–121, footnote 33. Glueck 1940.

398

luciani

ing. The latter mastabah tomb, excavated in the 1930s, is said to be “sunk partly into the floor-level of the dry moat between the two outer fortification walls of Period ii” but Glueck firmly interpreted its function and chronology in connection to the main Iron Age phase of the Tell Kheleifeh ‘Four-Room Building’. The 1999 excavation, on the contrary, specifies that this second mastabah tomb is stratigraphically underneath the red clay soil which was set up around the foundations of the earliest building level in Tell Kheleifeh and must, therefore, be the eldest evidence. However, there is no published statement on a possible date for this structure. Mussell does not mention earlier pottery or the possibility that older material (i.e. ‘Midianite’ pottery) might have to be associated with this mastabah tomb. Nonetheless, knowing about the new stratigraphic evidence from Mussell’s excavation at Tell Kheleifeh and aware of the presence of the ‘Midianite’ pottery from the site, U. Avner suggested: “The earliest identified remains now are two mastaba tombs in the north probably dated to the thirteenth-twelfth century bce”.34 While I recognise this as a distinct possibility that would account for the presence of our qpw sherds in Tell Kheleifeh, the fact that Glueck found in the mastabah tomb he excavated—besides the human bones—also remains of two three-handled jars and camel bones35 speaks more in favour of an Iron Age dating (late 8th to 6th century bce) of at least the first tomb found. Furthermore, the Harvard Semitic Museum Tell Kheleifeh registries do not evidence any qpw sherds found in the northern part of the excavation36 (see below fig. 5). The 1999 excavations have proven a higher complexity in the connections between the structures excavated in the 1930s, and more numerous building layers than those recognised by the first investigations. However, it remains dubious whether the two mastabah tombs—even if more ancient than all other excavated architecture on the site—should really be interpreted as the source of our two handful of qpw sherds.37 The most recent excavations on the site have proven that even Iron Age Kheleifeh was significantly larger than

34 35 36

37

Avner 2008: 1707. Glueck 1940: 9–10. According to the extracts of the Tell Kheleifeh registries at the hsm and kindly put at my disposal by Joe Green, a listing and mapping of the find-spot of each sherd has been undertaken below in Tables 4 and 5, and fig. 5. By the look of them, except for # 1 below, our pottery fragments resemble more settlement sherdage than vessels stemming from graves. We would expect the latter to be larger or even complete shapes. qpw pottery has been found in funerary contexts in Taymāʾ, Tell Jdūr, Tell al-ʿAjjūl and Tell al-Farʿah South (see Intilia 2016 for updated references), however, none of these burials take the shape of mud-brick mastabahs. On the other

pottery from the “midianite heartland”?

399

thought at the beginning. Therefore, it is not to be excluded that the pre-first millennium bce pottery sherds found on the surface must stem from levels that were disturbed by those Iron Age building activities that reached deepest in the layered stratigraphy of the site.38 Only renewed excavations could bring reliable information on the precise location of these earlier strata in Tell Kheleifeh. The story of the division of finds from Tell Kheleifeh is more complex than we have room to explore in this context.39 Suffice it to say that after completion of the 1930s excavation, the Palestine Archaeological Museum (pam) now Rockefeller Museum, retained the government’s share of the finds from Tell Kheleifeh on behalf of the Department of Antiquities. In 1951 portions of those finds were transferred to the Jordan Archaeological Museum. Glueck shipped asor’s share to his home institution, the Hebrew Union College in Cincinnati. In the 1970s the Nelson Glueck Archive was transferred from there to the Harvard Semitic Museum.40

Painted Pottery from Tell Kheleifeh at the Harvard Semitic Museum (For description and illustration of the pieces see Tables 1–3 and fig. 1–2) In March 2016, I was able to spend a week at the Harvard Semitic Museum, checking records and boxes in order to verify what the five sherds published by Glueck as ‘Edomite’ in 1967 and re-interpreted as ‘Midianite’ by Rothen-

38

39

40

hand, the lack of mud-brick mastabah tombs associated with qpw pottery may also be due simply to the state of our, admittedly still limited, knowledge. If, however, with Y. Aharoni (see above footnote 32), we hypothesise that the earlier settlement was located outside the Iron Age site (now also Finkelstein 2014: 135), it becomes even more difficult to account for these ancient sherds being redeposited on the surface (Stratum iv). On the contrary, Iron Age building activities could be responsible for indenting lower, preceding deposits and bringing up the pertinent pottery. Tables 4 and 5, and fig. 5 below, show that qpw sherds mostly come from the central-southern part of the excavation, where mostly late (Strata iii? and iv) structures are located and could have burrowed earlier layers, thus making the hypothesis that qpw sherds must be residual most likely (Finkelstein 2014: 122). I owe to Joe Greene all clarifications on this matter. For details on the story of the division of the Tell Kheleifeh finds and the Nelson Glueck Archive see McKenzie 2003 and 2013. In 1946 some finds went also to the Anthropology Department of the National Museum of Natural History of the Smithsonian Institution. For details on excavation records see Pratico 1993a: 9–15. “Harvard’s artifact collection is also the largest of the three and constitutes a representative assemblage of the site’s material culture”: Pratico 1993a: 11.

400

luciani

berg and Glass in 1983 really looked like. After having excavated and examined several hundred ‘Midianite’, or better qpw sherds from the site of Qurayyah41 and before that from Square q3 in Taymāʾ,42 I trusted a new assessment was possible. Moreover, methodologically, when dealing with pottery, direct examination is irreplaceable.43 To my surprise I could physically locate only one of the five published sherds I was looking for (# 5, see below). In one instance, I could establish that the large, fragmentary painted jug that was Glueck’s best piece, is listed as one of the vessels that has gone into the possessions of the Palestine Archaeological Museum (# 1 below). The whereabouts of all the other fragments (# 2–4 below) could not be ascertained.44 However, to my even greater surprise, in Box 39 of the Harvard Semitic Museum Tell Kheleifeh cabinets, I could identify five additional sherds that are surely to be attributed to the qpw ceramic class.45 I will briefly review the first four (# 1–4) for which black-and-white photos exist and have been recognised by different scholars46 as being ‘Midianite’, i.e. qpw sherds. Fragment # 5, part of that original lot, belongs to the holdings of the Harvard Semitic Museum and I re-publish it in this contribution with colour photographs and line drawing.

41 42 43

44

45

46

Luciani 2014a, 2014b, 2016a, forthcoming a, Luciani and Alsaud 2015, 2016, Luciani et al. 2015. Luciani 2013, Luciani and Machel forthcoming. See below footnote 106 and Sherratt 2013. On the importance of oral transmission of knowledge on pottery and learning the ropes in the field see Didier 2013, Luciani 2014c and Luciani 2015 and for the past see Wright 1970: 27 and 29. Thanks to Joe Greeene’s kind help not only Tell Kheleifeh boxes at the Harvard Semitic Museum containing decorated pottery (i.e., Boxes # 14, 19, 20, 39, 54, 55) were checked by the writer but also the paper lists of the Palestine Archaeological Museum and the Jordan Archaeological Museum as well as the electronic lists kept at the Harvard Semitic Museum of objects from ‘Ezion-Geber’ stored at the Smithsonian Institution in Washington d.c. No corresponding number or description could be found to match the sherds we were looking for. On the problems of artefact retrieval already in the 1980s see Pratico 1933a: 15. [Addendum: qpw sherds # 12, 13 and 14 were found respectively in Box 40, the first one, and Box 32, the latter two. They had not been registered as painted sherds in the hsm electronic lists at my disposal. They were spotted through systematic control of all boxes after I had established what qpw pottery looked like in Tell Kheleifeh]. The Harvard Semitic Museum has recently re-arranged this material. All specimens in Box 39, including the five sherds I identified as qpw, were generically labeled “Edomite/Midianite pot handle sherds”. Rothenberg and Glass 1983: 75–76, Intilia 2016: 205 and passim.

pottery from the “midianite heartland”?

401

A sixth ‘Midianite’ sherd must be discussed here as it is mentioned in a previous publication47 but could not be located either in the Harvard Semitic Museum boxes nor identified in any of the finds lists from Tell Kheleifeh/EzionGeber at either the Palestine Archaeological Museum, the Jordan Archaeological Museum nor the Smithsonian Institution.48 To the best of my knowledge it has not been illustrated in any publication. Pratico (1993a: 49) writes: “five body sherds and one sherd that preserves a small section of a rim are identified as Midianite based on their fabrics and geometric motifs”.49 Since this description does not correspond either in number or type to the lot published by Glueck (1967) or to the sherds retrieved in the Harvard Semitic Museum, or to the material listed as being in the three institutions with finds from the site, it cannot be excluded that Pratico’s statement refers to an additional group of Qurayyah Painted Ware sherds.50

47 48

49

50

Pratico 1993a: 49. The Smithsonian holdings were very recently revised by Joe Greene. There is a single instance of painted vessel (reg. no. 417), a complete, small globular jug with a rather coarse, vertical-strokes-in-a-horizontal-double-band, polychrome painted decoration located on the shoulder (the background of the bands is white, the strokes are dark and an additional horizontal band is red-and-black). Joe kindly sent me two pictures of this vessel. However, the decoration though generically ‘similar’ has no precise parallel in the qpw repertory and resembles more Iron Age straight-bands designs, obtained at the wheel (e.g., Oakeshott 1983). But even more telling is the vessel’s fabric. Besides being very dark in colour and not displaying the shale inclusions that are known for qpw, it features multiple chipping on the surface, apparently due to the bursting of carbonate / calcite tempering. The latter components are normally not present in the petrography of qpw (Rothenberg and Glass 1983: 113). Of course, in absence of direct analysis of the vessel, and judging only from photographs, this exclusion from the qpw group may need to be reviewed in the future. But Pratico in a previous passage in the same publication (1993a: 47) had identified four sherds as Midianite: Glueck 1967: fig. 1.2 (= fig. 5.1) and fig. 4.35, leaving fig. 4.2 (our # 5 here) out. I tried to understand the grounds for this discrepancy between what is published and what I personally observed by enquiring of G. Pratico via email. I have not received a reply before publication. On the one hand, the difference between the tally on Pratico (1993a: 47 and 49) could lead to think that the count of six sherds is just due to a typo. On the other hand, his description of the shapes is different from the ones published by Glueck and therefore there is a chance it refers to a never previously described, additional lot of qpw pottery. This would mean a significant increase of qpw material from the site of Tell Kheleifeh. And indeed the Addendum at the end of this article underscores the potential of such a discovery to occur in the future.

402

luciani

Finally, the five sherds (# 7–11) that I could for the first time positively identify with qpw have neither been mentioned nor published before. It must be stated at the outset that classifying these pieces as qpw depended mainly on the direct macroscopic51 autopsy of the fabrics, which were found to be identical to the ones attested on the site of Qurayyah and completely different from ones used in the Iron Age pottery from Tell Kheleifeh i examined in the Harvard Semitic Museum boxes. The emphasis I place on identification through the fabric is due to two factors. First, as Rothenberg and Glass remarked, clay fabrics and inclusions of qpw are really different from all other pottery encountered in Syria and the Levant and are therefore distinctive. Furthermore, while there are a number of—especially bichrome—painted patterns that are specific of and only found in qpw and therefore easily recognisable as such (e.g., rows of small red squares with dark dot in the middle,52 rows of concentric half circles at the bottom of a decoration frieze,53 a red/black ‘x’ with dark dots in each angle,54 etc.), when the specimens in question are small, the surfaces not well preserved, the shape or the type of decoration are generic (e.g. monochrome cross-hatched, simple monochrome straight or wavy bands and the like) then only fabrics are a reliable ground for a positive identification. Because of the critical role we assign to direct examination of the fabric (see below Table 3), the following comments on specimens # 1–4, those fragments that we could not analyse personally, are forcefully based only on description of shapes and painted decoration motifs as visible on the old published photographs. However, as will be shown below, the evidence for parallels with other qpw specimens is so rich that their identification with this pottery class can be made with a high degree of confidence.

51

52 53 54

A simple magnifying glass (×3) was used, too. In Table 3 below, I have not registered the minimal sand component at times visible in the fabric because I could not establish whether it was residual or intentionally added. This would need specific archaeometric analyses. Sherds # 12–14 listed in the Addendum have now been inspected directly and their fabric checked but not described in detail in this article. Hausleiter 2014: fig. 8a; Rothenberg 1988: fig. 6: 22; idem 1998: fig. 3: 20 and fig. 5 (colour); Intilia 2016: fig. 5: 7–9. Rothenberg 1998: fig. 4.31; Intilia 2016: fig. 14: 1–4. Parr et al. 1970: fig. 15.12; Rothenberg 1972: fig. xxiii lower left; idem 1998: fig. 3.7.

pottery from the “midianite heartland”?

403

Previously Published qpw Fragments from Tell Kheleifeh # 1 (see below Table 1): This “jug”55 features a shape not yet documented in qpw repertories and is not very much alike the Timnaʿ specimen it was compared to.56 However, the description of its “light creamy-buff ware with numerous grits” and the presence of a candelabra-like pattern—more than the generic trellis pattern—make me inclined to confirm previous identifications57 of this vessel as qpw. The comparison Glueck offered with the painted pattern on a differently shaped large jar does not seem fitting.58 But monochrome crosshatched decoration framed in vertical or horizontal bands59 combined with the candelabra-like pattern60 is attested in qpw assemblages. # 2 (see below Table 1): The lower part of this pattern, the semi-circle with concentric dot on top of converging lines alternating a couple of parallel vertical lines is very typical of qpw and found repeatedly on straight-sided bowls and globular juglets.61 Based on the description (since the photo is difficult to read) the upper part of the design could be similar to the “v” painted motif on the outer surface of a body sherd from Qurayyah and on the internal part of a bowl from Taymāʾ.62 The design is so specific and unique that it can only be interpreted as a qpw piece. # 3 (see below Table 1): Decoration of similar painted dark spikes on the neck of a jug was recognised very early on several specimens from sites ranging

55 56 57 58 59

60 61

62

Glueck 1967: 11. Glueck 1967: fig. 1 no. 1. Rothenberg and Glass 1983: 75–76, Pratico 1993a: 47, and Intilia 2016: 251. Glueck 1967: fig. 5.3. Ingraham et al. 1981: fig. 78: 1; Hausleiter in preparation: photos qu 10_0127 and qu 10_0128. I am very grateful to Arnulf Hausleiter for allowing me to see and refer to the work and material that stems from sampling undertaken during the geomorphological and archaeological survey of the site of Qurayyah carried out by the Saudi-German team in 2008 and 2009 and now in preparation for publication. Luciani 2014b: fig. 7, Hausleiter in preparation: photos qu 10_0167 and qu 10_0168. Both from Site 2 at Timnaʿ. Rothenberg 1967: illustration on p. xii; idem 1972: fig. 31.1 and photo no. 49; Rothenberg and Glass 1983: fig. 4.3 and fig. 6.7. This motif is attested also in Qurayyah (Luciani 2014a and Hausleiter in preparation: photos qu 10_0173 and qu 10_0174 [for the lower part of the design]) and Tell Qudayrāt/Kadesh Barnea (?) (Bernick-Greenberg 2007: pl. 11.7 no. 12). Qurayyah: Luciani 2014a: no. qu.14.1.7. For Taymāʾ: Hausleiter 2014: fig. 7a, lower right.

404

luciani

from Qurayyah to Tell al-Farʿah South and now again Tell al-ʿAjjūl63 and rapidly acknowledged64 as a very typical qpw motif. # 4 (see below Table 1): This photo is not very clear. And while the motif with “two groups of four vertical lines each”65 is well attested,66 they are more often framed by a horizontal not a vertical row of dots, as Glueck describes it. Because of the difficulty in interpreting the photo and the lack of perfectly matching parallels the identification of this sherd with qpw should be considered likely but not entirely sure. # 6: as no illustration is provided no conclusion can be reached about this piece. Even its existence could be doubted.67 See Table 1 on pp. 406–409.

Previously Unpublished qpw Pottery from Tell Kheleifeh at the Harvard Semitic Museum # 5 (see Tables 1–2, fig. 1 top): Also in this case, the old photograph would not have allowed a reliable identification. However, because this artefact could be retrieved in the Harvard Semitic Museum reserves, it was possible to correctly document it and analyse it. The piece is chipped in numerous parts. Furthermore, this is the only pottery fragment among those from Tell Kheleifeh that features on the outer surface a rather thick, light beige slip. Traces of a similarly coloured slip, but not as thick as the one on the outer surface, are present also on part of the inner surface. While shape and fabric of this flat base are perfectly compatible with an interpretation as qpw, the outer coating, and to a certain extent the bichrome painting look unusual, though not foreign. A possible general comparison for the painted motif could be drawn with a published base from Qurayyah itself68 but several aspects of the details differ. The

63 64 65 66 67 68

Intilia 2016: fig. 14 with all pertinent references. Dayton 1972. Glueck 1967: 16. Rothenberg 1998: fig. 3.19, and fig. 4.5. See above footnotes 44 and 50. Ingraham et al. 1981: fig. 78.10. This bowl is slipped, however, and features black paint in the interior. Also the motif displays not only long, tapering spikes but also parts of a cross-hatched metope, something our piece # 5 has not preserved. Another parallel is with Hausleiter in preparation: photos qu 10_0047 and qu 10_0048; in this case, however,

pottery from the “midianite heartland”?

405

most compelling comparison, not of the thick slip, but of the somewhat carelessly painted bichrome tapering spikes is with a complete, flat based beaker found on the surface in the Fields area in Qurayyah, during the 2015 excavation campaign.69 Because of the inaccurate stroke of the painting and the type of motif, we forwarded the hypothesis that this vessel represented a later development of the core qpw production in Qurayyah. However, since its context is not datable, this must remain conjectural for now. The attribution to the qpw group of this Tell Kheleifeh piece seems nonetheless very likely because of its fabric, and its combination with an attested shape and bichrome painted decoration. # 7 (see below Table 2 and fig. 1 middle): In the case of this narrow neck of jug(let), both the fabric with the visible red shale tempering, the slip and its colour zoning70 (due to the kiln firing process) as well as the bichrome painted decoration are all so typical of qpw that the identification may be considered assured. The painted motif showing a combination of lighter, reddish pattern (mostly zig-zag) with dark, brown to black either dots or triangles is well attested in pottery from Timnaʿ and considered typical.71 Comparisons can be found also in Lachish and Tell Qudayrāt.72 At least one specimen exactly mirroring the reddish trellis pattern interspersed with red drops comes from Qurayyah and is still unpublished.73 # 8 (see above Table 2 and below fig. 1 bottom): monochrome, cross-hatched decoration like the one attested on this body fragment of a globular vessel is less typical of qpw than would be the bichrome motifs discussed thus far. When it does appear, the cross-hatched motif, furthermore, is framed in a horizontal or vertical band74 or a metope. Traces of a line framing the cross-hatched design are not recognisable on this piece, but this is possibly due to the small size

69 70 71 72

73 74

the vertical, bichrome decoration is on the inside of a flat-based bowl. It, too, seems accompanied by a cross-hatched butterfly (?) design. Luciani and Alsaud 2015: 4, fig. 3 and see below footnote 118. For colour zoning as very common feature in qpw, see Rothenberg and Glass 1983: 103. Rothenberg 1998: fig. 4.16. For Lachish: Singer-Avitz 2004: fig. 20.55 no. 2 and 20.56 no. 2. In the case of Tell Qudayrāt, though, the painting is not on a jar neck but on a globular vessel’s shoulder, BernickGreenberg 2007: 140 (and vol. 2, 10–11), pl. 11.7 no. 10. For a dating of the ‘Midianite’ sherds from Tell Qudayrāt/Kadesh Barnea (?) in the 13th–11th century bce see Intilia 2016: 199– 200. Hausleiter in preparation: photos qu 10_0070 and qu 10_0071. Ingraham et al. 1981: fig. 78.1, Rothenberg 1998: fig. 7.4, 8.1.

406 table 1

luciani Previously published Qurayyah painted ware fragments from Tell Kheleifeh

Count Reg. no.

Vessel fragment

Place of publication

#1

Almost complete jug (missing base)

Wright 1959: 104 Glueck 1967: 11–13, fn. 14

#2

248

3016

“Wheel-made, slightly everted rim, tapering neck, single, broken loop-handle, rounded body curving down”

“light creamy-buff ware, with numerous grits, medium baked”

Body sherd, globular bowl

Glueck 1967: 15, fn. 31 “creamy-buff ware with numerous tiny grits, creamy-gray slip-covered surface, showing separate horizontal lines of pebble burnishing”

#3

7079

Outturned rim and neck of jug

Glueck 1967: 13, fn. 28 and p. 15, fn. 34

“shoulder, tapering neck, and outturned rim of a wheel-made jug”

“creamy buff, fairly well levigated ware, containing numerous tiny grits and some small ones”

407

pottery from the “midianite heartland”?

Place of illustration

Place of storage

Wright 1959: photo fig. 16b Photo: Glueck 1967: 9, fig. 1, no. 2; Drawing: Glueck 1967: 19 fig. 5, no. 1a

Palestine Archaeological Museum (now Rockefeller “a complex series of horizontal panels with trellis designs, and Museum) groups of vertical lines with “x” designs between some of them. To the left of it is a candelabra-like pattern, also with lines of dark brown paint” Glueck 1967: 16 fig. 4, no. 5

Whereabouts unknown75

“On the upper part, between two bordering lines, a sharply canted, zigzag, dark brown horizontal line, with open ‘v’s, alternating in regular or inverted position, occupying the triangular open spaces created by the zigzag line. Beneath is a horiz. row of round dots of paint, followed farther down by two more horiz. lines. The lowest one borders a series of vertical lines, the one on the left extending up to the bottom of the lower border of the zigzag line panel, with a more or less circular pendant, in which a small painted dot is set, occupying each corner angle created by the two outer vertical lines meeting the lowest horizontal one” Glueck 1967: 16 fig. 4, no. 3 “heavy creamy-buff slip. The top of the rim is decorated with narrow, long, horizontal rectangles of blackish reddish-brown paint, with a single, horizontal line immediately beneath the outturned rim. Below it, are two horizontal bands, separated by a row of circular blackish brown dots, with this same decoration being repeated on the top of the shoulder of the jug. The central painted panel, extending horizontally upward from a horizontal line of blackish-brown paint at the bottom of the tapering neck, consists of a series of long, narrow, inverted triangles, whose bases merge at the top. Each of the bottom halves of the intervening triangular spaces separating the inverted painted triangles, contains three vertical, stalagmite-like lines of blackish-brown paint”

75

See above footnote 44.

Whereabouts unknown

Illustrations (not to scale)

408

luciani

table 1

Previously published Qurayyah painted ware fragments from Tell Kheleifeh (cont.)

Count Reg. no.

Vessel fragment

Place of publication

#4

Rim, neck and shoulder of jug

Glueck 1967: 13, fn. 27

“outcurving shoulder, tapering neck, and slightly outturned rim [of] jug”

“light brownish-buff slip, horizontal and vertical lines of dark reddish-brown paint”

Flat base

Glueck 1967: 15, fn. 33

“wheel-made”

“comparatively coarse, grayish buff ware with numerous tiny and some medium grits”

?

Pratico 1993a: 49

#5

#6

10,082

10,083

?

of this specimen.76 Therefore, on the basis of decoration alone, it would be very difficult to attain a reliable attribution to the qpw. However, both the surface finishing with slip on the inside and out and the type of matt painting as well as the fabric, featuring very visible grit, are very homogeneous with other known painted pottery specimens from Qurayyah and are very different from the Iron Age pottery from Tell Kheleifeh seen in the Harvard Semitic Museum boxes.77 Furthermore, this type of decoration does not seem to appear in other painted groups, e.g. so-called ‘Edomite’ pottery, the latter displaying a yet simpler pattern of banded painted decoration.78 The published evidence that feature a cross-hatched decoration similar to our # 8 shows it occurring both on the straight sides of a deep bowl—a shape most likely different from

76 77

78

It is possible that the line defining the band containing the cross-hatch was on the part of the sherd that broke off. If anything, its decoration could look like so-called Sanaʿiyeh Ware, the Iron Age production well attested at Taymāʾ, (for example Hausleiter 2014: 407–409), but the fabric resembles more qpw. Oakeshott 1983, Bienkowski 2002.

409

pottery from the “midianite heartland”?

Place of illustration

Place of storage

Glueck 1967: fig. 4, no. 4

Whereabouts unknown

Illustrations (not to scale)

“on the neck are two groups of four vertical lines each, separated by a vertical row of dots, with the entire panel bordered by horizontal lines of paint. Above and below this panel are two horizontal lines of paint, with another painted design beginning at the bottom of the shoulder” Glueck 1967: 16 fig. 4, no. 2

Harvard Semitic Museum

“creamy buff slip originally covered all of the outer surface. Long, narrow, inverted isosceles triangles of dark reddishbrown paint are visible on the remnant of the bottom side of the vessel, with their points overlapping the outer edge of the flat base, which, too, was originally slip-covered and painted” none79

Whereabouts unknown

?

the one this fragment was part of—but also as bands on globular juglets.80 There is, however, at least one published parallel in a jug which comes from Site 2 at Timnaʿ and one unpublished fragment of body sherd of a globular vessel from Qurayyah that both in decoration and in shape strongly resemble the piece discussed here81 and make its classification as qpw convincing. # 9 (see above Table 2 and below fig. 2 top): The straight neck of a jug with inner groove and a monochrome, simple banded painted decoration on the outer surface and characterised by the typical, grit/shale tempered fabric, finds exact parallels both in shape and decoration design in specimens from Qurayyah82 and in painted design from a fragment from Timnaʿ.83 Its allocation to the qpw pottery class is certain.

79 80 81 82 83

See above discussion and footnote 50. See references in footnote 61 above. For Timnaʿ: Rothenberg and Glass 1983: fig. 6.7; for Qurayyah: Hausleiter in preparation Photos qu 10_0096 and qu 10_0097, the central fragment. Ingraham et al. 1981: fig. 78.20. Avner 2014: 115: fig. 7.10.

410 table 2

Register number (same as 10,083 above # 5)

#7

#8

10,057

10,093

luciani Previously unpublished qpw pottery from Tell Kheleifeh at the Harvard Semitic Museum Vessel fragment

Shape type and diameter

Fabric number and colour variant

Surface type out

Surface type in

Base

Flat base of jug(let)?

4

Thick Slip

Thin Slip

Body sherd

Body sherd

ø: 10 cm

(7.5yr 6/2 = pinkish grey)

Neck of jug

1

ø: 6 cm

(10yr 7/3 = very pale brown)

Closed container, globular jar?

5

ø: 20 cm #9

# 10

# 11

11,016 (a)

11,016 (b)

No id

Rim

Rim

Body sherd

Vertical neck of small jar

(10yr 8/3 = very pale brown) Slip

Slip

Slip

Slip

Slip

Thin Slip

Slip

Slip

Self slip? Not well preserved

Self slip

(10yr 6/3 = pale brown) 2

ø: 12,4 cm

(10yr 7/4 = very pale brown)

Deep bowl

3

ø: 20 cm

(10yr 7/4 = very pale brown)

Closed shape, small 1 globular jar? (5yr 8/2 = pinkish ø: c. 11 cm white)

411

pottery from the “midianite heartland”?

Decoration type

Surface colour out

Surface colour in

Technique

Bichrome brown (5yr 3/4 = dark reddish brown) and black (5yr 2.5/1 = black) painting

(10yr 7/4 = very pale brown)

(7.5yr 6/2 = pinkish grey)

Wheel-made

Bichrome red (10yr 4/6 = dark yellowish brown) and black (2.5r 3/0) painting

(10r 5/1 = reddish grey)

(7.5yr 6/6 = reddish yellow)

Wheel-made

Monochrome brown (10yr 3/2 = very dark greyish brown) painting

(10yr 6/3 = pale brown)

(10yr 6/3 = pale brown)

Wheel-made

Monochrome brown (7.5yr 3/2 = dark brown) painting

(7.5yr 6/6 = reddish yellow)

(10yr 7/4 = very pale brown)

Wheel-made

Monochrome? In: brown (5yr 5/6 = yellowish red) painting. Out: brown (5yr 3/3 = dark reddish brown) painting

(10yr 7/4 = very pale brown)

(10yr 7/3 = very Wheel-made pale brown)

Monochrome black (5yr 2.5/1 = black) painting

(5yr 8/3= pink)

(5yr 5/4 = reddish brown)

Wheel-made

Illustrations (Not to scale)

412

luciani

# 10 (see above Table 2 and below fig. 2 middle): The decoration on this round-sided, simple bowl with a cross-hatched design framed by lines in a vertical band is rather well attested and can be found both on open84 and closed shapes. The different colours of the decoration are probably due more to colour zoning than to intentional bichromic decoration. For its fabric, shape85 and decoration, attributing this specimen to the qpw group is wholly uncontroversial. # 11 (see above Table 2 and below fig. 2 bottom): Both inner and outer surface of this piece are badly weathered, even if the type of paint and design of the decoration can be made out with reasonable confidence. The fabric resembles the finer ones attested at Qurayyah. The paint was possibly more lustrous than the usual monochrome matt paint seen in qpw, but this is well attested on small globular juglets86 of this pottery class. Even if the decoration design is not very specific, some comparisons can be found for it in Qurayyah87 although they are not exactly identical and the shape of the vessel is different. However, considering the combination of fabric and paint, this specimen can be classed as qpw. In sum, there are a minimum of ten documented pottery sherds88—and a potential additional one—from Tell Kheleifeh that may be identified (in seven cases surely, in three extremely likely) as qpw because of the type and design of their decoration but even more so because this is combined with a specific and unique fabric that we recognise as identical to the one produced in the Ḥijāz.

Defining qpw and Its Date of Use One consequence of lowering the chronology of the buildings in Tell Kheleifeh from the 10th to the late 8th–6th century bce, has been that the duration of qpw has been stretched into continuing into the late Iron Age.89 But how can

84 85 86

87 88 89

Ingraham et al. 1981: fig. 78.1. For simple bowl constituting the majority of the qpw shape repertory see Parr 1982: 128. One of the well-preserved globular jugs we are referring to here is from Taymāʾ (al-Najm 2007: 39 first vessel from left), the other is unprovenanced and bichrome (Rothenberg 1998: 175 fig. 5). Both display a more lustrous paint than the majority of qpw vessels. Parr et al. 1970: fig. 16.3 and 4 and Ingraham et al. 1981: fig. 79.25; Hausleiter in preparation Photos qu 10_0123 and qu 10_0124 (and qu 10_0092 and qu 10_0093)? But see above footnote 5 for the final tally of qpw sherds from Tell Kheleifeh. Bimson and Tebes 2009 [2010] with previous references for the 8th century bce. For

pottery from the “midianite heartland”? table 3

413

Fabric types in the qpw sherds from Tell Kheleifeh at the Harvard Semitic Museum

Fabric no.

Fabric description

1

Clay: homogeneous clay, very slightly to slightly porous; Inclusions: all are mineral, most likely shale: (a) small red frequent, (b) medium red very frequent, (c) sub-angular medium black-grey and red grit (shale) rare.

2

Clay: homogeneous clay, very slightly to slightly porous; Inclusions: all are mineral, most likely shale: (a) small red frequent, (b) medium red very frequent, (c) sub-angular medium red grit (shale) frequent.

3

Clay: homogeneous clay, very slightly to slightly porous; Inclusions: all are mineral, most likely shale: (a) small red frequent, (b) medium red very frequent, (c) sub-angular medium black grit (shale) very frequent

4

Clay: homogeneous clay, very slightly to slightly porous; Inclusions: all are mineral, most likely shale: (a) small red frequent, (b) medium red very frequent, (c) with sub-angular medium red grit (shale) very frequent.

5

Clay: homogenous porous clay; Inclusions: all are mineral, most likely shale: (a) medium sub-angular black grit (shale) very frequent, (b) small black grit (shale) frequent.

we correctly frame chronologically the qpw phenomenon nowadays? In order to reach a coherent date for the appearance and use of qpw, I will discuss both re-evaluations of old results and new research.90 In his 2014 publication of the soundings carried out in the mid-1980s inside the Timnaʿ Temple (i.e. Site 200), Avner presents us with a revised, detailed analysis of the site’s stratigraphy. As is well-known, the several dozen qpw sherds found in the so-called Egyptian Temple at Timnaʿ have represented the very first and for a long time almost sole chronological anchoring for

90

partially older, but still Iron Age dates (10th century bce) for ‘Midianite’ pottery see Smith and Levy 2014: 412. Avner 2014, Erikson-Gini 2014, Singer and Avitz 2014, and Mumford 2015 on older data; Hausleiter 2014, Levy et al. 2014b, Luciani and Alsaud 2016, Luciani forthcoming a and forthcoming b on new research.

414

luciani

table 4

Find-spot of qpw sherds # 1–5 and # 7–10 according to the hand-written Tell Kheleifeh registries in the Harvard Semitic Museum

Reg.

Date

Entry in Tell Kheleifeh registry

#1

248

1937–1938, p. 27

Neck of broken, painted jar, l:12; 50cm Room 34

#2

3,016

1939 April 25, p. 1 Fragment of painted jug. Rm. 50, debris of iii (?)

#3

7,079

1940, p. 19

Painted sherd. Street 90, debris

#7

10,057

1940, p. 41

Painted sherd. Rm 30

#4

10,082

1940, p. 42

Painted sherd. Rm 125

#5

10,083

1940, p. 42

Fragment of bottom of jug, painted. Rm. 50

#8

10,093

1940, p. 43

Painted sherd. Debris north of Rm. 43

1940, p. 45

Painted sherds. Same as above [i.e., Trench. Square j 14. Debris]

# 9 11,016 (a) # 10 11,016 (b)

this pottery production. Site 200 features different imported materials, but the inscribed objects especially—with the names of Pharaos from Seti i to Ramesses ii and down to Ramesses v—have been most extensively, and sometimes superficially, used for pegging the date of the temple, and by extension, of qpw. It would be useless for me to repeat here in detail Avner’s assessment. For brevity it is enough to say that they point to the presence of not two or three but eight different phases of construction and use of the ‘temple’, only for the preRoman period. His observations stem both from a thorough re-examination of previous publications and their incongruences but also from carrying out three small soundings (Probes 1–3) inside the very premises of the ‘temple’. This was necessary for preparing conservation measures for the site. It must be underlined that the probes are very limited in extent91 and yet they represent the first detailed investigations of the stratigraphy of the ‘temple’ and as such carry an

91

Avner 2014: 111–113: Probe 1: 30 × 60 cm; Probe 2: between Wall 3 and the line of masseboth but the dimensions are not explicitly stated; Probe 3: 120×50cm. It seems that the plan of

pottery from the “midianite heartland”?

415

intrinsically higher information potential. They have proven critical for checking the sequence as known from published reports. His main finding is the fact that the ‘Egyptian temple’ was stricto sensu not a temple, and that both stratigraphically and chronologically, the square enclosure where most Egyptian finds (inscribed and not) were recovered was only added to a pre-existing local sanctuary, after this had already existed for some time. Furthermore, the presence of an Egyptian chapel or kiosk—as Avner dubs it more aptly than ‘temple’—was but an intermezzo for the local sanctuary that continued to function after this ‘foreign’ phase. Avner is still inclined to date the construction of the Egyptian chapel to sometime during the reign of Ramesses ii, even if at least one inscribed scarab of Seti i had reached the sanctuary. But even more important is the evidence that qpw sherds—together with so-called Negevite ware and local wheel-made pottery—are present in Avner’s Phase 1,92 underneath the Phase 2 red pavement, the phase preceding the construction of the Egyptian naos in Phase 3 (at the time, most probably, of Ramesses ii). This information, while not enabling us to be any more precise on the entire duration of qpw pottery is important because it allows—actually it suggests—that the painted vessels may have been in use and have reached the southern ʿArabah already before the 13th century bce. The 13th–12th century chronological peg from Site 200 in Timnaʿ may thus come to be regarded as a terminus ad quem or terminus post quem non for the beginning of use of qpw in the southern ʿArabah. As P. Parr recognised twenty years ago: “strictly speaking, there is nothing to preclude an earlier date [than the thirteenth century bce] for this ware”.93 This early dating of qpw to the full Late Bronze Age is confirmed by at least one additional site in the vicinity of the temple and belonging to the same mining district so intensively investigated by Rothenberg:94 Site 2. According to recent re-excavation by T. Erikson-Gini, radiocarbon dates for this site range from the 15th to the 13th century bce.95 And even if some have cast doubts on the reliability of this data,96 even critics have acknowledged that a date in

92 93 94 95 96

the probes on fig. 1 (ibidem: 106) is designed to point out the general area where the small trenches were dug. Avner 2014: 113–116, especially 115, fig. 7.10. Parr 1996: 214. Rothenberg 1967: xii–xxviii and idem 1972: 71–111. Erikson and Gini 2014: esp. 55. Because of the distortion that ‘old wood effect’ may potentially have caused on these dates. However, in the context of mining sites and operations, in the absence of long-standing, sumptuous buildings, i.e. the potential source of construction wood to be recycled, I fail to see that the phenomenon of reuse of old wood should constitute a major concern.

416

figure 1.5, 7–8

luciani

Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (with graphic scales). pencil drawing and photos: m. luciani; digital drawing: l. machel. courtesy of the harvard semitic museum.

pottery from the “midianite heartland”?

figure 2.9–11

Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (with graphic scales). pencil drawing and photos: m. luciani; digital drawing: l. machel. courtesy of the harvard semitic museum.

417

418

luciani

the 13th century for the life of Site 2 is granted.97 The qpw assemblage found on Site 2 both during old and new excavations is plenty and impeccable as far as its identification with the northern Ḥijāzī ware is concerned. Interestingly enough, it features twice that motif with “the semi-circle with concentric dot on top of converging lines alternating a couple of parallel vertical lines”98 that I have described as very typical of qpw (present in Qurayyah and Tell Qudayrāt) and is attested on our # 2 sherd found by Glueck at Tell Kheleifeh. This motif does not appear to be attested in published pottery from Timnaʿ Sites 200, 198, 3 or 30, but it is definitely too early to say whether this might be evidence of chronological anteriority. Further recent research in the southern ʿArabah has pointed out to Late Bronze Age occupation (later 13th and 12th century bce) also on Timnaʿ Site 3 with single, but recognisable99 qpw sherds. A sizable collection of seemingly qpw pottery stems also from old research at Site 30100 in Timnaʿ. Renewed excavations at this site have produced two possible101 qpw sherds and radiocarbon dates pointing to the late 12th to early 10th century bce. If current revisions indicate that qpw is a Late Bronze Age phenomenon, from the 15th to 12th century bce (see below the data from Taymāʾ and footnote 115) with possible extensions into the 11th century bce, it is important to keep in mind that “dating archaeometallurgical deposits in the southern ʿArabah based on ceramic assemblages is problematic, as the ceramic finds are usually slim, as demonstrated by the new excavations at Site 30 and previous excavations at

97

98 99 100 101

Yagel et al. 2016: 15 footnote 3. I wonder whether the presence/absence of pig bones (Site 2 vs. Site 200) should not be better read as to suggest chronological anteriority of Site 2 to Site 200 rather than ethnic affiliation (ibidem: 14). In the latter case it is very difficult to explain their complete absence from Site 200, an ‘Egyptian’ site if there ever was one in the ʿArabah (contra Sapir-Hen and Ben-Yosef 2014: 787) even admitting potential cultic vs. non-cultic uses. See above footnote 62 for references. Because published as photos, Yagel et al. 2016: fig. 12, especially no. 5–7 seem convincing as qpw specimens. Rothenberg 1980: Abb. 210. Ben-Yosef et al. 2012: fig. 11 and 12. It is very commendable that also photographs of the two attested sherds have been published. In both cases neither shape nor decoration are extremely typical of qpw but the surface of ibidem fig. 12a does put it on the range. The line drawings published by B. Rothenberg (see previous footnote) feature some shapes and decorations that look typical (no. 7–9, 11, 12, 15) and some that look slightly different (younger?) (no. 1–5). I would frame Ben-Yosef et al. 2012: fig. 11 and 12 among the latter group.

pottery from the “midianite heartland”?

419

this site and elsewhere in the region”.102 This assertion cannot but be wholly embraced. But I wish here to underline the entire range of implications this argument brings forth, since the logic works both ways. If single, mostly rather weathered and/or difficult to determine sherds are not reliable as dating tools, also radiocarbon determinations of the metallurgical activities on a certain site are not going to be decisive103 in dating the last occurrence of those sherds and/or of an entire pottery class—here qpw—the sherds purportedly belong to. With all the potential for these single occurrences for being either a further, local (?) development of earlier qpw or more simply residual104 fragments, radiocarbon dates on the strata they were found in, tell us two things: 1/ when those metallurgical activities took place and 2/ that those sherds—unless they are intrusive—may not have come to be deposited after this time. But they still may have been produced and used at different periods before the time indicated by the measured radiocarbon date. The Southern ʿArabah data has been connected with research further north, in the Wādī Faynān. The Edom Lowlands Regional Archaeology Project (elrap) succeeds in documenting an Iron Age assemblage, which features several painted specimens (called ‘Edomite’ pottery),105 but only single sherds (three through petrographic analysis and two possible additional ones) that may be identified as qpw.106 The latter all stem from Khirbat an-Nuḥās

102 103

104

105

106

Ben-Yosef et al. 2012: 62. The pattern of decreasing quantities of qpw in sites that are dated later than the Late Bronze Age (now recognised by Yagel et al. 2016: 15) is an additional ground for questioning the hypothesis of a continued use of qpw well into the 10th century bce. For the often underestimated phenomenon of residual pottery in archaeological excavations and its role in our defining the relationship between systemic vs. archaeological contexts see Luciani 2014c: 11. No field archaeologist can thoroughly exclude limited, earlier use of the site, e.g., in the vicinity but outside of the excavated trenches. Smith and Levy 2014. This pottery features overall similarities to the Iron Age assemblages from Tell Kheleifeh I saw at the Harvard Semitic Museum (on the date of this pottery see now Finkelstein 2014: 118–122). With Edom being an ill-defined entity (Ben-Yosef et al. 2012: 64 footnote 17 and Finkelstein 2005 and 2010: 15 and also Kafafi 2014), the designation ‘Edomite’ for the material culture of Southern Jordan is suboptimal. Since these sherds are made available through publication only as line drawings and they are very small it is really difficult to be sure about their determination. As pointed out recently by S. Sherratt (2013: 643–644 and footnote 30): “… we need the kind of information that can only be obtained from fully illustrated publication (with photographs, not only scaled-down line drawings)” … “We need to know what it looks and feels like: in other words, not only the details of shapes and decoration involved, but also of fabric consistencies, surface treatments and techniques of construction, and the extent to which

420

luciani

Area m,107 where 13th century bce radiocarbon dates have been obtained.108 Therefore, in view of the data discussed above, the other sherds found in the context of the elrap published as qpw, mostly single body sherds of limited size—even allowing for regional and chronological varieties of this pottery tradition that might still not be well known to us—are more safely read as residual109 rather than proving a 10th century bce date for qpw. Further confirmation of a Late Bronze Age date for qpw comes from yet further afar, i.e. from the site of its northernmost attestation. In the Amman Airport excavation110 several bowls and at least three fragments of open and two of closed shapes111 of qpw were uncovered.112 The site was found and partially excavated by a bulldozer and it revealed a sequence of different layers as well as imported finds stemming from as early as the second millennium bce. Notwithstanding this non-ideal find situation, the association of the qpw pottery sherds with at the latest Late Helladic iiib sherds seems granted.113 Since also the local pottery assemblage points to the Late Bronze ii(b), i.e. the

107 108

109

110 111 112 113

these vary both within and between sites, within and between regions, and over time”. To give a further example: Yagel et al.’s 2016 determination of fig. 12: no. 5–7 sherds as qpw is comprehensible only because the accompanying published photo is very explicit on their fabric. For the reader this identification would not have been intelligible if only line drawings had been provided. See also Singer-Avitz 2014: 136–137 and footnote 101 above. Smith and Levy 2014: Table 4.7 no. 9 and Table 4.8 no. 7 and 8. Levy et al. 2014a: 89, 131–151 and especially p. 150 Table 2.86; Smith and Levy 2014: 414. Area s has single sherds which might be qpw and has dates from the late 11th century (but at least one date also localized in the very late 13th to mid-12th century bce, p. 170 Table 2.11, date OxA-12169 = 1124–1024 / 1206–1002 cal bce). The fact that qpw is found mostly outside the buildings seems to confirm this interpretation (Smith and Levy 2014: 448). For a similar interpretation of Khirbat an-Nuḥās as a fort built after and overlapping the copper production layers (and its synchronism with Tell Kheleifeh), see Finkelstein 2005, esp. 123. Hankey 1995: 181 “except for am 6245 (Qurraya [sic] ware) and am6279 the pottery would be at home in most lb ii sites in Palestine”. Hankey 1995: pl. 14.4 left hand side. Hankey 1995 does not indicate explicitly the total sum of fragments. Mumford 2015 pins them at 9+. Mumford 2015: 190. Even if he writes: “The absence of Late Helladic/Mycenaean iii c pottery (c. 1190–1050 bc) and the predominance of lb iib potsherds (Kafafi 1983: 36) suggest that occupation did not extend much beyond this period, or that such imports had virtually ceased to reach this region. Kafafi (1983: 35–36) admitted, however, that some of the regional, Levantine pottery may extend into Iron i, thereby allowing for minimal occupation at the advent of Iron Age ia”.

pottery from the “midianite heartland”?

421

13th century bce,114 it seems that this should be taken as the terminus ad quem or terminus post quem non for the attestations of qpw at the Amman Airport site. Data from Taymāʾ now encases stratigraphically and chronologically the origin and outset of qpw after the initial production of Red Burnished/Barbotine pottery, and its final use before Early Iron Age (Area o) painted ware kicks in, with dates ranging from the late 15th to the 14th centuries calBCE.115 Similarly, the picture emerging from recent field research in Qurayyah only reinforces this time-frame. During the 2015 excavation campaign of the Saudi-Austrian Joint Archaeological Project at Qurayyah,116 we excavated and sampled one of the pottery kilns identified by previous surveys.117 The kiln (our Area a), which produced masses of simple, undecorated and decorated qpw pottery—often with beautiful, figurative painted motifs featuring different, mostly wild animals— has given us four very coherent radiocarbon dates: two for the beginning of the Late Bronze Age (16th to early 15th century calBCE) as the phase of use of the pyrotechonological installation. The other two belong to the time after the pottery production activities ended and the pit of the kiln was reused for the intentional secondary disposal of human remains. These later human bones, fittingly, have produced radiocarbon dates in the 13th century to early 12th calBCE.118 Even if these are but a few (all in all six)119 radiocarbon determinations and the very first ones from the site of Qurayyah, so that utter caution is called for,

114 115 116 117 118

119

Kafafi 1983. Hausleiter 2014: 406 and esp. 407 and Luciani 2014b. Luciani and Alsaud 2015 and 2016. Parr et al. 1970 and Ingraham et al. 1981, Luciani et al. 2015. We do not have pottery associated with these secondary burials but plenty of faience beads. However, on the surface of the site and residual in the Nabataean period layers (Area d) we do find painted pottery that looks slightly different from qpw produced in the kiln, with coarser design. Anyhow, since their context of provenance is not datable yet and we are dealing but with single pieces, it is too early to say whether we can really speak of chronological differences (see above footnotes 70–71 and discussion of # 5). Be this as it may, both Parr (1982: 28: “it covers a somewhat wider chronological range than originally thought”) and Rothenberg and Glass (1983: 101 and 119 footnote 26) recognised that there might be different time ranges of use of this ware and therefore for dating the sites where it was found. One radiocarbon determination was made on the bioapatite of a human femur from the burial building (associated with Barbotine Pottery) in Area b, two on tamarisk from overlapping layers of use of the pottery kiln, two on the bones buried in the kiln after its phase of use, all in Area a and one stems from tamarisk from settlement layers from

422

luciani

we must underline that all are strongly consistent with our stratigraphic results, systematically featuring older dates for the samples stemming from deposits (sus) that are stratigraphically underlying and younger ones for overlapping sus. Not only human bones but also tamarisk twigs may be considered shortlived samples, since the latter was used as fuel for the pottery kiln and an ‘old wood’ effect may reasonably be ruled out.120 While these dates for the production of qpw pottery are very early compared to the ones we are used to from past research, not only do they tally with those from Taymāʾ,121 they also agree with the fact that both in Taymāʾ and in Qurayyah bichrome and monochrome painted qpw follow closely in time a transitional production that depends from local, earlier, grey/brown and red burnished productions, the so-called Barbotine Pottery. The latter is a Middle Bronze age122 ceramic manufacture—mostly producing metal skeumorphs— attested in Qurayyah,123 in several locations within Taymāʾ124 and also in Khuraybah/Dadan, within the oasis of Al-ʿUlā.125 Both in Taymāʾ and Qurayyah it can now be shown that the qpw repertory developed locally a painted decoration by expanding on themes that were initially designed within the Barbotine pottery tradition as decoration in relief and using bichromic effect. One specimen from Timnaʿ clearly illustrates this genetic link126 between the elder Barbotine pottery (fig. 3a a specimen from Taymāʾ) and the younger

120 121

122

123

124 125

126

Area d. All dates were measured at the laboratory of the Center of Applied Isotope Studies University of Athens, ga, usa, calibrated with OxCal v.4.2.4 Bronk Ramsey 2013 and are published in detail in Luciani and Alsaud 2016 and Luciani forthcoming a and b. See above footnote 96. Hausleiter 2014: 406–408 (“15th/14th century bce or even earlier” for Barbotine pottery and 13th–12th century bce for qpw) and for a reassessment pointing to higher dates on the basis of new measurements now Hausleiter and Zur 2016, 153–154 for Red Burnished and Barbotine and Luciani 2014b and Luciani and Machel forthcoming for qpw. Hausleiter 2014: 406 and footnote 115 above. In Qurayyah we have one date from the bioapatite of human bones associated with this pottery and it points to the 17th century bce (Luciani forthcoming b), for references see footnote 119 above. Parr et al. 1970, Luciani 2014b, fig. 10 central fragment and eadem 2016a, fig. 10 central fragment. The Saudi-Austrian excavations have found evidence also of the production of this ware in Qurayyah, see Luciani and Alsaud 2015 and 2016, Luciani forthcoming b. Hausleiter 2014. Al-Shehry 2014: 205, photo 88 top one; 220, photo 103 top one, 106 top one; 266 photo 154 bottom piece and 476, line drawing 13; 513 line drawing 39?; 516 line drawing 49; 525 line drawing 525?; 542 line drawing 156. On the basis of one fragment alone, it is difficult to say whether it must be considered transitional also from the chronological point of view. Only when we will be able to discern

pottery from the “midianite heartland”?

figure 3

423

Barbotine pottery from Taymāʾ (a) compared to qpw pottery from Timnaʿ (b). a: Hausleiter 2014: fig. 5a; b: Rothenberg 1988: pl. 18 (colour) and fig. 5, no. 6 (line drawing).

white-on-red painted decoration on a qpw fragment (fig. 3b from Timnaʿ Site 200).127 Solidly framing qpw as a Late Bronze Age phenomenon in Qurayyah at least, and most likely an early Late Bronze one at that, allows us not only to account for the several instances in which this pottery is associated with contexts with second millennium dates (Timnaʿ Sites 200, 2 and 3, Taymāʾ, Qurayyah, Lachish, Amman Airport, Tell el-ʿAjjūl, Tell al-Farʿah South, Tell Jdūr128) but also to explain the obvious similarities with all painted pottery productions that emerged in the Southern Levant at the transition between the end of the Middle Bronze and the beginning of the Late Bronze Age such as: Levantine/Cypriot Bichrome Painted, Chocolate-on-White or Late Cypriot productions such as White Painted i as well as New Kingdom Egyptian faience.129 The resemblances between qpw and these other Eastern Mediterranean pottery assemblages,130 though never spelling imitation,131 are clearly recognis-

127

128 129 130 131

how the production of qpw changed over time will we be able to say whether the Timnaʿ Site 200 assemblage reflects the older phase, in use already before the Egyptian presence, or belongs to the younger one, while containing, e.g. ‘archaic’ pieces. Two specimens from the 2015 campaign in Qurayyah display a Barbotine motive realised not in relief but with white paint. Other fragments also feature both Barbotine and finger painting decoration on the same fragment (Luciani and Alsaud 2015). In Square q3 in Taymāʾ there was at least one small, hole mouth ‘Barbotine shape’ displaying painted qpw decoration (Luciani and Machel forthcoming). The latter is very likely to be considered chronologically transitional between Barbotine pottery and qpw. For details see Luciani forthcoming a. For detailed references and chronological reassessment see Intilia 2016. Parr 1982: 129, and footnotes 21 and 22; idem 1988: 74 and footnote 13. Similarities with Cypriot White Painted i productions had not been noticed before, and are discussed in Luciani forthcoming a. Peter Parr (1979, 1982, 1988, 1996) was adamant—and rightly so—on this aspect from the beginning and never tired of underlining that qpw was in no way the outcome

424

luciani

able and have been remarked upon several times.132 It is true that the degree and essence of these parallels was poorly understood in the past. It was precisely the combination of an obviously local production (fabrics, vessel shapes and distinctive regional motifs) and widespread use of a ‘second millennium bce Mediterranean syntax’ coupled with single foreign motifs that made clarifying the relationship with other painted wares so challenging.133 We cannot but underline the absolute originality, the individual—at once both regional and international—character of the qpw manufacture.134 With the new radiocarbon determinations, the supposed 200-year chronological gap135 that was seen as a serious obstacle for these comparisons may prove not to have existed. While it must be confirmed that no Aegean or Cypriot pottery136 or Egyptian faience vessels of the second millennium bce have been found in any of the oases of north-west Arabia thus far, it is now sure that the local population had long been acquainted with the Southern Levantine world and its material culture. As the Syro-Levantine bronze weapons found in Taymāʾ137 testify, the northern Ḥijāz started sharing elements of the material culture of the Southern Levant already at the beginning of the second millennium bce at the latest, so that seeing a development of local but related— though never identical!—bichrome painted pottery production at the turn from the first half to the second half of that millennium should not come as a surprise. Whatever increasing radiocarbon determinations will eventually confirm the initial date of production of qpw to have been, Qurayyah has thus far not produced any dates on the final period of use of its distinctive pottery. But the fact that Taymāʾ documents a shift to an altogether different painted assemblage138 in the 10th century calBCE is relevant for understanding the

132 133 134 135 136 137 138

of straightforward imitation of other Mediterranean productions but one developed locally (their producers, “genuine North Arabians”, Parr 1979: 40). Our data fully confirms this initial estimate both for the make-up of the pottery manufacture and now also its chronology. Parr et al. 1970, Dayton 1972 (but see corrections in Parr 1988), Parr 1979, 1982, 1988 and 1996. Parr 1988 passim. The mechanisms that have brought to this peculiar mixture will be investigated in detail once the qpw assemblage from Qurayyah will be documented more extensively. Parr 1982: 129. Confirming Parr 1988: 74 and 1996: 215 and contra Dayton 1972: 27–29. Hausleiter 2015, Hausleiter and Zur 2016: esp. 148–150, 151, fig. 12. This early Iron Age pottery has thus far not been found in Qurayyah. Because this assem-

pottery from the “midianite heartland”?

425

developments in pottery production in the northern Ḥijāz. It remains to be seen which part of this material cultural history applies to the areas north of this region, i.e. the Southern Levant, where qpw is attested. However, a recent review of the published qpw corpus has brought L. Singer-Avitz to suggest that “there are no data confirming the notion that the qpw survived for several centuries, long after its manufacture at its place of origin had come to an end”.139 Therefore, the chronological gap between the latest possible date for qpw— most likely not significantly later than the 11th century bce—and the late 8th century bce date (at the earliest) for the Tell Kheleifeh architecture must lead us to interpret the dozen qpw sherds found there as the residual material traces of a Late Bronze Age140 occupation, chronologically and functionally distinct from late Iron Age building activities at the site. As such, Tell Kheleifeh does not offer any evidence for extending the duration of qpw into the late Iron Age.

Tell Kheleifeh as Part of the ‘Midianite Heartland’? As we have seen above, Late Bronze Age Tell Kheleifeh was by no means isolated in the Southern ʿArabah, where a number of sites from the Timnaʿ complex (Sites 200, 2, 3, etc.) to Yotvata to the north, Jazīrat Firʿawn to the south on the coast of the Red Sea featured comparable pottery. However, to expand significantly on the weight and meaning of the Tell Kheleifeh pottery record from the Harvard Semitic Museum, composed as it is of only slightly over two handfuls of sherds, requires utmost consideration. It may best be done

139

140

blage is also painted and it partially re-uses some of the simpler qpw motifs, it has in the past at times been confused with qpw and partially with later Iron Age, so-called Sanaʿiyeh pottery (e.g., Bawden 1983 as recognised by Parr 1988). It is now well documented and better understood from the excavation of Area o in Taymāʾ, Hausleiter 2014: 403 “(14c dates ranging from the 12th to 9th centuries bc with a peak in the 10th century bc)”. Singer-Avitz 2014: 138 and footnote 121: “The homogenous nature of qpw, expressed by its material, shapes and decoration (Rothenberg and Glass 1983: 102), may also indicate that it was used for only a brief period of time”. More data is needed to unravel the relationship between a substantially homogenous pottery manufacture and its actual production span. With our work in Qurayyah, we aim at clarifying this and other aspects in the near future. Or at the very latest transitional Late Bronze / Early Iron Age. For the qpw sherds indicating a “settlement that predated the fortress” see also Singer-Avitz 2014: 132 and Finkelstein 2014: 122 “The earliest finds at Tell el-Kheleifeh—the Qurayyah Ware sherds—cannot be affiliated with any architectural remains” and ibidem for the presence of a chronological gap before the Iron Age levels.

426

luciani

by considering two issues: (1) The position of qpw from Qurayyah within the wider Southern Levantine/Eastern Mediterranean region; (2) The role of qpw in sites to the north of Qurayyah, from the Southern ʿArabah through the Gulf of ʿAqaba to the Red Sea where qpw is identical141 to the northern Ḥijāzī ware but constituted only a portion of the local assemblages. The fact that this painted pottery production, notwithstanding reverberating contemporaneous Eastern Mediterranean manufactures, featured a fully original and strongly regional elaboration of local and foreign motifs is revealing. Parr spoke of it in terms of “hybridity”142 long before this became the buzz word in post-colonial archaeological studies it is being used as now.143 However, it is difficult to find an apt term to describe this material, unencumbered by theoretical frames of ill-defined, unequal economic and political relationships that are dominating the field in our days.144 We interpret qpw as the material correlate of a distinctive identity localised in the northern Ḥijāz. This identity may have been defined through political, social, economic, performative, and not necessarily exclusively ethnic components nor represent a homogeneous or single element. The qpw pottery assemblage points to an upgrading of the interconnectedness and connectivity145 of this region with the wider Levantine

141

142

143 144

145

Whether this identity or similarity was due to the fact that the pottery was imported or produced locally by knowledgeable/Northern Ḥijāzī potters still remains to be established (see Daszkiewicz in Hausleiter 2014). The Tell Kheleifeh sherds, anyhow, seem to perfectly fit the make-up of the qpw assemblage from Qurayyah, located 150km to the south-east, in the Northern Ḥijāz. Parr (1982: 129) proposed that the Aegean motifs we find in qpw were mediated through Egyptian faience bowls and had not been seen directly. I am grateful to Robert Koehl for accepting to review together qpw pottery and discuss with me its potential degree of ‘Aegean-ness’. The development of qpw from local wares, the references to bichromic design and to some typically Cypriot, White Slip i motifs do not appear to have been transmitted through Egypt and are a hint that contacts were there and the process of selection was intentional and distinctive (see Luciani forthcoming a). For a critical review of the term, see Stockhammer 2013. See different contributions on this theme in van Pelt 2013. We are just starting to investigate the socio-political and economic structure of the Northern Ḥijāzī oases in the second millennium bce and superimposing a model at this stage risks uselessly muddying the picture (see Luciani 2016a: 43 footnote 168). As of now, e.g., the argument that Egypt must have played a stimulating role, or was even the prime mover, in the formation of these urban oases (Parr 1988) is not borne by any visible traces in the material culture before Ramesses iii (Sperveslage and Eichmann 2012), most probably centuries after the oases were created and thrived as large centres. From the previous Middle Bronze Age import of metal weapons to be displayed in elite

pottery from the “midianite heartland”?

427

world. It was a community that both wished to partake of the Eastern Mediterranean context and reiterate and reinforce, through its pottery repertory, the borders of its uniqueness. The settlement pattern of this polity, stretched across the northern Ḥijāz and the Southern ʿArabah,146 seemed to comprise: (a) large oases/urban147 centres (e.g. Qurayyah) and (b) smaller sites with specific functions (e.g., exploiting mining resources and/or with ceremonial functions, such as Timnaʿ), the latter possibly not settlements born out of organic growth148 and (c) nonmining/metallurgical sites, such as Tell Kheleifeh. Was the latter’s function controlling sea and land trade routes on behalf of the polity to its south-west, as has been surmised before149 or rather to its south-east? The similarity in the pottery record—even if made so far of only a dozen sherds—illustrates a link to the northern Ḥijāzī polities rather than the Egyptian one. This geographically extended formation need not have been structured based on comprehensive territorial control, and might have relied on high mobility of at least part of its members in a more multi-faceted, multi-layered political structure than we are used to think of for territorial or even ethnic states.150 Until field research will have produced a critical mass of data, this remains a hypothesis. Was this the ‘Midianite Heartland’? Our answer cannot be a positive one, as the term Midian seems both anachronistic and inadequate151 for describ-

146

147 148

149 150 151

‘warrior’ graves to Late Bronze Age visible echoes in the pottery assemblages (Luciani forthcoming a). Connectivity has already been discussed in Luciani 2016a: 22 and footnote 16. Avner (2008: 1704) points out that the Eilat region, notwithstanding its hyper-arid and very varied environmental conditions, was continuously settled as opposed to alternatingly in neighbouring Negev Highlands and Sinai surely due to its copper resources but also the network of international roads that crossed it, making it an economically and geopolitically strategic area. See Parr 1988 and 1996 on this theme. Yagel et al. 2016. But if Site 2 at Timnaʿ and even the sanctuary at Site 200 significantly predate Egyptian presence, may they have grown naturally as part of the Northern Ḥijāzī polity and have been taken over by Egypt under the Ramessides afterwards? Singer-Avitz 2008, and Munford 2015: 113 and footnote 61. Luciani 2016a. Biblical references more than referring to a Late Bronze Age context mostly appear to describe a mature Iron Age world, with fortified cities ‘east of Canaan’, kings, camel riding tribes and iron implements, etc. (Eissfeldt 1968, Albright 1970, Liverani 2005 and Chan in press). Whether the exact date of narratives mentioning the Midianites, such as the e.g., story of Joseph, are surely ‘post-exilic’ (Liverani 2005: 268), or are intentionally using the term Midianites to really talk about Ishmaelites (Revell 2001), whether other

428

luciani

ing the second millennium bce material record. But even when we do have contemporaneous written records,152 we should be wary of re-introducing old paradigms and the ‘pots = people’ equation, that has been proven fallacious before. So, how should we weigh the value of pottery alone—qpw—in these functional sites? Was it, as has been claimed for Timnaʿ “the special pottery favoured by a sizeable group of ‘foreigners’”?153 Contrary to the Timnaʿ sanctuary Site 200,154 it is not known what part of the entire Tell Kheleifeh Late Bronze Age assemblage our qpw sherds represented. However, I am not sure an ethnic reading of pottery is the right interpretation, not even for the different types of assemblages found in Timnaʿ. Differences might just as well have been functional: tableware shapes (jugs and bowls) were painted (= qpw), and other functions (food storage, cooking, food transport, ritual, etc.) were performed through specific assemblages (wheel-made simple ware, so-called Negevite ware, Egyptian faience, etc.). I would, therefore, propose that Timnaʿ and Tell Kheleifeh shared tableware and feasting (?) practices that were current in Qurayyah and underscored their cultural identity with the northern Ḥijāz thereby underlining both otherness and belonging to other assemblages of the area (e.g., Bichrome Painted Ware, Cypriot or Chocolate-on-White). New research in Qurayyah and on old material from Tell Kheleifeh push us to take a fresh look on north-west Arabia as a specific southern extension of the Levant, at once distinct and yet partaking of its material cultural milieu.

152

153 154

mentions of Midian are to be dated to the 8th/7th or to the 5th century bce does not change their substantial posteriority to the qpw material. A reanalysis of ‘Midian’ in an Iron Age chronological frame could possibly concur in better defining the essence of Edom, though not necessarily its geographical borders (see above footnote 105) or material culture. Coeval Egyptian sources, potentially dealing with this area/people, must be handled with great care, because of the typology of the records (celebratory inscription) and for their being generic and external to the region proper we are dealing with (e.g. the Shasu, see Sperveslage 2013 with references). Parr 1982: 129. Rothenberg 1988: 92: qpw represents 25 % of the entire assemblage in Site 200, not broken down in different strata. Singer-Avitz (2004: 1281) has looked at qpw’s presence in the different layers at this Timnaʿ site and determined that the majority (37) of fragments “cannot be attributed to any specific strata”. This is an additional indication that the layers were severely disturbed. Only 15 could be attributed either to Stratum iii (lba) or ii (early ia). However, now that we are also conscious of how the Site 200 stratigraphy needs to be revised (Avner 2014 and above footnote 92) pinpointing the exact percentage of qpw as opposed to other wares in each assemblage/stratum remains a challenge.

pottery from the “midianite heartland”?

429

But we could formulate it also the other way around: the presence of at least a dozen sherds of the qpw ceramic class in a site where the pertinent strata have not even been touched by excavation may be interpreted as hint that the Late Bronze Age settlement on Tell Kheleifeh must have been consistent. Definitive proof of this may be reached only through new excavation. And unravelling what the function of the site might have been on the basis of the pottery evidence alone is challenging: its role as relay station in the copper trade still remains conjectural. Be that as it may, the presence of Tell Kheleifeh at the core of an area that fully shared the decorated table wares produced in Qurayyah may suggest we might need to see the Southern ʿArabah as part of the material world of north Arabian oases. Present-day researchers who have been reassessing older scholarship155 have rightly made it a point of underlying how re-appraisals are not meant to be outright criticism of the work of pioneers of field research. Much more it is part and parcel of the dynamics of a evolving discipline, methodological development that is destined to be criticised with time, just as our current re-evaluation of the old evidence will be subject to improvement in the future. I fully share this attitude and I hope that the words that precede can stand to prove it. But I would like to take the lessons we learn from revising past contributions one step further by reflecting that it is only by taking the time for rigorous scrutiny of the evidence—archaeological and historical each in its own terms—that we can we avoid producing hasty historical syntheses that concur in blurring rather than clarifying our reconstruction of the past.

Addendum After this article was already completed, renewed, systematic inspections of the storerooms at the Harvard Semitic Museum revealed three additional sherds, which, on the basis of the photos sent to me, I would feel confident in interpreting as qpw (see below fig. 4 and Table 5). The find spots of the additional sherds, as recorded in the Tell Kheleifeh field register, is given in fig. 5. These additional sherds only strengthen the argument for the closeness of the pottery repertories from Tell Kheleifeh to those from the northern Ḥijāz and for a significant, not-yet-investigated Late Bronze Age layer at this site. A hands-on examination has now taken place and allows to confirm this identification. # 12

155

Pratico 1993a, Avner 2014, Erikson-Gini 2014, Levy et al. 2014b, Ben-Yosef et al. 2012, Yagel et al. 2016, etc.

430

luciani

table 5

Find-spot of qpw sherds # 12–14 according to the hand-written Tell Kheleifeh registries in the Harvard Semitic Museum

Reg.

Date

Entry in Tell Kheleifeh registry

# 12 8016

1940, p. 22

Handle with two lines of paint. Top of Max l: 8.7cm; w: mound between Rooms 50 and 80 2.4cm; max w: 3cm

# 13 11,047

1940, p. 47

Painted sherd. [Trench] j 16

Max h: 5.1cm; max w. 3.1cm

# 14 11,094

1940, p. 50

Painted sherd. Square “i” 15 at about 35cm below the surface. April 22

Max h: 7.7cm; max w. 7.5cm

figure 4.12–14

Dimensions

Additional Qurayyah Painted Ware pottery sherds from Tell Kheleifeh in the Harvard Semitic Museum (not to scale) photos: j. greene. courtesy of the harvard semitic museum

is an oval-section handle with monochrome painted stripes; # 13 is a flat base (of jar?) with bichrome painted stripes; # 14 is a body sherd with bichrome painted butterfly motif.

References Albright, W.F. 1970. “Midianite Donkey Caravans.” In Translating & Understanding the Old Testament. Essays in Honor of Herbert Gordon May, 197–205, edited by H.T. Frank and W.L. Reed, 197–205. Nashville-New York: Abingdon Press. Avner, U. 2008. “Eilat Region.” In The New Encyclopedia of Archaeological Excavations in the Holy Land, Supplementary Volume, edited by E. Stern, 1704–1711. Jerusalem: Israel Exploration Society; Washington: Biblical Archaeology Society. Avner, U. 2014. “Egyptian Timna—Reconsidered.” In Tebes 2014, 103–162.

pottery from the “midianite heartland”?

figure 5

Modified ( from Pratico 1993a: pl. 2) general plan of Tell Keleifeh showing the localization of qpw sherds (yellow: single fragment; red: two fragments). For description of the find-spot see above Tables 4 and 5.

431

432

luciani

Bawden, G. 1983. “Painted Pottery of Tayma and Problems of Cultural Chronology in Northwest Arabia.” In Sawyer and Clines 1983, 37–52. Ben-Yosef, E., R. Shaar, L. Tauxe, and H. Ron 2012. “A New Chronological Framework for Iron Age Copper Production at Timna (Israel).” The Bulletin of the American Schools of Oriental Research 367: 31–71. Bernick-Greenberg, H. 2007. “The Ceramic Assemblages and the Wheel-made Pottery Technology.” In Excavations at Kadesh Barnea (Tell el-Qudeirat) 1976–1982, edited by R. Cohen and H. Bernick-Greenberg, 131–185. (Israel Antiquities Authority Reports 34). Jerusalem: Israel Antiquities Authority. Bienkowski, P. 2002. Busayra. Excavations by Christal-M. Bennett 1971–1980. (British Academy Monographs in Archaeology 13). Oxford: Oxford University Press and British Academy. Bimson, J.J., and J.M. Tebes. 2009 [2010]. “Timna Revisited: Egyptian Chronology and the Copper Mines of the Southern Arabah.” Antiquo Oriente 7: 75–118. Chan, A. In press. “A Historical Perspective on the ‘Midianite’ Question of Qurayyah Painted Ware.” In Tayma i—Studies on Palaeoenvironment, Archaeology and History, edited by A. Hausleiter, and R. Eichmann. Dayton, J.E. 1972. “The Midianite Pottery.” In Proceedings of the Seminar for Arabian Studies 5: 25–37. Didier, A. 2013. La production céramique du Makran (Pakistan) à l’âge du Bronze ancien. Contribution à l’étude des peuplements des régions indo-iraniennes. (Mémoires des missions archéologiques françaises en Asie centrale et en Asie moyenne 14). Paris: De Boccard. Di Vito, R.A. 1993. “The Tell Kheleifeh Inscriptions.” In Pratico 1993a: 51–63. Eissfeldt, O. (with contributions by W.F. Albright). 1968. “Protektorat der Midianiter über ihre Nachbarn im letzten Viertel des 2. Jahrtausends v. Chr.” Journal of Biblical Literature 87: 383–393. Erickson-Gini, T. 2014. “Timna Site 2 Revisited.” In Tebes 2014: 47–83. Finkelstein, I. 2005. “Khirbet en-Nahas, Edom And Biblical History.” Tel Aviv 32: 119– 125. Finkelstein, I. 2010. “A Great United Monarchy? Archaeological and Historical Perspectives.” In One God-One Cult-One Nation: Archaeological and Biblical Perspectives, edited by R.G. Kratz and H. Spieckermann, 3–28. (Beiheft zur Zeitschrift für die alttestamentliche Wissenschaft 405). Berlin: De Gruyter. Finkelstein, I. 2014. “The Archaeology of Tell el-Kheleifeh And the History of Eziongeber/Elath.” Semitica 56: 105–136. Frank, F. 1934. “Aus der ʿAraba. Reiseberichte, i.” Zeitschrift des Deutschen PalästinaVereins 57 (3/4): 191–280. Glueck, N. 1935. Explorations in Eastern Palestine ii. (Annual of the American Schools of Oriental Research 15). New Haven: American Schools of Oriental Research.

pottery from the “midianite heartland”?

433

Glueck, N. 1938. “The Topography and History of Ezion-geber and Elat.” Bulletin of the American Schools of Oriental Research 72: 2–13. Glueck, N. 1939a. Exploration in Eastern Palestine iii (Annual of the American Schools of Oriental Research 18/19). New Haven: American Schools of Oriental Research. Glueck, N. 1939b. “The Second Season of Excavation at Tell el-Kheleifeh (Ezion-geber: Elat).” Bulletin of the American Schools of Oriental Research 75: 1, 8–22. Glueck, N. 1940. “The Third Season of Excavation at Tell el-Kheleifeh.” Bulletin of the American Schools of Oriental Research 79: 2–18. Glueck, N. 1965. “Ezion-geber.” Biblical Archaeologist 28 (3): 69–87. Glueck, N. 1967. “Some Edomite Pottery from Tell el-Kheleifeh.”Bulletin of the American Schools of Oriental Research 188: 8–38. Glueck, N. 1993. “Kheleifeh, Tell el-. Identification; Excavations.” In Stern 1993: 867–869. Glueck, N., and W.F. Albright. 1938. “The First Campaign at Tell el-Kheleifeh (Eziongeber).” Bulletin of the American Schools of Oriental Research 71: 3–18. Hankey, V. 1995. “A Late Bronze Age Temple at Amman Airport: Small Finds and Pottery Discovered in 1955.” In Trade, Contact, and the Movement of Peoples in the Eastern Mediterranean. Studies in Honour of J. Basil Hennessy, edited by S. Bourke and J.-P. Descoeudres, 169–185 (Mediterranean Archaeology Supplement 3). Sydney: meditarch. Hausleiter, A. 2014 (with a contribution by Małgorzata Daszkiewicz). “Pottery Groups of the Late 2nd / Early 1st Millennia bc in Northwest Arabia and New Evidence from the Excavations at Tayma.” In Luciani and Hausleiter 2014: 399–434. Hausleiter, A. 2015. “Tayma, Saudi-Arabien Rettungsgrabungen im Gräberfeld von alNasim. Die Arbeiten des Jahres 2014.” e-Forschungsberichte 2015 des Deutschen Archäologischen Instituts 2015 (2): 74–76. Hausleiter, A. In preparation. “Qurayyah Survey. 2008–2009: Summary Report.” Atlal. Hausleiter, A., and A. Zur. 2016 “Tayma in the Bronze Age (c. 2,000bce): Settlement and Funerary Landscapes.” In Luciani 2016b: 135–173. Ingraham, M.L., T.D. Johnson, B. Rihani, and I. Shatla. 1981. “Saudi Arabian Comprehensive Archaeological Survey Program: Preliminary Report on a Reconnaissance Survey of the Northwestern Province (with a note on a brief survey of the northern province).” Atlal 5: 59–84. Intilia, A. 2016. “Qurayyah Painted Ware: A Reassessment of 40 Years of Research on its Origins, Chronology and Distribution.” In Luciani 2016b: 172–256. Kafafi, Z.A. 1983. “The Local Pottery.” In The Amman Airport Excavations, 1976, edited by L.G. Herr, 33–46. (Annual of the American Schools of Oriental Research 47–48). Philadelphia: American Schools of Oriental Research. Kafafi, Z.A. 2014. “New Insights on the Copper Mines of Wadi Faynan/Jordan.” Palestine Exploration Quarterly 146 (4): 263–280. Levy, T.E., M. Najjar, T. Higham, Y. Arbel, A. Muniz, E. Ben-Yosef, N.G. Smith, M. Beherec,

434

luciani

A. Gidding, I.W. Jones, D. Frese, C. Smitheram, and M. Robinson. 2014a. “Excavations at Khirbat en-Nahas 2002–2009. An Iron Age Copper Production Center in the Lowlands of Edom.” In Levy et al. 2014b: 89–246. Levy, T.E., M. Najjar, and E. Ben-Yosef (ed.). 2014b. New Insights into the Iron Age Archaeology of Edom, Southern Jordan—Surveys, Excavations and Research from the Edom Lowlands Regional Archaeology Project (elrap). Los Angeles: The Cotsen Institute of Archaeology Press. Liverani, M. 2005. Israel’s History and the History of Israel. London: Routledge. Luciani, M. 2013. “Anlage c4_08: Arbeitsbericht 2011–13 Area q.” In dfg-Tayma Antrag, edited by R. Eichmann, 1–13. Berlin. [in English] Luciani, M. 2014a. Report on the Archaeological Visit to Qurayyah, Tabuk Province, Saudi Arabia, February 2014. Report Submitted to the Saudi Commission for Tourism and Antiquities. Riyadh. Luciani, M. 2014b. “Qurayyah in Northwestern Arabia. Archaeological Research between Levant and Hejaz.” Poster presented at the 9th International Congress on the Archaeology of the Ancient Near East, June 9–13, 2014, University of Basel, Abstracts book: 153. Accessed September 5, 2016. https://univie.academia.edu/MartaLuciani Luciani, M. 2014c. “Key Issues in the Investigation and Study of Late Bronze Age Pottery.” In Luciani and Hausleiter 2014, 1–17. Luciani, M. 2015. “Review of Didier 2013.” Paléorient 41 (1): 223–225. Luciani, M. 2016a. “Mobility, Contacts and the Definition of Culture(s) in New Archaeological Research in Northwest Arabia.” In Luciani 2016b: 21–56. Luciani, M. 2016b. (ed.), The Archaeology of North Arabia. Oases and Landscapes. Proceedings of the International Congress held at the University of Vienna, December, 5th–8th 2013 (orea Archaeological Series 4). Vienna: Austrian Academy of Sciences. Luciani, M. Forthcoming a. “Transitions in Material Culture of the 2nd Millennium bce: The Middle Bronze to Late Bronze Age Shift Seen from Northwest Arabia.” In Material Worlds: Interdisciplinary Approaches to Contacts and Exchange in the Ancient Near East. Proceedings of the Workshop held at the Institute for the Study of the Ancient World (isaw), New York, March 7th, 2016, edited by A. Hausleiter. Luciani, M. Forthcoming b. “On the Formation of ‘Urban’ Oases in Arabia: New Evidence from the North West.” In The Archaeology of the Arabian Peninsula 2: Connecting the evidence. Paper presented at the 10th icaane Workshop, edited by M. Luciani. Vienna, April 25, 2016. (orea Archaeological Series). Vienna: Austrian Academy of Sciences. Luciani, M., and A.S. Alsaud. 2015. Joint Saudi Arabian-Austrian Archaeological Expedition at Qurayyah, 2015 Campaign, 4th and Final Weekly Report. Report presented to the Saudi Commission for Tourism and Antiquities. Dec. 12th, 2015, 1–23. Riyadh. Luciani, M., and A.S. Alsaud. 2016. “Qurayyah 2015. 1st Report on the Joint Saudi Arabian-Austrian Archaeological Project.” Atlal 27, 1–46 (in press).

pottery from the “midianite heartland”?

435

Luciani, M., and A. Hausleiter (ed.). 2014. In cooperation with C. Beuger. Recent Trends in the Study of Late Bronze Age Ceramics in Syro-Mesopotamia and Neighbouring Regions. Proceedings of the International Workshop Held in Berlin, November, 2–5, 2006. (Orient Archäologie 32). Berlin: Rahden. Luciani, M., A. Hausleiter, M. al-Njem, S. Messal, and F. Lüth. 2015. “Saudi-Austrian Joint Archaeological Project at Qurayyah. Geomagnetic Survey 2015.” Report presented to the Saudi Commission for Tourism and Antiquities, 1–9. Riyadh. Luciani, M., and L. Machel. Forthcoming. “Square q3: results of the 2011–2014 Campaigns.” In A. Hausleiter, R. Eichmann, M.H. al-Najem, and S.F. al-Said, “Report on the 2014 Excavation Campaign at Tayma.” Atlal. McKenzie, J.S. 2003. “A Nabataean Temple in the Museum Basement: Khirbet et-Tannur Re-excavated. The asor Nelson Glueck Archive at the Semitic Museum.” In Semitic Museum News. Harvard University 6 (3): 1, 4–9. McKenzie, J.S. 2013. “Introduction and Appendix 1.2: The Tell el-Kheleifeh Division and Shipping.” In The Nabataean Temple at Khirbet et-Tannur, Jordan. Volume 1— Architecture and Religion. Final Report on Nelson Glueck’s 1937 Excavation, edited by J.S. McKenzie, J.A. Greene, A.T. Reyes, C.S. Alexander, D.G. Barrett, B. Gilmour, J.F. Healey, M. O’Hea, N. Schibille, S.G. Schmid, W. Wetterstrom, and S. Whitcher Kansa, 1–38. (Annual of the American Schools of Oriental Research 67). Boston, ma: American Schools of Oriental Research in collaboration with University of Oxford: Manar al-Athar. Meshel, Z. 1975. “On the Problem of Tell el-Kheleifeh, Elath and Ezion-Geber.” Eretz Israel 12: 49–56. [in Hebrew] Muhly, J.D. 1984. “Timna and King Salomon.” Bibliotheca Orientalis 41: 275–292. Mumford, G. 2015. “The Amman Airport Structure: A Re-Assessment of its Date-Range, Function and Overall Role in the Levant.” In Walls of the Prince: Egyptian Interactions with Southwest Asia in Antiquity, Essays in Honour of John S. Holladay, Jr, edited by T.P. Harrison, E.B. Banning, and S. Klassen, 89–198. (Culture and History of the Ancient Near East 77). Leiden-Boston: Brill. Mussell, M.-L. 1999. “Tell el-Kheleifeh.” American Center of Oriental Research Newsletter 11: 5–6. Mussell, M.-L. 2000. “Tell el-Kheleifeh.” American Journal of Archaeology 104 (Section “Archaeology in Jordan”, edited by V. Egan, P.M. Bikai, and K. Zamora): 561– 588. al-Najm, M.H. (ed.) 2007. Mašrūʿ al-biʿṯah al-ʾaṯariyyah al-ʾalmāniyyah al-muštarakah li-l-tanqīb ʿan ʾāṯār taymāʾ. Ar-Riyāḍ: Deputy Ministry of Antiquities and Museums; Deutsches Archäologisches Institut. [in Arabic] Oakeshott, M.F. 1983. “The Edomite Pottery.” In Sawyer and Clines: 53–63. Parr, P.J. 1979. “Archaeological Sources for the Early History of North-West Arabia.” In Studies in the History of Arabia. Proceedings of the First International Symposium on

436

luciani

Studies on History of Arabia, 23rd–28th of April, 1977, Sponsored by the Department of History, Faculty of Arts, University of Riyadh, edited by A.M. Abdallah, S. al-Sakkar, and R.T. Mortel, 37–44. Riyadh: Department of History, Faculty of Arts, University of Riyadh. Parr, P.J. 1982. “Contacts between North West Arabia and Jordan in the Late Bronze and Iron Ages.” In Studies in the History and Archaeology of Jordan i, edited by A. Hadidi, 127–133. Amman: Department of Antiquities of Jordan. Parr, P.J. 1988. “Pottery of the Late Second Millennium b.c. from North West Arabia and its Historical Implications.” In Araby the Blest. Studies in Arabian Archaeology, edited by D.T. Potts, 73–89. (The Carsten Niebuhr Institute of Ancient Near Eastern Studies Publications 7). Copenhagen: Carsten Niebuhr Institute of Ancient Near Eastern Studies: University of Copenhagen, Museum Tusculanum Press. Parr, P.J. 1996. “Further Reflections on Late Second Millennium Settlement in North West Arabia.” In Retrieving the Past. Essays on Archaeological Research and Methodology in Honor of Gus W. van Beek, edited by J.D. Seger, 213–218. Winona Lake: Eisenbrauns. Parr, P.J., G.L. Harding, and J.E. Dayton. 1970. “Preliminary Survey in n.w. Arabia, 1968.” Bulletin of the Institute of Archaeology, University of London 8/9: 193–242. van Pelt, W.P. (ed.). 2013. Archaeology and Cultural Mixture. (Archaeological Review from Cambridge 28.1 April 2013). Petrie, W.M.F. 1892. Ten Years’ Digging in Egypt, 1881–1891. London: Religious Tract Society. Pratico, G.D. 1985. “Nelson Glueck’s 1938–1940 Excavations at Tell el-Kheleifeh. A Reappraisal.” Bulletin of the American Schools of Oriental Research 259: 1–32. Pratico, G.D. 1992. “Kheleifeh, Tell el-.” In The Anchor Bible Dictionary, edited by D.N. Freedman, vol. 4, 33–34. New York-London-Toronto-Sydney-Auckland: Doubleday. Pratico, G.D. 1993a. Nelson Glueck’s 1938–1940 Excavations at Tell el-Kheleifeh. A Reappraisal. (American Schools of Oriental Research Archaeological Reports 3). Atlanta, ga: Scholars Press. Pratico, G.D. 1993b. “Kheleifeh, Tell el-. A reappraisal of Glueck’s conclusions.” In Stern 1993: 869–870. Pratico, G.D. 1997. “Kheleifeh, Tell el-.” In The Oxford Encyclopedia of Archaeology in the Near East, edited by E.M. Meyers, vol. 3, 293–294. New York-Oxford: Oxford University Press. Doi: 9780199892280. Revell, E.J. 2001. “Midian and Ishmael in Genesis 37: Synonyms in the Joseph Story.” In The World of the Aramaeans ii. Studies in History and Archaeology in Honour of PaulEugène Dion, edited by P.M.M. Daviau, J.W. Wevers, and M. Weigel, 70–91. (Journal for the Study of the Old Testament. Supplement Series 325). Sheffield: Sheffield Academic Press.

pottery from the “midianite heartland”?

437

Rothenberg, B. 1967. Negev. Archaeology in the Negev and the Arabah. Ramat Gan: Massada Press Tel Aviv. [in Hebrew] Rothenberg, B. 1972. Timna. Valley of the Biblical Copper Mines. London: Thames and Hudson. Rothenberg, B. 1980. “Die Archäologie des Verhüttungslagers Site 30.” In Antikes Kupfer im Timna-Tal. 4000 Jahre Bergbau und Verhüttung in der Arabah (Israel), edited by H.G. Conrad and B. Rothenberg, 187–213. Bochum: Vereinigung der Freunde von Kunst und Kultur im Bergbau. Rothenberg, B. 1988. The Egyptian Mining Temple at Timna. Researches in the Arabah 1959–1984. The Arabah Project Sponsored by Stiftung Volkswagenwerk, vol. i. London: Institute for Archaeo-Metallurgical Studies; Farnborough, Hants.: Thames & Hudson. Rothenberg, B. 1998. “Who were the ‘Midianite’ Copper Miners of the Arabah. About the ‘Midianite Enigma’.” In Metallurgica Antiqua. In Honour of Hans-Gert Bachmann and Robert Maddin, edited by T. Rehren, A. Hauptmann, and J.D. Muhly, 197–212. (Der Anschnitt, Beiheft 8). Bochum: Deutsches Bergbau-Museum. Rothenberg, B., and J. Glass. 1983. “The Midianite Pottery.” In Sawyer and Clines: 65– 124. Sapir-Hen, L., and E. Ben-Yosef. 2014. “The Socioeconomic Status of Iron Age Metalworkers: Animal Economy in the ‘Slaves’ Hill’, Timna, Israel.” Antiquity 88: 775–790. Sawyer, J.F.A., and D.J.A. Clines (ed.). 1983. Midian, Moab and Edom. The History and Archaeology of Late Bronze and Iron Age Jordan and North-West Arabia, 65–124 (Journal of the Study of the Old Testament, Supplement Series 24). Sheffield: Journal of the Study of the Old Testament Press. al-Shehry, M.M. 2014. Faẖẖār dādān ẖilāl al-mawsamayn al-ẖāmis 1429h wa-l-sādis 1430h. Dirāsah taḥlīliyyah. (Silsilat dirāsāt ʾaṯariyyah muḥakkamah 27). Ar-Riyāḍ: alhayʾah al-ʿāmmah li-l-siyāḥah wa-l-ʾāṯār [Saudi Commission for Tourism and Antiquities]. [in Arabic] Sherratt, S. 2013. “The Philistines and Other ‘Sea Peoples’ in Text and Archaeology.” In The Philistines and other “Sea Peoples” in Text and Archaeology, edited by A.E. Killebrew and G. Lehmann, 619–644. Atlanta, ga: Society of Biblical Literature. Singer-Avitz, L. 2004. “Section f: The Qurayyah Painted Ware.” In The renewed archaeological Excavations at Lachish (1973–1994), edited by D. Ussishkin, 1280–1287. Tel Aviv: Emery and Claire Yass Publications in Archaeology. Singer-Avitz, L. 2008. “The Earliest Settlement at Kadesh Barnea.” Tel Aviv 35 (1): 73–81. Singer-Avitz, L. 2014. “The Date of the Qurayyah Painted Ware in the Southern Levant.” Antiquo Oriente 12: 123–148. Smith, N.G., and T.E. Levy. 2014. “Iron Age Ceramics from Edom. New Typology.” In Levy et al. 2014b: 297–460. Sperveslage, G. 2013. “Ägyptische Einflüsse auf der Arabischen Halbinsel in vorislamis-

438

luciani

cher Zeit am Beispiel der Oase von Tayma.” Zeitschrift für Orient-Archäologie 6: 234– 252. Sperveslage, G., and R. Eichmann. 2012. “Egyptian Cultural Impact on North-West Arabia in the Second and First Millennia bc.” Proceedings of the Seminar for Arabian Studies 42: 371–383. Stern, E. (ed.). 1993. The New Encyclopedia of Archaeological Excavations in the Holy Land. Jerusalem: Israel Exploration Society & Carta; New York: Simon & Schuster. Stockhammer, P.W. 2013. “From Hybridity to Entanglement, from Essentialism to Practice.” In van Pelt 2013, 11–28. Tebes, J.M. (ed.). 2014. Unearthing the Wilderness. Studies on the History and Archaeology of the Negev and Edom in the Iron Age. (Ancient Near Eastern Studies Supplement 45). Leuven-Paris-Walpole: Peeters. Wright, G.E. 1959. “Is Glueck’s Aim to Prove that the Bible is True?”Biblical Archaeologist 22: 101–108. Wright, G.E. 1970. “The Phenomenon of American Archaeology in the Near East.” In Essays in Honor of Nelson Glueck. Near Eastern Archaeology in the Twentieth Century, edited by J.A. Sanders, 3–40. Garden City, ny: Doubleday. Yagel, O.A., E. Ben-Yosef, and P.T. Craddock. 2016. “Late Bronze Age Copper Production in Timna: New Evidence from Site 3.” Levant 48: 1–19.

chapter 21

A Caravan Merchant Family of ‘Antioch on the Chrysorhoas’. A Glimpse of Hellenistic Gerasa as a Caravanserai Ina Kehrberg(-Ostrasz)

I have been told it is the year of the camel—how fitting indeed, if less so for this very modest contribution than considering the multitude of images of camels Michael has recorded on his surveys of ancient inscriptions and rock carvings in the deserts of the Ḥawrān.

