The Aramaeans: Their Ancient History, Culture, Religion 9042908599, 9789042908598

Citation preview

ORIENTALIA LOVANIENSIA ANALECTA The Aramaeans Their Ancient History, Culture, Religion

PEETERS

Bar-Rakkab, king of Sam'al, seated on his throne, with his scribe. Orthostat in dolerite from Zincirli, ea. 730 B.C., 1.13 m high, I.15 m large: Staatliche Museen zu Berlin, VA 2817 (Courtesy PreuBischer Kulturbesitz, Photo: J. Liepe).

ORIENTALIA LOVANIENSIA ANALECTA ---100---

THE ARAMAEANS THEIR ANCIENT HISTORY, CULTURE, RELIGION

BY EDWARD LIPINSKI

UITGEVERIJ PEETERS and DEPARTEMENT OOSTERSE STUDIES LEUVEN - PARIS - STERLING, VIRGINIA 2000

Library of Congress Cataloging-in-Publication Data

Lipinski, Edward The Aramaeans: their ancient history, culture, religion I by E. Lipinski. p. cm. -- (Orientalia Lovaniensia analecta; 100) Includes index. ISBN 90-429-0859-9 I. Arameans--History. 2. Middle East--History--To 622. I. Title. II. Series. DS59.A7L57 2000 939.4--dc2 l

00-046525

© 2000, Peeters Publishers & Department of Oriental Studies Bondgenotenlaan 153, B-3000 Leuven/Louvain (Belgium) All rights reserved, including the rights to translate or to reproduce this book or parts thereof in any form. D. 2000/0602/64 ISBN 90-429-0859-9 (Peeters, Leuven)

CONTENTS PREFACE

11

ABBREVIATIONS

15

I. Aramaean Pre-history and Proto-history .

25

I. Early Occurrences of Aram

26

2. The Land of Qir

40

3. Aramaean Tribes in the 13 th Century

45

4. The Etymology of the Name Aram

51

II. Aram and the Hebrew Forefathers . III. Lage

55 77

I. Territory .

82

2. History

98

IV. Nisibis and the Temanites

109

I. Territory

I 10

2. History

Ill

V. Gozan or Bet-Bagyan and BalI!}

119

I. Territory

123

2. History .

128

VI. Bet-Zammani

135

I. Territory.

137

2. History

151

6

CONTENTS

VII. Bet-'Adini

163

1. Territory.

165

2. History .

183

VIII. Arpad or Bet-Gus

195

1. Territory

199

2. History .

211

IX. Kittik or Bet-�ulliil

221

1. Territory

221

2. History .

230

X. Sam'al or *Yu'add1

233

1. Territory

237

2. History .

238

XI. J:Iamath and Lugath

249

1. Territory.

255 255

a) Tell Afis and ljagarik b) Shalmaneser III's Campaign in 853 B.C. c) Inscribed Slabs from J:lama

d) Campaigns of Shalmaneser III (848 B.C.), Ashurnasirpal II, and Adad-nirari III e) Inscriptions of Tiglath-pileser III f) Hieroglyphic Luwian Inscriptions

258 266

280 286 296

2. History .

299

XII. �oba or Bet-ReJ:tob.

319

I. Territory

319

2. History .

330

CONTENTS

XIII. Aram Damascus

7

347

1. Territory

348

2. History

367

XIV. Aramaeans in Babylonia .

409

I. I l th-9 th Centuries B.C.

409

2. The Chaldaeans

4 16

3. Aramaeans and North-Arabians .

422

4. Tii'manu and tfa�allii

425

5. Puqiidu

429

6. 'Utii or 'ltii'

437

7. Rupii' and Lab(a)dudu .

439

8. Tiglath-pileser III's List

44 1

9 . tJamaranu and Lu!}ii'atu

442

10. Rabbi-'11, Na�ir, Guliis.

446

1 1. Nabatu or Nabayatu.

448

12. Ra!}1qu, Kap1ri, Rummiilutu

450

13. 'Adinye, Gibr1ye, 'Ubiidu

452

14. Guriimu

453

15. tJindaru

455

16. Dunanu, Nilqu, Rad1ye

458

17. 'Ubiilu .

460

18. Karma'

461

19. Amlatu and Damiinu

462

20. Ru'a' and Qab1' .

464

2 1. Li'itta'u

467

22. Marusu, 'Ammatu, ljaggaranu

468

8

CONTENTS

23. Naqri and TanTye

470

25. tfallatu, Ya'as-'11, Yadaqqu, Malagu

479

24. Gambiilu

472

26. Gurasimmu, Udda, Gurru, 'Ubayanat, Dagga, Yakimanu 482 27. Synthesis .

xv.

Nomadism, Royalty, Dignitaries

1. Nomadism and Royalty

485 491 491

2. Dignitaries

500

XVI. Society and Economy .

515

1. Agriculture and Stock-Raising

516

2. Exploitation of Forests.

524

4. Production of Luxury Goods .

531

3. Urban Society 5. Metalwork 6. Textiles

7. Trade

8. Hired Labour Force 9. Women

XVII.Law 1. Laws

526

535

539 543 551

554 557 557

2. Judicial Records .

565

4. Conveyances

575

3. Law of Persons

5. Obligations a) Loans

b) Leases

c) Delivery and Work Contracts

569

580 581 586 591

CONTENTS

6. Receipts.

7. International Treaties XVIII. Religion 1. Cult of Betyls .

2. Ancestral Cult .

3. 'Attar.

9

593 595

599

599 605 607

4. 'II and His Hypostases

614

6. Moon-god

620

5. Resheph and Ru$a 7. Sun-deity

8. Storm-god

9. Burial Practices

617

623

626 636

GENERAL INDEX

641

BIBLICAL SOURCES

689

LIST OF TEXT FIGURES

695

LIST OF MAPS

697

PREFACE The aim of the present work is to propose a comprehensive synthesis of the ancient Aramaean history down to the incorporation of the Ara­ maean states in the Neo-Assyrian Empire, which in turn was progres­ sively Aramaized. Its purpose is thus to investigate the history of the Aramaean states and tribes from the obscure last third of the secortd mil­ lennium B.C. to the 8th and 7th centuries B.C., when the Aramaeans lost their independence. In order to present the Aramaean history during some six centuries, it was necessary to examine a wide range of sources. Besides the written material, like Old Aramaic inscriptions, cuneiform and hieroglyphic Luwian texts, Hebrew, Aramaic, Syriac, Arabic, Greek, and Latin sources of various types, one should mention the archaeological and geographical evidence as well. The preparation of the present work has started with a course on ancient Aramaean religion, taught in 1969-70 and followed by a series of articles on subjects related to ancient Ara­ maean history. This work was interrupted for more than a decade in 1981 owing to other activities, but mainly in view of the expected pub­ lication of the Royal Inscriptions of Mesopotamia from the Assyrian Periods (RIMA) by A. K. Grayson, and of The Inscriptions of Tiglath­ pileser Ill by H. Tadmor. The discovery of new inscriptions, like those from Tell Fagariya and Tell al-Qaqi, and an improved understanding of other texts changed - sometimes in a radical way - the author's con­ ception of the early Aramaean history. Since this study was first undertaken some valuable research on the ancient Aramaean history has appeared. However, the approach of authors dealing with the subject was either narrower than the intent of the present work or proceeding from a point of view which was exter­ nal to the Aramaean world. Thus, the work Die Beziehungen Altisraels zu den aramiiischen Staaten in der israelitisch-judiiischen Konigszeit (Frankfurt 1989), by G.G.G. Reinhold, aims at examining the relations of Israel and of Juda with the neighbouring Aramaean states, while the excellent study by J. Kah-jin Kuan, Neo-Assyrian Historical Inscrip­ tions and Syria-Palestine (Hong Kong 1995), analyzes the Neo-Assyr­ ian sources in order to establish their contribution to our knowledge of the history of the Levant. A similar approach is found in the study of

12

PREFACE

S. Ponchia, L'Assiria e gli stati transeufratici nella prima meta dell 'Vlll sec. a. C. (Padova 1991 ), which is limited to the 8th century B.C. The work by H. Sader, Les £tats arameens de Syrie depuis leur fondation jusqu 'a leur transformation en provinces assyriennes (Beirut

1987), is centered on the Aramaean states of Syria, but its main scope is the presentation of their political history. Decidedly wider is the approach of P.-E. Dion, Les Arameens a /'age du fer: histoire politique et structures sociales (Paris 1997). However, questions of geographical history and literary criticism of sources are only dealt with occasion­ ally, while religion, presented thirty years ago by J. Hoftijzer, Religio Aramaica (Leiden 1968), and even the history of the Aramaeans in Babylonia are completely left aside. Special attention will be paid in the present work to linguistic features of the available evidence. Ety­ mological, phonological, and morphological data can indeed help resolving also historical questions. Our discussion will focus on political and geographical history, but particular chapters will be devoted to institutions, economy, and reli­ gion. The geographical area with which we shall be concerned is rather large, since it comprises Syria, southeastem Anatolia, and Mesopotamia. Our procedure for this investigation will be as follows. In the first chap­ ter, we shall examine Aramaean pre-history and proto-history. The sec­ ond chapter shall deal with the question of the alleged relationship between the Hebrew forefathers and the ancient Aramaeans. Then, in chapters III to XIII, we shall give a relatively accurate description of the territory of each historically attested Aramaean group or state from the Lower ljabiir and the Middle Euphrates to the Bega' Valley and the Oasis of Damascus, and we shall present a detailed narrative of political events in each group or state. The situation of the Aramaean tribes in Babylonia will be considered in chapter XIV, also in relation to the Chaldaeans and to the North-Arabian tribes. Chapter XV shall examine the institutions of the Aramaean states on the background of their nomadic origins. Chapter XVI will deal with the economy, chapter XVII, with legal practices, and chapter XVIII, with the religion of the ancient Aramaeans. The book will conclude with a general index and an index of biblical sources, which are not easily retraceable by means of the general index. The present book attempts to take us back to the beginnings. We see the Aramaeans as a nation that represents one of our cultural ancestors, as one of the points of departure for us in the West. We think it is nec­ essary to consider how the impulses of the western civilization origi-

PREFACE

13

nated in the Middle East, although the currently accessible sources are noted for their paucity, partiality, and obscurity. On the basis of these sources, we cannot expect to reach decisive or incontestable results in this research. Previous research in this field has been done generally in circumscribed areas, mainly in relation to the biblical accounts which are no primary historical sources but literary compositions. The underlying sources of the biblical text may sometimes be detected, but the actual compilation, adjusted by several redactional layers, cannot serve as frame for the reconstruction of the Aramaean history, not even that of Bet-Ret:iob and of Damascus which had common borders with Israel. On the other hand, the Aramaean history of the 9th and 8th centuries stands in the shadow of Assyria's expansion. Moreover, it is dominated by the Assyrian perspective, given that the chronology and the most useful information is provided by the Assyrian royal sources. The resulting pic­ ture is understandably partial and it can be corrected only by an increased number of epigraphical non-Assyrian sources and by more methodically conducted excavations. Yet, this study should succeed in exposing the present state of our knowledge as a prelude to future efforts, hopefully supported by more material that excavations and other research might make available. Some particular problems concern any book on the history of the Ara­ maeans, since it has to face the spelling of personal names and of place names. Alphabetic and cuneiform orthography were unable to indicate all the phonemes of the early Aramaic language. A reconstruction of the early pronunciation was often preferred here to a mechanical transcrip­ tion 1, unless non-Aramaic forms have become familiar to the English reader. These will retain the generally accepted English form. The dis­ tinction of macron and circumflex (e.g. a and a) in the notation of long vowels has been avoided, since there is no indication that this practice corresponds to actual phonemic reality2 . Therefore, it seems best to retain a single length indicated by a macron. As for the modern Arabic place names, their transcription generally follows the written form with the classical voweling, although the colloquial pronunciation is often 1 Neither the partly conventional transcription of PNA was followed, also because it is not fully consistent, as shown e.g. by '"A!tar", transcribed with!, which never appears in Aramaic names. 2 As rightly observed by G. BuccELLATI, Akkadian, in R. HETZRON (ed.), The Semitic Languages, London-New York 1997, p. 69-99 (see p. 70). Besides, the circumflex is gen­ erally used in transcriptions of Biblical Hebrew and Aramaic to indicate the presence of a vowel letter in manuscripts. This does not always mean that the vowel is actually long and does not imply that length is caused by contraction, as in Assyriological practice.

14

PREFACE

somewhat different. Names familiar to the English reader retain the usual English form. In any historical study, chronology is an important factor. The chronology of the first millennium B.C. presents the advantage of being relatively well established, especially thanks to the Assyrian and Baby­ lonian evidence. Year dates in this study are given in terms of the Julian calender. A year cited simply as "853" stands often for 853/852, since the Assyro-Babylonian New Year fell close to the vernal equinox. In accordance with Mesopotamian custom, the regnal years listed for Assyrian and Babylonian kings are considered to begin with their first full year in office and exclude the accession year. The author wishes to acknowledge his debt to those who have helped him in setting up this complicated piece of printing on computer, in drawing the ground of the maps, and in preparing the indices. He also expresses his deep appreciation for the work accomplished by Peeters Publishers and the Orientaliste typography.

ABBREVIATIONS AAA

=

Annals of Archaeology and Anthropology, Liver­

AA(A)S

AASOR

= =

Annales Archeologiques (Arabes) Syriennes.

ABEL, Geographie

=

ABL

=

ACFP2

=

ActAnt ADAJ ADD

= = =

AfO, Beih.

AfO

AHw

= = =

AJA AJSL

= =

AnAnt

ALASP

= =

ANEP

=

ANET

=

AnOr

= =

AO.

AnSt

= =

AOAT AOS

= =

ANRW

pool.

Annuals of the American Schools of Oriental Research. F.M. ABEL, Geographie de la Palestine 1-11, Paris 1933-38. R.F. HARPER, Assyrian and Babylonian letters

belonging to the Kouyunjik Collections of the British Museum I-XIV, London-Chicago 1892-1914, as well as L. WATERMAN, Royal Correspondence of the Assyrian Empire 1-11, Ann Arbor 1930. Atti de/ I/ Congresso lnternazionale di Studi Fenici e Punici, Roma 1991. Acta Antiqua Academiae Scientiarum Hungaricae. Annual of the Department of Antiquities of Jordan. C.H.W. JOHNS, Assyrian Deeds and Documents 1-11, Cuneiform Texts, Cambridge 1898-1901. Archiv fur Orientforschung. Archiv fiir Orientforschung, Beiheft.

W. VON SODEN, Akkadisches Handworterbuch I-III, Wiesbaden 1965-81.

American Journal of Archaeology. American Journal of Semitic languages and Litera­ tures. Abhandlungen zur Literatur Alt-Syriens-Paliistinas.

Anatolia Antiqua (Bibliotheque de l'Institut Fram;:ais

d'Etudes Anatoliennes d'Istanbul). J.B. PRITCHARD, The Ancient Near East in Pictures relating to the Old Testament, 2nd ed., Princeton 1969. J.B. PRITCHARD (ed.), The Ancient Near Eastern Texts relating to the Old Testament, 3rd ed., Princeton 1969. Analecta Orientalia. H. TEMPORINI - W. HAASE (eds.), Aufstieg und Niedergang der romischen Welt, Berlin-New York.

Anatolian Studies.

Inventory numbers of the Ancient Near Eastern sec­ tion in the Louvre Museum. Alter Orient und Altes Testament. American Oriental Series.

16

ABBREVIATIONS

APAW.PH

=

APN

=

ARES

ARM(D ARRIM

= = =

ARU

=

AS

= =

BAH

BA

= =

BaM BASOR

= =

BE

=

BE.FAR

=

BiOr BJPES

= = = = = =

ASAE

BM.

BMB BN

BRINKMAN, PKB

Abhandlungen der Preussischen Akademie der Wis­ senschaften. Philosophisch-historische Klasse, Berlin. K.L. TALLQVIST, Assyrian Personal Names, Helsingfors 1914. Archivi Reali di Ebia. Studi. Archives royales de Mari (transcrites et traduites).

Annual Review of the Royal Inscriptions of Mesopotamia Project. J. KOHLER - A. UNGNAD, Assyrische Rechtsurkunden, Leipzig 1913. Assyriological Studies.

Annales du Service des Antiquites de l'Egypte. (The) Biblical Archaeologist.

Bibliotheque archeologique et historique de l'Institut Fran�ais d'Archeologie de Beyrouth.

Baghdader Mitteilungen. Bulletin of the American Schools of Oriental Research.

The Babylonian Expedition of the University of Pennsylvania. Series A: Cuneiform Texts. Bibliotheque des Ecoles Fran�aises d' Athenes et de Rome.

Bibliotheca Orientalis. Bulletin of the Jewish Palestine Exploration Society.

Inventory numbers of the British Museum.

Bulletin du Musee de Beyrouth. Biblische Notizen. J.A. BRINKMAN, A Political History of Post-Kassite Babylonia: ll58-722 B.C. (AnOr 43), Rome 1968. Bulletin of the School of Oriental and African Studies.

BSOAS

BT. BTAVO BTS BUNNENS, Syria

= = = = =

BZAW

=

CAD

=

The Assyrian Dictionary of the Oriental Institute of the University of Chicago, Chicago-Gliickstadt

CAH CBQ

CCCM

= = =

The Cambridge Ancient History, Cambridge. The Catholic Biblical Quarterly.

CCSL

=

Inventory numbers of the Balliwlit tablets. Beihefte zum Tiibinger Atlas des Vorderen Orients. Beiruter Texte und Studien. G. BUNNENS (ed.), Essays on Syria in the Iron Age (Ancient Near Eastern Studies. Supplement 7), Lou­ vain 2000. Beihefte zur Zeitschrift fiir die alttestamentliche Wissenschaft.

1956 ff.

Corpus Christianorum. Continuatio Mediaevalis, Tumhout. Corpus Christianorum. Series Latina, Tumhout.

ABBREVIATIONS

17

CHLI

=

CIG CIL CIS II

= = =

CRAI

=

csco CSHB CT

= = =

CTN II

=

CTN III

=

DaM DBS DCPP

= Damaszener Mitteilungen. = Dictionnaire de la Bible. Supplement, Paris 1 928 ff. = . E. LIPINSKI (ed.), Dictionnaire de la civilisation

DDD

=

DEB

=

DeZ DION, Arameens

= =

DJD

= =

DNWSI

DUSSAUD, Topographie EA

=

El EM EpAn

= = =

EPRO

J.D. HAWKINS, Corpus of Hieroglyphic Luwian Inscriptions I, Berlin, forthcoming. Corpus lnscriptionum Graecarum, Berlin 1 828-77.

Corpus lnscriptionum Latinarum. Corpus lnscriptionum Semiticarum. Pars II inscrip­ tiones Aramaicas continens, Paris 1 889 ff. Comptes rendus de l 'Academie des Inscriptions et Belles Lettres.

Corpus Scriptorum Christianorum Orientalium. Corpus Scriptorum Historiae Byzantinae

Cuneiform Texts from Babylonian, etc., Tablets in the British Museum. J.N. P0STGATE, The Governor 's Palace Archive

(Cuneiform Texts from Nimrud II), London 1 973. S. DALLEY - J.N. PosTGATE, The Tablets from Fort Shalmaneser (Cuneiform Texts from Nimrud III), London 1 984.

phenicienne et punique, Tumhout 1 992.

K. VAN DER T00RN - B. BECKING - P.W. VAN DER HORST (eds.), Dictionary of Deities and Demons in the Bible, Leiden 1 995. Dictionnaire encyclopedique de la Bible, Tumhout 1 987. Inventory numbers of the Dayr az-Zawr Museum. P.-E. DION, Les Arameens a / 'age du fer: histoire politique et structures sociales (Etudes bibliques, n.s. 34), Paris 1 997. Discoveries in the Judaean Desert. J. HOFTIJZER - K. JONGELING, Dictionary of the North-West Semitic Inscriptions, Leiden 1 995. = R. DussAUD, Topographie historique de la Syrie antique et medievale (BAH 4), Paris 1 927. The El-Amama tablets numbered according to J.A. KNUDTZ0N, Die El-Amarna-Tafeln (VAB 2), Leipzig 1 9 1 5 ; A.F. RAINEY, El Amarna Tablets 359-379 (AOAT 8), 2nd ed., Kevelaer-Neukirchen-Vluyn 1 978; W.L. MORAN, Les lettres d 'El Amarna (LAPO 1 3), Paris 1 987; ID., The Amarna Letters, Baltimore 1 992. Encyclopaedia Judaica I-XVI, Jerusalem 197 1 -72. 'En�lqlopedlyii miqrii 'lt I-IX, Jerusalem 1 950-88. Epigraphica Anatolica.