∵ Excavations of Gerasa’s city wall foundations from 2000 to 2002 brought to light compelling evidence that challenged the previous 3rd–4th century ad date of the city wall. Irrefutable quantitative archaeological data from foundation contexts affirmed that constructions had begun instead in the early second century ad. Thus the ‘Jerash City Walls Project’ with the aim to settle the argument of the first city wall date by examining the wall foundations and their archaeological contexts at south, north, east and west strategic points of each sector of the city wall was successfully concluded. To the excavators’ surprise, the major discovery in the 2001 season was not a wall foundation trench.1 Beneath northern wall tumble was a layer of dirt and Roman pottery litter accumulated near and against the remaining standing lower courses of the city wall which had been built directly on levelled grounds of rocky outcrops. Excavating further laid bare the extent of the rocky terrain and revealed a set of hypogean tombs, a secret the builders of the Roman city wall had not known. One of the passages led to a sealed rock-cut chamber protruding from underneath the city wall and the modern road built on top of the wall remains. The excavation of the dromos, still filled with its original dirt fill, exposed a neatly carved closed doorway to the burial chamber, obviously

1 Kehrberg 2006; idem 2004; Kehrberg and Manley 2002a to 2002d.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

440

kehrberg(-ostrasz)

of an earlier history of Gerasa. The doorway was blocked tightly with roughly hewn large rocks wedged and held in place by an earth-clay binder, a commonly shared feature in the Mediterranean Hellenistic world (see the references cited in n. 1). The first glimpse into the undisturbed burial chamber showed that it had been tailored for a single burial: osteological examination revealed the prepuberty age of a child of about 8–10 years of age and thus the gender could not be determined. The ceramic objects and other accoutrements of the burial as well as the pottery sherds in the dromos fill confirmed that the interment took place in the Late Hellenistic period, to be precise in the second half of the second century bc, corroborated by a coin of Demetrius i found at the feet of the deceased or curved end of the chamber. Most grave goods—especially a pottery rhyton, lagynos and bull model, as well as the gold-leaf wreath, a game set of glass counters with glass astragals, and an iron strigil convey the intended message that the child belonged to a wealthy family and the aspiring middle class of the Late Hellenistic town echoing the Mediterranean world. Excepting the pottery rhyton which finds its contemporary twin in Amman in a layer beneath the Roman Hercules Temple, the pottery models or zoomorphic vessels introduced in preliminary publications (see n. 1) are unique finds unparalleled in Jordan and indeed beyond its modern borders. Yet, these very vases—and the lagynos—provide the link which proves Gerasa’s cultural ties under Seleucid rule with other regions still sharing cultural Ptolemaic influence.2 In addition, the unique three camel models or toys and toy clepsydra (fig. 1 and 2) allow some speculative insight into the young owner’s family. They are not specifically made burial offerings or ritual vessels as part of the lagynophoria like the bull model adorned with a tenia for sacrifice, the ritualistic rhyton, and lagynos. Could it be that these toys imitated life, telling us that the deceased was the child of a well-to-do merchant who owned a caravan trading in oil and wine? The set of glass astragals and counters are also toys of the deceased child and have been used in his or her life-time as the strigil, the camels and clepsydra—similar to animal models or toys in contemporary children’s tombs from the Hellenistic cemetery of Ptolemaic-Punic Carthage discovered underneath the Roman Odeon (see n. 2). The clepsydra on figure 2.2 began its Egyptian origin as a water clock and was later adapted for retrieving water. What is significant is the association of our clepsydra with the camel models: here we have models carrying amphorae and representing caravans trading in wine from Rhodes as identified by the

2 Kehrberg 2006; Kehrberg and Manley 2002c.

a caravan merchant family

figure 1

Fig. 1:1a–d, 1:2, 1:3: toy camels objects nos. 3, 10, and 14, jcw01/109: season 2001, n-w wall trench 109, pre-city wall context, intact Late Hellenistic hypogean tomb (cf. Kehrberg 2004; 2006; Kehrberg and Manley 2002a; 2002c). Fig. 1:4: photos of camels at Petra, Nov. 2005 (photos i. k–o)

441

442

figure 2

kehrberg(-ostrasz)

Fig. 2:1a–1c: two handle fragments and a base fragment from jcwp seasons 2000 and 2002: a/ jcw02/507 and c/ jcw02/508: season 2002, west wall trench 500, pre-city wall context (cf. Kehrberg & Manley 2003); b/ jcw00/27: season 2000, s-w wall trench 2000, pre-city wall context (cf. Kehrberg and Manley 2001). Fig. 2:2: clepsydra, object no. 6, jcw01/109: season 2001, n-w wall trench 100, pre-city wall context, intact Late Hellenistic hypogean tomb (cf. Kehrberg 2004; 2006; Kehrberg and Manley 2002a, 2002c) (photos and drawings i. k-o).

a caravan merchant family

443

amphorae strapped onto the camels. They mimic the real amphorae found at Jarash (fig. 2.1a–c) and in other cities like Umm Qays, Pella and Amman at that time.3 Like the amphorae, the pottery clepsydra is a realistic copy of a bronze vessel, faithfully rendering the metalwork and colour. An X-ray showed that the basket handle was indeed tubular and open-ended where it joins the body, copying its function to extract and release water from shallow puddles through its sieve-like bottom. Phoenician travellers have used just such utensil on their travels across arid lands. It is a personal item, carried on the belt and readily at hand should the caravan come across precious water. Similar water retrievers were found on earlier Iron Age sites in Jordan not far from Jarash. In the same historical and geographical contexts, one is also familiar with examples of zoomorphic vases—horses and mostly donkeys—carrying jars and baskets: like our Hellenistic toy models, and contemporary Punic examples from the children’s graves, they are wheel-made and the modelled applied features give them an individual character. Here is not the place to give a literary account of archaeological discoveries and historical research about ancient Jarash published over the last two centuries. Nor is there room to discuss depictions of camels at other contemporary sites famed for being camel caravan trading posts or caravanserais, especially Petra and Palmyra. Most readers will be familiar with the spectacular discoveries of over life-size reliefs of camels and their owners/drivers on the rock face along the Sīq of Petra. However weathered, the realistic renditions mirror the equally well known and much published camel caravan reliefs at Palmyra.4 The subject is not about established caravanserais, and their copious publications need no further reference in this context. What is being posited here is beginning with or going back to Rostovtzeff’s proposal published in 1932 that Gerasa was a caravanserai, which did not find much resonance among his contemporaries and subsequent research. I would like to present ‘circumstantial’ evidence supporting his claim. Much archaeological material evidence has since been unearthed and published which has confirmed Josephus’ description of Gerasa as a thriving town of the new Late Hellenistic middle class: tradesmen, agricultural producers and other merchants and these camel toys belong here; they represent camels of a caravan owned by a merchant from Gerasa. After this somewhat lengthy albeit necessary introduction, the speculative aspect of this modest paper is the actual rendition of the camel models. It was not until I had a close look at live camels at Petra in 2005, the camels shown

3 Kehrberg 1997: fig. 7 (no. 14 stamped amphora handle). 4 Yon 2002: 359, fig. 26, camel relief.

444

kehrberg(-ostrasz)

on figure 1.4, that I realised how closely the toys resembled the animals. If you compare the potter’s renditions of the curved neck, the mouth and eyes with their incised long lashes, one is struck by the potter’s intimite knowledge and uncanny portrayal of the camel. Indeed, the toys can be seen as an exaggeration of the anatomical and physiological features which characterise the camel. This close attention to detail does not stop here. The legs, in spite of their trunklike appearance, also show unmistakable features like the knobbly knees and flat, spread out softly padded toes. In addition it is the quirky posture and movement, if I may call it such, of the camel that is captured in each model, the ‘walk’, the head posture, the lips and almost roving eyes each recall the live camels as seen on figure 1.4. This astoundingly realistic detailing of camels suggests that the potter was indeed familiar with the animal. Familiarity with such detail can only come from observing the animal at close quarters and frequently for an extended time. The certain element of caricature or almost comical rendition of the typical features which, apart from the hump, personify the camel, can also be interpreted as intimacy with a camel’s behaviour. There may be other toy models buried underneath Roman, Byzantine, and Islamic Jarash. Our tomb of the young child is only a chance find. One may safely posit that the Gerasene potter would have been commissioned by the father of the child, the merchant trader who owned a caravan as portrayed by the toys. The toy clepsydra supports this speculative claim. It is, therefore, not unreasonable to suggest that the toy camels are circumstantial evidence hinting at Gerasa having been indeed a caravanserai, as posited in 1932 by Rostofzeff. The actual, mostly fragmentary amphorae finds in contemporary contexts of Hellenistic Gerasa, of which examples are shown on figure 2.1a–c, can now be seen as corroborative evidence rather than mere vessels purchased at the market. Animal bone studies were not very advanced in earlier excavations, but it is hoped that new osteological examinations of all bones from excavations at Jarash will add substantive evidence. Camel figurines are known from a Roman potter’s workshop at Gerasa, published by Iliffe in 1946, and excavations have since brought to light more figurines from other contexts. A Byzantine potter at the hippodrome depicted a camel on a sixth century Jarash Bowl.5 In other words, camels are not an unknown feature throughout the town’s history, as can be seen in Bankes’ (member of Buckingham’s expedition) watercolour of the Jarash Oval Piazza painted in 1816. But these are different representations and the figurines, mostly small mould-made pottery plaques to be placed on a

5 Kehrberg 1989: fig. 4.22.

a caravan merchant family

445

dado, can be regarded as decor adorning households like other figurines; one might call it a ‘fashion’. Whilst the above interpretation may be regarded as somewhat fanciful, it is however undeniable that true rendition, and more so skilful caricatures of animals, and people, are usually associated with familiarity of the subject in all its aspects. The torsos of the toy camels are not significant. The wheel-made barrel of the body made firing easier and provided the basic structure for a relatively inexpensive model where the visible aspects characterising the model, and its load covering most of the body, were the most important features. Other attention to detail again paralleled by live camels are the platted tail, the reins and the straps tying the vessels, and indeed the vessels themselves, which can be dated to the same period. When considering the very existence of the large family tomb complex, discussed in cited references, with its rich accoutrements for a young child affirming the family’s wealthy, not aristocratic, status as Gerasene citizens, the subjective reading of the material evidence may gain credence.

Catalogue, jcwp01.109 (Tomb Group) Fig. 1:1a–1d jcw01.109, object no. 3: camel model (grey) / zoomorphic vase, model with four amphorae, one crater. Condition: intact, very slight surface damage (abrasion) of rock roof fall, one handle of amphora mended. Cleaned and mended by Gabriel Humbert. Measurements of intact object: max. height 17.5cm, max. length 24.8cm, max. width 17.4cm. Clay/ware: slightly overfired (or intentional grey surface effect, rather an error since it is uneven); overfired to pale orangy brown (buff colour; it would have originally fired correctly to light yellowish orange); white, black, brown inclusions; slightly gritty in parts; grey to blackish grey slip-wash splashed as on other camels, especially no. 10. Method: wheelmade and modelled, as other camels no. 10 and 14. Description: no applied straps as other camels, incisions on tail (herring-bone), head and eyes (anatomical details); bottom of crater pierced, amphorae Rhodian and crater Late Hellenistic same as models no. 10 and 14. Fig. 1.2 jcw01.109, object no. 14: camel model / zoomorphic vase, model with four amphorae, one crater. Condition: almost intact; missing: upper tip of nostril or mouth, tip of tail, one leg; slight surface damage (chipping) of roof fall (rock fragments) from hypogean tomb (due to rain fall and tremors). Cleaned and

446

kehrberg(-ostrasz)

mended by Gabriel Humbert. Measurements of complete object: max. height 19cm (uneven proportions), max. length 20cm (head to tail); max. width 20 cm (amphora to amphora body). Clay/ware: same as camel no. 10, slightly brighter orange, same inclusions; same surface slip-wash but more unevenly applied and partly worn, especially the two back amphorae (from frequent handling before they became burial gifts). Method: wheel-made barrel-shaped body, truncated legs, crater and amphorae; solid rectangular piece for neck, solid head; applied ‘bridle’. Description: no straps applied as for no. 10. but incised markings for holster around head and under mouth on neck (tapestry or woven design as seen today for camels); objects not fastened by straps; ‘chameleon eyes’ (raised circular lids, incised radiating eyelashes); bottom of crater pierced; spout/airvent under neck. Amphorae and crater types as of no. 10. Fig. 1.3 jcw01.109, object no. 10: camel model / zoomorphic vase, model with two amphorae, one crater. Condition: almost intact; missing: handles of amphorae, applied mouthpiece of camel, tip of tail, one eye, bridle applique; one foot fragmented, mended; slight surface damage (chipping) of roof fall (rock fragments) from hypogean tomb (due to rain fall and tremors). Cleaned and mended by Gabriel Humbert. Measurements of complete object: max. height 20.5 cm (uneven proportions); max. length 18.8cm (head to tail); max. width 22.5cm (amphora to amphora body). Clay/ware: pinkish yellowish orange throughout; tiny to small white, black and red inclusions; red slip-wash throughout outer surface and inside of crater applied irregularly (like a splash wash painted and splashed on with a brush, which recalls Late Hellenistic ‘spatter ware’ of Tell Anafah); slightly unevenly fired. Method: wheel-made barrel-shaped body, truncated legs, crater and amphorae; solid rectangular piece for neck, solid head; applied ‘bridle’. Description: straps for fastening pots and tail incised patterns (decor and ‘hair’); one hole in bottom of crater joined onto back of camel; spout/airvent under neck. Amphorae size: height c. 12.5 cm; crater c. 6cm, amphorae Rhodian and crater types 2nd century bc/Late Hellenistic. Fig. 2.2 jcw01.109, object no. 6: clepsydra model. Condition: intact, slightly worn on sides from handling. Measurements: max. height 13.8cm (with handle); max. diam. 9.8cm (body; 9.9 on sides with handle attachments); max. diam. 6.5 cm (ring foot base); max. height of sphere 7.7cm (without handle, with foot). Clay/ware: very pale cream pinkish buff, soapy; very finely levigated with tiny black and white inclusions; pale orangy brown to dark brown and blackish brown slip-wash, worn off in patches on opposite sides (from frequent han-

a caravan merchant family

447

dling before they became burial gifts); burnished and shaved in parts. Method: wheel-made spherical body visible at base, tubular handle with pierced opening at central arc of handle; applied ring foot, inside ring foot base pierced (sieve-like) from outside. Description: decoration of circularly applied notching covering whole sphere (except at top where ‘straps’ pattern is apparent), handle and base, even ring foot (either imitation of notching from bronze hammering, typical method of metal worker in bronze, or—less likely—showing imitation of basketry on original bronze clepsydra).

References Buckingham, J.S. 1821. Travels in Palestine, through the Countries of Bashan and Gilead, East of the River Jordan: Including a Visit to the Cities of Geraza and Gamala in the Decapolis. London: Longman, Hurst, Rees, Orme and Brown. Iliffe, J. 1945. “Imperial Art in Trans-Jordan. Figurines and Lamps from a Potter’s Store at Jerash.” Quaterly of the Department of Antiquities in Palestine 11: 1–26. Kehrberg, I. 1989. “Selected Lamps and Pottery from the Hippodrome at Jerash.” Syria 66: 85–97. Kehrberg, I. 1997. “The Pottery and other Finds.” In Brenk, C., C. Jäggi, and H.R. Meier, “New Data of the Early Christian Cathedral of Gerasa: The Third Interim Report on the Jarash Cathedral Project 1996.” Annual of the Department of Antiquities of Jordan 41: 316–321, fig. 7. Kehrberg, I. 2004. “Late Hellenistic and Early Roman Pottery of Gerasa: A Commercial Enterprise in View of International Norms.” In Studies in the History and Archaeology of Jordan 8, 189–196. Amman: Department of Antiquities. Kehrberg, I. 2006. “A Late Hellenistic Link between Jordan and Cyprus: A View from Gerasa.” In Festschrift in Honour of J. Richard Green, edited by L.A. Beaumont, C.D. Barker, and E.A. Bollen, 299–306, pl. 37–38. (Mediterranean Archaeology 17). Sydney: The University of Sydney. Kehrberg, I. 2012. “Cypriot Hellenistic Lagynoi and International Connections.” In Aphrodite’s Island. Australian Archaeologists in Cyprus. The Cypriot Collection of the Nicholson Museum, edited by C. Barker, 59–64. Sydney: The University of Sydney. Kehrberg, I., and J. Manley (Jerash City Walls Project / jcwp reports): Kehrberg, I., and J. Manley. 2001. “New Archaeological Finds for the Dating of the first Gerasa City Wall.” Annual of the Department of Antiquities of Jordan 45: 437–446. Kehrberg, I., and J. Manley. 2002a. “The 2001 Season of the jcwp: A preliminary Report.” Annual of the Department of Antiquities of Jordan 46: 197–203. Kehrberg, I., and J. Manley. 2002b. “Gerasa/Jerash. Ina Kehrberg and John Manley, cbrl, Report: jcwp01 excavations.” American Journal of Archaeology 106 (3): 443–445.

448

kehrberg(-ostrasz)

Kehrberg, I., and J. Manley. 2002c. “La protégée du rempart de Gérasa.” Le Monde de la Bible 147: 54–55. Kehrberg, I., and J. Manley. 2002d. “The Earliest Hellenistic Tomb Find in Jerash or Antioch on the Chrysorhoas. The Jerash City Walls Project: Excavations 2001.” Occident and Orient 7 (1): 7–8. Kehrberg, I., and J. Manley. 2003. “The Jarash City Walls Project (jcwp) 2001–2003: Report of Preliminary Findings of the Second Season 21st September–14th October, 2002.” Annual of the Department of Antiquities of Jordan 47: 83–86. Kraeling, C.H. 1938. Gerasa: City of the Decapolis. New Haven, ct: American Schools of Oriental Research. Rostovtzeff, M. 1932. Caravan Cities. Oxford: The Clarendon Press. Yon, J.-B. 2002. Les notables de Palmyre (Bibliothèque archéologique et historique 163). Beyrouth: Ifpo.

chapter 22

The Visit of Mālik bin Muʿāwiyah, King of Kindah and Maḏḥiǧ to the Himyarite King Šammar Yuharʿiš in Maʾrib Mohammed Maraqten

Introduction Our Sabaic inscription bears the siglum mb 2006 i-54 (fig. 2) and records an offering of a bronze statuette dedicated to the Sabaean god Almaqah at the Awām temple/Maḥram Bilqīs, near Maʾrib, Yemen (Maraqten 2015: 109–135). Item Description: Base of a statuette, now missing, inscribed on one face. The inscription, canted slightly downwards and is engraved between double, lightly-incised horizontal double-lines. The face of the stone is polished; the top surface, where the statuette would have been attached, is broken away and the first line is mostly lost. The lower left corner is broken away and some parts of the stone edges are slightly damaged. The text is intact apart from few lost letters in the first line. Although the upper part of the inscription is partly lost, its reconstruction, as indicated in the transcription, is possible with reference to line 11. This 14 lines inscription has been discovered during the ninth afsm season, 2006 in the Annex of the Awām temple and is reused in the pavement. The measurements of this inscribed block are as follows: 35 × 22cm; letter height: 1.5 ×1.7cm. Almaqah curved club symbol occurs at beginning of line one and two; the upper part is broken. Present location of the inscription is the Awām Temple and still in situ (fig. 1, 2).

mb 2006 i-54 Transcription 1. 2. 3. 4.

[mlkm / bn / mʿw]yt / mlk kdt / wmḏḥgm / hqny ʾlmqhwṯhwnbʿlʾwm / ṣlmn / ḏḏhbn / bkn / mẓʾ

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

450

maraqten

figure 1 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Location of the inscription mb 2006 i-54

wnfṣ / ls¹tlmn / byd / mr ʾhmw / s²mr / yhrʿs² / mlk s¹bʾ / wḏrydn / bn / ys¹rm yhnʿm / mlk / s¹bʾ / wḏryd n / bhgrn / mryb / wlwzʾ / ʾlmqhṯhwnbʿlʾwm / ḫm r / ʿbdhw / mlkm / bn / mʿwyt ḥẓy / wrḍw / mrʾhmw / s²m r / yhrʿs² / mlk / s¹bʾ / wḏr ydn / bʾlmqhṯhwnbʿlʾwm Translation

1. 2. 3. 4. 5. 6. 7. 8.

(Mālikum son of Muʿāw)iyat king of Kiddat and Maḏḥigum has dedicated to Almaqah Ṯahwān, Lord of Awām, the bronze statue when he marched and hurried to gain security by the performance of loyalty with his lord Šammar Yuharʿiš, king of Sabaʾ and Ḏū-Raydān, son of Yāsirum Yuhanʿim, king of Sabaʾ and Ḏū-Raydān

the visit of mālik bin muʿāwiyah, king of kindah and maḏḥiǧ

figure 2 9. 10. 11. 12. 13. 14.

The inscription of Mālikum bin Muʿāwiyat (mb 2006 i-54)

in the city of Maʾrib. And may Almaqah Ṯahwān, Lord of Awām continue to grant for his servant Mālikum son of Muʿāwiyat the favour and goodwill of their lord Šammar Yuharʿiš, king of Sabaʾ and Ḏū-Raydān. By Almaqah Ṯahwān, Lord of Awām

451

452

maraqten

Commentary The restoration of the reading of the first line as mlkm bn mʿyt is certain since the name of this king is attested in line 11. This king occurs for the first time in the inscriptions and he is not known from other sources in particular from Qaryat Al-Faw inscriptions. The text mentions the title of Mālikum son of Muʿāwiyat as mlk kdt wmḏḥgm “king of Kiddat and Maḏḥigum”. This title is known from other Sabaean inscriptions which mentions mlk9 m bn bd mlk kdt w-mḏḥgm “Mālikum bin Budd, king of Kiddat and Maḏḥigum” (cias 39.11/o 2 n° 8/8–9 = Ja 2110/8–9).— The writing of kdt (< kndt) for Kindah is the usual form, which mentioned also in other inscriptions from Maḥram Bilqīs (e.g., Ja 576+Ja 577/2, passim; Ja 665/2, passim; Ry 506/4,5 = Murayghān 1/4,5; cias 39.11/o 2 no 8 = Ja 2110; AlSekaf 1985: 156–159).—The tribe mḏḥgm, or as it is known from the Arab sources as Maḏḥiǧ is also known from several inscriptions (Ja 665/2; Sh 32; AlSekaf 1985: 169 f.). Both tribes: kdt and mḏḥgm occur together in several inscriptions (i.g., Ir 32/2; Ja 660/2; Ja 665/2; 1028/7; ʿAbadān 1/28; Maʾsal 2/28 = Ry 510/28). The verb mẓʾ means “go, proceed, march” (Sab.Dict., 89; Maraqten 2014: 154).—The verb nfṣ means “march, march off” (Sab.Dict., 93). This verb could be understood as “to rush, hurry, quick”; see Arabic nafaṣa bi-l-kalimati: ʾatā bi-hā sarīʿan “He pronounced it (i.e. the word) quickly” (az-Zabīdī, Taǧ alʿArūs, 18/186). Hitherto, this verb nfṣ is attested only in military context in the Sabaic inscriptions of the Awām temple (Ir 13/27; j 631/22).—B-hgrn / mryb: The inscription reports that the visit of Mālikum son of Muʿāwiyat took place in the city of Maʾrib. Of significance is the writing form of the city of Maʾrib as mryb, which is not to be expected in this period and usually is written as mrb. Mryb is the common writing of the name of the city of Maʾrib in the early Sabaean period, until the end of the 3rd century ce. After that, it is no longer mentioned as mryb and we have only mrb, Mārib. This name is written in Classical Arabic as Maʾrib (Müller 1989: 559–567; Maraqten 2008: 107–144). In our text the formula: l-s¹tlmn b-yd mr 6 ʾhmw is an expression of peaceful loyalty of the king Mālikum son Muʿāwiyat, which has been confirmed by the visit of this king in Maʾrib to the Himyarite king Šammar Yuharʿiš.—The infinitive s¹tlmn is built from verbal form ftʿl, which derives from the S¹lm “sue for peace” (see Sab. Dict., 126), cf. the Sabaic phrase s¹tlm 6b-ʾlhn “gain security with (b-) deity” (Kortler 1/5–6). In general the verbal form s¹tlm means “to look for peace; gain security”. Accurately, the Sabaic term yd means “hand, main; fealty, loyalty” (Sab. Dict., 167; Sima 2001: 71–79) and in our text is an expression for loyalty. The

the visit of mālik bin muʿāwiyah, king of kindah and maḏḥiǧ

453

term yd occurs in a phrase as ʾḫḏ yd-k “to take your hand (of Almaqah), i.e. loyalty” (Zayd ʿInān 11/9–10). The phrase nzʿ ydm means “withdraw allegiance” (Ja 577/8); cf. also hʿd + yd “give back, restore loyalty to” (Sab. Dict., 22). The phrase l-s¹tlmn b-yd mr 6 ʾhmw in our text can be understood as follows: “in order to perform loyalty for his lord”. Apparently, the king Mālikum son Muʿāwiyat performed a number of loyalty measures during his visit to Maʾrib among them the dedication of an offering to the god Almaqah at the Awām temple.

Qaryat al-Faw in the Recently Discovered Sabaean Inscriptions at the Awām Temple The earliest history of qryt ḏt khlm/Qaryat Dhāt Kāhilum, modern Qaryat alFaw is obscure. Qaryat al-Faw was the royal headquarters of the realm of Kindah and is located in Wādī ad-Dawāsir on the edge of the Empty Quarter on the old trade route ca. 300 km north east of Naǧrān. This trade route is preceding towards the Arabian Gulf. Qaryat Dhāt Kāhilum seems to have flourished between the 3rd century bce and the end of the 3rd century ce. Although Qaryat al-Faw was known to scholars, little was known about the history of the site until the excavation of the King Saud University, Riyadh under the direction of A. Al-Ansāry at the beginning of the seventies of the last century and continued for several years (Al-Ansari 1982; Al-Anṣārī & Ṭayrān 2008: 97–106; Al-Ansari 2010: 311–317). The archaeological discoveries at Qaryat alFaw identified this site was qryt ḏt khlm/Qaryat Dhāt Kāhilum that mentioned in the Sabaean inscriptions of the Awām temple/Maḥram Bilqīs. This name, which relates to the site, was the main local god khlm/Kāhilum. The city god Kāhilm is known from several inscriptions discovered in Qaryat al-Faw itself as well as in graffiti from the surrounding area (Robin 1996: 665–714; al-Anṣāry & Ṭayrān 2008: 97–106). The name qryt ḏt khlm seems to have been given by the Sabaeans. In the inscriptions discovered in Qaryat al-Faw itself it was called: qryt, qryt ṭlw (dam 302 f 8 /7, 10); qryt khl and the term gntn seems to have been a designation of the Oasis of Qaryat al-Faw (dam 302 f 8/6; al-Anṣāry & Ṭayrān 2008: 97–106). Dozens of inscriptions written in the South Arabian Musnad script have been discovered at Qaryat al-Faw (Al-Ansari 1982: 24 f., 2010, Robin 2001: 509– 577; Kropp 1992: 55–67; dam 3 f 12; dam 2 f 12; dam 262 f 8; dam 302 f 8; nmr 3 f 23; nmr 887). Some of these inscriptions are clearly Minaic and give us good data about the trade relation with Mesopotamia or Syria-Palestine especially with Gaza (dam 262 f 8; dam 302 f 8). However, most of the epigraphic material

454

maraqten

of Al-Faw is written in a north Arabic tongue and in Musnad script (Al-Ansari 1982; Beeston 1979: 1–6; Macdonald 2008: 467f.; Robin 2001: 509–577; Al-Jallad 2014: 1–21). The question of the language of the Qaryat al-Faw inscriptions is still depending on preliminary reports of the excavators of the site. However, the publication of the discovered inscriptions at al-Faw would help to clarify this question (Beeston 1979: 1–6; Robin 2001: 509–577; Al-Anṣārī & Ṭayrān 2008: 97–106). Sabaean inscriptions contain virtuous data relating to Qaryat al-Faw and Kindah, and give light on the relations to ancient Yemeni kingdoms, which were occasionally hostile as it is mentioned in several inscriptions from the Awām temple/Maḥram Bilqīs, near Maʾrib. Qaryat al-Faw is mentioned in the Sabaean inscriptions of the 3rd century ce from the Awām primarily as qyrtm ḏt khlm or as qrytm and qrytn (Robin 1996: 665–714; al-Anṣāry & Ṭayrān 2008: 97–106). The uncovered inscriptions at the Awām temple have been a significant source for studying the history of this city. Until the recently excavations of the afsm at the Awām temple Qaryat Dhāt Kāhilum was known from several Sabaic inscriptions discovered by the afsm in the 1950es (Ja 634 = mb 2004 i68; Ja 635 = mb 2004 i-63; Ja 641 = mb 2004 i-50; Ja 665 = mb 2004 i-60; Ja 576 / 2; see also Ja 2110/ 8–9). Two military expeditions are mentioned against Qaryat Ḍāt Kāhilum carried out by the two kings: ʾĪlšaraḥ Yaḥḍub and his brother Yaʾzul Bayn (ca. 236– 260ce; Ja 576; Ja 2110). The newly discovered Sabaean inscriptions during the recent afsm excavations (1998–2006) mention Qaryat Ḏāt Kāhilum in relation with military expeditions carried out to central Arabia by the Sabaean and Ḥimyarite rulers.— qryt ḏt khlm is mentioned several times in the recently discovered epigraphic evidence at the Awām Temple/Maḥram Bilqīs (mb 2006 i-24; mb 2005 i-43; mb 2006 i-54) by the afsm in the last nine seasons (1998–2006). Two military expeditions were carried out by the Sabaean king Šāʿirum Awtar (ca. 210–230ce) against Qaryat Ḏāt Kāhilum. The first one is recorded by the military leaders of this king made dedications to Almaqah lord of Awām from the booty that they obtained from Qaryat Ḏāt Kāhilum (mb 2005 i-43). This 14-lines inscription was discovered recently in the excavations of the afsm in the season 2005 and reused in the pavement of the Annex. The second inscription (mb 2006 i-24) contains 32 lines and was recorded by a military leader of the king Sabaean king Šāʿirum Awtar and the Qayls of the Šaʿb of Tanʿumum and Tanʿamatum. It mentions that they dedicated to Almaqah Ṯahwān lord of Awām an offering of a bronze statue which they obtained as a booty from Qaryatum Ḏāt Kāhilum.

the visit of mālik bin muʿāwiyah, king of kindah and maḏḥiǧ

455

Recent epigraphic discoveries of the afsm at the Awām give us more lights on the relation of the kingdoms of Sabaʾ as well as of the Ḥimyarite kingdom with Central and North Arabia (e.g. mb 2004 i-127; mb 2004 i-163; 2005 i-104).

The Significance of the Inscription the King Mālikum bin Muʿāwiyat Our newly discovered inscription mentions the king Mālikum bin Muʿāwiyat, hitherto unknown king of the Kingdom of Kiddat/Kindah. This kingdom was a periphery state that emerged around the South Arabian kingdoms of Maʿīn and Sabaʾ and later on the border of the Kingdom of Sabaʾ and Ḏū-Raydān possibly in the first century bce (Retsö 2003: 557–563; Robin 2012: 59–129, Bukharin 2009: 64–80). Kindah played a central role in political, military and cultural issues in the pre-Islamic era and continued in the early Islamic era (Shahid 1986: 118–120; Lecker 1994: 333–356). According to the information gained from the inscriptions, Qaryat al-Faw was the capital of the kingdom of Kindah and Qaḥṭān possibly flourished from the first century to the second half of the 3rd Century ce. One king bear the title ḥgr bn ʿmrm mlk kdt “Ḥuǧr son of ʿĀmirum, king of Kiddat” ocuurs in a graffito found in Nafūd Musammā, north east of Jibāl Kawkab, near Najrān (Gajda 1996: 65–73). The rulers of this kingdom bear in the early period the title mlk kdt w-qhṭn “King of Kinda and Qaḥṭān” (e.g. dai Barʾān 2000–2001; Robin 2012: 59–129). These are known from local discovered inscriptions at the site Qaryat al-Faw itself and from the Sabaean inscriptions discovered at the Awām temple (Beeston 1986: 120). More light has been given recently on the chronology of the kings of Kinda (Nebes 2004: 273–281; Bukharin 2009: 64–80; Robin 2012: 59–129). Later on Kindah occurred in association with Maḏhig and the kings bear the title, mlk kdt w-mḏḥgm “king of Kiddat and Maḏḥigum” as in the case of Mālikum bin Muʿāwiyat, who seems to have been setteled also at Qaryat Ḏāt Kāhilum. Apparently, towards the end of the 3rd century ce the political situation has been changed and the famous tribe Kindah dominated the area of Central Arabia and established Qaryat al-Faw as its capital and reached its high rank realm at the end of the 3rd and the beginning of the 4th centuries ce. According to the Arabic sources, the Kindah seems to have comprised the bulk of the tribes of Central Arabia. It produced a certain unity among these tribes and dominated the region from northern Naǧrān to Central Arabia. It was a nomadic tribal confederation and had a different than the organised states of ancient South Arabia (Shahid 1986: 118). Arabic sources relate genealogy of the tribe

456

maraqten

Kindah to Kahlān b. Sabaʾ (Shahid 1986: 118; Robin 2012: 63f.). Already in the 5th century ce, the power of Kindah had begun to fall down and early in the 6th century ce, the Kindite authority on Arab tribes collapsed. The last Kindah king, the famous poet Imruʾ al-Qays, became a fugitive (Mumayiz 2005: 135–151; Robin 2012: 77–80). Our new inscription records a journey of Mālikum son Muʿāwiyat king of Kiddat and Maḏḥigum to Maʾrib. This is the first historical evidence of this king. Apparently, it was the custom that official visitors of the Sabaean and the Himyarite kings used to visit the Awām temple and perform an offering to Almaqah. The reason for the dedication of Mālikum son Muʿāwiyat seems to have been a part of the ceremonies when this king came to Maʾrib in order to achieve a peace accord with Šammar Yuharʿiš. This is expressed in our inscription as “to gain security by the performance of loyalty”. The visit of Mālikum son Muʿāwiyat was possibly an expression of a political loyalty, not only for the Šammar Yuharʿiš, but also for Almaqah. Of interest that the visit of the king Mālikum son Muʿāwiyat was in Maʾrib, which states that Maʾrib at the beginning of the Himyarite domination continue to be the political centre of the Himyarite kingdom or at least at the time of Šammar Yuharʿiš (ca. 287–312ce). However, in this period Maʾrib as a capital of the kingdom of Sabaʾ lost gradually its role as a cultural and religious centre and Ẓafār, the capital of Himyarite kingdom took its position (Müller 1989: 559–567). The visit of Mālikum son Muʿāwiyat king of Kiddat and Maḏḥigum to Maʾrib took place before the invasion of Hadramawt by Šammar Yuharʿiš since our inscription does not mention the long title of this king which known from other inscriptions as mlk s¹bʾ w-ḏrydn w-ḥḍrmwt w-ymnt “King of Sabaʾ and ḎūRaydān and Haḍramawt and Yamanat” (e.g. Ja 660/18–19). Accurately, the reign of Mālikum son Muʿāwiyat should have been fixed in the early reign period of Šammar Yuharʿiš, i.e. before bearing the long title and occupying Haḍramawt. This suggests that Mālikum son Muʿāwiyat might have governed between 290– 295ce (see also Robin 2012: 59, 65). Our text is significant for establishing the relations between kingdom of Sabaʾ and Ḏū-Raydān and the Arab tribes of Central Arabia. In this case, this visit seems to have had the purpose to establish the alliance of Kiddat and Maḏḥiǧ with the Himyarite kingdom and was a preparation for occupation of Haḍramawt by the Himyarites of which Kiddat military contingents took part (Ir 13). Our new inscription proves that the kingdom of Kiddat was integrated within the kingdom of Šammar Yuharʿiš. However, during the reign of the Ḥimyarite king Šammar Yuharʿiš (ca. 287–312 ce), the kingdom of Kindah was integrated within the Ḥimyarite empire and contingents of Kindah are mentioned in inscriptions from this period (Sh 32).

the visit of mālik bin muʿāwiyah, king of kindah and maḏḥiǧ

457

Abbreviations and Sigla ʿAbadān 1 afsm cias

Robin & Gajda 1994; Zwettler 2000: 230–238; Müller 2010: 50–54. American Foundation for the Study of Man. Corpus des inscriptions et antiquités sud-arabes, (Académie des Inscriptions et Belles-Lettres), Louvain, Peeters, 1977–1986. csai Corpus of South Arabian Inscriptions, http://dasi.humnet.unipi.it/index.php?id=42&prjId=1&corId=0& colId=0 (accessed September 14, 2017) dai Barʾān 2000–2001 Nebes 2004. dam Department of Archaeology Museum, King Saud University, Riyadh. dam 3 f 12 Al-Ghabban et al. 2010: no. 127. dam 2 f 12 Al-Ghabban et al. 2010: no. 128. dam 262 f 8 Al-Ghabban et al. 2010: no. 135. dam 302 f 8 Al-Ghabban et al. 2010: no. 136. Ir Inscriptions published in al-Iryānī 1990. Ja Inscriptions published in Jamme 1962. Ja 2110 Doe & Jamme 1968. Kortler 1/5–6 Müller 1978: 115–117. Maʾsal 2 = Ry 510; Ryckmans G. 1953. mb Registration siglum of inscriptions discovered by the afsm excavations at Maḥram Bilqīs (1998–2006). mb 2006 i-24; mb 2005 i-43; mb 2006 i-24; mb 2004 i-127; mb 2004 i-163; 2005 i-104 The publication of these inscriptions is in preparation by the author of this paper. Murayghān 1 = Ry 506; Ryckmans G. 1955; Müller 2010: 62. nmr National Museum, Riyadh nmr 3 f 23 Al-Ghabban et al. 2010: no. 129 nmr 887 Al-Ghabban et al. 2010: no. 130 Sab. Dict. Beeston et al. 1982 Sh 32 Inscriptions published in Sharafaddīn 2004; This inscription has been rediscovered by the afsm in 2004. Zayd ʿInān 11 Al-Iryānī 2005.

For further Abbreviations, sigla, and bibliography of cited inscriptions see Kitchen 2000 and csai.

458

maraqten

References Al-Ansari, A. 1982. Qaryat al-Fau: A Portrait of Pre-Islamic Civilisation in Saudi Arabia. London: Croom Helm. Al-Ansari, A. 2010. Qaryat al-Faw. Pages 311–317 in Al-Ghabban et al. 2010. Al-Anṣārī, A. & Ṭayrān, S. 2008. Qaryat al-Faw “madīnat al-maʿābid”. Pages 97–106 in in Al-Ansari, A.R., Muaikel, K.I. & Alsharekh, A.M., eds., The City in the Arab World in light of archaeological discoveries: Evolution and development, Riyadh: Adumatu, Abdul Rahman Al-Sudairy Foundation. Beeston, A.F.L. 1979. “Nemara and Faw”, Bulletin of the School of Oriental and African Studies 42: 1–6. Beeston, A.F.L. 1986. The relation of Kinda with Sabaʾ and Himyar, appendix to Kinda. Page 120 in The Encyclopaedia of Islam, 2nd Edition. Leiden: Brill. Beeston, A.F.L., Ghul, M.A., Müller, W.W. & Ryckmans, J. 1982. Sabaic Dictionary. Louvain-la-Neuve and Beirut: Editions Peeters and Librairie du Liban. Bukharin, M.D. 2009. Towards the earliest history of Kinda, Arabian Archaeology and Epigraphy 20: 64–80. Doe, D.B. & Jamme, J. 1968. New Sabaean Inscriptions from South Arabia, Journal of the Royal Asiatic Society: 1–28. Gajda, I. 1996. Ḥuǧr b. ʿAmr roi de Kinda et l’établissement de la domination ḥimyarite en Arabie centrale. Proceedings of the Seminar for Arabian Studies 26: 65–73. Al-Ghabban, A. et al., eds. 2010. Roads of Arabia: Archaeology and history of the kingdom of Saudi Arabia; [exhibition “Roads of Arabia: archaeology and history of the kingdom of Saudi Arabia”, Musée du Louvre, Paris 14 july–27 September, 2010]. Paris: Louvre Éditions; Paris: Somogy Art Publishers. al-Iryānī, M. 1990. Fī tārīḫ al-yaman. Nuqūš musnadiyyah wa- taʿlīqāt, 2. Edition. Ṣanʿāʾ: Markaz ad-Dirāsāt wal-Buḥūṯ al-Yamanī. Al-Iryānī, M. 2005. “Unshūda min Maḥram Bilqīs”, al-Thawabit 41, July, Ṣanʿāʾ, 64–106. Al-Jallad, A. 2014. On the genetic background of the Rbbl bn Hfʿm grave inscription at Qaryat al-Fāw, Bulletin of the School of Oriental and African Studies 77/3: 1–21. Jamme, A. 1962. Sabaean Inscriptions from Maḥram Bilqīs (Mârib), (Publications of the American Foundation for the Study of Man 3). Baltimore: The Johns Hopkins University Press. Kitchen, K.A. 2000. Documentation for Ancient Arabia. Part ii, Bibliographical Catalogue of Texts, (The world of Ancient Arabia Series). Liverpool: Liverpool University Press. Kropp, M. 1992. The Inscription Ghoneim AfO 27. 1980. A fortunate error, Proceedings of the Seminar for Arabian Studies 22: 55–67. Macdonald, M.C.A. 2008. “Old Arabic (Epigraphic)”. Pages 464–477 in Versteeg, K., ed., Encyclopedia of Arabic language and linguistics iii. Leiden: Brill.

the visit of mālik bin muʿāwiyah, king of kindah and maḏḥiǧ

459

Maraqten, M. [Al-Maraqṭan] 2008. Al-ʿĀṣima as-sabaʾiyya Maʾrib: Dirāsa fī tārīḫihā wabunyatihā al-ʾidāriyya wal-ʾiǧtimāʿiyya fī ḍawʾ an-nuqūš as-sabaʾiyyah. Pages 107–144 in Al-Ansari, A.R., Muaikel, K.I. & Alsharekh, A.M., eds., The City in the Arab World in light of archaeological discoveries: Evolution and development, Riyadh: Adumatu, Abdul Rahman Al-Sudairy Foundation. Maraqten, M. 2014. Altsüdarabische Texte auf Holzstäbchen: Epigraphische und kulturhistorische Untersuchungen. (Beiruter Texte und Studien 103). Beirut: Orient Institut; Würzburg: Ergon. Maraqten, M. 2015. Sacred spaces in ancient Yemen—The Awām Temple, Maʾrib: A case study. Pages 109–135 in Arbach, M. & Schiettecatte, J., (eds.), Proceedings of the 17e Rencontres Sabéennes: Pre-Islamic South Arabia and its Neighbours: New Developments of research (British Foundation for the Study of Arabia Monographs No. 16, bar International Series 2740). Oxford: Archaeopress. Müller, W.W. 1978. Sabaische Felsinschriften von der jemenitischen Grenze zur Rubʿ alḪālī. Pages 113–136 in Neue Ephemeris für semitische Epigraphik, ed. by R. Degen, R., Röllig, W. & Müller, W.W. Wiesbaden: Otto Harrossowitz. Müller, W.W. 1989. Mārib. Pages 559–567 in The Encyclopaedia of Islam. 2nd Edition, vi. Leiden: Brill. Müller, W.W. 2010. Sabäische Inschriften nach Ären datiert. Bibliographie, Texte und Glossar (Akademie der Wissenschaften und der Literatur—Mainz, Veröffentlichungen der Orientalischen Kommission, 53). Wiesbaden: Otto Harrassowitz. Mumayiz, I. 2005. Imruʾ al-Qays and Byzantium, Journal of Arabic Literature 36/2: 135– 151. Nebes, N. 2004. Ein Feldzugsbericht des Šāʿirum ʾAwtar in einer neuen Widmungsinschrift aus dem Barʾān-Tempel. Pages 281–283 in Sedov, A. (ed.), Scripta Yemenica. Issledovanija po Yuzhnoj Aravii. Sbornik nauchnyh statej v chest‘ 60-letija M.B. Piotrovskogo (Maison d’éditions « Vostochnaja literatura », Académie des Sciences de Russie). Moscow: Vostochna. Lecker, M. 1994. Kindah on the Eve of Islam and during the Ridda, Journal of the Royal Asiatic Society, Third séries, vol. 4: 333–356. Retsö, J. 2003. The Arabs in Antiquity. Their history from the Assyrians to the Umayyads. London: Routledge Curzon. Robin, Chr. 1996. Le Royaume Ḥujride, dit “royaume de Kinda”, entre Ḥimyar et Byzance, Comptes Rendus de l’Académie des Inscriptions et Belles Lettres, 1996: 665–714. Robin, Chr. 2001. Les inscriptions de l’Arabie Antique et les études Arabes, Arabica 48: 509–577. Robin, Chr. 2012. Les rois de Kinda. Pages 59–129 in Al-Helabi, A., Letsios, D., AlMoraekhi, M. & Al-Abduljabbar, A., (eds.), Arabia, Greece and Byzantium: Cultural Contacts in Ancient and Medieval Times, Part ii. Proceedings of the International Symposium on the Historical Relations between Arabia the Greek and Byzantine

460

maraqten

World (5th century bc–10th century ad) Riyadh, 6–10 December, 2010. Riyadh 2012 / ah1433: Robin, Chr. & Gajda, I. 1994. L’inscription du wādī ʿAbadān, Raydān 6: 113–137, 193–204. Ryckmans, G. 1953. Inscription sud-arabes—Dixième série, Le Muséon 66: 267–317. Ryckmans, G. 1955. Inscriptions sud-arabes. Douzième série, Le Muséon 68: 297–312. AlSekaf, A.A. 1985. La geographie tribale du Yemen antique, (Dissertation, Paris iii), Paris. Shahid, I. 1986. Kinda. Pages 118–120 in The Encyclopaedia of Islam, 2nd Edition v. Leiden: Brill. Sharafaddīn, A. 2004. Tārīḫ al-Yaman aṯ-ṯaqāfī, al-Qāhia: Maṭbaʿat al-kīlānī aṣ-ṣaġīr, 1967. Ṣanʿāʾ: Ǧāmiʿat Ṣanʿāʾ [reprint 2004]. Sima, A. 2001. Altsüdarabisch lb “Herz”, yd “Hand” und lsn “Zunge”, Acta Orientalia 62: 65–80. az-Zabīdī, Muḥammad b. Muḥammad al-Murtaḍā. Tāǧ al-ʿarūs min ǧawāhir al-qāmūs, 40 vols, Kuwait: Maṭbaʿat Ḥukūmat al-Kuwait, 1965–2001. Zwettler, M. 2000. Maʿadd in Late-Ancient Arabian Epigraphy and other pre-Islamic sources, Wiener Zeitschrift für die Kunde des Morgenlandes 90: 223–307.

chapter 23

Der rituelle Umzug des Yadaʿʾil Ḏarīḥ nach Ṣirwāḥ Norbert Nebes*

Für Michael – in dankbarer Erinnerung an die Diskussionen in Burmington und Weimar

∵ a Zu den prominenten, seit den Anfängen der Sabäistik bekannten Schriftzeugnissen gehört die unter dem Siglum 366 in das cih aufgenommene Monumentalinschrift des Yadaʿʾil Ḏarīḥ an der Mauer des ʾAlmaqah-Tempels in Ṣirwāḥ. Sie ist in drei gleichlautenden Fassungen an der Außenschale der äußeren Umfassungsmauer des Tempels angebracht, wobei die in den bisherigen Kampagnen gefundenen Fragmente darauf hindeuten1, daß sie in ebenso vielen Ausfertigungen an der Innenfassade der gekurvten Tempelmauer eingearbeitet war. Das Referenzschriftband befindet sich an der 22. Quaderlage der Nordwestfassade (Abb. 1–3) in einer – von außen gemessenen – Höhe von 6,35m. Von dem ḏḪ-Symbol zu Anfang und Ende begrenzt, weist es eine Länge von 12,5m mit einer Buchstabenhöhe von 22cm auf2. Der Text lautet in Umschrift:

* Die Abkürzungen der Inschriftensiglen folgen den bei N. Nebes (2016) 95 genannten Verzeichnissen sowie ibid. 96 f. – Dem ehemaligen Präsidenten der General Organization of Antiquities, Museum and Manuscripts (Sanaa, Jemen), Herrn Kollegen Yūsuf ʿAbdallāh, danke ich für die Publikationserlaubnis der folgenden Inschriften aus der Muḥāfaẓat Mārib. Frau Dr. Anne Multhoff, Jena, die das Manuskript kritisch durchgesehen hat, bin ich für substantielle Hinweise sehr zu Dank verpflichtet. 1 Die letzte Kampagne des Deutschen Archäologischen Instituts (dai) in Ṣirwāḥ datiert ins Jahr 2009. Frau Dr. Iris Gerlach, die Leiterin der Außenstelle Sanaa des dai und der Grabungen in Ṣirwāḥ von 2001 bis 2009, gilt mein verbindlicher Dank für die Überlassung der Abbildungen. 2 Die Angaben verdanke ich Herrn Dipl.-Ing. Mike Schnelle, Berlin.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

462

nebes

(Symbol) / ydʿʾl / ḏrḥ / bn / smhʿly / mkrb / sbʾ / gnʾ / byt / ʾlmqh / ywm / hʿ / ḥrmtm / šlṯtʾḏ / w-hwṣt / kl / gwm / ḏ-ʾlm / w-šymm / w-ḏ / ḥblm / w-ḥmrm / b-ʿṯtr / w-b / ʾlmqh / w-b / ḏt / ḥmym / w-b / ʿṯtršymm / (Symbol) Die Schwierigkeit des Textes liegt im Verständnis des Verbums hʿ im ersten Teil des ywm-Satzes, welches in der Vergangenheit verschiedene Interpretationen erfahren hat. W.W. Müller (1985) 659f., der die bisherigen Bearbeitungen zusammengestellt hat, übersetzt: Yadaʿʾil Dharih, der Sohn des Sumhuʿaliy, Mukarrib von Sabaʾ, hat ummauert den Tempel des Almaqah, als er zum dritten Mal ein Heiligtum ausführte und jede Gemeinde eines Gottes und Patrons und eines Bundes und Vertrags einrichtete. Bei ʿAthtar und bei Almaqah und bei Dhat Hamyim und bei ʿAthtar als Patron. Er begründet die Wiedergabe der strittigen Passage inhaltlich damit, daß nach dem ʾAwām in Mārib und dem Maʿrabum-Tempel in al-Masāǧid unweit südwestlich von Mārib das ʾAlmaqah-Heiligtum in Ṣirwāḥ der dritte Sakralbau des Yadaʿʾil sei, den der Mukarrib für ʾAlmaqah erbaut bzw. für den er eine Mauer errichtet habe3. Als alternative Übersetzungsvorschläge verweist W.W. Müller (1985) 660 Fn. f auf „als er dreimal ein (Räucher)opfer darbrachte“ und „als er dreimal im Heiligtum opferte“, welchletztere Bedeutung in Sab. Dict. s.r. hyʿ auch aufgenommen ist. Dort ist u.a. mit „run as a ritual act“ noch ein weiterer Vorschlag verzeichnet, der als auf M.A. Ghūl zurückgehend ausgewiesen ist4. In einer mit Fotos dokumentierten Wiederveröffentlichung der beiden zusammengehörigen Fragmente c 869+c 624 vom Südbau des Großen Dammes von Mārib, wo es heißt: .. / w-hyʿ / ḥr mtm / ṣrwḥ / … hat sich F. Bron (1983) 137 zu Recht für hyʿ im Sinne von M.A. Ghūls Vorschlag ausgesprochen, da sich anders die beiden nominalen Komplimente ḥrmtm und ṣrwḥ syntaktisch nicht sinnvoll einbinden lassen. In seiner nachgelassenen, von Omar al-Ghul herausgegebenen Dissertation schließlich hat sich M.A. Ghūl (1993) 300 ausführlicher zu hyʿ an unserer Stelle geäußert:

3 W.W. Müller (1985) 660 Fn. f und ausführlicher ders. (1980) 63f. sowie ihm folgend N. Nebes (2000) 299 f. 4 So bereits bei A.F.L. Beeston (1981) 23 f. erwähnt.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

463

The act indicated by hʿ/ ḥrmtm / šlṯtʾḏ was either an act of circumambulation round the sanctuary, ḥrmtm, or an act of walking or running between two landmarks, like the hurried walking between al-Ṣafā and al-Marwa, two rocks as their names indicate, in the Meccan pilgrimage (ibid.). Daß dieser Vorschlag in die richtige Richtung weist, wird durch eine viele Jahrhunderte später gesetzte, unveröffentlichte Altaraufschrift des Ḏamarʿalī Ḏarīḥ bestätigt5, auf der die unlängst vom Verfasser dieser Zeilen vorgeschlagene Neuübersetzung der fraglichen Passage in der Bauinschrift des Yadaʿʾil beruht6. Im folgenden soll die Inschrift des Ḏamarʿalī Ḏarīḥ vorgestellt, die Übersetzung i.e. begründet und der Kontext näher ausgeleuchtet werden, in den die Passage ywm / hʿ / ḥrmtm / šlṯtʾḏ zu stellen ist.

b Die vierzeilige Inschrift ist auf der Stirnseite eines rechteckigen Altarblocks mit erhöhten Außenkanten und zwei Auslässen auf der gegenüberliegenden Seite angebracht (Abb. 4–5). Sie wurde erstmals im März 1997 im Magazin der Muḥāfaẓat Mārib vom Verfasser dieser Zeilen aufgenommen. Der Aussage des damaligen Antikendirektors von Mārib, Muḥammad Muṣliḥ al-ʿIzzī, zufolge stammt der Altarblock aus Ṣirwāḥ. Im Januar 2004 befand er sich im unweit vom neuen Damm gelegenen, sogenannten Doguş-Camp, wo er von Dr. Burkhard Vogt, dem damaligen Leiter der Station Ṣanʿāʾ des dai, fotografiert wurde, dem die hier veröffentlichten Aufnahmen zu verdanken sind. Das Schriftfeld hat die Maße 75×34cm. Die Buchstabenhöhe beträgt 7,4 cm. Das Symbol zu Anfang von Zeile 1 stellt den Kelch einer Tulpe dar, dem eine schlechtere Ausführung unten angesetzt ist. Bezug genommen ist in verkürzter Form auf das ḏḪ-Symbol der sabäischen Mukarribe, wobei die unten angesetzte Ausführung der Tulpe deutlich in zwei ovalförmige Enden ausläuft, die die beiden Schlangenköpfe, die in den frühen Ausführungen dieses Symbols aus Ṣirwāḥ zu erkennen sind, andeuten sollen.

5 Zur Datierung des Ḏamarʿalī siehe unter Abschnitt B. Yadaʿʾil Ḏarīḥ wird herkömmlich nach Karibʾil Watar um die Mitte des 7. Jh. v. Chr. angesetzt. 6 N. Nebes (2011) 372: „als er zum dritten Mal einen rituellen Umzug (nach Ṣirwāḥ) unternahm“.

464

nebes

Muḥāfaẓat Mārib 11 Transkription 1. 2. 3. 4.

(Symbol) ḏmrʿly / ḏrḥ / mlk / sb{ʾ} [/ w-ḏ-] rydn / bn / krbʾl / wtr / šm tqyf / ʿfwn / w-{n}bl{n} / {y}wm / h {y}{ʿ} / ḥrmt / ʿdy / ṣrw{ḥ} Übersetzung (1) Ḏamarʿalī Ḏarīḥ, der König von Sabaʾ [und ḏū] (2) Raydān, der Sohn des Karibʾil Watar, hat (3) den Qayf-Altar der ʿfwn und nbln errichtet, als er (4) einen rituellen Umzug nach Ṣirwāḥ unternahm.

Kommentar Zeile 1f. Ḏamarʿalī Ḏarīḥ, der Sohn des Karibʾil Watar, wird gegen 80 n. Chr. angesetzt, für welchen Zeitraum auf die Ausführungen von C.J. Robin (1994) insbes. 108 (Tabelle) verwiesen sei. Weitere von diesem König gesetzte (und publizierte7) Inschriften sind mafray Mušǧiʿ 32, Muḥāfaẓat Mārib 20 (s.u.), j 2954, Ry 544, Muḥāfaẓat Mārib 24 (s.u.), ferner das im Nordbau des Großen Dammes von Mārib verbaute Fragment c 729 = dai gdn 2002–30 sowie die zweizeilige Sockelaufschrift auf dem linken, bei W. Daum (2000) 227 abgebildeten Foto. Zeile 2f. Das an dieser Stelle erstmals bezeugte Nomen tqyf ist vermutlich als Nomen actionis dem 02-Stamm der Verbalwurzel qyf zuzuordnen, vgl. analoge Bildungen wie z.B. tšqr „completion, finishing“ in Sab. Dict. 133 und andere bei P. Stein (2003) 60. In Verbindung mit dem Verbum š(y)m, welches auch für die Errichtung von Altären verwendet wird (z.B. c 382/1), ist die ganze Phrase mit „eine Qayf-Konstruktion errichten“ zu übersetzen. Mit dem Substantiv qyf bzw. qf ist oft nicht die Gedenkstele, womit das Wort herkömmlich übersetzt wird, sondern, wie in Sab. Dict. s.r. angegeben, jede Art von Kultstein gemeint, welche Bedeutung auch für einen Altar wie in unserem Fall zutrifft8. Die zahlreichen Beispiele für das Nomen q(y)f zeigen, daß es zumeist als erster Bestandteil in 7 Siehe Abschnitt c. 8 Zur Problematik der genauen Bestimmung von qyf vgl. C.J. Robin (2012) 32f. (mit den entsprechenden Beispielen), demzufolge der Qayf des Ǧabal al-Miʿsāl den Felsen („le piton dʾalMiʿsāl“) mit den darauf angebrachten Inschriften bezeichnen kann.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

465

eine Status-constructus-Fügung eingebunden ist, deren Rektum, sofern es sich um einen Namen handelt, in der Hauptsache von einem Götternamen9, seltener von dem Namen eines Herrschers10, aber auch von einem Ortsnamen11 gebildet werden kann12. Wie diesen und anderen analogen Bildungen zu entnehmen ist, können mit ʿfwn und nbln nicht die Namen von Altären gemeint sein, an denen die Inschrift angebracht ist, da man in einem solchen Fall von der Grammatik her eine appositionelle Fügung mit Determination des ersten Bestandteils, wie tqyfnhn / ʿfwn / w-nbln, und keine Status-constructus-Verbindung erwartet13. Zwei andere Möglichkeiten kommen jedoch in Betracht: Bei ʿfwn und nbln handelt es sich entweder um Ortsnamen bzw. um die Kultplätze, auf denen Ḏamarʿalī eine Qayf-Konstruktion errichtet hat, oder aber es handelt sich um Substantive in einer ganz konkreten Bedeutung. Aufschluß erhalten wir anhand zweier weiterer Stellen, an denen sich ʿfwn und nbln bzw. ʿfwn (ohne nbln) bislang nachweisen lassen: Unter den kleinen Fragmenten aus der Sammlung Eduard Glaser ist unter dem Siglum Gl 936+Gl 935 ein Fragment aufgenommen, welches von der Bearbeiterin H. Tschinkowitz wie folgt transkribiert wird: .. {d}. / qyf / my{d}n / ʿfwn / wḥbln / {b}……n / ywm / hyʿ / ḥrmt[m] Auch wenn eine Fotografie des Abklatsches, der bereits von M. Höfner (1944) 43 als „schlecht zu lesen“ bezeichnet wird, nicht veröffentlicht ist und der Wortlaut daher i.e. nicht überprüft werden kann, so ist der Bezug zu unserer Altarinschrift aufgrund des abschließenden Passus ywm / hyʿ / ḥrmt[m] und der beiden Nomina ʿfwn und ḥbln offensichtlich, wobei letzteres zweifellos zu nbln zu

9 10 11

12 13

Z. B. c 11/1 (qyf / šms-hmw), c 458 (qyf / ʿṯtr / w-sḥr). Z. B. c 390/2 (qf / ykrbmlk). r 4177/3 (qyf / ḫlfy / nwmm). Eine Status-constructus-Verbindung mit qyf und einem folgenden Ortsnamen kann auch Bestandteil von Gottestitulaturen sein, wie bei den Erscheinungsformen der Göttin Šams, die als „Herrin des Qayf von ršm“ (nnag 6/29 = j 627/29: bʿlt / qyf / ršm), als „Herrin des Qayf von wynn und ʿlfqn“ (r 3958/11: bʿlt / qyf / wynn / w-ʿlfqn) oder am Ǧabal al-Miʿsāl (siehe vorherige Fn. 8) als „Herrin des Qayf des (Berges) Šiḥrārum“ (al-Miʿsāl 4/1: bʿlt / qyf / šḥrrm) angesprochen wird. Vgl. auch die Beispiele für das Verbum q(y)f bei A. Multhoff (2011) 659f. Siehe unten die entsprechenden Ausführungen zu mwṯb und auch C.J. Robin (2012) 63 Fn. 195, der sich in Zusammenhang mit mwṯb / ʿfwn (s.u.) in diesem Sinne geäußert hat.

466

nebes

korrigieren ist. Mit einiger Sicherheit können wir daher annehmen, daß dieses Fragment Bestandteil einer weiteren, nicht wortidentischen Fassung unserer Altarinschrift ist und an einer anderen Ausfertigung einer Qayf-Konstruktion, voraussichtlich ebenfalls an einem Altar, angebracht war. Ob dieser ebenfalls von Ḏamarʿalī anläßlich desselben rituellen Umzuges aufgestellt wurde, läßt sich nicht mit Sicherheit beantworten. Auch ist wenig mit der Fundortangabe „Dimne“ anzufangen, die von H. Tschinkowitz nicht näher erläutert wird. In der Tabelle des Verzeichnisses des Glaser-Nachlasses bei M. Höfner (1944) 15 sowie 44 (s.o.) ist dieser Ort unter den Glaserschen Herkunfts- bzw. Fundorten singulär und ohne weitere Parallele. In der ibid. 15 verzeichneten Tabelle wird er zwischen den aus „Ḥaṣbā14 ben Nadā“ und aus „ʿOrqūb el Masʾabein“ stammenden Fundnummern aufgeführt, welche Orte bei E. Glaser (1913) Karte 1 (im Anhang) keine 20km südwestlich von Mārib entfernt am Oberlauf des Wadi Quṭūṭa eingetragen sind. Die zweite Inschrift, in der lediglich ʿfwn bezeugt ist, ist ebenfalls nur als Glaser-Abklatsch (Gl a 731 a–d) vorhanden. Es handelt sich dabei um eine Widmung des Halakʾamar, der sich als Sohn des Karibʾil Watar Yuhanʿim bezeichnet, der auch der Vater des Ḏamarʿalī Ḏarīḥ unserer Inschrift ist15. Gerichtet ist die Dedikation an ʿAṯtar ḏū Ḏībān, dem der mwṯb / ʿfwn gestiftet wird, bei dem es sich um eine Art Steinthron oder aber auch um eine steinerne Sitzbank gehandelt haben wird. Wie schon bei tqyf / ʿfwn / w-nbln kann auch in dieser Nominalfügung ʿfwn hier nicht den Namen des Mawṯab bezeichnen, wie M. Höfner in ihrem Kommentar zur Stelle annimmt und wogegen sich bereits C.J. Robin (2012) 63 Fn. 195 ausgesprochen hat16. Der Ort, an dem Eduard Gla14 15

16

In der Karte von E. Glaser (s. u.) als „Naṣbā“ angegeben. Demnach müssen Ḏamarʿalī und Halakʾamar Brüder gewesen sein, wenn nicht Halakʾamar der eigentliche Personenname ist, den Ḏamarʿalī vor seiner Thronbesteigung getragen hat. Grammatisch ist dies durch die Status-constructus-Verbindung eindeutig festgelegt. Eine Durchsicht der Beispiele für mwṯb zeigt, daß hier sehr wohl zwischen einer appositionellen Fügung und einer Status-constructus-Verbindung unterschieden wird. Im ersten Fall handelt es sich um determiniertes mwṯbn bzw. pronominal determiniertes Bezugswort mwṯb- mit folgendem Namen, womit dann in der Tat der Eigenname der „Sitzgelegenheit“ gemeint ist, so in j 600/8: mwṯbn / yhgl „der Thron yhgl“, c 308/4f.: mwṯb-hmw / yhgl „ihr Thron yhgl“, c 660/3: mwṯb-hmw / ġyln „ihr ʿSitzʾ Ġaylān“. Im zweiten Fall, der Status-constructus-Verbindung, kann das Rektum nicht als Name des vorausgehenden Bezugswortes mwṯb aufgefaßt werden, vgl. z. B. y.90.b.a 3 (aus dem Barrān-Tempel in Yaṯill/Barāqiš): mwṯb / qhltn „die (steinerne) Sitzbank der Versammlung (sc. des ʿAṯtar)“, ferner Kamna 24/5: mwṯb / šwʿn „der Thron der ʿGefolgschaftʾ“ und Kamna 25/6f.: mwṯb / šwʿ / ʿṯtr „der Thron der ʿGefolgschaftʾ des ʿAṯtar“.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

467

ser die Inschrift auf diesem „hohen stelenförmigen Stein“17 aufgenommen hat, wird mit „Dēret ibn Qauzān“ angegeben, welches in Karte 1 (im Anhang) bei E. Glaser (1913) ca. 15km nordwestlich von Mārib gelegen auf dem Weg ins Wadi Raġwān verzeichnet ist. Dort wird dann auch der Ort zu suchen sein, an dem die dem ʿAṯtar ḏū Ḏībān gewidmete steinerne „Sitzgelegenheit“ aufgestellt war. Fassen wir zusammen: In zwei Inschriften, Muḥāfaẓat Mārib 11 und Gl 936+ Gl 935, werden in Verbindung mit einer Qayf-Konstruktion, anläßlich eines ḥrmt-Umzuges die Begriffe ʿfwn und nbln genannt. In einer dritten Inschrift, Gl a 731 a–d, wird lediglich ʿfwn in Zusammenhang mit einem Mawṯab erwähnt. Für alle drei Inschriften sind unterschiedliche Herkunftsangaben überliefert: Muḥāfaẓat Mārib 11 kommt aus Ṣirwāḥ, Gl a 731 a–d aus einem Ort, ca. 15km nordwestlich von Mārib auf dem Weg ins Wadi Raġwān, und Gl 936+Gl 935 stammt – dies ist allerdings nur eine Vermutung – vom Oberlauf des Wadi Quṭūṭa, ca. 20km südwestlich von Mārib. Wenn wir nun annehmen, daß ʿfwn und nbln Ortsnamen bezeichnen, so ist ʿfwn in Gl a 731 a–d, wo mwṯb / ʿfwn zu lesen ist, noch am ehesten sicher zu lokalisieren und aufgrund der Glaserschen Angabe „Dēret ibn Qauzān“ ca. 15km nordwestlich von Mārib zu suchen. Da es sich bei ʿfwn an allen drei Stellen vermutlich um ein und denselben Ort handelt, müßte der ḥrmt-Umzug des Ḏamarʿalī nach Ṣirwāḥ den Umweg über einen Ort namens ʿfwn genommen haben, an dem zu seinen Lebzeiten oder wenig später ein „Sitz“ für ʿAṯtar ḏū Ḏībān aufgestellt wurde, dessen Heiligtum sich in verlängerter nördlicher Richtung am Ǧabal al-Lawḏ befindet. Mehr Plausibilität kann dagegen die andere Möglichkeit für sich beanspruchen. Sie geht auf einen Vorschlag von C.J. Robin (2012) 63 Fn. 195 zurück, wonach mit ʿfwn „une catégorie de personnes visitant le sanctuaire“ gemeint sein könnte. C.J. Robin ibid. hat diesen Vorschlag nicht weiter ausgeführt. Es ist aber offensichtlich, daß er sich an klassisch–arabisch ʿafā orientiert, für das die Lexika u.a. die Bedeutung „zu jmd. kommen, um eine Wohltat zu erbitten“ verzeichnen18. Demnach handelt es sich bei ʿfwn um die Teilnehmer des ḥrmt-Umzugs, die sich auf dem Hinweg zum Heiligtum befinden und möglicherweise am Qayf-Altar ein Opfer bringen, bevor sie in das Heiligtum eintreten. Dementsprechend sind dann unter den nbln die Prozessionsteilnehmer zu verstehen, die aus dem Heiligtum „entlassen“19 sind und die den Rückweg 17 18 19

M. Höfner in den Ausführungen zum Fundort von Gl a 731 a–d; vgl. auch M. Höfner (1944) 69. Vgl. z. B. Lane 2093c: „ʿafawtuhū (…) I came to him seeking, or demanding, his beneficence“. In diesem Sinne – und damit durchaus mit der Grundbedeutung „schicken, senden“ vereinbar – ist nbl aufzufassen. nbln und ʿfwn wären beide als Partizipien anzusetzen.

468

nebes

nach Mārib antreten. Diese Interpretation hat zum einen den Vorteil, daß es sich bei der Qayf-Konstruktion der ʿfwn und nbln um ein und denselben Altar handelt, der vom König anläßlich seines Umzug aufgestellt wurde, zum anderen, daß die notwendige Verknüpfung mit konkreten Orten entfällt, über die der Umzug geführt haben muß. So ist es durchaus denkbar, daß eine QayfKonstruktion der ʿfwn und nbln auch vor anderen Heiligtümern aufgestellt war, die Ziel eines rituellen Umzugs gewesen sind20. Zeile 3f. Daß der Passage hyʿ / ḥrmt / ʿdy / ṣrwḥ die obige Übersetzung zugrunde zu legen und dieser auch den anderen Wiedergaben des Passus in c 366 der Vorzug zu geben ist, geht aus den obigen Ausführungen unschwer hervor und ist auch inhaltlich begründet, da die bisherigen Übersetzungen von c 366 – auf unsere Altarinschrift bezogen – wenig Sinn ergäben, wonach Ḏamarʿalī eine Qayf-Konstruktion anderenorts errichtet, wenn er eine Tempelmauer in Ṣirwāḥ baut oder ein Räucheropfer daselbst darbringt. Obige Übersetzung erfolgt jedoch keineswegs nur aus dem Kontext heraus, sondern orientiert sich an der Grundbedeutung von hʿ, hyʿ, die als „gehen, (ver)laufen“ anzusetzen ist21. Mit der Grundbedeutung ohne weiteres vereinbar ist die modifizierte Wiedergabe der paronomastischen Wendung hʿ / mhyʿ mit „einen rituellen Umzug durchführen“22. Ikonographisch umgesetzt ist ein solcher Umzug auf dem Bronzerelief ym 13981, auf dem eine Prozessionsszene von sechs nach links schreitenden Kriegern mit Reliefbögen in der Linken und abgeschlagenen Händen in der Rechten zu sehen ist23. Daß es sich dabei um eine solche auch wirklich handelt, geht eindeutig aus der unmittelbar darunter angebrachten Inschrift mhyʿ / ʾwsʿṯt / bn / yhʿn / ʿhmyn hervor, wobei mit dem Schlüsselwort mhyʿ nichts anderes als ein ritueller Umzug gemeint sein kann, den der ʿĀhimit ʾAwsʿaṯt, der Sohn des Yuhaʿīn, in der dargestellten Szene veranstaltet. Die genaue Bedeutung der Verbindung hyʿ / ḥrmtm ist dagegen weitaus weniger durchsichtig.

20

21

22 23

Der Qayf-Altar steht in Ṣirwāḥ nicht notwendigerweise vor dem Eingang in den Tempelkomplex, sondern kann auch in näherer und weiterer Umgebung errichtet worden sein. Sicher ist, daß er nicht direkt im Heiligtum aufgestellt war. Vgl. dazu ausführlich die Diskussion bei A.F.L. Beeston (1981) 21–24 und N. Nebes (2000) 298–301, Beispiele auch bei A. Multhoff (2011) 405–407. Von dieser Grundbedeutung leitet A.F.L. Beeston (1981) 22 den „causative sense“ im Sinne von „carry out, execute (an action, an instruction)“ (ibid.) ab, den hyʿ (j 831/2) und das zugehörige Nomen actionis mhyʿ (c 338/10) zweifellos haben können, vgl. Sab. Dict. 57 und N. Nebes (2000) 299. r 3946/7: ywm / hʿ / mhyʿ / lqẓ / ʿṯtr / ḏ-fṣd, siehe zuletzt N. Nebes (2016) 86 mit Fn. 311. Siehe ausführlich I. Gerlach (2010) insbes. 267–269 und 279–284.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

469

Im Unterschied zum Nomen loci mḥrm kann ḥrmt, wenn ich richtig sehe, zwar nicht das architektonisch definierte Heiligtum, jedoch, wenn auch selten, den heiligen Bezirk bezeichnen24. Doch auch diese Bedeutung kann für ḥrmt in den hier in Rede stehenden Stellen nicht veranschlagt werden. In diesem Fall hätten wir in dem Monumentalinschriftenfragment w-hyʿ / ḥr mtm / ṣrwḥ (c 869+c 624, s.o.) von zwei präpositionslosen Nomina auszugehen, die beide im Satz die Funktion von adverbialen Ortsbestimmungen, und zwar einer adverbialen Richtungsangabe („zu einem heiligen Bezirk“) bzw. einer adverbialen Ortsangabe („in Ṣirwāḥ“), übernähmen, was syntaktisch überaus ungewöhnlich wäre, ganz abgesehen von den inhaltlichen Vorbehalten, die gegen eine derartige Interpretation vorzubringen sind25. Die einzig sinnvolle Möglichkeit, die auch der Syntax des Satzes einigermaßen Rechnung trägt, besteht darin, ḥrmtm – analog zu mhyʿ in obigem Beispiel aus r 3946/7 – als inneres Objekt zu hyʿ zu nehmen, womit der Passus mit „einen ḥrmt-Umzug nach Ṣirwāḥ durchführen“ zu übersetzen ist. Daß mit der Phrase h(y)ʿ / ḥrmtm der Umzug zu einem Heiligtum und auch zu einem heiligen Bezirk impliziert wird, ist offensichtlich, doch liegt einer derartigen Auffassung die syntaktische Analyse von ḥrmtm als innerem Objekt und nicht als präpositionsloser, adverbialer – und darüber hinaus indeterminierter – Umstandsbestimmung („zu einem Heiligtum“) zugrunde. Die Gegenüberstellung der beiden Passagen (w-)hyʿ / ḥr mtm / ṣrwḥ und ({y}wm /) h{y}{ʿ} / ḥrmt / ʿdy / ṣrw{ḥ} zeigt deutlich die Entsprechung der einzelnen Satzglieder, wobei die präpositionslose Adverbialbestimmung ṣrwḥ in der früheren Version26 durch den Präpositionalausdruck ʿdy / ṣrwḥ in der späteren Fassung analytisch auskommentiert wird. Auffällig ist allerdings in der Version des Ḏamarʿalī das Fehlen der Mimation bei ḥrmt, so daß eine Status-constructus-Verbindung mit einer Präpositionalangabe als Rectum nicht völlig ausgeschlossen werden kann. Doch bietet die Version mit ḥrmtm und ṣrwḥ zwei voneinander unabhängige Satzglieder und damit eine Konstruktion, die in der späteren Formulierung – lediglich mit präpositionaler Auskommentierung der Richtungsangabe (ʿdy / ṣrwḥ) – beibehalten worden sein wird, so daß das Fehlen der Mimation in der Altarinschrift als grammatische Unkorrektheit wohl eher in Kauf genommen werden kann. 24

25 26

So r 4176/7: b-ḥrmt / ʾtmn „im heiligen Bezirk von Utmān“ (W.W. Müller); ḥrmtn als Name des Heiligtums der Gottheit Sāmiʿ ist bezeugt in c 282/3. Unklar und unsicher ist ḥrmt in c 290/6, c 292/2 und Gl 1628/4. So wäre die Frage zu beantworten, wie in dem konkreten Fall eines rituellen Umzugs nach Ṣirwāḥ die indeterminierte Form ḥrmtm zu begründen wäre. Das Fragment stammt spätestens aus dem 6. Jh. v. Chr., so daß zwischen beiden Versionen ein Abstand von mindestens über 500 Jahren anzusetzen ist.

470

nebes

In diesem Zusammenhang ist noch die abschließende Frage zu stellen, die bereits M.A. Ghūl aufgeworfen hat, der sie aber im Falle des Yadaʿʾil Ḏarīḥ offen läßt, nämlich ob der Umzug innerhalb des Heiligtums oder von einem zum anderen Ort stattgefunden hat27. Aus den voraufgegangenen Ausführungen ist hinlänglich deutlich geworden, daß es sich wohl weniger um einen Umzug innerhalb des Heiligtums in Ṣirwāḥ als vielmehr um einen Umzug der beiden Herrscher nach Ṣirwāḥ gehandelt hat, der vermutlich von einem Heiligtum in Mārib seinen Ausgang genommen haben wird. Dabei spielt es für unsere Interpretation keine Rolle, ob Yadaʿʾil in seiner Eigenschaft als Mukarrib „zum dritten Mal“ oder „dreimal“ – und in letzterem Fall in entsprechend zeitlichem Abstand, wie šlṯtʾḏ korrekterweise wiederzugeben ist28, – den Umzug nach Ṣirwāḥ durchgeführt hat.

c Namentlich von Ḏamarʿalī Ḏarīḥ besitzen wir eine Reihe weiterer Inschriften, die dieser König anläßlich ritueller Handlungen gesetzt hat, die wir aus der Zeit der Mukarribe des 8. und 7. Jh. v. Chr. kennen. So kommen nach Auskunft von C. Robin29 (mindestens) sechs Inschriften vom Ǧabal al-Lawḏ, die Ḏamarʿalī Ḏarīḥ (vermutlich alle) im Zusammenhang mit der Ausrichtung des rituellen Mahls und dem „Feueropfer“ in dem dortigen Heiligtum trḥ gesetzt hat. Ausführlich ist dies auf mindestens zwei Ausfertigungen eines Tischaltars vom Ǧabal al-Lawḏ dokumentiert, von denen sich einer heute in der Muḥāfaẓat Mārib befindet30:

27 28

29 30

Siehe oben das ausführliche Zitat aus M.A. Ghūl (1993) 300. So zu Recht angesetzt von P. Stein (2003) 241 mit Verweis auf das Ugaritische. – Ob es eine besondere Bewandtnis damit hat, daß der Umzug dreimal ausgeführt worden ist, läßt sich schwer sagen. Auffallend ist in jedem Fall, daß die dreimalige Ausführung von Kulthandlungen auch an anderer Stelle begegnet, so das dreimalige Feuer(opfer) im Heiligtum trḥ am Ǧabal al-Lawḏ (m 48 nach C. Robin, J.-F. Breton (1982) 614 (s.u.)) und das dreimalige Schlachtopfer des Karibʾil Watar für ʿAṯtar (r 3945/1). C. Robin, J.-F. Breton (1982) 606 und 613. Der andere mit der Inschrift mafray Mušǧiʿ 32 ist veröffentlicht von C. Robin in C. Robin, J.-F. Breton (1982) 609 mit Foto und Teilübersetzung. Im Unterschied zu der in der Muḥāfaẓat Mārib befindlichen Altarinschrift weist mafray Mušǧiʿ 32 eine leicht veränderte Zeilendisposition auf, auch sind die beiden Auslässe des Altars auf Seite c nicht mehr vorhanden, siehe ibid. Abbildung c.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

471

Muḥāfaẓat Mārib 20 (Abb. 6–9) Tischaltar mit erhöhten Außenkanten, zwei Auslässe auf Seite c; Kalkstein, 49×40×7cm, Buchstabenhöhe: 6–7cm. a b c d

ḏmrʿly / ḏrḥ / mlk / sbʾ / w-ḏ-ryd n / hqny / ʾlʾlt / kwrn [/ yw] m / ʿl [Auslaß] y / tr [Auslaß] ḥ / bk n / ʾlm / ʿṯtr / w-hnr / b-trḥ (a) Ḏamarʿalī Ḏarīḥ, der König von Sabaʾ und ḏū Raydān, (b) hat den Göttern des (Berges) kwrn31 (diesen Altar) gewidmet, als (c) er zum Heiligtum trḥ hinaufstieg, als (d) er dem ʿAṯtar ein rituelles Mahl ausrichtete und in trḥ ein Feuer(opfer) anzündete.

Einen weiteren Qayf-Altar hat Ḏamarʿalī Ḏarīḥ anläßlich der rituellen Jagd des ʿAṯtar und Karwam und damit anläßlich einer rituellen Handlung aufgestellt, die bislang ausschließlich für die Mukarribe der frühen Zeit nachgewiesen ist32. Der Altar befindet sich heute ebenfalls in der Muḥāfaẓat Mārib, die darauf angebrachte Inschrift lautet33: Muḥāfaẓat Mārib 24 Altarblock mit vertiefter Wanne (50×37cm), Auslaß auf der der Inschrift gegenüber liegenden Seite, zwei kreisrunde Einlassungen mit ca. 20 cm Durchmesser an der Oberseite; 97×31×45cm, Buchstabenhöhe 8 cm. 1. 2. 3.

ḏmrʿly / ḏrḥ / mlk / sbʾ / w-ḏ-rydn / bn / kr[bʾ] l / wtr / bny / qyf / šrqn / b-wṭḥtn / ywm / ṣy d / ʿṯtr / w-krw{m}

31 32

Antiker Name des Ǧabal al-Lawḏ. Für Yiṯaʿʾamar Bayyin bin Sumuhūʿalī: y.85.aq/7, r 4177/4, Schm/Mārib 23; für Karibʾil Watar: r 3946/7 (ṣyd / krwm [ohne ʿṯtr]); Weiteres zur Jagd des ʿAṯtar und Karwam bei W.W. Müller (1986). In der Felsinschrift Ry 544/3 f. des Ḏamarʿalī Ḏarīḥ vom Naqīl Šiǧaʿ, ca. 45km nördlich von Ṣanʿāʾ, die von der Jagd des Königs in dieser Gegend berichtet, ist der Passus ywm / ṣyd / ṣyd / ʿṯtr / w-kr (4) wm ebenfalls enthalten, allerdings orientiert sich der Text nicht mehr am Formular und an der Diktion der altsabäischen Zeit, wie dies für die im folgenden mitgeteilte Altarinschrift mit gewissen Abstrichen der Fall ist. So ist dort beispielsweise u. a. die Zahl der erjagten Tiere genannt.

33

472

nebes

(1) Ḏamarʿalī Ḏarīḥ, der König von Sabaʾ und ḏū Raydān, der Sohn des Kari[bʾi]l (2) Watar, hat den Gedenkaltar des ⟨ʿAṭtar⟩ Šāriqān34 in wṭḥtn35 errichtet, als er die rituelle Jagd (3) des ʿAṯtar und Karwam ausführte36. Die Ausrichtung des rituellen Gastmahls für ʿAṯtar ḏū Ḏībān am Ǧabal al-Lawḏ und das Feuer(opfer) im dortigen Heiligtum trḥ, die rituelle Jagd des ʿAṯtar und Karwam sind allesamt rituelle Handlungen, die zur Kult- und damit zur Herrschaftspraxis der frühen Mukarribe gehören und diesen auch vorbehalten waren. Daß Ḏamarʿalī und seine unmittelbaren Vorgänger diese Rituale zu Beginn des 1. Jh. wieder reaktiviert haben, hat sicherlich seine politischen Gründe gehabt. Einen deutlichen Hinweis in diese Richtung liefert die Inschrift des Ḏamarʿalī Watar Yuhanʿim vom Ǧabal al-Lawḏ, den C. Robin (1994) 108 in der ersten Hälfte des 1. Jh. Chr. ansetzt und in dessen dynastischer Nachfolge Ḏamarʿalī an dritter Stelle steht. Die Inschrift ist, auch wenn es nicht expressis verbis festgehalten ist, anläßlich des rituellen Gastmahls gesetzt und enthält die Bundesschließungformel, in der der Machtanspruch der frühen Mukarribe über die tribalen Gemeinschaften in Südarabien formuliert wird37. Wie C. Robin überzeugend dargelegt hat38, stehen die Reaktivierung der Bundesschließungsformel und die Rituale am Ǧabal al-Lawḏ zu Beginn des 1. Jh. in direktem Zusammenhang mit der Vereinigung der Königshäuser Sabaʾ in Mārib und ḏū Raydān in Ẓafār unter sabäischer Ägide. In Verbindung damit sind aber

34 35 36 37

38

Ebenso ohne ʿAṭtar vgl. z. B. c 452/1,3, Schm/Sir 41, c 453, r 3930. Bislang nicht bezeugter Ort, der möglicherweise, worauf mich Anne Multhoff hinweist, in fragmentarischem c 588/a2 vorliegt. Hier ohne paronomastische Figur (ṣ(y)d / ṣ(y)d) konstruiert. Die Inschrift ist erstmals von C. Robin in C. Robin, J.-F. Breton (1982) insbes. Fig. 13 (Foto) unter dem Siglum mafray Mušǧiʿ 18 publiziert. Sie ist auf einer Stele mit vertieftem Register angebracht, bei der es sich möglicherweise um die Rückenlehne einer steinernen Sitzgelegenheit handelt. Die Transkription und Übersetzung lauten nach dem Foto bei W. Daum (2000) 229, der die vollständigere Inschrift bietet: (1) mwṯ{b} {/} [ḏ]m{r}[ʿ]{l}[ y / w] (2) tr / yhnʿm / mlk (3) sbʾ / w-ḏ-rydn / bn (4) smhʿly / ḏrḥ / yw (5) m / ʾlm / ʿṯtr / ḏ-ḏb (6) {n} / w-{h}{n}r-hw / b-tr (7) ḥ / ywm / hwṣt / k (8) l / gwm / ḏ-ʾlm / w-šym (9) m / w-ḏ-ḥblm / w-ḥmrm „Der ʿSitzʾ des [Ḏ]amar[ʿ]al[ī] [Wa]tar Yuhanʿim, des Königs von Sabaʾ und ḏū Raydān, als er dem ʿAṯtar ḏū Ḏībān ein Gastmahl ausrichtete und ihm ein Feuer(opfer) im (Heiligtum) trḥ anzündete, als er über jede Stammesgemeinschaft eines Gottes und eines göttlichen Patrons und eines Bundes und Vertrags gebot“. Zur Bundesschließungsformel und ihren Implikationen zuletzt N. Nebes (2016) 68–72. C. Robin, J.-F. Breton (1982) 619f. mit Verweis auf Muḥammad Bāfaqīh, der die Bestandteile des Königsnamens Ḏamarʿalī Watar Yuhanʿim als politisches Programm deutet.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

473

auch andere Kult- und Herrschaftspraktiken des 8. und 7. Jh. von den sabäischen Königen in Mārib wieder ins Leben gerufen worden, wie die rituelle Jagd und der Umzug nach Ṣirwāḥ, wie am Beispiel des Ḏamarʿalī Ḏarīḥ gezeigt werden konnte. Für den ḥrmt-Umzug des Yadaʿʾil, der bislang nur für diesen Mukarrib bezeugt ist39, bedeutet dies aber, daß ein solcher Umzug zum festen Inventar der Kult- und Herrschaftspraxis des 8. und 7. Jh. v. Chr. gehört haben muß, wenn er über 700 Jahre später von einem sabäischen König wieder aufgegriffen wird.

Zitierte Literatur Beeston, A.F.L. 1981. „Two Epigraphic South Arabian Roots: hyʿ and krb.“ In Al-Hudhud. Festschrift Maria Höfner zum 80. Geburtstag, edited by R.G. Stiegner, 21–34. Graz: Karl-Franzens-Universität Graz. Bron, F. 1983. „Inscriptions de la digue de Mârib.“ Aula Orientalis 1: 137–153. Daum, W. 2000. „Der heilige Berg Arabiens.“ In Im Land der Königin von Saba. Kunstschätze aus dem antiken Jemen, Staatliches Museum für Völkerkunde München, edited by W. Daum, W.W. Müller, N. Nebes, 223–232. München: Staatliches Museum für Völkerkunde München. Ghūl, M.A. 1993. Early Southern Arabian Languages and Classical Arabic Sources. A Critical Examination of Literary and Lexicographical Sources by Comparison with the Inscriptions, edited by O. Al-Ghul. Irbid: Yarmouk University. Gerlach, I. 2000. „Zur Übernahme altorientalischer Motive in die Kunst Südarabiens. Eine reliefierte Bronzeplatte aus dem Jemen.“ Baghdader Mitteilungen 31: 259–295. Glaser, E. 1913. „Eduard Glasers Reise nach Mârib“, edited by D.H. von Müller and N. Rhodokanakis. (Sammlung Eduard Glaser 1). Wien: A. Hölder. Höfner, M. 1944. Die Sammlung Eduard Glaser. Verzeichnis des Glaser-Nachlasses, sonstiger südarabischer Materialbestände und einer Sammlung anderer semitischer Inschriften. Akademie der Wissenschaften in Wien, Philosophisch–historische Klasse, Sitzungsberichte, 222. Band, 5. Abhandlung. Wien: Akademie der Wissenschaften. Müller, W.W. 1980. „Altsüdarabische Miszellen (i).“ Raydān 3: 53–73. Müller, W.W. 1985. „Altsüdarabische und frühnordarabische Inschriften.“ In Texte aus der Umwelt des Alten Testaments, Bd. 1, Rechts- und Wirtschaftsurkunden, Historisch–

39

Der Autor von c 869+c 624 ist unbekannt. Durchaus wahrscheinlich ist, daß das im Monumentalduktus ausgeführte Fragment mit seinen 19cm hohen Buchstaben, welches Teil eines längeren Inschriftenbandes ist, ebenfalls von einem Mukarrib stammt.

474

nebes

chronologische Texte, Lieferung 6, Historisch–chronologische Texte iii, edited by O. Kaiser, 651–668. Gütersloh: Gütersloher Verlagshaus. Müller, W.W. 1986. „krwm im Lichte einer neuentdeckten sabäischen Jagdinschrift aus der Oase von Mārib.“ Archäologische Berichte aus dem Yemen 3: 101–107. Multhoff, A. 2011. Die Verbalstammbildung im Sabäischen. Dissertation Jena. Nebes, N. 2000. „Zu den Inschriften auf einer reliefierten Bronzeplatte aus dem Jemen.“ Baghdader Mitteilungen 31: 297–308. Nebes, N. 2011. „Altsüdarabische Bauinschriften.“ In Grab-, Sarg-, Bau- und Votivinschriften, Texte aus der Umwelt des Alten Testaments, Neue Folge, vol. 6, edited by B. Janowski and D. Schwemer, 367–387. Gütersloh: Gütersloher Verlagshaus. Nebes, N. 2016. Der Tatenbericht des Yiṯaʿʾamar Watar bin Yakrubmalik aus Ṣirwāḥ ( Jemen). Zur Geschichte Südarabiens im frühen 1. Jahrtausend vor Christus. Mit einem archäologischen Beitrag von I. Gerlach und M. Schnelle. (Epigraphische Forschungen auf der Arabischen Halbinsel Bd. 7). Tübingen: Wasmuth Verlag. Robin, C.J. 1994. „Yashhurʾil Yuharʿish, fils d’Abiyasaʿ, Mukarrib du Ḥaḍramawt.“Raydān 6: 101–111. Robin, C.J. 2012. „Matériaux pour une typologie des divinités arabiques et de leurs représentations.“ In Dieux et déesses d’Arabie. Images et représentations. Actes de la table ronde tenue au Collège de France (Paris) les 1er et 2 octobre 2007, edited by I. Sachet, C.J. Robin, 7–118. Paris: DeBoccard. Robin, C.J., Breton, J.-F. 1982. „Le sanctuaire préislamique du Ǧabal al-Lawḏ (NordYémen).“ Académie des Inscriptions et Belles-Lettres. Comptes Rendus des séances de l’année 1982, juillet–octobre, 590–629. Sab. Dict.: A.F.L. Beeston, M.A. Ghul, W.W. Müller, and J. Ryckmans. 1982. Sabaic Dictionary (English–French–Arabic). (Publication of the University of Sanaa, yar). Louvain-la-Neuve/Beyrouth: Peeters. Stein, P. 2003. Untersuchungen zur Phonologie und Morphologie des Sabäischen. (Epigraphische Forschungen auf der Arabischen Halbinsel Bd. 3). Rahden/Westf: vml Vlg Marie Leidorf.