Etudes preliminaires aux religions orientales dans !'Empire romain.

ABBREVIATIONS

18

ESE

Erls

= =

ES/ EVO

= =

Excavations and Surveys in Israel. Egitto e Vicino Oriente.

FGH

=

FHG

=

F. JACOBY (ed.), Fragmente der griechischen Historiker, Berlin-Leiden 1 923-58. C. MOLLER (ed.), Fragmenta Historicorum Graecorum, Paris 1 84 1 -70.

GAG

=

GCS

= = =

GGA GGM

HARDING, Arabian Names

Eretz-Israel. M. LIDZBARSKI, Ephemeris fur semitische Epigraphik I-III, Giessen 1 900- 1 5 .

W. VON SODEN, Grundriss der akkadischen Grammatik, Roma 1 952; Ergiinzungsheft, Roma 1 969 (3rd

ed. with W.R. MAYER, Roma 1 995). Die griechischen christlichen Schriftsteller.

Gottingische gelehrte Anzeigen.

C. MOLLER (ed.), Geographi Graeci Minores 1-11, Paris 1 855-6 1 . = G . LANKESTER HARDING, An Index and Con­ cordance of Pre-Islamic Arabian Names and Inscriptions (Near and Middle East Series 8),

Toronto 1 97 1 .

HistOcc

=

HistOr

=

HTR HUCA

= =

Recueil des Historiens des Croisades. Historiens occidentaux, Paris 1 844- 1 895. Recueil des Historiens des Croisades. Historiens orientaux: Historiens arabes, Paris 1 872- 1 906. Harvard Theological Review. Hebrew Union College Annual.

ICC

=

The International Critical Commentary.

JG

IEJ

= =

IGLS

=

IM

=

/OS

=

Israel Oriental Studies.

= =

Journal Asiatique. Journal of the American Oriental Society. M. JASTROW, A Dictionary of the Targumim, the Tal­ mud Babli and Yerushalmi, and the Midrashic Liter­ ature 1-11, New York 1 886- 1 903 . Journal of Biblical Literature.

/CUR

IM.

JA JAOS

JASTROW

JBL

=

=

=

lnscriptiones Christianae Urbis Romae septimo saeculo antiquiores, Roma 1 862- 1 956. Israel Exploration Journal. Inscriptiones Graecae, Berlin 1 873 ff.; Editio minor

(/02), Berlin 1 924 ff.

Inscriptions grecques et latines de la Syrie, Paris 1 929 ff.

lstanbuler Mitteilungen.

Inventory numbers of the Iraq Museum, Baghdad.

ABBREVIATIONS

JCS JEOL

= =

JESHO

=

JKF JNES JPOS JRAS JSS

=

K.

=

KAI

=

KBV

=

KBo

=

KTU

=

KUB

=

LAPO

=

= =

=

19

Journal of Cuneiform Studies. Jaarbericht van het Vooraziatisch-Egyptisch Genoot­ schap "Ex Oriente Lux". Journal of the Economic and Social History of the Orient. Jahrbuchfar kleinasiatische Forschung. Journal of Near Eastern Studies. Journal of the Palestine Oriental Society. Journal of the Royal Asiatic Society. Journal of Semitic Studies. Inventory numbers of the Kuyundjik collection in the British Museum. H. DONNER - W. ROLLIG, Kanaaniiische und aramiii­ sche lnschriften, Wiesbaden 1 962-64 (3rd ed., 1 97 1 76). L. KOHLER - w. BAUMGARTNER et al., Hebriiisches und aramiiisches Lexikon 1-V, 3rd ed., Leiden 1 967-95. Keilschrifttexte aus Boghazkoi, Leipzig - Berlin 1 9 1 6 ff. M. DIETRICH - 0. LORETZ - J. SANMARTfN, The Cuneiform Alphabetic Texts from Ugarit, Ras lbn Hani and Other Places (KTU: second, enlarged edi­ tion), Miinster 1 995. Keilschrifturkunden aus Boghazkoi, Berlin 1 92 1 ff.

= =

Litteratures anciennes du Proche-Orient. E. LIPINSKI, Semitic Languages: Outline of a Com­ parative Grammar (OLA 80), Leuven 1 997.

MAD MAIBL

=

MAMA MARI MDOG

=

Materials for the Assyrian Dictionary. Memoires presentes par divers savants a l'Academie des Inscriptions et Belles-Lettres. Monumenta Asiae Minoris antiqua, London 1 928 ff. Mari. Annales de recherches interdisciplinaires. Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin. Materiali epigrafici di Ebia. A.R. MILLARD, The Eponyms of the Assyrian Empire: 910-612 BC (State Archives of Assyria Studies 2), Helsinki 1 994. Melanges de l 'Universite Saint-Joseph, Beyrouth. Mitteilungen der Vorderasiatisch-Agyptischen Gesellschaft.

LIPINSKI,

MEE

Semitic

MILLARD,

=

=

=

Eponyms

= =

MUSJ MVAG

= =

N. NABU ND.

= = =

Inventory numbers of the Nippur excavations. Nouvelles assyriologiques breves et utilitaires. Inventory numbers of the Nimrud excavations.

ABBREVIATIONS

20

NEAEHL

=

NESE Nippur IV

= =

Not. Dign. Or.

=

0.

=

OA

OTZEN, Hama

= = = = = = = =

PARP0LA, Toponyms

=

PBS

=

PEQ PJ PNA

= = =

PRU

=

PSBA PW

= =

RA RAO

= =

RB RES

RGTC

= = =

RGTC 8

=

080

OIP OLA

OLP OLZ Or

E. STERN (ed.), The New Encyclopedia of Archaeo­ logical Excavations in the Holy Land, Jerusalem 1 993. Neue Ephemeris Jar semitische Epigraphik. St.W. COLE, Nippur IV. The Early Neo-Babylonian Governor's A rchive from Nippur (OIP 1 1 4), Chicago 1 996. Notitia dignitatum utriusque imperii, ed. by 0. SEECK, Berlin 1 876: Oriens. Inventory numbers of the Near Eastern collections in the Royal Museum of Art and History, Brussels.

Oriens Antiquus.

Oriens Biblicus et Orientalis. Oriental Institute Publications. Orientalia Lovaniensia. Analecta.

Orientalia Lovaniensia. Periodica. Orientalistische Literaturzeitung. Orientalia. Nova series. B. OTZEN, The Aramaic Inscriptions, in P.J. Rus M.L. BUHL (eds.), Hama IU2. Les objets de la perio­ de dite syro-hittite (age du Fer), Kjijbenhavn 1 990, p. 267-3 1 8. S. PARP0LA, Neo-Assyrian Toponyms (AOAT 6), Kevelaer - Neukirchen - Vluyn 1 970. University of Pennsylvania, the Museum. Publica­ tions of the Babylonian Section.

Palestine Exploration Quarterly. Paliistinajahrbuch. K. RADNER (ed.), The Prosopography of the Neo­ Assyrian Empire 1/1 -2, Helsinki 1 998-99; H.D. BAKER (ed.), The Prosopography of the Neo-Assyrian Empire 11/1, Helsinki 2000. J. NouGAYR0L / Ch. VIR0LLEAUD, Le Palais royal d'Ugarit II-VI, Paris 1 955-70. Proceedings of the Society of Biblical Archaeology. A. PAULY - G. W1ssowA et al. (eds.), Real-Ency­ clopiidie der classischen Altertumswissenschaft. Revue d'assyriologie et d'archeologie orientale. Ch. CLERM0NT-GANNEAU, Recueil d 'archeologie orientale I-VIII, Paris 1 885- 1 92 I . Revue biblique. Repertoire d'epigraphie semitique, Paris 1 905 ff.

Repertoire geographique des textes cuneifontJes, Wiesbaden 1 974 ff. R. ZAOOK, Geographical Names according to New and Late Babylonian Texts (RGTC 8), Wiesbaden 1 985.

ABBREVIATIONS

RHA RHR RIDA RIMA I

= = = =

RIMA II

=

RIMA III

=

R/MB II

=

Rf.A

=

RS.

=

RSF

= = =

RSO

RTP

21

Revue hittite et asianique. Revue de l 'histoire des religions. Revue internationale des droits de l 'Antiquite. A.K. GRAYSON, Assyrian Rulers of the Third and Second Millennia BC (to J I 15 BC) (Royal Inscrip-

tions of Mesopotamia. Assyrian Periods I ), Toronto 1987. A.K. GRAYSON, Assyrian Rulers of the Early First Millennium BC I: 1 1 14-859 BC (Royal Inscriptions of Mesopotamia. Assyrian Periods 2), Toronto 1991. A.K. GRAYSON, Assyrian Rulers of the Early First Millennium BC II: 858-745 BC (Royal Inscriptions of Mesopotamia. Assyrian Periods 3), Toronto 1996. G. FRAME, Rulers of Babylonia: From the Second

Dynasty of /sin to the End of Assyrian Domination: 1 157-612 BC (Royal Inscriptions of Mesopotamia.

Babylonian Periods 2), Toronto 1995.

Reallexikon der Assyriologie und vorderasiatischen Archiiologie. Inventory numbers of the Ras Shamra-Ugarit excavations.

Rivista di Studi Fenici. Rivista degli Studi Orientali.

H. INGHOLT - H. SEYRIG - J. STARCKY - A. CAQUOT,

Recueil des tesseres de Palmyre (BAH 58), Paris 1955.

SAA I

=

SAA II

=

SAA III

=

SAA IV

=

SAA V

=

SAA VI

=

SAA VII

=

SAA X

=

S. PARPOLA, The Correspondence of Sargon //, Part / (State Archives of Assyria I), Helsinki 1987. PARPOLA - K. WATANABE, Neo-Assyrian Treaties and Loyalty Oaths (State Archives of Assyria 11), Helsinki 1988. A. LIVINGSTONE, Court Poetry and Literary Miscellanea (State Archives of Assyria III), Helsinki 1989. I. STARR, Queries to the Sungod (State Archives of Assyria IV), Helsinki 1990. G.B. LANFRANCHI - S. PARPOLA, The Correspondence of Sargon //, Part /I (State Archives of Assyria V), Helsinki 1990. Th. KWASMAN - S. PARPOLA, Legal Transactions of the Royal Court of Nineveh, Part I (State Archives of Assyria VI), Helsinki 1991. F.M. FALES - J.N. PosTGATE, Imperial Administrative Records, Part I (State Archives of Assyria VII), Helsinki 1992. S. PARPOLA, Letters from Assyrian and Babylonian Scholars (State Archives of Assyria X), Helsinki 1993.

s.

ABBREVIATIONS

22

SAA XI

=

SAA XII

=

SAA XIII

=

SAAB Sabaic Dictionary

= =

SADER, £tats arameens Safaitic Cairns

=

SATO I

=

SAID II

=

Sanherib

=

SAOC Sargon II

= =

Ann. Bull Cyl. Displ. Pavem. Rei. R. XIV SC SEL

= = = = = = = = =

Sennacherib

=

SLA

=

SOAW. PH

=

SPAW

=

F.M. FALES - J.N. PosTGATE, Imperial Administra­ tive Records, Part II (State Archives of Assyria XI), Helsinki I 995. L. KATAJA - R. WHITING, Grants, Decrees and Gifts of the Neo-Assyrian Period (State Archives of Assyria XII), Helsinki 1 995. St.W. COLE - P. MACHINIST, Letters from Priests to the Kings Esarhaddon and Assurbanipal (State Archives of Assyria XIII), Helsinki 1 998. State Archives of Assyria Bulletin. A.F.L. BEESTON - M.A. GHUL - W.W. MOLLER - J. RYCKMANS, Sabaic Dictionary I Dictionnaire sabeen, Louvain-la-Neuve - Beyrouth 1 982. = H.S. SADER, Les £tats arameens de Syrie depuis leur fondation jusqu 'a leur transformation en provinces assyriennes (BTS 36), Beirut 1 987. F.V. WINNETT - G.L. HARDING, Inscriptions from Fifty Safaitic Cairns (Near and Middle East Series 9), Toronto 1 978. E. LIPINSKI, Studies in Aramaic Inscriptions and Onomastics I (OLA I ), Leuven 1 975. E. LIPINSKI, Studies in Aramaic Inscriptions and Onomastics II (OLA 57), Leuven 1 994. E. FRAHM, Einleitung in die Sanherib-lnschriften (AfO, Beih. 26), Wien 1 997. Studies in Ancient Oriental Civilization. A. FUCHS, Die Jnschriften Sargons II. aus Khorsabad, Gottingen 1 994. The Annals. Bull Inscriptions. Cylinder Inscriptions. Display Inscriptions. Pavement Inscriptions. Inscriptions on Reliefs. The Display Inscription of Room XIV. Sources Chretiennes, Paris. Studi Epigrafici e Liguistici sul Vicino Oriente Antico. D.D. LUCKENBILL, The Annals of Sennacherib (OIP 2), Chicago 1 924. R.H. PFEIFFER, State Letters of Assyria (AOS 6), New Haven 1 935. Sitzungsberichte der Osterreichischen Akademie der Wissenschaften in Wien. Philosophisch-historische Klasse. Sitzungsberichte der Preussichen Akademie der Wissenschaften.

ABBREVIATIONS

23

TAD I-IV

=

TCL TCS

TGI

= = =

Th WA T I-VIII

=

Textes cuneifonnes du Louvre. Texts from Cuneiform Sources. K. GALLING (ed.), Textbuch zur Geschichte Israels, 2nd ed., Tiibingen 1968. J. BOTIERWECK - H. RINGGREN - H.-J. FABRY (eds.),

Tigl. Ill

=

Stuttgart 1970-95. H. TADMOR, The Inscriptions of Tiglath-pileser Ill,

TM.

TPR TRE

= = =

TSSI I-III

=

Inventory numbers of the Tell Mardikh excavations. Theologische Realenzyklopiidie, Berlin-New York 1977 ff. J.C.L. GIBSON, Textbook of Syrian Semitic Inscrip­ tions I. Hebrew and Moabite Inscriptions, 2nd ed., Oxford I 973; II. Aramaic Inscriptions, Oxford 1975; III. Phoenician Inscriptions, Oxford 1982.

UET

=

Ur Excavations. Texts.

VAB VAS

=

VT

=

Vorderasiatische Bibliothek. Vorderasiatische Schriftdenkmaler der Koniglichen/ Staatlichen Museen zu Berlin.

WA

=

WILD, Ortsnamen

=

WVDOG

WO

= =

WZKM

=

YOS

=

Yale Oriental Series.

ZA

=

ZADOK, WSB

=

Zeitschrift fur Assyriologie und vorderasiatische A rchiiologie. R. ZADOK, On West Semites in Babylonia during the Cha/dean and Achaemenian Periods: An Onomastic Study, 2nd ed., Jerusalem 1 978.

UF

=

=

B. PORTEN - A. Y ARDEN!, Textbook of Aramaic Doc­ uments from Ancient Egypt I. Letters, Jerusalem 1986; II. Contracts, Jerusalem 1989; III. Literature, Accounts, Lists, Jerusalem 1993; IV. Ostraca and Assorted Inscriptions, Jerusalem 1999.

Theologisches Worterbuch zum Alten Testament,

King of Assyria. Critical Edition with Introduction, Translation, and Commentary, Jerusalem 1994. Terqa Preliminary Reports.

Ugarit-Forschungen.

Vetus Testamentum.

Siglum of Western Asiatic Antiquities, British Museum, London. S. WILD, Libanesische Ortsnamen (BTS 9), Beirut 1973.

Die Welt des Orients.

Wissenschaftliche Veri:iffentlichungen tschen Orient-Gesellschaft.

der

Deu­

Wiener Zeitschrift fur die Kunde des Morgenlandes.

ABBREVIATIONS

24

ZAH ZAS

= =

ZA W ZDMG

=

ZDPV ZPE

=

=

=

Zeitschrift fiir Althebraistik. Zeitschrift fiir iigyptische Sprache und Altertums­ kunde. Zeitschrift fiir die alttestamentliche Wissenschaft. Zeitschrift der Deutschen Morgenliindischen Gesellschaft. Zeitschrift des Deutschen Paliistina-Vereins. Zeitschrift fiir Papyrologie und Epigraphik.

CHAPTER I ARAMAEAN PRE-HISTORY AND PROTO-HISTORY The Levant underwent significant changes and transformations between 1200 and 900 B.C., a period which corresponds to the end of the Late Bronze Age and to Iron Age I. The North-Syrian and Canaanite city-state system of the Bronze Age was then replaced by an ethno-polit­ ical structure in which the various regions of the Levant were inhabited by different peoples. This change was accompanied by the collapse of the Hittite empire, by a considerable shrinking of the Assyrian power basis, and by the evanescence of the Egyptian control in Syro-Phoenicia and in Canaan, with concomitant and widespread destructions of the urban centres. The consequence was the abrupt end of historical records provided by the cuneiform archives of Hattusha, Emar, and Ugarit, which were not replaced by a sufficient amount of reliable indigenous sources. The immediate cause of the momentous changes in the Levant seem to have been the large-scale migrations that occurred at the end of the Bronze Age and were probably brought about by a severe and pro­ tracted famine in Anatolia. Syria and Palestine were affected directly by the arrival of displaced "Sea Peoples", that were moving along the Mediterranean coast, and by the migrations of the Aramaeans, who spread inland all over the Fertile Crescent where they had already occu­ pied steppes, high grounds, and fringes of arable land, apparently with­ out being noticed by the sedentary and literate populations of the region. I.J. Gelb could state in 196 1 that, " staying on safe historical ground, the first real reference . . . to the Aramaeans" is to be found " in the inscriptions of Tiglath-pileser 1" 1 • However, the occurrences of the name Aram in cuneiform texts of the second part of the third millennium and in those of the early and mid-second millennium B.C., as well as the sys­ tematic use of the appellation At,zlamu to designate the Aramaeans in the first millennium B.C., repeatedly raised the question of the Aramaean prehistory and of the Aramaean origins. Besides the early attestations of the name Aram, the supposed relations between the Aramaeans and the Alzlamu, the Sutu, and the Chaldaeans have to be dealt with. Also the 1 I.J. GELB, The Early History of the West Semitic Peoples, in JCS 1 5 ( 1 96 1 ), p. 27-47 (see p. 28, n. 5).

26

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

biblical traditions and the alleged Aramaean emigration from Qir belong to this range of problems.

1. EARLY OCCURRENCES OF ARAM

A toponym A-ra-muki is already attested in the third millennium B.C. at Ebla2 , in a list of geographical names first known from Tell Abu �alabI!} and partly attested also among the lexical texts from Uruk. These versions of the list must derive from an older source which was probably composed in Mesopotamia3, in a period when the often alleged Hurrian origin of the toponym4 can probably be dismissed. The place name ought to be considered as pre-Semitic, pre-Sumerian, and pre-Hur­ rian, just as many geographical names of Northern Syria in the third and second millennia B.C. that reveal a protopopulation of unknown linguis­ tic affiliation, neither Semitic nor Hurrian5 . Since A-ra-meki is apparently the genitive of A-ra-muki, the same place name may occur in a date for­ mula from the reign of Naram-Sin, found on two tablets from Tell ljafage6 :

2 G. PETTINATO, L 'atlante geografico de/ Vicino Oriente antico attestato ad Ebia e ad Abu Saliibikh, in Or 47 ( 1978), p. 50-73 and Pl. VII-XII (see p. 70, No. 233); ID., Testi /essicali monolingui de/la biblioteca L. 2769 (MEE 3 ), Napoli 198 1, p. 224, No. 56, XI, 4; cf. p. 237, No. 234. Also similar place names, as Arimuki, Arramuki, Ar- 'a-muki, occur in documents from Ebia, but they cannot be reasonably linked with the Aramaeans. Ref­ erences to the texts are collected in A. ARCHI - A. PtACENTINI - F. POMPONIO, / nomi di luogo dei testi di Ebia (ARES II), Roma 1993, p. I 09- 1 10, 150- 1 5 1, 168- 169, while ref­ erences and an attempt at locating the places, at least approximately, can be found in M. BONECHI, / nomi geografici dei testi di Ebia (RGTC 12/ 1 ), Wiesbaden 1993, p. 45-46, 505 1 , 56. Besides the area of Ebia itself, proposed for Arramuki , the region of the Middle Euphrates, between Carchemish and Emar, is taken into account by the author, who quotes earlier suggestions as well. 3 R. BIGGS, The Ebia Tablets. An Interim Perspective, in BA 43 ( 1980), p. 76-87 (see p. 84-85); ID., Ebia and Abu Salabikh, in L. CAGNI (ed.), La lingua di Ebia, Napoli 198 1, p. 12 1- 133 (see p. 130-132). 4 Thus, recently, R. ZADOK, Elements of Aramean Pre-History, in M. COGAN - I. EPH'AL (eds.), Ah, Assyria . . . Studies in Assyrian History and Ancient Near Eastern His­ toriography Presented to Hayim Tadmor (Scripta Hierosolymitana 33), Jerusalem 199 1, p. 104- 1 17 (see p. 106). 5 Cf. I.J. GELB, art. cit. (n. I ), p. 39-40; ID., Ebia and the Kish Civilization, in L. CAGNI (ed.), La lingua di Ebia, Napoli 198 1, p. 9-73 (see p. 64-66). 6 I.J. GELB, Sargonic Texts from the Diyala Region (MAD I ), Nos. 2 1 7 and 224. See also D.R. FRAYNE, Nariim-Sin. A, in Rl.A IX/3-4, Berlin 1999, p. 1 69-174, in particular p. 173, §4. 13.4.