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

abb. 1

Nordwestfassade des ʾAlmaqah-Tempels: c 366 (1) m. schnelle, dai

abb. 2

Nordwestfassade des ʾAlmaqah-Tempels: c 366 (2) m. schnelle, dai

475

476

nebes

abb. 3

Nordwestfassade des ʾAlmaqah-Tempels: c 366 (3) m. schnelle, dai

abb. 4

Muḥāfaẓat Mārib 11 b. vogt, dai

der rituelle umzug des yadaʿʾil ḏarīḥ nach ṣirwāḥ

abb. 5

Muḥāfaẓat Mārib 11 b. vogt, dai

abb. 6

Muḥāfaẓat Mārib 20/a m. kramer, dai

abb. 7

Muḥāfaẓat Mārib 20/b m. kramer, dai

477

478

nebes

abb. 8

Muḥāfaẓat Mārib 20/c m. kramer, dai

abb. 9

Muḥāfaẓat Mārib 20/d m. kramer, dai

chapter 24

Sedentism of Arabs in the 8th–4th Centuries bc Israel Ephʿal

The term ‘Arab(s)’ occurs in ancient Near Eastern sources from the mid-ninth century bc onwards. Its precise meaning and etymology are obscure and require further examination.1 In Akkadian, Hebrew and Aramaic sources of the 9th–5th centuries bc, this term usually refers to nomads who raised camels and sheep, dwelled in tents and unfortified temporary camps, and lived in the deserts of northern Sinai, North Arabia and the Syro-Arabian Desert. These groups came in touch with the sedentary populations of the Fertile Crescent. The groups to which the term ‘Arabs’ is applied in the sources of the 8th– 7th centuries bc are the Qedarites, the people of Sumu-AN/il, Idibilu (biblical Adbeel), Thamūd, Ibadidi, Marsimani, Ephah and Mibsam.2 We do not know whether the inhabitants of the North Arabian oases (the largest of which were Taymāʾ, Dadan = al-ʿUlā and Dūmah = Dumat al-Jandal, al-Jawf) were considered ‘Arabs’ in ancient times. Our sources reflect the relations of the ‘Arabs’ with the settled frontier lands and the political bodies in the cup of the Fertile Crescent, on both sides of the annual average 200mm isohyet. They show evidence of the seasonal penetration of nomads into the settled lands, particularly in times of scarcity of rain. This penetration sometimes took place peacefully (though under supervision)3 and sometimes by force,4 and it generally ended with the nomads’ withdrawal

1 However, statements such as “The term ‘Arab’ originally designated a community which was the special property of a semi-divine hero, leading them as a kind of police force during certain renewal festivals” (Retsö 2003: 624) are not supported by any real evidence. 2 See Ephʿal 1982: 5 notes 3–5. According to Esarhaddon’s Qaqun inscription (rinap 4 102), the Assyrian army in its campaign against Egypt in 671bc was supported by the people of Mibsam that provided camels loaded with water skins for the army on the march in the Sinai Desert (the inscription is to be published by Elnathan Weissert, to whom I am indebted for providing me with a reconstructed text). In another inscription of Esarhaddon, the owners of these camels are described as Arabs (see rinap 4 34 rev. 1–2). In the biblical genealogies Mibsam appears as a son of Ishmael (Genesis 25: 13; iChronicles 1: 29). 3 For example Numbers 21: 21–22; 20: 17; cf. saa i 82; 177 rev. 7–9; 178. 4 For example Judges 6: 3–5, 11.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

480

ephʿal

into the desert. In several cases, however, the Arab penetration did not end with their withdrawal but with their settling in the settled areas, as is discussed below.

a

Babylonia, 8th Century bc

The first military campaign of Sennacherib, king of Assyria, was conducted in 703bc against the Chaldeans of Babylonia led by Merodach-baladan. A full account of this campaign is recorded in cylinder inscriptions discovered at Nineveh, Assur and Tarbiṣu (rinap 4 no. 1 and 213). The inscriptions contain a list of 81–86 walled towns (ālāni bît dūrāni) in southern Babylonia conquered and looted by the Assyrian army along with hundreds of minor settlements within their borders. The names of some of these towns—Dur-Abiyataʾ in the territory of Bit Dakkuri (1: 37; 213: 37), and Dur-Uayyit and Dur-Birdada in the territory of Bit Amukani (1: 43–44; 213: 44)—indicate the ethnic origin of their inhabitants: The personal names that appear as components of these toponyms, after the prefix Dur, are known to be the names of Arab leaders mentioned in the inscriptions of Ashurbanipal, king of Assyria. They are as follows:5 – Uayteʾ (Arabic wyṯʿ) son of Hazael king of the Arabs (Letter to the God i 3, ii 56, iii 33); – Uayteʾ king of the Arabs (a vii 83, 123; viii 25, 93); – Uayteʾ son of Birdada king of the Arabs (a viii 1–2; Letter to the God iv 2–3);

5 The inscriptions of Ashurbanipal are marked here according to biwa. On the Letter to the God see biwa pp. 76–82. The proper names in this article are spelled according to three writing systems: Akkadian, Aramaic and Arabic. The following table presents the correspondence between relevant consonants in these systems:

Akkadian

Aramaic

Arabic

t p ḫ ʾ

t p ḥ ʾ, ʿ

t, ṯ f ḥ, ḫ ʾ,ʿ

sedentism of arabs in the 8th–4th centuries bc

481

– Abiyataʾ (Arabic ʾbyṯʿ) son of Teʾri the Qedarite (a ix 16–17; Letter to the God iv 2–4);6 – cf. also Uayteʾ king of Sumu-AN/il (iit 113; line 119 of this inscription has Yauteʾ [Arabic ywṯʿ] king of Sumu-AN/il). The list also includes the town of Qidrina in the territory of Bit Dakkuri. The above-mentioned toponyms indicate the presence of permanent Arab settlements in western Babylonia in the second half of the eighth century bc. The existence of walled towns with Arabic names, which were centres for unwalled hamlets in their environs, not only reflects the extent and intensity of the Arab penetration into Babylonia, but also suggests that this development was the result of a prolonged process. Even if we assume that these towns were neither founded nor built by the Arabs but came under their control and in time underwent a name change, we have to suppose that the process of Arab penetration and settlement in Babylonia began at least some decades before Sennacherib’s first campaign.

b

Babylonia, 6th–5th Centuries bc

Toponyms and references in legal and economic documents indicate the existence in Babylonia, in the 6th–5th centuries bc, of small rural settlements with a clear connection to Arabs.7 In the Bit Murašû archive from Nippur (454– 404bc), we find: – lúšaknu ša lú Arbaia pbs 2/1 48;8 – ḫaṭru ša lú Arbaia pbs 2/1 48;9 – cf. also uruArbā (Ar-ba-a) tmh 2/3 147.

6 On the Arab leader Abiyateʾ son of Teʾri, without the designation “the Qedarite”, see also a vii 97, viii 31; b viii 32. 7 This section, like the previous one, contains only toponyms, not personal names. On people in Babylonia described as “pn the Arab” or whose personal names contain elements that may derive from Arabic, see Zadok 1981: 45–54. 8 The name of this official is not stated in this document. On the separation here of MušezibBêl from the lúšaknu ša lú Arbaia, see Stolper 1985: 85–86. 9 On the term ḫaṭru that covers rural settlements that were composed of groups of a common ethnic or professional background and were subject to the land-for-service system, see Jursa 2010: 247, 251; and Stolper 1985, esp. p. 25, 70–72, 97–100.

482

ephʿal

In the Nippur region but not in the Bit Murašû archive: – Ālu- ša- lú Arbaia be 8/1 26 (563bc), 50 (547bc);10 – Eqel arkat lú Arabaia (= domain of the Arabs) (496 bc);11 – cf. also uruQidari be 8/1 65 (534bc).

c

Idumea, 4th Century bc

In the last 25 years, about 2000 Aramaic ostraca dated 362–301 bc and originating in Idumea have come to light (about half of them have been published so far).12 They contain about 600 personal names, among which about sixty-eight are typical theophoric Edomite (all of them contain the component Qaws, qws—the name of the Edomite national god). They bear witness to the presence of people of Edomite origin in the southern Shephelah and the Beer-sheva Valley before the Hellenistic period. It appears that when the time came, at the beginning of the Hellenistic period (or perhaps earlier), to define, for administrative purposes, the region known in the Hellenistic and Roman periods as Idumea, it was done according to the ethnic makeup of the population that had already crystallised in the region in the Persian period. Beside the Edomite personal names one finds in the Idumea ostraca, there was another group of personal names that can be defined as ‘Arabic’ on the basis of their linguistic features. These are: 1.

Names deriving from roots that occur in the ancient Arabic onomasticon and are not found in Aramaic, Hebrew and Phoenician (such as whb, rather than yhb;13 and ytʿ < yṯʿ, rather than yšʿ).14 From this distinction, it

10

The toponym Ālu- ša-mAbi-ilaia occurs in be 8/1 51, 68, 72; tmh 2/3 90 from the years 547– 531 bc. On the possibility that it is a variant of Ālu- ša- Arbaia see Ephʿal 1982: 189–190. Joannès and Lemaire 1996: 48. For details of the collections of Idumean ostraca and the relevant publications, see tao, vol. 1: xv–xxii. On the origins of Idumea see Ephʿal 2003. ʾlwhb, whbʾl, whbw, whby and qwswhb in the Idumea ostraca. On Arabic names with the root whb, see Harding 1971: 651–653. Cf. also mUabu (= whb) who “incited all the Arabs” to rebel against Yataʾ/Yautaʾ (= yṯʿ/ywṯʿ) son of Hazael king of the Arabs (rinap 4 no. 1 iv 23–24; no. 31 rev. 7–8; no. 97: 16). ytwʿ, ytyʾl, ytyʿw, lytʿ, ʿmytʿ, qwsytʿ and qwslytʿ in the Idumea ostraca. The corresponding forms of the Arabic yṯʿ are Aramaic ysʿ and Hebrew yšʿ; see Ephʿal 1973; on the Aramaic form cf. hdysʿy in the Tell Fekheriye inscription. For Arabic names with the yṯʿ root, see Harding 1971: 658–659.

11 12 13

14

sedentism of arabs in the 8th–4th centuries bc

2. 3. 4. 5.

483

appears that the name qwswhb, containing both the Edomite theophoric component qws and the typical Arabic component whb, can be classified as Arabic. Personal names in the quṭaylu form, such as šʿydw, ʿbydw, nhyrw, zbydw. Personal names with an -w ending, such as ʿzyzw, ʿbdw, nmrw, mlkw, ḥlfw, zydw. Personal names with a -t ending, such as yʿft, ḥlft. Personal names with a -n ending, such as ʿdrn, mṭrn, ḥlfn, zydn.

I am aware that we may have a methodological problem here. While the differentiation of the Arabs in sections a–b rests on their explicit designation as ‘Arabs’, i.e., on anthropological criteria, here it rests on linguistic ones. I am also aware of the jubilarian’s reluctance to draw sweeping historical conclusions from the study of personal names.15 I must say, however, that in the present stage of our knowledge I cannot find a better explanation for the Idumean personal names that were given here.

Conclusion The data given in this article reflect the final stages of processes whose beginnings and stages of development are unknown. The evidence regarding these data indicates that we are not dealing with a unique incident but rather with a repeated phenomenon. All the data regarding the ‘Arabs’ under discussion relate to sedentary populations. This leads us to examine the validity of the usual definition ‘Arab = nomad, bedouin’. In view of this problem, I would like to make the following remarks: 1.

2.

15 16

One wonders whether nomadism was the only feature characterising the ancient Arabs. Could they not have had other common features, such as language, origin, religion?16 In certain cases, such as those dealt with above, the term ‘Arab(s)’ may indicate descent from Arabs rather than the lifestyle of those who bear this designation. An analogous phenomenon is the use of the term Yehû-

See Macdonald 1999: 253–256. This point has already been raised by Livingstone 1989. The present article provides additional data to support such a possibility.

484

ephʿal

dî(m), already in biblical times. While in the seventh and sixth centuries bc this term was applied to inhabitants of Judah, in the Persian period it was used for the Jews in Elephantine (Egypt) and for the heroes of the book of Esther whose ancestors had not lived in Palestine for several generations.

Appendix: On the Identification of Geshem the Arab and the ‘Arabs’ in the Book of Nehemiah Nehemiah’s three adversaries when the wall of Jerusalem was constructed were Geshem the Arab (Neh. 2: 19; 6: 1–2, 6), Sanballat the Horonite, and Tobiah “the Ammonite servant”. Extra-biblical sources indicate that Sanballat was governor of (the province of) Samaria, and Tobiah was a person of high rank in Transjordan and influential in Judah, suggesting a status with the Persian authorities and an influence in Palestine like that of Nehemiah. Similar qualities can reasonably be attributed to Geshem the Arab. His influence is attested by Sanballat’s ‘open letter’ to Nehemiah, which begins “Word has reached the nations, and Geshem (gšmw) too says that you and the Jews are planning to rebel, etc.” (Neh. 6: 6). Only a person solidly established in the Persian administration could make such a charge against Nehemiah, whose position in Judah was authorised by the Persian court. The identification of Geshem the Arab very much depends on understanding the nature of his rivalry with Nehemiah. While we know that Sanballat and Tobiah were involved with the Temple and with the Jerusalem leadership and were hostile to Nehemiah, probably due to his isolationist policy, we have no hint of a link between Geshem, the Temple and Jerusalem. The potential friction between him and Nehemiah can therefore be ascribed to economic-administrative factors, not to Judah’s internal situation. This has led to the assumption that Geshem was involved in the long-range North Arabian trade and was threatened by an over-powerful Judah. This viewpoint led to the proposed identification of the Geshem referred to in two inscriptions of the Achaemenid period with Nehemiah’s rival, Geshem the Arab. 1.

The date-formula in JSLih 349 from Dadan (al-ʿUlā) reads: “In the time of Gšm b. Šhr and ʿbd governor of Dadan ( fḥt ddn), in the reig[n of] … (brʾ[y] …).” The use of Gšm b. Šhr’s name as a time-indicator points to a prominent person, of high rank in the Dadan region, and possibly beyond. With regard to his title, however, we know only that he was not governor of Dadan.

sedentism of arabs in the 8th–4th centuries bc

figure 1

2.

485

Map showing the location of the sites mentioned in the text

One of the silver bowls that, most probably, were discovered at Tell alMaskhūṭah in Wadi Tumilāt, has an Aramaic inscription zy qynw br gšm mlk qdr qrb lhnʾlt, “That which Qaynū son of Gešem, king of Qedar, offered to han-ʾIlāt.”17

These identifications, though chronologically feasible, are tentative. In our present state of knowledge, we must exercise caution, especially regarding the opinion that Geshem, Nehemiah’s rival, the king of Qedar and the leader men-

17

See Rabinowitz 1956; Dumbrell 1971.

486

ephʿal

tioned in JSLih 349, are the same person. This opinion depends on the assumption that 1/ the Qedarites controlled Dadan and its environs in the 5th century bc (if that was the case, why was the title “king of Qedar” not appended to the name of Gšm b. Šhr in JSLih 349?), and that 2/ there existed, under Achaemenid rule, an organised, politically influential economic ‘empire’ of huge proportions, extending from Dadan to southern Judah and the approaches to Egypt. Considering, however, the size and geographical location of the province of Judah, it is difficult to envisage that the rivalry between Geshem and Nehemiah was about control of the routes of Arabian trade, as Judah’s borders never extended to these routes and thus Judah was not a competitor and could not endanger them. In any case, there is no written or archaeological evidence of Judah’s involvement in such activities. The tracing of dozens of Arabic names in the Idumea ostraca provides a more feasible possibility of identifying Geshem the Arab: In Neh. 4: 1, we read: “When Sanballat and Tobiah and the Arabs, the Ammonites and the Ashdodites heard that healing had come to the walls of Jerusalem, and that the breached parts had begun to be filled, it made them very angry.” This reaction can be easily explained, in view of the above discussion, by the need to prevent any damage to the interests of Judah’s neighbours rather than to those of the Arabs in northern Arabia or the approaches to Egypt. The distinction of the Arabic names in the Idumea ostraca supports the location of their owners in the Shephelah, west of the province of Judah, and helps to link them to the ‘Arabs’ of the book of Nehemiah. The location of Geshem the Arab in that region gains additional support from his and Sanballat’s invitation to Nehemiah: “Come, let us get together in Kephirim in the Ono Valley”, i.e., in the northern Shephelah. In the light of these data, though lacking explicit evidence, one may connect Geshem the Arab with Idumea and regard him as a South Palestinian local leader (governor? ‘king’?), and see the rivalry between him and Nehemiah as the result of bad neighbourly relations such as a struggle over land and local interests.18 Following this analysis, we may conclude that the author of the Book of Nehemiah already identified the ‘Arabs’ as inhabitants of Idumea.

18

Additional arguments for identifying Geshem the Arab with a North Arabian figure (rather than with an Idumean one, as suggested here), see Graf 2015: 294–296. Graf relates the Arab presence in 4th century bc southern Palestine to the Arabian incense trade. One should notice, however, that none of the Idumea ostraca bears witness to such an activity.

sedentism of arabs in the 8th–4th centuries bc

487

Sigla be 8/1 biwa JSLih pbs 2/1 rinap 4 saa i tao tmh 2/3

Clay 1908. Borger 1996. Dadanitic inscriptions published in Jaussen and Savignac 1909–1922. Clay 1912. Leichty 2011. Parpola 1987. Porten and Yardeni 2014. Krückmann 1933.

References Borger, R. 1996. Beiträge zum Inschriftenwerk Assurbanipals. Wiesbaden: Harrassowitz. Clay, A.T. 1908. Legal and Commercial Transactions, Dated in the Assyrian, Neo-Babylonian and Persian Periods Chiefly from Nippur. Philadelphia: University of Pennsylvania. Clay, A.T. 1912. Business Documents of Murashu Sons of Nippur Dated in the Reign of Darius ii. Philadelphia: University Museum. Dumbrell, W.J. 1971. “The Tell el-Maskhuṭa Bowls and the ‘Kingdom’ of Qedar in the Persian Period.” Bulletin of the American Schools of Oriental Research 203: 33–44. Ephʿal, I. 1973. “On the Identification of the Israelite Exiles in the Assyrian Empire.” In Excavations and Studies: Essays in Honour of Prof. Sh. Yeivin, edited by Y. Aharoni, 201–203. Tel Aviv: Tel Aviv University. [in Hebrew]. Ephʿal, I. 1982. The Ancient Arabs. Jerusalem: Magnes Press; Leiden: Brill. Ephʿal, I. 2003. “The Origins of Idumaea.” Qadmoniot 36: 77–79. [in Hebrew] Graf, D. 2015. Arabs in Palestine from the Neo-Assyrian to the Persian Periods, aram 27: 283–299. Harding, G.L. 1971. An Index and Concordance of Pre-Islamic Arabian Names and Inscriptions. Toronto: University of Toronto Press. Jaussen, A., and R. Savignac. 1909–1922. Mission archéologique en Arabie. Paris: E. Leroux-P. Geuthner. Joannès, F., and A. Lemaire. 1996. “Contrats babyloniens d’époque achéménide du Bît-abî-râm avec une épigraphe araméenne.” Revue d’assyriologie et d’archéologie orientale 90: 41–60. Jursa, M. 2010. Aspects of Economic History of Babylonia in the First Millennium bc. Economic Geography, Economic Mentalities, Agriculture, the Use of Money and the Problem of Economic Growth. (Veröffentlichungen zur Wirtschaftsgeschichte Babyloniens im 1. Jahrtausend v.Chr. 4). Münster: Ugarit-Verlag.

488

ephʿal

Krückmann, O. 1933. Neubabylonische Rechts- und Verwaltungstexte, autographiert und mit inventarverzeichnis und namenlisten versehen. Leipzig: J.C. Hinrichs. Leichty, E. 2011. The Royal Inscriptions of Esarhaddon, King of Assyria (680–669bc), Winona Lake, in: Eisenbrauns. Livingstone, A. 1989. “Arabians in Babylonia/Babylonians in Arabia: Some Reflections à propos New and Old Evidence.” In L’Arabie préislamique et son environnement historique et culturel. Actes du colloque de Strasbourg, 24–27 juin 1987, Université des Sciences Humaines de Strasbourg, edited by T. Fahd, 97–105. (Travaux du Centre de Recherche sur le Proche-Orient et la Grèce Antiques 10). Leiden: Brill. Macdonald, M.C.A. 1999. “Personal Names in the Nabataean Realm. A Review Article.” Journal of Semitic Studies 44: 251–289. Naveh, J. 1979. “The Aramaic Ostraca from Tel Beer-sheba (Seasons 1971–1976).” Tel Aviv 6: 182–195. Parpola, S. 1987. The Correspondence of Sargon ii: Letters from Assyria and the West. Helsinki: Helsinki University Press. Porten, B., and A. Yardeni. 2014. Textbook of Aramaic Ostraca from Idumea. Vol. 1. Dossiers 1–10: 401 Commodity Chits. Winona Lake, in: Eisenbrauns. Rabinowitz, I. 1956. “Aramaic Inscriptions of the Fifth Century bce from a North-Arab Shrine in Egypt.” Journal of Near Eastern Studies 15: 1–9. Retsö, J. 2003. The Arabs in Antiquity. London: Routledge. Stolper, M.W. 1985. Entrepreneurs and Empire: The Murašû Archive, the Murašû Firm, and Persian Rule in Babylonia. Istanbul: Nederlands Historisch-Archaeologisch Instituut te Istanbul. Yardeni, A. 2014. “Concordance of Aramaic Ostraca from Idumea.” In a cd attached to tao. Zadok, R. 1981. “Arabians in Mesopotamia during the Late-Assyrian, Chaldean, Achaemenian and Hellenistic Periods Chiefly According to the Cuneiform Sources.” Zeitschrift der deutschen morgenländischen Gesellschaft 131: 42–84.

chapter 25

Reflections on Arab Leadership in Late Antiquity* Greg Fisher

1

Writing the History of Arab Leaders

Information about Arab leaders in late antiquity is provided by a variety of sources. These include Greek, Latin, and Syriac texts, contemporary to the leaders themselves and produced mostly by writers in the Roman empire. There are inscriptions in a variety of languages and scripts, together with data provided by modern archaeological studies of sites scattered throughout the Middle East. There is also a group of Syriac, Arabic, and Persian texts produced after the Muslim conquests.1 With a few exceptions, this disparate selection of source material provides the reader with the viewpoint of the state2—and principally, for contemporary sources, the viewpoint of those in the Roman empire. The Roman sources are so dominant that it has proven challenging to write a history sympathetic to the ‘Arab’ perspective,3 or even to put together a coherent account of Arabs in the Persian empire.4 Adopting a comparative approach is one way to deal with this difficult group of sources: thinking about the Arabs in terms of other frontier peoples (such as those in late antique Egypt or western Europe) has proven useful,5 as has the application of anthropological theory to create a framework for historical enquiry.6 This essay builds on these comparative efforts to provide a consideration of Arab leadership in the late antique period. It examines a number of individuals

* I am most grateful to Elizabeth DePalma Digeser, Philip Salzman, and Philip Wood for their comments and criticisms of the arguments laid out in this paper. 1 For a recent, comprehensive view of the sources see Fisher 2015. 2 States are also understood here from a structural perspective as political entities dependent on institutions not normally found in tribes, such as a bureaucracy, and where power is vested in a ‘recognized authority claiming legitimate and exclusive power’ (Hourani 1990, 306). See too Salzman 2014, 83–85, and Gellner 1983, 445. 3 The most ambitious effort is that of Irfan Shahîd, beginning with Shahîd 1984. 4 See now Fisher and Wood 2016. 5 Fisher 2011 and Dijkstra and Fisher 2014. 6 Dijkstra and Fisher 2014; Genequand and Robin 2015; Haldon and Conrad 2004; Szuchman 2008; Leder and Streck 2005; see also Whittow 1999.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

490

fisher

active between c. 350 and 600ce as allies of the Roman and Persian empires, described by our sources as ‘Roman Arabs’ or ‘Persian Arabs’,7 and places them within a framework derived from anthropological studies of tribal8 leaders and against the backdrop provided by the careers of certain western Germanic ‘barbarian’ chiefs in late antique Europe.9 The careers of ancient Arab leaders are virtually unknown outside their interactions with the states and empires of antiquity—the Hellenistic kingdoms, the Parthians and Sasanians, the Republic and empire, and the kingdoms of South Arabia. Writing about Arab leadership thus means that the relationship between state and tribe cannot be avoided. While there is a plethora of anthropological research volumes that examine the way that these two political entities interacted, only rarely do they consider the ancient world in any detail.10 In part this is because studies of tribal societies have usually depended on observation and analysis of living subjects. Aside from the impossibility of carrying out such an enterprise for antiquity, it has also proven challenging to use modern studies retrospectively to theorise about the past. There is no guarantee that ancient tribes behaved anything like their modern counterparts, and it is well known that circumstances of observation and study between the 7

8

9 10

These terms are used here in preference to Ghassānid/Jafnid or Lakhmid/ Naṣrid, to reflect the perception of the ancient sources, whose focus was on individual leaders rather than groups of people. For discussion of the terminological issues see Fisher 2015, 6–8, and Hoyland 2014. The literature on this troublesome term is vast: for the purposes of this essay, the term ‘tribe’ is understood in structural terms, referring to an entity that may be decentralized and non-hierarchical, where individual members may “assume great freedom” (Salzman 2014, 82), or as a “set of ideas about how people think about themselves as a series of social, economic and political groupings that provide livelihood and profits, and the development and defence of these” (Lancaster and Lancaster 2004, 53). See Fisher 2015, 5–6; Salzman 2014, 83–85; Szuchman 2008; Lancaster and Lancaster 2015, 53–56, 67– 68; Donner 1989, 80. It is a common misconception that Arab tribes were all nomads: for a classic exposé, see Shaw 1982, and for a recent and effective counter see Michael Macdonald in Fisher 2015, ch. 1. There is insufficient evidence to determine whether the Arab leaders discussed in this article were nomads, settled, or somewhere in between. See further, Fisher 2011, ch. 3., and see also extensive discussions in Salzman 1980, Ginat and Khazanov 1998, and Khazanov 1994, to cite just three useful sources on this difficult topic. See also Wickham 2005: 305–306. This term is used here for the sake of familiarity and convenience, to denote non-Romans on the periphery of the empire. See now Maas 2012. Szuchman 2008 and Leder and Streck 2005 are conspicuous exceptions. See for general studies of state/tribe relations with some ancient content Khoury and Kostiner 1990; Tapper 1983; Mundy and Musallam 2000.

reflections on arab leadership in late antiquity

491

two periods are very different. For example, years ago Rudi Lindner showed how modern scholars, influenced by the ineffectual impact of tribal military power in modern studies, dismissed the military strength of ancient tribes. In antiquity tribal raiders easily evaded imperial troops during times of sanction or punishment, but in more recent times militant tribesmen easily fell prey to the “modern state and its technology of administration.”11 Aircraft and mechanised transport could locate groups of people who previously had the ability to travel beyond the reach of state authorities, while the chief could be “closely tied to the whim of the government.”12 Lindner also pointed to the permitting and granting process, which could sometimes be used by contemporary state authorities to steer researchers towards those situations deemed ‘suitable’ for scholarly study.13 No such process existed in antiquity, where writers such as Ammianus or Procopius described tribal activities in a variety of situations, including times of open and savage conflict between the tribe and the state. These writers were not anthropologists in the ‘modern’ sense, and placed their subjects within a range of ethnographic and literary frameworks specific to their worldview and time period. Despite these drawbacks, modern studies can provide a valuable interpretative tool for the often sparse, biased, incomplete, or cursory ancient sources.14 Modern accounts of tribal leadership, for example, those of the Rwala of Jordan or the Qashqaʾi of Iran, or analyses of older ethnographic accounts by travellers in the Ottoman-controlled Middle East such as Musil, Doughty, or Burkhardt, relate a variety of attributes and skills which find parallels in ancient texts. Those who led became chiefs based on their charisma, confidence, and strength, access to information, a range of skills or attributes, and their “known

11

12 13

14

For example, as reported by Malalas, 18.32 (p. 445) or Evagrius, 5.20. See Lindner 1982, 691–696, and Salzman 2008. As a counter, see Kostiner 1990, 226. See too Khazanov 1994, 222–223. Lindner 1982, 696. Lindner 1982, 691: “Consider for a moment how the ethnographer begins a field study: the first step (after getting a grant) is to obtain a research permit from the appropriate government authorities. This bureaucratic process is an aid to researchers, for it prevents their stumbling upon a field of local strife, and it also helps the government to screen its guests for “inappropriate” behavior. The import of this procedure for our analysis lies in the fact that no government will grant a researcher permission to live and work among a tribe that is not under its effective control.” See Lancaster and Lancaster 2004, 29: “some appreciation of the constructions of bedouin identities, and the moral premises through which bedouin organized economic and political activities, does allow a more informed reading [of ancient situations].”

492

fisher

reputation” and prestige.15 “Public opinion”, the consent of the tribe—whose own egalitarian moral premise conferred a sense of “individual autonomy”— kept chiefs in their positions.16 Tribal leadership was in addition stimulated and strengthened, and in some circumstances even ‘created’, by external factors, such as political pressure caused by the interaction between tribe and the state.17 Reputation and prestige, which were vital for successful leadership, could be conferred through a number of interrelated skills, attributes, and virtues, including: – an ability to mediate, negotiate, and solve problems18 – moral values, such as providing advice, hospitality, and benefaction19 – adapting to circumstance and protecting the tribe20 Several of these characteristics are applicable not only to Arab chiefs, as I will explore below, but to other leaders in the late antique world, such as the Ostrogothic king, Theodoric, or the Frankish leader, Clovis. Modern accounts of Theodoric’s rule draw a clear distinction between his self-presentation to Goths, and the rather different way in which he portrayed himself to the Italians following his defeat of Odoacer in 493. For the Italians, he chose a conservative approach, defending Roman laws and customs and upholding the traditions of the state to bolster his ‘Roman’ credentials; elite members of the Italian population responded in kind, calling him princeps and in at least one case, augustus.21 Yet for the Goths he possessed a “resolutely Gothic identity” as a rex or dux, a war leader of great reputation, who won and retained the loyalty of his people through performing feats of valour on the battlefield, and protecting them 15

16

17 18 19 20 21

Lancaster and Lancaster 2004, 34–35, 39, and 2015, 58–59; Salzman 2004a, 319; Salzman 2004b, 77–101; Fisher 2014, 282–286; Lancaster 1997, 43ff.; Garthwaite 2009, 39; cf. Grunebaum 1963, 11. Lancaster and Lancaster 2004, 30; Salzman 2004a, 301, 307–308; Salzman 2004b, 77–101, reflecting in particular on Barth 1961; Beck 1986, 200–234; Fisher 2014, 282; Gellner 1990, 111; Grunebaum 1963, 11. Salzman 1983; Strathern 1983, 451; Hourani 1990, 305. See too Caton 1990, 102–103; Paul 2003, 33; Beck 1986. Lancaster and Lancaster 2004, 34–35, and 2015, 64; Lancaster 1997, 87; Salzman 2004b, 90– 91; Beck 1986; Salzman 2004a, 301; Garthwaite 2009, 37. Salzman 2004a, 79; Lancaster and Lancaster 2004, 34, 41, and 2015, 58, 64; Tapper 1983, 56; Beck 1986, 214–215; Garthwaite 2009, 38–39. Lancaster and Lancaster 2004, 34, 39, and 2015, 71; Caton 1990, 102; Salzman 1974; Salzman 2004a, 319–325. Sarris 2011, 106.

reflections on arab leadership in late antiquity

493

from outside threats.22 His acclamation as king by the Ostrogothic army (and not by the Roman emperor) demonstrated the importance of tribal consent.23 Elsewhere, the Frankish king, Clovis, is portrayed at St. Martin’s Church as a benefactor, distributing largesse to his followers and using such acts to build and maintain his standing.24 Clovis’ famous appearance as a ‘new Constantine’25 in the text of Gregory of Tours parallels Theodoric’s deliberate approach to dealing with the Roman senate, smoothing interactions with a population long familiar with the iconography and practise of imperial rule. On the other hand, the careful performance of Clovis outside St. Martin’s Church can be interpreted as an effort to meet the expectations of a rather different group of people, accustomed to conferring respect and status on generous benefactors. And on the battlefield, Clovis, like Theodoric, preferred to be seen a fearsome adversary, fighting Romans, Alamanni, factions amongst the Burgundians, the Visigoths, and various internal rivals, successfully spreading Frankish power.26 Generosity, military success, and the ability to appease different groups of people were all ways in which tribal leaders became men of reputation. In the west, men like Theodoric and Clovis laid the foundations for kingdoms that played an important role conferring the political and cultural inheritance of Rome on medieval Europe. They also helped to transmit a distinctive model of kingship into the Middle Ages, one that spoke to the expectations of the different traditions comingled in the late antique west. In the east, the Arabs built complex relationships with the Romans and Persians that presented some similarities with the state/tribe interactions of their Germanic counterparts, specifically in the centralisation of power around individual leaders.27 In the west, this process became one of state formation that led to the emergence of several discrete kingdoms.28 Even if the Arab leaders of late antiquity fell prey to the dominant strength of their imperial patrons, a brief examination of their history, set against the main points itemised above, shows how strong Arab leaders—militant warriors, generous benefactors, effective mediators, men of reputation who developed links with the divine—flourished as allies of

22 23 24 25 26 27 28

Procopius, 5.1.25; Heather 1995; Sarris 2011, 107. The image of the powerful warrior is provided (for example) by Cassiodorus, 1.24. Sarris 2011, 102. Gregory of Tours, 2.38. Gregory of Tours, 2.31. Sarris 2011, 121; James 1988, 78–91. Fisher 2011, chs. 3 and 5. See Khazanov 1994; al-Rasheed 1987; “The Process of Chiefdom-Formation”; Tapper 1990, 49, 64–66.

494

fisher

Rome and Persia. The way that such figures developed within the framework of imperial power highlights points of comparison between the ‘Arab east’ and the ‘Germanic west’, and shows how one of the most recognisable features of the early Islamic world—strong Arab leadership—had its roots in the RomanoPersian Middle East.

2

Mediators and Negotiators

The ability to negotiate effectively and to solve problems between different groups is a key attribute of successful tribal leaders. Not only does it help to foster social peace within the tribe, it also reinforces the chief’s authority and prestige; failure, on the other hand, can have dramatic consequences. The skill of mediation appears in a range of guises in ancient sources, but is usually visible where tribal leaders bartered with the state for funding, support, protection from others, and for the state’s use of its supply of manpower, an extremely valuable commodity.29 While tribal chiefs could, from time to time, choose to remind the state of their military power, applying a proven ability to mediate may have ended up being a more profitable way of dealing with state authorities.30 Ecclesiastical historians portrayed Queen Mavia’s revolt in the spring of 377 as a struggle between an Arian emperor (Valens) and a Nicene orthodox Christian, Mavia, desperate to protect her people from views “opposed to the faith of Christ”.31 After her husband died, the treaty with Rome was cancelled, and Mavia went to war. In the most detailed account, that of Sozomen, it is not clear exactly why she took this action, but clues are to be found in the way that the war ended. Mavia’s campaign was very successful, claiming the dignity (and nearly the life) of the magister militum per Orientem. A Roman embassy was rebuffed with instructions that Mavia would only consent to peace if she received a Nicene bishop, Moses, without any interference from the Arian patriarch of Alexandria, Lucius. It seems likely that the question of the bishop predated peace negotiations, and that Mavia had gone to war to protect the tribe’s desire for a bishop of their choice. In any event, Mavia ultimately obtained Moses, and she had done so by leveraging her most precious resource—her 29 30 31

As Gellner 1990, 113, notes, this resource often “arrived, fully trained … with recognized leaders and a familiarity with the terrain in which they were to be deployed”. Cf. Salzman 2004a, 319, discussed in Fisher 2014, 284. Sozomen, 6.38: from the speech of Moses, Mavia’s choice for bishop, in response to the Arian Lucius, patriarch of Alexandria.

reflections on arab leadership in late antiquity

495

troops. She later supplied soldiers to defend Constantinople in the aftermath of the Gothic victory at Adrianople (378).32 Running under the surface of the passionately Christian sources that relate this sequence of events is an account of a negotiation: the manpower that Mavia controlled could be deployed for the benefit of the state, withheld, or used against it, depending on the circumstances. A new treaty between Rome and Mavia could function once the queen had decided to “live in peace” with her imperial sponsor; cooperation would be more effective than further militarized defiance. Mavia’s ability to adapt to the circumstances helped to ensure her success as a leader.33 Mavia’s revolt provides a useful example of how a tribal leader could adapt a skill at mediation to trade for a more attractive relationship with the state. A similar case is provided by the story, also told by Sozomen, of Zokomos, who approached a holy man in the Roman empire for help with his wife’s inability to conceive. Divine intercession was provided, a baby was born, and baptism quickly followed; Sozomen ends his narrative by stating that Zokomos and his people subsequently became fearsome military allies of the empire. Underneath the rhetorical and didactic elements of the story runs a parallel narrative found throughout the late antique Roman world, where Christianisation generated political advantage for Rome in the shape of military alliances and peace along the frontier. This is also a story of negotiation, where baptism and inclusion in the Roman commonwealth was traded for a commitment to the state.34 Descriptions of state/tribe negotiations are found disproportionately in texts dealing with religious issues. In part this is due to the fascination of ecclesiastical historians with the ability of Christianity to ‘civilise’, helping the barbarians cast off their savage past as “wolves” and obtain membership in the “rational flock of Christ”.35 The focus on Christian elements in stories about Arabs also reflects the conspicuous role assumed by Arab leaders as arbiters in the religious disputes of the state in the sixth century. Interestingly, Arabs carried out this function largely as laymen: Aspebetos, an early fifth-century émigré from the Persian empire, is the only Arab outsider reported by the sources to be consecrated; after arranging for his people to be settled in Roman

32 33

34 35

Socrates, 5.1. Cf. Lancaster and Lancaster 2015, 65; Irons 1979, 366, characterises this as “peaceful negotiation and adjudication backed up by the possibility of armed conflict.” See Lewis 2000 on the balance between resistance and collaboration in the context of the relationship between the Rwala and the Ottoman empire; see also Tapper 1990, 67, on this sort of adaptation. Sozomen, 6.38. Cyril of Scythopolis, Life of St. Euthymius, 15 (trans. Price).

496

fisher

Palestine, he took office, assumed the name Peter, and later appeared at the Council of Ephesus in 431.36 But others, staying aloof from church hierarchies, emerged as valued problem-solvers for the religious quarrels that plagued the Roman world. Al-Ḥārith (529–569) and his son al-Mundhir (569–582) are the best examples of this phenomenon, called upon by Miaphysite leaders, the emperor, and others to handle delicate negotiations: – al-Ḥārith playing a role (perhaps smaller than reported by John of Ephesus) to obtain Jacob Baradeus and Theodore for the Miaphysites (542)37 – al-Ḥārith mediating between the Miaphysite leadership and two Tritheist bishops, Conon and Eugenius (c. 569)38 – al-Ḥārith planning to convene a meeting in the province of Arabia to mediate the dispute between Tritheists and Miaphysites (c. 569)39 – al-Ḥārith in Constantinople, discussing the same problem (c. 569)40 – al-Mundhir attempting to solve a bitter dispute between different Miaphysite factions, using the ‘camp of the Ṭayyāyē’ (i.e. al-Mundhir’s camp) as neutral ground (578 and afterwards).41 – al-Mundhir invited to Constantinople by Tiberius ii to chair a meeting designed to solve the same dispute (March 580)42 – an Arab leader named Jafna, continuing mediation for this dispute at the church of St. Sergius at Gabitha/Jābiya (c. 590)43 Away from the arena of religious politics, we also find Arab mediators resolving other problems: – al-Ḥārith’s brother Abū Karib mediating a financial dispute, recorded in the Petra Papyri (c. 574)44 – Abū Jafna Nuʿmān b. al-Mundhir acting as a go-between for Maurice and Khusrau ii in the context of Persia’s civil war (c. 591)45

36 37 38 39 40 41 42 43 44 45

Cyril of Scythopolis, Life of St. Euthymius, 20. John of Ephesus, Life of Jacob and Theodore, 153–154. Chabot 1952, § 39 (following the numbering of Millar 2009). Chabot 1952, § 39. Chabot 1952, § 40. John of Ephesus, Ecclesiastical History 3.4.36/pp. 216–217. John of Ephesus, Ecclesiastical History 3.4.39/pp. 218–220. Michael the Syrian, 10.22/pp. 383–384. See Kaimio 2001. Chron. 1234, p. 215.

reflections on arab leadership in late antiquity

497

Sources for similar actions carried out by Arab leaders allied with Persia are not as forthcoming. However, the studied neutrality of the Persian Arab leader al-Mundhir (504–554)46—caught between calls for the persecution of Christians from the Ḥimyarite king Joseph and the stubbornness of Christians in his own retinue—can also be interpreted as a type of mediation. The Persian Arab leaders were careful not to antagonize the Christian community at alḤīra in Iraq, even while they carried out violent anti-Christian raids in Roman territory that targeted churches, convents, and monasteries.47 Successful mediation conferred prestige, but that status could be easily lost. One of the most difficult negotiations that the Roman Arabs faced involved the dispute between the troublesome patriarch of Alexandria, Damian (578–606) and the supporters of Paul ‘the Black’ (d. 578) and Paul’s successor in Antioch, Peter (581–591). Tiberius ii (574–582) called on al-Mundhir in 580 to solve this bitter quarrel, but after concluding a favorable agreement, he was apparently double-crossed by Damian and the settlement collapsed.48 (Damian would then go on to make trouble for the Jafna mentioned above, also undermining his efforts). Al-Mundhir’s prestige received a second blow when he was accused of treachery during a military expedition with the Roman general and future emperor Maurice, shortly before the latter assumed the throne. His reputation in tatters, al-Mundhir was arrested and exiled in 582.49

3

Advisors and Benefactors

Ancient literary texts and inscriptions provide a number of examples of Arab leaders dispensing guidance, benefaction, and patronage. Naturally some of the evidence for advising falls within the context of mediation, but it is also visible as part of the close relationship enjoyed by the Roman and Persian Arab leaders with their imperial patrons. Persian Arabs, in particular, are repeatedly depicted as members of the royal inner circle, guiding the king’s battle plans and offering advice. In the early sixth-century war between Anastasius and Kavadh, the Chronicle of Ps.-Joshua shows the Persian Arab leader al-Nuʿmān discussing the forthcoming attack on Edessa, in which he was killed.50 Kavadh 46 47 48 49 50

No relation to the Roman Arab of the same name. Fisher and Wood forthcoming. John of Ephesus, Ecclesiastical History 3.4.39–43/pp. 218–225. Theophylact Simocatta, 3.17.5–10; John of Ephesus, Ecclesiastical History 3.3.40–43/ pp. 173–178. Ps.-Joshua the Stylite, 58.

498

fisher

relied heavily on al-Nuʿmān’s successor al-Mundhir for guidance: Procopius says that the king came to rely on him when he was “completely at a loss as to what he should do” and “felt confidence in no-one else”, and trusted in his plans.51 The sources are more reticent about the advisory role played by Roman Arabs with individual emperors, although Maurice clearly trusted Abū Jafna Nuʿmān (above), Tiberius ii placed great stock in al-Mundhir’s abilities, and Justinian’s decision to place al-Ḥārith in control of the Arab allies of the state in 529 was guided by an assessment of his abilities.52 There is more detailed evidence for the role played by Arabs as benefactors, men of generosity and charity. For the people under their control they provided plunder, seized on raiding expeditions carried out independently or as part of the ongoing wars between Rome and Persia.53 Such missions provided funds for distribution to the poor, as well as to churches and monasteries, and constituted an important way for tribal chiefs not only to uphold the moral values of the tribe, but also to convert the wealth derived from interactions with the state into a benefit for others.54 The donation of wealth to religious institutions may help to explain the close links between the Roman Arab leaders and rural churches and monasteries, situated in a broad arc from the region south of modern Amman up to Sergiopolis/al-Ruṣāfa, near Raqqa; a number of texts found throughout this zone are linked with the popular martyr cult of St. Sergius, and reflect the status of al-Ḥārith and his son, al-Mundhir. – Nitl (near Madaba): the church complex, dedicated to St. Sergius, contains a number of inscriptions, including one that probably refers to al-Ḥārith; it has even been argued that members of the Jafnid family were buried there, although this is unclear.55 – Tall al-ʿUmayrī East (southern Amman): discovered during engineering works in 2009, part of the damaged inscription of a church of St. Sergius reads: “Our Lord Jesus Christ, God of Saint Sergius protect the megaloprepes-

51 52 53 54

55

Procopius, 1.17.30, 40 (trans. Dewing). Procopius, 1.17.40–48. E.g. John of Ephesus, Ecclesiastical History 3.6.3/pp. 280–282 (raid by al-Mundhir against Qabus, c. 570). E.g. John of Ephesus, Ecclesiastical History 3.6.4/pp. 284–287 (raid by all-Mundhir against the Persians in 581). Converting “wealth into a reputation for generosity” was one way to mitigate the dangers that came with service to the state, such as the perception of ‘going over’ to the state itself. See al-Rasheed 1987, 36, for a modern example; for the moral dimension, Lancaster and Lancaster 2004, 59–60. Fisher 2015, 193–197, 332–342.

reflections on arab leadership in late antiquity

–

–

–

–

499

tatos [magnificentissimus] Almoundaros, the komes.” The text should be dated to sometime between 563/4 and 569.56 Qaṣr al-Ḥayr al-Gharbī (west of Palmyra): two of the five panels of the lintel inscription from the Roman building at the site (probably the monastery of Haliarum) refer to al-Ḥārith. Panel 4 refers to “the endoxotatos [gloriosissimus] Arethas the stratēlatos”, while panel 1 commemorates a visit by alḤārith. The text, according to the most recent examination by Pierre-Louis Gatier, is undated.57 al-Burj (Ḍumayr, east of Damascus). This undated text records that “Flavius Alamoundaros, paneuphēmos, patrikios and phylarch”58 constructed a martyrion.59 Sergiopolis/al-Ruṣāfa: in the so-called “al-Mundhir building” outside the walls of the city and close to the shrine of St. Sergius, a text that celebrates al-Mundhir himself. Undated.60 Sammāʾ (near Suwayda). A Christian text calling for “Lord God of St. George” to give his protection to Abū Karib, this is sometimes placed in the context of the Samaritan revolt (529).61

Other, unknown Arabs appear on dedicatory inscriptions on martyria at Zabad (512, northern Syria) and Ḥarrān (c. 567/8, southern Syria), suggesting that the examples connected with the Roman Arab leaders were part of a broader trend. The text at Ḥarrān, from a martyrion of St. John, explicitly identifes in Arabic a certain “Šaraḥīl son of Ẓālim” (Greek: “Asaraël, son of Talemos”) as the founder of the structure.62 At Zabad, a martyrion of St. Sergius, part of the Arabic text mentions a “Sirgū son of Saʿdū and Š/Syrw and Š/S{.}ygw”. In Greek it then records perhaps the same individual (“Sergius son of Sergius son of Sergius”) as one of several associated with the its construction. An “Abu Sergius” is also identified in the Syriac portion of the Zabad text as one of the rebuilders of the structure, presumably after it fell into disrepair.63

56 57 58 59 60 61 62 63

Bevan et al. 2015, 54 (trans. Bevan). Fisher 2015, 197–202, 320–323 (trans. Genequand, rev. Gatier); Gatier 2015, 196–201. On this term, see below. Fisher 2015, 347 (trans. Gatier). Fisher 2015, 202–205, 330–331. Fisher 2015, 324 (trans. Bevan). For these inscriptions see Fisher 2015, 347–350 (Greek and Syriac; trans. Bevan), 410–415 (Arabic; trans. Macdonald). A couple of fifth-century martyria from Syria are often associated with Arabs or Arab benefactors, although the links have been very difficult to prove. See Fisher 2015, 311–312.

500

fisher

The actions of Arab benefactors can be explained via a desire for assimilation to the elites of the late Roman world, traditionally responsible for such activities. Yet it is also possible that men like al-Ḥārith and al-Mundhir were adapting traditional duties of patronage and community support to Christian religious sites, or were deliberately using cultural symbols of the Roman world to support their political position. The lines are unclear. Nevertheless, the visibility of Christian sites in the epigraphic record is significant. The building and upkeep of churches, martyria, and monasteries, and other acts of patronage (visits or commemoration via inscriptions) were of clear benefit to the community at large, and offered a way to build reputation, prestige, and social and political standing, allowing al-Ḥārith and al-Mundhir to acquire a certain clout as figures of authority within religious communities. The association between the Roman Arab leaders and the cult of St. Sergius gave them a particularly advantageous position at its center at Sergiopolis/ Ruṣāfa in northern Syria. There, at this important site that lay at the confluence of several routes across the Roman east, they were highly visible: interpretations of the so-called “al-Mundhir building” in which the Roman Arab leader was commemorated range from church and audience hall to a “great shaykh’s seven-poled tent”—a space for communication and information gathering, for mediation, and for hospitality and demonstration of generosity.64 Such a space allowed room for meetings and discussions, for the gathering of information, and for al-Mundhir to show an image of power and authority that rested on his support from Constantinople and his ability to represent the tribe’s interests there.65 It is not a coincidence that when al-Mundhir came to conduct the most sensitive negotiation of his career—his return to Roman service after Justin ii ordered his assassination—he chose to do so at Ruṣāfa. His discussions with the patrician Justinian took place “before the tomb of the blessed Mar Sergius”, at a location with deep connections to the prestige of his family.66 In the Persian empire, it appears that the leading Arab dynasty at al-Ḥīra also engaged in similar acts of benefaction towards monasteries, churches, and individuals. The tenth/eleventh-century Chronicle of Seert suggests that Ḥasan, son of al-Nuʿmān (d. c. 602), was particularly well disposed towards the poor.67 64 65

66 67

Quote from Whittow 1999, 222. For interpretations of the building see Fisher 2015, 202– 205. Cf. Lancaster and Lancaster 2015, 65; Beck 1986; Salzman 2004b. The gathering of information was another important prop, enabling the leader to be “better informed than anyone else”, as noted by Lancaster 1997, 93. John of Ephesus, Ecclesiastical History 3.6.4/pp. 284–287. Chronicle of Seert, pp. 468–469.

reflections on arab leadership in late antiquity

501

The Arab-Islamic tradition records two monasteries at al-Ḥīra that bore the names of members of the ruling family. One was the Dayr Hind al-Kubrā, which Yāqūt (d. 1229) claims was built by Hind the Elder, wife of al-Mundhir (504– 554). Yāqūt copied an inscription that he purported to have seen on the Dayr Hind al-Kubrā, which dated its construction to the rule of Khusrau i (531–579). The other monastery was the Dayr Hind al-Ṣughrā, which the Chronicle of Seert suggests was built by the daughter (or perhaps sister) of the same al-Nuʿmān mentioned previously.68 Both the Chronicle of Seert and a Syriac source, the seventh-century Khuzistan Chronicle, say that Hind the Younger buried the catholicos Īshōʿyābh in the monastery named for her at al-Ḥīra, further hinting at the close links between the dominant Arab family there and local religious structures that parallel the more detailed information available for the Roman empire.69

4

Protecting the Tribe

Tribal chiefs were “instigators of defence”,70 responsible for sheltering the people under their control. Threats often originated from other tribal groups, against whom violence was one solution: Roman- and Persian-allied Arabs fought each other repeatedly, often without state support or sanction. But there were also other factors that emerged as part of the state/tribe relationship, forcing tribal leaders to adapt to changing circumstances and become middlemen, allowing the state access to the tribe (and offering the opportunity to control it), but also advancing the tribe’s political interests.71 The middleman was usually the tribal chief himself, a broker, spokesman and negotiator who acted as a 68 69 70 71

Yāqūt, Muʿjam 2. p. 709; Chronicle of Seert, p. 442. Khuzistan Chronicle, p. 16; Chronicle of Seert, p. 442. Lancaster and Lancaster 2015, 81. Lindner 1982, 699–711; Lancaster and Lancaster 2004, 61; Salzman 1996, 155; cf. Beck 1990, 215, and 217: “by assisting the state in its rule, tribal leaders were able to control the state’s access to the tribe and help defend the tribe against the state. But they also helped the state impose central order on the tribe and increase state power; thus they facilitated the state’s efforts to control or eliminate both tribal leaders and the tribe itself”. Salzman 2004b, 93, on the position of the tribal chief (sardar) in 1960s/70s Iran as “heavily involved in acting as a middleman between the tribe and the state authorities, serving as intermediary, mediator, and broker, and … able to draw on some state political and economic resources to bolster his position inside the tribe”. See too Khazanov 1994, 215–218; Donner 2005, 515; Salzman 2004a, 319; Caton 1990, 81, but also 82, where not all middlemen relationships followed this trajectory, or produced this sort of outcome.

502

fisher

pivot between state and tribe. He sometimes underwrote tribal loyalty for the state in return for a certain devolution of authority as well as funding, titles, and other forms of recognition. Middlemen held a great deal of control over which resources the state accessed and which were withheld, and could prove vital in difficult times.72 Furthermore, as far as the tribe was concerned, the position highlighted the chief’s position as a benefactor, since blandishments from the state, such as money, flowed under his control and at his instigation to those under his leadership. Acting as this interface between state and tribe allowed middlemen to obtain elite status in the ranks of the state. Such membership allowed tribal leaders to advance their political interests from a very beneficial position, but it also posed a very real threat to winning and keeping a reputation in the tribe, and indeed the position of middleman presented a constant threat of going over to the state and losing one’s position in the tribal milieu (even visits to centres of state power, such as the capital city, could trigger this sort of distancing).73 Furthermore, the middleman “became the reluctant bottom man” as far as the state was concerned, even if he remained the individual with the greatest prestige in the tribe.74 This situation produced what Salzman has termed “the middleman’s dilemma”, placing tribal leaders in situations where they were, from time to time, morally responsible for the enactment of potentially unpopular state policies.75 The Roman Arab leaders appear to have mitigated some of these problems through their support of anti-Chalcedonian Miaphysite Christianity, which allowed them to engage with the dominant religion of the Roman empire while assuming a position of leadership in provinces where Miaphysitism was popular, such as Arabia. On the other hand, as the Roman Arab leader al-Mundhir found out in 582, elite status in the ranks of the

72

73

74 75

The Sunni Awakening in Iraq was dependent on the willingness of tribal chiefs in Anbar to act as middlemen for their tribespeople, committing their manpower and moral authority to support the us Surge. See Kilcullen 2016, 40–41, referring to the Anbar tribes as “cultural entrepreneurs”; cf. Salzman 2008, 181–182. Salzman 2004b, 100; Barth 1961, 90, on the Basseri and their relationship with Iran: “by being the leader of the tribesmen, a power factor in the province, … [the chief] can become a member of the privileged urban elite, and is thus able to defend the interests of the tribe within the sedentary hierarchy of authority, from a vantage point which is [otherwise] unattainable”; Lancaster and Lancaster 2015, 73. Salzman 2004a, 325. Salzman 2004a, 320–322. The closest corollary to this in the examples discussed in this essay is the Roman decision to provide Mavia with an unsuitable bishop, which she then chose not to support.

reflections on arab leadership in late antiquity

503

state counted for nothing when its members wanted to reinforce his position as the “bottom man”. Al-Mundhir could not compete in the Roman court, where, despite his titles, prestige, and record of service, he was shut out of its networks and swiftly eliminated at the hands of his own patrons.76 Middlemen are easily visible in a variety of ancient sources. In Roman sources, this prominence is due in part to the empire’s preference for working with non-Roman elites (e.g. tribal leaders) rather than with corporate bodies of people (e.g. the tribe); this preference focused contact and communication around single individuals, not groups, who then became the natural focus for discussions about resources, policy, and other matters of interest between the two parties. Tribal middlemen appear regularly throughout Roman imperial history. In North Africa the Romans recognised principes gentium, tribal leaders who might win Roman citizenship for themselves and their families in return for keeping the peace.77 In North Africa the Romans used their own military officers as praefecti, political agents who represented imperial interests to local tribes. Eventually, this arrangement was reversed, and praefecti were chosen from the tribal elites themselves to represent tribal interests to the state.78 Along the Rhine and Danube frontiers, the Romans co-opted Germanic reges, tying them to the imperial administrative system and military apparatus; expectations of frontier peace and military service were swapped for trade, diplomatic contacts, and funding.79 In the east, the Romans recognised Arab leaders as ‘phylarchs’, a term in use from the time of Strabo onwards to refer principally to Arab chiefs. By late antiquity, phylarchs had become part of the Roman bureaucracy, usually organised by province and expected to perform tasks similar to those carried out by the Germanic leaders in Europe or the principes gentium in North Africa.80 Roman Arab phylarchs such as al-Ḥārith, for example, fought in Rome’s wars and maintained peace amongst other groups along the frontier. In return for this service, al-Ḥārith and his family were rewarded with funding and membership in the illustres, the Roman elite. The Persian Arab leaders who appear in Roman sources appear to have enjoyed a similar sort of relationship with the Sasanian kings, fighting in Persia’s wars in exchange for royal support for their activities, which included raiding the 76 77 78 79 80

Fisher 2011, 173–184. For a recent study see Fisher and Drost 2017; for the bestowal of citizenship, see Seston and Euzennat 1961 and Modéran 2003, 481–505. E.g. Lepelley 1974 and Modéran 2003, 481–483. As outlined for example by Heather 2001. Grouchevoy 1995 remains the classic study; see also the Introduction to Dijkstra and Fisher 2014.

504

fisher

rich lands of Syria and pushing their influence, perhaps, into the northern part of the Arabian Peninsula.81 Middlemen had long been a component of successful client management programs, and were not at all unique to the Roman empire. Texts from the kingdom of Mari in Syria show that it established its own middlemen (sugāgū) to manage the tribal peoples along its periphery.82 Tribal middlemen were well placed to exercise a range of leadership skills such as mediation, benefaction, and advising, but the position of middleman also allowed the tribal leader to protect the interests of the tribe, because the key resource that the state normally wanted, manpower, was typically controlled by a single individual—the middleman. Because Roman authorities, in particular, had chosen to access tribal manpower in this manner, they consequently had only a limited ability to control it. Blandishments such as titles and funding could keep middlemen pliable, but the threat of military action if the state pushed too far was always there. Thus, while Roman accounts of state/tribe interactions predictably offer a rather one-side view of client management practises, often casting allies in a subordinate light, this was not always true from the tribal perspective. Roman sources took a dim view of Mavia’s unwarranted backlash against the legitimate authorities in the region. They stress that the treaty between Mavia’s husband and the emperor was cancelled on the former’s death, and artfully omit the actual cause for her rebellion, which has the effect of highlighting the moral authority of the Roman response. Valens, as an heretical Arian from the point of view of the ecclesiastical historians, comes off badly in these accounts, but the authors who tell the story of Mavia’s revolt do not question the prerogative of the state to make treaties with treacherous barbarian allies, and punish them if they stray from the expected line. From the perspective of a tribal middleman, however, Mavia’s revolt looks rather different. Her interests and those of the tribe threatened by the state, she deployed the most effective resource she possessed—the manpower under her control—to protect those interests. Once Roman authorities showed themselves prepared to cooperate after she successfully engaged the Roman military in battle, she not only willingly returned to alliance with Rome, but cemented her loyalty via a marriage alliance between her daughter and the Roman commander in the east, before deploying her forces to assist in the defence of Constantinople in 378.83 Mavia controlled both the withholding and granting of the military power of the tribe, 81 82 83

Fisher 2011, 92–95. Mari: Matthews 1978, 135–136, and Fleming 2008; see too Fabietti 2000, 84 and Tapper 1990, 66. Socrates 4.36.

reflections on arab leadership in late antiquity

505

defending her integrity and that of her people in the process. Mavia was also, perhaps, something of an opportunist, likely aware of events taking place west of Constantinople in the Balkans; it was a favourable time to risk a forceful negotiation. The activities of later Arab leaders repeat some of these elements, showing that men like al-Ḥārith and his son al-Mundhir were quite capable of acting opportunistically and making use of the resource most desired by the state in order to improve their prospects. Examples of tribal leaders protecting their interests can therefore be found in instances involving military power. Raiding was part of tribal culture.84 Some raids were occasioned by drought or the struggle for resources, although imperial alliances lessened the need for such activities.85 Nevertheless, both Roman and Persian Arabs appear to have taken a dim view on efforts to restrain raiding, which could be tremendously lucrative and provided ways for chiefs to redistribute wealth. Roman and Persian Arabs do though appear to have given some concessions to their patrons, by directing raids mostly at each other and by suppressing localised feuding in their respective spheres of influence. Unwilling to surrender an important part of their political autonomy (and potential leverage),86 they exploited the ongoing tensions between Rome and Persia for hostage taking, harrying, and pillage. Persian Arab leaders, most notably al-Mundhir, routinely raided Roman territory on missions that, even if they were generally tolerated by the Persian king, had little to do with Persian military campaigns. Roman sources describe the amount of money that al-Mundhir made in these expeditions, principally from kidnapping and ransoming hostages.87 His Roman Arab opponents likewise exploited the regularity of frontier violence and raiding to enrich themselves and fight their enemies without imperial supervision.88 Earlier, during the war between Anastasius and Kavadh, the anonymous author of the Chronicle of Ps.-Joshua the Stylite noted bitterly how Arab leaders opportunistically supplemented their income in the chaos sweeping Mesopotamia, saying to his abbot that “this war was the cause of much enrichment for the Tayyaye [Arabs] of both sides … they did as they pleased in both empires.”89 On this occasion

84 85 86 87 88 89

Betts and Russell 2000, 31–32; Salzman 2004b, 93; Salzman 2008, 136–137. For Arab raiding in antiquity, see Lenski 2011 and Fisher 2015, 217–231, 287–296. E.g. Barṣauma, Letter 2 = Syn. Or. pp. 526–527 (ce 484/5), and Marcellinus Comes addit. s.a. 536.11 (ce 536). cf. Fabietti 2000, 84–85. E.g. Ps.-Zachariah, 8.5a, and Malalas, p. 445 and pp. 460–461. Procopius, 2.19.26 and 2.28.12–13. Ps.-Joshua the Stylite, 51–53, 79 (trans. Trombley and Watt).

506

fisher

imperial authorities took swift punitive action, hanging some Arab leaders and putting others to the sword.90 Eventually the steady growth in unsanctioned tribal military activity caused both Roman and Persian authorities to ban intertribal violence in the treaty of 561/2.91 Even then, Arab leaders remained understandably reluctant to give up an important source of leverage, forcing imperial authorities to resort to desperate measures to control their errant allies. Not long after the treaty was agreed, the Roman Arab leader al-Mundhir demanded gold from the Roman emperor Justin ii to hire mercenaries, but the apparent haughtiness of the request went over badly in Constantinople and the emperor ordered alMundhir’s assassination. When the attempt was botched, the Arab leader took one of the few remaining options available to him: unwilling or unable to attack Roman forces as Mavia had done, he decided, instead, to go on strike, withdrawing his forces from Roman service and leaving parts of the frontier undefended.92 In his absence the Persian Arabs took full advantage to raid and harass Roman territory, and only when al-Mundhir was reconciled at Sergiopolis/Ruṣāfa did Persian raiding come to a temporary end. While al-Mundhir’s actions come across as almost petulant, they actually show a remarkably successful renegotiation of the relationship between himself and the Roman emperor. Al-Mundhir’s action was different from that taken by Mavia, but his sole control over a valuable resource—devolved to him as middleman by Rome’s own policies—allowed him to make the point. Both tribal strength and his own prestige received ample augmentation from the withering defeat al-Mundhir subsequently inflicted on the Persian Arabs.93 To the Romans and Persians, the military activities of Arab leaders could be a nuisance that required legislation. Arab leaders could be characterised as disloyal, uncivilised, prone to individual action and to disobeying orders; the state, suspicious of acts of tribal independence, preferred a tribe that was politically subordinate. Our sources reflect this perspective.94 Client management policies did offer clear benefits to the state: middlemen could be captured, killed, exiled, imprisoned, or replaced, as indeed happened with both

90 91 92

93 94

Ps.-Joshua the Stylite, 88. Fisher 2011, 119–122. John of Ephesus, Ecclesiastical History 3.6.4/p. 284; cf. Tapper 1990: 67, on strategies of adaptation that might “alternate … between acceptance of indirect rule, military resistance, and avoidance”, as well as Hole 2008, 262. John of Ephesus, Ecclesiastical History, 3.6.4/pp. 284–287. cf. Lindner 1982, 696, and Velud 2000, 79.

reflections on arab leadership in late antiquity

507

the Roman and Persian Arab leaders (below).95 But from a different viewpoint, tribal middlemen were taking advantage of the way that the state had chosen to manage them. Steven Caton summarized Lois Beck’s 1986 study of the Qashqaʾi of southwest Iran by noting: [Tribal] elites were in a certain sense created by the state in order that it may rule a difficult and remote population more efficiently. This elite was responsible for collecting taxes, conscripting young men into the state’s army, and implementing policies; in short, it was the state’s indirect ruler. However, the elite also acted to safeguard the interests of its constituents in the face of encroaching state power; and having been elevated to power by the state, elites could be its potential enemies.96 As middlemen of the Roman and Persian empires, Arab leaders defended themselves and their people, furthered their interests, and also reminded their patrons, from time to time, that the relationship between the two parties was not that one-sided after all. In a two-state world where tribal manpower was a commodity that could easily be transferred to a new patron, this gave tribal leaders tremendous leverage.97

5

Prestige, Reputation, and Status

Prestige was perhaps the most vital of all currencies of political power for tribal leaders. It could be won through all of the activities discussed so far: on the battlefield, defeating the enemies of the tribe and defending its interests; through participation in the political life of the state and through being seen to have access to the state’s hierarchy and its resources; through successfully negotiating solutions to the state’s problems. Prestige was reflected in actions of benefaction and generosity—the construction of public buildings, or the donation of plunder and booty to followers, institutions, and the needy. The reputation of tribal leaders is also visible away from the titles and ranks given by the state: the Arabic inscription from Jabal Says in Syria gives al-Ḥārith the title of malik, ‘king’, a title also used by at least one leader of Kinda, Ḥujr, acting as a deputy of the Ḥimyarite kings in central Arabia; several other texts 95 96 97

cf. Beck 1990, 217. Caton 1990, 102. cf. Gellner 1990, 113: “outside powers [took] an active interest in controlling [tribes] or denying such control to others.”

508

fisher

or inscriptions reveal the use of the title elsewhere in the east.98 Its appearance in connection with individuals who were also actors in state hierarchies is particularly significant. In the case of al-Ḥārith (and Ḥujr, perhaps), being middleman averted the contradiction of being both independent ‘kings’ and agents of the state; in the west, to cite just one example, another middleman, the Frankish king Mallobaudes, was both “king of the Franks” and comes domesticorum.99 Yet prestige was also the currency: it could be earned from a variety of sources, but could be lost in a single blow and without it little was possible. When prestige was lost, the results were fatal not only to tribal leaders, but sometimes also to the interests of the tribe. After al-Mundhir was exiled in 582 when both his political standing and credibility as a religious mediator vanished, the hitherto stable alliance between the Arabs and the Romans collapsed and was never reconstituted.100 In Persia, not long afterwards, Khusrau ii abruptly terminated his own agreement with the Arabs of al-Ḥīra, likely due to a marked lack of support during the civil war during the early part of his rule. The withdrawal of imperial favour had a dramatic impact on the Persian Arab leadership, and the Arab alliance with Ctesiphon, active in various guises for over three hundred years, collapsed.101 It is noteworthy that the Roman and Persian Arab pacts with their sponsors fell apart because of external causes, not because the leaders lost the “public opinion” of their tribe. This fact highlights the important way in which reputation within the tribe was in part attained through association with prominent individuals in the state hierarchy. In the sixth century, Arab leaders enjoyed a close relationship with their imperial patrons: al-Ḥārith, endoxotatos, patrikios, and stratēlatos, was personally chosen by Justinian, and shortly before his death he enjoyed an audience with the emperor to establish his son al-Mundhir as his successor. The discovery of the Tall al-ʿUmayrī inscription, recording the rank of megaloprepestatos (subordinate to endoxotatos) held by al-Mundhir, and likely dated to c. 563/4, suggests that the Romans had given some thought about how to structure Arab leadership within their own social and political hierarchy. Al-Mundhir would also be given patrician rank and hold the very senior grade in the illustres of paneuphēmos. His uncle Abū Karib was also a mem-

98 99 100 101

Fisher 2015, 412 (Jabal Says); 144 (Ḥujr); 324 (Abū Karib); other examples include, for e.g. Avner et al. 2013. Ammianus, 31.10.6. See van der Steen 2008, 110, for a 19th c. comparison. Fisher 2015, 262–266. Fisher 2015, 270–272.

reflections on arab leadership in late antiquity

509

ber of the illustres, as were al-Mundhir’s own children.102 A significant level of standing came from imperial patronage.103 Persian Arab leaders, most notably al-Mundhir, were also close to the throne, although it is unclear whether they similarly earned title or rank amongst the Sasanian aristocracy. However, the name of Aspebetos (above) is usually thought to be a corruption of spāhbed, “master of horse”, suggesting that some titles may have been available to Arab leaders in the Persian state.104 From the perspective of the state, it was preferable for the position of the chief to be “as closely tied to the whim of the government as to his prestige among the tribesmen.”105 The titles and ranks were part of the mechanism used by the Roman empire to bind al-Ḥārith and his family irrevocably to this “whim”. Another important source of prestige, though, came from their links to the divine. This was derived in part from the association between the Roman Arabs and important religious figures such as Jacob Baradeus, as well as a correlation of sorts between the activities of certain Arab leaders and the holy men of the late antique east. Arab leaders and holy men were mediators, who often depended on continued public success to earn and maintain their reputation. This was especially true of the monks, wandering hermits, and others who appear in Christianisation narratives of Arabs between the fourth and sixth centuries. These men and women held no specific office, possessing instead a more abstract authority derived from their links to the divine and their abilities as fixers or facilitators, solving all manner of problems (typically via healing miracles), and delivering access to resources both secular and holy (food, shelter, Roman officials, the Eucharist).106 Once this primary contact was complete and Arab groups adopted Christianity, these abstract types of holy men faded from the picture, and the role of facilitation was devolved onto men like al-Ḥārith. Bishops and priests, men like Jacob Baradeus, bolstered the religious cachet of Arab leaders, and as the latter became linked with important religious communities, such as the Miaphysites of the eastern provinces, they increasingly assumed a role not only as political middlemen but also as religious middlemen. In this respect they also began to look like moderators at church councils, and their skill at brokering religious compro102 103

104 105 106

For a discussion of the relevant texts and a table of ranks and titles see Bevan et al. 2015. The Ottoman leaders pursued a similar policy, granting the title of pasha to the leader of the Rwala in 1892 (Lewis 2000, 37); in Mandate Syria, bedouin leaders were granted the title of Chevalier de la Légion d’ Honneur (Velud 2000, 66–67). Fisher 2015, 305. Lindner 1982, 696. Keeping in mind the archetype of the holy man as outlined by Brown 1971.

510

fisher

mises blurred the lines not only between Christian holy men and Arab leaders, but also between the stature of Arab leaders and the religious authority of Roman elites, including the emperor, who also acted as brokers in religious disputes.

6

East and West?

Even a brief, generalised application of some of the main points about leadership drawn from modern anthropological studies yields some interesting results. First of all, it shows that a ‘tribal’ reading of the sources offers a nuanced and complex view the activities of Arab leaders, balancing the dominance of Roman material. Second, putting together the elements discussed in the sections above delineates a familiar image of ancient Arab leadership: a warrior figure of status and political prestige, adept at facilitation and intercession; a benefactor, strongly linked with a powerful religion. Although rather diluted in its sixth-century form compared to what was to come, there are some parallels between this combination of military, political, and religious authority and characterisations of the Prophet, the power of the caliphs, and even the existence of what Gellner has called ‘holy lineages’ amongst Middle Eastern tribes, men who claimed a link to the Prophet and offered leadership and arbitration.107 Yet it is also a figure familiar to the Roman world, bringing to mind the position of the Roman emperor, a military and political figure whose position as a broker for Christians was established by Constantine. Role models of a powerful military elite were also widely available, due to the militarisation of the late Roman aristocracy.108 What influenced the development of Arab leadership, helping to produce the figure as I have outlined it here? Two elements in particular can be underlined, particularly as they present subjects for further detailed study. First, Roman policies of client management/indirect rule were probably as influential in the east as they were in the west, centering interaction around tribal elites who then became middlemen. Thinking along the lines of Beck’s study of the Qashqaʾi, it might even be possible to make the argument that al-Ḥārith’s position was a creation of the Roman state, after Justinian took the unprecedented step of installing him “in command of as many clans as possible” in an 107

108

E.g. Hoyland 2012, 1058. Cf. Lapidus 1990, 34–35, specifically on the way that the ‘religious patriarch’ acquired military and administrative authority (caliphs); Gellner 1990, 112. For the religious/political authority of the caliph see Crone and Hinds 1986. See Halsall 2007, 101–110.

reflections on arab leadership in late antiquity

511

attempt to deal with the Persian Arab al-Mundhir.109 In any case, al-Ḥārith and his son found a ready market for their ability to mediate, which they adapted to a broader canvas, and ample opportunity to sell the power of the tribe to the Romans for relentless bouts of interstate warfare. Those episodes also created plentiful chances for enrichment via hostage taking, plundering, and then the bolstering of an image of a generous lord via the distribution of wealth. The Roman propensity for interpersonal links between emperors or imperial authorities and others with whom they developed treaties and alliances was also important: in a world where prestige could be parlayed into authority, the personal support of the emperor was immensely useful. Arab leaders were not in the same position as the Germanic chiefs to create similar kingdoms, but the relentless focus on individual Arab middlemen had the inevitable effect of crystallising their power and bolstering their status. Under different circumstances, this may have fuelled an even more noticeable degree of social and political differentiation already visible in our sources (e.g. ranks in the illustres) and triggered large scale political changes.110 In the meantime, the repeated successes on the battlefield and in mediation between disputing parties created a cycle of steadily increasing influence that was eventually ended when the vital pillar on which it all rested—prestige—was threatened. Another important factor for the development of late antique Arab leadership was the way that Christianity was introduced to and adopted by Arab tribes. For example, some holy men made a distinct effort to incorporate preexisting tribal structures into religious practise, such as Aḥūdemmeh, active in the Persian empire in the mid-sixth century. Aḥūdemmeh targeted the Arab tribes of the Jazīra, installing amongst them deacons and priests and then encouraging participation in religious ceremonies by naming churches after individual tribes, mapping tribal identities onto the nascent ecclesiastical structure.111 Such associations, reflected in the Roman empire via the way that men like al-Ḥārith and his son patronised community churches such as those at Nitl and Tall al-ʿUmayrī, created a network of what Elizabeth Fowden has termed “fixed points”. Such places allowed for itinerant leaders to meet people and exchange information; perhaps the most recognisable of all was the “alMundhir building” at Ruṣāfa, built for interaction and the display of status, and situated close to a holy shrine which was an important pilgrimage site not only

109 110 111

Procopius, 1.17.47. Cf. Salzman 1983, 282, associating indirect rule policies with the alteration of tribal political structures; see also Salzman 2004b, 100, and Irons 1979, 362. Life of Aḥūdemmeh, pp. 26–27.

512

fisher

for Christians in the Roman empire but also for those in Persia.112 Noticeably, fixed points appear at other points in the history of the Arab world. The shrines of the Cyrenaican saints (marabouts) for example, dotted the tribal landscape of early modern Libya and parts of Egypt, providing places for pilgrimage and gatherings, and acting as neutral ground for the resolution of disputes. The Sanusi order of Cyrenaica (late 19th/early 20th c.) built on this network, creating “lodges”, particularly along tribal boundaries, fulfilling much the same sort of purpose but also helping to offer a political unity to the different tribes under their leadership.113 In seventh- and eighth-century Syria, the Umayyad caliphs used numerous sites that lay at the interface between the steppe and the desert as places for meeting and communication, to facilitate their relationship with Arab tribal leaders.114 Some of these places, such as Ruṣāfa, were part of the network of fixed points in Roman Syria and the province of Arabia. In the sixth century, management of these networks, which were overwhelmingly Christian in nature, helped to devolve a significant level of religious authority onto the Roman Arab leaders. This helped to reinforce their status, and provided other Arabs with prominent role models for participation in the Christian life of the Roman (and also Persian) empires, likely encouraging others like the mysterious figures who appear on the martyria at Zabad and Ḥarrān. The ease with which al-Ḥārith and al-Mundhir framed their experience as Roman allies in tribal terms, acting as mediators and conduits, particularly in religious matters, shows that it was not just the political landscape of the late antique east, but also its religious culture, that offered fertile ground for Arab leaders to hone their skills and build their status. It is not a surprise that there are some distinct similarities between Arab leaders and other barbarians: Arab leaders are rarely found in dialogue with Germanic kings, but the Ostrogothic king Theodoric remains a useful foil. Peter Sarris recently characterised Theodoric as an itinerant king, taking “great care to maintain regular face-to-face contact with his chief military retainers and their retinues”. Theodoric held these audiences at a “network of lavishlydecorated royal palaces”, where he gave out “donatives and rewards” and listened attentively to “grievances”. Sarris concludes that Theodoric’s leadership was “military in tone”, and further notes the importance of supporting the Arian church (and importantly, not antagonising the Catholic church) in sustaining his position.115 This image of Theodoric rests on many of the leadership 112 113 114 115

For ‘fixed points’ see Fowden 2015, and discussion of the theory in Fisher 2015, 286–287. Evans-Pritchard 1949, 62–89. See the recent discussion in Fisher 2011, 205–210. Sarris 2011, 105, 108.

reflections on arab leadership in late antiquity

513

attributes discussed in this essay: the ability to advise and mediate, the importance of military leadership, generosity, and the use of fixed points where community leadership and benefaction could be expressed. An association with religious authorities also played an important role. As a final issue for further study: on the basis of ancient sources, it is difficult to judge what congruence (if any) exists between the image of the Arab leader in texts and inscriptions, and their actual activities. In his critique of Frederik Barth’s ethnography of the Basseri of southern Iran (Nomads of South Persia, 1961), Philip Salzman drew a line between “image and reality”, seeing the Basseri chiefs act in certain ways for the consumption of certain parties, but quite differently in other situations. Salzman suggested that Barth was drawn in by the image of the chief as a powerful autocrat, acting in an “imperial manner”— a picture “based rather too much upon style and image and too little upon substance and reality.”116 The chief in fact possessed remarkably little power, and was dependent for his success on his work as a middleman.117 Consent, not coercion, provided stable rule. Salzman points to other situations in which anthropologists and visitors were offered carefully cultivated images of tribal leaders, designed to show that they were more powerful than they actually were.118 The relevance of this problem for writing about ancient Arab leadership is found in Salzman’s explanation for the “imperial manner” of the Basseri chief: Nothing served the Basseri better in their complex and plural social environment, where they were surrounded by agents of the Persian government, by urban populations, by sedentary villagers, and by other tribes, than the picture of a solidary and unified tribe unquestionably following the commands of their strong and unchallenged ruler. The Basseri projected this image, they broadcasted it, they asserted it, and they may even have believed it. So, too, did the chief himself … This “presentation of (collective) self” was the façade exhibited by the Basseri to the wider world, to encourage respect and fear and discourage opposition and aggression.119

116 117 118

119

Salzman 2004b, 79. Salzman 2004b, 100. E.g. Lady Anne Blunt visiting the Rwala in 1879; Lancaster’s description of the ostentatious display provided for external consumption at a feast. See Salzman 2004b, 92, and Lancaster 1997, 82–83. Salzman 2004b, 90.

514

fisher

Beck’s study of the Qashqaʾi of southwest Iran similarly noted the ability of the different leaders “to symbolically manipulate the image of Qashqaʾi indomitableness before potential allies and enemies.”120 It is quite possible that the image of the powerful Arab leader, presented here, was similarly manicured—or, as Lindner phrases it: … if a suitably impressive outsider appeared and asked someone [of the tribe] to “take me to your leader” or wondered “who’s in charge here?” his hosts would have to choose a respondent.121 Roman and Persian Arab leaders grew in power in a manner proportionate to their intensified role in the activities of the state, and their public image— particularly the inscriptions and texts actually produced or influenced by the chiefs themselves—probably reflects this fact. It is thus quite conceivable that the image of a strong warrior, a man of generosity, a mediator between religious factions, was in fact exaggerated: generated in response to the relationship between Rome, Persia, and their Arab allies, and this construct might spread to include other frontier allies, such as the Germanic barbarians. If anything, this problem actually heightens the importance of the influence of state actors on one of the most significant elements across the late antique spectrum, the rise of non-Roman (or non-imperial) leaders on the periphery of the state. Theodoric may not be that different from figures such as al-Ḥārith or al-Mundhir; indeed, all stood at transition points between late antique barbarian kingship and the emergent types of leadership that would follow in the early medieval period. Historians of the west have long looked at client management policies and the development of Germanic kingship as key elements in the debate over continuity and change in the history of Europe. Similarly, while there are clear discontinuities between Arab tribal leadership in the sixth century and the imperial power of the caliphs, some of the origins of strong Arab leadership in the ancient Middle East seem to be located deep in the pre-Islamic period. Indeed, William Montgomery Watt suggested long ago that South Arabian kingship inspired models of early Islamic governance, but I would argue here that the development of Arab leadership is more intimately connected than we have previously thought to Romano-Persian imperial policy, and the role of states in producing and sustaining powerful tribal leaders.122

120 121 122

The assessment of Caton 1990, 102. Lindner 1982, 699. Hourani 1990, 305. See Watt 1961.

reflections on arab leadership in late antiquity

515

Bibliography Primary Sources Ammianus = Ammianus Marcellinus, History, translated by John Rolfe, 3 vols. Cambridge, ma: Harvard University Press, 1939–1950. Barṣauma = Synodicon orientale, edited by Jean-Baptiste Chabot. Paris: Imprimerie Nationale, 1902–. Cassiodorus = Variarum, edited by James W. Halporn. Corpus Christianorum, Series Latina, vol. 46. Turnhout: Brepols, 1973. Chabot, Jean-Baptiste, ed. Documenta ad origines Monophysitarum illustrandas. Leuven: Secrétariat du CorpusSCO, 1952. Chron. 1234 = Chronicle to 1234, Chronicon anonymum ad a.c. 1234 pertinens, edited and translated by Jean-Baptiste Chabot. Corpus scriptorum Christianorum Orientalium. Scriptores Syri, 55–56 Leuven: Peeters, 1904. Chronicle of Khuzistan = Chronicon Anonymum, edited and translated by Ignzaio Guidi. Leuven: Secrétariat du CorpusSCO, 1903. Chronicle of Seert = Histoire nestorienne inédite, edited and translated by Addai Scher. Patrologia Orientalis 4, 5, 7 and 13. Turnhout: Editions Brepols, 1950–1973. Cyril of Scythopolis, Life of St. Euthymius = The Lives of the Monks of Palestine, translated by Richard Price. Kalamazoo: Cistercian Publications, 1991. Evagrius = The Ecclesiastical History of Evagrius Scholasticus, translated by Michael Whitby. Liverpool: Liverpool University Press, 2000. Gregory of Tours = Historia Francorum, edited by Bruno Krusch. Monumenta Germaniae Historica, Scriptores Rerum Merovingicarum, vol. 1. Hanover: Impensis Bibliopolii Hahnani, 1965. John of Ephesus, Life of Jacob and Theodore = Lives of the Eastern Saints, edited and translated by Ernest Brooks. Paris: Firmin-Didot, 1923–1924. John of Ephesus, Ecclesiastical History = Iohannis Ephesini Historiae ecclesiasticae pars tertia, edited and translated by Ernest Brooks, 2 vols. Corpus scriptorum Christianorum Orientalium. Scriptores Syri vol. 3, 54–55. Paris: E Typographeo Reipublicae, 1935–1936. Ps.-Joshua the Stylite = The Chronicle of Pseudo-Joshua the Stylite, trans. and introd. Trombley, Frank and Watt, John. Liverpool: Liverpool University Press, 2000. Life of Aḥūdemmeh = Histoires d’Aḥoudemmeh et de Marouta, trans. François Nau, Patrilogia Orientalis 3. Paris: Firmin-Didot, 1903. Malalas = The Chronicle of John Malalas, translated by Elizabeth Jeffreys, Michael Jeffreys, and Roger Scott. Melbourne: Australian Association for Byzantine Studies, 1986. Marcellinus Comes = Chronicon, edited and translated by Brian Croke. Sydney: Australian Association for Byzantine Studies, 1995.

516

fisher

Michael the Syrian = Chronique, edited and translated by Jean-Baptiste Chabot, 4 vols. Paris: Ernest Leroux, 1899–1910. Procopius = Procopius, translated by Henry Dewing, 7 vols. Cambridge, ma: Harvard University Press, 1914–1954. Socrates Scholasticus = Kirchengeschichte, edited by Günther Hansen. Berlin: Akademie Verlag, 2011. Sozomen = Kirchengeschichte, edited by Joseph Bidez, revised by Günther Hansen. Berlin: Akademie Verlag, 1995. Theophylact Simocatta = The History of Theophylact Simocatta, ed. and trans. Whitby, Mary and Whitby, Michael. Oxford: Oxford University Press, 1986. Yāqūt = Muʿjam al-buldān, edited by Ferdinand Wüstenfeld, 6 vols. Leipzig: F.A. Brockhaus, 1866–1873. Ps.–Zachariah = The Chronicle of Pseudo-Zachariah Rhetor. Church and War in Late Antiquity, edited by Geoffrey Greatrex and translated by Robert Phenix and Cornelia Horn, with an introduction by Sebastian Brock and Witold Witakowski. Liverpool: Liverpool University Press, 2010.

Secondary Sources Avner, U., L. Nehmé, and Ch. Robin. 2013. “A rock inscription mentioning Thaʿlaba, an Arab king from Ghassān.” Arabian Archaeology and Epigraphy 24/2: 237–256. Barth, Frederik. 1961. Nomads of South Persia. The Basseri Tribe of the Khamseh Confederacy. Long Grove, il: Waveland Press. Beck, L. 1986. The Qashqaiʾi of Iran. New Haven: Yale University Press. Beck, L. 1990. “Tribes and the State in Nineteenth- and Twentieth-Century Iran.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 185–225. Berkeley: University of California Press. Betts, A. and K. Russell. 2000. “Prehistoric and Historic Pastoral Strategies in the Syrian Steppe.” In The Transformation of Nomadic Society in the Arab East, edited by M. Mundy and B. Musallam, 24–32. (University of Cambridge oriental publications, vol. 58). Cambridge: Cambridge University Press. Bevan, G., G. Fisher, and D. Genequand. 2015. “The late antique church at Tall alʿUmayrī East. New evidence for the Jafnid family and the cult of St. Sergius in northern Jordan.” Bulletin of the American Schools of Oriental Research 373: 49– 68. Brown, P. 1971. “The Rise and Function of the Holy Man in Late Antiquity.” Journal of Roman Studies 61: 80–101. Caton, S.C. 1990. “Anthropological Theories of Tribe and State Formation in the Middle East: Ideology and the Semiotics of Power.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 74–108. Berkeley: University of California Press.

reflections on arab leadership in late antiquity

517

Crone, P. and M. Hinds. 1986. God’s Caliph. Religious Authority in the First Centuries of Islam. (University of Cambridge Oriental Publications 37). Cambridge: Cambridge University Press. Dijkstra, J. and G. Fisher (eds.). 2014. Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers. (Late antique history and religion 8). Leuven: Peeters. Donner, F. 1989 “The Role of Nomads in the Near East in Late Antiquity (400–800c.e.).” In Tradition and Innovation in Late Antiquity, edited by F. Clover and S.R. Humphreys, 73–88. (Wisconsin Studies in Classics). Madison, wi: University of Wisconsin Press. Donner, F. 2005. “The Background to Islam.” In The Cambridge Companion to the Age of Justinian, edited by M. Maas, 510–534. Cambridge: Cambridge University Press. Evans-Pritchard, E. 1949. The Sanusi of Cyrenaica. Oxford: Oxford University Press. Fabietti, U. 2000. “State Policies and Bedouin Adaptations in Saudi Arabia, 1900–1980.” In The Transformation of Nomadic Society in the Arab East, edited by M. Mundy and B. Musallam, 82–89. (University of Cambridge oriental publications, vol. 58). Cambridge: Cambridge University Press. Fisher, G. 2011. Between Empires: Arabs, Romans and Sasanians in Late Antiquity. (Oxford Classical Monographs). Oxford: Oxford University Press. Fisher, G. 2014. “State and tribe in late antique Arabia: a comparative view.” In Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers, edited by J. Dijkstra and G. Fisher, 279–295. (Late antique history and religion 8). Leuven: Peeters. Fisher, G. (ed.). 2015. Arabs and Empires Before Islam. Oxford: Oxford University Press. Fisher, G. and A. Drost. 2017. “Structures of power in late antique borderlands: Arabs, Romans, and Berbers.” In European and American Borderlands. A New Comparative Approach, edited by J.W.I. Lee and M. North, 33–82. Lincoln: University of Nebraska Press. Fisher, G. and Ph. Wood. 2016. “Writing the history of the ‘Persian Arabs’: the pre-Islamic perspective on the ‘Naṣrids’ of al-Ḥīrah.” Iranian Studies 49/2: 247–290. Fisher, G. and Ph. Wood. Forthcoming. “Arabia and the Late Antique East.” Fleming, D. 2008. “Kingship of City and Tribe Conjoined: Zimri-Lim at Mari.” In Nomads, Tribes, and the State in the Ancient Near East, edited by J. Szuchman, 227– 240. (University of Chicago Oriental Institute seminars, vol. 5). Chicago: Oriental Institute of the University of Chicago. Fowden, E.K. 2015 “Rural converters among the Arabs.” In Conversion in Late Antiquity: Christianity, Islam, and Beyond, edited by D. Schwartz, N. McLynn, and A. Papaconstantinou, 175–196. Farnham: Ashgate. Garthwaite, G. 2009. Khans and Shahs. A History of the Bakhtiyari Tribe in Iran. London: ib Tauris.

518

fisher

Gatier, P.-L. 2015. “Les Jafnides dans l’épigraphie grecque au vie siècle.” In Les Jafnides. Des rois arabes au service de Byzance (vie siècle de l’ère chrétienne), edited by D. Genequand and Ch. Robin, 193–222. Paris: de Boccard. Gellner, E. 1983. “The Tribal Society and Its Enemies.” In The Conflict of Tribe and State in Iran and Afghanistan, edited by R. Tapper, 436–448. London: Croom Helm. Gellner, E. 1990. “Tribalism and State in the Middle East.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 109–126. Berkeley: University of California Press. Genequand, D. and Ch. Robin (eds.). 2015. Les Jafnides. Des rois arabes au service de Byzance (vie siècle de l’ère chrétienne). Paris: de Boccard. Ginat, J. and A. Khazanov. 1998. Changing Nomads in a Changing World. Brighton: Sussex Academic Press. Grouchevoy, A.G. 1995. “Trois ‘niveaux’ de phylarques. Étude terminologique sur les relations de Rome et de Byzance avec les Arabes avant l’Islam.” Syria 72: 105– 131. Grunebaum, G. 1963. “The nature of Arab unity before Islam.” Arabica 10: 4–19. Haldon, J. and L. Conrad (eds.). 2004. Elites Old and New in the Byzantine and Early Islamic Near East. (The Byzantine and Early Islamic Near East, vol. 6). Princeton: Darwin Press. Halsall, G. 2007. Barbarian Migrations and the Roman West, 376–568. (Cambridge Medieval Textbooks). Cambridge: Cambridge University Press. Heather, P. 1995. “Theoderic, King of the Goths.” Early Medieval Europe 4.2: 145–173. Heather, P. 2001. “The Late Roman Art of Client Management: Imperial Defence in the Fourth Century West.” In The Transformation of Frontiers: From Late Antiquity to the Carolingians, edited by W. Pohl, H. Reimitz, and I. Wood, 15–68. (Transformation of the Roman World, vol. 10). Leiden: Brill. Hole, F. 2008. “Pastoral Mobility as an Adaptation.” In Nomads, Tribes, and the State in the Ancient Near East, edited by J. Szuchman, 261–284. (University of Chicago Oriental Institute seminars, vol. 5). Chicago: Oriental Institute of the University of Chicago. Hourani, A. 1990. “Tribes and States in Islamic History.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 303–311. Berkeley: University of California Press. Hoyland, R. 2012. “Early Islam as a Late Antique Religion.” In The Oxford Handbook of Late Antiquity, edited by S. Johnson, 1053–1077. Oxford: Oxford University Press. Hoyland, R. 2014. “Insider and outsider sources: historiographical reflections on late antique Arabia.” In Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers, edited by J. Dijkstra and G. Fisher, 265–278. (Late antique history and religion 8). Leuven: Peeters. Irons, W. 1979. “Political Stratification Among Pastoral Nomads.” In Pastoral Production

reflections on arab leadership in late antiquity

519

and Society, edited by L’Equipe Ecologie et Anthropologie de Société Pastorale, 361– 374. Cambridge: Cambridge University Press. James, E. 1988. The Franks. (Peoples of Europe). Oxford: Blackwell. Kaimio, M. 2001. “Petra inv. 83: A Settlement of a Dispute.” In Atti del xxii congresso internazionale di papirologia, edited by I. Andorlini, 719–724. Florence: Istituto Papirolgico G. Vitelli. Khazanov, A. 1994. Nomads and the Outside World, translated by Julia Crookenden. Madison, wi: University of Wisconsin Press. Khoury, Ph. and Kostiner, J. (eds.). 1990. Tribes and State Formation in the Middle East. Berkeley: University of California Press. Kilcullen, D. 2016. Blood Year. Islamic State and the Failures of the War on Terror. London: Hurst and Co. Kostiner, J. 1990. “Transforming Dualities: Tribe and State Formation in Saudi Arabia.” In Tribes and State Formation in the Middle East, edited by P. Khoury and J. Kostiner, 226–251. Berkeley: University of California Press. Lancaster, W. 1997. The Rwala Bedouin Today. Prospect Heights: Waveland2. Lancaster, W. and F. Lancaster. 2004. “Concepts of Leadership in Bedouin Society.” In Elites Old and New in the Byzantine and Early Islamic Near East, edited by J. Haldon and L. Conrad, 29–61. (The Byzantine and Early Islamic Near East, vol. 6). Princeton: Darwin Press. Lancaster, W. and F. Lancaster. 2015. “Concepts of Tribe, Tribal Confederation, and Tribal Leadership.” In Les Jafnides. Des rois arabes au service de Byzance (vie siècle de l’ère chrétienne), edited by D. Genequand and Ch. Robin, 53–77. Paris: de Boccard. Lapidus, I. 1990. “Tribes and State Formation in Islamic History.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 25–47. Berkeley: University of California Press. Leder, S. and B. Streck (eds.). 2005. Shifts and Drifts in Nomad-Sedentary Relations. (Nomaden und Sesshafte, vol. 2). Wiesbaden: Reichert. Lenski, N. 2011. “Captivity and slavery among the Saracens in Late Antiquity (ca. 250– 630ce).” Antiquité Tardive 19: 237–266. Lepelley, C. 1974. “Le préfecture de tribu dans l’Afrique du Bas-Empire.” In Mélanges d’histoire ancienne offerts à William Seston, 285–295. (Publications de la Sorbonne). Paris: E. de Boccard. Lewis, N. 2000. “The Syrian Steppe during the Last Century of Ottoman Rule: Hauran and the Palmyrena.” In The Transformation of Nomadic Society in the Arab East, edited by M. Mundy and B. Musallam, 33–43. (University of Cambridge oriental publications, vol. 58). Cambridge: Cambridge University Press. Lindner, R. 1982. “What Was a Nomadic Tribe?” Comparative Studies in Society and History 24: 689–711. Maas, M. 2012. “Barbarians in Late Antiquity: Problems and Approaches.” In The Oxford

520

fisher

Handbook to Late Antiquity, edited by S. Johnson, 60–91. Oxford: Oxford University Press. Matthews, V. 1978. Pastoral Nomadism in the Mari Kingdom (ca. 1830–1760b.c.). (Dissertation series: American Schools of Oriental Research, vol. 3). Cambridge, ma: American Schools of Oriental Research. Millar, F. 2009. “Christian monasticism in Roman Arabia at the birth of Mahomet.” Semitica et Classica 2: 97–115. Modéran, Y. 2003. Les Maures et l’Afrique romaine (ive–viie siècle). (Bibliothèque des écoles françaises d’Athènes et de Rome, vol. 314). Rome: Ecole française de Rome. Mundy, M. and B. Musallam (eds.). 2000. The Transformation of Nomadic Society in the Arab East. (University of Cambridge oriental publications, vol. 58). Cambridge: Cambridge University Press. Paul, J. 2003. “The State and the Military—A Nomadic Perspective.” In Militär und Staatlichkeit. Beiträge des Kolloquiums am 29. und 30.04.2002, edited by I. Schneider, 25–68. (Orientwissenschaftliche Hefte 12; Mitteilungen des sfb “Differenz und Integration” vol. 5). Halle. al-Rasheed, M. 1987. “The Process of Chiefdom-Formation as a Function of Nomadic/ Sedentary Interaction. The Case of the Shammar Nomads of North Arabia.” Cambridge Anthropology 12: 32–40. Salzman, Ph. 1974. “Tribal chiefs as middlemen: the politics of encapsulation in the Middle East.” Anthropological Quarterly 47/2: 203–210. Salzman, Ph. (ed.). 1980. When Nomads Settle. Processes of Sedentarization as Adaptation and Response. New York: Prager. Salzman, Ph. 1983. “Why Tribes Have Chiefs: A Case from Baluchistan.” In The Conflict of Tribe and State in Iran and Afghanistan, edited by R. Tapper, 262–283. London: Croom Helm. Salzman, Ph. 1996. “Peasant Pastoralism.” In The Anthropology of Tribal and Peasant Pastoral Societies: The Dialectics of Social Cohesion and Fragmentation, edited by U. Fabietti and Ph. Salzman, 149–166. Pavia: Collegio Ghislieri. Salzman, Ph. 2004a. Black Tents of Baluchistan. Washington dc: Smithsonian. Salzman, Ph. 2004b. Pastoralists. Equality, Hierarchy, and the State. (Alexander Street Anthropology). Boulder, co: Westview Press. Salzman, Ph. 2008. Culture and Conflict in the Middle East. Amherst, ny: Humanity Books. Salzman, Ph. 2014. “The Meeting of the Twain: Tribe and State.” In Inside and Out. Interactions Between Rome and the Peoples of the Arabian and Egyptian Frontiers, edited by J. Dijkstra and G. Fisher, 79–88. (Late antique history and religion 8) Leuven: Peeters. Sarris, P. 2011. Empires of Faith: The Fall of Rome to the Rise of Islam, 500–700. (Oxford history of medieval Europe). Oxford: Oxford University Press.

reflections on arab leadership in late antiquity

521

Seston, W. and M. Euzennat. 1961. “La citoyenneté romaine au temps de Marc Aurèle et de Commode d’après la Tabula Banasitana.” Comptes Rendus de l’Académie des Inscriptions et Belles Lettres: 317–323. Shahîd, I. 1984. Rome and the Arabs. A Prolegomenon to the Study of Byzantium and the Arabs. Washington, dc: Dumbarton Oaks. Shaw, B. 1982. “‘Eaters of Flesh, Drinkers of Milk’: The Ancient Mediterranean Ideology of the Pastoral Nomad.” Ancient Society 13/14: 5–31. Strathern, A. 1983. “Tribe and State: Some Concluding Remarks.” In The Conflict of Tribe and State in Iran and Afghanistan, edited by R. Tapper, 449–453. London: Croom Helm. Szuchman, J. (ed.). 2008. Nomads, Tribes, and the State in the Ancient Near East. (University of Chicago Oriental Institute seminars, vol. 5). Chicago: Oriental Institute of the University of Chicago. van der Steen, E. 2008. “Tribal Societies in the Nineteenth Century: A Model.” In Nomads, Tribes, and the State in the Ancient Near East, edited by J. Szuchman, 105– 118. (University of Chicago Oriental Institute seminars, vol. 5). Chicago: Oriental Institute of the University of Chicago. Tapper, R. (ed.). 1983. The Conflict of Tribe and State in Iran and Afghanistan. London: Croom Helm. Tapper, R. 1990. “Anthropologists, Historians, and Tribespeople on Tribe and State Formation in the Middle East.” In Tribes and State Formation in the Middle East, edited by Ph. Khoury and J. Kostiner, 48–73. Berkeley: University of California Press. Velud, C. 2000. “French Mandate Policy in the Syrian Steppe.” In The Transformation of Nomadic Society in the Arab East, edited by M. Mundy and B. Musallam, 63– 81. (University of Cambridge oriental publications, vol. 58). Cambridge: Cambridge University Press. Watt, W.M. 1961. Islam and the Integration of Society. London: Routledge. Whittow, M. 1999. “Rome and the Jafnids: Writing the History of a 6th-c. Tribal Dynasty.” In The Roman and Byzantine Near East: Some Recent Archaeological Research, vol. 2, edited by J. Humphrey, 207–224. Portsmouth, r.i.: Journal of Roman Archaeology. Wickham, Ch. 2005. Framing the Early Middle Ages: Europe and the Mediterranean, 400– 800. Oxford: Oxford University Press.

chapter 26

A Paradise in the Desert: Iram at the Intersection of One Thousand and One Nights, Quranic Exegesis, and Arabian History Orhan Elmaz

The tales from One Thousand and One Nights consolidated some long-standing ideas and preconceptions of the Muslim world to the European reader and introduced some other, novel ones. This paper, however, will look into the creation of one of the tales by analysing references to the Arabian tribes of ʿĀd and Thamūd and the riches of the legendardy king Shaddād b. ʿĀd and his proverbial and paradisiacal city called Iram. Since Iram is mentioned in the Quran, setting off from Quranic exegesis and early Arabic poetry, this paper will also discuss how Iram developed into a motif in Islamicate literatures, and address the question of how the embellished tale of The City of Many-columned Iram made its way into the Nights. Although the goal of this paper is not to identify the geographical location of Iram or—as Lawrence of Arabia called it—‘Atlantis of the Sands’. There have been enough suggestions to do so, and while some locate it in Damascus in Syria and others in Alexandria in Egypt or Ubar or Wabar in the Sultanate of Oman (see Fiennes 1992, and Clapp 1999), Harold Glidden suggested identifying Iram with Jabal Ramm in Jordan already in 1939. It is worth mentioning that there is epigraphic corroboration from a Hismaic text found by Zayadine and Farès-Drappeau in 1997 that suggests to locate Iram and the ʿĀd around Wādī Ramm in northern Arabia (see Healey 2001:56ff.) notwithstanding some uncertainty (Macdonald 2000:73n141, 76n171). As of 2011, the International Council on Monuments and Sites (icomos) considered this evidence as arguable theory rather than scientific proof, asking for more substantiation for this identification (icomos 2011:45), but the area is protected and listed as the unesco World Heritage Site “Wadi Rum” since. Wādī Ramm also regained some of its fame since 2014 through the British-Jordanian film Theeb, Jordan’s first Oscar nomination, and the centennial anniversary of the Great Arab Revolt of 1916, during which the film takes place. The ʿĀd are mentioned in the Quran which refers to them 24 times, mostly listing their name along that of other annihilated people’s, while longer passages dealing with the ʿĀd and their postdiluvian prophet Hūd are to be found

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

a paradise in the desert

523

in q 11:50–60, and q 26:123–140. The point of departure for this paper is the passage q 89:6–8 in which Iram occurs. It seems that a whole embellished narrative has evolved or reemerged around this enigmatic name (see Elmaz 2011:23) and that this narrative yet again was reduced into a recurring reference (talmīḥ) with different connotations in Arabic, Persian, Ottoman Turkish, and Urdu poetry. The verse group q 89:6–8 reads: ʾa-lam tara kayfa faʿala rabbuka bi-ʿĀdin (6) ʾIrama dhāti l-ʿimādi (7) allatī lam yukhlaq mithluhā fī l-bilādi (8). Abdel Haleem translates “Have you considered how your Lord dealt with [the people of] ʿAd, of Iram, [the city] of lofty pillars, whose like has never been made in any land” (q 89:6–8). In doing so, he turns Iram into a unique place to which the ʿĀd relate or which they inhabit, and this is to be found in numerous other translations (see The Parallel Quran) with the exception of al-Hilālī & Khan (1998), who understand the ʿĀd to be of unique build and translate: Saw you (O Muhammad ‫ )صلى الل ّٰه عليه وسلم‬not how your Lord dealt with ʿÂd (people) Of Iram (who were very tall) like (lofty) pillars, The like of which were not created in the land? This variation in meaning strongly suggests turning to Quranic exegesis in order to try and understand the connotations of these verses. It is worth noting that a native of Mosul, Abū Bakr al-Naqqāsh (d. 962), dedicated a whole exegetical work of unknown length entitled Iram dhāt al-ʿimād in which he dealt with Iram and past nations (ʿUthmān 2009:184, Bakıxanov’s (d. 1847) history of Azerbaijan bears a similar name). Neuenkirchen recently undertook the task of reading this passage as a Biblical echo by proposing the correspondence of Iram to a Biblical Ḫīrām,1 the self-deifying king of Tyre in the Midrash. He considers the passage in the Quran to be an appropriation of the story of Ḫīrām into Arabian culture, and the parallels might be notable. However, aside from the very problematic sound change (Neuenkirchen 2013:690f.), this hypothesis is refutable based on the verse’s syntax: Iram is not separated from ʿĀd by the conjunction wāw (like the subsequent Thamūd and Firʿawn) in any of the documented readings (see Corpus Coranicum), and dhāt is of feminine gender which does not befit a male individual (Neuenkirchen 2013:692–694).

1 A legend about Ḫīrām who built Solomon’s Temple is also relevant to the genesis of free masonry since he was killed for not divulging the Master Masons’ secret passwords, see Frank 1987: 240 f.

524

elmaz

Nevertheless, Neuenkirchen gives a good overview of exegetical developments regarding Iram in the first part of his article. Since following up on the origin of the Iram myth lies outside the scope of this paper, reference will only be made to the Quranic commentary of polymath al-Ṭabarī (d. 923) who offers several possible interpretations of these verses. Among them is the view that Iram can be interpreted as a place name— with possible identifications as Alexandria or Damascus, or rather al-Shiḥr in Hadhramaut.2 Yet, as a word, ʾiram could mean “people (or nation)” (umma), “old” (al-qadīma), or “perishing” (from the verb ʾarama, yaʾrimu, ʾarm “to take away, remove”). Furthermore, Iram could denote a tribe (qabīla) of the ʿĀd people, which would derive from Iram as the name of an ancestor ( jadd) of the ʿĀd, a descendant of Noah (see also Wheeler 2002:65 f.). Yet, al-Ṭabarī prefers Iram as a toponym or a demonym to anything else, and clearly prefers the ethnonym for a tribe of the ʿĀd (al-Ṭabarī 1994 vii:516 f.). Regarding the “pillars” (ʿimād), al-Ṭabarī relates that they could describe the height or the strength of the ʿĀd, or they could also refer to the nomadic character of the ʿĀd if understood as “[tent]-poles”. Overall, al-Ṭabarī prefers and suggests the view that the ʿĀd of Iram [a branch of the ʿĀd; cf. “first ʿĀd” in q 53:50] were of unique build. The historian Ibn Khaldūn also holds this view in his most significant work, al-Muqaddima, in which he refutes circulating fictitious stories (Goodman 1999:205f.) about Iram in Quranic exegesis: Others go so far in their crazy talk as to maintain that the city lies hidden from sensual perception and can be discovered only by trained (magicians) and sorcerers. All these are assumptions that would better be termed nonsense. rosenthal 2005:17 f.

In q 6:129, the ʿĀd are said to have built fortresses, hoping to become immortal, and in q 41:16 they are portrayed as too proud about their strength, obviously forgetting that God is stronger than them and in a position to annihilate them for their disbelief in him and for rejecting their prophet Hūd. Their extermination (qatl ʿĀd) has become proverbial, and can be found in a hadith of Saʿīd (b. Jubayr) who said “if I reach them, I will not leave one of them alive (lit. la-ʾaqtulannahum qatla ʿĀd “I will kill them the way ʿĀd were killed”)”, and another three times in al-Ṭabari’s History as qatla ʿĀd wa-Iram (al-Ṭabarī

2 Neuenkirchen (2013:665) seems to have overseen this information in al-Ṭabarī’s exegesis since he claims that Iram as a Yemeni place name does not “appear in exegeses strictu sensu”.

a paradise in the desert

525

1968–1971 ii:354, v:124, vi:95), and a few times in al-Ṭabari’s Tafsīr, for example in the section about q 2:89. This phrase has recently been cited in Jihadist debates concerning criticism of isis (Islamic State of Iraq and Syria) which are considered by some Jihadi ideologists as Khawarij who should be exterminated according to the Prophet Muhammad (see Heller 2016). However legendary a place Iram might have been, there are many references to Iram and ʿĀd in pre-Islamic and Islamic poetry.3 The early anthology of mostly pre-Islamic poetry, the so-called Mufaḍḍaliyyāt, contain corroboration of Iram in pre-Islamic poetry: a verse, which has been attributed to Ṣuraym b. Maʿshar al-Taghlibī who probably lived in the second half of the sixth century (Sezgin 1996:150) and who is better known as Ufnūn al-Taghlibī. Lyall has edited the verse in question as law annanī kuntu min ʿĀdin wa-min Iramin / rubbītu fīhim wa-Luqmānin wa-min Jadani, and translated it as “If I had been of the race of ʿĀd and of Iram, brought up among them, and of Luqmān and [Dhū] Jadan” (see Lyall 1921–1924 i:525 v.4, and ii:203 v.4). This verse is of utmost significance, as it demonstrates that Iram is used to identify a group of people as do ʿĀd, and [qawm] Luqmān and [qawm Dhī] Jadan. The poet is ultimately lamenting his non-noble origins as he was not born into any of these tribes. The mukhaḍram poet Labīd (d. 661) who composed poetry in pre-Islamic times and lived to see the Prophet Muḥammad, has a short poem rhyming on d (quḍiya l-umūru wa-unjiza l-mawʿūdu) which must date well after Muhammad’s emigration to Medina (and Labīd’s conversion to Islam), in which he praises God and talks about his advanced age and weakness. The third verse of this poem contains a reference to Iram and ʿĀd along with Thamūd: wa-laqad balat ʾIramun wa-ʿĀdun kaydahū / wa-la-qad balathu baʿda dhāka Thamūdu “Iram and ʿĀd tasted His punishment, and Thamūd did so later”. (see Sharḥ dīwān Labīd b. Rabīʿa al-ʿĀmirī 1962:34 n.5, v.3; cf. Bravmann 1972:84n2). This verse demonstrates more clearly that Iram is understood to be an ethnonym rather than a toponym. The same can be observed in al-Būṣīrī’s (d. 1294) Mantle Ode—in praise of Prophet Muhammad for curing the Sufi poet of paralysis in his dream by wrapping him in a cloak—in v. 90ff. (Sperl 1996:401): 90 91

Yet eulogy can never hope to fathom / The noble traits and virtues that were his: Signs of truth from the All-merciful, both newly formed, / And, as attributes of the Eternal One, eternal,

3 Horovitz refers to a couple of verses in which Iram occurs (see Horovitz 1926:89f.).

526 92

93

elmaz

Of timeless import, giving news of Judgement Day, / And of the days of Iram and ʿĀd (lam taqtarin bi-zamānin wa-hiya tukhbirunā / ʿani l-maʿādi wa-ʿan ʿĀdin wa-ʿan Irami), Remaining ours for ever and so surpassing / All former prophets’ wonders which came but lasted not

In Shawqī’s modern mantle ode, the nasīb closes on Iram (Pinckney Stetkevych 2010:121): lam aghshu maghnāki ʾillā fī ghuḍūni karā / maghnāki ʾabʿadu li-lmushtāqi min ʾIrami “I never visited your place except in my sleep / Your place, for him who yearns for you, is further away than Iram.” This is by far not the only reference to Iram in modern Arabic literature where Iram is mostly if not solely used as a mythical place like in Adonis’s Aghānī Mihyār al-Dimashqī (1961; Mihyar of Damascus: His Songs). Jubran Khalil Jubran’s last major work in Arabic, the prose drama Iram dhāt al-ʿImād (1921, see Bushrui 2010:186; Iram, City of Lofty Pillars),4 is believed to be the first introduction of Iram into modern Arabic literature. Jubran uses mystical language and imagery to reflect the Sufi concept of the unity of being (waḥdat al-wujūd; Bushrui & Jenkins 1998:212–216) and tells us of a prophetess called Āmina who has entered the wondrous city of Iram. Similarily, Nasīb ʿArīḍa (d. 1946) sings about his longing to unify with God using Iram as a symbol of the pursuit towards an ideal or the divine (Durakovic 1997:68) and “pining for what is not” (Abu Haidar 1996:13) in his poem ʿAlā Ṭarīq Iram (1925; On the Way to Iram) and alMusāfir (The Traveler; see Jubran 2007:71f.). There are many more examples in contemporary Arabic poetry, for instance the Saudi poet Muḥammad alDuḥaymī, one of the foremost contemporary poets from the Gulf, mentions in Wǝsh b-ṣadri?! (What is in my heart?!, also entitled as “Qaṣīdat Shahrazād”, see al-Duḥaymī) that the history of Iram dhāt al-ʿimād is still buried in geography. Another narrative that refers to Iram not as an ethnonym but as a personal name is more comparable to pre-Islamic and Classical Arabic poetry. It is the story of the two legendary soothsayers Saṭīḥ and Shiqq, who have to describe Rabīʿa b. Naṣr’s alarming dream to him in order to gain his trust before interpreting it. This dream is meant to explain the establishment of the (ethnically South Arabian) Lakhmid kingdom in Iraq following the Abyssinian invasion. They prophesy a certain Iram of Dhū Yazan to come from Aden and to not leave a single Abyssinian in Yemen, though his dominion shall be cut short with the

4 Davide Cincis adopted Jubran’s drama into the short film Iram, la città dalle alte colonne in 2005, which won the Critics Award at the akab Short Movie Festival in 2006.

a paradise in the desert

527

advent of a prophet [from among the Quraysh, i.e. Prophet Muhammad] (The History of al-Ṭabarī 1999:178–182). Based on these sources, it seems that Iram is a tribal name or a personal name rather than the name of a city buried in the sands of South Arabia, especially in older literature. But how and when does this myth of a City of Lofty Pillars emerge? Khan (2015:109) suggested that Iram became known to the Western world through the Arabian Nights. Strangely, the first translation into a European language, which is Galland’s translation—or rather ‘edition’ as it is not very accurate throughout (see Osigus & Grotzfeld 2005:50)—does not include any reference to South Arabia. More specifically, it does not include stories,5 which feature Yemen or Iram or the tribes of ʿĀd or Thamūd (see Table 1). Yet, Galland includes 61 Tale of Qamar al-Zamân and Budûr but he does not refer to the great treasure which Qamar al-Zamân, right after finding the lost jewel, comes across as being from “the time of ʿÂd and Thamûd”. The comparison with Burton’s translation in Table 2 shall illustrate this significant difference. The question that arises is: why would Galland omit mentioning ʿĀd and Thamūd, when the Nights lives on mystery and referring to them would take the reader back more than a millenium? Galland should have been familiar with the myth around Iram,6 for he completed d’Herbelot de Molainville’s monumental Bibliothèque orientale, ou dictionnaire universel contenant tout ce qui regarde la connoissance des peuples de l’Orient in 1697 which contains entries on “Ad” (Herbelot 1697:51f.), “Iram” (Herbelot 1697:498), “Scheddad” (Herbelot 1697:780), and “Thamoud” (Herbelot 1697:1023). Therefore, it seems likely that Burton’s and Galland’s translations might rely on different sources, for the story 70 The City of Many-columned Iram is not only absent from Galland’s translation. It is also not contained in Mahdi’s edition, the Wortley-Montague manuscript (1764), Habicht’s edition (1825), the Reinhardt manuscript (1831) and Mardrus’s translation (1899–1904). Similarly, a tale about another legendary city, the city of Shaddād b. ʿĀd’s son Qūsh (cf. Arabian Nights Encyclopedia

5 Three of the stories given above and which refer to Yemen in one way or another (178 The Story of Jânshâh, 210 History of Gharîb and His Brother ʿAjîb, 357 Sultan of al-Yaman and His Three Sons) have been claimed to be of Jewish origin (see Gottheil & Jacobs 1902; Chauvin 1899:16, 19 f., and Perles 1873:61–69). These and some other stories have lead Chauvin to believe that the author of the late Egyptian recension of One Thousand and One Nights was Jewish but this theory has only a few adherents today (see Sadan 2004). 6 A later and extended edition of Galland’s translation by Jules Janin (Paris 1839) does include a story about Scheddad and how he built Iram (dxlve nuit in vol. iv: 442–445).

528

elmaz

table 1

Story in Burton’s translation 39

Burton’s source

The Tale of King ‘Umar ibn al-Nuʿmân and His Sons Sharrkân and Daw’ al-Makân The City of Many-columned Iram and ʿAbdullâh ibn Abî Qilâba The Man of al-Yaman and His Six-Slave girls The Vizier of al-Yaman and His Young Brother How Abû Hasan Brake Wind ʿAlî the Cairene and the Haunted House in Baghdad The Story of Jânshâh The City of Brass History of Gharîb and His Brother ʿAjîb Sayf al-Mulûk and Princess Badîʿat al-Jamâl Maʿrûf the Cobbler and His Wife Fâtima

Burton from the second Calcutta edition

342

Tale of King Ins ibn Qays and His Daughter with the Son of King al-ʿAbbâs

Burton from the Breslau edition

357 375 385

Sultan of al-Yaman and His Three Sons Story of the King of al-Yaman and His Three Sons History of al-Hajjâj ibn Yûsuf and the Young Sayyid

Burton from the WortleyMontague manuscript

423

The Story of ʿAlî Jawharî

Habicht translation

450 485

The Keys of Destiny Men in the Judgement of Their Wives

Mardrus translation

509

The Story of the Two Viziers and Their Children

524 526 536

Alexander the Great and the Search for the Water of Life The Story of King Sabâ The Story of Sayf ibn Dhî Yazan

70 84 108 144 155 178 180 210 229 262

Wortley-Montague manuscript

Reinhardt manuscript

2004:146), namely tale 180 The City of Brass, is missing from Galland’s translation, Mahdi’s edition, the Wortley-Montague manuscript (1764) and Habicht’s

a paradise in the desert

529

table 2

Camaralzaman prit une cognée, & alla mettre la main à l’œuvre. Comme il coupoit une branche de la racine, il donna un coup sur quelque chose, qui résista, & qui fit un grand bruit. En écartant la terre, il découvrit une grande plaque de bronze, sous laquelle il trouva un escalier de dix degrés. Il descendit aussitôt, & quand il fut au bas, il vit un caveau de deux à trois toises en carré, où il compta cinquante grands Vases de bronze, rangés à l’entour, chacun avec un couvercle. Il les découvrit tous l’un après l’autre, & il n’y en eut pas un qui ne fût plein de poudre d’or. Il sortit du caveau, extrêmement joyeux de la découverte d’un trésor si riche, remit la plaque sur l’escalier, & acheva de déraciner l’arbre, en attendant le retour du jardinier. (Galland 1704–1717:279–281)

At daybreak he rose to his work and, girding his middle with a cord of palm-fibre, took hatchet and basket and walked down the length of the garden, till he came to a carob-tree and struck the axe into its roots. The blow rang and resounded; so he cleared away the soil from the place and discovered a trap-door and raised it. And Shahrazad perceived the dawn of day and ceased saying her permitted say. Now, when it was the Two Hundred and Fourteenth Night, she said, It hath reached me, O auspicious King, that when Kamar al-Zaman raised the trap-door, he found a winding stair, which he descended and came to an ancient vault of the time of Ad and Thamúd, hewn out of the rock. Round the vault stood many brazen vessels of the bigness of a great oil-jar which he found full of gleaming red gold: whereupon he said to himself, “Verily sorrow is gone and solace is come!” Then he mounted from the souterrain to the garden and, replacing the trap-door as it was before, busied himself in conducting water to the trees till the last of the day, when the gardener came back … (Burton 1885–1888 iii:294)

edition (1825). Interestingly, there are manuscript story-collections in which the tale of Iram follows that of the City of Brass directly which might be due to the redactors (see Pinault 1992:211). However, although both ʿĀd related stories are missing from Mahdi’s edition, it does include a paraphrase of q 89:6–10 right at its very beginning, praising and characterising God as the one “who annihilated the Thamūd and ʿĀd and Pharaoh ‘of the stakes’” (wa-ʾahlaka qawm Thamūd wa-ʿĀd wa-Firʿawn dhū awtād, see Mahdi 1995) in its initial praise formula. It is the tale of The City of Many-columned Iram that offers the most elaborate description of the wealth and grandeur of Shaddād’s creation of Paradise on Earth (for a summary of different narratives including details, see Hamdouni Alami 2011:149–155). This tale features ʿAbdallāh b. Abī Qilāba who goes to look for his lost camel in the Yemeni desert and comes across an abandoned city built out of gold and silver and various precious stones. In order to prove

530

elmaz

what he experienced was real, he takes some of the riches and goes directly to the Caliph, Muʿāwiya at that time, and reports to him. The Caliph asks for Kaʿb al-Aḥbār to confirm the credibility of ʿAbdallāh’s report and he tells him about Iram and that one person would find it one day and that that person—and he points at ʿAbdallāh—is him. It is worth noting here that Kaʿb al-Aḥbār is also credited with localising Iram in South Arabia (Piotrovskij 1977:19). This tale must have had a major impact in shaping people’s imagination of Iram as a paradisiacal land of plenty, and Iram has produced a huge body of literature about its very nature as well as its location, which al-Rubayʿī (2000) discusses at a large scale. How this tale made its way into the field of Quranic exegesis is unclear, however, it seems likely that the earliest exegesis that contains this tale in an elaborate form is that of al-Ṭabarānī (d. 971, see Neuenkirchen 2013:675). The material itself goes back to Wahb b. Munabbih (d. between 725 and 737), a prolific writer on prophets and their stories (cf. Lebedev 1973:104), and the tale of ʿAbdallāh b. Qilāba, wonderful as it is, fits well in the collection of the Nights. Lane expressed his doubts about some of the material (Lane 1839–1841 ii:331n117): “Here it becomes necessary to mention, that many anecdotes are inserted among the longer stories of the Thousand and One Nights. They are chiefly, I believe, extracted from other works, and many are historical; but most of them are very inferior in interest to the longer stories.” Thus, he almost solved the mystery around Iram and how it made its way into post-Galland editions and translations of the Nights. To my knowledge, nobody has mentioned that the tale of Many-Columned Iram actually contains two marked quotations: one by a certain al-Shaʿbī, and one by al-Thaʿlabī (d. 1035/1036), both introduced by qāla. In fact, the whole text of this tale is to be found almost verbatim in Kharīdat al-ʿajāʾib wafīrat al-gharāʾib by Sirāj al-Dīn b. al-Wardī (d. 1457)7 in his discussion of al-Aḥqāf (al-Wardī 2008:154–159), the dwelling place of ʿĀd according to the Quran (q 46:21). His version of the tale is a little longer and contains a couple of lines more poetry. In addition, it does not have the mistakes that the second edition of Alf layla wa-layla published by Dār Ṣādir in Beirut in 2008 still comes with. At the beginning of the tale in the 276th night (Alf Layla wa-layla 2008: i.505), one reads for instance sallaytu nafsī (Burton: “composing my mind”; colloquial for salaltu; Ibn al-Wardī: istalaltu) instead of sallaytu sayfī (“I drew my sword”). Following shortly, one reads marʿūb dhāhil

7 Habicht’s edition is quite a different text and it is based on al-Shaʿbī’s Kitāb siyar al-mulūk, cf. al-Ibshīhī’s (d. 1466) al-Mustaṭraf fī kull fann mustaẓraf (1992 ii:167f.).

a paradise in the desert

531

al-labb instead of marʿūb dhāhib al-labb (Burton: “in a flutter of fear and dazed with surprise and affright”). Both mistakes are also found in the Būlāq edition of 1835 (i.451) and Macnaghten’s second Calcutta edition (1839: ii.142). Lebedev, who studied South Arabian traces and influences on the Nights, suggested that tales with references to South Arabia were incorporated in Egypt as early as the 9th century (Lebedev 1973:112). For the inclusion of the tale of Many-Columned Iram one has to suggest a late date for its inclusion into the corpus, most probably after Ibn al-Wardī and after Galland but before the first Būlāq edition of 1835 and the Breslau edition of 1825–1843. Both editions contain the legend around Shaddād’s replication of Paradise on Earth, however, latter edition focuses mostly on the description of the building of the paradisiacal and sybartic city. ʿAbd Allāh b. Qilāba al-Anṣārī is only mentioned towards the end as the only person who has ever found Iram while looking for his lost camel in the vicinity of ʿAdan (Breslau 1825–1843: vii.174). Nomen est omen? In view of the educational context of the first Calcutta edition entitled Ḥikāyāt miʾat layla min alf layla wa-layla (cf. Mahdi 1995:91), a likely candidate would have been its editor, Shaykh Aḥmad b. Muḥammad Shīrwānī al-Yamanī (“Shuekh Uhmud bin Moohummud Shirwanee ool Yumunee”), a Yemeni teacher of Arabic who had headed to Calcutta from Egypt. However, neither does Iram feature in this extensively edited version of the Nights for the purpose of teaching and learning Arabic, nor does his anthology of literature Nafḥat al-Yaman which was published in 1841 contain any legendary material about Shaddād b. ʿĀd or Iram. It is certain, however, that somebody must have edited a version of the tale the way it is told in Ibn al-Wardī’s book, Kharīdat al-ʿajāʾib wafīrat al-gharāʾib, including the quotations that it comes with, for inclusion in the corpus of the Nights. This must have then indeed taken place in Egypt, as Lebedev assumed, for the tale of ʿAbdallāh b. Qilāba was marvellous and stunning enough to extend the corpus to comprise tales for literally One Thousand and One Nights. This might also explain, why unlike Galland’s translation, even a cursory look at Burton’s monumental translation, which is based on the second Calcutta edition, reveals various references to South Arabia. In Ali the Persian, a Kurdish boy took his wallet and claimed ownership of all of its contents. In the following entertaining disputes about the contents of the bag in front of a judge, not only Shaddād’s and Khosrow i’s (Anushirwan) palaces are mentioned as items but also Many-Columned Iram; to name but a few of the other items found between China and Sudan, the wallet is claimed to contain “a thousand rogues and pimps and horse-courses and stables and mosques and baths and a builder and a carpenter and a plank and a nail” (Burton 1885–1888 iv:149–152, 294th– 296th night).

532

elmaz

The History of Gharīb and his Brother ʿAjīb features Gharīb meeting a 340year-old man from the tribe of the ʿĀd8 in a cave. The man had lived to see the ʿĀd and the Thamūd perish but was spared punishment for their disbelief himself. He invites Gharīb to Abrahamitic monotheism which the latter embraces by pronouncing: “There is no God but God and Abraham is the Friend of God.” (Burton 1885–1888 vi:269 627th night). Here Burton makes a very remarkable note that shows how interested early Orientalists were in Arabian history and folklore: The First Adites, I have said, did not all perish: a few believers retired with the prophet Hud (Heber?) to Hazramaut. The Second Adites, who had Marib of the Dam for capital and Lukman for king, were dispersed by the Flood of Al-Yaman. Their dynasty lasted a thousand years, the exodus taking place according to De Sacy in a.d. 150–170 or shortly after a.d. 100 (C. de Perceval), and was overthrown by Yaʿarub bin Kahtán, the first Arabist; see Night dcxxv. In Sayf al-Mulūk and Badīʿat al-Jamāl (Burton 1885–1888: vii.331–334,336,370– 372, viii.3), Sayf al-Mulūk falls in love with a portrait of Badīʿat al-Jamāl, the daughter of the King of the Kings of the jinn, who live in the city of Babel and sojourn in the garden of Iram, Son of ʿĀd the Greater. Because nobody knows where the garden of Iram is, Sayf al-Mulūk decides to set out and search for his beloved by himself. He goes through several adventures, which are quite similar to those of Sindbad the Seaman, and finally they meet by ways of magic in Sarandîb (hence we speak of “serendipity”), and then depart for the garden of Iram together. One should note here that in some versions of the epic of the same name, Köroġlu, the hero of Central Asian epics about a historical Anatolian rebel of the Jalālī resistance movement against the Safavid Shah ʿAbbās i (1581–1628), marries two fairies (peri in Turkic languages from Persian) which he brings— or in Turkmen and Uzbek versions: abducts (see Reichl 1992:156)—from the Garden of Iram (Korogly 1995:304, 306) or Iranbak (for Iram-bag) in a Kazakh version (see Karryev 1968:242). This seems to be a projection of the Quranic Houris into the earthly paradise and reflects the idea of a garden of fairies, and indeed, in Kazakh folklore Irambaq is “a wonderful and mysterious place, a land

8 It is worth noting that there is a miniature in a Timurid manuscript of Persian poet Neẓāmī’s (d. 1209) Khamseh showing Alexander visiting a hermit in a cave beneath a fortress that is identified as Iram (see Renard 1999:39).

a paradise in the desert

533

of beautiful girls and fairies” (tañğajayıp, jumbaq jer, sulw qızdar men periler mekeni, Qıraubaeva 2010:42). However, the liaison with a fairy can have terrific outcomes like Depegöz, the terrific cyclops of the Turkic epic Dede Korkut with strong parallels to Polyphemus (see Dede Korkut 2013:123 f.). In Hasan of Bassorah (Burton 1885–1888 viii:30, 786th night), a certain Hasan ends up at a palace after the evil Magian Bahrām abandoned him to his fate. He has not seen anything like it and wonders whether it is Many-Columned Iram shrouded in legend that he has found. Similar to the extermination phrase mentioned above (qatl ʿĀd wa-Iram), in Abdullah the Fisherman and Abdullah the Merman, a king damns the syndic of the jewellers “with the damnation of Ád and Thamúd” (Burton 1885–1888 ix:174 943rd night). The Thamūd were given a she-camel as a sign of God but they slaughtered it instead of allowing it to graze and drink water in peace (cf. q 7:73–79, q 11:61–68, q 26:141–159). Even this camel features the Nights as it is mentioned in a riddle within the story of Abu al-Ḥusn and his Slave-Girl Tawaddud (Burton 1885–1888: v.235 457th night): “What are the five that ate and drank, yet came not out of loins nor womb?”—“Adam, and Simeon, and Salih’s she-camel, and Ishmael’s ram, and the bird that Abu Bakr the Truth-teller saw in the cave.” Maʿrūf, a hard-working cobbler living in Cairo, finds a seal-ring in the story of Maʿrūf the Cobbler and his wife Fāṭima. When he rubs the ring, a jinni appears and explains that this was the seal-ring of Shaddād, who built Many-Columned Iram. The jinni serves whoever rubs the ring and as one can predict, suspense is achieved throughout the story by changing ownership of the ring. (Burton 1885–1888: x.29 996th night). It is worth noting that in Abu Mohammed Hight Lazybones, Burton identifies the City of Brass that is referred to cursorily as the “Moroccan City”9 and mentions that India had an Iram, too, using Iram as the epitome of a lost city like Atlantis (cf. Burton 1885–1888: iv.176 304th night). In his terminal essay, Burton (1885–1888: x.159) further explicates this idea: And the very depths of melancholy, of majestic pathos and of true sublimity are reached in Many-Columned Iram (vol. iv. 113) and the City of Brass (vol. vi. 83): the metrical part of the latter shows a luxury of woe; it is one long wail of despair which echoes long and loud in the hearer’s heart.

9 It seems to be accepted that the “City of Brass” type stories belong to the Maghrib, see Blatherwick 2016, p. 203 f.

534

elmaz

Likewise, other translations and editions than Burton’s include different references, for instance in the story of Jûdar and the Moor Mahmûd in Weil’s edition, the palace of Shaddād b. ʿĀd, the founder of many-columned Iram, is described as the biggest and most beautiful palace on Earth (Weil 1837–1841 iv:670 915th night). In Habicht’s edition, ʿAlī Jawharī’s son escapes by the skin of his teeth from being sacrificed to idols in China, when a demon suddenly takes him to an underground palace which resembles the ‘Paradise of Shaddād’: er fasste sie hierauf in seine Arme, die Erde öffnete sich unter ihren Füßen, und sie befanden sich plötzlich in einem prächtigen Palast, dessen Glanz nicht seinesgleichen hatte: Denn er war von Edelsteinen erbaut, ruhte auf Säulen von Smaragd und war von weiten Gärten umgeben, welche denen des Paradieses von Schedad glichen. habicht 1825 xvi:116 519th night

Reinhardt’s manuscript contains the Story of King Sabâ, which features a city in the Maghrib which the great ʿĀd have built (Reinhardt 1831 i:463b–468a, cf. The Arabian Nights Encyclopedia 2004:354–356), and the Story of Alexander the Great and the Search for the Water of Life, in which Iram appears as a prophet (Reinhardt 1831 i:429a–443b). Latter might be a blurred reference to Persian Alexander romances in which Alexander in his search for the Water of Life comes across the Garden of Iram, for instance in Neẓāmī’s Eskandarnāmeh (see Wilberforce-Clarke 1881). References to the ‘Water of Life’ and the ‘Garden of Iram’ abound in Neẓāmī’s Haft Paykar (see Meisami-Scott 1995) and other works of Classical Persian literature. For instance, they are mentioned together in Ḥafeẓ’s (d. 1390) ghazal Khoshtar ze ʿeysh o ṣoḥbat o bāġ o bahār chist? (“What’s sweeter than a garden and good talk?”, see Davis 2013): “The Water of Life, the Garden of Eram—/ What could these blessings mean | But heart-delighting wine that’s poured and drunk / Beside some pretty stream?” A unique story featuring Iram is the story about The Keys of Destiny in Mardrus’s translation. It mentions a manuscript that is written in a certain script that only one sheikh can read, and it hails from Iram, the marvellous city of Shaddād b. ʿĀd. A harbour to all the riches in the world, Iram is reached in a very spectacular way by human flight.10 One might compare this tale to the King of Black’s Tale of the Princess of the First Clime in Neẓāmī’s Haft Paykar, 10

Human flight is possible only after applying Phoenix blood and a serpent’s brain and heart are rubbed on one’s back and shoulders. Since the episode of a man growing wings and taking off to fly to Iram and landing there is quite spectacular, it shall be given in full: “Et soudain il se leva sur ses deux pieds, se dépouilla les épaules comme le font les pèlerins de

a paradise in the desert

535

where a king travels by way of Sīmorgh, a fantastical bird, to land in a garden that is likened to that of Iram (Meisami-Scott 1995:114 f.). Trying to put the tale of Many-Columned Iram as it is included in the Būlāq edition of 1835 into a category is not a trivial task. Essentially, ʿAbdallāh b. Qilāba gets lost in the desert but this is not a so-called Schreckmärchen, a scare tale, in which children lose their way in a forest or their parents abandon them. It is rather a tale of displacement comparable to those occurring in the Odyssey, the Aeneid, Sindbad the Sailor or the Voyage of Saint Brendan (Scharfe 2011:38b). As to treasure tales, one would expect the protagonist to lay a trail in order to find their way back (Scharfe 2011:40) which is absent in the story of ʿAbdallāh b. Qilāba, either. This makes the story difficult to believe for as a Bedouin looking for his lost camel in the desert, he got lost himself and came across Iram. Having taken some of its treasures to prove he found this lost city, he went right back to the caliph Muʿāwiya. The tale does not contain any information about his points of orientation, how he found his way back, or even whether the events of this story are taking place in daylight or whether he spent the night somewhere in order to rest and try to find his way home the next morning (Scharfe 2011:41). Among modern, colloquial Arabic tales involving getting lost, one can refer to the Geschichte von dem Baum der Verirrten, which Enno Littmann recorded in the Syrian Desert. After having lost her camel in the desert, a girl puts her steering stick into the ground to leave a mark for her father in order to help

la Mecque au départ, et, trempant un bout de sa ceinture dans le sang du Phénix mélangé à la cervelle et au cœur du serpent, il m’ordonna de lui en frotter le dos et les épaules. Et je me mis en devoir d’ exécuter l’ ordre. Et, au fur et à mesure que je le frottais, je vis la peau de son dos et de ses épaules se gonfler et éclater, pour en laisser lentement sortir des ailes qui, grandissant à vue d’ œil, descendirent bientôt jusqu’à terre. Et le Bédouin les agita avec force, à même le sol, et tout d’ un coup, prenant son élan, il s’éleva dans les airs. Et moi, préférant mille morts plutôt que d’ être abandonné en ces lieux sinistres, je fis appel à ce qui me restait de force et de courage, et je me cramponnai fortement à la ceinture de mon maître, dont le bout pendait par bonheur. Et je fus emporté avec lui hors de cette vallée noire d’ où je n’ espérais plus sortir. Et nous arrivâmes dans la région des nuages. Or je ne puis te dire, ô mon seigneur, combien de temps dura notre course aérienne. Mais je sais que nous nous trouvâmes bientôt au-dessus d’une plaine immense dont l’ horizon était au loin fermé par une enceinte de cristal bleu. Et le sol de cette plaine semblait formé de poudre d’ or, et ses cailloux de pierres précieuses. Et au milieu de cette plaine s’ élevait une ville remplie de palais et de jardins. Et mon maître s’ écria: « Voici Aram-aux-Colonnes!» Et, cessant de mouvoir ses ailes, et les étendant largement immobiles, il se laissa descendre, et moi avec lui. Et nous touchâmes le sol, au pied même des murailles de la cité de Scheddad, fils d’Aâd. Et les ailes de mon maître diminuèrent peu à peu et disparurent.” (Mardrus 1899–1904 xii:153f.)

536

elmaz

finding her. After one year or several years, her father finds her and on their way home, she sees that the stick has grown into a tree which became known as the “tree of the lost [girl]” (shajarat al-tāyiha, see Littmann 1916:25). The tale of ʿAbdallāh b. Qilāba has none of the features of tales of getting lost which one would expect. El-Shamy (2004:894f.) has classified the tale of Iram as pertaining to type 1645d “Perilous Journey in Search of Treasure Trove”, although clearly there is no danger involved in the story that is included in the Nights. While Walter Savage Landor’s Gebir (1798) and Robert Southey’s poem Thalaba the Destroyer (1801) feature Iram, and the latter describes the apocalyptic destruction of Shaddād’s city (Sharafuddin 1996:11, 53), in contemporary English expatriate literature, there are two novels that include perilous journeys to discover Iram. One of them is Eric L. Stone iii’s “historical-fictional novel” The Lost City of Iram: A geographic adventure through time (2012), which is about the quest to discover Iram set in the dawn of World War One featuring a young archaeologist named fittingly Thomas Edward Lawrence. A second example is Clive F. Sorrell’s Kawthar: Death Shadows Those Who Seek the Sweet Waters of Paradise (2012) and its protagonists are a married Scottish geologist searching for water on the Arabian peninsula and his wife who buys a pendant on a market which proves to be a map showing the way to Iram. Sorrell locates Iram “on the major spice route from Asia to Babylon and legend has it that it was the wealthiest city in the world” (2012:60), and he elaborates on the city’s beauty (2012:254). He includes a poetic quote from Burton’s translation of the tale of The City of Many-Columned Iram (Burton 1885–1888 iv:255), obviously taking inspiration for his description of Iram and probably for the whole novel from this very tale, which was our point of departure. Bencheikh (1990:78) describes Iram as a “ville-rêve”, a dream city that epitomises a paradise right here and now, hidden in the will of God. In Islamicate cultures, it seems therefore to be common practice to compare gardens and greenery to Iram. For example, Muhammad Iqbal (1877–1938) described the advent of spring in the opening verse of his Saqi Namah with the verse: huā xaimāzan kāravān-e bahār / Iram ban gayā dāman-e kohasār Spring’s caravan has pitched its tent / At the foot of the mountain, making it look like the fabled garden of Iram. see matthews 1993:177, 192

Moreover, the Quranic description of Paradise is the main source of inspiration for arts and architecture, which try to reflect infinity, beauty, calm and peace (Tonna 1990:195f.). Therefore, it is not surprising to find Iram in monu-

a paradise in the desert

537

mental inscriptions using eschatological imagery, which might even represent the apex of this genre. Two inscriptions in Chanderi in the state of Madhya Pradesh in India praise various parts of a complex built during three years of the reign of Sultan Ghiyāth al-Dīn Khaljī of Malwa (r. 1469–1500). The inscription commemorates the builder, minister Ṭaġī, son of Fakhr, and disciple of Shāh Madār (Falasch 2015:73), who constructed “a step-well, a mosque and a garden for the sake of God and for winning the pleasure of His Prophet” (Abdur Rahim 2000:35). Regarding the year of completion, the inscription gives the letters “jīm, ṣād and ḍād”, and the sum of the these letters’ numerical values yields the year 893a.h. (1487–1488ce). The mosque is not extant but the stepwell (bāʾolī in Urdu and Hindi, see Burton-Page 2008:65 f.) with its 32 (battīsī) steps—hence the local name battīsī-bāoŕī—is the largest step-well in Chanderi. The author of the inscription to the right of the entrance to the complex, Ibn Ḥamīd, praises the water of the well as the ‘Water of Life’ referring to Khiḍr and for being as sweet as sugar but still sweeter than sugar from Egypt or China. He goes on describing the well as the sacred well of Zamzam in Mecca and the heavenly river Kawthar, and the garden as Paradise itself. 9 10

11

12 13

14

How can Ibn Ḥamīd sing his praises as are due? And hence, I should now proceed to describe the step-well. The limpid water of Khiḍr (i.e. water of life) was concealed in the world from the eye of man; Now this great chief has produced it in this region. May Allāh grant us its drink! How sweet is the water which is sweeter than sugar-candy! Or perhaps it is the syrup of jujube, since it is as sweet as sugar. The sugar of Egypt and sugar-candy of China are not so sweet. How can sugar-cane come up to its sweetness even though it is sweet? Near the great stream (i.e. step-well), he has also constructed a mosque; This (i.e. mosque) is Kaʿba and that (i.e. well) is Zamzam, and its enclosure is like the verdant garden. All around the sweet smelling garden … flowers on all sides. But perhaps Paradise has appeared in the world for, that is Tuba (lote tree) and this is Kauthar. abdur rahim & desai 1966:68

Having praised this step-well in terms of best-possible qualities in this world and bringing in some eschatological terminology from the Quran, in the other inscription, to the left of the entrance of Battisi Baori, the author of the descrip-

538

elmaz

tion of the complex makes use of more terms pertaining to the Hereafter. The well-water is compared to springs and streams in the Quranic description of Paradise: salsabīl (q 76:18) is a fountain in Paradise from which the righteous will be given to drink ginger-spiced water, while tasnīm (q 83:27) is a fountain from which those brought near will drink.11 The inscription suggests that the mosque in the complex could be mistaken to be al-Aqṣā in Jerusalem, where the Prophet Muḥammad was taken during his night journey before his ascent to Heaven to legitimise his prophethood, the Kaʿba in Mecca, or—last but not least—Iram! Youths of Paradise and proverbial houris might descend onto the mosque while the garden is paradisiacal, bearing roses like imperial gardens in Byzantium or China. Kawthar, a river in Paradise the water of which is “more delicious than honey—is of a brighter whiteness than milk or snow, and runs over precious stones and pearls, with banks of gold and silver” (see Radscheidt “Springs and Fountains”) is jealous of the purity of the well-water, while Zamzam is envious of it: 5 6

7

8

O Lord, is this the stream of bounty or is it the limpid Salsabīl? O fortune, is it flowing water or honey from Tasnīm? Its mosque—is it (the mosque) Aqṣā, or Kaʿba (maʿmūr), or (Iram) the house of pillars (built by Shaddad)? Is it a descending-place for unbearded youths of Paradise (mahbaṭ-e ġelmān), or for its keeper, or for its black-eyed houries? Its garden—is it the meadow of Paradise or the garden of heaven (baġ-e ḫold)? Its rose-garden—is it the rose-garden of Byzantium or China? At times, the heart of the stream of Kauthar turns into water due to jealousy of its pure water; at other times, Zamzam is immersed in perspiration upto its forehead on account of envy. abdur rahim & desai 1966:71

The passages quoted above are taken from rather simple inscriptions and obviously, they are meant to remind people of the Hereafter using positive connotations, and to mark the complex as an oasis of peace and calm for prayer and contemplation. Having exhausted the most prominent earthly places to compare the complex to, the author helped himself to points of comparison in the Hereafter before climaxing in the most significant places al-Aqṣā and Mecca, and sybaritic Iram. This description has many parallels to the later

11

Cf. Radscheidt, Springs and Fountains, eq.

a paradise in the desert

539

praise poem of the ‘Garden of Eram’ at Rayy by the Persian poet Fatḥ ʿAli Khān Ṣabā of Kāshān (d. 1822/23; see Wilber 1962:44–51). Iram features quite a few poems of the Sultan of Ottoman poets, Sultânüşşuarâ Bâḳî (Maḥmūd ʿAbdülbâḳî, d. 1600) as a paradisiacal garden. This, as was mentioned above, can be attributed to Persian influence, and Sanāʾī (d. 1131) gives an early example for the use of the Garden of Iram as a reference to ‘Paradise on Earth’: Rāhī chunān guzāshtam / bāgh-i Iram pindāstham / az ṣabr tukhmī kāshtam / āmad be-bar baʿda l-taʿab | rūz āmade darmān-i man / āsūde az gham jān-i man / zi khayme-i jānān-i man / āmad be-gūsh-i man shaghab I went a path, I deemed it Paradise [lit. the Garden of Iram], of patience I had sown a grain, that bore fruit after much pain | My remedy had come, my soul was free of grief, from the tent of my friend a ting came to my ear. bürgel 2005:306

Among the imagery elements in Bâḳî’s panegyric poems (ḳasîde, abbr. k) and ghazals (abbr. g), there are references to the “rose-garden of Iram” (gülşen-i İrem in k1/38; gülistân-i İrem in g318/2), the “[worldly] garden of Iram” (bâġı İrem; k2/1, k16/11, k27/11, g 67/2, g426/4) as opposed to the heavenly rose gardens (gülzâr-ı cinân), the “garden of Iram” (bûstân-ı İrem; g147/1), and the “peacock of Iram”12 (İrem tâvûsı; k14/3, g414/2, g498/2) that strolls around like the beloved’s curl dangling around their face.13 12

13

Although the image of a ‘peacock of Iram’ looks most original at first, it seems to relate to Aḥmed Pâşâ’s (d. 1497) description of Mehmed the Conqueror’s son, Prince Mustafa (1450–1474), who died in mysterious ways, in his elegy as ḳaṣr-ı İrem kebûteri “dove of the Irem palace” (Alpaslan 1987:12). In one of his gazels, he also speaks of a tûtî-i bâğı İrem “parrot of the Garden of Iram” (g 237/8), and in others, he refers to peacocks. For references to peacocks as legendary birds in Ottoman Turkish poetry, see Eskigün 2006:73 f. The image of an Iram-peacock seems to be unique to Ottoman poetry since in Persian poetry it is probably only found once, in a verse by the last of the classical Persian poets, Qāʾānī (d. 1854): “It is a sin to relate (liken) a snake to the peacock of Iram, / (just as) it is wrong to liken a thorn to the heavenly rose-garden” (Mār rā nesbat-gonah bāshad be ṭāvus-e Eram / khār rā sheb[ā]hat khaṭā bāshad be golzār-e jenān). Unfortunately, Keleṣ has not listed any references to Iram in her overview of citations of sura 89 in Ottoman poetry (2016:267–272). The reference is: Keleṣ, Reyhan. 2016. Divan Ṣiirinde Âyet ve Hadis İktibasları. Istanbul: Kitabevi.

540

elmaz

In order to highlight more unique usages of the Iram motif, a small selection of these verses shall be given in translation below. The Sultan’s beauty turns the world into the Garden of Iram in k1/38; his beauty renders the whole world heavenly beautiful:

k1/38 With the rose of your beauty the world turns into the Garden of Iram

Ḥüsnüñ güliyle bâġ-ı cihân gülşen-i İrem

Everywhere there are a thousand nightingales and a hundred different songs14

Her sû hezâr bülbül ü ṣad gûne dâstân

In k2/1, the Garden of Iram is personified and given agency, only to be portrayed as jealous of the city, which was honoured by the Sultan’s visit. This line is only worthy of a sultan’s arrival, and proved to be the first poem Bâḳî dedicated to Sultan Süleyman (r. 1520–1566) with many more were to follow.

k2/1 The Sultan of the World’s visit made the city

İtdi şehri şeref-i maḳdem-i Sulṭân-ı cihân

The object of jealousy of the Garden of Iram and of the heavenly rose-gardens15

Reşk-i bâġ-ı İrem ü ġayret-i gülzâr-ı cinân

14

15

This poem is among the 204 poems which Hammer-Purgstall (1825:6) has translated into German, and it reads: “Deine Schönheit strahlt als Sonne/ Und erfüllt die Welt mit Wonne | Weltflur machest Du zu Eden, Nachtigall sagt tausend Reden”. Von Hammer-Purgstall translates this as: “Es kam und adelte die Stadt heut der Sultan/ Verwandelnd sie in Edens Flur und Gülistan” (ibid.).

541

a paradise in the desert

In g67/1–2, even the supposed paradisiacal garden of Iram is not good enough for a lover to bear separation from their beloved:

g 67/1–2 My heart is like a bird, looking for a Ṭūbā [a tree in Paradise] to alight,

Mürġ-i dil konmağa bir ḳâmet-i Ṭûbâ gözedür

It is not impressed by a cypress [a beauty] or boxwood [gracefully grown young person], it is looking for something better

Serv ü şimşâdı begenmez ḳatı aʿlâ gözedür

Yet, even if its home were the Garden of Iram, it would not rest,

Menzilin bâġ-ı İrem kılsalar itmez ârâm

For the lover will look for a shelter like your heavenly district

Ehl-i dil cennet-i kûyûñ gibi meʾvâ gözedür

Interestingly, in g426/4, Bâḳî refers to the Garden of Eden as the earthly paradise bâg-ı İrem:

g426/4 While there was eternal life in the Garden of Iram, he came out of it

Mülk-i bâḳî variken bâġ-ı İremden çıkdı

Yet, mankind’s pride and joy is this transitory world (lit. perishable rose-garden)

Yine maġrûr olur âdem bu fenâ gülzâra

He also achieves a pleasant simile in comparing the beloved’s dangling locks to a peacock strutting about in the Garden of Iram.

542

elmaz

g498/2 (cf. g414/2) Your lock dangles around your cheeks,

Ġam-ı [read: ḫam-ı] zülfüñ ki devr eyler yanağuñ

Like a peacock of Iram strutting about

Îrem [sic!] ṭâvûsıdur cevlâne geldi

Iram also features in other Ottoman poems. For instance, in his description of Istanbul, Şehr-engiz, the minor Ottoman poet with the pen-name Faḳîrî (d. after 1535) from present-day Tetovo (trk. Kalkandelen) in Macedonia praises Istanbul with its flourishing buildings to be like the Garden of Iram, while attributing its inhabitants’ happiness to its pleasant breeze (see Çalis-Kural 2014:98, 127): İrem bâgı gibi her beyti maʿmûr/ Nesîm-i ḫulḳı eyler ḫalkı mesrûr (see Levend 1957:98). Having mentioned a poet originating from today’s Macedonia, reference should also be made to a mystical ghazal by Nezim Frakulla (d. 1760), otherwise known as Nezim Berati. He was one of the popular poets who wrote in Albanian using the Arabic script (Bejtexhinj), and in a typical dialogue between rose and nightingale, the rose gets annoyed and trying to compliment her, the nightingale addresses the rose as a bud of the rose-garden of Iram (see Hamiti 2008:182):

Thashë: epu si të të them, Ej gonxhë bagi irem, Tha: as thuaj bëmë qerem, Ej bylbyl, Nezimi talib

So I said: “Then how should I ask of you, Oh bud of the rose garden of Iram?” [The rose] said: “Are you saying ‘have mercy’, Oh nightingale, Nezim, the seeker.”

There seems to be no better verse to exemplify the motif of Iram in poetry than the following verse from Eremya Chelebi’s (d. 1695) Armeno-Turkish narrative poem known as the “Jewish Bride”. Essentially, the different versions circulating in Armenian, Turkish and Greek tell about the Jewish girl Mergata’s elopement from Istanbul with the young Christian-Orthodox Albanian baker Dimo and her conversion to Christianity. In one of the several hundred verses, she is depicted beautifully as dandering in a garden as follows: Kızın salınır Eflahde/ Serve manend ol dem khramende/ Baghi Irem ya gülshende

a paradise in the desert

543

In Walachia your daughter is swaying like a cypress who walks about pompously in the Garden of Iram (or in a rose garden) bodrogligeti 1983:551

This demonstrates clearly how a term from Arabian lore through its use in the Quran became the archetype of a garden replicating Paradise, and by transcending linguistic borders, it soon featured Persian, Ottoman, and Urdu poetry and developed into new images. It is thus striking but not surprising that in 17th century Ottoman Istanbul it even transcended religious boundaries and made its way into this poem of “a literary minority on social minorities”, as Ambros has succinctly described it (2016:156). The aim of this paper was to analyse the role and function that Iram gained in literature. The Quran mentions Iram dhāt al-ʿimād juxtaposed to ʿĀd (q 89:6– 7), locating Iram in Ancient Arabian tradition. In early Arabic poetry, Iram seems to have been understood as a name of a person, a tribe, or the dwellingplace of the ʿĀd. In modern Arabic literature, Iram is used to refer to a mystical place, maybe far away and unattainable, a city of dreams, and not to a person or a tribe. Numerous commentaries on the Quran from al-Ṭabarānī (d. 971) on, include the marvellous tale of ʿAbdallāh b. Qilāba who was the only person to find Iram, Shaddād b. ʿĀd’s extravagant city as a replication of Paradise on Earth. Similar to what is the case with Atlantis, the notion of Iram as a lost city has motivated people to try and actually find it, which has lead to several geographical identifications. However, denoting ‘Paradise on Earth’, in significant works of Persian, Ottoman, and Urdu poetry among other Islamicate literatures like in Nezim Berati’s Albanian ghazal, there are references to an Islamic appropriation of this paradise as the paradisiacal “Garden of Iram”. The lost and hidden but luxurious and mystical city or garden of Iram also features in tales of the Nights. In some further embellished tales of the Nights, jinn inhabit it, while beautiful fairies inhabit the garden of Iram in Turkic folklore. In three of Bâḳî’s Ottoman poems one finds a peacock strutting about in the garden of Iram, and the next stepup from the inhabited garden must have been the sea since there used to be a lake called Baḥr-e Eram (“sea of Iram”) in northern Iran until two hundred years ago. Thus, it does not come as a surprise that gardens were oftentimes compared to the garden of Iram as exemplified in the two inscriptions in Chanderi, drawing heavily on eschatological imagery, because gardens were the center of leisure and entertainment in the Islamic world. Yet, as if the idea of Shaddād’s Iram as Paradise on Earth was not enough, some rulers actually named gardens

544

elmaz

that they had planned and built “Iram”. Like modern Shiraz, which is home to a Bāġ-e Eram, Istanbul had an İrem bâğı at the Saʿd-âbâd complex (see ÇalisKural 2014:217–221). Evliya Çelebi (d. 1682) uses in his travelogue the adjective şeddâdî, derived from Shaddād [b. ʿĀd] to describe huge buildings (Karateke 2013:22, 32, 37, 42, 46, 48, 53, 56, 76), and according to him, there was also a Garden of Iram in the vicinity of Buda, the “Pecy-i dilcuy Seremi bagh-i irem Garden” (see Conan 2007:199), or at least something resembling it. This paper did not discuss the origins of the history of ʿĀd and related Iram, or where Iram is or was located. Rather, it demonstrated how lively and dynamic Iram was in the creative mind of people, who transformed the sybaritic city into a paradisiacal garden, and even inhabited it with jinns, fairies, and birds—an admittedly much more appealing image than that of an abandoned city. A proper name mentioned only once in the Quran, did not only add flavour to many a poem and monumental inscription. Owing much to its inclusion in the corpus of One Thousand and One Nights it is still inspiring literary fiction and more recently video-games such as Uncharted 3: Drake’s Deception (2011) to this day, in the East as well as in the West.

References Abdel Haleem, Muhammad. 2010. The Quran. Oxford: Oxford University Press. Abdur Rahim, Syed & Ziyauddin A. Desai. 1966. Inscriptions of the Sultans of Malwa. Epigraphia Indica—Arabic and Persian Supplement 1964–65 (1966): 45–78. Abdur Rahim, Syed. 2000. Arabic, Persian, and Urdu Inscriptions of Central India. New Delhi: Sundeep Prakashan. Abu Haidar, Jareer. 1996. “Romanticism and Modern Arabic Literature,” In Tradition and Modernity in Arabic Language and Literature, edited by James M. Smart, 3–17. London: Curzon Press. Al-Duḥaymī, Muḥammad. Wǝsh b-ṣadri?! Online at http://www.adab.com/folk/ modules.php?name=Sh3er&doWhat=shqas&qid=87840&r=&rc=6. [Last accessed 1 May 2016]. Al-Hilālī, Muhammad T. & Khān, Muhammad M. 1998 (1419a.h.) The Noble Qurʾan. English translation of the meanings and commentary. Medina: King Fahd Complex for the Printing of the Holy Qurʾan. Al-Ibshīhī, Shihāb al-Dīn. 1992. Al-Mustaṭraf fī kull fann mustaẓraf. Beirut: Dār maktabat al-ḥayāt li-l-ṭibāʿa wa-l-nashr. Al-Rubayʿī, Fāḍil. 2000. Iram dhāt al-ʿimād. Min Makka ilā Ūrushalīm: al-baḥth ʿan aljanna. Beirut: Riad El-Rayyes Books. Al-Shirwānī, Aḥmad b. Muḥammad al-Yamanī. 1841. Nafḥat al-Yaman fī-mā yazūl bidhikrih al-shajan. Hooghly: Medical Press.

a paradise in the desert

545

Al-Shirwānī, Aḥmad b. Muḥammad al-Yamanī. 1814–1818. The Arabian Nights Entertainments in the original Arabic. 2 vols. Calcutta: Hindoostanee Press. Al-Ṭabarī, Muḥammad b. Jarīr b. Yazīd. 1968–1971. Tārīkh al-rusul wa-l-mulūk, edited by Muḥammad Abū al-Faḍl Ibrāhīm. Cairo: Dār al-Maʿārif bi-Miṣr. Al-Ṭabarī, Muḥammad b. Jarīr b. Yazīd. 1994. Tafsīr al-Ṭabarī, ed. Bashshār A. Maʿrūf & ʿIṣām F. al-Ḥarastānī. Beirut: Muʾassasat al-Risāla. Al-Wardī, Sirāj al-Dīn b. 2008. Kharīdat al-ʿajāʾib wafīrat al-gharāʾib. Cairo: Maktabat al-thaqāfa al-dīniyya. Alf layla wa-layla, ed. ʿAfīf Ḥaṭūm. 2008. 2 vols. Beirut: Dār Ṣādir. Alpaslan, Ali. 1987. Ahmet Paşa. Ankara: Kültür ve Turizm Bakanlığı. Ambros, Edith Gülçin. Frivolity and Flirtation, in Ottoman Women in Public Space, edited by Ebru Boyar and Kate Fleet, pp. 150–186. Leiden: Brill. Bakıxanov, Abbasqulu ağa. 2015. Gülüstani-İrǝm. Baku: Kitab Klubu. Bencheikh, Jamel Eddine. 1990. «Iram ou la clameur de Dieu. Le mythe et le verset». Revue du monde musulman et de la Méditerranée 58.1: 70–81. Blatherwick, Helen. 2016. Prophets, gods and kings in Sīrat Sayf ibn Dhī Yazan: an intertextual reading of an Egyptian popular epic. Leiden: Brill. Bodrogligeti, Andras J.E. 1983. Review of Eremya Chelebi Kömürjian’s Armeno-Turkish Poem “The Jewish Bride” by A.K. Sanjian, A. Tietze. International Journal of Middle East Studies 15.4: 545–556. Bravmann, Meïr Max. 1972. The Spiritual Background of Early Islam: Studies in Ancient Arab Concepts. Breslau 1825–1843. Hâdhâ Kitâb Alf layla wa-layla min al-mubtadaʾ ilâ al-muntahâ/Tausend und Eine Nacht. Arabisch. Nach einer Handschrift aus Tunis. Herausgegeben von Dr. Maximilian Habicht (and H.L. Fleischer). 12 vols. Breslau. Būlāq 1835. Alf layla wa-layla. Al-Ṭabʿa al-ūlā. 2 vols. Cairo. Bürgel, Johann C. 2005. “The Mighty Beloved: Images and Structures of Power in the Ghazal from Arabicto Urdu”, in Ghazal as world literature 1. Transformations of a literary genre, ed. by Thomas Bauer and Angelika Neuwirth, pp. 283–309. Beirut: Ergon Verlag. Burton, Richard F. 1885–1888. A Plain and Literal Translation of the Arabian Nights’ Entertainments, Now Entituled the Book of the Thousand Nights and a Night. Printed by the Burton Club for private subscribers only. Online at: http://burtoniana.org/ books/1885-Arabian%20Nights/index.htm. [Last access 1 May 2016] Burton-Page, John. 2008. Indian Islamic architecture: forms and typologies, sites and monuments, edited y George Michell. Leiden: Brill. Bushrui, Suheil & Jenkins, Joe. 1998. Kahlil Gibran. Man and Poet. A new biography. Oxford: Oneworld Publications. Bushrui, Suheil. 2010. “Jubrān Khalīl Jubrān”, In Essays in Arabic Literary Biography 1850–1950, edited by Roger Allen, pp. 175–191. Wiesbaden: Harrassowitz.

546

elmaz

Çalış-Kural, B. Deniz. 2014. Şehrengiz, urban rituals and deviant Sufi mysticism in Ottoman Istanbul. Farnham: Ashgate. Chauvin, Victor. 1899. La récension égyptienne des Mille et une nuits. Brussels: Office de publicité. Clapp, Nicholas. 1998. The Road to Ubar: Finding the Atlantis of the Sands. New York, ny: Houghton Mifflin. Conan, Michel. 2007. Middle East garden traditions: unity and diversity. Questions, methods and resources in a multicultural perspective. Washington: D.C., Dumbarton Oaks Research Library and Collection. Corpus Coranicum. Variae Lectiones Coranicae, hg. von der Berlin-Brandenburgischen Akademie der Wissenschaften durch Michael Marx unter Mitarbeit von Suad Hodzic, Edin Mahmutovic und Jens Sauer mit Beiträgen von Feras Krimsti und Ismail Mohr. Betaversion: Stand 1.5.2016. Online at http://corpuscoranicum.de/lesarten/ index/sure/89/vers/7. [Last accessed 1 May 2016] Davis, Dick. 2013. Faces of love Hafez and the poets of Shiraz. New York, ny: Penguin Books. Dede Korkut. 2013. The Book of Dede Korkut: A Turkish Epic, ed. by Faruk Sümer, Ahmet E. Uysal, Warren S. Walker. Austin, tx: University of Texas Press. Divân-ı Baki Efendi, [1176, i.e. 1762 or 1763]. Online at: https://catalog.hathitrust.org/ Record/006818621. [Last access 1 May 2016] Duraković, Esad. 1997. Poetika arapske književnosti u sad: prožimanje književnih tradicija. Sarajevo: Zid. Elmaz, Orhan. 2011. Studien zu den koranischen Hapaxlegomena unikaler Wurzeln. Wiesbaden: Harrassowitz. El-Shamy, Hasan M. 2004. Types of the folktale in the Arab world: a demographically oriented tale-type index. Bloomington, in: Indiana University Press. Eskigün, Kübra. 2006. Klasik Türk şiirinde efsanevi kuşlar. Yüksek lisans tezi, Kahramanmaraş Sütçü İmam Üniversitesi. Falasch, Ute. 2015. Heiligkeit und Mobilität: die Madāriyya Sufibruderschaft und ihr Gründer Badīʿ al-Dīn Shāh Madār in Indien. 15–19. Jahrhundert. Berlin: Lit Verlag. Fiennes, Ranulph. 1992. Atlantis of the Sands: The Search for the Lost City of Ubar. London: Signet. Frank, Ulrich. 1987. „Freimaurer“. Enzyklopädie des Märchens 5:240–246. Galland, Antoine. 1704–1717. Les mille et une nuits. Contes arabes. Lyon: Antoine Briasson. Glidden, Harold. Koranic Iram. Legendary and Historical. Bulletin of the American Schools of Oriental Research 73 (1939): 13–15. Goodman, Lenn Evan. 1999. Jewish and Islamic Philosophy: Crosspollinations in the Classic Age. New Brunswick, nj: Rutgers University Press. Habicht, Max. 1825. Tausend und Eine Nacht. Arabische Erzählungen. Zum erstenmal aus

a paradise in the desert

547

einer Tunesischen Handschrift ergänzt und vollständig übersezt von Max. Habicht, F.H. von der Hagen und Karl Schall. 15 vols. Breslau: Josef Max und Komp. Hamdouni Alami, Mohammed. 2011. Art and Architecture in the Islamic Tradition. London: I.B. Tauris. Hamiti, Abdullah. 2008. Nezim Frakulla dhe divani i tij Shqip. Skopje: Logos-A. Hammer, Joseph von. 1825. Baki’s, des größten türkischen Lyrikers Diwan. Vienna: C.F. Beck’sche Buchhandlung. Healey, John F. 2001. The religion of the Nabataeans: a Conspectus. Leiden: Brill. Heller, Sam. Al-Qaeda’s Ayman Al-Zawahri Plays Politics. Gandhara—Radio Free Europe/ Radio Liberty. May 20, 2016. Online at http://gandhara.rferl.org/a/zawahiri -politics/27747624.html. Herbelot de Molainville, Barthélemy d’. 1697. Bibliothèque orientale, ou Dictionnaire universel contenant généralement tout ce qui regarde la connaissance des peuples de l’Orient, ed. by Antoine Galland. Online at http://catalogue.bnf.fr/ark:/12148/ cb305951566. [Last accessed 1 May 2016]. Horovitz, Josef. 1926. Koranische Untersuchungen. Berlin: Walter de Gruyter. icomos. 2011. Evaluations of nominations of cultural and mixed properties to the World Heritage List. icomos report for the World Heritage Committee. 35th ordinary session. unesco. Jayyusi, Salma K. & Tinglev, Christopher. 1977. Trends and Movements in Modern Arabic Poetry. Leiden: Brill. Gottheil, Richard & Jacobs, Joseph. 1902. “Arabian Nights”. The Jewish Encyclopedia, vol. 2:44f. New York: Funk and Wagnalls. Jubran, Sulaiman. 2007. Classical Elements in Mahjar Poetry. Journal of Arabic Literature 38.1 (2007):67–77. Karateke, Hakan. 2013. Evliyā Çelebī’s Journey from Bursa to the Dardanelles and Edirne: From the Fifth Book of the Seyāḥatnāme. Leiden: Brill. Karryev, Baymukhamed A. 1968. Ėpicheskie skazanija ο Ker-ogly u tjurkojazychnykh narodov. Moscow: Nauka. Khan, Jalal Uddin. 2015. Readings in Oriental Literature: Arabian, Indian, and Islamic. Newcastle upon Tyne: Cambridge Scholars Publishing. Korogly, Chalyk G. 1995. „Köroġlu“. Enzyklopädie des Märchens 8:304–307. Küçük, Sabahattin. 1994. Bâkî dîvânı. Ankara: Türk Dil Kurumu. Lane, Edward W. 1839–1841. The Thousand and One Nights. Commonly called, in England The Arabian Nights’ Entertainments. A new translation from the Arabic, with copious notes. 3 vols. London: Charles Knight and Co. Lebedev, Viktor. Sledy juzhnoarabskoj fol’klornoj tradicii v skazkakh 1001 nochi. Narody Azii i Afriki 1973.1: 102–113. Levend, Agâh Sırrı. 1957. Türk edebiyatında şehr-engizler ve şehr-engizlerde İstanbul. Istanbul: Baha Matbaası.

548

elmaz

Littmann, Enno. 1916. Märchen und Legenden aus der syrisch-arabischen Wüste. Nachrichten von der Königlichen Gesellschaft der Wissenschaften zu Göttingen. Philologisch-historische Klasse aus dem Jahre 1915:1–27. Berlin: Weidmann’sche Buchhandlung. Lyall, Charles J. 1921–1924. The Mufaḍḍaliyāt: an anthology of ancient Arabian odes according to the recension. 3 vols. Oxford: Oxford University Press. Macdonald, Michael C.A. “Reflections on the linguistic map of pre-Islamic Arabia”. Arabian Archaeology and Epigraphy 11 (2000): 28–79. Mahdi, Muhsin. 1995. The thousand and one nights. Leiden: Brill. Mardrus, Joseph C. 1899–1904. Le Livre des Mille Nuits et Une Nuit. Traduction littérale et complète du texte Arabe. 16 vols. Paris: Éditions de la Revue Blanche. Matthews, David J. 1993. Iqbal, a Selection of the Urdu Verse. New Delhi: Heritage. Mcnaghten, William H. 1839–1842. The Alif Laila Or Book of the Thousand Nights and One Night. Commonly known as “The Arabian Nights’ Entertainments.” 4 vols. Calcutta. Meisami-Scott, Julie. 1995. Haft Paykar: A Medieval Persian Romance. Oxford: Oxford University Press. Mihyar of Damascus: His Songs. Poems by Adonis, tr. by Adnan Haydar & Michael Beard. 2008. Rochester, ny: boa Editions. Neuenkirchen, Paul. 2013. Biblical Elements in Koran 89, 6–8 and its Exegeses: A New Interpretation of “Iram of the Pillars”. Arabica 60.6: 651–700. Osigus, Anke & Heinz Grotzfeld. 2005. „Europäische Übersetzungen“, In Dreihundert Jahre 1001 Nacht in Europa. Ein Begleitheft zur Ausstellung in Münster, Tübingen und Gotha, edited by Anke Osigus in Zusammenarbeit mit Heinz Grotzfeld, pp. 49–64. Münster: lit. Pinault, David. 1992. Story-Telling Techniques in the Arabian Nights. Leiden: Brill. Perles, Joseph. 1873. “Rabbinische Agada’s in 1001 Nacht: Ein Beitrag zur Geschichte der Wanderung orientalischer Märchen.” Monatsschrift für Geschichte und Wissenschaft des Judenthums 22:14–34, 61–85, 116–125. Pinckney Stetkevych, Suzanne. 2010. The Mantle Odes: Arabic Praise Poems to the Prophet Muhammad. Bloomington, id: Indiana University Press. Piotrovskij, Mikhail B. 1977. Predanie o khimjaritskom care As’ade al-Kamile. Moscow: Nauka. Qıraubaeva, Alma. 2010. Şığıstıq qissa-dastandar. Almaty: Öner. Radscheit, Matthias. 2016. “Springs and Fountains.” Encyclopaedia of the Qurʾān. General Editor: Jane Dammen McAuliffe, Georgetown University, Washington dc. Brill Online, 2016. Reference. University of St Andrews. 1 May 2016 http://referenceworks .brillonline.com/entries/encyclopaedia-of-the-quran/springs-and-fountains -EQCOM_00194. Reichl, Karl. 1992. Turkic Oral Epic Poetry. Traditions, Forms, Poetic Structure. London: Garland Publishing.

a paradise in the desert

549

Renard, John. 1999. Islam and the heroic image: themes in literature and the visual arts. Macon: Mercer University Press. Retsö, Jan. 2003. The Arabs in Antiquity: Their History from the Assyrians to the Umayyads. Oxon: Routledge Curzon. Rosenthal, Franz. 2005. The Muqaddimah. An Introduction to History. The Classic Islamic History of the World. Ibn Khaldûn. Princeton, nj: Princeton University Press. Sadan, Joseph. 2004. “The Arabian Nights and the Jews”. The Arabian Nights Encyclopedia, edited by Ulrich Marzolph & Richard van Leeuwen with the collaboration of Hassan Wassouf. Oxford: abc-Clio. Scharfe, Martin. 2011. „Verirren“. Enzyklopädie des Märchens 14:36–42. Sezgin, Fuat. 1996. Geschichte des arabischen Schrifttums, Band ii: Poesie. Bis ca. 430h. Leiden: Brill. Sharafuddin, Mohammed. 1996. Islam and Romantic Orientalism: Literary Encounters with the Orient. London: I.B. Tauris. Sharḥ dīwān Labīd b. Rabīʿa al-ʿĀmirī, ed. Iḥsān ʿAbbās. 1962. Kuwait: Wizārat al-iʿlām bi-dawlat al-Kuwait. Sorrell, Clive F. 2012. Kawthar: Death Shadows Those Who Seek the Sweet Waters of Paradise. N.p.: authorhouse. Sperl, Stefan. 1996. “Al-Būṣīrī (d. c. 1296). The Burda in praise of the Prophet Muḥammad”, In Qasida poetry in Islamic Asia and Africa. Vol. ii. Eulogy’s bounty, meaning’s abundance. An Anthology, edited by Stefan Sperl & Christopher Shackle, 388–412. Leiden: Brill. Stone iii, Eric L. 2012. The Lost City of Iram: A geographic adventure through time (2012). N.p.: CreateSpace Independent Publishing Platform. The Arabian Nights Encyclopedia, ed. Ulrich Marzolph and Richard van Leeuwen with the collaboration of Hassan Wassouf. 2004. Oxford: abc-Clio. The History of al-Ṭabarī. Volume v. The Sāsānids, the Byzantines, the Lakmids, and Yemen, tr. C.E. Bosworth. 1999. Albany, ny: State University of New York Press. The Parallel Quran. Copyright Clay Chip Smith. Online at http://www.clay.smith.name/ Parallel_Quran.2004.03.21.htm. [Last access 1 May 2016] Tonna, Jo. 1990. The Poetics of Arab-Islamic Architecture. Muqarnas 7 (1990):182–197. ʿUthmān, ʿAbd al-Jawād Sālim. Abū Bakr al-Naqqāsh al-Mawṣilī. Shuyūkhuh, talāmīdhuh, muʾallafātuh. Dirāsāt Mawṣiliyya 23 (1430ah/ 2009): 173–193. Weil, Gustav. 1837–1841. Tausend und eine Nacht. Arabische Erzählungen. Zum Erstenmale aus dem arabischen Urtext übersetzt von Dr. Gustav Weil. 4 vols. Wheeler, Brannon. 2002. Prophets in the Quran: An Introduction to the Quran and Muslim Exegesis. London: Continuum. Wilber, Donald. 1962. Persian Gardens & Garden Pavilions. Rutland, vt: C.E. Tuttle Co. Wilberforce Clarke, Henry. 1881. The Sikandar nāma, e bara; or, Book of Alexander the Great. London: W.H. Allen.

550

elmaz

Wortley-Montague (1764). 1995. Neue Erzählungen aus den tausendundein Nächten. Die in anderen Versionen von ‘1001 Nacht’ nicht enthaltenen Geschichten der WortleyMontague-Handschrift der Oxforder Bodleian Library. Aus dem arabischen Urtext übertragen und erläutert von Felix Tauer. Frankfurt am Main: Insel Verlag.

chapter 27

Mourning for the Dead and the Beginning of Idolatry in the Kitāb al-Aṣnām and the Spelunca Thesaurorum—an Unknown Parallel to Sūrat at-Takāṯur (q102)? Konstantin M. Klein

In September 2014 the supervisors of Medina’s Prophet’s Mosque reviewed a proposal to destroy Muḥammad’s grave and to remove his remains to the nearby Baqīʿ al-Ġarqad cemetery, where they would be interred anonymously. Based on Saudi Arabia’s strict adherence to the teachings of Muḥammad ibn ʿAbd al-Wahhāb,1 which consider visits to saints’ tombs (ziyārat al-qubūr) an expression of idolatry (širk), these plans ran the risk of inflaming tensions between Sunni and Shiite pilgrims to the holy places. Two months earlier, in July 2014, isis militants destroyed the mosque of the prophet Yūnus in Mosul for similar reasons. In the argumentation against the veneration of tombs, one Qurʾānic prohibition against this practice is particularly important, q102:1–2: “Rivalry in worldly gain has distracted you, / Until you have come to the graves” (alhākumu t-takāṯur / ḥattā zurtumu l-maqābir).2 The interpretation of these two verses has puzzled scholars and stirred debate:3 Following Rudi Paret, Angelika Neuwirth suggested that “takāṯur” meant a special form of increase, namely increase originating from the competition for (a family’s) fame (as in takāṯur fī-l-awlād, the striving for (more/many) sons).4 In general, it seems clear that takāṯur alludes to an “obsessive pursuit of material gains” similar to q106. In the second verse both zāra and maqābir are hapax legomena in the Qurʾān. However, it is very sensible to assume that the plural maqābir denotes not just ordinary tombs, but such which were somehow more visible or prominent.5 1 On Muḥammad ibn ʿAbd al-Wahhāb’s concept of idolatry, cf. Hawting 1999: 63–65 and 81–82. 2 The translation is taken from Jones 2007: 586. 3 For an overview, cf. the discussion in Neuwirth 2011: 125–128. Künstlinger 1931: 619 suggested a connection between q102:1 and Luke 12:15. Cf. also the commentary on q102 by Nicolai Sinai, accessible online at http://corpuscoranicum.de/kommentar/index/sure/102/vers/1. 4 Cf. Paret 1977: 519–520 and Neuwirth 2011: 127. 5 Cf. Crewswell 1952: i,110–111 and 138, who argued that the Arabs in the time of jâhiliyya had no interest in commemorating their dead with buildings. Cf. also Leisten 1990: 12.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

552

klein

The general meaning of the second verse does not seem to be “until you reach the graves,” i.e. “until you die,” as Nicolai Sinai has argued, but rather describe the custom of visiting graves or graveyards with a specific intention. In any case, q102:1–2 and similar early prohibitions have been considered as a reaction to a flourishing and widely practiced cult (of ancestors?) in Arabia, which Muḥammad objected to.6 However, while recent scholarship, first and foremost Berlin’s ongoing Corpus Coranicum project, aims at interpreting the Qurʾān in the context of its historical development and in the light of non-Qurʾānic texts, so far no sufficiently fitting parallel to q102:1–2 has been found. This contribution argues that the two verses should indeed be understood in the context of an Arabian ancestor cult and suggests a parallel text deriving from the Syriac New Testament apocryphal traditions on the burial of Adam in a legendarily famous cave, containing the “treasures” frankincense, myrrh, and gold.

∵ It is tempting to connect the Qurʾānic prohibitions against the visiting of graves to the literary evidence of the Safaitic inscriptions and in particular to the countless mentions of the verb wgm, which is both among the most common and hotly debated words within this corpus. It is usually translated as “grieving”7 or, as Kerstin Eksell has more recently argued, “to perform a (ritual) act of mourning.”8 One could also connect the prohibitions to the

6 On the early Islamic attitudes towards funerary architecture in ḥadīṯ literature, cf. Leisten 1990: 12–14 and esp. notes 9–12 on prohibitions against praying in cemeteries and by graves, interdictions against visiting them or sitting next to them, and, in particular against the practice of al-bināʾ ʿalā l-qubūr, “building on graves.” 7 Albert Jamme has argued that the verb only means “to grieve” or “to mourn” and not, as often found, “to place a stone on a cairn” or as cis translated “signum posuit”, cf. Jamme 1967: 168– 172. The assumption that the action described by the Safaitic verb wgm denotes the ritual act of laying down a stone derives not only from the archaeological and epigraphical evidence of cairns with inscriptions written, of course, on stones, but also from the Arab noun wajam, pl. awjām (“a heap of stones in desert, a sign”), however, it is impossible to ascertain when the Arabic noun took on the this specific meaning, which may or may not have been a part of the meaning of the Safaitic verb wgm. 8 Cf. Eksell 2005: 167–172. The more specific meaning suggested by Kerstin Eksell shares clear similarities with the meaning of the Arabic verb wagama (“to remain silent with downcast eyes in grief or anger”)—which denotes a ritualised act following upon an emotion (mostly: grief), cf. also Jamme 1967: 168–169. Al-Jallad 2015: 351 chose the more general sense of the

mourning for the dead and the beginning of idolatry

553

archaeological evidence of stone structures such as so-called kites, wheels, pendants, walls or cairns in the Jordanian Basalt desert, among them the Cairn of Hāniʾ, which was the first cairn to be excavated and is, not least because of the extensive studies it attracted, the most famous epigraphical locus in the region.9 The initial construction of the cairn was a rectangular structure built from field stones to which a second circuit had been added towards the north, east and south. Finally the space between the walls and the courtyard was filled up with stones, probably over a longer period of time.10 172 Safaitic inscriptions (as well as a Latin one, which might point to an early dating of the cairn) have been discovered, most of which published by G.L. Harding in his study of this cairn in 1953. Hāniʾ is mentioned in 97 of the inscriptions, of which 18 were written by his close relatives (brothers, nephews, and cousins— it seems that Hāniʾ died without any direct descendants of his own) and about 79 by his friends, members of 21 different lineages.11 Most inscriptions state that the person either helped to build the cairn or “grieved” (wgm) for the deceased, who, in all likelihood died of natural causes.12 Since no inscriptions from any children of Hāniʾ were discovered and because of the funerary goods (a staff, a jar tentatively identified as a begging bowl, a spoon, and a water-skin) and not least because of the number of people who came to pay their respects, Harding assumed that Hāniʾ was some kind of holy man or dervish.13 However, while Hāniʾ’s grave certainly was the object of some sort of veneration and was visited over a certain period of time by family members and friends, the cairn is both geographically and (more importantly) chronologically too far removed from the Qurʾānic prohibition of q102 to combine this evidence. Compared to the early evidence for ancestor(?) worship in Safaitic inscriptions and cairns, it is more sensible to search for examples in early Islamic traditions on idolatry, first and foremost in Ibn al-Kalbī’s Kitāb al-Aṣnām, on

9

10 11 12

13

word in his lexicon, i.e. “grieving,” quoting an example which cannot be connected to the death of a human being (which would be specified by adding the passive participle trḥ “dead” or qtl “killed” after the name). The original report of the excavation and epigraphical survey was published by Harding 1953: 8–56. It should be noted that while some cairns, such as Hāniʾ’s, contain burials, others clearly do not, cf. Kennedy 2012: 483. Cf. Kennedy 2012: 484–485. All numbers are taken from Kennedy 2012: 484. Cf. Shanklin/Dark 1953: 59 and Kennedy, “Cairn of Hānī,” 486–487, against Harding 1953: 8. G.L. Harding pointed to the custom, still practised among the Beduins of the 1950s, to build cairns only over those who were killed, not those who died of a natural cause. Cf. Harding 1953: 9.

554

klein

which almost all later traditions concerning the veneration of idols are based.14 The author, Hišām ibn Muḥammad ibn as-Sāʾib al-Kalbī (b. c. 737ad), was a polymath from Kūfa where he also received his education. After spending some time in Baġdād during the reign of the ʿAbbāsid caliph al-Mahdī, he returned to his hometown where he died in 819 or 821ad. Like his father before him, he wrote on a large variety of topics but was particularly renowned as a specialist on genealogies. An-Nadīm’s Fihrist knows 141 works attributed to him, and Yāqūt records even more, some 150 titles. It is likely that most of these works were rather short treatises, of this large number only few are extant today: his collection of genealogical tables, Ǧamharat an-nasab, survives only partly, whereas one short treatise on the most famous horses of the pre-Islamic Arabs, Kitāb ansāb al-ḫayl, and the Kitāb al-Aṣnām survive completely. At a time when knowledge of the pre-Islamic past became fashionable again,15 Ibn al-Kalbī collected all information on idols, cult places, and cultic rituals available to him in this short treatise. However, following the fast expansion of Islam in the Arabian Peninsula and its struggle against polytheistic traditions, knowledge on pre-Islamic religion had suffered a veritable damnatio memoriae, so that the information gathered by the author is rather sparse and often difficult to date.16 Gerald Hawting has called this perception and the overall historicity of Ibn al-Kalbī’s and other traditionalists’ information into question. He has argued that, while the Qurʾān merely provided obscure hints and allusions to a pagan environment, texts such as the Kitāb al-Aṣnām, originated mainly in order to provide more detail for the Qurʾānic descriptions and to connect a variety of cults and deities to Mecca and the Ḥiǧāz so that the Qurʾānic accusations against idolatry became more credible.17 While Hawting’s argument is compelling, it has also attracted a considerable amount of criticism. The historicity of the cult of certain deities is not at the centre of this contribution,

14

15 16

17

Ibn al-Kalbī himself should probably not be considered as the author of the text as we have it today which is clearly an amalgamation of different texts. It is more sensible to attribute the authorship to the group of his disciples or even to later generations of their disciples. For simplicity’s sake, however, in this contribution, the text will be referenced by the name to which it is traditionally attributed, cf. Hawting 1999: 88–89 and 91–92. Cf. Nyberg 1939: 348–349. Cf. Hawting 1999: 5, calling the early Muslim literature which provides us with details about pre-Islamic idolatry “largely stereotypical and formulaic” and its value as evidence “questionable.” Cf. Hawting 1999, passim, esp. 62–63, 68, 95 and 110: “They [works such as the Kitāb alAṣnām] should not be understood, as it seems they often are, as a collection of authentically Arabian ideas and traditions.”

mourning for the dead and the beginning of idolatry

555

which focuses on a neglected source of Ibn al-Kalbī’s reports on how idolatry entered this world and which might also offer an explanation for the two first verses of q102. In the Kitāb al-Aṣnām Ibn al-Kalbī transmits two different stories concerning the origin of idolatry. One of them is placed at the very beginning of the work, and has attracted considerable scholarly interest:18 The reason which led them [the pre-Islamic Arabs of Mecca] to the worship of images and stones was the following: No one from the Banū Ismāʿīl left Mecca without carrying away with him a stone from the stones of the ḥaram as a token of reverence to it, and as a sign of deep affection to Mecca. Wherever he settled he would erect that stone and circumambulate it in the same manner he used to circumambulate the Kaʿba [before his departure from Mecca], seeking thereby its blessing and affirming his deep affection for the Sacred House [i.e. the Kaʿba]. In fact, the Arabs still venerate the Kaʿba and Mecca and journey to them in order to perform the pilgrimage and visitation, conforming thereby to the time honoured custom which they inherited from Ibrāhīm and Ismāʿīl. In time this led them to the worship of whatever took their fancy, and caused them to forget their former worship. They exchanged the religion of Ibrāhīm and Ismāʿīl for another. Consequently they took to the worship of images, becoming like the nations before them. They sought and determined what the people of Nūḥ had worshiped of these images and adopted the worship of those which were still remembered among them. The story explains the spread of idolatry with a model of the dissemination of beliefs in the ancestral era of the ancient Arabs and with a corruption of the true faith due to human forgetfulness.19 Here (as well as in the second report), the invention of the wrong cult practice is attributed to a certain ʿAmr ibn Luḥayy, a man who had seized chieftainship (siyāda) over Mecca and had corrupted the true faith once introduced by Ibrāhīm and Ismāʿīl. According to the first report, ʿAmr travelled to al-Balqāʾ in Syria,20 where he learned about 18

19 20

Kitāb al-Aṣnām 4,1–9, (translation adapted from Faris 1952: 4). On the two reports, cf. Athamina 2004: 193–195, Hawting 1999: 91, and Nyberg 1939: 352–353 as well as 353–354 on the genesis of the text. On this report, cf., e.g. Hawting 1999: 24–25 and 93. On the localisation of al-Balqāʾ, cf. Stummer 1944: 382–383. The story of ʿAmr ibn Luḥayy’s visit to Syria, including a healing, shows a general similarity to the healing of Naaman in 2 Kings 5, as noted by Klinke-Rosenberger 1941: 79 n. 52.