EARLY OCCURRENCES OF ARAM

in I

Kh. 1934, 22

in [ I

Kh. 1934, 25

27

dNa-ra-am-dE[N.ZU]

dNa-ra-am-[dEN.Z]U

"In the year (when) Naram-Sin

. .. ..... Ba-ba ENSf Si-mu-ur4-ri-im ki Dub-ul ENSf [A-r]a-meki ik-mi-ME

······ Ba-ba ENSf Si-mu-ur4 [Dub]-ul ENSf A-ra-[meki]

captured Baba, the ruler of Simurrum, (and ) Dubul, the ruler of Aramu".

MU

MU]

ik-mi-u

Simurrum, later S imurrum, should be located somewhere between Altin Koprii, on the Lesser Zab, and the Diyiila river, most likely on the (A.b-i) Sirwiin, as upper Diyiila is called7 • The Sirwiin river seems indeed to have been named after the ancient city of Simurrum, a neighbour to the land of Lullubi, which lay in the district of Sulaimaniyya, as indi­ cated by the rock-reliefs of Naram-S1n8 • The city of Aramu has to be looked for probably in the same region, also on the Diyiila river, but south of the point where the river broke through the Gebel I:Iamrin9• Aram was believed to be mentioned also in a copy of a Naram-S1n' s inscription published in 19 1 1 by Fr. Thureau-Dangin and widely dis­ cussed by scholars 10 until E. Sollberger made it clear, in 1970 , that EN A-ra-am was no "lord of Aram" but the preterite en 'aram of the Old 7 D.R. FRAYNE, On the Location of Simurrum, in G.D. YOUNG - M.W. CHAVALAS - R.E. AVERBECK (eds.), Crossing Boundaries and Linking Horizons. Studies in Honor of M. C. Astour, Bethesda 1 997, p. 243-269. See also E. F. WEIDNER, Simurrum und 'Zaban, in AfO 1 5 ( 1 945-5 1 ), p. 75-80; A. GoITZF., ljulibar of Duddul, in JNES 1 2 ( 1 953), p. 1 1 4- 1 24 (see p. 1 20 and 1 23); W. W. HALLO, Simurrum and the Hurrian Frontier, in RHA 36 ( 1 978 ( 1 980]), p. 7 1 -83, with map. M S. SMITH, Early History of Assyria, London 1 928, p. 96-97; E.A. SPEISER, Southern Kurdistan in the Annals of Ashurnasirpal and To-day (AASOR 8), New Haven 1 928, p. 8 and map; C.J. EDMONDS, Kurds, Turks, and Arabs, London 1 957, p. 360 and map fac­ ing p. 440; lo., Some Ancient Monuments on the Iraqi-Persian Border, in Iraq 28 ( 1 966), p. 1 59- 1 63 and Pl. XLV-XLVIII (see Pl. XLVb). 9 D.R. FRAYNE, The Early Dynastic List of Geographical Names (AOS 74), New Haven 1 992, p. 70; lo., art. cit. (n. 7), p. 263. IO F. THUREAU-DANGIN, Une inscription de Naram-Sin, in RA 8 ( 1 9 1 1 ), p. 1 99-200. Thureau-Dangin regarded his interpretation as partly uncertain and warned against a too rush identification of A-ra-am with the Aramaeans. E. DHORME, Abraham dans le cadre de l 'histoire, 2nd art., in RB 37 ( 1 928), p. 48 1 -5 1 1 (see p. 487-488 = Recueil Edouard Dhorme, Paris I 95 1 , p. 2 1 9-220), did not resist the temptation of identifying this prob­ lematic A-ra-am with the Aramaeans. See also B. HROZNY, Naram-Sin et ses ennemis d'apres un texte hittite, in Archiv Orientalni I ( 1 929), p. 65-76, in particular p. 75-76, and A. DUPONT-SOMMER, Sur les debuts de l 'histoire arameenne, in Congress Volume: Copenhagen 1 953 (VTS I ), Leiden 1 953, p. 40-49. Instead, I.J. GELB, Inscriptions from Alishar and Vicinity (OIP 27), Chicago 1 936, p. 6, n. 60, and J. LEWY, Tabor, Tibar, Atabyros, in HUCA 23/1 ( 1 950-5 1 ), p. 357-386 (see p. 368-376), have identified A-ra-am with Armanum.

28

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

Akkadian verb na 'iirum, "to smite", "to strike", with the morpheme - am of the ventive 1 1 • In consequence, the much discussed passage of this text should be read [dNa-ra-am-d]EN.ZU LUGAL [ki-ib-ra-t]im ar-ba-im i­ nu Jjur-sa-ma-atki en-a-ra-am it AM in qdb-ld-ni Di-ba-ar sa-du-im su4ma u-sa-am-q{- it-su, "Naram-Sin, king of the four (world's) quarters, when he had smitten ljursamat and he himself had killed a wild bull in the middle of the mountain Dibar .. . " The location of the land or town of ljursamat and of Mount Dibar is unknown, because it is unlikely that ljursamat has something in common - apart its name - with the North-Anatolian city of ljursamma 1 2• However, a city A-ra-miki is mentioned after As-nunki , i.e. Esnunna, present-day Tell Asmar, in a Sumerian administrative document from the Ur-III period, which lists animals brought to the central cattle-park at Puzris-Dagan, near Nippur, at the site now called Drehem 1 3 , while an Old Babylonian tablet refers to Ar-ra-muki in Subartu 14• Since Esnunna lies east of the Tigris, while the Subarians of that time probably lived in the east-Tigridic foothills of the Kurdish mountains, A-ra-miki and possibly Ar-ra-muki must be the same city as the A-ra-meki mentioned with Simur­ rum at the time of Naram-Sin. In any case, these texts clearly refer to a town and A-ra-meki was governed in the 22nd century B.C. by an ens{ or issi 'akkum, while most Aramaeans of the 12th- 1 l th centuries B.C. were still half-nomadic tribesmen of the Gazira and of the Syrian steppe. There can hardly be any relation between these various ethnic entities. Nevertheless, the mention of the town of A-ra-miki in the Ur-III docu­ ment and the almost contemporaneous reference to an employee called A-ra-mu in another text from Puzris-Dagan 1 5 were considered in some quarters as a sufficient evidence of Aramaean presence in the east­ Tigridic region of Mesopotamia 1 6• However, a distinction should be I I E. SOLLBERGER, Princes fantomes, in RA 64 ( 1970), p. 173- 1 74; ID. - J.-R. KUPPER, Inscriptions royales sumeriennes et akkadiennes (LAPO 2), Paris 1 97 1 p. 108, § IIA4f. 1 2 H. OrrEN, ljursam(m)a, in Rf.A IV, Berlin-New York 1972-75, p. 521b. 13 The text was published by P. DEIMEL, Miszellen, in Orientalia 2 ( 1920), p. 62 (Wen­ gler 22). Cf. D.O. EDZARD - G. FARBER, Die Orts- und Gewdssernamen der Zeit der 3. Dynastie von Ur (RGTC 2), Wiesbaden 1974, p. 1 5; D.R. FRAYNE, art. cit. (n. 7), p. 263. 14 J.J. FINKELSTEIN, Subartu and the Subarians in Old Babylonian Sources, in JCS 9 ( 1955), p. 1-7 (see p. 2). 15 C.E. KEISER, Cuneiform Bullae of the Third Millennium B.C. (Babylonian Records in the Library of J. Pierpont Morgan 3), New York 19 14, No. 159. 16 The relative cautiousness shown by N. SCHNEIDER, Aram und die Aramiier in Ur 111-Zeit, in Biblica 30 ( 1 949), p. 1 09- 1 1 1 , is conspicuously missing in the articles by S. MosCATI, Su/le origini degli Aramei, in RSO 26 ( 1951), p. 1 6-22, and A. DUPONT-SOM­ MER, art. cit. (n. I 0), p. 40, 43.

4000 m 3000 m 2000 m 1 500 m 1000 m 500 m 200 m 100 m O rn

Q. Nal_)l • •Tamad -

.,.

-l

,l

t

�l

"Aqaba

• Timna

-

�9.;

".J>..

0

• Palmyra

100

200

400

500 km

I . Middle East.