556

klein

the veneration of idols and then re-imported this cult to Mecca. The main aim of this story is to demonstrate that despite of all these corruptions of the true religion of Ibrāhīm, Islam is still identical with this primordial faith (or, at least, its direct successor): In spite of the polytheistic idolatry which had spread among the Arabs of Mecca, there were still survivals of the original faith of Ibrāhīm and Ismāʿīl, e.g. the veneration and circumambulation of the Kaʿba, the rituals of ḥajj itself as well as the lesser pilgrimage, the ʿumra, the sacrificial offerings, or the standing upon ʿArafa and Muzdalifa.21 A characteristic feature of the Kitāb al-Aṣnām is that it aims at demonstrating that all idols listed within the text had a connection to Mecca—even if these idols were originally located far away from it. The figure of ʿAmr ibn Luḥayy is vital in these narrations, since it was he who was responsible for bringing exactly these deities, the gods of Noah mentioned in the Qurʾān, to Mecca, from their alleged far-away locations in Yemen. Yemen.22 These deities, Wadd, Suwāʿ, Yaġūṯ, Yaʿūq, and Nasr, also figure prominently in the second and longer account on the origin of idolatry, which shows more associations with a cult of ancestors:23 The beginning of the worship of idols was that, when Ādam died the sons of Šīṯ ibn Ādam buried him in a cave in the whereon Adam alighted (when he was sent) to the land of Hind. (The name of the mountain is Nawḏ, and it is the most fertile mountain in all the world. […]). The Banū Šīṯ were wont to visit the body of Ādam in the cave in order to pay their respect to him and offer their prayers for his soul. Thereupon one of the Banū Qābīl ibn Ādam said: “O Banū Qābīl! Verily the Banū Šīṯ have a dawār [circuit] which they circumambulate in veneration, but you have none.” Consequently he carved for them an idol, and was, therefore, the first to make such. […] Wadd, Suwāʿ, Yaġūṯ, Yaʿūq, and Nasr were righteous people who died within one month of one another, and their relatives were griefstricken over them. Then a man from the Banū Qābīl addressed their relatives saying, “O my fellow-citizens! Shall I produce for you five statues after the image? However, I cannot impart life to them.” They answered: “Yes.” Thereupon he carved unto them five statues after the image of [their departed relatives], and placed them [over their graves]. Then it came to pass that each would visit [the grave of] his brother, uncle, or cousin, pay

21 22 23

Cf. Hawting 1999: 36 with Kitāb al-Aṣnām 4,10–11 and Aḥbār Makka i,116–117. Cf. Hawting 1999: 128. Kitāb al-Aṣnām 31,8–34,2. Cf. on this report Hawting 1999: 94 as well as Nyberg 1939: 364– 365 against Wellhausen ³1961: 14–16.

mourning for the dead and the beginning of idolatry

557

his respect and circumambulate around them, until the first generation passed away. The statues were made during the age of Yarād ibn Mahlāʾīl ibn Qaynān ibn Anūsh ibn Šīṯ ibn Ādam. Then another generation [of people] followed, who worshipped these statues with more veneration than the first one. After them there was the third generation [of people], and they said: “Our forefathers venerated these statues for no other reason than the desire to enjoy their intercession before God.” Consequently they worshipped them, and became far gone in disbelief. Thereupon God sent unto them Idrīs as a prophet, who is Aḥnūḫ ibn Yarād ibn Mahlāʾīl. He urged them to [return to] his religion, however, they called him a liar. Therefore God lifted him up to a place on high. Their cause continued to wax strong until Nūḥ ibn Lāmak ibn Matūšalaḥ ibn Aḥnūḫ grew up. And God sent him as a prophet. He was, then, 480 years old. For a period of one hundred and twenty years he advised them to venerate God. However, they did not obey him and called him a liar. Thereupon God commanded him to build the ark. Nūḥ completed it and entered therein when he was 600 years old. And everyone drowned. Afterwards, he lived for another 350 years. The flood rose and covered the entire earth. Between Ādam and Nūḥ were 2200 years. The waters of the flood washed down the idols from the mountain Nawḏ to the land below. The current and the waves of the waters and the tide were rising from land to land, until [the waters] cast out the idols in the vicinity of Ǧudda. Then the waters flow off, and they [the statues] were left on the coast. The wind blew sand over them until they were completely covered. […] ʿAmr ibn Luḥayy was a kāhin, […] He had conquered Mecca, drove the Ǧurhum out of the city. He took over the custody of the House. […] ʿAmr proceeded to the shores of Ǧudda dug them [the idols] out, carried them until he reached Tihāma. The time of pilgrimage arrived, and he summoned all the Arabs to the worship of these idols. This second account does also explain the origins of idolatry with a corruption, however, this time it is not a corruption of the true faith due to human forgetfulness, but rather a corruption by a rival cult, originating from the descendants of Adam via his wicked son Cain (Qābīl). While the descendants of Adam via his righteous son Seth command over their ancestor’s tomb in a cave which they revere, the sons of Cain receive commemorative statues of their relatives—this happens, at least in the beginning, without an evil intention. However, as time passed, they began venerating these statues in the hope that they might intercede with God on their behalf. The Qurʾān repeatedly states that intercession (šafāʿa) by angels, jinn, other deities, or, as in this case, sacred ancestors, has no

558

klein

power and, moreover, is a form of association with God, hence idolatry.24 The idea of statues turning into objects of worship in the course of time, which is fundamental to Ibn al-Kalbī’s account, can be found most prominently in the Biblical Book of Wisdom:25 For the idea of making idols was the beginning of fornication, and the invention of them was the corruption of life […]. For through the vanity of men they entered the world, and therefore their speedy end has been planned. For a father, consumed with grief at an untimely bereavement, made an image of his child, who had been suddenly taken from him; and he now honoured as a god what was once a dead human being, and handed on to his dependents secret rites and initiations. Then the ungodly custom, grown strong with time, was kept as a law, and at the command of monarchs graven images were worshiped. When men could not honour monarchs in their presence, since they lived at a distance, they imagined their appearance far away, and made a visible image of the king whom they honoured, so that by their zeal they might flatter the absent one as though present. Then the ambition of the craftsman impelled even those who did not know the king to intensify their worship. For he, perhaps wishing to please his ruler, skilfully forced the likeness to take more beautiful form, and the multitude, attracted by the charm of his work, now regarded as an object of worship the one whom shortly before they had honoured as a man. And this became a hidden trap for mankind, because men, in bondage to misfortune or to royal authority, bestowed on objects of stone or wood the name that ought not to be shared. A similar way of thinking can be observed throughout the Kitāb al-Aṣnām and in other traditionalist works, e.g. in the legend on the couple Isāf and Nāʾila, who fornicated in the sacred precinct of the Kaʿba. They were petrified and their images, turned to stone, set up in Mecca as a warning for the inhabitants not to commit the same sin. However, as a perversion of the original intention, the statues became objects of veneration in themselves.26 The idea that

24 25 26

Cf. q6:94, 10:18, 30:13 and 72:6 (which has the jinn themselves as speakers) with Hawting 1999: 52–53. Wisdom 14:12–21, cf. also Hawting 1999: 101–103 and Stummer 1944: 385. On the statues of Isāf and Nāʾila, cf. Kitāb al-Aṣnām 6 as well as Aḥbār Makka i,119–122 with Hawting 1999: 101–103.

mourning for the dead and the beginning of idolatry

559

real historical events or personages over the course of time turned into mythological accounts or, in the case of venerated persons, into deities, is an old concept called euhemerism (after the Greek mythographer Euhemerus), and there are many examples for it ranging from Herodotus to Snorri Sturluson’s Prose Edda.27 However, as has been observed by Friedrich Stummer in his critique on Klinke-Rosenberger’s translation of the Kitāb al-Aṣnām, neither the account of the Book of Wisdom, nor similar ones by Christian authors, such as Minucius Felix, Lactantius, and Isidore of Seville, offer a suitable parallel to Ibn al-Kalbī’s second account on the euhemeric origin of paganism. Especially the author of the Book of Wisdom did not aim at explaining the origin of an ancestor or family cult, but rather the origin of the usage of images resulting from affectionate mourning; the critique is clearly directed against images, a common topos of Jewish and Christian polemics, and not against any potential ancestor cult.28 Similarly, the early Christian accounts are similar to that by Ibn al-Kalbī, yet not quite the same.29 Gerald Hawting pointed to a Jewish legend, collected by Louis Ginzberg, in which Enos wanted to demonstrate to his people how God created Adam and made a figure of dust and clay. In the moment when Enos tried to show how God had breathed life into the clay figurine, Satan animated it himself.30 While this is the only account which connects the origin of idolatry to the descendants of Noah/Nūḥ and also mentions the aspect of animation (which, however, in the Kitāb al-Aṣnām, is said to be impossible), it can only vaguely count as a parallel to Ibn al-Kalbī’s version.

∵ Ibn al-Kalbī’s second account has several characteristic traits: (1) the mention of various generations (2) with ‘exact’ numbers of years stated, (3) the lengthy genealogy of Noah (Nūḥ ibn Lāmak ibn Matūšalaḥ ibn Aḥnūḫ ibn Yarād ibn

27 28 29

30

On euhemerism, cf. Buxton 1994, Cooke 1927 and Winiarczyk 2002. Cf. Stummer 1944: 384. Cf. Stummer 1944: 385–387, esp. 385: “Doch muß gesagt werden, daß diese Stellen trotz aller nahen Verwandtschaft nicht ganz auf der gleichen Ebene liegen.” Stummer also pointed to the fifth-century account by Fabius Planciades Fulgentius (based on the thirdcentury author Diophantis) on a certain Egyptian called Syrophanes, who offered flower and incense sacrifices to his only and untimely-deceased son, as well as to Firmicus Maternus (c. 348), who offered a similar explanation for the origin of the cult of the god Liber. Cf. Ginzberg 1911–1938: i,22 with Hawting 1999: 103–104 with further references to similar stories in Ginzberg’s Legends of the Jews.

560

klein

Mahlāʾīl ibn Qaynān ibn Anūsh ibn Šīṯ ibn Ādam), (4) the burial of Adam in a cave (5) and the subsequent cultic veneration of this tomb, (6) the mention of a mountain in or on which the cave is situated, (7) the imitation of this cult by the descendants of Cain (Banū Qābīl), and (8) God’s repeated warnings to them, (9) Noah’s construction of the Ark, and finally, (10) the Deluge. It is not surprising that several of these elements can be found in Louis Ginzberg’s collections of Talmudic legends, since all these elements belong to a set of Jewish and Christian legends on Adam, which found their most famous literary expression in a Syriac text, the so-called Spelunca Thesaurorum or “Cave of Treasures.” The text is relatively well-studied, and it is surprising that the parallels between the report in the Kitāb al-Aṣnām and this source have never been explicitly mentioned. To my knowledge, this has only been alluded to in passing by H.S. Nyberg, who, however, did not discuss the matter any further.31 All elements of the version of Ibn al-Kalbī can be found in the Syriac text, whose main purpose is to demonstrate how Jesus Christ was descended from Adam. This becomes evident from the title of several manuscripts which call the text “The Book of the order of the succession of generations.” The more common title, Spelunca Thesaurorum seems to be a double allusion: it is both a repertoire of literary treasures and describes the legendary cave in which Adam was buried along with the offerings he had carried with him from the land of Paradise, i.e. gold, myrrh and frankincense, later to be vital in the birth narrative of the Lord. While the descendants of Adam’s son Seth dwell on the mountain, where also Adam and Eve were made to live after the expulsion from Paradise, they are contrasted with the descendants of Cain, who live at the foot of the mountain: The children of Seth live a life like angels, venerating the tombs of their patriarchal ancestors who are all buried in the magic cave, while the children of Cain are wicked, and constantly try to lure the righteous children of Seth away from the mountain.32 They do this by inventing all sorts of innovations (mostly musical instruments such as harps, flutes, and whistles), but also by presenting themselves quite open for fornication as well as other cardinal vices such as lasciviousness, drunkenness, dancing and singing, but also dying of cloths, horse racing, or henna tattoos. Despite continuous warnings and admonitions uttered by every leading patriarch (i.e. the male lineage from Seth down to Noah) not to do so, over the generations more and more descendants of Seth are tempted to go down from the mountain and visit the descendants of Cain. 31 32

Cf. Nyberg 1939: 357. Cf. Toepel 2006: 207–210, who investigated parallel traditions in Josephus, Antiquitates 1,67–73, which, however, has much less to say on the contrast between the two groups of Adam’s descendants.

mourning for the dead and the beginning of idolatry

561

Once they had slept with the beautiful and tempting women there, and wanted to return up to the holy mountain, its stones became fire: Having defiled their souls, God did not permit them to ascend to the holy place. Eventually, when Noah was on the mountain with only his closest family members, his wife and three sons and their wives, he built the Ark upon God’s command, removed the revered (and true) tombs from the cave and brought them into the Ark. God’s caused the heavens to release the rain, and the Deluge drowned the wicked descendants of Cain. The Syriac Spelunca Thesaurorum would only account as a highly similar parallel, which, however, was still somewhat removed from the Kitāb al-Aṣnām, were it not for the fact that this text was—in parts—translated quite early into Arabic.33 There are almost fifty manuscripts which give different titles to the translation, Kitāb al-Majāll as well as Apocalypse of Peter seems to be the oldest among them.34 The main difference, however, between Ibn al-Kalbī’s version of the story and the Spelunca Thesaurorum is that the intention to create images which later on become venerated as idols is not connected to the time before the Deluge in Ibn al-Kalbī’s source which clearly states that God sent raging winds which tore away the images and their worshippers. The text goes on to mention that some assert that in the time of Abraham’s father Terah there was a deluge of wind, just to declare that this is nonsense, for image-worship happened after the Deluge of water, and the Deluge was not sent upon them for the worship of images but because there was so much corruption on the earth among the children of Cain.35 The Spelunca Thesaurorum and its Arabic adaptation, the Kitāb al-Majāll, expand on certain defining moments in the origins of idolatry; at first, rather generally:36

33

34

35 36

Cf. Götze 1924:51 (suggesting a translation towards the end of Umayyad rule) as well as passim for the text’s influence on (in varying degrees) on Ps-Dionysios Areopagita, PsMethodios of Patara, Ps-Epiphanios of Cyprus, and, especially, the Ethiopic Gadla Adam. Other passages of the Arabic version of the Spelunca Thesaurorum were used by at-Tabarī and al-Yaʿqūbī, cf. Götze 1924, 61 and Minov 2009: 16. Cf. Götze 1922: 23–37 and Toepel 2006: 9–13. The Arabic ‫مجال‬, as a plural form of ‫مجلة‬, seems to go back to Hebrew ‫( מגלה‬book scroll), via the Syriac ‫ܐ‬狏‫ܡܓܠ‬. However, it cannot be excluded that the title derives from ‫جلا‬, ii. “to reveal”, hence revelation. On the Kitāb alMajāll in general, cf. Roggema 2007: 133–137 as well as 138–140 (for her dating of the entire compilation to ʿAbbāsid times), and Grypeou 2007: 116–117. Cf. Wallis Budge 1927: 29–31. Cf. also Ri 2000: 313.

562

klein

Spelunca Thesaurorum (ed. Ri 1987) 25,13–14 (Ms.Oc.).

Kitāb al-Majāll (ed. Gibson 1901) fol. 119a

.‫ܘܢ‬煿‫ ܡܢ‬煟‫ܬ ܗܘܐ ܚ‬焏‫ܐ ܕܡ‬煟‫وكان احدهم اذا مات صنع اهله صنما على شبهه ܡܚ‬ .‫ܬܗ‬熏‫ܡ‬煟‫ ܒ‬焏‫ ܨܠܡ‬煿‫ ܗܘܘ ܠ‬爯‫ܝ‬煟‫ܥܒ‬ 犯‫ ܢܥܒ‬焏‫ܗ ܕܠ‬犯‫ ܩܒ‬爏‫ ܗܘܘ ܥ‬爯‫ܘܣܝܡܝ‬ 狏‫ ܘܐܙܕܪܥ‬.‫ܗܘܢ‬狏‫ ܒܝܢ‬爯‫ ܡ‬.煿‫ܗܕܢ‬熏‫ܥ‬ 狏‫ ܘܐܬܡܠܝ‬.‫ܘܟ‬煟‫ܐ ܒܟܠ‬狏‫ܗܕܐ ܒܝܫ‬ ‫ܬܐ ܕܕܟܪܐ‬熏‫ܡ‬煟‫ܟܪ ܒ‬狏‫ ܦ‬焏‫ܐܪܥ‬ .‫ܐ‬狏‫ܘܢܩ̈ܒ‬ When one of them died, his people made an image in his likeness, and put it upon his tomb, lest, so they said, his memory should be forgotten among them. And this evil matter was disseminated everywhere, and the earth was filled with idols in the form of males and females.

‫ وامتلت‬.‫ونصبوه على قبره ليلا ينقطع ذكره‬ ‫الارض خطايا وكثر فيها الاوثان المصنوعة على‬ .‫تماثيل الذكران والاناث‬

When one of them died, his people made an image in his likeness, and put it upon his tomb, lest his memory should be cut off. The earth was filled with sins, and idols were multiplied in it, made in the likenesses of males and females.

The second mention, following shortly after the first one, locates the creation of images in the time of Serug, the grandfather of Abraham:37

Spelunca Thesaurorum (ed. Ri 1987) 26,2–3

犯‫ܝ‬狏‫ ܥ‬焯‫ ܘܛ‬煟‫ܐ ܚ‬犯‫ ܓܒ‬煿‫ܘܗܘܐ ܒ‬ 煿‫ ܠ‬煟‫ ܘܥܒ‬.焏‫ܘ ܙܒܢ‬煿‫ ܒ‬狏‫ ܘܡܝ‬.‫ܗܘܐ‬ ‫ܬܗ‬熏‫ܡ‬煟‫ ܒ‬焏‫ ܕܕܗܒ‬焏‫ܗ ܨܠܡ‬犯‫ܒ‬ ‫ ܘܐܘܬܒ‬.‫ܥ‬犯‫ ܩܒ‬爯‫ ܡ‬爏‫ ܠܥ‬爟‫ܘܣ‬ .‫ܘܗܝ‬犯‫ ܕܢܛ‬煟‫ ܚ‬焏‫ܥܠܝܡ‬

37

Cf. Ri 2000: 314–315.

Kitāb al-Majāll (ed. Gibson 1901) fol. 119b ‫ فصنع ابنه‬.‫وكان بدوه ان رجلا موسرا توفى‬ ‫صنما من ذهب ونصبه على قبره على رسم اهل‬ .‫ ووكل به رجلا شابا يحفظه‬.‫عصره‬

mourning for the dead and the beginning of idolatry

563

Spelunca Thesaurorum (ed. Ri 1987) 26,2–3

Kitāb al-Majāll (ed. Gibson 1901) fol. 119b

In this time, there was a man who was very rich, and he died. His son made an image of his likeness for him, and he placed it on his tomb, and he appointed a young man to guard it.

The beginning of it was that a rich man died; his son made a golden idol of him and placed it upon his tomb as a mark to the people of his age, and appointed a young man to guard it.

Especially the latter tale unfolds the deadly implications of idolatry, when the devil entered into the image of the deceased father and began to speak with his voice. A little later, robbers took all that belonged to the mourning son, who, nevertheless, continued to visit the grave of his father, whose devilish voice tricked the credulous mourner into sacrificing his own son, i.e. the grandson of the deceased.

∵ While there are certain differences in the arrangement of the narratives, the Syriac Spelunca Thesaurorum via its Arabic translation, the Kitāb al-Majāll, appears to be the direct source of Ibn al-Kalbī’s second account on how idolatry entered this world. This has implications for the two enigmatic first verses of q102: Either, following Gerald Hawting, Ibn al-Kalbī’s text was only invented to explain the verses; however, since the Arabic version is not a genuine text or compilation of early Islamic times, but a free translation of a Christian Syriac text, this appears to be highly unlikely. Or, and this seems to be the more likely case, Ibn al-Kalbī incorporated an originally Christian narrative into his Muslim explanation of pre-Islamic idolatry, thus preserving a genuine story perhaps going back to the environment of Muḥammad and the genesis of the Qurʾān (labelled “Umwelttexte” in Berlin’s Corpus Coranicum project). In the latter case, versions of the Arabic text of the Kitāb al-Majāll must have been circulating in early Islamic times and influenced theological thinking of the age, just as previous scholarship has assumed in the first place. The narrative transmitted in the Kitāb al-Aṣnām may provide a clue for both the Qurʾānic terms “worldly gain” (takāṯur) as well as “graves” (maqābir)—and especially for the close proximity in which they are mentioned together. While Muslim tradition is relatively vague on what kind of intercession (šafāʿa) exactly is

564

klein

to be hoped for by those venerating images, it seems sensible to assume that these aspirations were not only restricted to personal aspects, such as family matters (e.g. “fertility”, i.e. takāṯur fī-l-awlād), but also financial and mercantile gain and profit (i.e. takāṯur proper). Be it at different positions within the overall narratives, both the Kitāb al-Majāll and the Kitāb al-Aṣnām connect these hopes with the custom of visiting the tombs of ancestors—be it direct ones or, rather, those whose memory has long been forgotten and turned into worship.38

Bibliography Athamina, Kh. 2004. “Abraham in Islamic perspective: reflections on the development of Monotheism in pre-Islamic Arabia.” Der Islam 81: 184–205. Buxton, R. 1994. Imaginary Greece: The Contexts of Mythology. Cambridge: Cambridge University Press. Cooke, J. 1927. “Euhemerism: a mediaeval interpretation of Classical paganism.” Speculum 2: 394–410. Crewswell, K.A.C. 1952. Muslim Architecture of Egypt i, Ikhshids and Fatimids, a.d.939– 1171. Oxford: Oxford University Press. Eksell, K. 2005. “The verb wjm in Safaitic inscriptions.” In Alltagsleben und materielle Kultur in der arabischen Sprache und Literatur: fs Heinz Grotzfeld, edited by Thomas Bauer and Ulrike Stehli, 163–172. Wiesbaden: Harrassowitz. Faris, N. 1952. trans., The book of idols. Princeton: Princeton University Press. Gibson, M. 1901. Apocrypha Arabica. London: Clay and Sons. Ginzberg, L. 1911–1938. The legends of the Jews. Philadelphia: Jewish Publication Society of America. Götze, A. 1922. Die Schatzhöhle. Überlieferung und Quellen. Heidelberg: Winter. Götze, A. 1924. “Die Nachwirkungen der Schatzhöhle.” Zeitschrift für Semitistik und verwandte Gebiete 2: 49–94. Grypeou, E. 2007. “The re-written Bible in Arabic: the Paradise story and its exegesis

38

To end on a rather speculative final note, we can imagine that this practice was not too dissimilar to the inscription of a certain ʾsd son of ẓnn son of ṯlm, who in all likelihood might have had no personal recollection of or connection to Hāniʾ, but still hoped to invoke the divine powers of al-Lāt and Ḏū š-Šarā in the presence of the widely visible cairn, f-h-LT w-DŠR slm w-qbll, to grant him security and reunion with his loved ones. Cf. Harding 1953: 23 (no. 42) with Al-Jallad 2015: 333 for the meaning of qbll.

mourning for the dead and the beginning of idolatry

565

in the Arabic Apocalypse of Peter.” In The Bible in Arab Christianity, edited by David Thomas, 113–129. Leiden/Boston: Brill. Harding, G.L. 1953. “The Cairn of Haniʾ.” adaj 2: 8–56. Hawting, G. 1999. The idea of idolatry and the emergence of Islam: from polemic to history. Cambridge: Cambridge University Press. Al-Jallad, A. 2015. An outline of the grammar of the Safaitic inscriptions. Leiden/Boston: Brill. Jamme, A. 1967. “The Safaitic verb wgm.” Orientalia 36: 159–172. Jones, A. 2007. The Qurʾān. Exeter: Gibb Memorial Trust. Kennedy, D. 2012. “The Cairn of Hānī: significance, present condition and context.”adaj 56: 483–505. Klinke-Rosenberger, R. 1941. Das Götzenbuch—Kitāb al-Aṣnām des Ibn al-Kalbī: Übersetzung mit Einleitung und Kommentar. Leipzig: Harrassowitz. Künstlinger, D. 1931. “Einiges über die Namen und die Freuden des ḳurānischen Paradieses.” bsoas 6: 617–632. Leisten, T. 1990. “Between Orthodoxy and exegesis: some aspects of attitudes in the Shariʿa toward funerary architecture.” Muqarnas 7: 12–22. Malḥas, R., (ed). 1983. Abu ʾl-Walīd al-Azraqī: Aḥbār Makka wa-mā ǧāʾa fīhā min al-āṯār. Beirut/Mecca: Dār al-Andalus/Dār aṯ-Ṯaqāfa. Minov, S. 2009. “Адам и Ева в сирийской «Пещере сокровищ».” Символ 55: 9–46. Neuwirth, A. 2011. Der Koran. Band 1—Frühmekkanische Suren. Poetische Prophetie. Handkommentar mit Übersetzung. Berlin: Verlag der Weltreligionen im Insel Verlag. Nyberg, H.S. 1939. “Bemerkungen zum “Buch der Götzenbilder” von Ibn al-Kalbī.” In δραγμα: fs Martin Nilsson, edited by Krister Hanell, 346–366. Lund: Gleerup. Ri, A.S. (ed). 1987. La Caverne des Trésors: les deux recensions syriacques. Louvain: Peeters. Ri, A.S. (ed). 2000. Commentaire de la Caverne des Trésors: étude sur l’histoire du texte et de ses sources. Louvain: Peeters. Roggema, B. 2007. “Biblical Exegesis and Interreligious Polemics in the Arabic Apocalypse of Peter—The Book of the Rolls.” In: The Bible in Arab Christianity, edited by David Thomas, 131–150. Leiden/Boston: Brill. Paret, R. 1977. Der Koran: Kommentar und Konkordanz. Stuttgart: Kohlhammer. Shanklin, W. and Dark, A. 1953. “Anthropological studies on two skulls from Jordan.” adaj 2: 57–61. Stummer, F. 1944. “Bemerkungen zum Götzenbuch des Ibn al-Kalbî.”zdmg 98: 377–394. Toepel, A. 2006. Die Adam- und Seth-Legenden im syrischen Buch der Schatzhöhle. Eine quellenkritische Untersuchung. Louvain: Peeters. Wallis Budge, E. 1927. The Book of the Cave of Treasures: a history of the Patriarchs and the Kings their successors from the Creation to the Crucifixion of Christ. London: The Religious Tract Society.

566

klein

Wellhausen, J. 1961. Reste arabischen Heidentums. Berlin: DeGruyter³. Winiarczyk, M. 2002. Euhemeros von Messene: Leben, Werk und Nachwirkung. Munich/ Leipzig: Saur.

chapter 28

Temple Inscriptions and “the Death of the God(s)” John F. Healey

In a volume dedicated to Alan Millard (2005), to which Michael Macdonald also contributed, I discussed the tradition of legal injunctions being displayed in public places in the Near East of the Hellenistic-Roman period. I present here some footnotes to that discussion as a small contribution to the honouring of a scholar whose work has had a profound effect not only in the study of ancient Arabia, but also in adjacent fields of study, including Aramaic epigraphy in general and Nabataean epigraphy in particular.

a

The Jerusalem Inscriptions

Josephus tells us in several places about inscriptions on the balustrade of the Jerusalem Temple, in Greek and Latin, warning gentiles from entering the courtyard to which Jews alone were admitted. He also tells us that the Roman authorities gave the Temple authorities the power to execute transgressors, even if they were Roman (War v, 194; vi, 124–128; Ant. xv, 417; viii, 95; xii, 145– 146). Thus Titus, according to War vi, 125–126, asked: Was it not you [the Jews] that ranged along [the balustrade] those slabs, engraved in Greek characters and in our own [i.e. in Latin], proclaiming that none may pass the barrier? And did we not permit you to put to death any who passed it, even were he a Roman? Trans. thackeray 1928: 215

The Greek text of one of these inscriptions, found in 1871, was an amazing confirmation of what Josephus says (Clermont-Ganneau 1872) and another fragment (with slightly different wording) was discovered in 1935 (Iliffe 1936). The many discussions of these inscriptions include an important article by Bickerman (1945–1946), the Corpus Inscriptionum Iudaicarum (Frey 1952: no. 1400, 328–330) and ciip (§2, 42–45, with full bibliography). The Clermont-Ganneau stone is on public display in the Istanbul Archaeological Museum (Inv. 2196 t):

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

568

healey

No alien may enter within the balustrade around the sanctuary and the enclosure. Whoever is caught, on himself shall be put blame for the death which will ensue. bickerman 1946–1947: 388

As Bickerman pointed out, “The punishment of ritual transgressions was generally left to the divinity” (1946–1947: 394). On the other hand, the section of Josephus War vi quoted above implies that the Temple authorities had the power to impose capital punishment in such cases after a trial. He concluded that the final sentence in the inscription effectively means “or else!”, though specifying death as the consequence. Here we may note the occurrence in Rabbinic discussions of the phrase myth bydy šmym, “death by the hand(s) of heaven”, the concept being that the actual outcome is left to God (Bickerman 1946–1947: 395–398; Kaizer 2006: 150, citing the Babylonian Talmud bSanh. 52b). The idea of restriction of entry to temples is quite well attested in the GraecoRoman world. Samothrace provides examples dated c. 200 ce with inscriptions in Greek or in both Greek and Latin placed near the entrance to a temple: ἀμύητον μὴ εἰσιέναι εἰς τὸ ἱερόν The uninitiated may not enter the temple. fraser 1960: 117–118, no. 62, pl. xxiv; see also 118–120, no. 63, pl. xxiv [the bilingual]; clinton 2003: 61–62

Another oft-cited example comes from Tralles in Lydia (Robert 1937: 415–416; Poljakov 1989: 200, no. 245). In my 2005 publication I also mentioned, from nearer to Jerusalem, a Greek inscription probably of the third century ce and now in the British Museum (bm: 1903.0422.1), referring perhaps to Baʿal Ḥermōn (Judges 3: 3; i Chron. 5: 23), from the temple on the summit of Mount Hermon. This has been subsequently re-edited by J. Aliquot, whose Greek text I reproduce here (Aliquot 2008a: 72– 73, no. 40): κατὰ κέλευσιν θεοῦ μεγίστου κ[ὲ] (= καὶ) ἁγίου, ὑ (= οἱ) ὀμνύοντες ἐντεῦθεν By the order of the greatest and holy god. From here (inwards) only those who take the oath. clermont-ganneau 1903; aliquot 2008a: 72–73, no. 40; see also aliquot 2008b: 82–84

temple inscriptions and “the death of the god(s)”

569

In these cases, the prohibition relates to the exclusion of initiates or professed devotees. In the case of Mt Hermon aspects of the possible cult and some literary allusions to it are discussed by Aliquot (2008b). In Jerusalem the issue of entry to the sacred area is one of ethnicity (“Jewishness”), though there were other restrictions imposed also on Jews entering the Temple (arising, e.g., from gender and from purity regulations). It always seemed surprising to me that so few inscriptions were known from the Jerusalem temple, e.g. inscriptions recording donations to the Temple. Of course the destruction of the Temple in 70ce would account for this to some extent, but even so the comparison, for example, with the temples in Palmyra explains the surprise of the discovery of the Jerusalem inscription: finding such an inscription in Palmyra would be much less remarkable. In the Palmyrene temples and especially the Temple of Baʿalshamīn we have a whole series of donation inscriptions, referring to porticoes, entablatures and other architectural features. Thus in Dunant 1971 we have dedications of porticoes (with columns, entablatures and roofs: no. 1–8, 9 [walls and roofs]), columns (10, 11, 12, 13 [columns and roof], 14 [columns, entablature, roof], 15, 16, 17 [columns, entablature and roof], 18, 19 [capitals (?), entablature, roof], 20 [column, capital, entablature, roof]), not to mention altars and statues, which would not in any case be expected to be found as common features in Jerusalem. The Temple of Bēl at Palmyra is less rich in building inscriptions but similar are Yon (2012) no. 16, 18, 21 (bronze doors), 23 (doors, ḥammānā, andrōn roof), 24 and inscriptions dedicated to building workers (e.g. 37–38). For the Temple of Nabū see Yon (2012) no. 183 and 184 (porticoes) and for the Temple of Bēl Ḥammōn on Jabal Munṭār see Inv. xii, 48 and 49. We may note also Yon (2012) no. 324 (referring to a banquetting-room) from the Afqā spring. This is not, however, the end of the story. The publication of the Corpus Inscriptionum Iudaeae et Palestinae in recent years has collected in one place the epigraphic remains from the Temple Mount, some of which have come to light recently (Cotton et al. 2010; henceforth abbreviated as ciip). Here, in addition to the Josephus inscription (ciip § 2), there is ciip § 3, a badly damaged Greek inscription first published in 1983, probably from the Herodian Temple and recording the donation of funds for a pavement (18– 17 bce). What is particularly interesting here from the point of view of comparison with temples at places like Palmyra is the fact that a donor (here a donor from Rhodes, Paris son of Akeson) is recording a donation to the Temple, specifically, it seems, in relation to the installation of a pavement (στρῶσις). This fragmentary inscription appears to come from the destruction debris of the Herodian Temple and may relate to some part of the Herodian pavement. Such inscriptions, we must assume, were commonplace in the Temple area, though

570

healey

there are possibly particular features of Jerusalem which need to be taken into account (see Schwartz 2009: on §3 see his p. 77). Again ciip § 4 is an Aramaic inscription of the 1st century ce mentioning a personal name, “Mattiya son of …” and it too probably related to a donation to the work of Temple-building. ciip §5 is in Hebrew and is of similar date, but is in its function very different: it appears to mark a particular place, the byt htqyʿh, the place where the trumpet is to be sounded, while ciip §6 and § 7, also of the first century ce, are not so clearly related to the Temple itself, containing only fragmentary personal names. ciip §8, which is given a broader dating of 1st century bce to 1st century ce, appears to be directly related to the Temple cult. It is on the handle of a limestone jar bearing the word qrbn, “offering, sacrifice”. There are other more obscure items of a similar general type from the Temple Mount area in the Corpus: ciip §10, 11, 12, 13, 16, 17. ciip §9 is a long Greek inscription dated to the 1st century bce to 1st century ce. This is the well-known inscription of Theodotos, found with other building fragments south of the Temple mount and first published in the 1920s. The inscription indicates that the building provided hospitality, as well as a place εἰς ἀνά[γ]νωσ[ι]ν νόμου καὶ εἰς [δ]ιδαχ[ὴ]ν ἐντολῶν, “for the reading of the Law and the teaching of the commandments”, especially for foreign visitors to Jerusalem, probably while the Temple still existed. These inscriptions from the Temple area suggest that Herod’s Temple was provided with many dedicatory inscriptions, including many commemorating particular donations for particular purposes. This is hinted at in Luke 21: 5, where Jesus is asked to comment on the ἀναθήματα decorating the Temple. The Theodotos inscription also provides a link with the many later synagogue inscriptions (see the collections in Naveh 1978 and Fitzmyer-Harrington 1994, 151–303). These can be seen as a continuation of the epigraphic habit which had been current in the Temple itself until its destruction. Again many of these inscriptions refer to particular donations, which in a Roman context would have been termed acts of euergetism (but see Schwartz 2009). For example the Beth Gubrin synagogue inscription of the 4th–6th centuries ce: Remembered for good be Kyrios [son of She]ʿay—Repose (for his) soul!— son of Auxentios, who purchased this column in honour of the congregation. Peace! frey 1952: no. 1195, 232–233; naveh 1978: no. 71, 109–112; trans. fitzmyerharrington 1994: no. a12, 259

temple inscriptions and “the death of the god(s)”

b

571

Jerusalem, Hatra and “the Death of the God(s)”

The Jerusalem inscription reported by Josephus is also remarkable for the nature of the punishment which is indicated for transgressors, i.e. the fact that there will be an ensuing death, though it is not specified what form that death would take. The report in Josephus (above) suggests that the Romans devolved jurisdiction to the local Temple authorities in this matter. How the injunction actually worked (or was imagined to work) in Jerusalem is, however, unknown. In my 2005 paper in the Millard Festschrift I made comparisons with similar punishments enjoined by legal inscriptions at Hatra. Shortly thereafter Kaizer published an important article on capital punishment at Hatra (2006). There is no need to rehearse his excellent discussion, but one of the loose ends left both by his treatment of the Hatra inscriptions and by mine was the phrase “the death of the god(s)”, threatened in four inscriptions as a divine sanction against wrong-doers. All four inscriptions were found at city gates of Hatra, the northern and eastern gates, rather than in particular temples, but they clearly relate to temple regulations and were on public display. Only one of the four (no. 342 below) mentions specific deities, referring to Nergol initially as source of the decree and then to Māran, Mārtan and Barmārēn. The relevant parts of the four inscriptions in question read as follows (generally following Healey 2009: no. 77–79, 302–309): Hatra No. 342 (1) hkyn psqw nrgl … wsnṭrwq mlkʾ (4) … dy kwl zmrtʾ (5) wqyntʾ dy mrn wmrtn (6) [w]brmryn dy tšbw[q] mqmh (7) wtʾzy[l m]n dkʾ bmwtʾ (8) dy ʾlhʾ [tmwt w]dy lnpq (9) zmrt[ʾ …] lh gbʾ (10) bmwtʾ dy ʾ[l]hʾ lmwt hw (11) [g]brʾ … Thus have decided Nergol … and King Sanaṭrūq … that any female musician or singer of Māran and Mārtan and Barmārēn who leaves her place and departs from office will die by the death of the god(s) and he who enables the (said) female musician to go out … for her a hiding-place, by the death of the god(s) that man will die.

572

healey

Hatra No. 344 (4) šlhy ʾnš dlzbyn lk[p]ʾ … (9) wʾyn lzby[n] mnhwn … lmwt mwtʾ {dy} (13) ʾlhʾ It is forbidden (?) that anyone should buy stone [etc.] … and if he sells any of them …, he will die the death of the god(s). Hatra No. 343 (Dated 151–152ce; No. 336 is Almost Identical) (5) … kwl dlgnwb lgw mn mlʾ hdyn (6) wlgw mn šwrʾ bryʾ ʾyn gbrʾ (7) hw gwyʾ lqṭyl bmwtʾ dy (8) ʾlhʾ wʾyn gbrʾ hw bryʾ (9) lrgym … anyone who steals within this entrance-ramp (?) and within the outer wall [of the city], if he is a native he will be killed by the death of the god(s) and if he is an outsider he will be stoned to death. Two details should be noted immediately. Firstly, the alternation between the verb qṭl (343: 7, also 336: 10) and the verb mwt (342: 10; 344: 12). Whereas lmwt could be either “he will die” or “let him die”, not necessarily implying the imposition of capital punishment, the phrase lqṭyl means “he will be killed” or “let him be killed”, which necessarily involves an action taken by someone to cause the offender’s death. It follows that lmwt too must be understood in this way: it involves an executioner executing a death sentence (though it seems just about possible to imagine “plague”, e.g., acting as executioner with the verb qṭl used: “he will be killed by plague”). A second issue which affects the discussion which follows concerns ʾlhʾ. So far as the form is concerned, it is impossible to be sure whether ʾlhʾ is to be read as a singular or a plural and interpreted as “god” or “gods”. Beyer translated it as a plural (1998: ad loc. and 145), while Vattioni, Aggoula and Kaizer translated it as singular (Vattioni 1994: 64, etc.; Aggoula 1991: 158, etc.; Kaizer 2006: 143, etc.). Favouring slightly the plural interpretation is the fact that no. 344: 8 makes it clear that the rule in question applies to the byt ʾlhʾ, which must, in the context, mean the whole sacred precinct, “the temple of the gods”, and no specific deity is mentioned in the inscription. Also, in no. 343 and 336 there is reference at the beginning of each to bmlkʾ dy ʾlhʾ without further specification. The lack of specification again makes it more likely that the meaning is “by the counsel of the gods”. If these inscriptions had been found in the temple of a particular god, there would be less of a difficulty here: the specification of which god was involved would be unnecessary, as it was in the Greek inscription of 227ce from

temple inscriptions and “the death of the god(s)”

573

Iraq cited by Kaizer (2006: 147; see Seyrig 1954: no. 1, 212–214), which names no particular god: … ἐκ κελεύσεος θεοῦ κὲ (= καὶ) ἀρχόντων …, “… in accordance with the command of the god and of the officials …”. Inscription no. 343 was found, however, at Hatra’s eastern gate, not in a temple, and no. 336 at the northern gate. If “counsel of the gods” is correct, ʾlhʾ would almost certainly be plural also in line 10: the gods authorized the law in question and the gods would act to enforce it. In no. 342, where deities are mentioned, ʾlhʾ could refer to all of them rather than just one: Nergol appears at the beginning of the text, but the divine triad intervenes before the ambiguous form ʾlhʾ. This constitutes the argument in favour of “the death of the gods”. The issue cannot, however, be regarded as resolved and the interpretation of the word ʾlhʾ as singular would be attractive in the context of some cuneiform evidence which is adduced below. But before turning to that evidence we may note that the Hatra inscriptions refer to a much clearer alternative to “the death of the god(s)” in the form of stoning to death (no. 343 and 336 above and also no. 281). Even before the discovery of the Hatra inscriptions, stoning to death was already known to be a feature of Hatran legal practice, because it is specifically mentioned by Bardaiṣān (154–222ce), who in The Book of the Laws of Countries (Drijvers 1965) gives long lists of typical actions of the pagans of different regions. His list is a strange mixture of confused reports (such as the idea that the Germans and Gauls practised male same-sex marriage: Drijvers 1965: 48, l. 6–10), reports which may have a grain of truth in them but are oversimplified or exaggerated in some way (the Persians marry their sisters: Drijvers 1965: 42, l. 21–23), and reports which are very specific and plausible. With regard to the latter category we can reasonably assume that Bardaiṣān, who was a court philosopher of the Edessan king, Abgar the Great (177–212ce), knew first-hand about local customs in Edessa. So when he says that the Edessans stone to death their adulterous wives and sisters, he is probably telling us what he really knows. He advocates the supposedly more Christian and lenient approach of shunning them: I only use the adverb “supposedly” because shunning in a society like that of northern Mesopotamia, in which, one supposes, a shunned woman would be indigent and possibly reduced to prostitution, could not in modern terms be regarded as a particularly generous-minded reaction to deviation from society’s norms. But one of Bardaiṣān’s examples of how the pagans behave and the Christians, by implication, should not and do not imitate them (as the unavoidable result of being born in a particular place under particular constellations), is the statement that at Hatra thieves are stoned to death: “In Hatra there is the

574

healey

law that everyone who steals a trifle, if only water,1 is punished with stoning” (Drijvers 1965: 46, l. 17–18, and see 60, l. 12–13). This literary allusion is clearly confirmed by the Hatra inscriptions cited above, which specify stoning in relation to theft and other offences. Indeed the offences and their punishments were written up on walls for all (or many) to see. We may note, incidentally, that the verb rgm here and elsewhere cannot be adequately translated as “to stone” in English: in this kind of context it means not “to stone” but “to kill by stoning”. It is a form of capital punishment as it obviously is in the biblical and other contexts. So, e.g., in the Peshitta of Acts 7: 58–59 (= Greek λιθοβολέω). Two of the relevant Old Testament passages, Lev. 24: 14–16 and Num. 15: 32–36, are particularly interesting, since the community awaits the judgement of the Lord before deciding what punishment to impose, though the issue is not a matter of leaving it to the Lord to indicate which form of capital punishment should be used. So much is clear, but then we have an alternative punishment, the rather obscure threat of being killed by “the death of the god(s)”. The phrase “the death of the god(s)” could, if ʾlhʾ is taken as singular, mean “the death determined by oracle from the god in question, against whose temple the transgression was committed, and therefore in the hands of priests who would determine precisely how the offender would die, which might not be by stoning but by some other means”. This could be implemented through consulting the god mantically. Such a procedure would echo the consultations in Lev. 24: 14–16 and Num. 15: 32–36 (above) and the rather elaborate discussions of forms of capital punishment found in the Babylonian Talmud (see below). Dirven (2009: 58–60) connects “the death of the god [singular]” with NergalHerakles, nrgl dḥšptʾ, “Nergal, Lord of the Guards”. The title dḥšptʾ is of Iranian origin, giving Nergol the role of executioner, but, as Dirven notes, Barmārēn too can impose capital punishment by stoning (no. 281). The connection with Nergol/Nergal in our “death of the god(s)” texts would be more secure if the inscriptions in question came from a specific temple, such as the ones which have other Nergal-Herakles associations (Temples ix and x) rather than gateways. It is thus left uncertain which individual god would be involved. Alternatively, if ʾlhʾ were plural, the meaning of the phrase might be rather a generic “divine capital punishment”, “the death of the gods”, with the precise

1 The wording here is a little odd and there have been alternative readings, one of which would suggest “if it is only worth an obolus” (Tubach 1986: 228 and references).

temple inscriptions and “the death of the god(s)”

575

meaning clear to all readers, but not to us. In this case we can only speculate on what alternatives to stoning might have been available or usual—burning? strangulation? hanging? beheading? burying alive? crucifixion? Kaizer (2006: 149–150) mentions the possibility of death by pestilence. As he notes, however, such a death would not be a predictable and enforceable legal consequence of breaking the law: it would be contingent on external events, so it seems unlikely as a legal punishment. There is an echo of this vagueness in the Jerusalem inscription’s reference to “the death that will ensue”, a wording which is again unspecific. The Mishnah indicates four approved methods of capital punishment: stoning, burning, beheading and strangling (Mishnah Sanh. 7: 1) and Kaizer (2006: 150) adduces some interesting parallels in Jewish legal literature, especially “death by the hand of heaven” in the Babylonian Talmud, contrasted with “death by the hand of a human being” (bSanh. 52b; see Steinmetz 2008: 4–6; also bKeth. 37b). It may be noted that in at least one instance in the Talmud it is stated that this “death by the hand of heaven” is a less severe form of capital punishment (bB.M. 56a, and cf. bKeth. 37b). I want to add here another subset of evidence to this discussion and suggest a partial solution to this series of problems. “The death of the god” (the singular creating a better parallel) could be a late reflex of an earlier formula, until now unnoticed in this context (I think!), which occurs in Akkadian. The details of this earlier evidence are a little complex. cad m/ii, 318–319, within the entry for mūtu, “death”, records usages where this word is qualified by a following word: mūt šīmti, “natural death, death according to destiny, death at the natural or appointed time”, and mūtu ilīšu, also translated in cad as “natural death, the death (decreed) by his god” (but see Fincke below). The former also appears in the negative form: mūt lā šīmti, “unnatural death”. Two recent publications suggest that cad does not have the last word on this. Fincke (2013–2014) discusses the use of the phrase mūtu ilīšu in omen texts of the second and first millennia bce, where this kind of death, “the death of his god”, is predicted in apodoses. Fincke notes that cad had taken the kind of death involved to be the death decreed by his god and to refer to natural death, but she observes that both kinds of death appear in omen apodoses (even on the same tablet) and this she takes to show that there is some difference between the two types of death. mūt šīmti is clearly “natural death, death of destiny”, as also in the verbal phrase ana šīmti alāku, “to go to one’s fate/destiny”. mūtu ilīšu, therefore, cannot mean “natural death”. It might refer to death through sickness: Fincke notes the phrase qāt ilīšu, “the hand of his god”, in diagnostic apodoses referring to disease, the outcome suggested by particular

576

healey

symptoms.2 Some examples are to be found in the cad entry under mūtu 3’. In a Mari text, the death of Zuzu, former king of Apum, was caused by an illness like dropsy or because of an accident arising from the fact that his own personal god had turned against him: mūt ilīšu-ma umūt-ma (Fincke 2013–2014). Fincke’s note was taken up for further comment by Attinger (2014), who draws attention to a literary passage, Epic of Gilgamesh xii 146 f. (George 2003: i, 734–735), where the man who dies by “the death of his god” (George: “a natural death”) enjoys a state of repose in the afterlife, in relative comfort compared with less fortunate souls. This might at first seem to cast doubt on the idea that “the death of his god” is a punishment or negative outcome, as in Fincke’s examples. Attinger does not, however, draw this inference: he argues rather that death was always regarded in a negative way in Mesopotamian culture and hence there would be no specifically negative implications for the afterlife of someone who died through “the death of his god”. The basic contrast in the cuneiform sources seems to be between natural death and death outside the normal order of things. And this seems to be supported by mūt lā šīmti, “untimely death”, the opposite of mūt šīmti, which appears in the Tukulti-Ninurta Epic as follows: The [ ] have steadily cast down in untimely death, Their [wives] are become widows in undue season. … until I expose your hair fluttering behind you and you have [disappeared] to an untimely death, … translated by foster 2005: 306; in his line numbering iii [= a obv.], 7, 17

It seems that mūt šīmti is “natural death”, while both mūt lā šīmti and mūt ilīšu are not in this category. This Mesopotamian evidence immediately struck me for its possible relevance to the Hatran “death of the god”. In the Hatran context the situation is similar to that of an omen apodosis, since the laws are essentially casuistic and in the Hatran case too there is a contrast clearly intended by the reference to two different types of death, by stoning and by the “death of the god or gods”. In the Hatra context, both eventualities arise from legal infringements and are negative consequences to be experienced by different categories of offenders: on the one hand non-Hatrans, who are subject to stoning, on the other natives, who are to suffer the “death of the god(s)”.

2 Note the parallel with “the hand of heaven” in the Talmud, cited above.

temple inscriptions and “the death of the god(s)”

577

The cuneiform evidence gives us a possible context for the phrase mwtʾ dy ʾlhʾ. It does not, however, help us in working out what specific kind of death was implied in the phrase: as Kaizer noted, the only thing that is clear on this question is that mwtʾ dy ʾlhʾ was not death by stoning (Kaizer 2006: 149). To judge from the Mesopotamian parallels it might have been untimely and unnatural (i.e. not simply as the result of old age). Since, as Kaizer rightly noted, this could not be a formal punishment (being open to the possibility of never being implemented), an alternative, I suggest, might be that lqṭyl/lmwt bmwtʾ dy ʾlhʾ is actually a curse: “May he be killed/die out of due time, at the time determined by the god”. This would be a curse applied to Hatrans, contrasted with immediate stoning to death applied to non-Hatrans. In support of this approach one might cite the way that a number of Hatran inscriptions contain curse formulae beginning with bgn, “invocation/curse”, followed by the names of a particular god or a series of gods. Degen (1974) established that bgn ʿl … is always negative, so that “curse against …” or “anathema against …” is an appropriate translation.3 Some of these bgn-“curses” seem to operate on a purely religious level, e.g. directed against anyone who fails to display the appropriate piety when confronted by a memorial inscription which requires the reader to utter a blessing on the name of the person commemorated. Thus no. 23: “Remembered be so-and-so … and the curse of Shaḥru on whoever reads this inscription and does not say ‘Let him be remembered for good’” (and see also 53, 74, 101, 153 and Healey 1996: 184–185). Curses of this kind cannot strictly be regarded as legal in character (though the demarcation between legal and non-legal injunctions is not a clear one). But there is also in no. 23 a specified punishment for anyone who damages the inscription itself: “And whoever erases it, let Baʿalshamīn (or B. will) remove his possessions and his offspring from before him”. This too is to be interpreted as a curse. There are other cases where it is clear that a specific regulation rather than a general pious injunction is involved. Thus in no. 281, mentioned earlier, bgn mrn wmrtn wbrmryn …, “the curse/invocation of Māran, Mārtan and Barmārēn …”, is pronounced against thieves stealing from the property of the Barmārēn temple and leads to stoning to death. There is a clearly defined and legally enforceable punishment attached to the offence. No. 29 is another temple regulation, the well-known one apparently indicating that shoes were not to be worn in a

3 The root bgn is a little obscure, but Degen cites further contextual evidence from Imperial Aramaic, in the form of the Akkadian borrowing of bagani from Aramaic, Mandaic bgan and Syriac bgan.

578

healey

particular temple. This one does not involve mwtʾ dy ʾlhʾ, but only the bgn-curse of six named gods (for no. 23, 281 and 29 see Healey 2009: no. 67, 76 and 80). If a curse were the punishment for an offence, the question would remain of how this could have been worked out in practical terms. However, this may be a modern rather than an ancient problem! Curses were taken seriously in antiquity and we know that curses sometimes appear in legalistic contexts of this kind, as is clear also from the Nabataean tomb inscriptions, where fines and curses appear side by side in legal contexts (Healey 1993: e.g. no. 1: 4 and 8: 5; idem 2013: 173–174): one could well imagine that the curse of the Nabataean god Dushara might be a more severe threat than that of a heavy fine. In Hatra, if this suggestion is correct, a curse in the form “Let him die by the death of the god” would be a fate worse than stoning, implying a life lived under a continuous threat of divine retribution and almost certainly a life of exclusion from the community.

Sigla cad ciip Inv.

Chicago Assyrian Dictionary (see Oppenheim and Reiner 1977). Corpus Inscriptionum Iudaeae/Palestinae (see Cotton et al. 2010). Inventaire des inscriptions de Palmyre (i–xii).

References Aggoula, B. 1991. Inventaire des inscriptions hatréennes. (Bibliothèque archéologique et historique 139). Paris: P. Geuthner. Aliquot, J. 2008a. Inscriptions grecques et latines de la Syrie, t. xi, Mont Hermon (Liban et Syrie). (Bibliothèque archéologique et historique 183). Beyrouth: Ifpo. Aliquot, J. 2008b. “Sanctuaries and Villages on Mt Hermon during the Roman Period.” In The Variety of Religious Life in the Near East in the Hellenistic and Roman Periods, edited by T. Kaizer, 73–96. (Religions in the Graeco-Roman World 164). LeidenBoston, ma: Brill. Attinger, P. 2014. “mūt ilī-šu mâtu.” nabu 2014 (1, mars): 8–9, §05. Beyer, K. 1998. Die aramäischen Inschriften aus Assur, Hatra und dem übrigen Ostmesopotamien. Göttingen: Vandenhoeck & Ruprecht. Bickerman, E.J. 1946–1947. “The Warning Inscriptions of Herod’s Temple.” Jewish Quaterly Review 37: 387–405. Clermont-Ganneau, C. 1872. “Une stèle du Temple de Jérusalem.” Revue Archéologique 23: 214–234, 290–296 and pl.

temple inscriptions and “the death of the god(s)”

579

Clinton, K. 2003. “Stages of Initiation in the Eleusinian and Samothracian Mysteries.” In Greek Mysteries: the Archaeology of Ancient Greek Secret Cults, edited by M.B. Cosmopoulos, 50–78. London-New York: Routledge. Cotton, H.M., L. Di Segni, W. Eck, B. Isaac, A. Kushnir-Stein, H. Misgav, J. Price, I. Roll, and A. Yardeni (ed.). 2010. Corpus Inscriptionum Iudaeae/Palestinae. Vol. i Jerusalem. Part 1: 1–704. Berlin-New York: De Gruyter. Degen, R. 1974. “Zur Bedeutung von bgn in den Hatra Inschriften.” In Neue Ephemeris für Semitische Epigraphik 2, edited by R. Degen, W.W. Müller, and W. Röllig, 99–104. Wiesbaden: Harrassowitz. Dirven, L. 2009. “My Lord with his Dogs. Continuity and Change in the Cult of Nergal in Parthian Mesopotamia.” In Edessa in hellenistisch-römischer Zeit: Religion, Kultur und Politik zwischen Ost und West. Beiträge des internationalen Edessa-Symposiums in Halle and der Saale, 14.–17. Juli 2005, edited by L. Greisiger, C. Rammelt, and J. Tubach, 47–68. (Beiruter Texte und Studien 116). Beirut: Ergon Verlag, OrientInstitut Beirut. Drijvers, H.J.W. 1965. The Book of the Laws of Countries. Dialogue on Fate of Bardaisan of Edessa. Assen: van Gorcum. Dunant, C. 1971. Le sanctuaire de Baalshamin à Palmyre. Vol. iii: Les inscriptions (Bibliotheca Helvetica Romana x, 3). Rome: Institut suisse de Rome. Fincke, J.C. 2013–2014. “mūt ilī-šu mâtu, ‘to die a death (decreed by) his god’.”nabu 2013 (14, décembre): 124–125, §75. Fitzmyer, J.A., and D.J. Harrington. 1994. A Manual of Palestinian Aramaic Texts (Biblica et Orientalia 34). Rome: Pontificio Istituto Biblico. Foster, B.R. 2005 (third edition). Before the Muses: An Anthology of Akkadian Literature. Bethesda, md: cdl Press. Fraser, P.M. 1960. Samothrace 2.1. The Inscriptions on Stone. (Bollingen Series 60, 2.1). New York: Pantheon Books. Frey, J.-B. 1952. Corpus Inscriptionum Iudaicarum. ii: Asie—Afrique. (Sussidi allo Studio delle Antichità Cristiane 3). Roma, Città del Vaticano: Pontificio Istituto de Archeologia Cristiana. George, A.R. 2003. The Babylonian Gilgamesh Epic: Introduction, Critical Edition and Cuneiform Texts. 2 vol. Oxford: Oxford University Press. Healey, J.F. 1993. The Nabataean Tomb Inscriptions of Madaʾin Salih. (Journal of Semitic Studies Supplement 1). Oxford: Oxford University Press. Healey, J.F. 1996. “‘May He be Remembered for Good’: An Aramaic Formula.” In Targumic and Cognate Studies. Essays in Honour of Martin McNamara, edited by K.J. Cathcart and M. Maher, 177–186. (Journal for the Study of the Old Testament Supplement 230). Sheffield: Sheffield Academic Press. Repr. in Healey 2011: chapter xx. Healey, J.F. 2005. “The Writing on the Wall: Law in Aramaic Epigraphy.” In Writing and Ancient Near Eastern Society. Papers in Honour of Alan R. Millard, edited by

580

healey

P. Bienkowski, C. Mee, and E. Slater, 127–141. (Library of Hebrew Bible/Old Testament Studies 426). New York-London: T. & T. Clark. Repr. in Healey 2011: chapter xix. Healey, J.F. 2009. Aramaic Inscriptions and Documents of the Roman Period (Textbook of Syrian Semitic Inscriptions 4). Oxford: Oxford University Press. Healey, J.F. 2011. Law and Religion Between Petra and Edessa: Studies in Aramaic Epigraphy on the Roman Frontier. (Variorum Collected Studies Series cs966). FarnhamBurlington, vt: Ashgate-Variorum. Healey, J.F. 2013. “Fines and Curses: Law and Religion among the Nabataeans and their Neighbours.” In Law and Religion in the Eastern Mediterranean: From Antiquity to Early Islam, edited by A.C. Hagedorn and R.G. Kratz, 165–186. Oxford: Oxford University Press. Iliffe, J.H. 1936. “The θanatos Inscription from Herod’s Temple.” Quarterly of the Department of Antiquities in Palestine 6: 1–3. Kaizer, T. 2006. “Capital Punishment at Hatra: Gods, Magistrates and Laws in the Roman-Parthian Period.” Iraq 68: 139–153. Naveh, J. 1978. On Stone and Mosaic: the Aramaic and Hebrew Inscriptions from Ancient Synagogues. Tel Aviv: Sifriyat Maʿariv. [in Hebrew] Oppenheim, A.L., and E. Reiner (ed.) 1977. The Assyrian Dictionary of the Oriental Institute of the University of Chicago. m/i–ii. Chicago, il: Oriental Institute; Glückstadt: J.J. Augustin. Poljakov, F.B. 1989. Die Inschriften von Tralleis und Nysa. Teil i: Die Inschriften von Tralleis (Inschriften griechischer Städte aus Kleinasien 36, 1). Bonn: Dr Rudolf Habelt GmbH. Robert, L. 1937. Études anatoliennes: recherches sur les inscriptions grecques de l’Asie Mineure. Paris: De Boccard. Schwartz, S. 2009. “Euergetism in Josephus and the Epigraphic Culture of First-Century Jerusalem.” In From Hellenism to Islam: Cultural and Linguistic Change in the Roman Near East, edited by H.M. Cotton, R.G. Hoyland, J.J. Price, and D.J. Wasserstein, 75– 92. Cambridge: Cambridge University Press. Seyrig, H. 1954. “Antiquités syriennes. 58: Inscriptions grecques.” Syria 31: 212–224. Steinmetz, D. 2008. Punishment and Freedom. The Rabbinic Construction of Criminal Law. Philadelphia: University of Pennsylvania Press. Thackeray, H.St.J. 1928. Josephus: the Jewish War Books v–vii (Loeb Classical Library). Cambridge, ma-London: Harvard University Press. Tubach, J. 1986. Im Schatten des Sonnengottes. Der Sonnenkult in Edessa, Ḥarrān und Ḥaṭrā am Vorabend der christlichen Mission. Wiesbaden: Harrassowitz. Vattioni, F. 1994. Hatra. (Annali dell’Istituto universitario orientale di Napoli, Supplement 81). Napoli: Istituto Universitario Orientale. Yon, J.-B. 2012. Inscriptions grecques et latines de la Syrie, t. xvii, fascicule 1, Palmyre. (Bibliothèque archéologique et historique 195). Beyrouth: Ifpo.

chapter 29

‘The Conception of Jesus’* Hannah M. Cotton Paltiel

In every society the legal status of children depends on the legal status of the parents, as well as on the legitimacy of the marriage. There is a fundamental difference between the status of a child born altogether outside the marriage bond and that of a child born in what one may call a ‘defective marriage’, i.e. when the two people who conceived the child were forbidden by the laws of the land or of their religion to enter into a legal marriage. The first would be a fatherless child and inherit the legal status of the mother (sometimes with grim consequences); the latter, however, should not have been conceived in the first place, and the stigma of its birth could become a legal liability for the rest of its life. An instance of a forbidden relationship is that between a man and a married woman, i.e. another man’s wife. At what point is a woman considered to be legally married to a man under Jewish law? Under biblical (as well as under rabbinic) law marriage is effected in two stages, betrothal and marriage proper. The English term betrothal renders loosely the legal institution designated ʾerusin in the Bible and qiddushin in rab-

* Dear Michael: you have heard my Conception of Jesus more than once, and on several continents, but having now been turned into an article and gone through the superlative editing of David Wasserstein and Ari Paltiel, it is yours exclusively and forever: donum auctor dedit! Over the years of conceiving the Conception, I have inccurred very many debts, only two of which can be acknowledged here: I am indebted to Sacha Stern for a thorough and wonderful editing and criticism of an early draft, and for drawing my attention to a plethora of issues I had not been aware of, and to Alanna Nobbs of Macquarie University who must have liked the Conception well enough to bring me over to Australia to hear it once again. Needless to say, this article only skims the surface of the formidable issues of betrothal, marriage and divorce in Jewish law of the Second Temple period and its aftermath, to say nothing of the story of the Nativity and Early Christianity. The bibliography, intimidating for the expert, let alone for the layman, is poorly represented here. I was always amazed and intrigued by my audience’s response. I suspect that as a Jew I failed to take for granted what a born Christian would have done.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

582

cotton paltiel

binic sources.1 Betrothal in biblical and rabbinic law is constitutive: it creates a new legal status. Deuteronomy 22: 23–29 assumes throughout that there is a difference between the violation of the betrothed and that of the unbetrothed virgin: When a virgin is betrothed (μεμνηστευμένη lxx) to a man and another man comes upon her in the town and lies with her, you shall bring both of them out to the gate of that town and stone them to death; the girl because, although in the town, she did not cry for help, and the man because he dishonoured another man’s wife (τὴν γυναῖκα τοῦ πλησίον lxx) … [on the other hand] … when a man comes upon a virgin who is not pledged in marriage (ἥτις οὐ μεμνήστευται i.e. not bethrothed) and forces her to lie with him, and they are discovered, then the man who lies with her will give the girl’s father fifty pieces of silver, and she shall be his wife because he has dishonoured her. The ʾarusa, i.e., the ‘betrothed’ woman (from the same root as ʾerusin, ‘betrothal’) normally continued to live in her parents’ home,2 but otherwise, as pointed out above, her legal status approximated that of a married woman. There is nothing in biblical law about the option of terminating a betrothal (as distict from marriage) with a deed of divorce, although this becomes commonplace in rabbinic literature.3 The second stage, the nissuʾin, or marriage, is associated with the taking of the bride by the groom to his home. Some, not unreasonably, cast doubt on the existence of the two-stage marriage during the Second Temple period.4 Michael Satlow, for example, singles 1 For identifying references to editions of Jewish sources one may consult Starck and Stemberger, 1991. The following description relies heavily on and borrows freely from the authoritative discussion in Friedman, 1980, i: 192–215. 2 The difference in customs between Judaea (where the betrothed couple could have intimate relations in the bride’s parents’ house) and the Galilee need not concern us here; cf. Mishnah, Ketubbot 1.5; Mishnah, Yebamot, 4.10, and see Noam, 1994 (Hebrew). Variety does not end there; one should also consider the evidence in secterian writings which cannot be taken into account here, e.g. the facinating aticle by Shemesh, 1998. 3 Katzoff, 1996, makes the interesting observation that Roman law makes adultery punishable also in the case of betrothal: impune in ea stuprum committeretur: Divi Severus et Antoninus rescripserunt etiam in sponsa hoc idem vindicandum, quia neque matrimonium qualecumque nec spem matrimonii violare permittitur, Digest 48.5.14 (13).3. 4 As distinct from the rabbinic period, although no precise dates can be given.

‘the conception of jesus’

583

out the story of Miriam’s betrothal to Joseph in the gospel of Matthew (which he names ‘a form of inchoate marriage’), as our sole evidence for the existence of a legal status for the act of betrothal before the first half of the second century ad, i.e. before the so-called rabbinic period, and dismisses the story in the New Testament as a sign of the provincialism of Jews in rural Galilee.5 It is quite true that rabbinic Judaism as commonly conceived emerged later than the Second Temple period. The legal documents from the Judaean Desert, the latest of which coincided with the outbreak of the Bar Kokhba Revolt in 132,6 still display a variety of legal customs and the incursion of different legal systems which do not always overlap with what came to be halakhic law (Cotton, 1998; Cotton and Yardeni, 1997: 133–157).7 Thus, the dismissal of the constitutive force of ‘betrothal’ as a legal stage in the creation of a Jewish marriage, i.e. that it was impossible for a betrothed woman to marry another man without having first secured a deed of divorce from her fiancé, need not be prima facie unreasonable, especially since there is no mention in the Bible of a deed of divorce in the case of dissolving a betrothal. It could be argued that in post-biblical times the constitutive nature of betrothal faded—if it was at all observed. If the passage in Matthew were the sole evidence for betrothal being constitutive (which as we shall see is not the case), scepticism could not be ruled out. It could even be argued that the passage in Matthew reflects Deuteronomy rather than contemporary usage. However; scepticism should not be taken too far: the New Testament, although not a book of history, may well reflect current legal practice—and precisely where it wishes to convey its new message. The story of the nativity is found only in the gospels of Matthew and Luke.8 There is no conception narrative in the other two gospels.9 However, the birth 5 Cf. Satlow, 2001: 69–73. He states categorically that post-biblical Judaism did not practise the two-stage marriage: ‘during the entire Second Temple period (most?) Jews neither customarily betrothed … nor did they even have a firm understanding of what such a betrothal would mean. Instead they followed Greek practices, and understood the biblical institution of betrothal within their own Hellenistic contexts’ p. 69. 6 Cf. P.Yadin 27 (19 August 132); unless ‘year 4 to the destruction of the House of Israel’ dates to 140 rather than to 74; see Eshel, Eshel and Yardeni, 2009 (Hebrew) = Eshel, Eshel and Yardeni, 2011 (English). 7 Scepticism concerning the authority of the Rabbis could go too far; see below n. 35 on P.Hever 13. 8 For the evaluation of the individual gospels, see Millar, 1990 (= Millar 2006). 9 Michael Macdonald pointed out to me the existence of slightly different genealogies in Matt. 13:55: οὐχ οὗτός ἐστιν ὁ τοῦ τέκτονος υἱός; οὐχ ἡ μήτηρ αὐτοῦ λέγεται Μαριὰμ καὶ οἱ

584

cotton paltiel

narratives in the two gospels are not just different; they are irreconcilable. Merely in the light of historical information from external sources the sequence of events in Matthew is possible, but that in Luke is impossible. Perhaps both are post eventum fictions. However, to my mind these reservations do not affect the argument I wish to put forward here. Both accounts contain two premises: 1) that Maria/Miriam had already been betrothed to Joseph (μνηστευθείση), but was not yet married to him, and 2) that she was a virgin (παρθένος) when Jesus was conceived. Matt 1: 18–2510 And the birth of Jesus was like this. His mother, Maria, was betrothed (μνηστευθείσης τῆς μητρὸς αὐτοῦ) to Joseph, and before they came together (πρὶν ἢ συνελθεῖν αὐτοὺς), she was found to be pregnant by the holy spirit (εὑρέθη ἐν γαστρὶ ἔχουσα ἐκ Πνεύματος Ἁγίου). Joseph, her husband, being a righteous/law-abiding man (δίκαιος ὤν), and at the same time not wishing to put her to shame in public, decided to divorce her quietly (λάθρᾳ). When he had resolved on this—lo and behold!—an angel of God appeared to him in a dream saying: ‘Joseph, son of David, be not afraid to take Maria, your wife (τὴν γυναῖκά σου). The child was begotten in her of the holy spirit (τὸ γὰρ ἐν αὐτῇ γεννηθὲν ἐκ Πνεύματός ἐστιν Ἁγίου). She will bear a son, and you will call him Jesus, for he will save his people from their sins.11 All this has happened in order to fulfill what was said by the Lord through his prophet: “Behold the virgin will conceive and bear a son (Ἰδοὺ ἡ παρθένος ἐν γαστρὶ ἕξει καὶ τέξεται υἱόν), and they will call him Emmanuel”, which means “God is with us” ’. After Joseph woke up he did as he was instructed by the angel; he took his wife home with him.

10 11

ἀδελφοὶ αὐτοῦ Ἰάκωβος καὶ Ἰωσὴφ καὶ Σίμων καὶ Ἰούδας (‘Is not this the carpenter’s son? Is not his mother called Mary? And his brethren, James, and Joseph, and Simon, and Judas?), and in Mark 6:3: οὐχ οὗτός ἐστιν ὁ τέκτων, ὁ υἱὸς τῆς Μαρίας καὶ ἀδελφὸς Ἰακώβου καὶ Ἰωσῆτος καὶ Ἰούδα καὶ Σίμωνος; καὶ οὐκ εἰσὶν αἱ ἀδελφαὶ αὐτοῦ ὧδε πρὸς ἡμᾶς; καὶ ἐσκανδαλίζοντο ἐν αὐτῷ ‘Is not this the (son of the) carpenter, the son of Mary, the brother of James, and Joseph, and of Juda, and Simon? and are not his sisters here with us?’ See commentary in Luz, 2002, and Basser, 2009. The impact of the divine intervention is greatly enhanced if one recalls Genesis 22:11– 13, where the angel stays Abraham’s hand when he is about to slaughter his own son: μὴ ἐπιβάλῃς τὴν χεῖρά σου ἐπὶ τὸ παιδάριον μηδὲ ποιήσῃς αὐτῷ μηδέν (And he said: ‘Lay not thy hand upon the lad, neither do thou any thing unto him’).

‘the conception of jesus’

585

Luke 1: 25–35: And in the 6th month the angel Gabriel was sent by God to a city of the Galilee, called Nazareth, to a virgin, betrothed to a man (πρὸς παρθένον μεμνηστευμένην ἀνδρί) called Joseph from the house of David. And the name of the virgin was Miriam. Coming to her he said: ‘Greetings, blessed one, the Lord is with you’. She was disturbed by his saying and asked what this greeting meant. And the angel told her: ‘Be not afraid, for you have found favour with God. (31) Behold, you will conceive and bear a son, and you will call him Jesus etc. …’ (34) And Miriam said to the angel: ‘How can (will) this be since I do not know a man/a husband?’ (Πῶς ἔσται τοῦτο, ἐπεὶ ἄνδρα οὐ γινώσκω). And the angel said to her in reply: ‘The holy spirit will come upon you and the power of the most high will cast his shadow over you. It is on this account that the child to be born will be called holy, son of God’. It is true that the fact of Miriam being a betrothed virgin who has not yet ‘known’ a husband is played down in Luke; it seems to be there in order to emphasise the miraculous nature of her pregnancy. Nevertheless here too we are told that she was pregnant during her betrothal—and before her marriage. Joseph goes to Bethlehem ‘in order to register together with Miriam who was betrothed to him, who was pregnant’ (Luke 2:5: ἀπογράψασθαι σὺν Μαριὰμ τῇ μεμνηστευμένῃ αὐτῷ, οὔσῃ ἐγκύῳ). We should overlook the improbable need for those living in Nazereth to register in Bethlehem. The parallel with Egyptian census-returns indicates that the order was to return to one’s normal place of residence.12 But as we know there are good reasons—into which we need not enter here—for sending them to Bethlehem; but why make her engaged and pregnant? Finally, in Matthew in contrast to Luke, there are elements which clearly imply that her betrothal to Joseph had obvious legal implications. The angel who dissuades the latter from divorcing her, treats them as man and wife: ‘Joseph, son of David, be not afraid to take (home?) Maria, your wife’ (μὴ φοβηθῇς παραλαβεῖν Μαριὰμ τὴν γυναῖκά σου, Matt 1:20), and this becomes a fact: ‘he took his wife home with him’ (καὶ παρέλαβε τὴν γυναῖκα αὐτοῦ, Matt 1:24). Once Miriam’s pregnancy was discovered, Joseph, who had not known her in the biblical sense of the word (πρὶν ἢ συνελθεῖν αὐτούς), and certain that the child was not his, decided to divorce her. He chose a course which, as we have

12

For the census see Cotton, 2003 and Cotton, 1999.

586

cotton paltiel

seen above, went against the biblical injunction specified in Deut. 22:24: ‘When a virgin is betrothed to a man and another man comes upon her in the town and lies with her, you shall bring both of them out to the gate of that town and stone them to death’. One may regard this as an indication that the biblical injunction was no longer in force, or no longer practised. At all events, he decided to divorce her. Nevertheless, being a righteous/law-abiding man (δίκαιος ὤν),13 and wishing to spare her honour (καὶ μὴ θέλων αὐτὴν παραδειγματίσαι),14 Joseph chose to divorce her away from the public eye (λάθρα ἀπολῦσαι αὐτήν), but divorce her nonetheless. This implies of course that they were already legally bound together (Matt. 1:19), although, as pointed out earlier, nowhere in the Bible is divorce mentioned as a means to dissolve a betrothal, and certainly not when the woman became pregnant by someone other than the man who had betrothed her. Until very recently (see below), one would have assumed that divorce as a means of breaking up an engagement was a rabbinic invention, if only for the lack of any evidence to the contrary.15 However, what the angel instructs him to do: Ιωσήφ, υἱὸς Δαβίδ, μὴ φοβηθῇς παραλαβεῖν Μαριὰμ τὴν γυναῖκά σου, was surely more compromising than a divorce, for ‘taking her home’ meant passing over to the second stage in Jewish marriage; from now on it would be assumed that Joseph was the father of the child, even though the author, who without doubt wishes to establish the paternity of God, informs us that Joseph had no marital relations with her until she gave birth (οὐκ ἐγίνωσκεν αὐτήν, ἕως οὗ ἔτεκεν υἱόν).16 By entering upon the second stage of marriage, Joseph will become the legal father of the son in her womb. He will acknowledge the child as his by giving

13 14

15

16

There are very many different interpretations of δίκαιος ὤν which we need not get into here. However, it is hard to see how he could have spared her since divorce was not a private matter which could be performed away from the public eye. But again there is no need to enter into this here. As pointed out by Zaas, 2009: 127: ‘This makes the Gospel of Matthew … a key in the history of Jewish law. There is no earlier literary text, biblical or otherwise, that specifies a divorce for a broken engagement, whether, this is an innovation in rabbinic law or not’. See also Zaas, 2005. These two papers will be published (with others) in his forthcoming book, ‘Mary’s Divorce. Jewish-Legal Aspects of the Matthean Birth Narrative’. I owe the last observation to John Curran. The full text in Mattew 1: 24–25 reads: ἐγερθεὶς δὲ ὁ Ἰωσὴφ ἀπὸ τοῦ ὕπνου ἐποίησεν ὡς προσέταξεν αὐτῷ ὁ ἄγγελος Κυρίου, καὶ παρέλαβε τὴν γυναῖκα αὐτοῦ· καὶ οὐκ ἐγίνωσκεν αὐτήν, ἕως οὗ ἔτεκεν υἱόν; ‘he did as the angel of the Lord commanded him; he took her as his wife, but he did not know her (i.e. had no marital relations with her) until she had borne a son …’.

‘the conception of jesus’

587

him a name—an indirect, or symbolic, act of adoption so to speak. Adoption, as we know, did not exist as a legal institution in Jewish law at that time.17 The concept of virginal conception, i.e. a non-biological conception with no human intervention, strains belief. Nevertheless, this offers no reason to give preference to the new ‘feminist theological interpretation of the infancy narratives’ for which ‘The New Testament Infancy Narratives incorporate the tradition of Jesus’ illegitimate conception’, i.e. that behind the story of the nativity lurks an intervention in the form of rape or adultery.18 There is in fact nothing new about this interpretation except the association with the new feminism: already in the Jewish sources and in Origen’s Contra Celsum we find allegations of adultery with a Roman soldier, named Panthera,19 and Jesus is called Ben Pantera/Pandera and Ben Stada in Talmudic sources. There are many versions of the Jewish story of the nativity, designated ‫ תולדות ישו‬or ‫מעשי ישו‬, which were collected by Samuel Krauss in Das leben Jesu nach jüdischen Quellen.20 All of them have allegations of rape, adultery and so forth. The ‘new feminists’, like the old Jewish detractors of Jesus, argue that both gospels—whether by playing down the pregnancy of a betrothed woman, as Luke does, or by confronting it directly as does Matthew—are in fact replying to charges of adultery and illegitimacy.21 17 18 19

20

21

See Levin, 2006, who takes the combination of θεοῦ υἱός and son of David to reflect the dynastic adoptions of the Julio-Claudians. E.g. the late J. Schaberg, 1987. Cf. Contra Celsum 1.32. However, his Homilia in Lucam 6 (pg xiii, 1314–1315) (= pl xxvi, 230–231) shows that he makes no distinction between betrothal and marriage, and is no longer aware of the betrothed being forbidden to her husband: Debuit de ea virgine nasci, quae non solum sponsum habet, sed ut Mattheus scribit, iam viro tradita fuerat, licet eam vir necdum noscat, ne turpitudinem virginis habitus ipse monstraret, si virgo videretur utero tumenti. The bibliography is immense. For a convenient recent introduction see Meerson, Schäfer and Deutsch, 2001. I am grateful to M. Meerson for allowing me to read his forthcomg paper ‘llegitimate Jesus: Family Matters with “Toledot Yeshu”’. I am not sure whether the definition of mamzer (‫ )ממזר‬as the offspring of a betrothed or legally married woman, who conceived from another man, had entered Jewish law by that time. Nothing in the biblical passages mentioning the exclusion of the mamzer from the community actually defines how that status was achieved. All they tell us is the dire legal consequences of being one. Thus the strength at that time of the charge against Jesus’ illegitimate conception cannot be meaningfully evaluated. See for example Chilton, 2001, and a response by Quarles, 2004. I have learnt much from an unpublished paper by Yonathan Sagiv submitted to Professor Shlomo Naeh at the Hebrew University of Jerusalem.

588

cotton paltiel

But why should we assume that the concept of virginal conception represents Christian ‘apologetics’ in response to those allegations, rather than accept, with no less plausibility, that these allegations took their cue from the narrative of virginal conception? For indeed if the subtext was an illegitimate conception, one wonders why the ‘incriminating’ details were not left out of the two gospels altogether. The best way to meet allegations of adultery and illegitimacy would have been to hide the fact that Maria was pregnant between betrothal and marriage, and to find another way to make the child born to married parents a child of God through a different form of divine intervention.22 There are certainly examples in the Old Testament and the apocrypha which endow the child of a mortal but barren woman with godlike provenance.23 Luke (chapter 1) does precisely this with John the Baptist and his mother Elizabeth. For some reason, it would seem, the two gospels chose not to omit the ‘incriminating’ details in the case of the conception of Jesus. Even Luke, who does very little with them, leaves them there notwithstanding, en passant. I suspect that what was left was retained, or at least was not removed, for a reason. It makes much better sense to assume that the story, as we have it, with the so-called incriminating details, seemed the best way to make virginal conception plausible, or, even better, possible. A precondition—albeit insufficient without the miracle—for virginal conception was the fact that the couple, Joseph and Mary, were betrothed but not yet married, and thus forbidden to each other. Admittedly, this precondition would also constitute the ground for a claim of illegitimacy.24 Were she properly married, her pregnancy would arouse no suspicion; no allegation of illegitimacy could be made, and conversely, no

22

23

24

I cannot attempt here to put the Conception of Jesus in the context of miraculous conceptions and births in the Jewish, Christian and Pagan tradition. See for example the intriguing article by Böttrich, 2001. Although after Genesis 6:1–4, all intercourse between the ‘sons of God and the daughters of men’ seems to have come to an end: ‘And it came to pass, when men began to multiply on the face of the earth, and daughters were born unto them, that the sons of God saw the daughters of men that they were fair; and they took them wives of all which they chose. And the Lord said, my spirit shall not always strive with man, for that he also is flesh: yet his days shall be an hundred and twenty years. There were giants in the earth in those days; and also after that, when the sons of God came in unto the daughters of men, and they bare children to them, the same became mighty men which were of old, men of renown’. For the difference between Judaea and the Galilee see above n. 2.

‘the conception of jesus’

589

virginal conception would be believed. Were she not betrothed, her subsequent marriage to Joseph would imply that he is the biological father—since this is the remedy offered in Deut. 22:29 in such cases—and this too would be no good for sustaining the claim of virginal conception: ‘When a man comes upon a virgin who is not pledged in marriage (ἥτις οὐ μεμνήστευται lxx) and forces her to lie with him, and they are discovered, then the man who lies with her will give the girl’s father fifty pieces of silver, and she shall be his wife because he has dishonoured her’. Thus, only in the case of a betrothed virgin, who was expected to remain a virgin until her marriage, virginal conception or illegitimacy are possible alternatives. For the virginal conception the concept of betrothal meant that the woman was for the period between betrothal and marriage in a chastity vacuum; and for the illegitimacy charge—to maintain the argument of equal plausibility—it meant that she ought to have remained in a chastity vacuum. Of course we must not forget that Joseph, the fiancé/husband, came from the house of David, for only in this way could the combination of virginal conception be combined with descent from the house of David: ‘And Jacob bore Joseph, Maria’s husband, of whom Jesus called Christus was born’ (Matt. 1:16). Perhaps this and no other was the raison d’être for the entire construct: a son of God whose legal father is a descendant from the house of David, for only he could be the Messiah. Because of the crucial importance of the betrothal element, I believe that the evidence of the stories of the gospel for the existence at that time of the twostage marriage, i.e. for the constitutive force of betrothal, is far stronger than one would have otherwise thought. The existence of the the two-stage marriage should not be dismissed as restricted to some rustic Jews and to parochial Galilee (to use the terms employed by those who reject its existence). This was the state of the evidence until 2001. Since then we have new documentary evidence for the legal status of betrothal in Jewish law from secondcentury bc Egypt: the two-stage marriage is now confirmed in P.Politeuma 4 (Cowey and Maresch, 2001).25 The publication by James Cowey and Klaus Maresch in 2001 of twenty papyri from the politeuma of the Jews in Heracleopolis in Middle Egypt written between 144/3 and 133/2bc marks a turning point in the study of the history of Jewish settlement in Egypt: it constitutes our first unambiguous evidence for the use of Jewish law by Jews in Egypt. The absence of such evidence led

25

Cf. Kruse, 2010.

590

cotton paltiel

the editors of the Corpus Papyrorum Iudaicarum to conclude that ‘the laws and regulations forming the legal basis for the business life of the Jews are the common laws of the Greeks in Egypt … the family life of Alexandrian Jews, their marriages and divorces, were regulated by Greek contracts in accordance with the principles of Hellenistic law’.26 The new archive demonstrates clearly that they were wrong; Jewish law was used by Jews as their νόμος πολιτικός— certainly in cases which involved family and personal law as here.27 P.Politeuma 4 vindicates the claim that the two-stage marriage was practiced long after the Biblical period: betrothal created a new legal status and its dissolution called for a deed of divorce. P.Polit.Iud. 428 is a petition by a Jew named Philotas son of Philotas to the archons of the Jewish Politeuma of Herakleopolis from 134bc, recounting in some detail his betrothal to a woman called Nikaia daughter of Lysimachos29 and his subsequent betrayal: he was promised the bride, and a sworn agreement was made according to the law, but: ‘Not long afterwards, Lysimachos, without justification joined Nikaia to another man before having received from me the customary bill of divorce’: μετ̓ οὐ π̣ [ολὺν χ]ρ̣ο̣ν́ ον ὁ Λυcι ̣μ̣ ά̣χ̣ο̣υ̣ cυνήρμοκεν ἄν̣ευ̣ ̣ λόγου ἑτ̣ ε̣ ρ̣́ ωι ἀ̣ν̣δ̣ρὶ τὴν Νε̣ίκαιαν̣ πρὶν ἢ λαβεῖν παρ’ ἐμοῦ τὸ εἰθιcμένον τοῦ ἀποcταcίου τὸ βυβλίον. The phrase τὸ βιβλίον τοῦ ἀποστασίου is the technical term (so to speak) used in the Bible for a deed of divorce given by a husband to his wife.30 Admittedly, the Bible as pointed out above, does not mention a deed of divorce for the dissolution of betrothal.31 Be this as it may, this papyrus makes it clear that the betrothal created a legal bond which could only be dissolved by a deed of divorce, given by the fiancé.32 On the basis of this text, we should now accept the reconstruction of the fragmentary first lines of P.Enteux. 23 (= cpj i 128, ll. 2–3), [κατὰ τὸν νόμον 26 27 28 29

30 31 32

cpj i, 34, but see the whole discussion 33–47 (the Ptolemaic period) as well as the commentary on P.Ent. 23 = cpj i, 128 (218 bc). As already claimed by Mélèse-Modrzejewski, 1996: 313. See text, translation and a photo in Appendix 1 at the end. I once thought that this might have been a case of intermarriage, that Nikaia and her father were not Jews and thus unaware of the need to obtain a deed of repudiation from the man to whom she was engaged before getting involved with another one. But even if this were the case, there is no doubt at all that the betrayed groom did not make such an allowance for the gentiles. Philotas takes it for granted that he should have given her a deed of divorce before she could become free to marry the next man. Cf. Deut. 24:1 and 24:3: καὶ γράψει αὐτῇ βιβλίον ἀποστασίου καὶ δόσει εἰς τὰς χεῖρας αὐτῆς; cf. Jeremiah 3:8 βιβλίον ἀποστασίου and Isaiah 50:1 τὸ βιβλίον τοῦ ἀποστασίου. See pp. 2 and 3 above. See below, n. 35.

‘the conception of jesus’

591

π]ο̣λ̣ιτικὸν τῶν [Ἰου]δαίων ἔχειν με γυν[αῖκα], which shows that Jewish law was paramount in marriage contracts in Hellenistic Egypt (Cowey and Maresch, 2001: 60, n. 22). Another justification for the emphasis placed on the two-stage marriage, which does not contradict the one pursued so far, but rather complements it, may be suggested. It is possible that the first impulse for creating the entire construct came from the lxx text of Isaiah 7.14. The verse is cited by Matthew in v.23: Ἰδού, ἡ παρθένος ἐν γαστρὶ ἕξει, καὶ τέξεται υἱόν, καὶ καλέσουσι33 τὸ ὄνομα αὐτοῦ᾿Εμμανουήλ (‘Behold: the virgin will conceive in her womb and give birth to a son, and they will call him Immanuel’). The Greek παρθένος translates here the Hebrew term alma (‫)עלמה‬, an ambiguous term which can mean νεᾶνις as well as παρθένος, that is to say, a young woman as well as a virgin. However, the Septuagint παρθένος can only mean the latter. Having used the Greek text, Matthew understood that for the prophecy to be accurate Jesus/Emmanuel must be born of a virgin. The two-stage marriage, which required the preservation of the virginity of the mother in the time gap between betrothal and marriage (for as has been pointed out above, the betrothed woman was forbidden to her fiancé as well as to everyone else) suited the prerequisite for a virgin-mother for the messiah to perfection—in fact, it was indispensable for maintaining the divine conception in a Jewish context. Matthew did not need to invent a complex narrative device; in the two-stage marriage he had at hand one which was part of everyday life.

Conclusion There is no room here to look at the question of the identity of the authors and intended readers of the Gospels, nor of any earlier sources that the Gospels were drawing on. We cannot be certain of the cultural milieu and the traditions that the narratives in the Gospels used and reflect: the Galilee, the Jewish Diaspora, or particular pagan audiences in the Roman empire? These questions have been discussed endlessly without reaching definite conclusions. However, it is no longer the case that we have no evidence for the two-stage marriage outside the conception story in the New Testament, and that the story is strictly localized in cultural terms to provincial Galilee.34 On the contrary, the structure of the story, as related in the Gospels, appears to reflect a very particular context 33 34

Matthew uses the plural although the Septuagint like the Hebrew text ‫הנה העלמה הרה‬ ‫—שמו עמנואל—ויולדת בן וקראת‬uses the 2nd person singular. One can apply here Fergus Millar’s double-negative form of argument (Millar, 2002: 72): it

592

cotton paltiel

in which the nature and portentous consequences of betrothal were clearly understood, and heighten the narrative. Unlike the marriage contract which is merely a financial transaction with no constitutive power, acts of betrothal—and divorce35—are decidedly crucial for the legal status of the children born to a couple. And it stands to reason that the society concerned would pay more attention to these than to financial transactions, which are the essence of marriage contracts. The lack of uniformity in marriage contracts is hardly disturbing, and there is no need to follow some scholars in proselytising every single marriage contract written in Greek.36 But the constitutive force of betrothal in post-biblical Jewish society is vindicated by the evidence: by P.Politeuma 4, by Tosefta Ket. 4:937—and by the stories of the nativity. As for the latter: scholars can express no opinion about virginal conception. In this essay I have tried to explain the historical ingredients which contribute to the making of the narrative of the conception, the myth, the miracle or whatever one chooses to call it. A miracle is the interference of the supernatural with mundane reality. But how can one appreciate the miraculous without knowing the ordinary, the normal, the routine? I have therefore attempted to reveal the hard core of legal reality behind the New Testament conception narrative, which to my mind makes it all the more powerful on a symbolic plane. For those who believe that God could intervene in human affairs, this explanation takes nothing away; on the contrary, it makes the miracle all the more clear and striking. It supplies the nuances which would have been perfectly understood by contemporaries.

35

36

37

is not the case that we do not have a single piece of evidence from real life, i.e. documentary one, for the existence of the two-stage marriage in post-biblical Jewish society. The deed of divorce was always to be given by the fiancé/husband, as the editors of P.Politeuma 4 (p. 57) quite rightly point out. P.Hever 13 can no longer be used against the clear evidence of Mur 19, and in strict contradiction to rabbinic law, as though referring to a deed of divorce given by a wife to her husband, as once claimed by Cotton and Qimron, 1998. This claim has been convincingly refuted once and for all in an unpublished paper by Hillel Newman (given on 15 May 2014 in Haifa University) who has kindly allowed me to refer to his ms. We must return to Yardeni’s reading of the document as referring to a deed of divorce given, as expected, by the husband (djd 27, no. 13). Despite Mordzejewski, 1961, I agree with Yaron, 1964 that CPJud. 144 of the year 13bc is not a Jewish doument, but a mutual agreement between the couple to disslove their marriage. See also Jackson, 2010. See Wasserstein, 1989, who dismisses the interpretation of Ranon Katzoff of the papyrus in ‘Part ii: Legal Commentary’ in Lewis, Katzoff and Greenfield, 1987: 236–247, and response in Katzoff, 1991: 173–174. See now Cotton, 2013. On which see Kister, 2002.

593

‘the conception of jesus’

Appendix: P.Politeuma 4: Greek Text, Photo and Translation

figure 1 P.Politeuma 4 (recto) photo credit: © institut für papyrologie, ruprecht-karlsuniversität heidelberg

594

cotton paltiel

figure 2 P.Politeuma 4 (verso) photo credit: © institut für papyrologie, ruprecht-karlsuniversität heidelberg

595

‘the conception of jesus’

1

(ἔτουc) λϛ Χο(ιὰκ) ιθ̅ π̣ ερ̣ ι̣ ̣̀ γ̣άμ̣ ο̣υ̣ cυ(νετάξαμεν)παρα(γγεῖλαι) 2 τοῖc ἄρχοῦ[cι] 3 π̣ α̣ρα̣ ̣̀ Φ̣ ι ̣λ̣ ώ̣ το̣ ̣υ̣ τ̣ο̣υ̣̃ Φ̣ ι ̣λώτου 4 τῶν ἐκ τοῦ πολιτεύματοc. 5 ἐν τῶι ἐνεcτῶτι ἔ[τ]ει ἐμνηc6 τευcάμην Ν̣ ε̣ι ̣κ̣ α̣[ι]α̣ν̣ Λ̣ υ̣cι̣ ̣μ̣ ά̣7 χου καὶ τοῦ cη̣ μ̣ α̣[ι]ν̣ο̣μέν̣ο̣μενου̣ 8 αὐτῆc πατρὸc ὀμ[ό]σαντοc 9 δώc̣ει̣ ̣ν̣ ἐμοὶ αὐτ[ὴ]ν καὶ τὴν 10 cταθε̣ιc̣̃ α̣ ν ἐπὶ α̣[ὐ]τῆι ̣ φ̣ερ̣ ν̣ ̣ὴν, 11 ἐφ’ ἧι καμου̣̃ εὐδοκ̣ οῦντοc 12 οὕτωc̣ οὐ μ̣ ο̣ν́ ̣ο̣[ν] ὀριcμῶν 13 γενομέν̣ω̣ν̣ κα̣[τ]ὰ̣ κοινὸν 14 ἀλλὰ καὶ τῆc̣ κα̣τὰ̣ τὸν νό15 μον αποκ .. [.]. c̣ γ̣ενηεἰc εδ̣η̣λ̣ ου̣[..] … 16 θείcηc κα̣ι ̣̀ ἐπ̣ ι ̣̀ [τ]ο̣ύτοιc 17 ἀπαλλαγέντ̣ω̣ν̣ ἡμῶν. 18 μετ’ οὐ π̣ [ολὺν χ]ρ̣ον̣́ ον 19 ὁ Λυcι̣μ́ ̣ χ̣ο̣υ̣ cυνήρμοκεν 20 ἄν̣ευ̣ ̣ λόγου ἑτ̣ ε̣ ρ̣́ ωι ἀ̣ν̣δ̣ρὶ 21 τὴν Νε̣ίκαιαν̣ πρὶν ἢ λα22 βεῖν παρ’ ἐμοῦ τ̣ὸ̣ εἰθ̣ιcμέ23 νον τοῦ ἀποc̣τα̣ cίου 24 ⟦τὸ⟧ β̣υβλ̣ ίο̣ν. διὸ ἀξιῶ, 25 ἐὰν φάνῃται, cυντάξαι 26 γρ̣άψ̣αι τοῖc ἐν τῆι ̣ κώμηι ̣ 27 Ἰ ουδαίοιc π̣ αραγγεῖλαι τῶι 28 Λυcιμάχωι ἀπαντᾶν 29 ἐφ’ ὑμᾶc ἵν’ ἐὰ̣ ν ᾖι ̣.α̣ι ̣⟧η̣ ι οἷα 30 [γ]ρ̣α̣φ ́ ̣ ω̣ δ̣ι ̣α̣λ̣η̣(φθῆι) περὶ αὐ(τοῦ) κα(τὰ) τὸν νό(μον) ἐμ̣ ο̣ὶ δ’ ἐπα̣ναγ31 [κάcαι ca. 4] . [ca. 2] . [ca. 2] … ψ̣ …… χ̣ –––––––––

(1–4) Year 36. 19 Choiak. Regarding a marriage. We have given an order to issue a summons. To the archons from Philotas son of Philotas, (4) a member of the Politeuma. (5–17) In the current year I betrothed Nikaia daughter of Lysimachos. Her said father swore to give her to me along with the dowry laid down for her, and with which I was in agreement. So after not only vows(?) were exchanged between us but also the … according to the law (or the Law) … We parted on those terms. (18–24) Not long afterwards, Lysimachos without justification joined Nikaia to another man before having received from me the customary bill of divorce. (24–31) Therefore, I request, if you think it right, that you give the order to write to the Jews in the village to summon Lysimachos to appear before you, so that if the matter is as I write, his case may be decided according to the law (or the Law), and at the same time [he] may be forced … to me …

596

cotton paltiel

Bibliography Basser, H.W. 2009. The Mind behind the Gospels: Commentary to Matthew 1–14. Boston: Academic Studies Press. Böttrich, Ch. 2001: “Die vergessene Geburtsgeschichte: Mt 1–2 / Lk 1–2 und die wunderbare Geburt des Malchisedek.” In Jüdische Schriften in ihrem antik-jüdischen und unchristlichen Kontext, Studien zu den Judischen Schriften aus hellenistischenromischen Zeit i, eds. H. Lichtenberger and G.S. Oegema, 222–249. Gütersloh: Gütersloher Verlagshaus. Chilton, Bruce. 2001. “Jésus, le mamzer (Mt 1.18).” New Testament Studies 46: 222–227. Cotton, H.M., and A. Yardeni. 1997. Aramaic, Hebrew and Greek Texts from Nahal Hever and Other Sites. Oxford: Clarendon Press. Cotton, H.M. 1998. “The Rabbis and the Documents.” in The Jews in a Graeco-Roman World, ed. M. Goodman, 167–179. Oxford: Oxford University Press. Cotton, H.M. and E. Qimron, 1998. “XHev/Se ar 13 of 134 or 135: a wife’s renunciation of claims.” Journal of Jewish Studies 49: 108–118. Cotton, H.M. 1999. “Some Aspects of the Roman Administration of Judaea/Syria-Palaestina.” In Lokale Autonomie und römische Ordnungsmacht in den kaiserzeitlichen Provinzen vom 1.–3. Jh., Kolloquien des Historischen Kollegs, ed. W. Eck, 75–91. Munich: De Gruyter Oldenbourg. Cotton, H.M. 2003. “The Roman Census in the Papyri from the Judaean Desert and the Egyptian κατ’ οἰκίαν ἀπογραφή.” in A Climate of Creativity. Papers from a New York University conference marking the retirement of Baruch A. Levine, ed. Lawrence H. Schiffman, 105–122. Leiden-Boston: Brill. Cotton, H.M. 2013. “Change in Continuity in Late Legal Papyri from Palaestina Tertia: Nomos Hellênikos and Ethos Rômaikon.” In Jews, Christians, and the Roman Empire: The Poetics of Power in Late Antiquity, eds. Annette Y. Reed and Natalie Dohrmann, 209–221. Philadelphia: University of Pennsylvania Press. Cowey, J.M.S. and K. Maresch. 2001. Urkunden des Politeuma der Juden von Herakleopolis (144/3–133/2 v. Chr.) (P. Polit. Iud.), Papyrologia Coloniensia 29, 56–71, no. 4. Wiesbaden: Westdeutscher Verlag. Eshel, E., H. Eshel and A. Yardeni. 2009. “A Document from ‘Year Four of the Destruction of the House of Israel’ in which a widow declared that she Received all her rights.” Cathedra 132: 5–24. (Hebrew) (English version in Dead Ses Discoveries 18, 2011, 1–28). Fassberg, S. 2012. “Did Final i > e in the Language of Nahal Seʾelim 13?” Leshonenu, A Journal for the Study of the Hebrew Language and Cognate Subjects 74: 95–107. Friedman, Mordechai. 1980: Jewish marriage in Palestine: a Cairo Geniza study. Tel Aviv and New York: The Jewish Theological Seminary of America. Jackson, B.S. 2010: “Marriage and Divorce: From Social Institution to Halakhic Norms.” in The Dead Sea Scrolls. Texts and Context, ed. C. Hempel, 339–364. Leiden: Brill.

‘the conception of jesus’

597

Katzoff, R. 1991. “Papyrus Yadin 18 again: a rejoinder.” The Jewish Quarterly Review 82: 171–176. Katzoff, R. 1996. “Philo and Hillel on Violation of Betrothal in Alexandria,” The Jews in the Hellenistic-Roman Period World: Studies in Memory of Menachem Stern, eds. I.M. Gafni, A. Oppenheimer, D. Schwartz, Jerusalem, 1996, 39*–57*. Kister, M. 2002: “From Philotas to Hillel: ‘Betrothal’ Contracts and their Violation.” Scripta Classica Israelica 21: 57–60. Kruse, Th. 2010. “Das jüdische politeuma von Herakleopolis und die Integration fremder Ethnien im Ptolemäerreich.” In Volk und Demokratie im Altertum (Bremer Beiträge zur Altertumswissenschaft Band 1), eds. V.V. Dement’eva and T. Schmitt, 93–105. Göttingen: Edition Ruprecht. Levin, Yigal. 2006. “Jesus, ‘Son of God’ and ‘Son of David’: The ‘Adoption’ of Jesus into the Davidic Line.” Journal for the Study of The New Testament 28: 415–442. Lewis, N., R. Katzoff and J. Greenfield, 1987. “Papyrus Yadin 18.”Israel Exploration Journal 37: 229–250. Luz, Ulrich, Das Evangelium nach Matthäus ekk i/1: Mt 1–7, 2ed, 2002. Meerson, M., P. Schafer, and Y. Deutsch. 2011. Toledot Yeshu (“The Life Story of Jesus”) Revisited. Tübingen: Mohr Siebeck; cf. www.princeton.edu/judaic/special-projects/ toledot-yeshu Mélèse-Modrzejewski, Joseph. 1996. “Jewish Law and Hellenistic Legal Practice in the Light of Greek Papyri from Egypt.” In An Introduction to the History and Sources of Jewish Law, eds. N.S. Hecht et al., 75–99. Oxford: Clarendon Press. Millar, Fergus. 1983. “Epigraphy.” In Sources for Ancient History, ed. M. Crawford, 80–136. Cambridge 1983 (= Rome, the Greek World, and the East, Vol. i: The Roman Republic and the Augustan Revolution, eds. Hannah M. Cotton and Guy M. Rogers, 39–81. Chapel Hill and London: University of North Carolina Press). Millar, Fergus. 1990. “Reflections on the Trial of Jesus.” In A Tribute to Geza Vermes: Essays on Jewish and Christian Literature and History, eds. P.R. Davies and R.T. White, 355–381. Sheffield: Journal for the Study of the Old Testament (= Rome, the Greek World, and the East, Vol. iii: eds. Hannah M. Cotton and Guy M. Rogers, 139–163. Chapel Hill and London: University of North Carolina Press). Mordzejewski, J. 1961. “Les juifs et le droit hellénistique: Divorce et égalité des époux (CPJud. 144).” Iura 12: 162–193. Noam, Vered. 1994: “The Seventeenth of Elul in Megillat Taʾanit.” Zion 59: 433–444 (Hebrew). Quarles, Ch. 2004. “Review Essay, ‘Jesus as Mamzer’”, Bulletin for Biblical Research 14: 243–255. Satlow, Michael. 2001. Jewish Marriage in Antiquity. Princeton: Princeton University Press. Schaberg, Jane. 1987. The Illegitimacy of Jesus. A Feminist Theological Interpretation of the Infancy Narratives. San Francisco: Harper & Row.

598

cotton paltiel

Shemesh, Aharon. 1998, “4q271.3: A Key to Sectarian Matrimonial Law’” Journal of Jewish Studies 49: 244–263. Starck, H.L., and G. Stemberger. 1991: Introduction to the Talmud and Midrash. Edinburgh: t & t Clark. Wasserstein, A. 1989. “A Marriage Contract from the Province of Arabia Nova: Notes on Papyrus Yadin 18.” The Jewish Quarterly Review 80: 93–130. Yaron, R. 1967. “CPJud. 144.” Iura 18: 21–24. Zaas, Peter. 2005. “Spiritus Ex Machina: Jewish Legal Aspects of the Matthean Birth Narrative.” Jewish Law Association Studies 17: 295–302 Zaas, Peter. 2009. “Matthew’s Birth Story: An Early Milestone in the History of Jewish Marriage Law.” Biblical Theological Bulletin 39: 125–128.

part 3 Modern Dialects and Tribes

∵

chapter 30

Drink Long and Drink in Peace: Singing to Livestock at Water in Dhofar, Sultanate of Oman Miranda J. Morris and Sālim ʕAwaḏ̣ Ahmad al-Shaḥri

I first started making recordings of the Modern South Arabian Languages (msal) when I was working in Dhofar in the 1970s, initially in my spare time and with a cheap tape recorder, and later, in the 1980s, sponsored by the Diwan of Royal Court, with a vast heavy Uher 4000. In the 1970s I recorded various songs sung to livestock. Re-visiting these when digitising them for the Leverhulme project I am currently working on,1 Salim ʕAwaḏ̣ and I had the chance to do some work on some of them when he was living with us while taking an English course in St. Andrews. I did further work on them with others2 during a fieldtrip to Dhofar earlier this year. I soon realised that I had been fortunate indeed to record such songs when I did. The hard work associated with livestock husbandry—herding to pasture and water; constructing and mucking out byres and pens; catching, drying and transporting sardines as dry season feed; searching for fodder in times of drought—is now mostly carried out by immigrant labour, mainly from the Indian sub-continent. Although many livestock owners still resist ‘foreigners’ milking their livestock, even this last barrier is crumbling. Today too, trucks have replaced camels and the human back for transport, local factories provide all sorts of supplementary feed, and bales of (largely imported) hay are trucked along a network of tarmacked roads to various distribution points. And as for watering livestock, the Government of Oman has sunk wells and supplied storage tanks and drinking troughs throughout the region, and this has made the long treks to and from water also a thing of the past. The songs I recorded reflect a time when the traditional economy of the interior of Dhofar was still based on livestock. The animals, principally cat-

1 [Documentation and Ethnolinguistic Analysis of Modern South Arabian, website l http://www .leeds.ac.uk/arts/info/125219/modern_south_arabian_languages.] 2 We would like to thank here especially Muḥammad Saʕīd Baḳwɛ̄r al-Mašaykhi and Ṭufūl Farag Bait Saʕīd for their help with the cow songs of the central region. Sālim ʕAwaḏ̣ is from the dry eastern area around Sudḥ where cows are few.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

602

morris and al-shaḥri

tle and goats, were managed to maximise milk production, and and in order to conserve milk supplies, the majority of male young were slaughtered soon after birth. The milk was processed to make the highly-valued butter-oil which herders then traded for the necessities they were unable to produce themselves. The precious milk-animals were bred to give birth in the cooler autumn months when the pastures were at their richest after the summer monsoon, and it was at this time that most of the butter-oil was made. This was also a time of relative ease for the herding communities, a time for weddings and other social activities as families enjoyed this seasonal abundance of milk and meat. But the task of finding water and walking livestock to it was an unending and exhausting one. Although there are a series of permanent springs along the foothills of the Dhofar mountains, much of this water had to be shared with the settled agricultural communities of the plains, and elsewhere water sources were few and far between. They were often difficult to access, and could entail herding livestock down precipitous rocky trails to boulder-strewn gorges. In many areas it was common to spend most of a morning walking livestock to water, the rest of the day watering them (and having to take their turn with others watering at the same source), and the next day walking them back home, only to repeat the whole process in a day’s time. Both men and women undertook this task, though it was usually women who fetched the water for household consumption. There were many different types of water source. The terms for the most important ones are given below.3 1. Types of water source.

Śḥerɛ̄t4

English

Notes

ʕayhn (pl. ʕāntə)

spring

of any size; some are very small and some only seasonal

mih ʕod

permanent water

3 Abbreviations: sing., singular; pl., plural; dim. diminutive; def., definite; √, root; l., line (of a song); masc., masculine; fem., feminine; poet., in poetry; sp., in everyday speech; Ar., Arabic; trans., transitive; lit., literally. 4 The Śḥerɛ̄t in these two tables is that spoken in the eastern area.

603

drink long and drink in peace

Śḥerɛ̄t

English

Notes

mih merbo̍ t; mih maġtero̍ ḳ

permanent water

ġɔr (pl. ġabri̍n)

well

xohr (pl. xɛ̄ri̍tə)

coastal lagoon

ḳɔt (pl. ḳiźe̍t) (√/ḳlt/) fo̍ ḳɛ (pl. fi̍ḳi)

rock pool pool in conglomerate or in soft ground water scrape deep pool shallow pool flowing water shallow pool small but deep hole in rock that fills after rains

mih merbo̍ t: a broader stretch of water than mih ʕod above; mih maġtero̍ ḳ: a much smaller pool but a deep one (maġterto̍ ḳ, ‘that one could drown in’) uncommon in Dhofar outside the coastal settlements mostly small; the water increasingly brackish water as it nears the sea of any size sometimes has a small spring; often in a dry / seasonal flow bed

mḥe̍si (pl. mḥa̍bsi) žiyi̍t (pl. žɔ̍ btə) śɛ̣ t̄ (pl. śɔ̣ ̄ btə) jere̍t (pl. jirɛ̍r) śterɛ̍r (pl. śtero̍ r) ʕanṯi̍ḳ (pl. ʕunu̍ ṯuḳ)

ʕo̍ nə

cistern in the high plateaus

do̍ jeb (pl. dejo̍ b)

a steady drip or seepage of water from a rock

usually in a dry / seasonal flow bed a deep pool, larger than fo̍ ḳɛ above temporary, in the mountains anything from a slow trickle to a flowing stream can be only a puddle after rain the water is often syphoned out through a tube made by joining sections of Jatropha dhofarica [zəbrɔ̍ t] twigs from which the inner core has been removed basically a mist-trap in the monsoon: a ring of walling is constructed around or to one side of an isolated broad-crowned tree (as ṭayḳ, Ficus vasta); the interior is lined with mud or clay. Moisture from the mist and drizzle of the monsoon is trapped by the foliage and drips down into the cistern commonly inside a cave or at the back of a cliff ledge. Often a hollow is dug out beneath to collect the water, making a pool or fo̍ ḳɛ; beside it a larger area is hollowed out to make a mikiri̍f (see below)

604

morris and al-shaḥri

(cont.)

Śḥerɛ̄t

English

Notes

mikiri̍f (pl. mokoro̍ f )

(i) a basin dug out for livestock to drink from; (ii) walling constructed to trap water in a gully

muśɛ̣ ʕ (pl. mośọ ʕ)

the shallow, greening, often scummy pools that remain as floodwaters dry up bad, polluted water

the basin is dug beside a mḥe̍si or fo̍ ḳɛ, the rim built up with earth and stones, the interior lined with mud or clay or with a tanned cow-hide (rɛ̍kɛ), and the water is poured into it (sp. bik eku̍ rfk her fo̍ ḳɛ, ‘I’ve made a drinking trough for the pool’) herders try to keep their goats away from these, fearing that if the animals drink from it they will not drink enough of the clean water they are being herded to

mih iź xɔ̍ be̍

regarded as undrinkable except in emergencies

Other than springs, water for people and livestock depended on rainfall. Although herding families who lived within the monsoon-affected zone enjoyed summer months of more plentiful water, those outside this region had to struggle the year round to find adequate water for themselves and their animals. Most rain fell in the winter, during the ‘star seasons’5 of aġi̍l and lḥēmar (pl. elḥameru̍ n), and principally at night. The summer rains were not so reliable, but when they came they fell during the ‘star season’ of elkēḏeb in the early summer, a period called do̍ ṯɛ. Rain in summer tended to fall during the day, rarely at night. Some of the principal terms for rain are given below. 2. Rain

Śḥerɛ̄t

English

Notes

muse̍ (def. õse, pl. mla̍bsi) mɛ̍lsɛ̍ śiryũ̍ t (pl. śiru̍ m)

rain light rain downpour

in general dim. of muse̍ above heavy rain of short duration

5 As elsewhere in Southern Arabia, the year in Dhofar is divided into periods dominated by a specific star (see various publications of R.B. Serjeant and D. Varisco, for example).

605

drink long and drink in peace

Śḥerɛ̄t

English

Notes

śerɛ̄nu̍ t (pl. śerɛ̄ni̍tə) ṯaʕr (pl. ṯaʕyo̍ r)

cloudburst shower

dim. of śiryũ ̍t above; less heavy rain than śiryũ ̍t short-lived and often occurs more than once in a day; dim. ṯāʕa̍ r dim. rɛ̄si̍̌ ̃ š ̃

rəbs̃is̍ ̃

light rain; steady drizzle laġs (pl. laġyo̍ s and laġsi̍tə) mist, low cloud and drizzle mḥũr (√/ḥmr/) drops dripping from trees and shrubs nuś very light drizzle; light spots of moisture felt on the face enṭɛ̄f and birot d-eni̍ṭfən a few drops of rain

of the monsoon months; dim. lāġa̍ s during the monsoon and producing an interesting micro-climate for plants beneath the crown of the tree or bush usually during the monsoon months

seen as without value as they are rarely followed by proper rain

The extent, volume and effects of rain were a matter of consuming interest and the subject of detailed discussion whenever people met. The best rains of all were xɛ̄r jūr!, ‘well-being flows everywhere!’, or o šəxyu̍ r mes de̍ lo!, ‘no-one need ask about it!’ (i.e. its effects are visible to all). Enough rain to cause the dry river beds to flow was described as ḏhēb eśaʕi̍b ʔītə b-inṣani̍tə, ‘both large and small watercourses flowed’; ḏhēb enḥōrtə, ‘the larger flow-beds ran with water’; ḏhēb enḥārni̍tə b-šolo̍ l, ‘the small flow-beds and gullies ran with water’. Rain could be further described as muse̍ ed śọ rg, ‘-lasting rain’; muse̍ iź birdo̍t, ‘rain of large droplets’; muse̍ məbśọ ̍ t,6 ‘enough rain to bring up green shoots’, or just ṯiro̍ igderi̍t, ‘only a dampening of the ground’. The foregoing is to give some background to the songs7 that follow. Although the recordings were not made in the field, but were offered by the singers as examples of the songs they sung, I have heard such songs being sung to

6 √wśỵ . 7 The archive title of each song is given in a footnote so that anyone interested can go to the Endangered Language Archive (elar), soas, University of London, and listen to it.

606

morris and al-shaḥri

livestock at water. The herder sings to the animals to reassure them of his continuing presence, to keep them calm and so encourage them to drink their fill in peace. The songs are sung in a low voice, in an unhurried and meditative way. The song starts with the refrain, some variation of the “boyboy b-boyboy b-boyboy b-boy”, “bey-b-bey, bey-b-bey” theme. This is a signal to the animals to start drinking. The small number of usually rhyming lines are repeated, interspersed with the refrain, until all the animals have drunk. Then a special call is made to bring them up from the water, something like “aaaaaaaaahḥa̍ !”, or “he-he-ḥa̍ !” or “aḥa̍ ! derrrra̍ ʔ!”, ending on a rising note. The subject of the songs is principally about the animals the herder is watching over, but they also sang about their own anxieties and fears, or vented their anger or disappointment with those around them. Some examples of songs sung to cattle at water are given below. Only the lyrics are given, without the repetition or repeated refrain of the original recording. 1.8 The singer reminds his cows that they must drink deep now while they have the chance.

Śḥerɛ̄t9

English translation

1

ɛblīs o yertɛ̄sạ́ n bo̍ hum May the Devil not be among them (i.e. the / b-tǝbġi̍d b-ʕos̃ o rīs̃ cows) / So that you (f. sing.) leave before you have drunk enough!

2

b-doʕo̍ s̃ ʕer tfiṭi̍nhum / For surely you will remember them (i.e. the men ḳerɛ̍rɛ ḏ̣ ɛl̄ e̍ water sources) a / Tomorrow when the sun is well up.

3

her eteśʕi̍r aśaʕr / b-tenu̍ si l-iyũi

When you are pulling up dead, dry grasses / And have to go for two days without being taken to water.

8 1980s reel to reel tape Tape a7 Ali Musallim and Agham Ali_CJ_boyboysongatwater. 9 The Śḥerɛ̄t of songs 1–5 is that of the central region.

drink long and drink in peace

607

Notes l. 1 ɛblīs yertɛ̄sạ́ n bo̍ hum: i.e. upsetting the cows so they are unable to drink enough; there is also an alternative meaning: o yerṭɛ̄sạ́ n biš lo!, ‘Don’t trust him!’ l. 2 ḏ̣ ɛl̄ e̍ roughly 7–9 a.m. 2.10 The singer here seems to be (temporarily?) fed up with his cows—or perhaps they are not his and he is working for someone else.

Śḥerɛ̄t

English translation

1

het ol śink elhu̍ ti / her aġmi̍d ḏ̣ ime̍ b-ḥɛ̄r

You haven’t seen the cows / When they are thirsty and cold as the sun goes down!

2

ol b-ḥije̍t yno̍ f ʕak / b-o They are no good to you / Nor do they relieve yḳo̍ ṭaʕ ʕãk mši̍l you of your burden of debt.

Notes l. 1 ḥɛ̄r, √/ḥbr/, ‘to be, feel cold’ l. 2 ‘They are no good to you’, i.e. they have little or no milk, are not pregnant and are too thin or sickly to be worth selling l. 2 mšil, lit. ‘the load carried’, commonly used in poetry for the burden of debt. The herdsmen of the interior were nearly always in debt to people in the coastal towns and villages, mainly for the sardines they needed to feed their animals in the long dry season, but also for other necessities. This debt was commonly repaid in butter-oil, and many families from the town moved up to the mountains during the post-monsoon season to supervise the making of the butter-oil (in some cases, making it themselves with their host’s milk) which was owed them. 3. The singer of the song above then went on to sing another song, this time in praise of cows.

10

1980s reel to reel tape Tape a7 Ali Musallim and Agham Ali_CJ_boyboyatwater2.

608

morris and al-shaḥri

Śḥerɛ̄t

English translation

1

erḥamu̍ n k-elhu̍ ti / ʕayro̍ b iź ol ykīl

God be with the cows / Dracaena who (give) without measure!

2

iź yḳo̍ ṭaʕ ɛ̃ke̍blef / b-yḥõl ɛ̃ši̍l

Who lighten the load of one with many dependents / And bear the burden of debt.

3

men ṯeru̍ t yḳo̍ ṭaʕ mut / From (the sale) of two (cows) it is lessened by a hundred (silver dollars) / From the debt owed to men duhn eṢilīl (people in) the settlements of Ṣalālah.

4

b-ṯeru̍ t yũl aḳa̍ḥf / ḥa̍dǝn ṭad yḥĩl

And (the milk of) two cows fills up the round clay pot (so full) / That one man is barely able to lift it.

Notes l. 1 ʕayro̍ b is Dracaena serrulata, an important tree in earlier times, because very strong cordage could be made from its leaf fibre, one that was suitable for marine use as well as on land. Different types of livestock often have generic nicknames (see ms̃iroʕ in Song 5 below), and it is said that ʕayro̍ b is an example of this. In everyday speech ʕayri̍b (sing.) designates either a female frequently aborting in early pregnancy, or one currently unproductive, but this is unlikely to be the meaning here in a song in praise of the cows. l. 1 ykīl √/kyl/, to measure out (usually cereals) l. 2 ɛ̃ke̍blef: sp. ber iḏili̍n ḏ̣ irš mke̍blef, ‘so-and-so has many people dependent on him (in addition to his own immediate family)’ l. 3 eṢilīl, an interesting use of the plural form for the market town of Ṣɛlo̍ lt, the regional capital. This reflects reality since in earlier times Salalah (as the name appears on signposts today) consisted of a series of villages strung out along the coast: Daḥarīz, al-Ḥuṣn, ʕAwḳad and so on. So people would say: ʕaḳ ḥalīl iź ṣɛlo̍ lt, ‘in one of the settlements of Ṣalalah’ l. 4 ḳaḥf (pl. ḳaḥi̍f ), a round pot made from local clay, shaped by hand without use of a potter’s wheel

drink long and drink in peace

609

4.11 The singer here praises his cows but also points out what their care costs him.

Śḥerɛ̄t

English translation

1

erḥamu̍ n k-elhu̍ ti / axɛ̄ryu̍ n iź ṭīb / aḥaro̍ f ʕafiri̍tǝ

God be with the cows / Of the golden blood-line of Bēt Xɛryɛ̍n / Discs of red (gold)!

2

her ġotolḳo̍ t ere̍mnem When the wild sea runs high up the beach / For / lis̃ yġo̍ si aḥrɛ̄t / hes your (f. sing.) sake he wades to his knees in the eshi̍m ḥāri̍tǝ black waters / As the dark waves come crashing and rolling in.

Notes l. 1 axɛ̄ryu̍ n, the cow bloodline of Bēt Xɛryɛ̍n l. 1 aḥaro̍ f (sing. ḥarf), ‘discs of silver or gold worn by women and girls, often tied to the hair to hang down the forehead’ l. 2 yġo̍ si, ‘to wade into water up to the knees’; here either to scoop up the sardines tipped out of the girīf sardine drag net, or to help pulling the filled girīf net up the beach l. 2 aḥrɛ̄t, √/ḥwr/, ‘to be black, dark-coloured’; ḥɔ̄ r (masc.), ḥɛ̄ro̍ t (fem.), hɛ̄ri̍tə (pl.), ‘black, dark-coloured’, a common epithet for a dangerous sea l. 2 eshi̍m: se̍həm describes a rough sea with a succession of high waves running up the beach. However in the long Śherɛ̄t dūrūr poems, we also find the phrase eshi̍m ḥɛ̄ri̍tə used to describe large sharks and whales, and there may well be an echo of this here 5.12 A final example of a song praising a cow.

11 12

1980s reel to reel tape Tape a7 Ali Musallim and Agham Ali_CJ_boyboysongatwater2. 1982 Tape42_Tawi Atair_Kherut_SideAanon_EJ_sung_boyboy.wav; singer ʕAli Aḥmad Saʕīd Al-Shaḥri.

610

morris and al-shaḥri

Śḥerɛ̄t

English translation

1

li-ms̃iroʕ id ɛsmu̍ lk / iḏ I say the bismillāh for a cow / Which neither ol tero̍ keṣ́ b-o taʕɛ̄ṣ́ kicks out nor moves off (as I try to milk her).

2

b-ol ekifki̍f eḥilb / kɛs yəbġo̍ d id ertɛ̄ṣ́

Nor fidgets around at milking time. / He (i.e. the man doing the milking) leaves her well contented.

3

b-d-eśḅ o̍ ṭ nhɛ̍rɛ / yifteri̍sẹ́ n bis eḳɛ̄ṣ́

And offers the milk around at midday. / The herdsman is delighted with her.

Notes l. 1 cows in the eastern monsoon mountains are often referred to as ms̃iro̍ ʕ (in the central area bis̃ira̍ʕ) l. 1 taʕɛ̄ṣ √/ʕṣy/, as ḏinu le̍ʔ ʕīṣi̍t, ‘this cow won’t stand still to be milked (but tries to walk off)’ l. 2 ekifki̍f, as ʔoz ḏi̍nu ekifki̍f tɔʔ, ‘this goat is making it difficult for me to milk’ (i.e. forcing me continually to adjust the position of my hands and squatting position) (in the dialect of the central region ekilki̍l eḥilb) l. 2 kɛ̃s = ken + s, from her l. 3 nhɛ̍rɛ: the morning runs from fe̍jer, ‘sky beginning to lighten’, to ʕaḳ enjəho̍ t, ‘dawn’, to ḏ̣ ēlɛ̄, ‘around 7–9 a.m.’, to ḏɔ̄ l (√/ḏbl/), ‘around 9–10.30 a.m.’, to nhɛ̍rɛ, ‘around 10.30–11.30 a.m.’ to ḥum eyuhm, ‘the hottest time of day’ l. 3 eḳɛ̄s,̣́ √ḳbś,̣ to herd, to look after (cows, camels) To turn to songs sung to goats. 6.13 As well as asking God to look after his favourite goat, the singer also criticises those who lack generosity.

13

1980s reel to reel tape Tape 70 Bu Ammaar_EJ_sung_boyboy_togoatsatwater.

drink long and drink in peace

611

Śḥerɛ̄t

English translation

1

l-Ōdo̍ r ed-ɛsmu̍ lk / təmta̍ʕ men eśēkitǝ

I say the bismillāh for my goat of the Ōdo̍ r blood-line / May she enjoy freedom from all care and hardship!

2

se tḳo̍ ṭor l-ɛnɛ̄f / b-iźo̍ k She goes about on her own / And those others ešēki̍tǝ follow along behind.

3

helk en ʕar edīle̍m / b-enbe̍ idifiri̍tǝ

Oh keep us well away from those who don’t know how to behave and never share their food! / Those who wish only ill for others!

4

iźo̍ k il ol yḳu̍ rb ʕa̍śer / b-aʕbe̍r b-ġahi̍tǝ

Those who, instead of bringing people close in true friendship / Are the cause of brothers falling out.

5

b-in kun ḳili̍l ynu̍ faʕ / yṣof id ṣilḥi̍tǝ

For even if what offered is but little, it does good / If it is offered warmly, from the heart.

Notes l. 2 l-ɛnɛ̄f: sp. ‘on her own’ would be l-ɛ̄nu̍ f : is l-ɛnɛ̄f poetic licence? See also ʔɔz ʔinfi̍t, ‘a goat always in front, the one who leads the other goats’; and sp. ʕaḳ ɛnɛ̄f iź ɛ̍ru̍ n, ‘among great numbers of goats’ (the exact number and location unknown or unstated); but l. 2 b-iźok ešēki̍tǝ, poet. ‘following behind’; also ‘a group of goats which has peeled off from the main herd and gone their own way’ l. 3 helk en = Ar. yā waylanā!, ‘Oh woe is us!’ l. 3 ʕar edīle̍m: dilm l-ɛ̄ḳe̍t (masc.), diliũ ̍t l-ɛ̄ḳe̍t (fem.), glossed as ‘extremely mean, not doing the right thing, especially as regards offering hospitality (and even to the extent of concealing any food s/he has)’. In speech dini̍ś (masc.) and dini̍śt (fem.) is more common l. 3 b-enbe̍ idifiri̍tǝ: sp. helk ʕar enbe̍ b-nīt difiri̍t!, ‘Oh woe to you (m. sing.) wishing nothing but ill on others!’ The opposite is enīts erḥĩt, ‘she wishes others well.’ l. 5 yṣof: sp. šu̍ tum id-ṣifi̍t, ‘the skies cleared’; ḳɛlb ɛ-ber iḏili̍n ṣi̍fi, ‘so-and-so’s heart is pure and honourable’

612

morris and al-shaḥri

7.14 The singer deplores the ignorance of the younger generation and the lack of expertise of cattle people with goats.

Śḥerɛ̄t

English translation

1

sənle̍ Fu̍ dǝt inḳiśo̍ t / sənle̍ Fu̍ dǝt inḳiśo̍ t

The Lord take good care of Fu̍ dət who has grown so thin! / The Lord take good care of Fu̍ dət who has grown so thin!

2

ḏi̍nu ol jilyu̍ t b-ʕAbḳad This one has never been taken to the market in / b-ol b-jizḥa̍t rɛśɔ̄ t ʕAwḳad / Or been tethered to the criminal’s stake.

3

b-ol ḥaśị ro̍ t iḏ-iź ilhu̍ ti Nor is she a house-goat for cattle people / / her yaḥlo̍ bis ɛrśɔ̄ t Milked by boys.

4

b-is̃o̍kṯer inaʕi̍ts / b-yaʕmu̍ r id īź̃ ɔ̄ t

Who think her udders are full to bursting / And say (in admiration): “It’s full!”

Notes l. 1 sənle̍ …!, a common invocation in poetry: ‘God take care of …!’ l. 1 Fu̍ dǝt, a well-known blood-line of goats l. 2 jilyu̍ t, √ /jlb/, to take to market, passive voice l. 2 ʕAbḳad, this is the village of ʕAwḳad, today in western Salalah, and formerly the market town for the people of Bait Saʕīd from the mountains above (the tribe of the singers) l. 2 jizḥa̍t, in al-Huṣn, the seat of government in old Salalah, a pole to which minor criminals were tied and beaten, usually for theft ‘from the government’. This was a punishment for (former) slaves or those of slave descent, rather than for other townsmen or people from the interior l. 3 ḥaśị ro̍ t, ‘a goat in milk kept as a house-goat and fed mainly on household leftovers’ l. 3 ɛrśɔ̄ t, ‘boys up to the age of puberty, 14–15 years old’. l. 4 b-is̃o̍kṯer inaʕi̍ts, ‘who see her udder as filled with milk’. The singer is implying that anyone who knows anything about goats would recognise

14

1980s reel to reel tape Tape a7 Ali Musallim and Agham Ali_EJ_boyboyatwater3.

drink long and drink in peace

613

that his goat was not in full lactation, and the boys are thereby displaying their lack of experience with goats. l. 4 b-yaʕmu̍ r id-īź̃ ɔ̄ t (√/mly/) ‘and say “It’s full!” ’; something that no experienced older person would say, since such praise risks bringing down the Evil Eye 8.15 The singer here is not so enamoured of his goats.

Śḥerɛ̄t

English translation

1

helk aġmi̍d õliš / ʕayḳe̍b iź o ḥalɔ̄ n

To be pitied is the man whose animals, as the sun goes down / Are (like) birds which give no milk.

2

iź men śiryũ̍ t taxs̃u̍sə̃ n Who go to pieces after a fierce downpour / And / u-men śɛrb təgerɔ̄ n get mange from small ticks.

3

hel aġmi̍d eṣɛlḥi̍tə / terteḳo̍ dən b-təḏɔ̄ rn

4

b-ḏ-axa̍fk bi̍sən ne̍hel / But if you spend the first night with them in a ḏ̣ ɛr ešɔ̄ k aʕrɔ̄ n place unknown to them / Crowd around and are all over you.

Who, if they are in good health at sunset / Kick up their heels and are still dissatisfied.

Notes l. 1 ʕayḳe̍b (sing. ʕayśị t̍ ), ‘birds’: here he compares his goats to birds because like them, and unlike cows and camels, they suffer badly in severe weather l. 2 taxs̃u̍sə̃ n, lit. ‘to decompose, to crumble away’, as heru̍ m ber xis̃, ‘the trees were rotten’; mgiffo̍ t ḏ-īro̍ t xos̃sõ ̍ t, ‘a corpse which had already decomposed’ l. 2 śɛrb, ‘a species of small tick’, believed by herdsmen to be the cause of sarcoptic mange in goats and camels l. 2 təgerɔ̄ n, √ /grb/, ‘to have, catch mange’ (which is infectious)

15

1982 Tape36_Jufa_1_IV_SideA_anonman_EJ_beybey_poemsungtolivestockatwater. The version transcribed here has been corrected by Sālim ʕAwaḏ̣ .

614

morris and al-shaḥri

l. 3 təḏɔ̄ rn, ‘to be continually dissatisfied, to constantly grumble’, as seh kels ḏɛ̄ro̍ t, ‘she’s always dissatisfied’ l. 4 ne̍hel, when transhuming, the first night spent away from the homestead: animals tend to be restless and nervous on their first night in unfamiliar surroundings l. 4 ḏ̣ ɛr ešɔ̄ k, lit. on your back l. 4 aʕrɔ̄ n, sp. aʕrōn ɛru̍ n ḏ̣ ɛr šɔ̄ k, lit. ‘the goats crowded around your (sing.) back’; ɛru̍ n ḏ-aʕri̍b ḏ̣ ɛr šīyen (pl.), lit. ‘the goats crowded around our backs’ 9.16 In this song the singer rails at sorcery.

Śḥerɛ̄t

English translation

1

ber ero̍ dən ḳot mulu̍ i / We took them to water at a pool slow to fill / b-enfo̍ śən Ãḥsu̍ n And we took them off again in the late afternoon to Ãḥsu̍ n.

2

b-iź fižḥāri̍tə he̍te / ɛ̃śe̍bki tġaśũn

The coven of witches (make it) a dangerous place. / They infect it with (their) wicked deeds.

3

kam men śe id ber fre̍riš / b-enḥa̍n ed aḥtũ̍ n

How many (lives) have you (witches) taken off with you! / And we know that for a fact!

Notes l. 1 ero̍ dən, √/wrd/, ‘to take to water’ l. 1 mulu̍ i: mulu̍ i (masc.); miliyo̍ t (fem.), ‘slow; taking a lot of time to do something’ l. 2 fižḥāri̍tə, dim. form; (sing. fižḥe̍r, pl. fiže̍bḥar), ‘an area high on the escarpment face’, the usual meaning. But there is a more specialised and local meaning here: the iź fižḥāri̍tə is a clan of witches (words never to be spoken in front of any woman) l. 2 he̍te, sp. igderi̍t he̍te, ‘a place of danger (especially from wolves and leopards)’ l. 2 ɛ̃śe̍bki < mśe̍bki, ‘evil, sorcery; wicked actions’

16

1982 Tape40_IV_Jufa_SideA_anon_EJ_sung_boyboy6_atwater.

drink long and drink in peace

615

l. 2 tġaśũn, √/ġśm/, ‘to infect someone, pass on a germ’, as seh ġiśyũt, ‘she passed on the germ’; also ‘to find, discover’ l. 3 kam: in sp. ‘how many’; ‘how much’ is mśe̍ l. 3 fre̍riš, ‘to make fly; to pick up and carry off’ l. 3 ed-aḥtũ̍ n,√/ḥtm/, ‘to be certain, sure of something’ 10.17 This singer describes water sources available to his goats.

Śḥerɛ̄t

English translation

1

o tis̃ti̍ḳen iź Śo̍ jo / mih You won’t drink at Śo̍ jo / It’s brackish water, you ġaro̍ r ol teśẹ ro̍ gən can’t quench your thirst with it. (i.e. even if there is plenty of it)

2

ʕar iźo̍ k id ṭaʕru̍ n / iź men ser eḏəhētə

3

lə-hes bles̃ tertɛ̄sẹ́ n It’s no use trying to satisfy your thirst from them bo̍ hum / b-tġa̍d b-ʕos̃ o / And you’ll go off with your thirst unquenched. rīs̃

4

b-seh ʕar e-ḳerɛ̍rɛ / b-de̍ śelo̍ ṯ eśe̍lṯ

But instead (you’ll have to wait) for tomorrow / For the one who takes the animals to water every third day.

5

ehe̍lk aḥāri̍tə / woh! boy b-boy-b-boy b-boy b-bo derrrrrrrrrrra̍ !

Oh the exhaustion of (mucking out) the kids quarters! / woh! boy b-boy-b-boy b-boy b-bo derrrrrrrrrrra̍ !

Or those full of sediment / Left behind by the floodwaters.

Notes l. 2 ṭaʕru̍ n, sing. ṭaʕri̍t, sediment l. 2 eḏəhētə, √/ḏhb/, floodwaters l. 3 bles̃ < Ar. dialect bilāš, ‘for nothing; it is useless, pointless’; bles̃ o ti̍ṣer bo̍ hum!, ‘there’s no point trying to stop them!’

17

1982 Tape40_IV_Jufa_SideA_anon_EJ_sung_boyboy7_atwater.

616

morris and al-shaḥri

l. 3 bo̍ hum, from them: i.e. the brackish and the dirty waters: goats are notoriously fastidious about the water they drink l. 3 ʕos̃ = ʕād + s̃, you (fem.) have not yet l. 4 śelo̍ ṯ, to take animals to water every third day (śhəli̍ṯ, ‘three’; eśe̍lṯ, ‘three (days)’). It is suggested that in time the sediment will sink to the bottom and the water above will have cleared l. 5 helk, exhaustion, physically demanding work; lob ʕar hilko̍t!, ‘she was really exhausted’ l. 5 aḥāri̍tə (sing. ḥohr), quarters for young kids, of stone or brushwood or in small caves 11.18 The singer longs for rains and expresses his preference for goats over camels.

Śḥerɛ̄t

English translation

1

l-Eṣʕo̍ r d-esmu̍ lk / I say the bismillāh for (my goat) Eṣʕo̍ r / emborṯo̍ t men ser aʕū̃ i Inherited down the generations from my ancestors.

2

mu̍ tnik əl-õse̍ do̍ ṯe / b-ǝl-ʔɛ̄rśị t̍ ə kel esūi

3

b-ṭob ed Ḳo̍ hof tǝklɛ̄n / How I wish they (i.e. my goats) were coming men kənu̍ ʕal b-ũlūi back at dusk to Ḳo̍ hof / From the lower slopes of the peaks along the winding tracks!

4

bǝli̍ enfǝśi̍nsǝn ǝlEśạ ʕi̍tǝ / ḥaṣ-ɛ̄r es̃o̍sọ́ r egūi

5

ed l-i̍bdel b-erēḳ / b-ol So that they could have a change from the waste b-aʕdo̍ b lǝ-kūi paper and cardboard (of the coastal villages and towns) / And no longer circled around the low rocky outcrops (searching for something to eat)

18

I long for the summer rains / And for all the lands to receive the same (rain).

Or that we were taking them in the afternoon to Ešạ ʕi̍tə / At a time when the gullies running down the escarpment were greened over.

1982 Tape40_IV_Jufa_SideA_anon_EJ_sung_boyboy2_atwater.

drink long and drink in peace

Śḥerɛ̄t

617

English translation

6

ed s̃aḥḳōn eśiro̍ stə / ol They are so dear to me, the ones with the b-ɛ̄ḳro̍ s̃ lǝ-śū̃ n snipped ears! / They will never be sold for silver dollars.

7

her neṣre̍fhum lǝfigǝru̍ ti / b-enḳele̍b ed ũxūi

To be spent on the camels of the stony desert / And the rest put into pockets.

Notes emborṯo̍ t, √/wrṯ/, inherited aʕū̃ i, √ /ʕwm/, ʕum (pl. ʕim) grandfather; ancestor esūi, √ /swy/, to be equal ṭob, how I wish. Ar. yā laita Ḳo̍ hof (i) proper name; (ii) high-ceilinged cave or overhang at the a base of a cliff kənu̍ ʕal (sing. kənʕolo̍ t), lower slopes of a mountain range; small hills at the base of a mountain range ũlūi, √/lwy/, to twist and turn; to bend (trans.) enfǝśi̍nsǝn < enfe̍ś, ‘to take animals off in the late afternoon’ egūi (sing. jōt), in other dialects of Śḥerɛ̄t ‘holes; hollows; depressions’, but in the areas east of Mirbāṭ it has the specific meaning of ‘gullies, in shape like inverted Vs, which run down from the lip of the escarpment of Samḥān above’ erēḳ, goats that spend their time in villages and towns eat waste paper and cardboard (which indeed have some commercial value in a time of drought) aʕdo̍ b, in this area, rocky outcrops emerging from the flat dry plain ( jerbīb) s̃aḥḳōn, √/ s̃ + ḥḳb/, to be dear (to), be beloved (by) eśiro̍ stə, goats, like other livestock in Dhofar, are marked to designate ownership; śirȯstə are slanting cuts made from the middle of the outside edge of the ear downwards ɛ̄ḳro̍ s̃ (sing. ḳɛrs̃), Maria Theresa silver dollars, a former currency of the region lǝ-śū̃ n, √/śʔm/, to sell figǝru̍ ti (fem. pl.), camels of the fe̍ger, the stony desert areas to the north of the Dhofar mountains ũxūi (sing. maxbe̍), √/xby/, pockets

618

morris and al-shaḥri

12.19 However, some songs are just for amusement, or to poke fun at someone else, as here.

Śḥerɛ̄t

English translation

1

ʕar kun tśɛm ʔɔzis̃ / u-berkel etefḳīš

You should sell your she-goat / So you could wear a dress of polished indigo.

2

b-fəlo̍ gero̍ b id tũr / ɛ̃śto̍ ḏe̍̍nu tetīš

Or instead (buy) a gerāb of dates / And spend the winter eating them.

3

b-fəlo̍ firi̍k əl-ʕayni / yo̍ gzer Ber Ḥadīd ilīš

Or instead leave your husband tonight. / Let Ber Ḥadīd have to promise to slaughter his cow (to bring you back)!

4

b-fəlo̍ bdīlis̃ b-yo̍ šfoḳ / Or instead let him swop you for another and b-yaġṭo̍ b aʕādīš marry her / And the Lord rid him of his enemy!

Notes l. 1 ʕar kun, it should be that. Ar. wājib ʕala … l. 1 berkel, a dress of indigo cloth polished to make it shine; sp. more commonly ḳebḳe̍b l. 2 gero̍ b < Ar. jarāb, a large, oblong woven container of pressed dates l. 3 yo̍ gzer: if a husband wants to bring his angered wife back, especially if she left her children behind, a group of his associates go to wherever she has taken refuge to try and persuade her to return, promising that her husband is penitent and will slaughter a precious animal (such as a cow) for her on her return l. 4 b-yaġṭo̍ b aʕādīš, sp. ɛ iḏili̍n yaġṭo̍ b aʕādīk!, ‘Oh so-and-so, God rid you of your enemy!’ 14.20 And a song may reflect an old feud. Here the singer mocks a man (the ‘goat’ of the song) whom he considers to be of little value.

19 20

1983 Tape54_IV_Fushi_II_SideA_anon_EJ_sung_boyboy_atwater (second song). 1982 Tape40_IV_Jufa_SideA_anon_EJ_sung_boyboy5_atwater (first song).

drink long and drink in peace

Śḥerɛ̄t

619

English translation

1

ʕot ol s̃eḳhiyo̍ t lǝ-ho̍ jer She (goat) no longer spends the midday hours / b-ol hɛ̄s elfa̍x in burning heat / Nor do the flames of the sun at its zenith reach her.

2

ʕāls Ber Ḥadīd ɛ̄r ʕãyɛ̍r (For) her owner is Ber Ḥadīd ɛ̄r ʕãyɛ̍r / And she / b-iḏbəli̍ts Ɛ̄ḳa̍ḥ is under the protection of Ɛ̄ḳa̍ḥ

3

ol ḳehe̍b se̍ris teje̍bri / b-ol ḳīśị t̍ d-ɛ̃da̍ḥ

4

ke̍se heryɛ̍n mente̍briṯ He (i.e. the thief) only found a little shrivelled / b-ol śaġa̍niš ekēla̍ḥ plant / He wasn’t impaled on a protruding knot (of wood).

People didn’t gather from here and there at midday to discuss her (theft and disappearance) / Nor was any great amount of money demanded for her in compensation.

Notes l. 1 The meaning of the line is that this man doesn’t know the meaning of hardship: he is untried and untested l. 1 ol hɛ̄s, sp. lob ʕar lə-hɛ̄ tun ho̍ jer!, ‘the real heat of the midday sun is indeed upon us!’ l. 1 elfa̍x, poet., here the same as ho̍ jer, ‘heat of the midday sun’, but sp. elfa̍ x men ḥiṭ ḏo̍ hun!, ‘all there is for you is that (leftover) rice!’ l. 2 i.e. he has two powerful protectors l. 3 teje̍bri, sp. təjərɔʔ ḏə-yɔ, ‘groups of people for a major event (as a death, a serious problem that needs to be discussed etc.)’. Here the implication is that no-one cares what has happened to this ‘goat’. l. 3 ḳīśị t̍ , here the implication is that the ‘goat’ was not worth much l. 3 d-ɛ̃da̍ḥ, sp. ol mi̍daḥ lo, ‘not many, not much’: ol ḥu̍ ṣelk mi̍daḥ lo, ‘I didn’t get much’ l. 4 heryɛ̍n, dim. of herɛ̍m, ‘bush, shrub, plant’. Here the thief found only a worthless goat: without his two protectors this ‘goat’ is nothing l. 4 śaġa̍niš, lit. ‘he impaled it (a fish) on the hook’. For stealing such a poor and little-valued ‘goat’ the thief’s punishment will not be severe (‘being impaled on a hard knot of wood’)

620

morris and al-shaḥri

13.21 I only give a single example of a camel watering song here, as the Śḥerɛ̄t of these is heavily influenced by the Mehri language (which is interesting in itself, but not under discussion here). Camel watering songs are strongly rhythmical and are sung in a loud, forceful manner. They consist of short single lines interspersed with variations on a “hēʔ wə-hēba̍ y” refrain. The recording here opens with the spoken: ḥa-nərɛ̍ (in Śḥerɛ̄t, the singer’s native language is Śḥerɛ̄t), ‘We’re going to water,’ and this is followed by the call made to encourage the camels to start drinking: “wēy! wēy”. The song ends with the calls “həbš! həbš! derrrr-ḥa̍ ʔ! derrrr-ḥa̍ ʔ! derrrr-ḥa̍ ʔ!” and much clicking of the tongue, blowing of raspberries between pursed lips and other sounds which camels recognise as the signal to come up from water. For reasons of space I present the lines grouped in threes with the refrain where it occurs.

Mehri and Śḥerɛ̄t

English translation

1

ele̍-wə-hēʔ b-hēba̍ y / hay b-Il-Fāṭir ehēba̍ y / ber Māšawm ehēba̍ y

ele̍-wə-hēʔ b-hēba̍ y / Come along (my shecamel) Il-Fāṭir! ehēba̍ y / Of the bloodline of Māšawm ehēba̍ y

2

ed ešġa̍l li-hēba̍ y / ber resuwm ahēba̍ y / men b-aġahl enhēba̍ y

One who works hard li-hēba̍ y / Pure blooded ahēba̍ y / (Come up) from below (to the water)! enhēba̍ y

3

ḥa̍nat yawm ohēba̍ y / hay b-Il-Fāṭir ehēba̍ y / ber Māšawm ehēba̍ y

The day (for watering) has come ohēba̍ y / Come along (my she-camel) Il-Fāṭir! ehēba̍ y / Of the bloodline of Māšawm ehēba̍ y

4

men b-āġahl enhēba̍ y / (Come up) from below enhēba̍ y / The day (for ḥa̍nat yawm ohēba̍ y / watering) has come ohēba̍ y / One who works ed ešġa̍l li-hēba̍ y hard li-hēba̍ y

5

ber resuwm ehēba̍ y / Pure blooded ehēba̍ y / With her pregnancy ḏ̣ ɛr ãḥto̍ ls ehēba̍ y / hes ehēba̍ y / When she gives (a lot of) milk ehēba̍ y ederr ehēba̍ y

21

1982 Tape40_IV_Jufa_SideA_Bu Ammaar_sung2_wateringcamels.

drink long and drink in peace

Mehri and Śḥerɛ̄t

621

English translation

6

seh xaṭayft ehēba̍ y / She is thin ehēba̍ y / Like a crescent moon hes eśe̍her ehēba̍ y / ḏ̣ ɛr ehēba̍ y / With her pregnancy ehēba̍ y ãḥto̍ ls ehēba̍ y

7

hes ede̍rr ehēba̍ y / seh xaṭayft ehēba̍ y / hes eśe̍her ehēba̍ y

When she gives (a lot of) milk ehēba̍ y / She is thin ehēba̍ y / Like a crescent moon ehēba̍ y

Notes l. 1 ber Māšawm, of the blood line of a Kathiri camel stallion, Maʕšūm; in Śḥerɛ̄t the /ʕ/ would be vocalised l. 2 ber resuwm, of pure blood; sp. bis risũt rḥãt, ‘she is of good blood’; the /uw/ vowel is distinctive in Mehri l. 4 ḥa̍nat yawm, the day has come; sp. lob ʕar ḥanu̍ t yũk!, ‘indeed your time has come!’ (e.g. the day I will take my revenge on you) l. 5 ãḥto̍ ls √/ḥtl/, to tie up; sp. aḥtili̍t ɛḏ̣ iro̍ b, aḥtili̍t aśaʕr, ‘a tied bundle of firewood / of hay’. But also, as here, pregnancy: sp. hi̍ni ãḥte̍ls, ‘what is in her belly is for me’ (said by the seller of a pregnant cow or camel to the buyer) l. 5 hes ede̍rr, (poet.) when she gives a lot of milk; sp. dərr, ‘to flow (as sap, milk)’; edre̍r, to get a lot of milk at a single milking Songs like these are no longer heard today. I was near a foothills spring in Dhofar earlier this year and heard high, long drawn out calls coming from the wooded slopes above as a herder brought the goats down to drink at the spring. The calls were like nothing I had heard before. This was not surprising, since when the goatherd appeared he turned out not to be a Dhofari, and his calls were in his own language. I hope the few examples presented here give some idea of what we have all lost. They are important because ordinary, everyday songs such as these, and others sung when soothing a child to sleep, churning butter, weeding, grinding grain, tanning leather or fishing and so on, tell us so much about a way of life which has passed. They tell us about the feelings and the preoccupations of those who did more than just survive in this often harsh and unforgiving environment, but also managed to bring humour, wit and poetry into their and their listeners’ lives.

chapter 31

South Arabian Sibilants and the Śḥerɛ̄t s̃ ~ š Contrast* Alex Bellem and Janet C.E. Watson

The consonant systems of the individual Modern South Arabian languages pose a challenge to linguists working on these languages. In comparison with scholarship on the phonology and phonetics of other Semitic languages, there is surprisingly little work published on Modern South Arabian phonology and phonetics, and even less of this is grounded in primary first-hand data. In this context, we present in this article a study of Śḥerɛ̄t s̃,1 taking as a starting point previously published accounts of the Śḥerɛ̄t voiceless sibilants. Fieldwork conducted by Watson indicated that aspects of the little that the scholarly world knows and perpetuates about Śḥerɛ̄t s̃ were incorrect, so to this end we present in this article a revised account of this consonant. In exploring the question of the origin of the Śḥerɛ̄t s̃, we start by outlining the

* The authors are indebted to Barry Heselwood for his time, help and expertise with the electropalatography and to Miranda Morris for making her sound files available to us. We also thank Ayman Ghummad for his patient assistance with epg investigations conducted in 2015. We would also like to thank our language consultants, and in particular the following, without whose participation the work would not have been possible: Ali al-Mahri, Saeed al-Mahri, Mohammad al-Mahri, Khalid Selim Ruwaya, Faysal Bakhit al-Mahri and Muhammad Ahmad Ali Amar al-Gid al-Mahri. Finally, thank you to Watson’s ‘whatsapp’ consultants Abdullah Musallam al-Mahri, Ahmad Hardan and Khalid Ruweya al-Mahri for checking data themselves, and in the case of Abdullah, with Śḥerɛ̄t speakers in the field. We are grateful to the Leverhulme Trust for project grant rpg–2012–599, which enabled this work to be carried out. 1 There is a considerable amount of variation in the terms for this language, as noted by Simeone-Senelle (1997:379). It is commonly known as Jibbāli (in Arabic and much western literature on the language) or Geblɛ̄t (in Śḥerɛ̄t), but this designation is, in our opinion, incorrect. In Śḥerɛ̄t and other msal of the mainland, the mountains of Jabal Al Qara and Jabal Al Qamar that receive the monsoon rains are known as śḥɛr (Śḥerɛ̄t, Hobyōt), śḥayr (Mehri), śhɛr (Baṭḥari) and hence the name of the language, śḥerɛ̄t in Śḥerɛ̄t and baʕlīt śḥayr in Mehri (a term that describes both the Śḥerɛ̄t-speaking people and the language), denotes the language of the people who reside in the monsoon-affected mountains. These mountains are distinguished from the dry mountain ranges further east and west, which are known as gyɛl in Śḥerɛ̄t.

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

623

historical correspondences of the Modern South Arabian (henceforth msa) sibilants—including Ancient South Arabian2 (asa) s1, s2, s3—and equivalences that are known in contemporary msa. We then move on to a synchronic view of Śḥerɛ̄t s̃, which is a typologically unusual sound, recognised for Central Śḥerɛ̄t, but previously said not to be present in Eastern Śḥerɛ̄t. Contrary to what was previously perpetuated in the scholarly literature, we show that the s̃ is indeed also a feature of Eastern Śḥerɛ̄t. We outline the phonological status of these sibilants, and then present evidence of how the s̃ is actually realised by different speakers, with perceptual, with visual, and with articulatory (palatographic) evidence of the articulation of the sibilants in Śḥerɛ̄t. It becomes clear that there is considerable variation, but that it is not phonetically simply a rounded counterpart of š, and we argue that it is most often alveo-palatal for speakers who have a clear distinction.

1

Background

1.1 Outline of the Modern South Arabian Languages The Modern South Arabian (msa) languages are a group of related south Semitic languages spoken in the southern part of the Arabian Peninsula, predominantly in Yemen and Oman. Traditionally, they were classified as southeastern members of the South Semitic branch of West Semitic along with asa and Ethio-Semitic languages, with Arabic forming the other branch of South Semitic group.3 Under a revised classification4 accepted by many (but not all) Semitists, the msa languages are now considered to be eastern members of the South Semitic branch of West Semitic, while the Ethio-Semitic languages are the western members; the main difference being that Arabic and asa are now considered to be less closely related to msa languages than previously, being

2 We use the term adopted by Michael Macdonald (e.g. Macdonald 2000:30), noting that he rejects the nomenclatures Epigraphic South Arabian (as describing “a language group by the materials on which it is written”) and Old South Arabian (as misleading, since it incorrectly implies direct ancestry of msa). As Macdonald notes, asa is a collective label covering both the Ṣayhadic languages (Sabaic, Maḏābic, Qatabanic and Ḥaḍramitic, which are the languages traditionally known as ‘Epigraphic’ South Arabian, Altsüdarabisch, etc.) and the nonṢayhadic languages—the other ancient languages of the region of which not much is known. 3 This view prevailed throughout much of the twentieth century, e.g. Brockelmann (1961), Moscati et al. (1964), Ullendorff (1970), Bergstrasser (1983). For discussion, see Faber (1997), Huehnergard (1995, forthcoming). 4 Started by the work of Robert Hetzron (e.g. 1974, 1976).

624

bellem and watson

part of the Central branch of West Semitic, along with Northwest Semitic languages such as Hebrew, Aramaic, and so on.5 The six msa languages are Mehri, Śḥerɛ̄t, Ḥarsūsi, Baṭḥari, Soqoṭri and Hobyōt. Those with the largest number of speakers are Mehri (also the most geographically widespread), Soqoṭri and Śḥerɛ̄t, although estimates vary considerably. For Mehri it is particularly difficult to estimate speaker numbers, partly because there are no reliable census figures for ethnic groups, partly because many Mehri tribal members do not actually speak Mehri, and partly because the speech community is spread from eastern Yemen (Qishn, as the west-most point of the Mahra), to Dhofar in Oman, and north into Saudi Arabia. Watson (2012:1) notes reasonable estimates of 100–180,000, although some contemporary estimates set the lower end as low as 75,000 speakers.6 For Śḥerɛ̄t, estimates are also not easy because many Śḥerɛ̄t speakers are from the Mahra tribes, for one reason, although a reasonable estimate might be around 30,000 (Watson 2012:1).7 Hobyōt and Ḥarsūsi are both estimated to have under 1000 speakers,8 and Baṭḥari now under 20 (Miranda Morris, p.c.).9 Tom Johnstone, author of the Mehri Lexicon, Jibbāli Lexicon and Ḥarsūsi Lexicon,10 believed that all the msa languages were severely endangered, stating in 1982 that Mehri, Śḥerɛ̄t, and Ḥarsūsi were “in grave danger of dying out” and that “all three languages will have been replaced [by Arabic] within ten years”;11 fortunately, both Mehri and Śḥerɛ̄t are still spoken, although the predominance of Arabic threatens both. There is renewed interest in intangible cultural heritage, as evidenced in the collection of oral poetry, tales and histories by both native speakers and non-native speakers, and this may help to prevent further weakening of the languages, particularly with the advent of social media and the 5

6 7 8 9 10 11

This is based on the latest work by Huehnergard (2017), cf. also Huehnergard (1995), Huehnergard & Rubin (2011). A similar (revised) classification, but with some crucial differences, is to be found in Faber (1997), who notes that her model is based on the revisions of Rodgers (1991) and Huehnergard (1992) to the work by Robert Hetzron previously cited. See Rubin (2010b:8). Rubin is more optimistic, giving 30–50,000 (2010b:8; 2014:3). Watson (2012:1), Rubin (2010b:8). Watson (2012:1). This is down from the 300 speakers estimated by Morris (1983:130), see Simeone-Senelle (2011:1075). Johnstone (1987, 1981, 1977). Quoted from “Jebali Dictionary to Be Ready by 1982: Prof. Johnstone Tries To Record Dying Languages”, in the Times of Oman, January 17, 1980, a cutting of which is to be found in the archive that Johnstone bequeathed to Durham University, and which is now the University’s Palace Green Library (joh31).

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

625

development of mobile phone technology, meaning that at least some of the msa languages are now commonly being written, even if mostly informally. 12The earliest major work on the msa languages arising from primary fieldwork was the documentation of the Vienna expedition in the early twentieth century, published as Jahn (1905) and Bittner (1909–1914). Following this was the pioneering work of T.M. Johnstone that culminated in the lexica of Mehri, Ḥarsūsi and Śḥerɛ̄t (the Jibbāli Lexicon), as noted above, but which was cut short by his sudden and untimely death. The study of these languages was further developed by fieldwork conducted by Antoine Lonnet and MarieClaude Simeone-Senelle,13 and on Soqoṭri by Vitaly Naumkin and Vladimir Porkhomovsky.14 Fieldwork on Mehri was begun by Alexander Sima, but this was also curtailed by his tragically early death while conducting fieldwork in eastern Yemen in 2004.15 Work on Mehri has continued with fieldwork conducted by Watson16 and by members of the French-led OmanSAM team. A Śḥerɛ̄t grammar based on Johnstone’s materials but with fieldwork conducted with native-speaker informants in the us has also recently been published17 by Aaron Rubin, a habilitation on the msa verb system has been completed by Julien Dufour, and Watson has been undertaking fieldwork on Śḥerɛ̄t in Oman since 2013. There have recently been two large projects investigating various aspects of the msa languages and working with primary material: a Leverhulme-funded project led by Watson, with fieldwork in Oman being conducted by her, Domenyk Eades and Miranda Morris, and the resulting, large corpus of data being archived with elar;18 and the OmanSAM team funded by the anr. In addition, in 2015, a team led by Viktor Naumkin and Leonid Kogan published the Corpus of Soqotri Oral Literature.19 Other work on msa languages based on previously published or otherwise available material has also contributed considerably to the body of literature.20 12

13 14 15 16 17 18 19 20

An up-to-date, comprehensive bibliography of sources on msa languages is available in the “Resources” section of the msal project website: http://www.leeds.ac.uk/arts/info/ 125219/modern_south_arabian_languages/2376/resources (last accessed 17th June 2016). E.g. Lonnet & Simeone-Senelle (1997). Naumkin & Porkhomovsky (1981, 1995, etc.). See the work published in Sima (2009) and the “Introduction” by Watson (pp. 1–28). Producing a number of outputs, particularly Watson (2012). Rubin (2014). Endangered Languages Archive, based at soas in London http://elar.soas.ac.uk/ (last accessed 17th June 2016). Naumkin et al. (2014). Such as (but not limited to) Leslau (1938), Wagner (1953), Stroomer (1999, 2004), Rubin (2010a), as a selection of the larger works, in addition to various smaller articles.

626

bellem and watson

table 1

Proto-Semitic and Ancient South Arabian sibilant correspondences and allophones across Modern South Arabian

ProtoSemitic

Ancient South Arabian21

Śḥerɛ̄t

Mehri

Ḥarsūsi

Soqotri

Lateral

*ś *ṣ́

s2 ⟨š⟩ ?[ɬ] ṣ́ ⟨ḍ⟩

ś ś2̣ 2

ś ṣ́

ś ṣ́

ś ṣ́

Sibilant

*s *dz *ṣ *ts24

s1 ⟨s⟩ [š/s] z ṣ s3 ⟨ś⟩ [ts]?

š, s̃23 z ṣ s

š, h z ṣ, ṣ̌ s

š, h z ṣ s

š, h z ṣ s

Velar

*k *g *ḳ

k g ḳ

k g,25 ž, z̃ ḳ, s̃,̣ ṣ̌

k g, ž ḳ, ṣ̌

k g ḳ, š2̣ 6

k g, ž ḳ, ṣ̌

1.2 Historical Outline Table 1 shows the system of sibilants and laterals across four msa languages, with the corresponding historical antecedents reconstructed for Proto-Semitic, and the corresponding asa consonants, including the problematic sibilant s1— s2—s3 series.27

21 22 23 24 25 26

27

Among others, see in particular Nebes & Stein (2004), Stein (2011), Gragg (1997); also, Beeston (1962, 1984:8–9). This is often transcribed in the literature as ⟨ź⟩ or ⟨ẓ́⟩, e.g. Johnstone (inter alia), Rubin (2010a, 2014), Kogan (2011), among others. Note that s̃ also arises synchronically as an allophone of k. Traditionally, s3 was assumed to have had the value [s], but recent work increasingly points towards [ts]. This is comprehensively discussed in Kogan (2011: 65–69). This is often rather palatalised, gy, and in Western Śḥerɛ̄t and for some Eastern Śḥerɛ̄t speakers is j (ipa [ʤ]). Rubin (2014:26) notes j only for Western Śḥerɛ̄t. Swiggers (1981:360) comments on its extremely limited distribution in Ḥarsūsi; here, we should add that ṣ̌ has very limited distribution in all the msa languages and is found in comparatively few lexemes. Among other sources, including the classics such as Moscati et al. (1964), see in particular the overview of Huehnergard (1995), alongside Kogan (2011), Goldenberg (2013), Nebes & Stein (2004), Macdonald (2004:499), Ehret (1995:481–482), Steiner (1977).

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

627

As the table shows, the sibilants include the Śḥerɛ̄t palatal sibilant series, with s̃, usually noted as a marginal phoneme, since it is said to occur only in Central Śḥerɛ̄t, and its emphatic counterpart s̃.̣ Synchronically, Śḥerɛ̄t s̃ is a lexical segment (i.e. a phoneme in its own right), as well as being derived via palatalisation of /k/ (i.e. it also occurs as an allophone of /k/). Śḥerɛ̄t emphatic ṣ̃ is a lexical segment (independent phoneme), as well as being derived via palatalisation of /ḳ/ (i.e. it also occurs as an allophone of /ḳ/); the lexical ṣ̃ has very limited distribution.28 The Mehri emphatic ṣ̌ is the emphatic counterpart of š and is similarly very rare. Since this emphatic has such a limited distribution, we have limited speech data for Śḥerɛ̄t that includes it, particularly visual data (video), and due to space constraints we confine the focus of the current paper primarily to s̃ and its contrast with š and s. We have included the historical Semitic and the Modern South Arabian correspondences and allophones to highlight the following questions: how do the Śḥerɛ̄t sibilants fit into the Semitic and the msa sound systems, and where does s̃ come from? One answer to the latter is that the diachronic and synchronic correspondences reveal what is not there. That is, there is no obvious historical precedent to the Śḥerɛ̄t s̃. Further, looking synchronically across both Semitic more widely, and across the other msa languages is no more enlightening, since there is no clue here either of any obvious origin for the Śḥerɛ̄t s̃. This indicates that although—as we demonstrate below—Śḥerɛ̄t s̃ has phonemic status (for many speakers), this must have arisen historically through a phonological process. The logical conclusion is that at some historical point, early Śḥerɛ̄t or an ancestor language variety would have developed a process of contextual palatalisation, perhaps of *k, such that this historical phoneme (perhaps *k) would have had (at least) two allophones: [k] and (something similar to) s̃; it would seem that at this earlier stage, this ‘perhaps *k’ did not palatalise to *š, since that would have resulted in a complete merger with the pre-existing š. This process must have ‘fossilised’ and become opaque, such that s̃ had a predictable distribution, but could no longer be said to be the output of a synchronic phonological process.29 At some later stage, perhaps due to vowel shift, or perhaps due to koineisation or heavy borrowing from another language variety, a later (modern) stage of Śḥerɛ̄t again developed a context where k could be

28 29

This is marginal, and rare. As indicated in table 1, it derived historically as an allophone of ḳ (Rubin 2014:26). The reader is referred here also to the discussion and correspondences given in Kogan (2011:107) as well as Lonnet & Simeone-Senelle (1997).

628

bellem and watson

found in a palatalising environment, and developed another—synchronic— process of palatalisation. Therefore, Śḥerɛ̄t s̃ has no direct external or historical origin, but arises through a historical stage of palatalisation (the now-lexical s̃) as well as resulting from a synchronic process of palatalisation of k.

2

Previous Studies of msa Sibilants

With respect to the phonology (and in particular the consonantal systems) of the msa languages, the work of T.M. Johnstone was pioneering, and has at the very least informed, if not constituted the basis of almost all subsequent work on the phonology and phonetics of the Modern South Arabian languages. His main informants for Śḥerɛ̄t, Sālim Bakhīt and Ali Musallam al-Mahri,30 have also worked with subsequent scholars investigating aspects of Mehri and Śḥerɛ̄t and thus informed later descriptions of the phonetics and phonology of these two msa languages. Of special relevance to the current article are Johnstone’s descriptions of the sibilants, particularly the s̃ of Śḥerɛ̄t, the existence of which was first reported by Johnstone in a paper he delivered to the Third International Hamito-Semitic Congress in London in 1978 (later published as Johnstone, 1984). This paper was entitled “New Sibilant Phonemes in the Modern South Arabian Languages of Dhofar”, and he discusses therein four sibilants that he has found in the Central dialect of Śḥerɛ̄t, and which had hitherto been unrecorded: s̃, z͂ , s̃,̣ ẕ́. He explicitly contrasts both Eastern Śḥerɛ̄t and Mehreyyet (Dhofari) Mehri, in which he has not found these sounds. The first of these, s̃, he notes may be both phonemic—as the result of the splitting of Śḥerɛ̄t š into š and s̃— and allophonic (through palatalisation of k), as we also noted in the previous section. He then describes the articulation of s̃: “pronounced with the blade of the tongue on the hard palate and the lips protruding: the breath is forced between the blade of the tongue and the hard palate.” The status of the voiced counterpart z͂ he describes as only an allophonic variant of g;31 it is noteworthy that in his Jibbāli Lexicon (1981, henceforth jl), he does not mention a z͂ variant of g/ j. The ṣ̃ is the emphatic counterpart of s̃, and has both lexical status and

30

31

Ali Musallam was a native speaker of Mehri who learnt (Eastern) Śḥerɛ̄t during his childhood, and he was initially Johnstone’s informant for Śḥerɛ̄t before Johnstone met Sālim Bakhīt, a speaker of Central Śḥerɛ̄t. Johnstone (1984) uses ⟨j⟩ to transcribe this segment, unlike in jl, where he uses ⟨g⟩.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

629

may be derived through palatalisation of ḳ; he finds the lexical ṣ̃ to be the Central equivalent to Eastern Śḥerɛ̄t (and Mehreyyet) š.̣ Here we focus on s̃, with some discussion of its emphatic counterpart s̃,̣ while the z͂ we leave for future work. All these sounds are highly variant, and appear to be dependent on several phonological contexts (segmental context, prosodic context, and phrasal context) as well as the sociolinguistic context of the speaker. Thus, the reports of the z͂ allophones need detailed, careful testing in a large number of phonological contexts with a range of speakers. Having established this focus, it is also worth quoting in full here two comments that Johnstone presents in jl (1981:xiv), since this underlies what has been said about Śḥerɛ̄t s̃ in the 35 years since. Firstly, he comments that he has found that “The C[entral] dialect of Jibbāli has two phonemes: š and s̃, where E[astern] has only one, namely š.” Secondly, he describes the articulation of s̃ as follows: The C s̃ is pronounced with approximately the same tongue position as š but there is no contact between the top [sic] of the tongue and the alveolum. The air is pushed out over the tongue and the lips are simultaneously rounded and pouted. It is a tribute to Johnstone’s contribution to the field of msa studies that these observations underlie what is still generally believed today about Śḥerɛ̄t s̃. We note that it is his jl description that seems to have taken hold, and that there is some small difference in his original description of the articulation of s̃ as regards the blade of the tongue and the palate, which is actually rather hard to effect, and which he has apparently amended slightly in the jl. Nevertheless, the two points that one may draw from the above are that (1) s̃ is only found in Central Śḥerɛ̄t, and (2) s̃ is a labialised palato-alveolar, and as such produced like š with lip-rounding, but with less tongue–alveolum contact. The force of his observations concerning s̃ that we quoted above—and as summarised in (1) and (2)—may be seen in subsequent reiterations of these two points. For instance, details given by Simeone-Senelle (2011:1077) are that the “Central dialect of Jibbali … has a phoneme /s̃/ (labialized š) contrasting with /š/”, while “the Eastern dialect … has only /š/”. Rubin (2014:26) provides the following: “The phonemes s̃ and š are distinguished only among some speakers of cj. Otherwise, both are pronounced as š”, and he also gives a description of the sound s̃ by quoting verbatim (p. 26) from Johnstone’s jl (1981:xiv) the passage we cited above.32 Similarly, Lonnet & Simeone-Senelle (1997:364) state that 32

It is noteworthy that Rubin has evidently not confirmed the s̃ ~ š contrast amongst the

630

bellem and watson

Central but not Eastern Śḥerɛ̄t displays strongly labialised sibilants s̃, s̃,̣ z͂ (“le jibbali du centre, non celui de l’est, présente des sifflantes particulières … fortement labialisées”), and they also quote (in translation) Johnstone’s jl description of the articulation of s̃ given above. Dufour (2016:24) describes /š/ as closer to [ɕ] than [ʃ],33 and states that “Le /s̃/ en diffère principalement par la labialisation, bien que la zone de contact entre la langue et le palais y paraisse plus réduite que pour /š/”. In this current article, we aim to advance knowledge of the Śḥerɛ̄t sibilant system by challenging assertions (1) and (2) above. To this end, in the following section of this paper, we discuss the results of our investigation into the sibilants of Śḥerɛ̄t, with a focus on s̃, and we add to Johnstone’s discoveries by showing how s̃ is in fact found in the speech of (at least some) Eastern Śḥerɛ̄t speakers,34 and that its production is not similar to the hushing (chuintante) sibilant š, but rather seems to be closer to the hissing (sifflante) sibilant s.

3

Analysis of Śḥerɛ̄t Voiceless Sibilants

In this section, we discuss the results of an instrumental phonetic study of the Śḥerɛ̄t voiceless sibilants, in which we investigated the articulatory angle through an electropalatographic study, in addition to visual evidence in the form of video; we also conducted some acoustic analysis. 3.1 An epg Analysis of Śḥerɛ̄t Sibilants Electropalatography (epg) is a way of showing oral articulation by recording articulator contact. The speaker is fitted with a false palate made of acrylic resin, into which are embedded 62 silver contacts,35 arranged in rows to cover the major zones of the palate, as shown on the left in Figure 1, below. As the palate-wearer speaks, each electrode records contact, which is sampled by the software at up to 100 frames per second, and subsequently shows as black-

33 34 35

native speakers whom he consulted. We observe that his informants are all heritage speakers of Śḥerɛ̄t and younger generation, which is a point that we return to in section 4 on the sociolinguistic situation. We also note that Rubin (2014:26) is quite correct in his understated observation that s̃ contrasts with š for only some speakers of Central Śḥerɛ̄t, which is confirmed by our own data, as discussed below. Which we show below to be incorrect. A finding also made recently by Dufour (2016). This is the most common type of palate used in Europe, the Reading Palate (see Wrench 2007 for discussion).

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

figure 1

631

A Reading palate (l), shown next to one frame of a palatogram (r) generated by ArticulateAssistant during j001’s articulation of s̃ of Śḥerɛ̄t /s̃ehey/ ‘tea’ 36

shaded squares on the resulting palatograms, as can be seen on the right in Figure 1. Looking across the palatograms generated for a recorded speech string allows the analyser to see real-time articulation via the recorded pattern of palate contact. With the help of Barry Heselwood, the authors conducted electropalatographic work in the phonetics laboratory at the University of Leeds between October 2012 and October 2015 with four male msa speakers aged in their 20s– 40s. The participants were m001 (a monolingual Mehri speaker from Rabkut), j003 (a bilingual Mehri / Śḥerɛ̄t speaker from Gabgabt and Salalah37), j001 (a bilingual Śḥerɛ̄t / Mehri speaker from Gabgabt and Salalah), and j043 (a monolingual Śḥerɛ̄t speaker from Sudh). In the current article, we discuss the implications of analysis of relevant data from the three Śḥerɛ̄t speakers, j003, j001 and j043. The data were recorded via a Reading palate (j001 and j003), and an Articulate palate (j043) using ArticulateAssistant, with accompanying wav files recorded simultaneously onto a pc via Audacity. Laryngography was also conducted in some cases, but this is not discussed here as it is not relevant to the current paper.

36 37

This frame shows asymmetrical tongue contact with the sides of the upper palate from rows 2–8, i.e. as far forward as alveolar. Place of birth and normal place of habitation.

632

bellem and watson

From the palatographic data recorded with the three Śḥerɛ̄t speakers, we looked at the following sounds for the purpose of the current study: 34 tokens of emphatic s̃,̣ 47 of s̃, 51 of š and 32 of s (a total of 164 tokens, in 151 individual palatograms, sampled at a rate of 100 per second). The phonological environment of the tokens varied in terms of position and syllable stress, as well as the segmental context (adjacent consonants and vowels). The only clear phonological environment that may have had an effect was that where the relevant consonant was followed by a rounded vowel, although this more obviously affected lip protrusion (visible on the videos and audibly perceptible) than the tongue gestures measured by palatography. One interesting result to emerge from a visual assessment of the palatograms was that the three speakers with whom we have been able to conduct palatography varied, apparently along the lines previously known, i.e. that the two speakers of Central Śḥerɛ̄t had a distinct s̃, which contrasted with s and š, while the one speaker of Eastern Śḥerɛ̄t who participated in palatographic experiments did not have a clear š ~ s̃ contrast. We will show in the following sections, however, that what appears to be dialectal variation is actually not the case; this demonstrates the problem with relying on data from too few speakers, particularly when working on under-studied or minority languages for which there is comparatively little data available. As noted above, one of the three epg participants—j043—is an Eastern Śḥerɛ̄t speaker. This speaker was the last of the three to participate in this work (in spring 2015); we examined 103 tokens of the target segments (s̃,̣ s̃, š, s) from 97 palatograms. The palatograms were not always easy to interpret, as they showed considerable variation in contact for the various consonants we looked at, and in some cases little discernible pattern was evident. However, one pattern was that in final position these four sibilants showed a tendency to lenite strongly; intervocalically there was also some tendency towards lenition, i.e. the tongue made far less contact with the palate, in many cases only the side rows towards the back (the palatal to velar region). The second main observation was that the wide variation we could see in the articulation of these consonants occurred amongst the tokens of individual consonants; in this speaker’s data, we did not find clear contact patterns that differentiated ṣ̃ and s̃ from š, and s was not as clear as expected, either, although for s there was evidence of some alveolar approximation with a wide (contactless) channel running through the oral cavity, while s̃,̣ s̃ and š were more likely to have some contact around the side in the palatal region. We did not find evidence in this data of this speaker having a contrast between s̃ and š. The earlier epg experiments with speakers j001 and j003 were far more conclusive, and the sibilants much clearer on the palatograms. Both of these

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

633

figure 2

Palatogram showing contact patterns during j003’s production of s̃es, extracted from /s̃es lo/ ‘not with her’ (left) and šeš, extracted from /okšeš lo/ ‘not fat’ (right)

figure 3

Palatogram showing contact patterns during j003’s production of s̃es, extracted from a token of the phrase /s̃es lo/ ‘not with her’ (a different token from fig. 2 above)

participants are speakers of Central Śḥerɛ̄t. For j003—our first epg subject— we found 18 tokens of the target segments (s̃,̣ s̃, š, s) from 11 palatograms. On visual assessment of the palatograms, while there was some variation evident, we could nevertheless see that generally s̃ had a similar oral contact pattern to s, with contact as far forward as the alveolar region, while š had a clearly different pattern with no tongue contact further forward than the palatal region. This is demonstrated in figures 2–3, above. The contact pattern for ṣ̃ was again generally distinct, with a tendency to have the narrowest (contactless) channel running from back to front of the mouth, and therefore a relatively greater area of tongue–palate contact than the other three sibilants, and with little to no contact in the velar and post-palatal regions, but contact towards the front (the alveolar region). The data from j003 are interesting for two reasons. One is that it bears out the reports of a contrast between s̃ and š in Central Śḥerɛ̄t; the other is that it indicates that s̃ is produced with an oral gesture that in terms of contact with the upper palate looks similar to s, and not to š. This should be considered in

634

figure 4

bellem and watson

Individual frames of cumulative epg contact data showing maximal constriction achieved during j001’s production of s̃ in the word /s̃ehey/ ‘tea’ (left), s in the word /derhis/ ‘young female goat’ (centre), and š in the word /šibʕet/ ‘seven’ (right)

the context of Johnstone’s description of s̃ as being like a labialised š with “no contact between the top [sic] of the tongue and the alveolum”. Lastly, the data from the third epg participant—j001—are particularly revealing. We examined 43 tokens of the target segments (s̃,̣ s̃, š, s) from 43 palatograms for this speaker, and these gave us the clearest data overall. Firstly, this speaker had a clear contrast between s̃ and š. The s̃ was produced with contact in the alveolar region, similarly to s, except that the area of contact was greater for s̃, particularly in the palatal region, and sometimes there was more narrowing of the channel. This means that for this speaker, s̃ is being produced with air forced through a narrow gap in the alveolar to post-alveolar region, compared with s, for which air flows through a (sometimes) less narrow gap in the alveolar zone of the oral cavity. This is demonstrated in figure 4, below. On the other hand, š is very distinct and clear for this speaker: this had the least tongue contact with the upper palate, and showed a wide (contactless) channel going through the velar and palatal region, and typically no contact around the alveolar region. This is to be seen on the right-most palatogram in figure 4. The data for j001 bear out the data of the other Central Śḥerɛ̄t speaker, j003, in showing a clear s̃ ~ š contrast, which is very marked in the articulatory pattern. It is also interesting that there appears to be some (much slighter) difference in contact pattern between s̃ and s, since although they both evidence alveolar contact, s̃ is sometimes produced with a narrower channel to force the air through, and greater alveo-palatal to palatal contact. The narrower channel may partially account for its perception as sounding ‘whistled’, as we note below in discussing the acoustic evidence.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

635

The epg data support our belief that, in articulatory terms, s̃ is better characterised as alveo-palatal (where š is palato-alveolar and s alveolar). That is, the greater degree of tongue contact in the palatal and alveolar regions evident in the production of s̃ in comparison with s is indicative of a consonant which is not strictly alveolar (as s), yet is clearly not palatal. Having examined the tongue articulation evidence, we move on in section 3.2 to visual evidence. 3.2 Video Evidence of Śḥerɛ̄t Sibilants It is very clear visually in Śḥerɛ̄t speech that for some speakers at least part of the oral tract is configured differently for s̃ than for both s and š. Even in fast connected speech, the listener will be aware of lip-spreading versus lip-protrusion affecting consonants and surrounding vowels, and that in the case of consonants, lip protrusion is not just evident during the production of those which are primarily labial or labio-velar. For s̃ there is lip protrusion that sometimes has a slight off-glide effect into the following vowel, particularly where this latter is a front vowel, as in the word s̃ighīm ‘to come in the morning’, for instance. By contrast, s very noticeably lacks this lip protrusion. We examined seven video recordings recorded individually with three speakers of Śḥerɛ̄t; two videos of j108 (Eastern Śḥerɛ̄t), recorded in a recording studio at the University of Leeds (October 2015), one video of m026 (a bilingual speaker of Mehri and Central Śḥerɛ̄t), recorded outdoors in Frankfurt, Germany (September 2014), and four videos of j014 (Central Śḥerɛ̄t), recorded in central Dhofar, Oman (2013–2014). Using Movie Maker, we examined individual frames of the recordings so that we could see the lip and jaw movements. It was clear that for these three speakers, the lip position was different for s̃ than for s and š, as described above. These data were interesting because one of the participants, j108, is a speaker of Eastern Śḥerɛ̄t, which was previously thought not to have a s̃ ~ š contrast.38 The video evidence showed incontrovertibly that this contrast is a feature of the speech of some Eastern Śḥerɛ̄t speakers.39 In our work, we had discovered by chance that speaker j108 had this contrast when talking with j043 about vocabulary in front of j108; j043 seemed unaware of a difference (which is borne out in the epg analysis discussed above, in which we could find no difference in tongue–palate contact for his s̃ ~ š); when j108 38

39

We were introduced to this speaker by chance when he came to the uk for a period of study. The time it takes to have a palate made up prevented us on that occasion from performing epg experiments. We intend to do this in the near future. This supports Dufour’s (2016) findings that the presence or absence of the s̃ ~ š contrast is sociolinguistic rather than geographic.

636

bellem and watson

figure 5

Three video stills showing lip position during articulation of: (1) s in /sīda/ ‘straight’ (j108); (2) s̃ in /s̃erbaʕ/ ‘cross!’ (j108); (3) s̃ in /s̃aʕgul/ ‘he hurries’ (m026)

joined the conversation, we noticed that he had a distinct s̃ ~ š contrast, which further questioning revealed that he was fully aware of. Analysis of the video recordings revealed that the lip shape for š seems the least stable, in that there is a tendency towards coalescence, such that the lip shape fairly often varies according to adjacent segments. The neutral state, as evidenced in recordings of j108 in the recording suite at the University of Leeds of š in the context of a, or elicited alone, is lip spread. With other speakers—natural, narrative speech—we happened to have no clear instances of š with a. In natural, normal-speed speech for all speakers we looked at, in producing š preceding a front (unrounded) vowel, the lips are most often spread wide; however, where š follows or precedes a back (rounded) vowel, there is a noticeable tendency for the lips to assume the non-spread, protruding position of the adjacent vowel. This contrasts with s̃ and s, which seem more stable in this respect, in particular s̃, which was always produced with protruding, non-spread lips which often took the shape of a brief ‘pout’ particularly of the lower lip. As noted above, where this s̃ preceded a front vowel, the effect was almost of a pouted palatal-sounding glide, i.e. s̃ɥighīm ‘to come in the morning’. Lastly, s in all contexts was generally produced with spread lips, without the lower lip pout. In figures 5 and 6 below, we present video frames showing different lip shape for s̃ in comparison with s and š. Figure 5 compares (1) s (left) with (2) s̃ (centre) for speaker j108 for Eastern Śḥerɛ̄t,40 and with (3) s̃ (right) for speaker m026 for Central Śḥerɛ̄t.41 These demonstrate the s̃ ~ š contrast that j108 has, although a speaker of Eastern Śḥerɛ̄t. Figure 6 compares (1) s (left) with (2) š (centre) and (3) s̃ (right) for speaker j014, Central Śḥerɛ̄t.42 These demonstrate the lip-spread position of s and—in this context—š, and the lip-pout position of s̃. 40 41 42

The recording was filmed at the University of Leeds, during an experiment that required the participant to discuss directions from one place to another. The recording of m026 was filmed in Frankfurt in September 2014. The frames were extracted from video recordings filmed in central Dhofar by Watson in February and November 2014.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

figure 6

637

Three video stills showing lip position of speaker j014 during articulation of: (1) s in /yiʕūrɛ̄s/ ‘they build it f. up’; (2) š in /bšaʕak/ ‘you m.s. have chopped it up’; (3) s̃ in /yis͂ir̍ ek/ ‘he does’

To conclude this section, we argue that the term ‘rounding’ is not appropriate for Śḥerɛ̄t s̃. It is lip protrusion in the form of a pouting gesture which is more evident than actual rounding, and the contrast of this lip position with the lip spreading evident with s and š. It is interesting also to note that when native speakers of Śḥerɛ̄t describe s̃ they typically do not point out the lip protrusion. This could perhaps be compared with English š, which is typically produced with lip protrusion, although not generally included as a ‘rounded’ consonant; this is markedly different from š in languages where it is not produced with such lip protrusion, but with spread lips. Therefore, it may not be that lip protrusion is as relevant to the identity of s̃ as lip spreading is to the identity of s and š. 3.3 An Acoustic Analysis of Śḥerɛ̄t Sibilants Finally, we move on to a brief outline of the results of some acoustic assessment we made of s, s̃ and š. This is a preliminary assessment of the data, and forms part of an ongoing study. The laboratory-recorded data focusing on sibilants were mostly not usable for acoustic analysis because of the interference of the palate, so for acoustic analysis we looked at recordings of natural speech data. From various audio sound files (in wav format), we extracted 20 tokens of s̃, 7 of s and 7 of š across seven different speakers of both Central and Eastern Śḥerɛ̄t. A brief outline of the initial assessment we made of the data showed that across the speakers, and given the nature of naturally recorded speech, overall, s and š looked as expected: the ‘hushing’ sibilant š has a more widespread aperiodic energy range which is less concentrated and goes much lower that s, with energy as low as c. 2000Hz and up to the 5000Hz bar. In certain positions (e.g. intervocalically), and where there is evident voicing, formants may be visible for at least part of the duration. In comparison, s is quite distinct, with concentrated aperiodic energy towards the top of the spectrum, particularly around the 4000Hz zone.