300

H. David de/.

P�rsian Gulf

Caspian Sea

�

30

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

made between the place name and the personal name, which shows no characteristic Semitic morpheme indicating the alleged derivation from a toponym, e.g. -i or -ay. On the contrary, the name seems to be based on an element appearing also in compound Hurrian anthroponyms and can therefore be considered as Hurrian. This conclusion can hardly be deduced from one-word names like A­ ra-am-mu 1 7, A-ra-am-me 1 8 , A-ra-ma 1 9, or A-ra-mu20, because the mean­ ing of aram- is unknown, but a name like A-ra-am-mu-su-ni2 1 is cer­ tainly Hurrian and means "Aram is the righteous one"22 • Considering the weakness of the r/1 phoneme in Hurrian and the occasional interchange r/1 in its notation23 , one can interpret the name of A-ra-am-ma-da-ra, presented as the earliest member of the first dynasty of Babylon24, also as Hurrian *Aram-adal-, "Aram is strong"25 • Besides, it seems that arim is a dialectal and perhaps a later variant of aram; it appears, among oth­ ers, in the same names A-ri-mu-su-[ni] 26 and in A-ri-im-a-da-a/27 . The weakness of r, which is sometimes elided altogether in the middle of a name, allows also for a Hurrian interpretation of the Old Babylonian name Mu-ti-a-ra-mu28, read Mutri-aramu in parallelism with Mu-ut-ri­ dTeJub(uh) and Mu-tara/i-STORM at Emar29 • The element 'arm vocalized 17 J.-M. DURAND, ARM XXI, Paris 1982, No. 399, 10; F. J0ANNES, Nouveaux memorandums, in Miscellanea Babylonica. Melanges offerts ii Maurice Birot, Paris 1985,

p. 97- 1 13 (see p. 101- 102: A 1 40 1 = No. 4, line 1 4): Mari (early 1 7th century); D. J. WISEMAN, The Alalakh Tablets, London 1953, p. 128b: names from Alalakh VII (early 16th century B.C.). 18 H. H. F!GULLA, Letters and Documents of the Old-Babylonian Period (UET 5), London 1953, No. 197, 4. 19 OJ. WISEMAN, loc. cit. (n. 17): name from Alalakh VII (early 16th century B.C.). 20 M. BIR0T, in ARM XVU I , Paris 1 979, p. 66: Mari (early 17th century B.C.). 21 D.J. WISEMAN, loc. cir. (n. 17): name from Alalakh VII. 22 E. LAROCHE, Glossaire de la langue hourrite II ( = RHA 35), Paris 1979, p. 1 73. 23 E.A. SPEISER, Introduction to Hurrian (AASOR 20), New Haven 1 94 1 , p. 27. 24 J.J. FINKELSTEIN, The Genealogy of the Hammurapi Dynasty, in JCS 20 ( 1 966), p. 95- 1 18 (see p. 96, line I ; cf. p. 98- 10 1 , I 14- 1 15). 25 E. LAROCHE, Glossaire de la langue hourrite I ( = RHA 34), Paris 1978, p. 35. The equation a-da-al-lu = ga-aI-ru, "strong", in a Babylonian list of synonyms shows that the meaning of adal- was known among the Assyro-Babylonian scribes. 26 O .J. WISEMAN, op. cit. (n. 17), p. 129b: from Alalakh IV ( 15th century B.C.). See also A-ri-mu-Ie (ibid., the same period). 27 M. BIROT, in ARM XVU I , p. 66: Mari, a slave (early 17th century B.C.). 28 L. LEGRAIN, Historical Fragments (PBS 13), Philadelphia 1922, No. 56, 4. 29 D. ARNAUD, Textes syriens de /'age du Bronze Recent, Sabadell-Barcelona 199 1, Nos. 20, 20; 36, 30; 72, 28; 76, 19; 84, I . I 0. Hieroglyphic seal impressions: H. GONNET, ibid., p. 202 (No. 36c), 206 (No. 76b), 207 (No. 84); etc. For the element mutri, cf. E. LAROCHE, op. cir. (n. 22), p. 17 1- 172, s. v. mudri, and see also mu-ti-be-eI­ Ia (ibid., p. 1 7 1 ).

EARLY OCCURRENCES OF ARAM

31

aram does not appear until now in Hurrian names from Ugarit, which belong to a later phase of the Hurrian onomastics, but the word 'arm occurs there in the alphabetic Hurrrian phrases '['u[k tzg 'arm Tf.P and tzg 'arm T£b '['u[k, which must mean "Sauska is the tzg 'arm of Tesub"30• Since the l 8th-l 7th centuries B.C. mark the very beginning of the Hurrian occupation of western Syria, the place name Ar-ra-muki attested on the territory of Ebia about the 23rd century B.C.3 1 must be brought in relation with the protopopulation of the country. However, we cannot suppose that people living in this region were speaking the same kind of language as the protopopulation of southern Kurdistan. Therefore, one might prefer to consider this Ar-ra-muki as a variant spelling of Ar- 'a­ mu ki , i.e. / 'Arhamu/, attested in the same area32 . As for Bn- 'Army33 , Ar-me-ia34, and eqlerme! a-ra-mi-ma35 in 13th cen­ tury texts from U garit, some authors were inclined to see in them a gen­ tilic meaning "Aramaean(s)"36, while others have rightly called this interpretation in question37 • In fact, Armeya is the patronymic of a man called Zi-di-ia, in one case, and, in the other, Armeya is a witness to a deed in which the two parties and all the witnesses bear Hurrian or Luwian names38 . In these circumstances, the proper name Armeya has to be explained either as a reduced form of the Hurrian component aramlarim or as the Luwian theophorous element Arma, "Moon god", both with the hypocoristic suffix -ya, which is well represented in Hur­ rian onomastics39• A similar explanation should be considered in the 3° KTU 1 . 1 48, 1 7 ; 1 . 1 49, 1 0. Cf. E. LAROCHE, in Ugaritica V, Paris 1 968, p. 5 1 7. The word tzg probably means "chief'; cf. ID., op. cit. (n. 22), p. 263, s. v. tdulji. ·1 1 See here above, p. 26, n. 2, in particular TM. 75.G. 1444, transliterated by G. PETTI­ NATO, Ebia, un impero inciso nel/ 'argilla, Milano 1979, p. 1 1 5. 32 TM. 75.G. 1 789, published by P. MANDER, Administrative Texts of the Archive L. 2769 (MEE I 0), Roma 1990, No. 3, I, 1 1 . .1.1 KTU 4.63, Ill, 22; 4.232, 9; 4.309, 1 0. See also Bn. 'Arm (KTU 4.232, 5) and 'Army (KTU 4.232, 7), probably the same village as 'Arm (KTU 4.750, 5). 34 J. NouGAYROL, PRU III, Paris 1 955, p. 35, RS. 1 5 .37, line 13; ID., Ugaritica V, Paris 1 968, p. 1 8 1 , No. 86, RS. 20. 1 76, line 25. 35 J. NouGAYROL, PRU III, p. 1 48, RS. 1 6. 1 78, line 1 0. 36 F. THUREAU-DANGIN, Une tablette bi/ingue de Ras Shamra, in RA 37 ( 1 940-4 1 ), p. 97- 1 1 8 (see p. 1 1 5, n. 6); A. DuPONT-SOMMER, art. cit. (n. 1 0), p. 46; J.-R. KUPPER, Les nomades en Mesopotamie au temps des rois de Mari, Paris 1 957, p. I 14; G.G.G. REINHOLD, Die Beziehungen Altisraels zu den aramiiischen Staaten in der israe/itisch­ judiiischen Konigszeit, Frankfurt a/M 1 989, p. 27. 37 Already M. LIVERANI, Storia di Ugarit (Studi Semitici 6), Roma 1962, p. 153- 154. 38 See here above, note 34. 39 E. LAROCHE, Les noms des Hittites, Paris 1 966, p. 349.

32

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

case of the early Urartian leader Ar(r)amu (9th centuryB.C.)40 , whose name would instead mean "Aramaean" according to some authors4 1 • As for the noun a-ra-mi-ma at Ugarit, it is the Ugaritic genitive plural in -ima of a noun 'aram, haram, or 'aram which qualifies a certain type of fields. It is used without any determinative and, in consequence, can hardly be considered as a gentilic. Since Babylonian arammu, South­ Arabian 'nn, and Arabic 'arim mean "dam", one could propose for eqletm•� a-ra-mi-ma the translation "fields with dams", i.e. "dammed fields". This interpretation is strongly supported by the Arabic derivative 'arim, "irrigated land", developed from the notion of fields surrounded by ridges of earth which retain water for irrigation42 . A derivative of the verbal root 'nn, "to heap", occurs in Ugaritic as a place name43 . Attempts have also been made in order to discover a mention of the Aramaeans in Egyptian texts. A toponym P3-lrm(w) occurs in a topo­ graphic list of Amenhotep III' s (ea. 1 386- 1 349 B.C.) funerary temple44 and in Papyrus Anastasi III from ea. 1 2 1 0 B.C.45 • Since this name does not have the "country" determinative, but rather the "man" determina­ tive, it was interpreted as "(Place) of the Aramaeans" and tentatively located somewhere in southern Syria, perhaps in the Beqa' Valley or in the oasis of Damascus46 • According to Papyrus Anastasi III, this was the district including the "Town of Memeptah". This indication seems to preclude any identification of P3-lnn(w) with a region of Syria. In fact, Egyptian suzerainty was not maintained in inner Syria under Memeptah, although there is clear evidence of the restoration of Egyptian power in 40 APN, p. 28a; PNA U I , p. 132-133. 41 M. VAN LOON, Urartian Art, istanbul 1966, p. 7, n. 24; p. 26, n. 1 3 1 ; M. SALVINI, La formation de l'Etat urarteen, in Hethitica 8 ( 1987), p. 393-4 1 1 (see p. 399-400); ID., Geschichte und Kultur der Urartiier, Dannstadt 1995, p. 26-27. 42 M.A. Gtt0L, Early Southern Arabian Languages and Classical Arabic Sources, Irbid 1993, p. 189- 193. This explanation is now more appealing than our previous attempt to divide a-ra-mi-ma into two distinct words; cf. E. LIPINSKI, Aramiier und Israel, in TRE III, Berlin-New York 1978, p. 590-599 (see p. 591). 4 3 KTU 4.68, 22: place name 'nn; KTU 4.33, 5, and 4.5 1, 13: gentilic 'nny. One should also mention the personal name 'nnn (KTU 4.93, II, 13), A r-mu-na in syllabic cuneiform (J. N0UGAYR0L, PRU Ill, p. 203, lines 10- 1 2 and 20). 44 E. EDEL, Die Ortsnamenlisten aus den Totentempel Amenophis //I., Bonn 1966, p. 28-29. 45 BM. 10246, rev. V, 5, facsimiled in G. M6LLER, Hieratische Lesestucke Ill, Leipzig 1935, p. 27. Transcription into hieroglyphic in A.H. GARDINER, Late-Egyptian Miscella­ nies (Bibliotheca Aegyptiaca 7), Brussels 1 937, p. 32. Passage in ANET, p. 258b-259a. 46 E. EDEL, op. cit. (n. 44), p. 28-29; M. GORG, Aram und Israel, in Vf 26 ( 1 976), p. 499-500; ID., Namenstudien III. Zurn Problem einer Fruhbezeugung von Aram, in BN 9 ( 1979), p. 7- 1 0, reprinted in ID., Beitriige zur Zeitgeschichte der Anfiinge Israels. Doku­ mente-Materialien-Notizen (Agypten und Altes Testament 2), Wiesbaden 1989, p. 157- 160.