638

bellem and watson

This gave us a point of comparison to start evaluating acoustic data for s̃. This consonant varies widely both by speaker and by token. In natural, normal-speed speech, and especially intervocalically, it is easily voiced and the relatively short duration makes it hard to measure. Investigation of the acoustics data is further complicated because some speakers43 do not appear to have a s̃ ~ š contrast (as we noted above for, e.g., j043), and some speakers who do have a contrast seem to vary actual production, at times producing a sibilant which has most of its acoustic energy in the higher range of the spectrum, more like the ‘hissing’ sibilant s, and at other times having a more spread-out energy range, at both high and mid zones of the spectrum, more like the ‘hushing’ sibilant š. This is not the case for all speakers, for instance speakers j00544 and j04145 consistently produced a s̃ that clearly contrasted with š and consistently produced it with an acoustic pattern more like that of s (bearing out the auditory impression that it was a ‘whistled’ or ‘hissing’ sibilant, not ‘hushing’). It is worth noting here the evidence of Śḥerɛ̄t speakers themselves, who describe s̃ as having a whistling sound; this demonstrates the auditory impression created by an acoustically higher frequency than is usual with š. Lastly, our assessment of the acoustic data led us to the further conclusion that a pouting lip position may also contribute to the lowering of the frequency of the s̃. We recall here that the acoustic pattern of š includes a much lower and more widely distributed frequency than that of s. Indeed, the lowering of the frequency of s̃ through lip-pouting may well be one reason for its merger with š for some speakers, and its variability for other speakers. To conclude this section, we sum up our findings that s̃ is not simply a rounded version of š, and that the s̃ ~ š contrast is a feature for some Eastern Śḥerɛ̄t speakers (contra the previous literature), while also not being a contrastive feature in the speech of all Central Śḥerɛ̄t speakers.

43 44 45

This applies to speakers of both Central and Eastern Śḥerɛ̄t. A female speaker of Central / Western Śḥerɛ̄t, recorded in Dhofar by Miranda Morris in 2013. A male speaker of Eastern Śḥerɛ̄t, recorded by Miranda Morris at various times since the early 1980s. The recordings include the one we analysed, which was produced in the laboratory at St Andrews in 1982.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

4

639

Sociolinguistic Situation

We have shown in this article that s̃ is indeed also a feature of the speech of (at least some) Śḥerɛ̄t speakers from the eastern part of the language area and that it is also variable in the speech of Central Śḥerɛ̄t speakers. However, there is considerable pressure for s̃ to neutralise with š, and this may be at least partly generational—that is, very generally, there may be more older speakers who have this contrast than younger speakers. For instance, while staying with Śḥerɛ̄t-speaking families during fieldwork in Dhofar, Watson was told by adult speakers that their mothers used to chastise them for using š in place of ‘the whistled’ s̃, and would instruct them to use the s̃ appropriately. Since gradual neutralisation causes inter-speaker variation, this is probably why Johnstone thought that s̃ was a feature of only the Central Śḥerɛ̄t consonant system. It is possible that s̃ had been retained to a greater degree in certain speech areas, although as we have shown in this paper, this includes Central and Eastern Śḥerɛ̄t areas. Further, we have noted quite some variation between speakers in terms of whether there is actually a s̃ ~ š contrast, and in those speakers who do have this contrast, we have noted variation in actual production of s̃. This explains why Johnstone may have found this contrast with only certain of his speakers (i.e. the Central Śḥerɛ̄t informants), and why the reports had not been challenged before now. Given the inter- and intra-speaker variability, we can see that s̃ is a rather unstable consonant. Whether this will eventually lead to a complete merger of s̃ with š, or whether s̃ is strongly enough tied to Śḥerɛ̄t identity—e.g. in the face of bilingualism with both Mehri and Arabic—is not evident. Therefore, we conclude this section by observing that one possible outcome of such instability as is evidenced in intra- and inter-speaker variation is loss of contrast (in this case merger of s̃ with š), but that it is equally possible that in the context of increased awareness of Śḥerɛ̄t cultural heritage and a move toward stronger identity politics, that if s̃ is sensed by speakers as an indicator of Śḥerɛ̄t identity, then it may remain distinct. We do not speculate further on the likelihood of either of these two alternatives here.

5

Conclusion

We have discussed in this paper the discovery of a contrast between s̃ and š for some speakers—but not all—in Eastern as well as Central Śḥerɛ̄t. We have also identified that the s̃ has a range of allophones, some of which are dependent on phonological context, and which have the effect of being for many speakers

640

bellem and watson

more s-like than š-like, both perceptually and in terms of tongue–palate configuration. We have shown that the lip shape of s̃ is very different from that of s, while the lip-shape of š is more variant. We have also shown that tongue contact is greater in the alveo-palatal region for s̃ than s, and we posit that ‘alveo-palatal’ is therefore a more accurate articulatory term than ‘alveolar’ or ‘palato-alveolar’. The discovery of a s̃ ~ š contrast for some speakers of Eastern Śḥerɛ̄t goes against what has been reported in the literature on Śḥerɛ̄t since Johnstone’s initial observations over 30 years ago. We conclude that the reason for his initial discovery of s̃ being confined to his Central Śḥerɛ̄t informants is likely to be the considerable intra- and inter-speaker variability. This variability necessitates investigation of a large amount of data from a range of speakers. These discoveries are rather exciting, and go some way to showing the enormous amount of work yet to be done on the phonology, phonetics and sociophonetics of the msa languages. To that end, in this paper, we also discussed the preliminary assessment of acoustic material focused on sibilants. We have begun to systematically analyse more of the acoustic data, and have identified that we also need to record more acoustic data in controlled settings from a range of speakers of the various dialects of Śḥerɛ̄t, continuing with obtaining elicited material focusing on sibilants in different phonological environments, and which can be analysed acoustically. Under the auspices of the current project (described in section 1, above), we have obtained a good amount of data, which we have begun to analyse, and we have identified what more is needed Clearly, there needs to be further experimental work carried out with a range of Eastern Śḥerɛ̄t speakers. There is definitely a s̃ ~ š contrast for some speakers of Eastern Śḥerɛ̄t, not just Central, but we have not yet obtained palatographic evidence of this, because of time and funding constraints.46 We are working towards rectifying this in the near future, funding permitted, to provide further evidence of the existence of this segment in Eastern Śḥerɛ̄t. We found a distinct s̃ ~ š contrast for speaker j108, but were not in a position to do epg work with him while he was here, and would need to obtain funds to bring him back to the uk. epg work cannot be done in the field. The other possibility would be to do traditional palatography in the field, and since the costs and time commitment of epg work make epg investigations so difficult, this is another possibility to explore. We also discussed the value of visual evidence, which can be captured on video, and which we demonstrated with

46

For palatography, funding would need to be found to cover the cost of bringing Śḥerɛ̄t speakers to the uk, where epg facilities are available (as, e.g., the University of Leeds), and for individual palates to be specially made to fit the oral cavity of individual speakers.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

641

comparable video stills presented in figures 5 and 6, above. The videos were filmed in a variety of contexts, most of which meant that the subject was at a distance from the camera, with the resulting films preserved with varying quality; additionally, almost all the videos were of either naturally spoken narrative or of a laboratory experiment focusing on set tasks, which meant that the phonological contexts were not controlled. The visual data are therefore an important initial study which indicate that future research should gather audio-visual data on sibilants specifically for the purpose of visual analysis of mouth / lip shape, with both recording conditions and linguistic context controlled. This will result in more controlled data which are visually clearer. Lastly, we note also that experimental work urgently needs to be carried out with speakers of Western Śḥerɛ̄t, for which there is almost no description in the literature.

References Beeston, A.F.L. 1962. A Descriptive Grammar of Epigraphic South Arabian. London: Luzac. Beeston, A.F.L. 1984. Sabaic Grammar. Manchester: Manchester University Press. Bergsträsser, Gotthelf. 1983 [1923]. Introduction to the Semitic Languages. Translated by P.T. Daniels. Winona Lake: Eisenbrauns. Bittner, M. 1909. Studien zur Laut- und Formenlehre der Mehri-Sprache in Südarabien. i. Zum Nomen im engeren Sinne. Wien: Hölder. Bittner, M. 1911. Studien zur Laut- und Formenlehre der Mehri-Sprache in Südarabien. ii. Zum Verbum. Wien: Hölder. Bittner, M. 1913. Studien zur Laut- und Formenlehre der Mehri-Sprache in Südarabien. iii. Zum Pronomen und zum Numerale. Wien: Hölder. Bittner, M. 1914. Studien zur Laut- und Formenlehre der Mehri-Sprache in Südarabien. iv. Zu den Partikeln. Wien: Hölder. Bravmann, M.M. 1977. Studies in Semitic Philology. Vol. 6 of Studies in Semitic Languages and Linguistics. Leiden: Brill. Brockelmann, Carl. 1961 [1926]. Grundriss der vergleichenden Grammatik der semitischen Sprache. Hildesheim: Georg Olms. Dufour, Julien. 2016. “Recherches sur le verbe sudarabique moderne.” Unpublished habilitation thesis. Ecole pratique des hautes études, Paris. Ehret, Christopher. 1995. Reconstructing Proto-Afroasiatic (Proto-Afrasian): Vowels, Tone, Consonants, and Vocabulary. Berkeley: University of California Press. Faber, Alice. 1997. “Genetic Subgrouping of the Semitic Languages.” In The Semitic Languages, edited by R. Hetzron, 3–15. London & New York: Routledge.

642

bellem and watson

Goldenberg, Gideon. 2013. Semitic Languages: Features, Structures, Relations, Processes. Oxford: Oxford University Press. Gragg, Gene. 1997. “Old South Arabian Phonology.” In Phonologies of Asia and Africa (Including the Caucasus), edited by A.S. Kaye. Vol. 1, 161–168. Winona Lake: Eisenbrauns. Hetzron, Robert. 1974. “La division des langues sémitiques.” In Actes du premier congrès international de linguistique sémitique et chamito-sémitique, Paris 16–19 juillet 1969, edited by A. Caquot & D. Cohen, 181–194. The Hague & Paris: Mouton. Hetzron, Robert. 1976. “Two Principles of Genetic Reconstruction.” Lingua 38: 89–108. Hetzron, Robert. 1997. The Semitic Languages. London & New York: Routledge. Huehnergard, John. 1992. “Languages of the Ancient Near East.” In The Anchor Bible Dictionary, edited by D.N. Freedman. Vol. 4, 155–170. Huehnergard, John. 1995. “Semitic Languages.” In Civilizations of the Ancient Near East, edited by J.M. Sasson, 2117–2134. New York: Simon & Schuster Macmillan. Huehnergard, John. 2017. “Arabic in Its Semitic Context.” In Arabic in Context: Celebrating 400 Years of Arabic at Leiden, edited by A. Al-Jallad. Leiden: Brill. 3–34. Huehnergard, John & Aaron Rubin. 2011. “Phyla and Waves: Models of Classification.” In The Semitic Languages: An International Handbook, edited by S. Weninger in collaboration with G. Khan, M. Streck & J.C.E. Watson, 259–278. Berlin: Walter de Gruyter. Jahn, Alfred. 1905. Grammatik der Mehri-Sprache in Südarabien. Wien: Hölder Johnstone, T.M. 1975a. “The Modern South Arabian languages.” Afroasiatic Linguistics 1: 93–121. Johnstone, T.M. 1975b. “Contrasting Articulations in the Modern South Arabian Languages.” In Hamito–Semitica: Proceedings of a Colloquium Held at the School of Oriental and African Studies, University of London, on the 18th, 19th and 20th of March 1970, edited by J. & Th. Bynon, 155–159. The Hague & Paris: Mouton. Johnstone, T.M. 1977. Ḥarsūsi Lexicon and English–Ḥarsūsi Word List. London: Oxford University Press. Johnstone, T.M. 1981. Jibbāli Lexicon. Oxford: Oxford University Press. Johnstone, T.M. 1984. “New Sibilant Phonemes in the Modern South Arabian Languages of Dhofar.” In Current Progress in Afro-Asiatic Linguistics. Papers of the Third International Hamito–Semitic Congress, edited by J. Bynon, 389–390. Amsterdam & Philadelphia: Benjamins. Johnstone, T.M. 1987. Mehri Lexicon and English–Mehri Word-list, with Index of the English Definitions in the Jibbāli Lexicon. Compiled by G.R. Smith. London: School of Oriental and African Studies. Kogan, Leonid. 2011. “Proto-Semitic Phonetics and Phonology.” In The Semitic Languages: An International Handbook, edited by S. Weninger in collaboration with G. Khan, M. Streck & J.C.E. Watson, 54–150. Berlin: Walter de Gruyter.

south arabian sibilants and the śḥerɛ̄ t s̃ ~ š contrast

643

Leslau, Wolf. 1938. Lexique Soqotri (Sudarabique Moderne). Paris: C. Klincksieck. Lonnet, Antoine & Marie-Claude Simeone-Senelle. 1997. “La phonologie des langues sudarabiques modernes.” In Phonologies of Asia and Africa (Including the Caucasus), edited by A.S. Kaye. Vol. 1, 337–372. Winona Lake: Eisenbrauns. Macdonald, M.C.A. 2000. “Reflections on the Linguistic Map of Pre-Islamic Arabia.” Arabian Archaeology and Epigraphy 11: 29–79. Copenhagen. Reprinted in M.C.A. Macdonald, Literacy and Identity in Pre-Islamic Arabia. Variorum Collected Studies Series. (Farnham: Ashgate Publishing, 2009). Macdonald, M.C.A. 2004. “Ancient North Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R.D. Woodard, 488–533. Cambridge: Cambridge University Press. Morris, Miranda. 1983. “Some Preliminary Remarks on a Collection of Poems and Songs of the Baṭāḥirah.” Journal of Oman Studies 6(1): 129–144. Moscati, Sabatino, Anton Spitaler, Edward Ullendorff & Wolfram von Soden. 1964. An Introduction to the Comparative Grammar of the Semitic Languages: Phonology and Morphology. Wiesbaden: Otto Harrassowitz. Naumkin, Vitaly & V.Ja. Porkhomovskij. 1981. Ocerki po etnolingvistike sokotry. Moscow: Nauka. Naumkin, Vitaly & V.Ja. Porkhomovskij. 1995. “Sokotriiskie fol’klornye teksty.” In Khadramaut. Arkheologicheskie, etnograficheskie i istoriko-kul’turnye issledovanija. Moscow. 442–468. Naumkin, Vitaly, Leonid Kogan, Dmitry Cherkashin, Maria Bulakh, Ekaterina Vizirova, ʿĪsa Gumʿān al-Daarhi & Aḥmad ʿĪsa al-Daarhi. 2014. Corpus of Soqotri Oral Literature. Vol. 1. Leiden: Brill. Nebes, Norbert & Peter Stein. 2004. “Ancient South Arabian.” In The Cambridge Encyclopedia of the World’s Ancient Languages, edited by R.D. Woodard, 454–487. Cambridge: Cambridge University Press. Rodgers, Jonathan. 1991. “The Subgrouping of the South Semitic Languages.” In Semitic Studies in Honor of Wolf Leslau, edited by A.S. Kaye. Vol. 2, 1323–1336. Wiesbaden: Harrassowitz. Rubin, Aaron D. 2010a. The Mehri Language of Oman. Leiden: Brill. Rubin, Aaron D. 2010b. A Brief Introduction to the Semitic Languages. Piscataway: Gorgias Press. Rubin, Aaron D. 2014. The Jibbali (Shaḥri) Language of Oman: Grammar and texts. Studies in Semitic Languages and Linguistics, Vol. 72. Leiden & Boston: Brill. Sima, Alexander. 2009. Mehri-Texte aus der jemenitischen Sharqīyah: Transkribiert unter Mitwirkung von Askari Saʾd Hugayrān. Edited, annotated & introduced by J.C.E. Watson & W. Arnold. Wiesbaden: Harrassowitz. Simeone-Senelle, Marie-Claude. 1997. “The Modern South Arabian Languages.” In The Semitic Languages, edited by R. Hetzron, 378–423. London & New York: Routledge.

644

bellem and watson

Simeone-Senelle, Marie-Claude 2011. “Modern South Arabian”. In The Semitic Languages: An International Handbook, edited by S. Weninger in collaboration with G. Khan, M. Streck & J.C.E. Watson, 1073–1113. Berlin: Walter de Gruyter. Stein, Peter. 2011. “Ancient South Arabian.” In The Semitic Languages: An International Handbook, edited by S. Weninger in collaboration with G. Khan, M. Streck & J.C.E. Watson, 1042–1073. Berlin: Walter de Gruyter. Steiner, Richard C. 1977. The Case for Fricative-Laterals in Proto-Semitic. American Oriental Series vol. 59. New Haven: American Oriental Society. Stroomer, Harry. 1999. Mehri Texts from Oman. Based on the Field Notes of T.M. Johnstone. Wiesbaden: Harrassowitz. Stroomer, Harry. 2004. Ḥarsūsi Texts from Oman. Based on the Field Notes of T.M. Johnstone. Wiesbaden: Harrassowitz. Swiggers, P. 1981. “A Phonological Analysis of the Ḥarsūsi Consonants.” Arabica 28: 358– 361. Ullendorff, Edward. 1970. “Comparative Semitics”. In Current Trends in Linguistics, edited by T. Sebeok. Vol. 6 of Linguistics in South West Asia and North Africa, 261– 273. The Hague: Mouton. Wagner, Ewald. 1953. Syntax der Mehri-Sprache unter Berücksichtigung auch der anderen Neusüdarabischen Sprachen. Berlin: Akademie-Verlag. Watson, Janet C.E. 2012. The Structure of Mehri. Wiesbaden: Harrassowitz. Wrench, Alan A. 2007. “Advances in epg Palate Design.” Advances in Speech–Language Pathology 9(1): 3–12.

chapter 32

Was There a “Bedouinisation” of Arabia? Probably Not, at Least in the Way It Has So Far Been Portrayed William Lancaster and Fidelity Lancaster

Introduction Michael Macdonald (2015) demolished the Caskel, Dostal, and Bulliett theory of the ‘Bedouinisation of Arabia’ using material from epigraphy and classical literature sources, and pointed out that Noth and others had questioned Caskel’s argument at the time. The ‘bedouinisation of Arabia’, according to Caskel (1953), occurred between c. 100ad when Arabia was peaceful, dominated by settled states, with some non-tribal nomads who were ‘shepherds near the towns’, and 6th–7th c. Arabia when ‘the bedouin form of society and ideology prevailed’. Macdonald demonstrated that Caskel’s evidence for this false dichotomy was based on errors, misunderstandings, and arguments without evidence. Dostal (1959) explained bedouinisation by defining bedouin as ‘camel herders accustomed to fighting as rider-warriors’, with the development of the šaddād camel saddle. Macdonald demonstrated there is no evidence that Arab nomads ever fought from camel back; camels were ridden to the battle field, where the men dismounted and fought on foot or from horseback. Bulliett (1975) took Dostal’s theory further, claiming that the use of the šaddād saddle produced an alteration in the balance of power in favour of nomads, which gave apparent support to Caskel’s theory that North Arabian bedouin were able to sweep to military and political domination of sedentary populations in the 6th century. Macdonald shows that this theory has no foundations and argues that what is known about nomadic life between ad100–500 suggests continuities in its structures and in its relations with sedentaries. Macdonald’s arguments from epigraphy and classical resources seem exemplary, being out of our spheres of competence, but his discussion of tribe, bedouin, and settled populations in trying to establish what the structures and processes of Arab populations were between the 1st and 6th centuries ad is unsatisfactory. Clear and unambiguous evidence for that period is obviously in short supply, and we are not competent to examine it adequately. What we are able to present is a paper with two purposes. The first focuses on ideas of

© koninklijke brill nv, leiden, 2018 | doi: 10.1163/97890043

646

lancaster

Arab tribes expressed by nomadic and settled tribespeople across the Arabian peninsula to various ethnographers and anthropologists, and their relations with members of other social formations. The second looks at the question of bedouinisation in general and uses local ideas of what ‘being badū’ means. In general we use the ethnographic present in the text, material comes mostly from the 19th and 20th centuries, and we describe a past or recent past time before the transformation of technology and global economics and politics; this allows for descriptions of tribespeoples’ ideas on events of centralisation. We are not concerned with the model of tribe and the putative historical development of society,1 nor in ‘philosophical anthropology’;2 these may have been the fields with which Caskel, Dostal and Bulliett were concerned. We argue that there was probably not a ‘bedouinisation’ of Arabia. The construction of ‘tribe’ as a general concept and of tribes in particular from peninsula-wide moral premises of individual autonomy and jural equality (variously expressed) was and is more relevant. Ideas of tribe, tribal premises, structures, and practices appear to be very old, much earlier than the time scale of the 1st–6th centuries ad put forward by Caskel, and to be almost endemic, which is why we would not use ‘tribalisation’, while ‘tribalism’ is encumbered with other meanings. It also seems that the repertoires of irrigated agriculture, arable farming, pastoralism, and fishing, plus exchange, are of considerable historical depth,3 and to have developed almost in parallel. Whether tribes are settled or nomadic, large or small, the manner in which they make their livelihoods, or the nature of the physical environments they use, seems to make little essential difference to their underlying premises, structures or processes, regardless of the local particular terms in which these general concepts are described. This accounts for some settled tribespeople to identify themselves as badū, because “we behave properly”. The common basis is that Arab tribal society is a moral society, founded on the equality of all before God, expressed as jural equality and individual autonomy but using a range of terms to describe/identify structural groups and processual practices. Although tribes are constructed from sets of named groups, recruited differently and with different functions, these groups are not corporate, as all jural and political practice is vested in the individual person. This is the reason why Arab tribes are conceptually, and many tribespeople actually, opposed to centralisation. Tribe is therefore concerned, not with ‘power over’ and vertical hierarchies,

1 Kuper 1982. 2 Pouillon 1996: 57–60. 3 Betts and Cropper 2013: 183–191; Cleuziou et al. 2002: 9–27; Cleuziou and Tosi 2012: 313–323.

was there a “bedouinisation” of arabia?

647

but with ‘power to’ and horizontal orderings of closeness and distance. Ideas of ‘bedouin’ and ‘tribe’ as oppositional ‘others’ can be constructed by agents of centralised authorities who regard legal and political practice to be vested in central institutions and their agents.

Ideas of Tribe Arab thought is descriptive, bi-ʾl-waṣf, not determinative.4 Roots of words lend themselves to flexible and fluid interpretation. Arabic is a poetic language and speakers use it both lexically and poetically, so words and terms are often multilayered.5 Tribespeople use language in literary ways rather than necessarily literal, they use metaphors and metonymy. All the presentations of self or identity are in terms of conceptual oppositions, badū/ḥaḍar, badū/šwāyah, qabīlī/ḍaʿīf, and so on, are constructed by their users from different referents and are shorthand descriptions of observable and public behaviour; they lack validity as theoretical models.6 Macdonald7 states that the Yemeni tribal system with strong territorial elements as, for example, set out by Dresch8 is and was quite different to the badū tribal system, as for example, of Ruwalah (Rwala) of northern Arabia. The terms used are different, Dresch’s Ḥāshid and Bakīl tribes talked of themselves as qabīlī, tribal, whereas Ruwalah used badū; but the referents of autonomy, protection, and right behaviour were similar and underpinned by the same premises of moral autonomy. This paper explores the concepts that underpin tribal structures and processes from tribes across the Arabian peninsula, and notes the differences between generalised concepts and their operation by individual persons of a specific named group in particular situations. The tribes are:9 Ruwalah, camel herding bedouin of northern Arabia; Ḥāshid and Bakīl,

4 Watt 1973: 315. 5 Lancaster and Lancaster 2011: 36–42. 6 Badū and ḥaḍar is the obvious example, frequently used in the literature to contrast nomadic camel herding tribal social structures with those of ḥaḍar, centralised settled populations with urban elite structures, and often drawn on ʿaṣabiyyah, “tribal solidarity” but also “bond of power”, see Khalidi 1994: 222–231, discussion of ibn Khaldun’s monumental work of the principles of historical change. 7 Macdonald 2015: 53. 8 Dresch 1989: 75–83. 9 Burckhardt 1967 [1831]; Musil 1928; Lancaster 1981; Dresch 1989; Hartley 1961; Lancaster and Lancaster 1992; idem 1995 and idem 2011; al-Juhany 1983; Cole 1975.

648

lancaster

settled agricultural tribes of north-eastern Yemen; Nahid, a settled agricultural tribe in Wādī Ḥaḍramawt; small fishing and herding tribes in Jaʿlān, southeast Oman; small fishing, herding, and agricultural tribes of Raʾs al-Khaymah Emirate and Omani Musandam; the tribal towns and villages of pre-Wahhābī Najd; camel herding sections of the ʾĀl Murrah tribe of south-eastern Arabia; and ʾAhl al-Karak, small herding and agricultural tribes of the northern Karak plateau. What is common to all these social groupings is that they are tribal, jural and political power is vested in the individual person, they manage their own affairs, live from their own resources, share common moral premises, and are opposed to any form of centralisation and state. Local ideas of tribe necessarily take into account the natural environments and climates. Looking at a map of the Arabian peninsula and climatic information, it is clear that few regions have the conditions necessary for intensive agriculture—adequate and reliable rainfall and suitable soils—that would support permanent, dense urban societies. But even where intensive agriculture for grains or date and other tree crops is practised, waters are taken to terraced fields or built gardens. Water for agriculture is nearly always described in terms of surface flood flows and silts from rains or occasionally from snowmelt, or from underground waters, and both need to be channelled to where they are needed. Areas with reliable agricultural production are usually in or close to mountain or plateau and valley10 regions, which have greater and more reliable amounts of rains, have suitable soils from weathering, often have natural underground water storage, and provide a range of micro-climates and microenvironments in relatively short distances. Most of the peninsula provides, from a range of geological and hydrological factors, a wide variety of ecological regions. Few of these are seen by their inhabitants to provide livelihood over the seasons and across the years without movement between three ecological areas, using a range of herding strategies, sometimes opportunistic grain or date cultivation, or fishing on coasts, whose seas are rich in fish species that migrate seasonally according to salinities and temperatures. So opportunities for livelihood that present themselves would be based on animal management; cultivation; fishing; plus the crafts that these would need; plus distribution of surpluses between these resource systems; plus jural processes. A glance at a geo-political map shows the peninsula between India and Iran on the east, Africa to the west, the Mediteranean to the north-west, and central Asia to the north. Exchange systems between states in those regions needed to

10

Al-Juhany 1983: 43–63; Hartley 1961: 12–18; Wilkinson 1977: 36–51; Lancaster and Lancaster 1999a: 129–166; idem 2011: 99–102, 137–147, 191–208.

was there a “bedouinisation” of arabia?

figure 1

649

Map of Arabian Peninsula, showing tribe locations and places mentioned in the paper

cross—or circumnavigate—Arabia. For societies of the coastal regions of Arabia, and to a lesser extent for those in the interior, there was ‘an outside’, or ‘outsides’, with whom relations were or might be necessary. Some11 have considered this to lead into ideas on the need for a centralising political structure of such a region. People of all the tribes and places mentioned above each saw their own area and region as providing secure livelihood and profits over time, and as beautiful, an example of God’s generosity to his people. Listening to narratives with their emphasis on location and personnel, to descriptions of how to reach various destinations, to possible resource management strategies in different locations, and to explications for aspects of access to resources, all demonstrate the range of knowledge of ecological regions, in general and in particular, of the plains, mountains, valleys, wetlands and shores of the Arabian

11

Wilkinson 1977: 122–134; Dresch 1989: 158–167; Serjeant 1981a [1962]: 50.

650

lancaster

peninsula. Speakers often had both a wide experience of different environmental areas and very detailed knowledge of particular environments. The depth of time for known resource strategies in these environments was assumed by tribespeople to be considerable, almost ‘for ever’, ‘as we have no knowledge of anything else, until now’. Few areas, of themselves, were considered capable of providing secure livelihood and surpluses through the seasons and over the years. Environmental regions differ in their rocks and soils, their waters, rains, climate, altitudes, plants and possible crops, the seasons of growth and harvest, seasons in which different domestic animals give birth, and so on.12 Movement between resource areas was essential, either as movements of people and their animals, or as exchanges of particular foods and other necessary products between groups living from different resource strategies and/or from other areas. Movement of groups of people with their animals and of people wishing to make exchanges of foods and goods required accepted jural practice between participants—which in turn required known and accepted identities as jural persons. A further common feature was and is the willing acceptance of living from one’s own resources, a focus on self-reliance and self-sufficiency, of providing what one needs: the ʾĀl Murrah attitude is described as austerity13 which with endurance, justice and hospitality, are the ideals of desert life; the Ruwalah emphasised that camel-herding enabled the daily expression of the moral life14 and described regions as being good to live in.15 Herders, farmers, and fishermen alike constantly emphasised inter-dependence between natural environments and people, and the need to be satisfied with what one has.16 Archaeologists working in the north of the peninsula17 and in the south18 date recognisable resource management to the 6–7th millennium bc, with a growth in technological and political complexity, and participation in long distance exchange systems across the peninsula towards the end of the 4th.

12

13 14 15 16 17 18

E.g., Musil 1928: 676–684, the large Ruwalah vocabulary for topographical features; Lancaster and Lancaster 1999a: 131–142, local terms for natural and enhanced water storage systems in Bādiyat ash-Shām; Lancaster and Lancaster 1997: 367–378, use of bādiyah areas by different tribal groupings through the year and over years, and local knowledge of plant ecology; Cole 1975: 32–36, 41–43, 46–48, 50, 54–58, ecology. Cole 1975: 53. Lancaster and Lancaster 1990: 178–179. Idem 1999a: 105, 108–109. Idem 1999a: 111–114; idem 2011: 1–2, 10–12, 36–44. Garrard et al. 1996. Cleuziou 2002; Cleuziou and Tosi 2012.

was there a “bedouinisation” of arabia?

651

Tribespeople considered that in the distant past, the peninsula was inhabited by people like themselves, there being no knowledge of any other form of social organisation. Archaeologists used tribe, tribal alliance to describe the social formations of the past because the terms were consistent with what appeared to have existed then and because local populations had no terms that indicated any other former socio-political identity. In the 3rd millennium bc, nomadic groups were always identified by tribe, there were scattered references to tribes within long-established urban communities, and patrilineal descent was the norm in Mesopotamian society.19 Strong tribal links across regions enabled and maintained the strong cultural continuity that existed in the ancient Near East.20 Hoyland21 described Arab tribal societies from the Bronze Age to the coming of Islam, using pre-Islamic poetry to inform comments from Classical authors and from inscriptions a thousand years earlier. His description of the social, jural, and political relations of Arab tribes has significant similarities to Arab tribes now. Such cultural continuity over time may be linked to the cultural solidarity across the peninsula noted by Serjeant22 in his comments on ṭāġūt, the moral imperatives or absolute givens of tribal life from Kuwait to Ḥaḍramawt and Yemen. Doumani’s23 research in Jabal Nāblus, Palestine, extends this idea to present “an alternative framework to centralisation: unity through cultural solidarity and local identification, not through political hegemony”. This is not to say that all Arab tribes are the same, each particular tribe develops its ideas of tribe and tribal behaviour from general moral premises or imperatives. Nor are tribes as particular manifestations unchanging over time,24 since tribes may be subject to the effects of external historical and geographical events while tribal reactions to such events would be consistent with the general moral premises.

19 20 21 22 23 24

Postgate 1992: 82–85, 92. Postgate 1994: 7. Hoyland 2001: 117–123. Serjeant 1981a [1962]: 41. Doumani 1995: 19–20. Dresch 1989: settled Ḥāshid and Bakīl tribes are geographically fixed, men and families need not be. Lancaster and Lancaster 2011: 469–470, from Shiḥūḥ tribesmen, incomers ‘become Shiḥūḥ’ by behaving properly; the tribe continues but not entirely by descent.

652

lancaster

Words for ‘Tribe’ Qabīlah and ʿašīrah describe ‘tribe’ as a socio-political group among other similar groups. The root qbl is “to receive hospitably, to accept” and the root ʿšr (3rd form) is “to associate closely with someone”. Wehr25 translates qabīlah as “tribe” and qabīl as “guarantor”, “surety”, and both words have been associated by some tribesmen, probably because they derive from the same root, in explicating qabīlah; Lane26 as “a body of men from one father and mother”, a descent group. Wehr27 translates ʿašīrah as “clan, kin, folk, tribe”; Lane as “a man’s kinsfolk, from the same father or ancestor, a synonym for qabīlah”. These synonymous terms linking descent, real and assumed, with moral or jural relations, are used for tribal confederation, tribe and major sections of a tribe across the peninsula. ʾĀl is also heard. Ruwalah often use ʾĀl Ruwalah when talking about themselves as a group, but use qabīlah when talking about Ruwalah as one tribe among others; Murrah is always ʾĀl Murrah28 as are all its named sections and sub-sections. For the meanings of ʾāl, Lane29 gives “relations and followers”; Wehr, “family, kindred, companions, partisans, people”; there is kinship and so descent but also affiliation and alliance, which accurately describes the composition of such groups. ʾAhl according to Wehr30 means “relatives, family, kin”; Lane, “the people of a house or dwelling, and of a town or village, and of a country”, and describes a family living together under one roof, groups of families who have come together and consider themselves as a tribe, and groups of tribes living in the same area and bound together by alliances. Thus, ʾAhl Raʾs al-Khaymah—the families of Raʾs al-Khaymah town who consider themselves to act as a tribe, and ʾAhālī al-Karak, the tribal groups living in the Karak region and linked by alliances. The Majālī, the leading family of northern Karak, said “we came as a family, ʾahl, to Karak but over time we have become a tribe, we are now ʾĀl Majālī”. The people of Tibnah, a village in northern Jordan, identify themselves as ʾAhālī Tibnah, “families of Tibnah”, or “ʿašāʾir Tibnah”, tribes of Tibnah, but never as ‘a tribe’, as they are an association of three descent groups.31 In the uae, a Shiḥūḥ commented that ʾAhl Salḥad and ʾAhl Rawḍah, 25 26 27 28 29 30 31

Wehr 1974: 739, 741. Lane 1984 [1863]: vol. 2, 2984. Wehr 1974: 614; Lane 1984 [1863]: vol. 1, 2053. Cole 1975: 23. Lane 1984 [1863]; Wehr 1974. Wehr 1974: 33; Lane 1984 [1863]: vol. 1, 121. Antoun 1972: 36 ff.

was there a “bedouinisation” of arabia?

653

sections of Banī Hadiyyah Shiḥūḥ, were unrelated descent groups that had come together and behaved as tribes, and so were tribes.32 Ideas already associated with ‘tribe’ involve groups of individuals that come together using descent (through males), affiliation and alliance; moral relations which lead to jural and political identities expressed through relations between such persons; histories; locations; ‘right’ behaviours; and group identities. Tribal identities also have referents of speech, dress, the way of sitting in the saddle, of walking, etc. Speaking of groups as kin was general Semitic practice.33 It did not describe personal descent but personal relations, those who shared closeness of interests, similarities of purposes and practice, and an example of kafāʾah, “like to like”,34 or “closeness and distance”, a peninsula-wide referent for many aspects of tribal social practice. Musil35 quoted Nūrī ibn Hazzaʾ al-Shaʿlān, šayẖ of the Ruwalah, in 1914, describing relations between the European powers of ww1 to visiting ʿAmarat šayẖs in tribal terms, ‘tribe’ being paralleled to named political and military units in sets of alliance and closeness or enmity and distance. In the 1970s, his grandson, Nūrī ibn Fawwāz al-Shaʿlān explicated alliances between Arabs, French and English in terms of the closeness between sections of the same tribe, and opposed by a different tribe of Germans.36 These explanations are recognised as insufficient, but allow conversations of two differently constructed political systems. Unlike European states, Arab tribes are not corporate, solidary or legal bodies, but aggregations of sets of ordered names, whose members are free, autonomous, individuals, which require moral balance between individuals and groupings expressed as moral relations. This moral balance is expressed in terms of a balance of honour; each has honour equal to that of any other, and honour must be defended. Honour defended equates with individual autonomy. 32 33 34

35

36

Lancaster and Lancaster 2011: 15–16. Dresch 1989: 111, n. 4, quoting Smith 1885: 14–16, and Beeston 1972: 257. Kafāʾah, kifāyah; Serjeant 1981b [1977]: 227, 240, 244 sees this as concerned only with the status of descent, expressed through permissible marriage partners, as the basis for the stratification of southern Arabian society. Ruwalah and others regard kifāyah as shared or similar interests or attitudes, Lancaster 1981: 47. Musil 1927: 441–443, The ʿAmarat šayẖs, like the Ruwalah, had considered all Europe as France and its people as French, but the ʿAmarat knew the English had occupied Baṣrah. So why were all Europeans called French if they did not belong to the French tribe? Nūrī drew a parallel with the ibn Saʿūd ruling a large territory and various tribes, and the French having under their sway other tribes like Germans, Austrians, English, Russians, and so on, some of whom were allies with the French and others in revolt. Lancaster 1981: 28–29.

654

lancaster

Infringements of honour create situations between individuals and groups of shame on one side and insult on the other which upset the balance, the peace, between them. Moral relations are broken and should be redressed. These situations must be redressed by compensation or restitution for damage to person or goods together with amends for the insult so that the balance is restored and honour defended. The equation of honour with morality, right behaviour, recalls the concern of tribes in the Jāhiliyyah with abstention from sin, and so a commitment to right behaviour.37 Tribes do not gather mechanically around men at dispute; men move constantly between and inside groupings using choices from within tribal practice. Arab tribal systems rely on fragmented answerability, with personal jural responsibility, and their customary law is transactional.38 Hartley regarded Arab tribes as essentially jural units while Burckhardt considered tribal political institutions as “natural” but their civic institutions to “indicate the work of a legislator”.39 The obligation to give protection to all who ask—and the ability to do so—might be seen as the single attribute that characterises Arab tribes, protection being the providing of defence and restitution for those unable or unwilling to provide their own, while the ability to defend tribal property and assets is a key factor in the description of tribe across the peninsula.

Tribe and Identity Dresch40 comments that western readers assume power and identity are equated “as if, for example, state and nation were the same, or as if a collective identity is only what it does. The present status of a social unit would be defined by a singular past, distinct and quite separate from that of similar units”. Tribes do not see themselves in this way, but define themselves against their equals, as are sections against sections, men against men. Dresch talks of “power” but in our experience tribespeople spoke of political “reputation” when describ-

37 38

39 40

Kister 1980a [1965]: 136; idem 1980b [1968]: 159. Dresch 2012b: 152. Dresch prefers “legalism” to describe the organising concepts of customary law based on local ideas; “legalism, at its simplest, is a discussion of moral order” (Dresch 2012b: 12). Ḥāshid and Bakīl tribes have documents as records of cases and as codes of customary law (Dresch 1989: 116), and he observed šayẖs as arbitrators and judges. Ruwalah (Musil 1928: 426–435), and ʾĀl Murrah (Cole 1975: 124–125), had no documents, but had judges and judicial procedures: Lancaster 1981: 73, 77, 95. Hartley 1961: 4; Burckhardt 1967 [1831]: 378–379. Dresch 1989: 354–355.

was there a “bedouinisation” of arabia?

655

ing tribes in relation to each other. Thus political reputation spreads in terms of tribal identity, which continues when the details of the events that led to an increase in reputation are no longer known. “Tribes are not overrun or destroyed by others” while the identity of men as tribesmen comes from their stable, named structural units alongside their actions with their fellows, and these actions are made sense of in terms grounded in the named units of ibn ʿamm, sections, and tribe. What is relevant to this paper is his statement that “What inequalities there may be between men have so far been registered as part of the history of states and Imams, not of tribal identity, and disorder has not been directly tied to collective self-definition”.

Tribes with Honour/Reputation, and Groups Without The tribes of Ḥāshid and Bakīl of northern Yemen, like Ruwalah of northern Arabia, use the term “ʾaṣl” as a defining part of their identity. For Ḥāshid and Bakīl, the term means “authenticity of descent”, and so honour, which distinguishes them from non-tribal, weak people.41 Ruwalah link the term to knowing their tribal genealogical links to the ʿAnazah confederation, a tribe old before the coming of Islam, and which explicitly refers to their defence of their assets and resources, giving protection to all who ask, and so having reputation.42 These tribes are large and well known. What about small tribes, or tribes without genealogical referents? Banī Ṣakhr in Jordan ignored genealogy and spoke of reputation gained from giving protection and defence of lands. The small tribes making up the ʾAhl al-Karak in the central Jordanian plateau talked of behaving rightly (being generous, giving protection, defending), having rooted customs (ʾuṣūl, pl. of ʾaṣl), and of defending and maintaining their assets of lands and waters. The settled Nahid of Ḥaḍramawt spoke of tribe in terms of political potential, a tribe was a group that could wage war, arrange truces, and conclude peace:43 but when the composition of tribal groups or the relationships between such groups was being discussed, kinship and descent came into the discussion, in which genealogical statements reflected the political order. ʾĀl Murrah of south-eastern Arabia are recognised as a noble tribe.44 Cole considered that while the tribe was an important military, political, social, and 41 42 43 44

Dresch 1989: 75. Lancaster 1981: 43–44, 119 uses “reputation”. Musil 1928: 471–472, uses “esteem” to convey the idea of the honour awarded by others to someone for their behaviour. Hartley 1961: 55–56 and echoed in Musil 1928: 506, 570 for Ruwalah. Cole 1975: 93–95.

656

lancaster

cultural unit, its base was economic, since the tribe was the unit that claimed sovereignty over its central pasturelands. Small tribes in Raʾs al-Khaymah and Shiḥūḥ and Ẓuhūriyyīn of Raʾs al-Khaymah and Omani Musandam considered genealogies as relatively unimportant, and emphasised other facets of ʾaṣl in the meanings of “original” and “to be or become firmly rooted”.45 They were ʾaṣl because their tribes had been the original developers of their lands and then because they behaved rightly in giving protection to those who asked, who then might “become firmly rooted”. “Behaving properly”, as a “good man” and the qualities required for this—generosity, courage, wisdom, prudence, foresight, being merciful and not necessarily demanding all one’s rights46—were apparent in all discussions of tribe. Tribes have lands from which their members live and which they defend. These lands may be described in terms of their waters: permanent, allowing for irrigated tree and crop agriculture; seasonal rains, capable of producing arable crops; or variable, unreliable rains that provide seasonal grazing. In these latter regions, few areas of themselves provided livelihood and profits throughout the year; tribes lived from animals and moved between three environmental areas.47 Some of these tribes describe their lands as dīrah, meaning “to make the rounds”. Ownership of resources is in terms of exclusive or preferential access to and control of land, developed48 and defined by tribal enterprise and politics, and rested in the persons of the users; Ruwalah say: “when we are here, this is our dīrah; when we go, our dīrah goes with us”. Dīrah49 is used variously. By Ruwalah, as the areas where they live and have owned or preferentially claimed assets, as the area of influence of a man of good reputation when present, and as a local administrative centre. Banī Ṣakhr use dīrah for the region where the tribe owns resources and is influential. ʾĀl Murrah use dīrah for the core area, “territories”, of the pastures they use in an arc northeast of Najrān, across the western and central areas of the Rubʿ al-Khālī, and north to the oasis of al-Aḥsāʾ, the lands they control through military conquest

45 46 47

48

49

Lane 1984 [1863]: vol. 1, 64; Lancaster and Lancaster 2011: 469–470. Musil 1928: 471–473. The need for movement by tribes recognised by classical Arab geographers: Hiyari 1975: 510; Lancaster and Lancaster 1999a: 97–128, 131–142, 182–185; Cole 1975: 39–53; Lancaster and Lancaster 2011: 31–32, 50–59, 93–94, 128–131, 135–136, 185–190, 235–241; idem 2002: 237– 244. In the Islamic tradition, ownership is concerned with the techniques of production, as the medium of production, land and water, comes from God, and only development that results in greater production confers ownership. Lancaster and Lancaster 1986; Bocco 1987; Cole 1975: 28–30, 46, 94.

was there a “bedouinisation” of arabia?

657

and displacement of other tribes. Tribespeople of Jaʿlān in south-east Oman used dīrah for the seasonally productive bases on the coasts and in the wadis between which members moved and defended in contrast to less productive areas used and shared with others.50 In Raʾs al-Khaymah Emirate and Omani Musandam, tribesmen living from arable farming, herding and fishing, in any combination, described the areas from which they made livelihood and profits as dīrah, whereas those who lived from irrigated agriculture and some herding used bilād,51 for their territorial areas, as do Ḥāshid and Bakīl in northern Yemen living from arable agriculture.52 Nahid lands in Ḥaḍramawt are not contiguous and so Nahid use “places”.53 Waṭan, “homeland” was used by a Saʿdiyyīn tribesman in Wādī ʿArabah in southern Jordan to describe the wells, gardens and grazing lands of the tribe, comparable to the descriptions of tribally owned oases by the classic Arab geographers. Ḥaram was used like dīrah by a Shiḥūḥ tribe in Wādī Ghalīlah, Raʾs al-Khaymah Emirate, for the area from the sea to the mountain top and the exclusive use of each faẖḏ, small section; and people pointed out that not all tribes had dīrah.54 Some tribes owned routes:55 tribal ownership of land along the Pilgrim route between Damascus and Mecca gave rights to ṣurrah payments and to act as carriers for the Pilgrimage. Ḥāshid and Bakīl, like other Yemeni tribes, controlled passage on roads as did Omani tribes. Hutaym (Wehr: hutāmah, “s.th. broken off, fragment, breakage”) tribes56 of northern Arabia, Kuwait and al-Aḥsāʾ had no lands and paid for protection to use the lands of others. “They had their chiefs and their social organisation, they live in tents and breed camels … yet they [as a tribe] are not held in esteem” although they [as individuals] had honour. Doughty considered the Hutaym to be more prosperous than bedouin whose territories they used because they did not have to defend themselves and saw “many poor families of Bedouin tribes living (for their more welfare) in the peaceable society of the Hutaym”. Hutaym live among bedouin tribes as qaṣārah, glossed by Musil as “neighbours”.57 This

50 51 52 53 54 55 56 57

Lancaster and Lancaster 1992: 149–150. Bilād, Lane 1984 [1863]: vol. 1, 246–247, “remained, stayed, dwelt”, “a country, land, district”. Dresch 1989: 81–82. Hartley 1961: 31. Lancaster and Lancaster 2011: 28–29. Wallin 1854: 123–124; Doughty 1936: 117; Dresch 1989: 187; Wilkinson 1987: 120–122. Musil 1928: 136; Dickson 1949: 110–112; Doughty 1936: 212–214. Musil 1928: 136, 267. Wehr 1974: 767, “to be inadequate”. Dresch 1989: 59–60, 118, jār as neighbour, protégé, in the context of protection taker and protection-giver. Dresch 2012a: 148, “the asymmetry of protégé and protector is basic”, as is “the distinction between compensation and amends”.

658

lancaster

crucial difference—the group needs to take protection—removed members of these tribes from “being like” tribes with honour/reputation, and for that reason, marriages did not take place between members of tribes with honour/reputation and those without. Hutaym identity described groups who, although like tribes, were not fully tribal in their political behaviour and did not defend themselves. Among settled tribes in the south of the peninsula and in the towns of Jabal Shammar, al-Qaṣīm, Najd, Oman, Ḥaḍramawt, and Yemen, where men of known tribal descent were the rulers, such groups were muzayyin or ḍaʿīf, lacking in authenticity of descent58 and so honour. Dresch59 noted that many low status people were richer than many tribesmen, that a sayyid family had earned their livings as tailors for three generations, and that individual traders had become tribesmen. Bujra60 examined mobility between the various groups of al-Ḥurayḍ society, which was possible in economic occupation, but rarely authenticated by marriages; examples were Mashāyikh, descendants of Ḥaḍramī saints and scholars, who had become farmers; a farming group tried to become tribal; farmers had become craftsmen or small traders; a tribesman had become a servant. In Raʾs al-Khaymah and Oman, there were bayādir, cultivators, who worked in the date gardens. Lorimer61 described them as a low social class, but local people said that being a baydar (sing.) was a perfectly honourable occupation; all it meant was that the man was without land or enough land in that year.62 He might well have land of his own in other places and employ a baydar to work it. A baydar looked after the dates and got a share, negotiated once the size of the harvest could be seen; he cultivated the owner’s vegetable garden, which supplied the owner’s household, and did what he liked with the rest— supplied his own family, sold them, whatever; he was fed, given a new set of clothes at ʿĪd, and a small wage. He could also work bits of unused land, parts of another date garden, and in the end save enough to buy land. A baydar had responsibilities, unlike a ẖādim who was paid a wage only. While working, the landowner protected his baydar, who often took the tribal name of his employer.

58

59 60 61 62

Dresch 1989: 119, 131–132, some weak groups in Khamir, a tribal town, claimed they were of tribal descent. Bujra 1971: 13, 41, lacking in known descent, tribal or religious, and lacking in capital. Dresch 1989: 134, 138, 152. Bujra 1971: 104–114. Lorimer 1908–1915: vol. 2, 1825; Lancaster and Lancaster 2011: 110–116. In his home area, poor rains, a shortage of land, income for marriage, compensation, buying land, etc.

was there a “bedouinisation” of arabia?

659

Honour, Group and Individual Honour is a complicated concept, constructed culturally; e.g. honour in Mediterranean societies is not the same as in Arab societies. Within Arab tribal societies, tribes differ in the words used for honour[able behaviour], while the behaviours so described and the purposes of actors are clearly associated though separated. Tribes are “like” but not the same. Ḥāshid and Bakīl of northern Yemen use several words to describe sorts of honour: šaraf, “honour as displayed to others”,63 the shared “honour”, common honour, is expressed in the idiom of shared male ancestry as the sons of a father and the men of a tribe; šaraf leads to ʿirḍ, “honour defended” of individuals and tribes; a “good name”, naqāʾ, also applies to individuals and to tribes, and has a jural connotation as the restoration of a good name was achieved by payment or by blood revenge; wajh, “face” or “honour as committed in specific instances” is used only of individuals, not of tribes.64 Ruwalah, as described by Musil,65 used naqāʾ, “good name” in two ways: one applied to individual men in providing the motivation for honourable behaviour and avoiding dishonour and “his face being blackened”; the other applied to tribes, and in the instance given, after peace had been made between tribes but one wants to continue fighting; if it attacked, it would “blacken its face” so it declares that it returns the good name or wipes it off, and may then attack the other without any injury to its honour. ʿIzz66 describes honour as “good qualities”, proper behaviour, bravery, prudence, generosity, etc. inherited through women and exemplified by men.67 Wajh is closely associated with the possession or loss of “a good name”, personal honour, by the practice or omission of behavioural obligations by an individual,68 where “face” can be

63 64 65 66

67

68

Dresch 1989: 439. Dresch 1989: 440, 79. Musil 1928: 451, 505–506. al-ʿizz fī ʾl-qanāʿah wa-l-ḏill fī ʾl-ṭamāʿah, the only inscription seen in Musandam dated 1132 ah/ad 1719–1920; “honour is in contentment (or moderation, frugality) and shame is in greed” (see Lancaster and Lancaster 2004b: pl. 6 on p. 136). Shiḥūḥ tribesmen said this described how they had lived in the mountains before modernity, Lancaster and Lancaster 2011: 169. Musil 1928: 547, men with little honour described as ḏalīl, glossed as a man who prefers his health and life above his honour; Musil 1928: 472, rajājīl, men who occupy themselves with women’s work; rijjarijjah, a coward who babbles like a woman. Musil 1928: 451–452, a wronged person blackened the face of his oppressor by tying a black rag to a stick or spear, and cried out in the camp, or rode round the tents of his kin,

660

lancaster

blackened.69 Wajh is also the personal honour that all have as individuals, whether they are men or women, Ruwalah, Hutaym, (former) slaves,70 free servants, or smiths.71 Lancaster72 noted honour to belong to all members of groups present at Ruwalah encampments as individuals, and described other aspects of honour in terms of personal and group reputation, šaraf, which Ruwalah presented in terms of right behaviour by autonomous (because honourable) individuals, which then accrued to named groups and tribe. Tribespeople gave examples of individual and group honourable and dishonourable behaviour, some of which Musil had included in his 1928 Manners and Customs of the Rwala Bedouin (in the 1970s this book was not known to Ruwalah) and others more recent or contemporary (but not included in Lancaster 1981 for political reasons). The settled and agricultural Ḥāshid and Bakīl of northern Yemen and camel herding Ruwalah of northern Saudi Arabia and southern Syria hold similar ideas about aspects of honour, which are also held by ʾaṣl tribes across the peninsula.73 Having honour, which all do, implies a man’s honour can be challenged, lost, or blackened; challenges to a man’s honour insult or shame, ʿayb him, and honour must be restored. A tribesman’s šaraf is broken by ʿayb.74 Nahid in Ḥaḍramawt75 used ʿayb in a technical sense for relations of war or peace between members of the alliance-making ʾāl or qabīlah group, as well as

69 70

71 72 73

74 75

shouting that he has been insulted by his protector’s failure to protect him, not guarding him effectually, not getting him adequate compensation, or being disregarded as a guest by him. Only after the wrongs have been righted, and amends paid for the insult to his honour, will the wronged person change the black cloth for a white one, and make a circuit of the camp, crying out that a reconciliation has been made. Dresch 1989: 95 described jiḏn; blackened stakes are planted by a wronged person against persons they claim have betrayed a formal obligation. Musil 1928: 277–278, the former social category of slaves had families, ʾahl, took vengeance among themselves and against tribesmen, and could give protection as his master could; Musil 1928: 82–83, the family a slave served were his ʿamām, his uncles, father’s brothers. Musil 1928: 438–440, wajh, face, also a category of protection where a man grants protection to a person regardless of whether he is present or not. Lancaster 1981: 12–13, 43–45, 73–76. Musil 1928: 238–239, sexual honour seems to be different, certainly between married couples; wives of men who were away for long periods and who entertained men in his absence or who took a new partner were not considered shameless by others; if her father and brothers did not mind, her husband or former husband should accept the situation with good grace, as they were the only persons the matter concerned. Dresch 1989: 39–41 quoting Serjeant 1981b [1977]: 228. Hartley 1961: 180–181.

was there a “bedouinisation” of arabia?

661

in the general sense of any individual’s shameful act. A tribesman can act in a way demeaning to himself, ʿayb, can have a ʿayb inflicted on him, or can inflict a ʿayb on others. ʿayb is blameworthy whether done or suffered, and the lexical equivalence between active and passive indicates a pervasive moral symmetry. A man cannot ‘have’ a ʿayb, and a tribe does not carry, in the eyes of others, any burden of specific faults to complement the noteworthy deeds people attribute to it. The faults of groups and persons have no hold on them, until someone else brings them up, at which point one is dealing again with a specific ʿayb, an insult that is here and now. Honour is nearly always a social given, whereas ʿayb is specific. Dresch’s discussion of honour and shame among Ḥāshid and Bakīl of northern Yemen is like that heard among Ruwalah in Jordan and northern Saudi Arabia, and among tribespeople in Raʾs al-Khaymah and Omani Musandam,76 and has a strong link to concepts and practice of jurality and of rule. This is where tribal honour, reputation as tribes that give protection, and keep the peace for all comes in—the weak groups, the Hutaym, or the muzayyin in Yemen and Ḥaḍramawt, cannot do this.

Terms for Groups within Tribe In the segmentary model of tribe, all segments are replicated in recruitment and function, but ethnographies show that this is not so in the Arabian peninsula. When an Arab tribesman describes tribe to an outsider, he often gives a series of increasingly inclusive named groups from himself, and sets these out horizontally. The horizontal placing is significant, indicating that power is fragmented among individuals, and that some are closer to him than others. Or he may describe tribe as a series of sets of ordered names more or less in balance, and in which the named groups of each structural set are recruited differently and have different functions. Individuals of the same tribe rarely describe these ordered sets in exactly the same way, especially for those names remote from them in time and space. An individual is born into his/her natal family and must know his/her descent77 in the male line for at least five generations, or more precisely the 76 77

Lancaster and Lancaster 2011: 298–301, 315, 321. Jabbur 1995: 264, Syrian tribes viewed with suspicion anyone ignorant of his genealogy, regarding such a person “as deficit in ambition and culture” (Jabbur’s work is in the Arab literary tradition). Altorki and Cole 1989: 23, of the three descent categories of ʿUnayzah, those of qabīlī tribal status were men and women of unquestionable tribal descent, mainly Banī Khālid, Subayʾ, or Banī Tamīm. Dresch 1989: 75, Ḥāshid and Bakīl of northern Yemen,

662

lancaster

five linkages described in the idiom of genealogy, between him/herself and the tribe. An individual knows the names of his forebears up to his greatgrandfather; they and their descendants make up the informal and unnamed working groups within which people live, work, marry and spend most of their lives. These groups are referred to as ʿāʾilah, ibn ʿamm, baṭn, bayt, dār, ʾahl, and doubtless others. The smallest named group and so the smallest identifiable group as a group is at the fifth generation, beyond actual named forebears and where the cohesion of blood descent is lost. The particular name of this group is not the actual fifth grandfather/ancestor but a name, and these groups are more or less stable over time.78 These named five-generation groups may be called ʾahl, farīj, farʿ, faẖḏ, bayt, dār, ibn ʿamm; or ẖamsah, the “five” [links liable for vengeance] or luzūm, the “obligated” [to take vengeance], since these five generation named groups include the group that protects an individual from injustice and suffers for his guilt.79 This vengeance/protection group is fixed in relation to the individual only, who is obligated to avenge or be avenged by those to whom he can make five links between himself and them. The ẖamsah/luzūm is thus restricted to a smaller number of men than the named five generation unit. Many tribes follow the rule of dropping down a generation when a man has a son, “following their son into the sixth generation”, which further limits the vengeance group. In addition, those on the edge of the vengeance group can move themselves out of blood vengeance by giving a pledge as among Ḥāshid and Bakīl; “what is due”, Ruwalah; or “the camel of sleep” among Majālī, limiting the group of active vengeance seekers even more.80 The groupings of more inclusion are successively subsection or section, then section or tribe, then tribe or confederation. Subsections and sections are often

78

79

80

tribesmen had authenticity of descent, which gave them honour, which non-tribal, weak people lacked. Musil 1928 lists of Ruwalah five-generation groups, made between 1908 and 1916, differ little from Lancaster’s, collected in the 1970s (Lancaster 1981: 193–195); also Musil 1908: 41–43, lists for southern Jordanian tribes compiled in 1906 and checked by Lancaster in the 1990s. Cole 1975: 85–91, ʾĀl Murrah regard this group of c. 50 households, a faẖḏ, as a military defence group primarily, the vengeance group, and important in organising migrations. Dresch 1989: 85–88, Ḥāshid and Bakīl seem to construct vengeance groups in other, often ambiguous, ways. Hartley 1961: 76–77, Nahid bayt or dār, five generation groups territorially expressed as villages, were co-responsible groups for vengeance and political and economic activities. Dresch 2012a: 157, pledge; idem 1989: 66–68, 84–88, vengeance, feud affecting groups; Musil 1928: 489–503, feud and groups; Lancaster 1981: 37, 76; idem 1999a: 66, 381–382, the camel of sleep, vengeance and groups; idem 2011: 315, 318–319, 321, 330–331.

was there a “bedouinisation” of arabia?

663

described81 as fractions, e.g. sixths or eighths; “branches”, farʿ; “split off units”, farīj, farīq; thighs, fuẖuḏ; “tribe”, qabīlah.82 Sections and subsections have particular names, and are recruited from continuity over time, although there are examples of groups of families or parts of subsections of one tribe leaving their own tribe and becoming members of another.83 Subsection and section names describe increasing orders of aggregation of sets within tribe, and so order and balance groupings within tribe. Each of these sections and subsections is named, identification of groupings within tribe becomes possible, and if necessary, more precise identification of individuals from the unnamed small family units may be made. Confrontations between members of named subsections or sections or tribes can be recounted in narratives or poems only in terms of the named blocks. A man will say “we, the riders of the Ruwalah, raided the Fidʿān (another ʿAnazah tribe) …”. On questioning, there were ten or twelve men from two Ruwalah tribal sections ( fuẖuḏ or qabīlah), from different ibn ʿamm within these two sections. Ruwalah is the only term that encompasses them all, just as Fidʿān does for their opponents. Tribe and major tribal section, qabīlah and qabīlah/fuẖuḏ are political and military groups. There are no accounts of whole sections confronting whole sections. Sections, like tribes and like subsections, do not act as corporate groups.84 Tribal sections have histories, and are associated preferentially with known areas in the different environmental areas where they have their assets. ʾĀl Murrah clans (major tribal sections) had no active economic function but are more important as political units,85 defending resources, seeking vengeance, settling legal disputes, and gaining conces81

82

83

84 85

Dresch 1989: 75–79, fractions. Lancaster and Lancaster 1999a: 66, branches. Musil 1928: 47– 48, split off units. Lancaster 1981: 28–29, Lancaster and Lancaster 1999a: 65–66, idem 2011: 13–14, Cole 1975: 85–91, thighs. Hartley 1961: 57–61, Lancaster and Lancaster 1992: 348; idem 2011: 13, tribe/qabīlah. As Dresch 1989: 78 so truly says, “The vocabulary denoting sections and subsections varies from place to place, but there is no privileged level of organisation which stands out in all circumstances, nor any standard distinction of terminology between one level and the next”. Musil 1928: 632–633, some of the Dughmān section of the Ruwalah joined a Shammar section, Musil 1928: 52–56. Lancaster 1981: 155–156, the Kawatzbah, a section of Qaḥṭān, moved into Ruwalah territory and eventually joined the Ruwalah as a new section. Lancaster and Lancaster 2011: 15, in Raʾs al-Khaymah Emirate, part of a Shiḥūḥ tribe moved and became part of the Ḥabūs. Dresch 1989: 108–109 discusses individuals moving tribe. Cole 1975: 87–88, ʾĀl Murrah, the lineage, faẖḏ, of c. 50 households is the only unit above the household that regularly comes together as a single group. Cole 1975: 91–93.

664

lancaster

sions from government. Political and military coalitions were between clans although actual coalitions never followed strict theory as groups allied without others who should have been included. Most ʾĀl Murrah clans had families considered as amiral standing, and three had provided the leader of all ʾĀl Murrah. This is like the Ruwalah major sections. This sounds vague and unsatisfactory. But when tribesmen are describing the named sets of groups comprising tribe to an outsider, ethnographer, they are simplifying massively from sets of ideas about jural responsibilities and political actions constructed and reconstructed over generations to removing the actual behaviours and motivations of individuals and the wider families of which they are part into the sets of names for the increasing aggregations of groups comprising tribe. Tribes vary in size. Hartley noted Nahid in 1961 to number c. 6,000 souls, a “goodsized” tribe by South Arabian standards. ʾĀl Murrah were reckoned to be c. 15,000 people. Dresch estimated the median population of Ḥāshid and Bakīl tribes as between 20 to 30,000, and the whole mass of the people he was concerned with to be upwards of half a million. Tribes in coastal Jaʿlān, Oman, were really small, probably in hundreds not thousands; the many small tribes in Raʾs al-Khaymah Emirate and Omani Musandam are possibly in the high hundreds to low thousands. Ruwalah in the 1970s estimated they had possibly around a quarter of a million souls86 and other northern ʿAnazah and Shammar tribes were probably somewhat less. Some Jaʿlān tribes claimed they were parts of well known, much larger Omani tribes who had moved, usually for political reasons, and who for current politics looked to one of the two large regional tribes. Wilkinson in the 1970s estimated Omani tribal falaj settlements to have populations varying from c. 1,000, with a reasonably sized falaj and 40 hectares of cultivated land; larger settlements of 2,500, and multiple settlements with several falaj, up to 8,000 people. Small tribes of the Sharā mountains in southern Jordan said they were separate tribes for social and economic activities, but for politics joined the much larger Ḥuwaytāt. Small Karakī tribes formed a political association of two alliances and neutrals. Tribes in Raʾs al-Khaymah Emirate and Omani Musandam, when necessary, had fairly stable long term political alliances that looked towards rulers on the coasts. Although tribes speak of territories and describe their boundaries, their members know that in most cases they also ‘use’ areas they do not ‘own’, and that the ability to move to seasonal resources or to markets is essential.

86

It can be difficult to establish who is being counted—members by descent, or by political association.

was there a “bedouinisation” of arabia?

665

Tribespeople spend most of their time in economic and social activities as members of their wider domestic groups or households, which are not part of the named sets of orders of tribe. Jural and political activities are the concern of the named sets. But as these political and jural concerns are acted upon by members of wider households, there is a difference between a description of political and jural concepts and the ways in which individuals of wider domestic groups put these into operation:87 Dresch analyses the “indeterminacy” of collective action and responsibility of Ḥāshid and Bakīl tribesmen; Hartley has a clear account of Nahid practices in such situations, and that individual tribesmen did not always act as expected; Musil and Lancaster illustrate Ruwalah individual autonomy in rejecting jural decisions. The flexibility of terms used to describe orders of aggregation seems to depend on context and audience, and may be used by an individual to account for a particular behaviour (e.g. Nahid have, in increasing order, bayt, qabīlah/ʾāl, qabīlah/ʾāl, qabīlah; Ruwalah have ibn ʿamm, fuẖuḏ/ibn ʿamm, fukuḏ/qabīlah, qabīlah; ʾĀl Murrah have bayt, faẖḏ, qabīlah, qabīlah). There is a clear distinction between a general concept which is accepted by a tribe or group of tribes and its practical operation in reality when active participants are members of much smaller restricted groupings identified at the smallest named group, but in fact members of particular families within that named unit.

Other Terms Describing Groupings of Tribespeople, jamāʿah and qawm Neither word has connotations of descent or jurality, and so cannot compete with tribe. Each describes groups of individual tribesmen or tribespeople. Jamāʿah is described by Lane88 as “a state of union, agreement, friendship, companionship”. Musil89 noted “when Prince ibn Shaʿlān (of the Ruwalah) speaks of jamāʿatī he is thinking of the various tribes who obey his orders and hasten to his assistance […]. Jamāʿah is almost the same as qawm”. Here, the terms carry political implications, although ‘obey’ is not wholly supported in Musil’s accounts of events between 1908 and 1916.90 Qawm as manifested in the 1970s among 87 88 89 90

Dresch 1989: 83–88, Ḥāshid and Bakīl; Hartley 1961: 85–86, 174–188; Musil 1928: 444–446; Lancaster 1981: 76, 142. Lane 1984 [1863]: vol. 1, 457. Musil 1928: 50. See Musil 1927: 386, 398–399, 426–443, North Arabian politics in December 1914. Musil,

666

lancaster

Ruwalah was those tribespeople who actively participated with a tribal leader in his undertakings and political enterprises.91 Members of leading families, lodged in tribal sections, draw on the purposes of domestic units and ibn ʿamm groups, which are largely concerned with honour, livelihood, and social relations, to enable short-term political action using the support of individual tribesmen. Qawms are named after the leader who attracts supporters for his purpose, they are not named in terms of the tribe or tribal section of its leader. On the other hand, the small families making up a farīj in Raʾs al-Khaymah town described themselves as a qawm, and in Bū Baqarah village on the Omani Bāṭinah coast, the small families of a farīj identified each component qawm by its tribal core—qawm Zaʿab, qawm Balush, qawm Jabrī, and so on.92 Wilkinson93 described qawm as the war group of Omani tribes, which on its much smaller scale would apply to the farīj of Raʾs al-Khaymah town and Bū Baqarah village. We do not recall having heard the term jamāʿah with Ruwalah, but the small tribes94 around Raʾs al-Ḥadd in Jaʿlān, south-eastern Oman used the

91

92 93 94

an Austrian agent, urged his friends among ʿAnazah šayẖs, Banī Ṣakhr and Ḥuwaytāt to make peace among themselves ready for jihād against the English. Tribal leaders said that their statements of loyalty to the Caliph in Constantinople were public statements with no meaning. Tribesmen said tribal leaders could obey whoever they wanted, but they guided themselves by the law of the desert. They were uninterested in the war, complaining of the hypocrisy of the government and its depredations on their livelihoods. Lancaster 1981: 15–18, 91–94, in this case, qawm an-Nūrī, qawm al-ʾAnwar and qawm alʾAwrens (name derived from T.E. Lawrence), all in the eastern ḥamād, were smuggling (customary trading which had become illegal because of imposed borders and customs duties) into Syria in retaliation for the Syrian government seizing tribal assets; their encampments kept the peace in the area and policed the border. The leaders of each encampment were not on speaking terms, and arranged meetings were held on neutral territory with agreed supporters, while tribesmen visited each freely and continually; family difficulties between leaders were no concern of theirs. Jaussen 1948 [1908]: 146 presented qawm as a state of enmity between two tribes and as a raiding party. Doughty 1936: vol. 1, 388, 390, used the term similarly. Lancaster and Lancaster 2011: 34. Wilkinson 1987: 124. Lorimer 1908–1915: 1406 and Miles 1994 [1919]: 404 listed Banī ʿAmr, Mawālikh, Muhayyir, Banī Ghazzāl, Ḥikmān, and Ḥarb as the tribes of the area; each claimed to have “always” been there, and to be part of a tribe living in other areas of Oman. Lancaster and Lancaster 1992: 347–348, each tribal group was small, made up of a number of buyūt, and each locality based on a watering place was described as used by members of two or more tribes, each having parts of two or more buyūt present.

was there a “bedouinisation” of arabia?

667

term constantly; every description of the conflicting and ambiguous relations between the tribes living in the wadis and beaches ended “but we are all one jamāʿah, which is the people who live, work and marry here”. Jamāʿah was used in a similar way by members of small tribes in Raʾs al-Khaymah Emirate, as an idea that constructed networks between small tribes through marriage (and so inheritance, resource management practices and exchange), together with right behaviour of contributing to and sharing with a larger community that identified by tribe alone.95 Tribes of the Karak plateau in Jordan spoke of jamāʿah as a family’s informal networks, created partly by marriages. Large tribes seem not to require jamāʿah.

Marriage Arab tribesmen say they marry their bint ʿamm, their father’s brother’s daughter, a man’s paternal first cousin. Bint ʿamm is also a classificatory term which includes daughters of a man’s father’s brothers, his grandfather’s brothers, and great-grandfather’s brothers. No instance of actual paternal first cousin marriage was known among the Ruwalah96 who considered such a marriage to be unnecessary as these groups were already close. Most marriages take place within three and five generation groups, pulling some parts closer and distancing others. Other marriages maintain links to more distant groups into other tribal sections and to other tribes, and these marriage links are repeated over time; a few marriages create links to new groups. These patterns were similar to those of small tribes in Jaʿlān and to Banī ʿAmr of Karak. Other tribes of Karak, like Majālī and Banī Ḥamīdah, had a few actual first cousin marriages, as did some of the small tribes on the coastal plain of Raʾs al-Khaymah Emirate.97 Dresch98 found a small proportion of actual first cousin marriage

95 96

97

98

Lancaster and Lancaster 2011: 24–25. Musil 1928: 137, “a girl should marry her nearest young relative whom it is permissible to marry, ibn al-ʿamm; this is generally a son of her father’s cousin,” i.e. her second cousin, and who in any case has to give his consent to any marriage she should make. Lancaster 1981: 43–57. Tribes with no actual first cousin marriages seem to be those living from herds or fishing and sea trading, that is, their assets are moveable, Lancaster and Lancaster 1992: 357; those with some actual first cousin marriages live from agricultural land, and so concerned with more fixed assets—although this ignores other possibilities, Lancaster and Lancaster 2011: 17. Dresch 1989: 286–287.

668

lancaster

among Ḥāshid and Bakīl of northern Yemen, as did Hartley99 for the Nahid in Ḥaḍramawt, while most bayt have roughly equal numbers of close and distant marriages. Women, after marriage, continue to be jural members of their natal group,100 which is how women direct and/or consolidate the active descent groups on the ground. Kifāyah, like to like, is important; marriages are made between equals not only in personal or family reputation, but also in the economic and political interests of each wider domestic group. Many tribespeople pointed out how important this wide range of marriages was in reducing risk by enabling the continuation of or initiating a geographical spread of access to varied resources and building networks inside the tribe and across tribes, reaching across land and seascapes.101

The Wider Domestic Group A definition of the domestic group, or household that shares from a common purse, produces livelihood, arranges marriages, and defends itself, has 99 100

101

Hartley 1961: 156–168. Men say women have no jural responsibility as they are not involved in vengeance, but see Lancaster and Lancaster 1999a: 382; in giving protection, but see Musil 1928: 447, 637 and Lancaster 1981: 63; or in active defence, but see Musil 1927: 185; idem 1928: 525, 573, 622. These are the responsibilities of tribesmen identified by male descent, and therefore women are not in genealogies but see Lancaster 1981: 63. Women are thus set aside from active participation in political affairs, they are ḥarām and it is from women that honour is transmitted. Tribes differ in their attitudes to women. For many, e.g. ʿAnazah, ʾĀl Murrah, Shiḥūḥ, Ẓuhūriyyīn, and Mazāriʿ of Raʾs al-Khaymah, Wahībah in Oman, and others, women are “free daughters of free tribes”, Doughty 1936: vol. 1, 272; Musil 1928: 238. Women made decisions, controlled their fertility, Lancaster and Lancaster 2011: 162, 164, 187, and in these tribes were not given in compensation for vengeance, Jaussen 1948 [1908]: 222. Musil 1928: 52–54, the renunciation, accepted by the father, by his wife of their only surviving son, saved from an ambush. Women are valuable and necessary economic partners, Lancaster 1981: 62–68: Lancaster and Lancaster 2011: 164, 214, 218. A Banī ʿAmr of Karak enumerated his links by marriage of himself, siblings, his children, and wife within the wider family, to another family of his tribal subsection, to another subsection of the major section; to another tribe living in the village, to three other tribes of the neighbourhood, to three more distant tribes in the region; these links spread from Jabal al-ʿArab in southern Syria to the northern Ḥijāz in Saudi Arabia. In Jaʿlān, south-east Oman, marriage ties linked local groups, and each family also had one or two marriages to unconnected tribes outside the normal area of use; Bayt Muḥammad of the ʿAmr had a tie to the Qanbus from Qumaylah to the south-west and one to Banī Jābir in the mountains behind Ṣūr, Lancaster and Lancaster 1992: 357.

was there a “bedouinisation” of arabia?

669

occupied many. The smallest named group is the five generation ibn ʿamm which is more inclusive than five actual descendants (as above), while the group that co-operates in production and consumption talks of itself as from a common grandfather, but is not named, nor is its membership consistent. Attempting to establish the domestic group responsible for production, Fabietti,102 working with Shammar tribespeople of central Arabia, arrived at this description; “the Bedouin domestic group [is] … a dynamic unit towards which converge resources originating from a variety of sectors, procured and organised by mobile individuals belonging to a parental group whose dimensions and composition are not definable a priori”. This is consistent with what was observed and discussed with Ruwalah, and with such apparently diverse tribal and familial productive groups103 as farmers in Jordan and Raʾs al-Khaymah Emirate, tribally descended merchants and entrepreneurs of Sukhnah in Syria, and fishermen in Raʾs al-Khaymah Emirate and south-east Oman. All of these groups moved seasonally and practised multi-resource economics. Dresch104 uses “household” for what appears to be a less fluid productive group among Ḥāshid and Bakīl, self sufficient arable farmers with cattle, living permanently in villages, while Hartley105 found “no clearly discernible work or consuming unit” in the flood watered farming of the Nahid.

Multi-Resource Economics Tribespeople give themselves an alternative identity from the resources by which they live and make their profits, maʿāš wa fāʾidah. A tribesman describes his economic activities as those that provide his profits, while he usually lives from a wider and more disparate set of resources.106 Livelihood and profits

102 103

104 105 106

Fabietti 1990: 242. Farmers in Jordan; Antoun 1972: 20–25, 103–110; Lancaster and Lancaster 1995. In Raʾs alKhaymah Emirate, Lancaster and Lancaster 2011: 147–190, 209–238. Tribally descended merchants and entrepreneurs of Sukhnah, Syria, Métral and Métral 1991: 153–171. Fishermen, Lancaster and Lancaster 2011: 61–68, and idem 1992: 347–357. Dresch 1989: 301, 308. Hartley 1961: 48. E.g. Musil 1928: 348, noted camel herding Ruwalah made profits selling camels; ibidem: 336, employed as herders; ibidem: 278–280, worked for ʿUqayl (ʿAgeyl) merchants as camel drivers. Schilcher 1991: 56, worked as harvest guards and transported grain. Musil 1928: 510–511, raided. Lancaster 1981: 122, a few owned oasis gardens. Lancaster and Lancaster 1999a: 121–122, some ʾAhl al Karak tribes farmed commercially and had sheep and goat

670

lancaster

come from ownership of resources and their development. In tribal customary law, ownership is rather preferential access to and claims over natural or God given resources. Development of resources confers ownership, or alternatively what is owned is the means to production. The resources so developed, according to their attributes and qualities, confer different sorts of ownership and varying rights of disposal over the resource itself and what it produces— permanent, seasonal, animal, plant, foods for people or for animals, crafts, enabling services, for sale or for home consumption. Tribal livelihoods and profits come basically from resources owned by tribal members, and in general, all members of all tribes own, as tribespeople, at least some of the resources they need for livelihood and profits, and their access to other resources come from their tribal jural identity. The development by tribal members of the natural resources of lands and waters of the general areas associated with them is at particular localised points or nodes, where the natural conditions are most favourable. These developed assets are protected by those who developed them and their heirs (or purchasers) who maintain them, and become seasonal or permanent living areas, described variously by their owners in particular areas. These terms have lexical as well as particular meanings, and include village (qaryah), farīj,107 ẖānah,108 or quṣūr.109 Over time, such assets tend to become

107

108

109

flocks, others concentrated on profitable flocks but also farmed. Lancaster and Lancaster 2011: 151–184, mountain tribes’ profits from grain, goats, honey, firewood, clarified butter and cheeses and ibidem: 209–235, grain, honey, dates, fruit, firewood, charcoal, tobacco; ibidem 2011: 134, sands’ tribes, from animals, bulls for well work in date gardens, carried goods, and escorted families summering on the Bāṭinah coast. Lancaster and Lancaster 1992: 348–354, Ḥarb made profits from fishing and fish trading and had goat flocks, Banī ʿAmr lived from fishing and herding camels and cattle; each made profits for individual families in different years. Lancaster and Lancaster 2011: 30–33, as in Raʾs al-Khaymah Emirate and Omani Musandam, the people (of two or more tribal sections), their fields, gardens, cisterns, and stores in the mountains and valleys. Lancaster and Lancaster 2011: 34, in Raʾs al-Khaymah town and coastal villages, the mixed populations and buildings of a neighbourhood. Al-Naim 2004: 193–207, it describes the people and buildings of a town neighbourhood. Musil 1928: 156, an encampment of no more than ten tents and their camels. Doughty 1936: vol. 2, 250, “a nomad hamlet” in the region of Khaybar. Lancaster and Lancaster 1999a: 267, on the Karak plateau, the word encompasses the lands, rainfed and irrigated, buildings, cisterns, and the descendants in the male line from a common ancestor, the original developer. Musil 1928: 160, an oasis in northern Saudi Arabia; Lancaster and Lancaster 1999a: 177, alJawf, the lands, wells, buildings and people from a common ancestor. Qaṣr is “any stone building”.

was there a “bedouinisation” of arabia?

671

divided, exchanged, abandoned or extended by those who work them, while the jurally responsible group continues to have an interest and rights expressed as šufʿah,110 the right of pre-emption, where an owner wishing to sell property must offer it to his/her co-owners first or only to them. The exception is purely commercial enterprises, such as date gardens, outside tribal areas. Although people speak of livelihood and profits from owned land (and Dresch111 notes that Ḥāshid and Bakīl call their fields māl/ʾamwāl, “wealth”), herds, boats etc., wealth is also regarded to be ‘social’ and founded on tribal social relations which protect and provide security.

Other Social Groups with Tribes Tribes share landscapes with other social groups. In northern Yemen, Dresch112 describes these as “estates of society within the tribal peace”; two estates of non-tribesmen have rights and obligations within the tribal territory of Ḥāshid and Bakīl. One is the “weak”, the ḍuʿafāʾ, people; the second, the “learned”, the sayyids who claim descent from the Prophet, and the judges, the qādīs. The weak people lack authenticity and reputation because of their lack of tribal ancestry, and lived from craft work, small market trading, and small services provision. Markets, where the weak provide goods and services, were often protected by or guaranteed by a tribe or section of a tribe.113 A tribe makes itself responsible for the protection of a given branch of a qādī family, and usually protects the place where the qādī family lives; the qādī family is hijrah from this tribe. Sayyid families are also hijrah, with claims to protection from tribesmen and provide an optional source of justice for tribesmen. Hijrah is glossed by Dresch114 as “set aside”, protected status, and “inviolability”, and linked to ḥaram, “an inviolable space”.

110

111 112 113 114

Dresch 1989: 81, Ḥāshid and Bakīl villages; Dresch 1989: 112, n. 11, the rules vary between different ecological zones, e.g. Dresch 1989: 286, in some villages, village territory is exclusive although outsiders can own individual fields. Al-Naim 2004, a farīj consists of social relationships transferred to physical structures; inheritance manages expansion or re-use, while šufʿah requires an owner to offer his buildings to his neighbours so that outsiders cannot move in. In the mountains and valleys of Raʾs al-Khaymah Emirate, šufʿah was implicit, as it was in Nahid villages in Ḥaḍramawt (Hartley 1961: 49). Dresch 1989: 277. Ibidem: 117–157. Ibidem: 127–129. Ibidem: 435, 146.

672

lancaster

In Ḥaḍramawt, Nahid villages share a tributary wadi or part of the main wadi with villages of one or more other tribes; there is one Nahid town, and some live in other towns. Although their villages are spoken of as being of one bayt, men of other Nahid groups and of other tribes live there, as well as wives from other tribes, and non-tribal labourers.115 Ruwalah tribespeople of northern Arabia had ‘strangers in the camp’,116 protected as guests or had ‘brother’ relations with tribesmen who protected them; either as guests or through ‘brothers’ they were within the tribal peace. One category of ‘strangers in the camp’ but not of a different estate of society were tribesmen from other tribes who arranged with Ruwaylīs to live as neighbours, quṣārah, and to protect each other against their own people. Strangers of a different estate of society, and with whom therefore there were no marriages, and who lived alongside tribespeople in encampments were slaves117 and smiths.118 Traders who came to camps were from towns. Small traders, described as Kubaysat,119 came to small encampments in the desert where they became the guest of the senior man, and so protected, and traded with the tribespeople around who bought clothing, cloth, coffee, cardamom and drugs on credit, and paid in the summers when they were in the settled areas. When tribes were in the settled areas, big traders sent messengers to the šayẖ to ask permission for their agents to go out and trade; when permission was granted, the traders’ agents were under the šayẖ’s protection, and took camels in exchange for clothing, grain, rice, coffee and so on, household goods, and weapons.120 There were also ʿUqayl (ʿAgeyl),121 the large wholesale merchant families based in Damascus, Ṭāʾif, Baghdad, Baṣrah, Cairo and Bombay. ʿUqayl had agents, mostly from al-Qaṣīm, to whom they lent money to buy up camels from camel herding tribes 115 116 117

118

119 120 121

Hartley 1961: 30–36. Musil 1928: 267–282. Ibidem 1928: 276–278, slaves were not chattels; they could not be sold or killed and were free to choose a new master; every married slave had his own tent, camels and weapons from his master, to be returned if he left; and was entitled to a share of raided camels. Musil 1928: 281–282, smiths shod horses, repaired and made weapons and metal goods; had rights to a camel for every horse and saddle taken on a raid, and some money. Smiths claimed vengeance among themselves. They never fought in wars, raided, or defended themselves as each smith family had a ‘brother’ among the tribesmen with whom they lived. Smiths across Arabia were related to each other. Musil 1928: 269, from Kubaysah, a small town near Hīt on the mid-Euphrates, although many did not. Ibidem: 60, 90, 94, foodstuffs; household goods, tent cloth, Musil 1928: 61, 64, 69–70; clothing, ibidem: 118–125; arms, ibidem: 132–134. Musil 1928: 278–281.

was there a “bedouinisation” of arabia?

673

for export to Egypt, Baṣrah or Kuwait. An agent hired assistants and bought tents, coffee, rice, and often arms for sale or exchange, and went to a šayẖ of the tribe from whom he wanted to buy with letters of recommendation and presents. He was then allowed to put up his tents in that šayẖ’s encampment or was directed to another; the šayẖ was his host, and his protector (although he did not provide board). Bedouin brought in their camels, usually selling for cash, but exchanging for arms. The bought camels were branded with the ʿUqaylī mark, put into herds, and ultimately, driven to the nearest large town with a camel market. On their journey, these camels might be taken by raiders, so the ʿUqayl had ‘brothers’,122 individuals in each large tribal section to whom they made an annual payment, and who returned to them every camel taken by a raider of that tribal section. Traders together with protection of a tribal šayẖ ‘created’ markets. The various quarters of desert fringe villages paid ẖuwah, brotherhood payments, to competent individuals from the major tribal sections, ‘brothers’, who had to return all the property of the villagers that his fellow tribesmen had raided;123 that is, ẖuwah “was the payment of a negotiable sum of money or goods to opt out of the economy of raiding”. During the 1970s, at ar-Rīshah, a Ruwalah encampment in eastern Jordan, Syrian townsmen and villagers lived with Ruwalah, some employed as drivers or servants, while others were traders in the sūq or mechanics. A Banī Ḥarb tribesman from Saudi Arabia originally came for protection and then stayed as a neighbour, and worked as a driver. There were two groups of Fiwāʿirah and one of ʾAhl al-Jabal who had sought protection while arranging compensation. All were under the šayẖ’s protection, and so in the tribal peace. ʾĀl Murrah mix with ʾĀl ʿAjmān, Muṭayr, Subayʾ and Dawāsir tribes, sheep herding tribes, gipsy tribes, and groups of peasants and merchants in the winter pastures which are outside their tribal territories.124 They use the urban markets of Hufūf and Najrān for buying what they need,125 now using money paid to them from government employment or subsidies, but earlier as direct

122

123 124 125

The ʾaẖ, brother, relation between a Ruwalah notable and a man from the smiths, ʿUqayl, or a ẖuwah paying village quarter that provided a person-to-person channel of ‘like’ by constructed relation but ‘not like’ by reality of birth for communications and for resolution of disputes. Dresch 1989: 140, some sayyids among Ḥāshid and Bakīl gave up their hijrah status after 1961 and became ‘brothers’ to tribes, making themselves tribesmen: Dresch 1989: 328–334, ‘brotherhood’ also used to make alliances or change tribal identity. Musil 1928: 60; Lancaster 1981: 121–124. Cole 1975: 48. Ibidem: 106–108.

674

lancaster

exchanges for protecting markets. In their own territories, the only strangers were raiders from other tribes. The people of Raʾs al-Khaymah Emirate and the neighbouring areas of Omani Musandam and the Bāṭinah coast describe themselves as a tribal society, who on occasion have relations with social groups who are not of tribe. One group is the rulers, merchants, administrators, and Islamic jurists, and their families; members of this group are outside the tribal peace. As rulers are the arbitrators of last resort for the tribal peace, they underpin this peace.126 Rulers in the past, and others on the coasts, had slaves127 originally from East Africa and Baluchistan who worked as guards, soldiers, delivering water, in pearling, fishing, and in gardens. The presence of slaves implies profitable commercial enterprises outside tribal economies and the tribal peace. In the mountain areas, all the populations were tribespeople doing their own work and providing their own peace. A third non-tribal group were the metal-workers,128 described as Baluch, as gipsies, zuṭṭūṭ, or as smiths, ḥaddāds. Smiths lived at places on the coastal plains and made circuits around the living areas on the coastal plains and into the foothills, but not into the mountains. At each place they stopped to do work that was needed, and exchanged metal goods for animals, their hosts protected them as guests.

Ruling, Keeping the Peace Ruling is described by tribespeople across the peninsula as essentially about keeping the peace so that all may go about their business in making their livings. Conflicts over tribespeople’s ideas of rule and specific aspects of rule were frequently given to account for wars and/or the splitting of tribes.129 For Ruwalah, rule requires presence and the effecting of a purpose, “just as we speak of the rule of the stars in the appearance of rainfall”. The purpose of rule for Ruwalah šayẖs is tribal peace and so security and tribal livelihood. Ruwalah leaders and tribespeople focus their discussions on ḥukm, glossed as

126 127

128

129

Lancaster and Lancaster 2011: 298–312, 333–346. Slaves in the north are described as ʿabd, slave and recognised to have property of their own and may be wealthy; in the south, as ẖādim, servant, glossed as employed, working without responsibilities. Lancaster and Lancaster 2011: 176, 259–261, a very few metalworkers were Shiḥūḥ tribesmen in Musandam; Serjeant 1981b [1977]: 232, n. 7, a knife handle bought in Ṣanʿāʾ with an inscription showing it had been made by a member of a sayyid house. Lancaster and Lancaster 1999a: 121–122; idem 2011: 332, 335–336.

was there a “bedouinisation” of arabia?

675

the maintenance or re-establishment of tribal and personal peace by mediation and arbitration in resolving disputes between persons and their supporters, and in providing advice, when requested. Ḥukm is exercised by ‘the power to’ act and make decisions by each individual in contrast to ‘the power over’ that states are considered to have over their citizens.130 At the same time, each person should, by ‘behaving properly’, lessen situations that might cause disputes. The honour that each tribesperson has and his/her concern to have a good reputation underpins desired behaviours—being hospitable and generous, giving protection, acting as a sponsor or guarantor, acknowledging the rights of others, thinking of the future—in encampments large and small. In a political system embedded in the moral concepts of jural equality and autonomy, where no-one can command another, what is leadership and who are leaders? Burckhardt131 stated that šayẖs had no actual authority: “his commands would be treated with contempt; but deference is paid to his advice, if the people regard him as a man skilled in public and private affairs”. Leaders are those with reputation and influence, those whom people choose to support or follow or consult, and these are the same men with whom those outside the tribal peace interact when necessary for particular purposes.132 Aspects of leadership may be shared between members of the shaykhly family, often in relations with particular groups of those outside the tribal peace, or in specific enterprises with other tribes. ‘The šayẖ’ declares war, arranges alliances, and concludes peace; on occasion, a tribe may have two šayẖs, one for diplomacy and the outside world, the other for tribal and military affairs.133 The shaykhly family is seen as in some way ‘set apart’ from the tribe as a whole, it is ‘of’ but ‘not exactly the same’.134 Ṣabīr, kabīr, are the ‘leaders’ of each encampment, the men with the most widely known reputation, the most experienced, and the best advisers.135 Dresch’s analysis136 of Ḥāshid and Bakīl tribes of northern Yemen focused on peace, its breaking and remaking, in the moral balance between tribesmen and tribal groups, and distinguished personal peace from tribal peace. ‘Personal peace’ came from a tribesman’s honour and underpinned the social representation of his person; ‘tribal peace’ came from the defence of tribal honour, which was a “responsibility, a right, and a prerogative”. In his 2012 paper, Dresch dis130 131 132 133 134 135 136

Lancaster 1981: 79–80. Burckhardt 1967 [1831]: 115–119. Lancaster and Lancaster 1999b: 158–166; idem 2004a; idem 2015. Musil 1928: 504–506, 570–571, 50–51; Lancaster 1981: 126. Lancaster and Lancaster 2015: 67. Lancaster 1981: 77–78; Lancaster and Lancaster 2004a: 36–37, 39–40. Dresch 1989: 62–64, 79.

676

lancaster

cusses Yemeni texts concerned with “the laws of the tribes”, ḥukm al-manʿ,137 which basically all dealt with the laws of protection, and duties and rights of escort, refuge, and hospitality, and so produced a peace between moral equals. Shayẖs and other notables of Ḥāshid and Bakīl act as guarantors, kafīl, and ‘give their faces’ on behalf of tribesmen, i.e., protect them. Ideas of hijrah,138 space that is set aside, are also used by šayẖs as well as judges and sayyids to provide protected places and may be permanent or temporary. Nahid of Wādī Ḥaḍramawt139 spoke of tribe as a group that waged war, arranged truces, and concluded peace; any group that did these was a tribe. The groupings within Nahid were different kinds of units with varying political competences. The smallest named group, the bayt or dār, was the basic jural unit, whose raison d’être was to mobilize in defence of its members. A number of bayt made up a qabīlah, the basic alliance making unit and war-waging unit, “and in this sense, a political isolate”.140 These units were grouped into qabīlah or ʾāl, jurisdictional units which were grouped into the qabīlah, the Nahid, and suprajurisdictional. Nahid lands are not contiguous and their villages, based around large cores of bayt members, are interspersed among villages of other tribes and of non-tribal groups; “it is virtually impossible to mobilise any sub-segments of any unit of higher order than the subclan [alliance-making qabīlah]”.141 War, ḥarb,142 is between members of qabāʾil alliances, often within Nahid, sometimes between Nahid and similar alliance groups of other local tribes, and resolved by ḥākims, arbitrators from the ʾāl or qabīlah of Nahid if inside the tribe, and by the ḥākims of the Nahid and of a neutral tribe if between Nahid and another local tribe. Three of the four jurisdictional qabāʾil have a ḥākim who was both a ruler and a judge, the ‘father’ of his clan, while the fourth 137

138

139 140 141 142

Dresch 2012a: 151, the substance of these laws thought to be pre-Islamic; a Ruwalah šayẖ was sure that tribal law was pre-Islamic. Ibidem: 147, manʿ is not standard Arabic, but in Yemeni material denotes a capacity to protect or defend. Musil 1928: 448–449, from Ruwalah, a manīʿ was an enemy who surrendered voluntarily to his captor, his māniʿ was his “pardoner”, who saved him and then protected him. Wehr, 926, manaʿa, “to protect, to guard”. Serjeant 1981b [1977]: 228–229, 244, hijrah a peculiarly Yemenite institution, concerned with protection and respect; and suggested that the protection the Prophet was accorded at Medina may modify understanding of the Islamic term hijrah as “flight”. Hartley 1961: 55. Ibidem: 64. Ibidem: 84. Lane 1984 [1863]: vol. 1, 540–541, ḥrb, 1st form, “he despoiled him of his property leaving him nothing, despoiled”; 3rd, 4th and 5th forms, where both parties move against each other, war.

was there a “bedouinisation” of arabia?

677

had two ḥākims, the senior ḥākim for the Nahid tribe. Though the two senior hākims handled disputes of any kind, one was known as an expert in matters of dishonour, ʿayb, revenge or nagāʾ, naqāʾ, and killing, qatl; the other was an authority on cases of land and water. Nahid say their tribe is divided into two factions, each supporting one of the senior ḥākims; there is agreement that the buyūt of each ḥākim are their friends, ʾaṣḥāb; but no one can be certain which buyūt belong to which bloc at any one time. Tribesmen potentially need both and so each commits himself as little as possible to either, until it becomes necessary. All tribesmen are ultimately bound to both ḥākims who themselves are bound to all their tribesmen.143 The small tribes of Raʾs al-Khaymah Emirate and Omani Musandam144 viewed keeping the tribal peace as the business of tribespeople, and only where the wider family representatives failed to mediate over difficulties between families, did people go to the rulers, the šuyūẖ of the ruling families, for arbitration. Rulers were the arbitrators of last resort, they were the ḥukm, the government. The arbitrator’s decision had to accepted, because there was no other authority; but if for some reason, one side would not accept the arbitration, that party would leave the region. Inter-tribal disputes might be sorted out between themselves, or by a ruler. A Mazrūʿī explained: Each tribal section of the Mazāriʿ acknowledged an outside šayẖ, who might be the Ruler of Raʾs al-Khaymah, Sharjah, Dubai, ʿAjmān and so on, and this outside šayẖ was the final arbitrator. That was their only function, and they arbitrated only when we went to them. They were a convenience not a necessity, because there were always other arbitrators. The Shiḥūḥ tribe had their own arbitrators, and ruled themselves through public assemblies; serious disputes resolved by the families involved were ratified by the šayẖ. In Jaʿlān, south-east Oman, the small tribes describe themselves as “working for” Banī Bū ʿAlī, or “working with” Banī Bū Ḥassan; the tribes of Banī Bū ʿAlī and Banī Bū Ḥassan consider themselves to be in opposition, and are regarded by others to be so. This opposition is at least partially expressed in attitudes to ruling; each assumes it would keep a tribal peace, but how each would do so is different. The Wahhābī Banī Bū ʿAlī seemed to focus on imposing their political attitudes on tribes living in their sphere of influence and tried

143 144

Hartley 1961: 114–120. Lancaster and Lancaster 2011: 311–320.

678

lancaster

to expand this influence to more coastal areas. The ʾibāḍī Banī Bū Ḥassan preferred to work with tribes in their wider hinterland and supported them against encroachment. Members of these small local tribes managed their varied resources across the land and seascapes by using jamāʿah networks as individuals, while defending tribal assets as tribesmen.145 ʾĀl Murrah146 tribal leaders were the ʿaynin, the eyes, of the tribe, respected and consulted by their tribesmen. They ruled by persuasion, and from their ability to reconcile conflicts within the tribe, and to represent the tribe in dealing with the outside world. Formerly, they were military leaders but now are intermediaries between tribe and government. The family of current ʾĀl Murrah tribal leaders rose to eminence from successful warfare against the other tribes in the desert regions of al-Aḥsāʾ and then by diplomatic relations initially with the Turkish government of al-Aḥsāʾ and after their defeat in 1911, with the ibn Saʿūd. Al-Juhany147 describes the settled society of Najd as tribal since “the survival, identity and welfare of an individual were closely linked to his family and clan”; tribespeople, knowing their descent, were the majority of the population and all settlement chiefs were tribesmen. H̱ ādiriyyīn, without tribal descent, were craftsmen, traders, and farmers, and there were slaves. Some ẖādiriyyīn were descended from former slaves, while local historians thought others had been tribesmen who had lost their sense of ʿaṣabiyyah (group feeling) over time, but loss of their capital, their ownership of agricultural land, probably through debt,148 would be as plausible. Najdī towns had two significant groups among the tribespeople, the ruʾasāʾ or šuyūẖ and the neighbours, jīrān. Ruʾasāʾ considered themselves the rightful owners of the settlement and the ḥimā, the protected grazing lands, that they established around it, based on the assumption that their ancestors had bought or developed the land. Therefore the ruʾasāʾ could grant, lease, or sell property in their domain to whom they wished. Ruʾasāʾ welcomed or invited in potential settlers, from their own tribe or others, to live and cultivate arable land as jīrān, which increased the numbers of men available for defence and increased their revenues.149 Ruʾasāʾ protected the persons and property of their jīrān, while the jīrān helped defend the town and gave part of their harvest to the chief. 145 146 147 148 149

Lancaster and Lancaster 1992: 354–355; idem 2002. Cole 1975: 95–101. Al-Juhany 1983: 173–182. Altorki and Cole 1989: 53, 66, debt was pervasive in the town, especially for share farmers; but this would have been Islamic šarīʿah law, not tribal customary law. Al-Juhany 1983: 77, 229 n. 27; the establishment of a new settlement which then competed

was there a “bedouinisation” of arabia?

679

Ruʾasāʾ had limited authority; their public decisions were based on the interests of their own family and tribal section, and of those of the settlement as a whole. Ruʾasāʾ had slaves and fidāwiyyīn or retainers who functioned as police, administrative assistants, and in raids.150 Except in the largest towns, customary law was usual, exercised by the chief or any other recognised competent person. Ruʾasāʾ collected a share of the inhabitants’ harvests in return for protection; the amounts varied according to the relation of a settler to the chiefly family, title to land, and the terms on which the land was farmed. Distribution of income from a settlement’s population to members of the chiefly family (as an inheritance to the descendants of the town’s founder and as they were the first to be called on for the defence of the town) might explain many of the disputes within ruling families. Al-Juhany151 found infrequent attacks by nomads on settlements, which he attributed to nomad-settler relations becoming regularised, with alliances between particular towns and tribes which “may indicate the establishment of protective relationships” and “mutual needs”. His sources make no mention of tribute paid by towns to nomadic groups, but “the settled people had ‘brothers’ (associates) from among the nomads who would recover all properties and animals taken by members of their own tribes”; travellers and traders paid tolls to nomadic chiefs or paid a tribal escort for safe passage for themselves and their goods.152 In Oman, Wilkinson153 considered settled people were as tribal as nomads; “they form a vigorous, more or less self-sufficient and self-regulating society whose notions of honour and shame, of clan and genealogy, have much in common with the badū”. In the interior of Oman, the ʾaflāj, the reliable channelling of underground water to the surface for the cultivation of date palms and other trees, and other crops, were crucial to settlement. Major settlements concentrated a whole series of ʾaflāj into clusters at favoured points on the main drainage lines154 complemented by seasonal rains and flood flows. A falaj forms an individual settlement but the tenure depends on communal ownership and division of the water flows. Agriculture is the main productive

150 151 152 153 154

with the old; the expulsion of a jīrān group from a settlement because it became too numerous; and a jīrān group which bought the settlement from the ruʾasāʾ. So not dissimilar to a qawm. Al-Juhany 1983: 269–272. This is like the processes already described for Ruwalah and others. Wilkinson 1987: 94. Ibidem: 24.

680

lancaster

resource of inner Oman; “mulk155 rules are the form of ownership … tribalism determines the means by which such rights are acquired or exploited” with the tribal system safeguarding the legal rights of small owners and share farmers.156 Macdonald’s157 concern with the sedentary tribesman whose capital was in land, and so subject to external authority in a way the nomad is not, may be dispelled by al-Juhany’s and Wilkinson’s descriptions of tribal relations and access to land in tribal oases, which would have used tribal customary law, not Islamic šarīʿah law. That is, jurality and the power to act were vested in the individual, not in a centralised authority. The banī ʿamm relation, seen by Macdonald158 as “the only group in bedouin society in which each member is totally responsible for each of the others” and that “of course, the responsibilities affect all aspects of life,” was the view also held by Serjeant159 for Arab tribes in general, “for maintenance of security the Arabians evolved their collective system, the tribe is responsible for the protection of each of its members and for their misdemeanours”. As we have seen, ‘tribe’ is not corporate, as powers are vested in the individual as individual honour/autonomy. Musil160 described this relation for ʿAnazah tribes, among whom it was prohibited to bind a captured thief, to attack after midnight until early before sunrise, and a fixed blood price, in addition to the rights of kinship with its duties of protecting each other’s neighbour/qaṣīr, guest/ḍayf, or fellow traveller/ẖawī. The relation could be extended by chiefs to unrelated tribes. Doughty161 included grazing on the pastures of allies or protectors as part of the rights of kinship. Jaussen162 listed banī ʿamm agreements among tribes east of the Dead Sea, some of which are identified by town, e.g. Karak, Shōbak, and Maʿān ash-Shāmiyyah. Musil and Jaussen, like tribespeople, pointed out that such agreements were of little purpose unless all actually wanted such arrangements and actively participated. The banī ʿamm relation within a tribe may be taken as the concept of what should happen as the right of each

155 156 157 158 159 160 161 162

Ibidem: 116, mulk here means “fully alienable”, such lands may be bought, sold, inherited and rented as the land has been made more productive by the construction of ʾaflāj. Wilkinson 1987: 118–119. Macdonald 2015: 54. Ibidem: 44. Serjeant 1996: 272, “the tribe is responsible for the protection of each of its members and for their misdemeanours”. Musil 1928: 46–47. Doughty 1936: vol. 1, 396. Jaussen 1948 [1908]: 152–153.

was there a “bedouinisation” of arabia?

681

member, but does not imply concerted action by the aggregated tribe on behalf of one individual. Individuals are members of smaller units, their families and their named five generation groups, in which he has rights and receives obligations. As has been indicated above, no grouping above the five generation named group actually acts as one cohesive body, not even in wars between tribes or between tribes and outside powers.163 Logistics work against this, but more so do individual decisions as to the justice of the cause, and in cases of wars with outside agencies, the value of having members on both sides. Tribespeople consider that ideas of tribe provide an adequate and satisfactory rule in keeping the peace so that all may make their livings, by “minding our own business and getting on with our own affairs”. Tribal leaders, rulers, are there to deal with the leaders of other tribes and with the agents of other forms of government, with centralised and urban authorities. Raiding was associated by Dostal164 with warriors riding camels, that is, aggressive camel herding bedouin tribes. Raiding was about camels and horses, as the means of raiding and as useful booty; sheep and goats (and young camels) being too slow to keep up with the raiders, and taking tents weighed down the returning raiders. Accounts of and poems about raids are legion.165 Raiding may be seen as the other side of protection giving, protection from raiding often being what protection seekers, sheep and goat herders or farming villagers, wanted. This protection-giving returns raided animals, sheep and goats, and goods to the herders and villagers, and is to do with aspects of ruling, keeping the peace, and is not the subject of poems and accounts. Raiding for camels and horses was against tribes similar to the raider’s own but with whom there were no close relations.166 Musil167 records many raids between parties from tribes in a very generalised banī ʿamm relation, but ʿAnazah tribes condemned raids between banī ʿamm made only at night. Raiding required daring, bravery, and endurance, and could bring booty and reputation. Raiding could be on a large or small scale. Small scale raiding was carried out by a couple of youths168 to get together assets that might be retained personally or contribute to the income of the wider household. For many camel herding tribes, raids were often organised by qawm leaders with customary or negotiated jural 163 164 165 166 167 168

Lewis 1987: 154; al-Rasheed 1996: 437, “Evidence points to the fact that in every war groups from one tribe joined different protagonists”. Dostal 1959: 11–34. Musil 1928: 512–537, 607–661; Ingham 1995: 121–140. Jaussen 1948 [1908]: 166. Musil 1928: 506–538, 641–661. Ibidem: 653–655; Lancaster 1981: 140–141.

682

lancaster

allocations and divisions of booty169 for small and large raiding parties. Raiding was concerned with booty, not fighting, noted by Burckhardt,170 as violence against men resulted in revenge and feuds. Some accounts of raids (though not the poems) describe unsuccessful or disastrous raids.171 Tribesmen pointed out that raiding had an environmental function, since it limited the accumulation of camels beyond the capacity of the owner to guard them, and so encouraged sustainable pastures. Successful raiding built up the economic resources of participants, and enabled generosity; successful leaders increased their political reputations as well, with recognition of good planning, foresight, and management. It also increased knowledge of distant regions. Raiding had a political function in inhibiting encroachment by hostile or neighbouring tribes into one’s own tribal area, or in threatening the resources of other tribes.172 Ruwalah described this sort of raiding as preliminary skirmishes between tribes who were neighbours at that time and did not get on well together, to lead to war.173 Raiding was not limited to camel herding tribes of the inner deserts. Tribes of mountain areas also raided; Jaussen174 noted horsemen from Karak raiding in the vicinity of Bāniyās in the Jawlān, Tirābīn descending into Wādī ʿArabah to raid tribes there, and Ḥajāyā or Rishāyidah going down into the Ghōr for raids on travellers. Tribes, such as ʾAhl al-Jabal, which lived from sheep and goats but had camels for riding and carrying, also raided.175 Mountain and coastal tribes of Raʾs al-Khaymah Emirate and Omani Musandam described three sorts of raiding. The most common was raiding by land or sea as preliminary skirmishes before settlement of disputes or escalation into serious fighting between local tribes. Raids were also a tactic used by groups from one tribe trying to move into the territory of another. There was also raiding by members of tribes from outside the local region, such as ʿUwaymir, Manāṣīr, and Durūʿ from Oman, Abu Dhabi and Saudi Arabia, who came to capture women and children for sale as servants in Saudi Arabia, or to take whatever they could get, which in local usage could be elided with robbery. Raiders were identifiable by tribe, robbers were not. For raiding by tribally identifiable known raiders, there were processes to regain goods and for compensation for death or injury 169 170 171 172 173 174 175

Burckhardt 1967 [1831]: 137–144; Musil 1928: 506–507, 510–511. Burckhardt 1967 [1831]: 134. Musil 1927: 181–182, 206–207, 247–248; Musil 1928: 637, 650–652, 653–655. Cole 1975: 94–95. Musil 1928: 603–618, war with the western tribes. Jaussen 1948 [1908]: 166. Musil 1928: 660.

was there a “bedouinisation” of arabia?

683

through mediation and arbitration by tribal leaders.176 Against robbers, by definition unidentifiable and outside tribal processes although they might be tribesmen, there was active defence and the building of towers and strong stores.177 Al-Juhany178 noted that the pre-Wahhābī tribal towns of Najd, who kept horses and camels on ḥimā, protected grazing areas around a town, raided each other frequently, as well as raiding nomad tribal groups. ʿUnayzah, a major town in al-Qaṣīm, north of Najd, was the nucleus of a politically autonomous emirate, and raiding between its component settlements on tribal lines existed.179 Doughty180 reported that the ʾamīr of Ḥāʾil, ibn Rashīd, raided tribes to weaken them before making them submit to his power and pay zakāt. Jubbah, a town on the southern edge of the Nufūd, was inhabited by families from one section of the Shammar tribe. Jubbah had gardens that grew grain and dates, watered from deep wells by camels, and owned “great numbers of camels herded by bedouin relatives or by their own herders on nearby pastures”. As Wahhābīs and supporters of ibn Rashīd, they raided the Sharārāt and other tribes north of the Nufūd; in one summer, they made five raids, taking over 2,000 camels.181 Owning camels (and horses) by tribal townsmen enabled raiding. Al-Jawf, a town on the north of the Nufūd, was made up of twelve quarters, each called sūq, market, inhabited by families of nomadic sections of four tribes and families from settled towns or villages; Wallin182 noted they had few camels and depended on their bedouin allies for trade which took place at the date harvest, and made no mention of raiding by them or against them. In Ḥaḍramawt, Hartley183 noted Nahid in their villages and town raided and looted before actual war broke out, and considered184 the two chief hakīms/arbiters ‘raided’ each other’s jurisdictions to increase their influence.

176

177 178 179 180 181 182 183 184

Walker 1994: 326. In 1950, the ʿUwaymir who raided Raʾs al-Khaymah were a small nomadic section from Oman and Abu Dhabi under bin Ḥamm; the leader of a much larger ʿUwaymir section between Tarīm in Ḥaḍamawt and Abu Dhabi sent letters to the rulers of Sharjah and ʿAjmān stating they no longer had any connection to bin Ḥamm and his ʿUwaymir. Lancaster and Lancaster 2011: 322–325, 328–330. Al-Juhany: 273–275. Altorki and Cole 1989: 16. Doughty 1936: vol. 2, 241. Wallin 1854: 162–163. Ibidem: 139–151. Hartley 1961: 176, 181. Ibidem: 93–95.

684

lancaster

To finish this section: tribespeople may stop being tribal. Herding tribesmen often say it is hard work to be badū, explained not as the hard work of herding itself, but the requirement to be ‘a good man’, to behave honourably.185 Some do leave the desert and settle. Others settle because they consistently behave so badly that few will co-operate with them;186 some tribes have formal exclusion procedures.187 Those who leave in such ways abandon their tribal identity. Others settle—or their ancestors settled—for unknown reasons but assumed for reasons such as protection after a killing; an association with a family of a settled tribe; decisions that agriculture was preferable. In some cases, tribal identity is kept, e.g. a Ruwalah whose ancestors settled some 300–400 years earlier in north Jordan, and a Sbaʿah whose ancestors had settled in al-Jawf “a long time ago”. Others leave their tribe after a killing, take protection with another distant tribe, marry and stay there, ‘becoming’ a member of that tribe; this seems more common among the small tribes of Raʾs al-Khaymah Emirate, who are not so concerned with descent as a condition of tribal membership.188

Bedouinisation: Descriptions, Conceptual Oppositions Separating tribally organised societies by degrees of movement—into settled, semi-settled, semi-nomadic, and nomadic—is descriptive and not definitive, whether made by western travellers and scholars, or as conceptual oppositions of badū and settled/ḥaḍar by Arab townsmen and tribesmen. Badū and ḥaḍar are words with flexible meanings. The lexical meanings of badū are linked to the desert, describing those who live in that physical environment. Tribespeople who live in deserts describe themselves as bedouin tribes because they live in and from deserts; because they live from animals that do well in deserts; because they move, so their herds have pastures and water; because they have particular moral premises and are free to express these fully; and in so doing, complement other tribes and other social groups in other environmental regions.

185 186 187 188

Musil 1928: 480–481, a poem composed by a Ruwaylī who had accepted someone as a neighbour, qaṣīr, eight years earlier and how much time he had spent on him. Lancaster 1981: 76–77. Ginat 1983. Outside the tribal context, for example in the government departments of the modern state, where people are citizens, tribal identity is not used: a man is so-and-so, son of soand-so.

was there a “bedouinisation” of arabia?

685

Bedouin was also used for those living in the countryside by the inhabitants of towns and cities. Arab historiography considered the countryside of Arab villages, steppes, and deserts as outside the realms of the civilised and of proper Islam, as in Goitein189 and Faroqhi.190 Small herding and fishing tribes of Jaʿlān, south-east Oman, considered themselves as badū primarily because they managed their own affairs as they were far from government, and secondarily, because they moved and had animals. Banī Ḥamīdah of the north Karak plateau in Jordan, asked whether they considered themselves badū or fallāḥ/cultivators and so ḥaḍar, said they were badū because their flocks were the sources of their profits, they cultivated only for their own use, and because they managed their own affairs. Recently, since the introduction of modern transport, nation states, oil, and global politics, tribespeople describe themselves as bedouin because they make their livelihoods from their own efforts and resources rather than relying on government agencies.191 ‘Settled’/ḥaḍar is not a necessarily bounded concept. The fact that you have a house does not mean you are always there, as Musil192 noted; when a settler leaves his house for a tent he is an Arab, and either badū or šwāya.193 Wilkinson194 for Oman pointed out that because people may not move around to make a living does not make them ‘settled’ any more than livestock herding makes them ‘nomads.’ Some ‘settled’ tribespeople, particularly in mountain areas who farm and herd, have houses in two, three or four places they use over the seasons.195 Altorki and Cole196 described how the settled population of ʿUnayzah were “for generations … on the move both in and out of the city”, with camel caravans, as small carriers, mercenaries, wage labourers, and trading to bedouin camps, the Yemen, Ḥijāz cities, ports in the Gulf and southern Iraq, greater Syria, Egypt, Bahrein, and India. Wallin197 commented on the frequent journeys to Ḥāʾil by the settled tribal inhabitants of Jubbah for trade. A third of the settlers at Jawf accompanied Ruwalah on their migration north,

189 190 191 192 193 194 195 196 197

Goitein 1967: 75. Faroqhi 1994: 67–69. Fabietti, unpublished lecture, 1994; Lancaster and Lancaster 1986: 30–31. Musil 1928: 44–45. Lancaster and Lancaster 1999a: 212–213; Lancaster and Lancaster in Colin et al., paragraph on Syria and the Arabian peninsula; Wilkinson 1977: 62–63, 206 n. 14, 209. Wilkinson 1977. Lancaster and Lancaster 1999a: 114–120; idem 2011: 152–153, 163–167. Altorki and Cole 1989: 67–82. Wallin 1854: 162–163.

686

lancaster

so they could work in the harvest in the Ḥawrān.198 Métral199 discussed the inter-relations between herding and agriculture for the traders of Sukhnah, ʿUmūr and Ḥadīdīn herders and Sbaʿah tribesmen in the ḥamād, the Jazīrah, the Ghāb and the Ḥawrān. Lancaster and Lancaster200 mention examples from the literature of sheep flocks owned by Syrian townsmen and villagers using desert pastures. Dickson201 commented on Muntafiq sheep herders using agricultural land in Iraq, and herding and marketing sheep and sheep products in Kuwait and eastern Saudi Arabia. Doughty202 noted traders’ movements between the Ḥijāz and the Maydān quarter of Damascus, as did Musil203 for small traders from the settlement of Kubaysah near Hīt in Iraq. Bujra204 noted “there has always been extensive movement of people and goods205 between villages and towns, between the plateau and the valleys, and between the coast and the interior of Hadramaut”. Migrants from the interior had to get to the coast before going overseas, and return by the same routes; cash remittances and large amounts of imported goods had to be carried from the coast to the interior. There was inland trade between Ḥaḍramawt, Yemen and other parts of south Arabia; and in Ḥaḍramawt itself, some valleys were flooded while others were dry, so grain and dates from flooded areas were carried to dry ones. Al-Juhany206 described the movements of settled groups from one tribal town in Najd to another, as well as to towns in southern Iraq, Kuwait, or al-Aḥsāʾ. The tribes of the coastal towns of Raʾs al-Khaymah moved constantly from one town to another207 for reasons such as disputes between themselves or with the ruler, taxation, external political or economic events, the effects of storms, or to join others close to themselves. Men from ʾAhl Raʾs al-Khaymah, Zaʿab and Ṭunayj of Raʾs al-Khaymah Emirate went to the summer pearling in the Gulf,208 as did some bedouin groups of Banī Yās and Manāṣīr.209 198 199 200 201 202 203 204 205

206 207 208 209

Musil 1927: 279; Lancaster and Lancaster 1999a: 280. Métral 1993: 195–222. Lancaster and Lancaster 1999a: 212–213. Dickson 1949: 545–549. Doughty 1936: vol. 1, 339. Musil 1928: 269. Bujra 1971: 7. These movements were enabled by tribal political systems; safe conduct by tribesmen or holy men; hawtah, guaranteed as neutral places by tribesmen; holy men as mediators; and tribesmen with camels. Al-Juhany 1983: 185–227. Kemball 1985 [1845]: 94, 101. Lancaster and Lancaster 2011: 56–59. Wilkinson 1977: 54, 56.

was there a “bedouinisation” of arabia?

687

‘Bedouin’ is similarly not necessarily bounded. Bedouin tribes, tribal sections and families could and did own agricultural land, and sometimes worked it themselves, sometimes with share partners. ʿAnazah owned date gardens at Khaybar in partnership with villagers, while the arable lands were owned only by villagers;210 the ‘ground rights’ enabled rights for tribesmen to be carriers for the Pilgrimage and to receive Pilgrimage payments.211 In the ḥarrah (black basalt desert) of Khaybar, camel and goat herding Mawāhib families from ʿAnazah regularly grew barley, wheat, pumpkins, melons, and a little tobacco on terraces, watered by channels from springs above.212 Members of the leading family of the Ruwalah owned oases; ʿAbdallah ibn Ṭalāl owned Ithrā, and the sons of Saṭṭām owned most of the gardens at Kāf.213 Nūrī ibn Hazzaʾ considered buying Ḥadīṯah,214 and later bought land at ʿAdrā, north-east of Damascus. The Nawāṣirah ibn ʿamm are said to have sold the oasis of Qārah to the ibn Dughmī, the leading family of the Dughmān section of the Ruwalah, for horses.215 Wetzstein216 mentioned tribesmen growing grain crops on land flooded by runoff waters; an ʾAhl al-Jabal family started in 1931 a farm in the Jordanian black basalt desert using snowmelt and floodwaters, which grew barley and melons.217 Grain farming by Banī Ṣakhr camel herders was mentioned by Tristram218 and by Banī Ṣakhr and other herding tribes of the region in the Ottoman 1538 register.219 Musil220 noted tribal families or sections owning agricultural land in oases or riverain lands, worked by share partners. ʾĀl Murrah bedouin221 owned agricultural plots in oases cultivated by settled farmers who often claimed descent from the same ancestors, and shared the harvest; while many sedentaries owned animals which they entrusted to bedouin for fattening on winter pastures, the herders having the right to any milk. These possibilities of individual families of nomads/herders owning or using agricultural land, and the families of settled people owning animals herded

210 211 212 213 214 215 216 217 218 219 220 221

Doughty 1936: vol. 2, 133–134. Ibidem: vol. 1, 117. Ibidem: vol. 1, 477–478. Musil 1927: 326. Ibidem: 332. Lancaster 1981: 122. Wetzstein 1860: 30–31. Lancaster and Lancaster 1999a: 146–148. Tristram 1873: 303, 306. Bakhit and Hmud 1989: 136. Musil 1927: 360. Cole 1975: 108.

688

lancaster

in desert areas are examples of the management of family assets and options in multi-resource economics. For the majority of tribespeople, herding tribes continued to herd over the centuries and the millennia just as farming tribes, whether arable or oasis farming, continued to farm. If families or small groups migrated, which happened often for reasons of unresolved disputes or sometimes for environmental changes, they tended to go to places like those they had left; either because they already had links to the people of the new area, or because it was empty and recognised as similar and so easy to develop. Virtually all agricultural areas require channelled flows of seasonal water and/or stored waters, and people seem very able to construct their water management for agriculture from a wide range of environments. Similarly, herders will tell you how different one region is from another for herding particular animals, but assume they can herd successfully in each. Macdonald,222 discussing Caskel’s assumption that most traders and oasis dwellers became bedouin, agrees with Dostal’s comments that it would be very difficult for a sedentary to become a nomadic pastoralist (and by implication, vice versa). However, individual tribesmen, whatever their source of livelihood, are not isolates; each has family, connections through women, through shared enterprises which might be with herders (and vice versa), through exchange and support networks; he was able to make jurally binding contracts with others who had different economic resources; because he was a tribesman, he was in a system of moral relations in which each had a right to livelihood if only of protection, food and shelter. Many oases owned domestic animals including camels, sometimes herded locally by oasis dwellers, sometimes away with nomadic herders; connections would be available. Tribespeople assume anyone can learn other techniques for making a living; one can make share arrangements, hire oneself to someone who knows, or acquire the desired resource and hire someone who knows how to use it, and learn that way. Tribespeople are brought up to notice and to remember, whether in person or by listening to others, details of environments, livelihood practices, jural relations, relationships. A tribesman explained:223 “We aren’t so concerned with owning things outright but about being able to use resources over which we have claims. Because we are all members of families who all have some sort of assets and access, there are no people without anything forever. Someone might lose his land or his herd or his job, but his wider family would offer help and in a while he or his children will have something else. That is why we put

222 223

Macdonald 2015: 51. Lancaster and Lancaster 1999a: 325.

was there a “bedouinisation” of arabia?

689

so much importance on being a good man, having a good reputation. Wealth is as much or more in social relationships as in goods and money”. In Raʾs al-Khaymah Emirate, tribesmen considered that when a man had no resources left, he became a small trader, as at least that way, he got fed and sheltered. There are examples of settled people becoming nomadic;224 in the 17th century, peasants from Ramlah abandoned their land, many becoming nomadic herders, transporters, or traders; urban-based military-administrative personnel became transporters and traders with tribal affiliations;225 families in the Euphrates valley returned to being fully nomadic;226 and an extended sheep-herding Palestinian family from near Irbid in the 1990s “became badū” when they took up nomadic herding. Many families switched between nomadic herding and grain cultivation in desert fringe areas depending on suitable rains, available land, and the relative prices of grain, sheep/goats, and dairy products.227 Nomads settling never seems to cause writers any problems; but managing waters and their channelling, soils and their preparation, the sowing and nurture of crops, harvesting, processing and storage also needs to be learnt. Khawāṭir bedouin in the sands of Raʾs al-Khaymah Emirate described how their grandfather hired Sharqiyyīn tribesmen from the mountains to show and help him to start farming; there were existing relations between the two groups, as the Khawāṭir escorted the Sharqiyyīn to their summer places in coastal date gardens, and protected them when the Sharqiyyīn grazed their animals in the sands in the winters.

Why Settlement? Why Nomadic? Settlement needs reliable, permanent resources; waters for drinking by man and beast, and for agriculture; soil and rock formations that can hold and store waters on the surface and underground; fertile soils that produce flourishing trees, shrubs, and perennial and annual growths; preferably near other ecological areas. Maps of the Arabian peninsula show only some regions with settlements; apart from Yemen, rainfall is unreliable, erratic, and low. Most of Yemen’s population is settled; the Dhofari and Bāṭinah coasts of Oman have settlements, as do the mountains of Inner Oman; the Gulf coast has settlements. 224 225 226 227

Seikaly 1984: 406. Doughty 1936: vol. 2, 157–159. D’ Hont 1991: 214. Burckhardt 1992 [1822]: e.g. 221; Robinson and Smith 1841: 176–180; Lewis 1987: 82ff.

690

lancaster

In inner Arabia, there are numerous settlements in Lower Najd, al-Aḥsāʾ, Jabal Shammar, al-Qaṣīm, and on the west, in the Ḥijāz. Otherwise, settlements are few, and the population nomadic. The settled populations are and were larger than those of nomads, while it is probable that tribal populations were greater than those of non-tribesmen, certainly outside the few big cities. There are a few accounts of the foundations of towns and villages. In northern Yemen, a tribal village was founded jointly by the eponymous ancestor of the tribal section and the ancestor of the Jews of the present population; all other families had moved in.228 In Ḥaḍramawt, the ancestral lineage of Nahid founded the original village,229 now in ruins and known as ʿAdan, “source”, “homeland”, more or less in the centre of the dispersed lineages of the ancestral section, and equidistant from the seats of the two main hakīms. Over time, various groups of kinsmen left as a result of quarrels and established villages on the sites they now occupy. The one remaining house at ʿAdan belonged to the headman of a very small patriline of the tribal section who has the specific jural function of searching for an acceptable arbiter when intra-tribal disputants will not agree, and so is essential for Nahid as a community. Ḥurayḍah230 was originally settled by farmers, ḥirṯān; when ʿUmar, the ancestor of the sayyid families arrived, these farmers were in fourteen tribes and at feud. ʿUmar mediated, established peace and remained. The town became a religious sanctuary, ḥawṭah, ruled by the sayyid families descended from ʿUmar, and led by two hereditary chiefs, with strong political relations with groups of four local tribes, with whom they marry. The townsmen of Ḥurayḍah and the tribesmen form a ‘moral community’231 in which all groups need each other and are recognised to do so. Altorki and Cole232 state “throughout much of history, Najd was an area characterised by a high degree of autonomy from central control … Communities like Ḥāʾil, Buraydah and ʿUnayzah were urban places with their own amirs under whose leadership their citizens provided for their own defence. These cities (and the villages attached to them) were like islands in that they were surrounded by vast areas of desert that was controlled by autonomous Bedouin tribes”. ʿUnayzah was developed by members of the Banī Khālid and Subayʾ tribes, and the rulers, ʿumarāʾ, were always tribesmen. Townspeople used three descent categories for political life: qabīlī, of recognised tribal descent; ḥaḍar, freeborn but not knowing a tribal descent; and ʿabīd, slaves. ʿUnayzah was a 228 229 230 231 232

Dresch 1989: 78. Hartley 1961: 104–105. Bujra 1971: 37–38. Ibidem: 116. Altorki and Cole 1989: 17, 23.

was there a “bedouinisation” of arabia?

691

headquarters of the ʿUqayl camel merchants, and trade was very important to the town and its people. Sukhnah, near Palmyra, was an open town, developed as an alliance between settlers and nomads, and had sayyid or sādah families.233 Like Ḥāʾil in Jabal Shammar,234 Sukhnah had gardens and springs, but both relied more on servicing trade routes235 into northern Syria and Turkey, and Saudi Arabia. There were series of symbiotic relations between towns, countrysides, deserts, mountains, and coasts, expressed as moral relations between persons who had particular occupations and jural identities. Settlement or nomadism by tribespeople was related to where they were in the land and seascapes, their perceptions of resources available to them, and their development of these. In some regions, northern Yemen, Ḥaḍramawt, inner Oman, the mountains of Raʾs al-Khaymah and Omani Musandam, parts of upland Ḥijāz and Jordan, tribal names appear to be constant for centuries, although individuals, families and small sections are known to have come in or left. In other regions, such as the desert areas of the Bilād ash-Shām, the eastern Ḥijāz and Najd, tribal names come and go across the centuries. In his descriptions of pre-Wahhābī Najd tribal settlements, al-Juhany noted236 many migrations in and out of the region by groups from nomadic and settled tribes, and while some nomad families and groups did settle, most migration in that region was by settled tribal groups. Local accounts of tribal migrations described families identified as parts of tribal sections with places in which they could see themselves continuing, more or less, to live as they did before. Individuals and their families might and did change livelihood patterns, on the whole larger groups did not. But the flexible and labile descriptions of tribal groups makes this distinction uncertain, since recent movements by known persons are spoken of as such whereas those of more distant times are identified by group names—but when these movements actually took place, the movers might have been families, wider domestic groups. The ability to move was seen by tribespeople as a right, as a guarantee of individual autonomy. People moved to where wealth (in the old sense of being self-sufficient) was, that is, where a family could make its living, and this had happened all the time. Moving and then making a living by a share partnership, 233

234 235 236

Métral and Métral 1991: 163. This is rare in the Bilād ash-Shām, but according to Peake (1958) such families had been important in Ramthah and for small tribal groups in north Jordan. Al-Rasheed 1991. De Boucheman 1939; Weuleresse 1946: 307; Métral and Métral 1991: 156. Al-Juhany 1983: 171–172, 186–187.

692

lancaster

whether in herding or farming, so that both were free and equal partners, was the manner in which tribespeople considered movement to be enabled.237 Reasons for migrations were various, and focused on moving from an area of poor conditions, be these environmental, economic, or political and jural, to an area thought to be better. Or this could be phrased as droughts; internal unresolved disputes; disputes with central authorities. Droughts/ jifāf were a fact of life, whichever livelihood strategy was followed, and there were customary ways of managing drought by herders and farmers.238 Centrally imposed political changes made some or many of these strategies unavailable to Banī Ṣakhr in the 1930s, the Karak tribes in the 1940s, and to Ruwalah in the late 1950s and early 60s.239 Market changes had made people more vulnerable since their capital had either lost value (camels), or had increased in value (agricultural land) and attractive to urban merchants supported by national laws rather than local customary law. Although these droughts were bad, and animals died, every informant insisted that people had not died from hunger. Tribesmen in the sands of Raʾs al-Khaymah Emirate240 recalled Yemeni camel herders coming in by agreement because of severe droughts in their home areas and who stayed for five or six years (Thesiger 1949 refers to this migration). They also said that the constant light dews in the sands were sufficient for the ġāf and samrah trees, even if there were no rains, and that tree areas were not grazed in good rain years. This is similar to ʾĀl Murrah, who left ʿabal,241 which remained green and edible for four years after one rain, in the Empty Quarter, as a reserve.242 Camels were seen as walking larders, while sheep and goat dairy products could be stored for months or years. In the Ruʾūs al-Jibāl mountains of Raʾs al-Khaymah Emirate and Omani Musandam, tribespeople expected localised storms to enable farming in the terraced fields of some mountains and wadis in some areas while others would

237 238

239 240 241 242

Lancaster and Lancaster 2006: 347. The multi-resource nature of most tribal livelihoods is relevant. Most farming families also had animals for milk, manure, and transport, and had fields or gardens in different locations to spread risk; had relations in nearby but different drainage systems on whom they could call for support; could move as migrant workers to better watered areas; and had stored grains or dates. Herders reduced flocks and herds, and moved to reserve areas or to distant better areas using customary agreements; often had stored dairy products; sought employment. Lancaster and Lancaster 1993: 223–246. Idem 2011: 133. ʿabal—calligonum crinitum, Mandaville 1990: 108, pl. 59. Cole 1975: 32–33.

was there a “bedouinisation” of arabia?

693

be dry. This was the reason to have fields in different drainage systems and at different altitudes. They considered they managed adequately in droughts as the goats fed on dry grazing and the trees, water remained in their cisterns and seeps, and they had grain in their granaries.243 In the valleys of the western Ḥajar, Raʾs al-Khaymah, old tribesmen244 recalled that when there were no rains for growing grain, they had eaten the grain in their stores, then shared that of those who still had some grain, then they had milk from the animals who browsed trees moistened by dews, dates watered from wells, and meat of male kids and gazelle. No one in the mountains ever went hungry, it was people in coastal towns who went hungry, people without capital who depended on their labour. Merchants were remembered as having been generous in times of need in the coastal towns. Nahid farmers in Ḥaḍramawt245 say they can manage for three years without a flood to their fields, from which they normally have three harvests in a year—wheat, barley, and sorghum. Their area has more regular floods than other local areas, and as they live with kin and affines, they help each other. Dates fruit a year after flood waters, so provide a crop when there is no grain harvest, and survive without water for three to four years—Nahid regard the date tree as the botanical equivalent to the camel. Normal grain harvests are good, and store well, providing the household for at least a year and often two. Nahid also control the market to which grain from western Yemen comes in years of local famines. Famine affected poorer, non-tribal labourers and tribal cameleers more. While drought and consequent famine meant some left the area for migrant work, Hartley makes no mention of migration by Nahid. Al-Juhany,246 using the pre-Wahhābī chronicles of Najd, noted there had been a considerable settlement of the nomadic population during the two and a half centuries that preceded the Wahhābī Reform Movement. Favourable ecological and climatic conditions in the first half of the period encouraged migrations to Upper Najd from the west initially by small groups of ʿAnazah and Muṭayr, followed by larger groups being harassed by the šarīf s of Mecca, and from the east by Banī Khālid; then by Dawāsir and Qaḥṭān from the south-west and then ʾĀl ʿAjmān and ʾĀl Murrah from the Najrān region. Some 243 244 245 246

Yields on these fields were 1:40 to 1:100, and kept for at least six years (Lancaster and Lancaster 2011; 148–149). Lancaster and Lancaster 2011: 225. Hartley 1961: 44. Al-Juhany 1983: 152–155. Population increase; this seems obvious, but in our experience tribal populations who herded or farmed themselves had a very slow rise in population. Leading families had more children, but lost many young men from raiding.

694

lancaster

tribes that had been in Najd moved to southern Iraq. In the second half of the period, droughts became more frequent; since the nomadic population had increased above the number that could be supported from the region’s natural resources, this led to fierce conflicts during the 17th century. Weaker and smaller groups were obliged to settle in Lower Najd. Some ʿAnazah and Shammar sections left for the Syrian desert, while other nomadic families settled, some establishing their own communities, others settling in old villages and towns of some of the older tribes that had gone to Southern Iraq. The nomads who settled were only a small part of the total nomad population, but a significant addition to the settled population. There are also reports of settled ʿAnazah families of Najdī towns and villages joining their nomadic relations247 in severe droughts. Droughts cause migration because there follows a dearth of profits and unsatisfactory livelihoods; while movement to other areas needed processes to acceptance. Neighbour, guest, protected status, share-farmer—all these and others are possible for tribespeople.248 Wilkinson,249 in Oman, considered that “life in the desert is so poor250 that the drive for the badū to improve their economic lot is paramount”, which leads 247 248

249

250

Seetzen 1810: 281–324. Lienhardt 2001: 146–147, described the movements of a family owning four trading boats and their ‘followers’, their boat crews and probably the crews’ families, 270 people. They had left the Persian coast for Khaṣab in Musandam, then gone to Qatar, then in the late 1940s to Failaka Island with the permission of Kuwaytī šayẖs, on condition they supported themselves and were of good behaviour. Wilkinson 1977: 189–207. This view is based on his use of the ‘frontier of settlement’ model, derived ultimately from ibn Khaldūn’s badū and ḥaḍar, in which there “is a tendency for the inhabitants of zones of lower economic potential to try and establish territorial rights in the nearest neighbouring area of greater land potential”, Wilkinson 1977: 64. While there are social mechanisms for this, the process is slow and conflict is limited; “It is therefore more profitable to view the occupants of the different natural regions of Oman as forming complementary groups specializing in the country’s different land-resources”, Wilkinson 1977: 66. Wilkinson 1977: 189. He continues “So it is that the nomads of south-east Arabia will do work that is despised … among other Arabs”, referencing Dostal 1967: “They will fish, go pearling, make charcoal, and even hew salt in order to supplement their basic living from livestock herding”. In our experience, tribespeople in that area (sections of Wahībah, Banī ʿAmr, Ḥarb, Ṭunayj, and Shiḥūḥ) making livelihood and profits from fishing do so because their dīrah are on the coasts; Zaʿab pearled because they were on the coast and had boats as they also traded by sea; charcoal making by mountain tribes in Raʾs al-Khaymah was a part of their economy, for cash in coastal towns. All saw themselves, and were seen by others, as honourable as they owned and defended their productive resources. What was dishonourable was the loss of these resources or inability to support one’s family.

was there a “bedouinisation” of arabia?

695

to some badū settling by acquiring agricultural land, easier in periods of political instability as there would be abandoned land to move into or be purchased. Badū groups did acquire outlying abandoned falaj/qanāt settlements; gained political control over outlying qanāt settlements where most of the existing population remained, and bought and/or bought into outlying settlements.251 He also mentions252 ex-settled groups who became nomadic herders as an economic choice or did so because they had lost their agricultural land. Unresolved disputes between tribesmen and tribal families may result in long term or permanent migration to another tribe. In Raʾs al-Khaymah Emirate and the region, the killer fled and took protection from a tribesman while compensation and amends were arranged; often the killer remained, worked, and married and ‘became’ a member of the protecting tribe. In other tribes, for example, the killer and his close male relations flee and take protection with someone of another tribe, the women remain and look after the family livelihood. In cases that remain unresolved, the only solution is for one party to move; the lack of resolution is often not the actual injury itself but a long standing complex of disagreements. The dissatisfied party moves to someone they already know and has a good reputation in a tribe like their own. Sometimes groups from a tribal section leave because of long standing difficulties, sometimes tribes broke apart because of differences in ideas of ruling between tribal leaders and tribesmen, and their constituent parts went their separate ways.253 Sometimes tribal sections move because they do not want to become involved in wars their tribe are fighting; this was the case with Banī ʿAmr in Jaʿlān, who moved from Suwayq on the Bāṭinah coast. Migration caused by difficulties with centralised authorities or their agents and tribal leaders are well known. Al-Juhany254 commented that the expansion of ʿAnazah groups into upper Najd in the 17th century was partly a response to harassment by the šarīf of Mecca, and their movement north-west a century later a response to the disruption accompanying Wahhābī expansion; a quarrel between the ruler of Raʾs al-Khaymah and a section of the Zaʿab tribe caused that section to move to Abu Dhabi. Ruwalah withdrew to their southern bases

251

252 253

254

Wilkinson 1977: 193, gives examples of qanāt land-holding tribal groups’ vulnerability to central authorities; but it is to be assumed that confrontations ended in mediated negotiations between tribal leaders and central authorities. Wilkinson 1977: 206, n. 14. Lancaster and Lancaster 2011: 551–554; Banī ʿAmr in Jordan, who ruled the Karak region in the 16th–18th centuries, lost power by their different sections quarrelling over ideas of rule, splitting, and many sections migrating. Al-Juhany 1983: 152–153.

696

lancaster

in Saudi Arabia from Syria because of the Baʿath government’s actions against tribes in general and in particular against Ruwalah.255 Tribes and their leaders do not consider relations with central governments as inevitably oppressive; co-operation for specific purposes, such as the safety of the roads,256 the passage of and supplies to the Pilgrimage,257 or trade between certain tribes and towns258 are examples, although customary arrangements were broken. Migration by ʿulamāʾ families from tribal towns in Najd, whose leaders insisted on the use of tribal customary law, to the larger towns where šarīʿah law was the norm, was known.259 Exchange and trade have been important in Arabia for millennia. We are out of our sphere of competence in the period of c. 100 ad to the early Islamic conquest, but from Lecker,260 by Late Antiquity the markets of Arabia were well-established, with merchants paying customs duties to regional tribal leaders as proxies of the competing Byzantine and Sassanian Empires. Merchants as a category can almost be opposed to tribe: merchants were of family and had family networks, but of no larger jural group; urban; literate; religious;261 upheld šarīʿah law; were not armed or extended protection; owned boats, land, and animals solely for commercial ventures; invested in enterprises using credit and debt; had wealth as money and goods; and were often close to rulers or as, on the Gulf coast and coastal Oman, were rulers; and assume a centralized authority. Merchant investment262 has been noted in ʾaflāj/qanāt for commercial agriculture in Oman, the acquisition of arable land from tribespeople in Jordan and Palestine, commercial fishing in Bahrein and Shihr, south Yemen, in ocean going dhows and Baṣrah date gardens, and pearling. Merchants were rich. In the early 1950s, a Kuwaiti merchant told Lienhardt263 it had been the pearl 255 256 257 258

259 260 261 262 263

Lancaster 1981: 112–116. Hiyari 1975; Bianquis 1991: 91; Irwin 1986: 49; Bakhit 1982: 204ff. Faroqhi 1994: 54–73; Bakhit 1982: 107–117, 204–226; Barbir 2014: 169. Dickson 1949: 49, described the musābilah of the Muṭayr, Ḥarb, Shammar, ʾAwāzim, and northern tribespeople with the towns of Kuwait, and the Ḍafīr tribes with the towns of Zubayr, Nāṣiriyyah, and Samāwah in south Iraq. Cole 1975: 106–108, 111, for ʾĀl Murrah with Hufūf and Najrān. Wilkinson 1977: 192–193, for Durūʿ tribespeople and the town of ʿIbrī. Al-Juhany 1983: 254. Lecker 2005: vii, 109–114. Bujra 1971: 21, 32, the ruling sayyid family of Ḥurayḍah, missionaries and religious scholars, had strong links to merchants. Al-Juhany 1983: 254. Wilkinson 1977: 124–126; Rogan 1992; Musil 1908: 86; Gubser 1973: 28; Serjeant 1995: 508, 193–197; Villiers 1940: 323–340; Lorimer 1908–1915: 2233, 2286; Heard-Bey 1996: 208–211. Lienhardt 1993: 99.

was there a “bedouinisation” of arabia?

697

merchants and boat owners of Kuwait who had made the wealth of Kuwait, and invited in šayẖs from Najd to maintain law and order, and look after the country’s defence; the people of Kuwait had allowed them to tax pearl fishing and levy customs duties at 4%, to give them an income. Other Gulf šayẖs, such as the Qawāsim of Raʾs al-Khaymah, were wealthy merchants264 during the 18th century with close links to the interior through the Naʿīm tribal šayẖs at Buraymī. Debt was a constant factor, seen along with profit-sharing by Lienhardt265 in the context of economic relations as the centre of the traditional economic system. He continued: I had it made very clear to me that they also belonged within moral relations … The moral situation lay in the relation between the owner of the boat and the sailors who worked the boat. This was a relation of patron and client, and the client had a moral expectation of of being helped and looked after by his patron, who was also, usually, his creditor. Because of profit sharing, the sailors were not strictly speaking the ‘employees’ of the boat owner. They were his ‘company of followers’, his jamāʿah.266 Credit is the other face of debt, and conversations in Raʾs al-Khaymah Emirate made it clear that many had found credit essential, and that creditors had been realistic and generous about repayment; as one elderly man remarked “generosity and friendship were the important matters, not the totals in the account book”. Altorki and Cole267 showed the importance of credit in the women’s market in ʿUnayzah. In the bad years of the late 1920s and early 1930s, “pearl merchants and boat owners had thought it dishonourable to seize the assets of debtors poorer than themselves”;268 Kanoo269 wrote that his grandfather, a merchant in Bahrein, had lost an immense amount of money but “gave to families who were finding hard to survive”. 264 265 266

267 268 269

Slot 1993: 250. Lienhardt 1993: 96–98. Tribal seamen in Raʾs al-Khaymah recognised the moral relation between boat owner/ creditor and sailors but not that of patron and client. Each needed the other; for the owner and his crew, there were alternative crew members and boat owners. They were moral equals of contributing what they had and in sharing risk and profit. The competent are assumed to be generous to and protective of the less able; generosity and protection are conceived of as diffuse relations rather than reciprocal. Altorki and Cole 1989: 147–154. Lienhardt 1993: 99. Kanoo 1997: 44, 46.

698

lancaster

Since merchants and rulers had wealth as money at their disposal, these groups are assumed to have the working capital for investment in infrastructure—major water capture and delivery, for example, the construction of forts, or ocean-going boats.270 The ʾaflāj or qanāt of Oman271 are a case in point, as are the massive systems of terracing, water storage and channelling on the western side of the peninsula.272 Underground water systems in Syria273 and Ḥijāz274 are like Omani ʾaflāj but called foggara in Syria and qanāt by Nasif. The origins of these systems have been associated with Sasanian Iran; Wilkinson275 is convinced of such an origin and that while tribal village organisations might maintain or restore such systems, they could not have developed them. Wilkinson linked mining and the making of tunnels with underground water systems; mining for copper is very early in Oman, and could have been connected to falaj construction. It is assumed that such construction would have taken huge amounts of investment and labour as well as skill, out of the reach of tribes but available to centralised authorities who could command labour.276 Hartley277 examined this view in relation to the large-scale irrigation system used by Nahid farming tribespeople, and found that “surprisingly little maintenance” was required, carried out by one or two men for a short time after each flood; major extensive repairs were made by non-tribal labourers (not numerous) hired by tribal leaders. Haj Ibrahim278 described villagers of Dayr ʿAṭiyyah digging a five kilometre channel, nine metres deep at its start, in 300 work days. Doughty279 saw a new well in Ḥāʾil fifteen fathoms (c. 25 m) deep that took fifteen workmen twenty days, and at Buraydah, a well that took three skilled men twenty-five days work. In Raʾs al-Khaymah Emirate and Omani Musandam, the northern and southern mountain systems have different hydrologies; in the north, there are a few springs and rainfall is channelled across mountain slopes for a kilometre or two to complexes of terraced fields and to a variety of cisterns; in the south, fields and gardens are in the valley bottoms, water is channelled to them from the slopes, there are springs filling holding tanks

270 271 272 273 274 275 276 277 278 279

Villiers 1940: 324, 341. Wilkinson 1977: 74–95. Hartley 1961: 37–44. Haj Ibrahim 1990. Nasif 1988. Wilkinson 1977: 122–134. Cleuziou 2002: 227–228; but see Lancaster and Lancaster 2012: 113–114, 120. Hartley 1961: 41–42. Haj Ibrahim 1990: 301. Doughty 1936: vol. 1, 301.

was there a “bedouinisation” of arabia?

699

channelled to gardens, and shallow wells. All these systems, of which this is a very simplified account, are instigated, maintained, and managed by tribesmen.280 Perhaps these examples are straightforward, although they depend on acute and sustained observation and a dexterity with stone. What about the deep wells on the plains at the feet of the northern mountains? These are said to be 40–50m deep, it is unknown when they were dug, but probably in the early Iron Age. A local tribesman281 said: In my opinion, the wells were started by men digging in wadi gravels for water. Very gradually, over time, the wells became deeper. This was partly from the repeated cleanings of the stuff brought down by floods and winds over the years and centuries, partly because the earth and gravels brought down from the mountains built up layers on the surface of the soil, very slowly over time. And partly from changes in the weather; rainfalls and storms change; sometimes there are periods of heavy storms, sometimes rain falls as rain. Sometimes there are long periods without rain so the depth of water in the wells would fall, and people would dig down deeper to the new level. So in all these ways, wells became deeper, and their depths now282 show how old they must be—thousands of years. A French archaeologist at an Iron Age site at Mleiha gave a virtually identical account. These processes would probably account for other deep wells in desert regions, such as those on the northern edge of the Nufūd.283 The same principle of repeated excavation and cleaning out over centuries may be seen in the maḥāfir in parts of the Bilād ash-Shām deserts,284 series of built pools at the ends of flood flows from rainfall. Some of the bigger complexes are fifty metres or more in diameter with walls four metres high, held over 100,000 cubic metres and would last for months, the smallest are holes dug in a drying rain pool. Early and mediaeval Arabic inscriptions are associated with some. Local tribesmen state that these structures would have been started long, long ago and become the size they are now from repeated cleanings out

280 281 282

283 284

Lancaster and Lancaster 2011: 137–149, northern mountains; 191–208, southern mountains. Ibidem 2011: 103–104. One such well was described as 35 men deep, and so required plaited leather ropes to haul up the water, another was 26 men deep; the numbers are the number of men required for cleaning the well, so multiplying numbers of men by average height does not give the depth of the well. Lancaster and Lancaster 1999a: 140–141. Agnew et al. 1995; Lancaster and Lancaster 1999a: 135–138.

700

lancaster

of sediments brought in on flows and blown in by winds. They also point out that to have a successful series of maḥāfir, the digger has to know the directions and speeds of flows into the area, which are complex and change according to exactly where rain fell, its duration and intensity. There are many other water catchment and storage facilities in deserts and in the countrysides made by tribespeople over the centuries and millennia285 using deep knowledge of their environments, social and jural relations. Tribespeople assume oasis gardens and qanāt water facilities would have been started by tribal families and groups as shares in joint enterprises, and to have been developed and elaborated over time, working with each micro-environment and a deep knowledge of local materials and the needs of crops and trees. Ownership comes from development to make something more productive; construction comes about from ownership (in the sense of privileged or uncontested access) of the place and its resources, and from contributions of skill, labour and materials which result in shares of the enterprise.

Centralising Structures Several authorities assume the need for a central authority or structure, as if tribes were incapable of providing peace, organising systems of production and exchange, or developing infrastructure such as the provision of water and storage facilities. If tribal practice provides security and support for the needy, enabling economic enterprises by shares and handing-on, then it is assumed to be power relations of patron and client rather than moral relations between autonomous and jurally equal individuals who are members of identifiable and known groupings. Across the peninsula, tribesman and šayẖ, shepherd and flock-owner, cultivator and landowner, boat-owner and fisherman who owns the nets, producer and merchant, both sides reiterate that patron-client relations cannot happen for two reasons. Firstly, it would be inconsistent with the underlying moral premises since protection and hospitality are open to all and from all, and are diffuse rather than reciprocal relations. The second, one-to-one relations are for a particular purpose, and were contracts for that occasion between two equal partners, who were accountable, and involved restitution and recompense for non-fulfilment. The equality of partners is possible because each brings capital to the enterprise and shares in the risks, neither can command the other, and both have to negotiate and agree.

285

Lancaster and Lancaster 1999a: 129–166; idem 2011: 99–105, 137–147, 191–208.

was there a “bedouinisation” of arabia?

701

Altorki and Cole286 considered the market acted as “a major centralising structure that linked the various systems of specialised production followed by bedouin, farmers, and craftspeople”. This has validity, but ignores the role of tribes as guarantors of markets which was common in northern Yemen,287 Ḥurayḍah in Ḥaḍramawt,288 Najrān and al-Aḥsāʾ,289 Khaybar,290 and the Ḥawrān.291 It also ignores the role of the networks demonstrated by Firestone292 in Palestine and the importance of share partnerships and handing on in these productive and distribution networks.293 Others assume centralising religious and/or political rule. Wilkinson294 considered the Sasanian empire initiated the ʾaflāj of Oman, expanded or restored by ʾibāḍī Islamic rulers and merchants. Serjeant295 sees “an ever-recurrent pattern of centralisation” exemplified by “an unending conflict between the secular organisation of tribal groups and the theocratic nuclei of the professional men of religion” and manifested at sacred enclaves, where both sides guarantee safety for all. The observance of Islam “could not be enforced except by a strong central authority”296 and thus “the Muslim state became the seemingly indispensable corollary of the Muslim religion”. ‘States’ imply cities, and Islam based itself in cities, people of the countryside and tribespeople could not be fully observant Muslims. Al-Juhany297 referred to the need for a central authority or unifying ideology for the tribal towns of Najd. Dresch298 describes Yemeni tribes as part of a larger whole, centred on the Zaydī Imams who were themselves sayyids. The influence of the successive Islamic governments of

286 287 288 289 290 291 292 293 294 295 296

297 298

Altorki and Cole 1989: 233. Dresch 1989: 123–130. Bujra 1971: 117. Cole 1975: 106. Doughty 1936: vol. 2, 205. Musil 1928: 270. Firestone 1975: 185–209. Lancaster and Lancaster 1999a: 43–45, 212–213, 300–301, 313–319; idem 2011: 237–282, for Raʾs al-Khaymah and region, where exchange and gifts reciprocated over time. Wilkinson 1977: 122–134. Serjeant 1981a [1962]: 41–42, 47–51. Goitein 1968: 31. Ibidem: 39, described the building of the Islamic community, the ʾummah, as based on the Arab tribe, “the only practical example of a community” on which he could model it. The crucial differences, imposed by Islam, was strict obedience to those in authority, and the settlement of disputes had to be settled by orderly (Islamic) legal procedure. Al-Juhany 1983: 272, 279. Dresch 1989: 158–167.

702

lancaster

Iraq, Syria and Egypt, and of the guardianship of the Holy Places is a constant in accounts of the relations between these entities and the countrysides and tribes. Tribes construct themselves as opposed to centralising forces or states, which themselves often construct tribes as social formations to be destroyed,299 marginalised,300 or incorporated as clients.301 This opposition is made by tribespeople as ḥukūmah, rule as ‘keeping the peace’ which allows tribesmen as individuals and with their supporters to use ‘power to’ activate customary jural processes and make agreements between jural equals, and dawlah, rule by a centralising state with ‘power over’ by officials and which removes ‘keeping the peace’ from individual tribesmen. This attitude is common to nomadic tribes, such as Ruwalah,302 but also common to long settled tribes in Yemen,303 Oman and Najd. Tribespeople’s dislike of centralised state government304 is embedded in the refusal of such states to countenance face-to-face moral relations between known jurally autonomous individuals where everyone is responsible for what they say and what they do,305 and are answerable to complainants.306 States incorporate the constituent groups of their populations as power relations delivered from the centre by officials and informed by an ideology.307 Another fundamental opposition between centralised political structures and tribal political concepts is over lands. Centralised polities regard all unowned—because undeveloped to be more profitable—land to be the property

299

300 301 302 303 304

305

306 307

Al-Juhany 1983: 281–301, Ibn Saʿūd’s Wahhābī state. Burckhardt 1967 [1831]: vol. 2, 117–120. Lancaster 1981: 112, the Syrian Baʿath government. Baram 1997: 29, 1–31, the Iraqi Baʿath government. Dresch 1989: 158, as in northern Yemen by the Zaydī Imamate. Velud 1995: 65; al-Azmeh 1986: 82, 86; Kostiner 1991: 234–235. Lancaster 1981: 128–131. Dresch 1989: 183–189; Wilkinson 1987: 36; al-Juhany 1983: 180. Exemplified by the comment about a man imprisoned for taking the law into his own hands; “it’s the government’s job to defend the interests of its citizens. But they won’t do it, and they don’t let you do it yourself”. Glubb 1959: 173, “There was less inequality in wealth and social position in the old insecure chaotic time that there was under the new theoretical ‘democracy’. The establishment of law and order resulted in the rich becoming richer and the poor growing poorer … the establishment of public security deprived the farmer of the power to threaten the usurer with violence”. Lancaster and Lancaster 1999a: 366–375; 2011: 312–317. In reality, there is some compromise between the two systems; Lancaster and Lancaster 1999a: 381–386, 1990s Jordan. Mandate Syria, Büssow 2011; Lewis 1987: 154; Thoumin 1936: 153. Yemen 1960s–1980s, Dresch 1989: 253–268.

was there a “bedouinisation” of arabia?

703

of the state: tribes regard all land as having value and use of some kind at some time for some groups, and that, in this sense, to be associated preferentially with tribally identified groups.308 As has been mentioned, many tribes live from and therefore ‘own’ different environmental areas over the seasons; so the same area is used by sequences of groups from different tribes.309 But at the same time, owning also embodies the defence of these lands and the keeping of the peace among users; and so ownership is vested in individuals as members of tribes. Owning and using different areas to which tribal individuals and families have well established claims are common, and for nomads or settled. The regions used seasonally have few developed assets, and used for grazing and/or for collecting natural products such as medicinal plants, firewood, or wood for camel sticks, etc. for use or sale, but are a vital resource310 in the overall economic resource management of the peninsula. However, it is possible to view tribe, rather than state, as a more realistic longterm social formation; a Ruwalah leader said “States depend for their income on external factors. Tribal wealth comes from its social cohesion and the assets of its people. These are material, like herds, land, wells, property in general, the resources available to people through tribal processes, and especially the generosity and good moral qualities of tribespeople. This tribal system, unlike a state, depends on itself, it continues since it is, at bottom, a moral system. So a tribe, if it maintains its system, will outlast a state. This has happened before. The tribe might not be exactly the same, members might have changed a bit, it might have another name, but it goes on”. Doumani311 regards the presentation of decentralised political arenas in local histories by families, whether of tribe or of town, to present “an alternative framework to centralisation: unity through cultural solidarity and local identification”. This view is consistent with local ideas and practice, and to be productive for those interested in the processes of historical changes. This paper has tried to examine ideas of tribes, badū and ḥaḍar, from tribespeople, through ethnographic and other scholarly writings. From these, it seems that building a theory of historical development of Arabia on the dualities of badū/ḥaḍar is not useful. Badū/ḥaḍar oppositions have their uses as shorthand, quick labels to describe but not to define or for analysis. The premises, constructs, processes, and practice of tribes, badū and ḥaḍar, are 308 309 310 311

E.g., Lancaster and Lancaster 2011: 354–355, 462. Lancaster and Lancaster 1999a: 106, 214–219; idem 1992: 354–355; idem 2002: 239–241, 246– 248. Ibidem: 170–175; idem 2011: 105–108, 232. Doumani 1995: 19–20.

704

lancaster

ignored whereas these are entirely relevant and show that it is tribe itself, as a general concept and as particular manifestations, that has created much of what has taken place in the Arabian peninsula across the millennia through the jural and moral relations engendered by ideas of tribe.

References Note that the series in which the works listed below are published are not systematically indicated. Agnew, C., E. Anderson, W. Lancaster, and F. Lancaster. 1995. “Mahafir: a Water Harvesting System in the Eastern Jordan (Badia) Desert.” GeoJournal 37 (1): 69–80. Altorki, S., and D.P. Cole. 1989. Arabian Oasis City: The Transformation of ʿUnayza. Austin: University of Texas Press. Antoun, R. 1972. Arab Village: a Social Structural Study of a Trans-Jordanian Peasant Community. Bloomington: Indiana University Press. al-Azmeh, A. 1986. “Wahhabite Policy.” In Arabia and the Gulf: From Traditional Society to Modern States, edited by I.R. Netton, 75–90. Totowa, nj: Barnes & Noble. Bakhit, M.A. 1982. The Ottoman Province of Damascus in the 16th century. Beirut: Librairie du Liban. Bakhit, M.A., and N.R. Hmud. 1wfl989. The Detailed Defter of Liwa Ajlun, no. 185, Istanbul: 1538. Amman: The University of Jordan Press. Baram, A. 1997. “Neo-tribalism in Iraq: Saddam Hussein’s Tribal Policies 1991–1996.” International Journal of Middle East Studies 29: 1–31. Barbir, K. 2014. Ottoman Rule in Damascus, 1708–1758. Princeton: Princeton University Press. Beeston, A.F.L. 1972. “Kingship in Ancient South Arabia.” Journal of Economic and Social History of the Orient 15: 256–268. Betts, A.V.G., and D. Cropper. 2013. “The Eastern Badia.” In The Later Pre-History of the Badia, edited by A.V.G. Betts, D. Cropper, L. Martin, and C. McCartney, 179–191, Oxford-Oakville: Oxbow Books. Bianquis, T. 1991. “Le Ḥawrān de l’avènement de l’Islam à la conquête ottomane.” In Le Djebel Al-ʿArab. Histoire et patrimoine au musée de Suweidāʾ, edited by J.-M. Dentzer and J. Dentzer-Feydy, 89–92. Paris: erc. Bisson, J. 1991. Le nomade, l’oasis et la ville. Actes de la Table Ronde tenue à Tours les 21– 22–23 septembre 1989. Tours: urbama. Bocco, R. 1987. “La notion de Dirah chez les tribus bédouines en Jordanie: le cas des Bani Sakhr.” In Terroirs et Sociétés, edited by B. Cannon, 193–214. (Études sur le monde arabe 2). Lyon: Maison de l’Orient.

was there a “bedouinisation” of arabia?

705

Bocco, R., R. Jaubert, and F. Métral (ed.). 1993. Steppes d’Arabie. États, pasteurs, agriculteurs et commerçants: le devenir des zones sèches. Paris-Genève: Presses universitaires de France. de Boucheman, A. 1939. Une petite cité caravanière: Suḫné. Damas: Institut français de Damas; Paris: E. Leroux. Bujra, A. 1971. The Politics of Stratification: a Study of Political Change in a South Arabian Town. Oxford: Clarendon Press. Büssow, J. 2011. “Negotiating the Future of a Bedouin Polity in Mandatory Syria: Internal and External Political Dynamics of the Sbaʾa-ʾAbada during the 1930s.” Nomadic Peoples 15: 70–95. Burckhardt, J.L. 1992 [1822]. Travels in Syria and the Holy Land. London: Darf Publishers. Burckhardt, J.L. 1967 [1831]. Notes on the Bedouins and Wahhabys. London-New York: Johnson Reprint Company. Cleuziou, S. 2002. “The Early Bronze Age of the Oman Peninsula: from Chronology to the Dialectics of Tribe and State Formation.” In Cleuziou et al. 2002, 191–246. Cleuziou, S., and M. Tosi. 2012. “Reasoning about Magan.” In Giraud et Gernez 2012, 313–323. Paris: De Boccard. Cleuziou, S., M. Tosi, and J. Zahrins. 2002. “Introduction.” In Cleuziou et al. 2002a, 9–27. Cleuziou, S. 2002a (ed.). Essays on the Late Prehistory of the Arabian Peninsula. Rome: Instituto Italiano per l’Africa e l’Oriente. Cole, D.P. 1975. Nomads of the Nomads: the Āl Murrah Bedouin of the Empty Quarter. Chicago: Aldine. Colin, G.S., Lancaster, W., Lancaster, F., and O. Jastrow. “Shāwiya.” In Encyclopaedia of Islam, Second Edition, edited by P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel, and W.P. Heinrichs. Accessed August 2, 2016. http://dx.doi.org/10.1163/1573-3912_islam_COM_1048 Dickson, H.R.P. 1949. The Arab of the Desert. London: Allen and Unwin. Dostal, W. 1959. “The Evolution of Bedouin Life.” In L’antica società beduina, edited by F. Gabrieli, 11–34. Rome: Centro di Studi Semitici. Doughty, C. 1936. Arabia Deserta. London: Jonathan Cape. Doumani, B. 1995. Rediscovering Palestine: Merchants and Peasants in Jabal Nablus, 1700–1900. Berkeley: University of California Press. Dresch, P. 1989. Tribes, Government and History in Yemen. Oxford: Clarendon Press. Dresch, P. 2012a. “Aspects of Non-State Law: Early Yemen and Perpetual Peace.” In Dresch and Skoda 2012, 145–172. Dresch, P. 2012b. “Legalism, Anthropology and History: a View from Part of Anthropology.” In Dresch and Skoda 2012, 1–38. Dresch, P., and H. Skoda (ed.) 2012. Legalism, Anthropology and History. Oxford: Oxford University Press. Fabietti, U. 1990. “Between Two Myths: Underproductivity and Development of the

706

lancaster

Bedouin Domestic Group.” In Sociétés pastorales et développement, edited by E. Bernus and F. Pouillon, 237–253. (Cahiers des Sciences Humaines 26). Paris: Éditions de l’orstom. Fabietti, U., and P.C. Salzman. 1996. The Anthropology of Tribal and Peasant Societies. Pavia: Collegipo Ghisliere. Faroqhi, S. 1994. Pilgrims and Sultans: the Hajj under the Ottomans. London: I.B. Tauris. Firestone, Y. 1975. “Production and Trade in an Islamic Context: Sharika Contracts in the Transitional Economy of Northwestern Samaria 1853–1943.” International Journal of Middle Eastern Studies 6: 185–209. Garrard, A., S. Colledge, and L. Martin. 1996. “The Emergence of Crop Cultivation and Caprine Herding in the ‘Marginal Zone’ of the Southern Levant.” In The Origins and Spread of Agriculture and Pastoralism in Eurasia, edited by D.R. Harris, 204–226. London: University College of London Press. Ginat, J. 1983. “Meshamas—the Outcast in Bedouin Society.” Nomadic Peoples 12: 26– 47. Giraud, J., and G. Gernez (ed.). 2012. Aux marges de l’archéologie. Hommage à Serge Cleuziou. Paris: De Boccard. Glubb, J.B. 1959. Britain and the Arabs: A Study of Fifty Years, 1908 to 1958. London: Hodder and Stoughton. Glubb, J.B. 1968. Studies in Islamic History and Institutions. Leiden: Brill. Gubser, P. 1973. Politics and Change in al-Karak, Jordan. Oxford: Oxford University Press. Haj Ibrahim, M. 1990. “Les techniques de captage et de conduite de l’eau aux fins d’irrigation et les modes de culture irriguée dans les ghoutas de l’Ouadi el Majarr (Kalamoun Nord): l’exemple de Deir-Atieh.” In Techniques et pratiques hydro-agricoles traditionelles en domaine irrigué. Approche pluridisciplinaire des modes de culture avant la motorisation en Syrie, edited by B. Geyer, 295–312. (Bibliothèque archéologique et historique 136). Paris: P. Geuthner. Hartley, J.L. 1961. “The Political Organization of an Arab Tribe of the Hadhramaut.” PhD diss., University of London. Heard-Bey, F. 1996. From Trucial States to United Arab Emirates. London: Longman. Hiyari, M.A. 1975. “The Origins and Development of the Amīrate of the Arabs during the seventh/thirteenth and eighth/fourteenth Centuries.” Bulletin of the School of Oriental and African Studies 38: 509–524. d’Hont, O. 1991. “Production pastorale et consommation citadine: lait et laitages en Syrie orientale.” In Bisson 1991, 213–221. Hoyland, R. 2001. Arabia and the Arabs from the Bronze Age to the Coming of Islam. London: Routledge. Ingham, B. 1995. “Texts in the Dialect of the Rwala of Northern Arabia.” In Dialectologia Arabica: A Collection of Articles in Honour of the Sixtieth Birthday of Professor Heikki Palva, 121–140. (Studia Orientalia 75). Helsinki: Finnish Oriental Society.

was there a “bedouinisation” of arabia?

707

Irwin, R. 1986. The Middle East in the Middle Ages: the Early Mamluk Sultanate, 1250–1382. London-Sydney: Croom Helm. Jabbur, J.S. 1995. The Bedouins and the Desert: Aspects of Nomadic Life in the Arab East, translated by L. Conrad, edited by S.J. Jabbur and L. Conrad. Albany, ny: State University of New York. Jaussen, A. 1948 [1908]. Coutumes des Arabes au Pays de Moab. Paris: Adrien Maisonneuve. al-Juhany, U.M. 1983. “The History of Najd Prior to the Wahhabis.” PhD diss., University of Washington. Kanoo, K.M. 1997. The House of Kanoo: A Century of an Arabian Family Business. London: London Centre of Arab Studies. Kemball, A.B. 1985 [1845]. “Memoranda on the Resources, Localities, and Relations of the Tribes Inhabiting the Arabian Shores of the Persian Gulf.” In Arabian Gulf Intelligence: Selections from the Records of the Bombay Government, New Series, no. xxiv, edited by R.H. Thomas, 91–119. Cambridge-New York: Oleander Press. Khalidi, T. 1994. Arab Historical Thought in the Classical Period. Cambridge: Cambridge University Press. Kister, M.J. 1980a [1965]. “Mecca and Tamīm (aspects of their relations).” In Studies in the Jāhaliyya and Early Islam, 113–163. Aldershot: Variorum Reprints. Kister, M.J. 1980b [1968]. “Al-Ḥīra. Some Notes on its Relations with Arabia.”Ibidem, 61– 93. Kostiner, J. 1991. “Transforming Dualities: Tribe and State in Saudi Arabia.” In Tribes and State Formation in the Middle East, edited by P.S. Khoury and J. Kostiner, 226–251. London: I.B. Tauris. Kuper, A. 1982. “Lineage Theory: a Critical Retrospect.” Annual Review of Anthropology 11: 71–95. Lancaster, W. 1981. The Rwala Bedouin Today. Cambridge: Cambridge University Press. Lancaster, W., and F. Lancaster. 1986. “The Concept of Territoriality among the Rwala Bedouin.” Nomadic Peoples 20: 41–48. Lancaster, W., and F. Lancaster. 1990. “Desert Devices: the Pastoral System of the Rwala Bedouin.” In The World of Pastoralism: Herding Systems in Comparative Perspective, edited by J.G. Galaty and D.L. Johnson, 177–194. New York: Guildford Press. Lancaster, W., and F. Lancaster. 1992. “Tribe, Community and the Concept of Access to Resources: Territorial Behaviour in South-east Jaʾalan.” In Mobility and Territoriality: Social and Spatial Boundaries among Foragers, Fishers, Pastoralists and Peripatetics, edited by A. Rao and M. Casimir, 343–363. New York-Oxford: Berg. Lancaster, W., and F. Lancaster. 1993. “Sécheresse et stratégies de reconversion économique chez les bedouins de Jordanie.” In Bocco et al. 1993, 223–246. Lancaster, W., and F. Lancaster. 1995. “Land Use and Population in the Area North of Karak.” Levant 27: 103–124.

708

lancaster

Lancaster, W., and F. Lancaster. 1997. “Indigenous Resource Management Systems in the Badia of the Bilad ash-Sham.” Journal of Arid Environments 35: 367–378. Lancaster, W., and F. Lancaster. 1999a. People, Land and Water in the Arab Middle East: Environments and Landscapes in the Bilad ash-Sham. Amsterdam: Harwood Academic Press. Lancaster, W., and F. Lancaster. 1999b. “On the Nature of Power in the Works of Orientalist Scholars and its Contribution to a History of Bedouin Society and NomadSedentary Relations in the Bilad ash-Sham.” In Antonin Jaussen, sciences sociales occidentales et patrimoine arabe, edited by G. Chatelard and M. Tarawneh, 143–175. Amman: cermoc. Lancaster, W., and F. Lancaster. 2002. “Trade and Stratified Society: Information from Jaʾalan, Oman.” In Cleuziou et al. 2002a, 237–253. Lancaster, W., and F. Lancaster. 2004a. “Concepts of Leadership in Bedouin Society.” In The Byzantine and Early Islamic Near East, edited by J. Haldon and L. Conrad, 29–61. Princeton: Darwin Press. Lancaster, W., and F. Lancaster. 2004b. “Ar Rawdhah in the Musandam.” Journal of Oman Studies 13: 127–137. Lancaster, W., and F. Lancaster. 2006. “Integration into Modernity: Some Tribal Rural Societies in the Bilad ash-Sham.” In Nomadic Societies in the Middle East and North Africa: Entering the 21st Century, edited by D. Chatty, 335–369. Leiden: Brill. Lancaster, W., and F. Lancaster. 2011. Honour is in Contentment: Life Before Oil in Ras alKhaimah (uae) and some Neighbouring Regions. Berlin: De Gruyter. Lancaster, W., and F. Lancaster. 2012. “Ethnography and Archaeology: Some Thoughts from Ra’s Al-Jinz.” In Giraud and Gernez 2012, 107–124. Lancaster, W., and F. Lancaster. 2015. “Concepts of Tribe, Tribal Confederation, and Tribal Leadership.” In Les Jafnides. Des rois arabes au service de Byzance (vie siècle de l’ère chrétienne), edited by D. Genequand and C.J. Robin, 53–77. (Orient & Méditerranée 17). Paris: De Boccard. Lane, E.W. 1984 [1863]. Arabic-English Lexicon. 2 vol. Cambridge: Islamic Texts Society. Lecker, M. 2005. People, Tribes and Society in Arabia. Ashgate: Variorum Collected Series. Lewis, N. 1987. Nomads and Settlers in Syria and Jordan, 1800–1980. Cambridge: Cambridge University Press. Lienhardt, P. 1993. Disorientations: A Study in Flux. Kuwait in the 1950s, edited by A. alShahi. Reading: Ithaca. Lienhardt, P. 2001. Shaikhdoms of Eastern Arabia, edited by A. al-Shahi. Basingstoke: Palgrave. Lorimer, J.G. 1908–1915. Gazetteer of the Persian Gulf, Oman, and Central Arabia. 2 vol. Calcutta: Government Printing Press.

was there a “bedouinisation” of arabia?

709

Macdonald, M. 2015. “Was there a ‘bedouinization’ of Arabia?” Der Islam 92: 43–74. Mandaville, J. 1990. Flora of Eastern Saudi Arabia. London-Riyadh: Kegan Paul International and National Commission for Wildlife Conservation and Development. Métral, F. 1993. “Élevage et agriculture dans l’oasis de Sukhné (Syrie): gestion de risques par les commercants-entrepreneurs.” In Bocco et al. 1993, 195–222. Métral, J., and F. Métral. 1991. “Une ville dans la steppe, la tribu dans la ville: Sukhné, Syrie.” In Bisson 1991, 153–171. Miles, S.B. 1994 [1919]. The Countries and Tribes of the Persian Gulf. Reading: Garnet Publishing. Musil, A. 1908. Arabia Petraea. Vol. 3. Ethnologischer Reisebericht. Vienna: A. Hölder. Musil, A. 1927. Arabia Deserta. A Topographical Itinerary. (Oriental Explorations and Studies 2). New York: American Geographical Society. Musil, A. 1928. The Manners and Customs of the Rwala Bedouins. (Oriental Explorations and Studies 6). New York: American Geographical Society. al-Naim, M.A. 2004. “The Dynamics of a Traditional Arab Town: the Case of Hofūf, Saudi Arabia.” Proceedings of the Seminar for Arabian Studies 34: 193–207. Nasif, A.A. 1988. Al-ʿUlā: An Historical and Archaeological Survey with Special Reference to its Irrigation System. Riyadh: King Saud University. Peake, F. 1958. The History of Jordan and its Tribes. Coral Gables, fl: University of Miami. Postgate, J.N. 1992. Early Mesopotamia: Society and Economy at the Dawn of History. London: Routlege. Postgate, J.N. 1994. “In Search of the First Empires.” Bulletin of the American Society for Oriental Research 293: 1–13. Pouillon, F. 1996. “Bédouins des lumières, bédouins romantiques: mouvement littéraire et enquête sociologique dans le voyage en Orient (xviiie–xixe siècles).” In Fabietti and Salzman 1996, 57–79. al-Rasheed, M. 1991. “Le pouvoir et l’économie caravanière dans une oasis de l’Arabie du Nord.” In Bisson 1991, 225–235. al-Rasheed, M. 1996. “Religious Revival and Tribal Cohesion: the Fragmentation of the Shammar under Wahhabism.” In Fabietti and Salzman 1996, 437–445. Robinson, E., and E. Smith. 1841. Biblical Researches in Palestine, Mount Sinai, and Arabia Petræa. London: John Murray. Rogan, E. 1992. “Moneylending and Capital Flows from Nablus, Damascus and Jerusalem to Qaḍāʾ al-Salṭ in the last Decades of Ottoman Rule.” In The Syrian Lands in the 19th and 20th centuries, the Common and Specific in the Historical Experience, edited by T. Phillipp, 239–260. Stuttgart: Franz Steiner. Schilcher, L. 1991. “The Grain Economy of Late Ottoman Syria.” In Landholding and Commercial Agriculture in the Middle East, edited by C. Keyder and F. Tabuk, 173– 196. Albany: State University of New York.

710

lancaster

Seetzen, U.J. 1810. “Mémoire pour servir à la connaissance des tribus arabes en Syrie et dans l’Arabie Déserte et Pétrée.” Annales des voyages, de la géographie et de l’histoire 8: 281–324. Seikaly, M.S. 1984. “Land Tenure in 17th Century Palestine: the Evidence from the alFatawa al-Khairiyya.” In Land Tenure and Social Transformation in the Middle East, edited by T. Khalidi, 397–408. Beirut: American University in Beirut. Serjeant, R.B. 1981a [1962]. “Haram and Hawtah, the Sacred Enclave in Arabia.” In Serjeant 1981c, 41–58. Serjeant, R.B. 1981b [1977]. “South Arabia.” In Serjeant 1981c, 226–247. Serjeant, R.B. (ed.) 1981c. Studies in Arabian History and Civilization. Aldershot: Variorum Reprints. Serjeant, R.B. 1995. “Fisher-folk and Fish-traps in al-Bahrain” and “Customary Law among the Fishermen of al-Shihr.” In Farmers and Fishermen of Arabia, edited by R.B. Serjeant. Aldershot: Variorum Reprints. Serjeant, R.B. 1996. “The Customary Law of South-west Arabia and Bedouin Justice in Jordan.” In Society and Trade in South Arabia, edited by G. Rex Smith, 269–280. Aldershot: Variorum Reprints. Slot, B.J. 1993. The Arabs of the Gulf, 1702–1784: An Alternative Approach to the Early History of the Arab Gulf States and the Arab Peoples of the Gulf, Mainly Based on Sources of the Dutch East India Company. Leidenschendam: Slot. Smith, W.R. 1885. Kinship and Marriage in Early Arabia. Cambridge: Cambridge University Press. Thoumin, R. 1936. Géographie humaine de la Syrie Centrale. Paris: E. Leroux. Tristram, H.B. 1873. The Land of Moab. Travels and Discoveries on the East Side of the Dead Sea and the Jordan. New York: Harper & Brothers. Velud, C. 1995. “Syrie: tribus, mouvement national et État mandataire (1920–1936).” In Tribus, tribalismes et États au Moyen-Orient, edited by R. Bocco and C. Velud, Maghreb-Machrek 147: 48–71. Paris: La documentation française. Villiers, A. 1940. Sons of Sindbad. London: Hodder and Stoughton. Walker, J. 1994. The United Arab Emirates: Internal Boundaries and the Boundary with Oman. 8 vol. Farnborough: Archive Editions. Wallin, G.A. “Narrative of a Journey from Cairo to Medina and Mecca, by Suez, Araba, Tawila, al-Jauf, Jubba, Hail, and Nejd, in 1845.” Journal of the Royal Geographical Society 24: 115–207. Watt, W.M. 1973. The Formative Period of Islamic Thought. Edinburgh: Edinburgh University Press. Weuleresse. R. 1946. Paysans de Syrie et du Proche-Orient. Paris: Gallimard. Wehr, H. 1974. A Dictionary of Modern Written Arabic. Wiesbaden: Harrassowitz. Wetzstein, J.G. 1860. Reisebericht über Hauran und die Trachonen nebst einem Anhange über die sabäischen Denkmäler in Ostsyrien. Berlin: Dietrich Reimer.

was there a “bedouinisation” of arabia?

711

Wilkinson, J.C. 1977. Water and Tribal Settlement in Oman. Oxford: Clarendon Press. Wilkinson, J.C. 1987. The Imamate Tradition of Oman. Cambridge: Cambridge University Press.

Index The individual contributions are distinct enough to allow readers to find most topics with ease. This index contains a list of keywords that occur in more than one contribution. Agrippa (king) 239–242, 246 Ancient North Arabian 64, 131, 134, 136, 202, 221, 225, 346 Ancient South Arabian 53, 55, 64, 94, 96, 113, 116–119, 121, 131, 141, 338, 623 Arabic route 334–336, 542 Assyrian 222, 375, 395–396, 479–480 Babylonian 364, 368, 373–374 Bedouin 81, 85, 274–275, 483, 491, 509, 535, 645–647, 657, 660, 669, 673, 680–681, 683–690, 701 Betyl 15 Bronze Age 116, 361, 365, 380, 382, 385, 415, 418–429, 651 Burial 365, 383, 385, 398, 421, 439–440, 446– 447, 552–553, 560 Camel

72–75, 85, 291, 372, 398, 427, 439– 446, 479, 529–535, 601, 610, 613, 616–617, 620–621, 645, 647–650, 657, 662, 669, 672–673, 681–703 Caravanserai 271, 290–291, 443–444. Christianity 327, 334–336, 495, 502, 509, 511, 542 Curse 41, 47, 49, 64, 70, 309, 577–578 Dadan 203–215 Dadanitic 58, 131, 202–215, 218–233, 338 Desert 85, 101, 119, 263, 270, 275, 280–281, 288, 291, 293, 327, 439, 479–480, 512, 529, 535, 552, 553, 583, 617, 650, 672– 673, 678, 682, 684–691, 694, 699, 700 Dūmah (=Dūmat al-Jandal) 362, 479 Dura Europos 286, 299–300, 302 Formularies

96, 99–102, 111–113, 132

Gerrha 32 Ghassanids 273, 275, 329–330, 490 Gulf (Arabian) 105, 453, 526, 685–686, 689, 696–697 Gulf (ʿAqabah) 393, 426

Graffiti 116–119, 130, 134–141, 161–163, 187–188, 218, 225, 230, 392, 453 Hatra 20, 45, 571–578 Hamad 666, 686 Ḥarrah 253–254, 262, 270, 687 Ḥawrān 54–55, 239, 243, 245, 253–254, 261, 263, 265–266, 283, 439, 686, 701 Hellenistic 442–446, 482, 490, 567, 578, 583, 590–591 Ḥīra 497, 500–501, 508 Identity 96, 426, 428, 492, 591, 639, 647, 654– 658, 669, 670, 678, 684 Iran/Iranian 29, 298–303, 319, 491, 501–502, 507, 513–514, 543, 574, 698 Jāhiliyyah 654 Jafnid 273, 275, 490, 498 Jawf 18, 97, 100–102, 271, 291, 292, 330, 334, 479, 670, 683–685 Jerusalem 484, 486, 538, 567, 571–575 Jew/Jewish 303, 484, 527, 542, 559, 560, 567, 569, 575, 581, 582–592, 690 Lakhmid 490, 526 Liḥyan/Liḥyanite 27, 202, 381 Literacy 28, 141, 346 Madāʾin Ṣāliḥ 4, 5, 7, 10, 16, 21, 218, 225 Maryamat 102–113 Medina 525, 551, 676 Migration 662, 685, 691–696 Military 452–456, 480, 491, 493–494, 497, 503–506, 510, 512–513, 645, 653–656, 662–664, 675, 678, 689 Modern South Arabian 343, 601, 622–628 Monumental inscriptions 214, 544 Nabataean 60, 218, 223, 227, 334, 335, 378, 381, 385, 421, 567, 578 Naṣrid 490 nomadism 483, 691

714 Oasis 131, 203–215, 218–219, 225, 232, 271 Old Arabic 210 Onomastics/onomasticon 28, 105, 107, 125, 127, 133, 136, 140, 231, 233, 482 Ostraca 395, 482, 486 Palmyra 443, 499, 569, 691 Persia 324, 482, 484, 487, 489–514, 539–543, 573, 694 Petra 4–21, 105, 286, 291–292, 441, 441–443 Punic 222, 225, 346–355, 440, 443 Poetry 85, 522–530, 539, 542, 607, 624, 651 Prophet 510, 527–538, 551, 557, 584, 671, 676 Proto-Semitic 202, 340, 342, 626 Province of Arabia 334–335, 496, 512 Qaḥṭān 455, 663, 693 Qatabanian 102–106, 120, 125, 136–141, 221, 230 Quran 52, 64, 67, 336, 342, 522–523, 536–544, 551–557, 563 Rock art 74–76, 81 Rome 240, 243, 247, 274, 302, 335, 493–506, 514 Rwala 491, 495, 509, 513, 647, 660

index Sanctuary 116, 119, 138, 139, 141, 323, 415, 427– 428, 463, 568, 690 Safaitic 41–64, 69–76, 81–87, 130, 132, 142, 207, 209, 221–232, 254, 265, 339, 342, 346, 552, 553 Scribal school 119, 214 Sedentary 479, 483, 502, 513, 645, 680, 688 Soldier 244, 246, 262–265, 281–283, 289–293, 304, 330, 495, 587, 674 Syria 84, 253, 318, 327, 330–335, 365, 397, 402, 453, 499–500, 504, 507, 512, 522, 525, 535, 555, 660, 666, 668–669, 685, 691, 696, 698, 702 Thamud 206, 225, 479, 522–529, 532–533 Thamudic 87, 219, 225 Trade 58, 105–106, 202, 230, 427, 429, 453, 484, 486, 495, 503, 602, 672, 678, 683– 696 Tomb 5, 7, 12, 51, 67, 206, 242, 243, 244, 247, 265, 257, 274–275, 288, 354, 395, 396– 399, 439–446, 500, 551, 557, 561–564, 578 Ṭayyāyē 496, 505