EARLY OCCURRENCES OF ARAM

33

southern Palestine47 . This is confirmed by the epithet "subduer of Gezer" given to Merneptah in an inscription at Amada48 . Besides, another topo­ graphic list of the same Amenhotep's temple mentions Damascus49, thus precluding the identification of P3-lrm(w) with this city or its oasis. A simple explanation of the toponym would consist in equating P3lrm with biblical P 'rn, in the central Sinai Peninsula, on the way from Midian to Egypt (I Kings 1 1, 1 8). This large desertic region50 was tra­ versed by the Egyptians who were exploiting the Timna' copper mine, in operation at least from the beginning of the 1 3th century to the time of Ramses V (1 145- 1141 B.C.)5 1 • In this hypothesis, the "Town of Mernep­ tah" mentioned in Papyrus Anastasi III was a manned strong-point on this difficult desert route through the Sinai Peninsula. This route may correspond to the Darb al-I:Iagg of the pilgrims travelling from Egypt to Mecca. It crosses the Tih desert, ancient P(h)aran, by way of Qal 'at an­ Nagl and A!-Tamad to reach Ras an-Naqb and the northern tip of the Gulf of 'Aqaba. The place called Phara on the Peutinger Table is undoubtedly Pharan and it should correspond to A!-Tamad, since the milage indicated there for the trip from Haila (Aila) to Pharan is 50 Roman miles52 • The difference m/n in the spelling of the toponym P3lrm/P 'rn does not constitute any difficulty, since the two phonemes fre­ quently alternate, especially at the end of a word53 . Since Egyptian r regularly corresponds to Semitic l, P3-lrm(w) might be identified also with the important oasis of Elim ( 'ylm), which was known in biblical times54 . It has long been suggested to locate it in WadI Garandal, near the western shore of the Sinai Peninsula55 , on the way to 47 This results from the so-called Israel Stela, studied by W. SPIEGELBERG, Der Siegeshymnus der Merneptah auf der Flinders-Petrie Stela, in ZAS 34 ( I 896), p. 1 -25. For the best reference to this text and its duplicate, see K.A. KITCHEN, Ramesside Inscrip­ tions IV, Oxford 1982, p. 12- 19. There are several translations, also in ANET, p. 376-378, with a reproduction in ANEP, Nos. 342-343. 48 For the inscription, where the numeral should be read "Year 5" (E. WENTE, in Erls 18 [ 1985), p. 62*, n. 2), see K.A. KITCHEN, op. cit. (n. 47), p. 34. Cf. also A.H. GARDINER, Egypt of the Pharaohs, Oxford 1 96 1, p. 273. 49 E. EDEL, op. cit. (n. 44), p. 1 1. 50 The limits of the desertic area are difficult to determine; cf. Y. AHARONI, The Land of the Bible. A Historical Geography, London 1967, p. 180-183. 51 8. ROTHENBERG, Timna: Valley of Biblical Copper Mines, London 1972; ID., The Egyptian Mining Temple at Timna, London 1988. 52 However, the well-watered Feiran oasis is pointed at by Ph. MAYERSON, The Clysma-Phara-Haila Road on the Peutinger Table, in Numismatic and Other Studies in Honor of B.L. Trell, Detroit 198 1, p. 1 67- 176 (see the map on p. 169). 53 LIPINSKI, Semitic § 1 1 .7; 27.30; 29.26; 3 1. 1 2; 36. 1, 5; 4 1 .35; 48.5; 6 1.2. 54

55

Ex. 15, 27; 16, 1 ; Numb. 33, 9- 10. ABEL, Geographie II, p. 2 10 and 3 12.

34

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

�erabi! al-ljadim where mining operations went on under Merneptah' s reign56. The place was known to Diodorus and Strabo57 , and to pilgrims in Late Roman times, as vividly expressed in the appendix to the ltine­ rarium Egeriae, which calls the site Arandara58 , and in the Antonini Pla­ centini ltinerarium, where this pleasant place, ubi applicuimus duos dies delectati, is called Surandala59 • However, no pre-Roman remains were discovered there and an alternative location was thus proposed for Elim at Bi'r Qatya, the largest oasis in northern Sinai, about l O km south of the Al-' Ar'is - Qaniara road60 • Extensive Nabataean ruins have been uncovered on the site, going back to the Hellenistic period61 , but there is no evidence pointing to Merneptah's time. Thus, the equation with P'rn seems at present to provide the best location for P3-lrm(w), that no indi­ cations link so far with the Aramaeans. This equation may also provide an explanation of the name. In fact, a descendant of Seir is called 'A ran in the Hebrew text of Gen. 36, 28 (= I Chron. 1 , 42) and A paµ in the Greek text, also in other versions and in some Hebrew manuscripts. Mount Seir is personified in the genealogy of Gen. 36, 20-30 as the fictitious ancestor of the indigenous tribes and clans living in the Negev and the Sinai Penin­ sula. Their location in this area is decidedly suggested by the juxta­ position of Mount Seir with Mount P(h)aran and Mount Sinai in early poetic texts, like Deut. 33, 2 and Hab. 3 , 3. Egyptian P3- lrm seems thus to refer to one of these tribes. In later times, apav occurs in 56 A.H. GARDINER - T.E. PEET, The Inscriptions of Sinai, 2nd ed. by J. CERNY, London 1952, Nos. 266-267A. 57 OIODORUS SICULUS, Bib/iotheca Historica III, 43 (1ap1voavw;); STRABO, Geogra­ phy XVI, 4, 18 (fop1vliafo1). 5K ltineraria et alia geographica (CCSL 175), Turnhout 1 965, p. 102, § 12. Also A!­ Tiir, in the southwestern part of the Sinai, was identified in the Late Antiquity with Elim; cf. P.-L. GATIER, Les traditions et / 'histoire du Sinai" du IV' au VII' siecle, in T. FAHD (ed.), l 'Arabie preislamique et son environnement historique et culture/, Leiden 1 989, p. 499-523, in particular p. 500-503. 59 ltineraria et alia geographica (n. 58), p. 150, §4 1, 2-7, and p. 172, §41, 8- 13. 60 M. DU BUIT, Elim, in DEB, Turnhout 1987, p. 402. 61 J. CLEDAT, Fouilles a Qasr Gheit (mai /91 1), in ASA£ 12 ( 19 12), p. 145-168 and Pl. I-III; Y. MARG0VSKY, Three Temples in Northern Sinai (in Hebrew), in Qadmoniot 4 ( 197 1), p. 19-20; E.D. OREN - E. NETZER, Settlements of the Roman Period at Qasarweit in Northern Sinai (in Hebrew), in Qadmoniot 10 ( 1978), p. 94-107; E.D. OREN, Survey of Northern Sinai, in Z. MESHEL - I. FINKELSTEIN (eds.), Sinai in Antiquity (in Hebrew), Tel Aviv 1980, p. 129- 146; ID., Excavations at Qasrawet in North- Western Sinai, in /El 32 ( 1 982), p. 203-2 1 I and Pl. 27-29; Y. TSAFRIR, Qa$rawet: Its Ancient Name and Inhabi­ tants, in IEJ 32 ( 1982), p. 2 12-214; E.D. OREN, Qa$rawet, in NEAEHL, Jerusalem 1993, vol. IV, p. 12 13- 12 1 8.

EARLY OCCURRENCES OF ARAM

35

Gebel Musa in the southern part of the Sinai Peninsula.

Ptolemy's Geography V, 16, 1 -3, also as a tribal name apavhm, but it is unlikely that this connotation echoes the Egyptian use of P3-irm with the determinative of "man", viz. the land "of the Aran/m" men. The first incontestable use of the name "Aramaeans" occurs in the inscriptions of Tiglath-pileser I of Assyria ( 11 14-1 076 B.C.), who is described as fighting the "Aramaean Ablamii", Ab-la-mi-i (genitive) kurAr-ma-ia"'c�62 • He engaged them "from the edge of the land Siibu to the city of Carchemish of the land ljatti", thus along the Middle Euphrates, crossed the river pursuing them, and conquered six of their settlements at the foot of Mount Bisri, - seventeen according to another inscription63 . The Aramaean activity around Mount Bisri and along the Middle Euphrates suggests that regions in the West have been raided by nomads as well, as far as Tell Afis which was destroyed in the 12th century B.C.64 . Summarizing his campaigns against the Aramaeans, Tiglath-pileser I says in his later annalistic texts: "I have crossed the Euphrates twenty­ eight times, twice in one year, in pursuit of the Aramaean Ablamii (k"'A b62 RIMA 11, text A.0.87.1, p. 23, lines 46-47; text A.0.87.2, p. 34, line [28]; text A.0.87.3, p. 37, lines 29-30; text A.0.87.4, p. 43, line 34. 63 Six: RIMA II, text A.0.87.1, p. 23, lines 43-63; text A.0.87.2, p. 34, lines 28-29. Seventeen: RIMA II, text A.0.87. 12, p. 59-60, lines 4'-8'. 64 Cf. F. VENTURI, Le premier age du Fer ii Tell Afis et en Syrie septentrionale, in BUNNENS, Syria, p. 53 1-562, in particular p. 558-562.

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

36

la-me-e ku,Ar-ma-a-iarnd). I brought about their defeat from the city Tad­ mar (Palmyra) of the land Amurru, Anat of the land Siibu, as far as the city Rapiqu of Karduniash (Babylonia). I brought their booty (and) pos­ sessions to my city Ashur"65 • A somewhat different picture of the events emerges from a fragment of a Middle Assyrian chronicle which describes a serious famine that occurred in the later part of Tiglath-pileser' s reign, probably in 1082 and 1081 66. The Assyrians, pressured by starving Aramaean tribes, took then refuge in Kirruri, in the Zagros area, northeast of Erbil67 , and the Ara­ maeans probably captured Nineveh before Tiglath-pileser I could respond to the threat by marching to Katmubu, on the eastern edge of the Tur 'Abdin68 , from which the Aramaeans had apparently pushed their way into Assyria proper by advancing down the Tigris river. The text, though very fragmentary and in parts obscure, certainly refers to the "houses" or clans of the Aramaeans, without characterizing them as Ablamu: 2') 3') 4') 5') 6') 7') 8') 9')

"[ . . . ], the people ate flesh of each other, [ . . . ] [ . . . ] the 'houses' of the Aramaeans [ . . . ] the relief, they seized the roads, [ . . . ] they conquered, they took [the . . . ] of Assyria. [ . . . ] to the mountains of Kirruri [ . . . ] life. [ . . . ] took their [gold], silver, and all their property. Marduk-nadin-abbe passed away; Marduk-sapik-zeri, [his son], entered [the house of his fathe]r. Marduk-nadin-abbe had reigned 1 8 years. 10') [ . . . ] the entire crops of the land of Assyria [ . . . ] 1 1 ') [ . . . ] became numerous, they seized (or: afflicted) the 'houses' of the Aramaeans, 1 2' ) [ . . . ] the side of the fortress of Nineveh, the country downstream [ . . . ] 1 3 ' ) [ . . . Tiglath-pil]eser, the king of Assyria, [went] to Katmubu"69.

65 RIMA II, text A.0.87.4, p. 43, lines 34-36; cf. text A.0.87.3, p. 37-38, lines 29-35, with the variant: " . . . their defeat from the foot of Mount Lebanon, the city Tadmar. . . ". "Lebanon" is probably an error for "Bisrf', which is mentioned in earlier inscriptions, ibid., text A.0.87. 1, p. 23, line 59; text A.0.87.2, p. 34, line 29. 66 The events are dated by the mention of the death of Marduk-nadin-alJIJe, king of Babylon (ea. 1099- 1082 B.C.), and that of the accession of his son, Marduk-sapik-zeri (ea. 108 1 - 1 069 B.C.); cf. J.A. BRINKMAN, Marduk-nadin-al:Jl:Je and Marduk-slipik-zeri, in Rl.A VII, Berlin 1987-90, p. 377 and 378. 67 For the geographical location of this area, see L.D. LEVINE, Kirruri, Kirriuri, in Rl.A V, Berlin 1 976-80, p. 606-607. 6H For the location of KatmulJu, see J.N. PosTGATE, Katmul:Ju, in Rl.A V, Berlin 197680, p. 487-488. 69 A.K. GRAYSON, Assyrian and Babylonian Chronicles (TCS 5), Locust Valley, N. Y., 1 975, p. 1 89. For a discussion of this passage, cf. H. TADMOR, The Decline of Empires

EARLY OCCURRENCES OF ARAM

37

The inscriptions of Ashur-bel-kala ( 1 073-1 056 B.C.), the second suc­ cessor of Tiglath-pileser I, simply refer to "the land of the Aramaeans" or to "the Aramaeans", KUR A-ri-me, KUR A-ri-mi, KUR A-ra-me, A-ra­ ma70, although they contain a few mentions of the Ablamii as well7 1 . The later usage of Ablamii in Assyro-Babylonian texts as an appellative for "Aramaic" and "Aramaean" clearly indicates that the Aramaeans were perceived in Mesopotamia as linked closely to the Ablamii72 , but the question of the exact relationship between the two names and the two groups cannot be answered in a definite way as yet. The Ablamii, who were nomadic or seminomadic tribes, partly set­ tled along the fringes of Syrian and Mesopotamian society, are men­ tioned in southern Babylonia as early as the second half of the 1 8th century B.C., at the time of Rim-Anum, when a text refers to "AIJlamaean messengers", DUMU.MES LU.KIN.GI4.A Ab-la-ma-iu73 • They appear later, in the 1 6th century B.C., at Tell ed-Der, the ancient Sippar-Amnanum, where the Ab-la-mu-u were expected to bring bar­ ley74 . Their name clearly reflects the nisba-fonnation of Semitic gen­ tilitial names75 and indicates that the appellation Ablamii was perceived among the sedentary population of Babylonia as a proper name. How­ ever, the etymology relates the word to Westsemitic glm76, known from Ugaritic (glm), Hebrew ( 'elem), Arabic (guliim), South-Arabian (glm), in Western Asia ea. 1200 B.C.E., in F.M. CROSS (ed.), Symposia Celebrating the Seventy­ Fifth Anniversary of the Founding of the American Schools of Oriental Research ( /900-

Cambridge, Mass, 1 979, p. 1 - 14 (see p. 1 2- 1 3). II, text A.0.89.3, p. 94, line 6'; text A.0.89.6, p. 98, line 7'; text A.0.89.7, p. 1 0 1 , col. III, I , 2; p. 102, lines 8, 10, 13, 1 8, 19, 2 1 , 22, 23; p. 103, lines 25, [26]. 30; text A.0.89.9, p. 1 07, line 4'. 7 1 RIMA II, text A.0.89.6, p. 98, ine 14'; text A.0.89.9, p. 107, line [9']. l 7 2 G.M. SCHWARTZ, The Origins of the Aramaeans in Syria and Northern Mesopotamia, in O.M.C. HAEX et al. (eds.), To the Euphrates and Beyond. Archaeologi­ cal Studies in Honour of M.N. van Loon, Rotterdam 1 989, p. 275-29 1 . 71 0. LORETZ, Die ASiRUM-Texte (I), in VF 1 0 ( 1 978), p. 1 2 1 - 160 (see p. 1 29 and 149, No. 20). I wish to thank my colleague K. Van Lerberghe for this reference. 74 K. VAN LERBERGHE - G. VoET, Sippar-Amniinum. The Ur- Utu Archive I (Mesopotamian History and Environment. Texts I ), Ghent I 99 1 , No. 87 (Di 227), lines 16-22. 75 LIPINSKI, Semitic, § 29.4 1 . 76 AHw, p. 2 1a; E. LIPINSKI, "Mon pere etait un Arameen errant'". L 'histoire, car­ refour des sciences bibliques et orientates, in OLP 20 ( 1989), p. 23-47 (see p. 32). The previously proposed etymology was rightly criticized by S. MOSCATI, The "Aramaean Ahlamu '", in JSS 4 ( 1 959), p. 303-307, who failed, however, to give an acceptable expla­ nation of the name. The difficulty in accepting Moscati's argument that Ablamu is a proper name for a tribal group was already pointed out by BRINKMAN, PKB, p. 277-278, n. 1799. 1975), 70

RIMA

38

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

and Aramaic ( 'ulim). This word means "lad", "boy", and its form Ab-la­ a-mu11 marking 'agliim with a long second vowel points to a so-called "broken" or "internal" plural, which in reality is a collective noun78 . The pattern 'af'iil may be used in Classical Arabic to express the plural of collective nouns of persons79, as 'an�iir(un), from na�r(un), that des­ ignated the groups of Mohammed' s Medinan followers who had granted him refuge after the Hegira. Similarly the 'agliimu must have been, initially, "bands of lads" belonging to the tribes that were moving along settled areas and between urban centres. The word is thus a com­ mon noun, but it does not belong as such to the Assyro-Babylonian vocabulary that borrowed it from the language of nomadic or semino­ madic populations. If this interpretation is correct, Ablamu was no proper name of a particular ethnic or linguistic group, but a nomadic designation of the raiding forces that were making forays or razzias for the capture of flocks, slaves, food supplies, etc.80 . In the language of the sedentary population, this word became an appellation of the mem­ bers of nomadic clans making such razzias. The Aramaeans have obvi­ ously belonged to this category of roaming tribes and they became its most conspicuous representatives towards the end of the second mil­ lennium B.C. The relation between the Aramaeans and the Sutaeans raises similar questions8 1 • According to the evidence provided by the Mari corre­ spondence in the 18th-1 7th centuries B.C., the Sutaeans were a con­ federation of nomadic tribes, active over the Syrian steppe to the west of the Middle Euphrates82 . Their presence in the Syro-Palestinian area in the 19th- 18th centuries B.C. is confirmed by the repeated mention 77 This spelling of the personal name Abliimu occurs in ARM XI, 208. References to this name can be found in ARM XVI/ I , p. 54; see also J.-R. KUPPER, op. cit. (n. 36), p. 1 09, n. I . 7H LIPINSKI, Semitic, § 3 1 .23 ff. 79 W. FISCHER, Grammatik des klassischen Arabisch (Porta linguarum Orientalium, N.S. 1 1 ), Wiesbaden 1 972, p. 50, §86, Anm. 3. 8 Compare A. JAUSSEN, Coutumes des Arabes au pays de Moab, Paris 1 908, p. 1 65- 1 66. 81 The fundamental study on the Sutaeans is that of J.-R. KUPPER, op. cit. (n. 36), p. 83- 1 45. See further ID., Suteens et ljapiru, in RA 55 ( 1 96 1 ), p. 1 97-200; M. HELTZER, The Suteans, Naples 1 98 1 , where the question of their identification with the Aramaeans is raised explicitly on p. 86-99. Cf. also a.a.a. REINHOLD, op. cit. (n. 36), p. 29-34. 82 The essential elements of this description can be found in M. ANBAR, Les tribus amurrites de Mari (080 1 08), Freiburg-aottingen 1 99 1 , p. 88-89, 97, 1 1 0, 1 1 5- 1 1 6, 1 1 7, 1 33- 1 34, 205-207. See also J.-M. DURAND, Documents epistolaires du palais de Mari I I (LAPO 1 7), Paris 1 998, p. 505-5 1 1 .

°

EARLY OCCURRENCES OF ARAM

39

of the Swtw in the Egyptian execration texts83 , while only scattered intruders appear in Babylonia at the time of the First Babylonian Dynasty84 . In a later period, in the l5 th- l3 th centuries, the Sutaeans appear in Syria and in Canaan85 , and they even reach the eastern Delta, at least if the S3sw of the Egyptian sources are to be identified with the Sutaeans86 • But this identification is questionable from a phonetic point of view 87 , and it hardly finds a support in the frequent mention of the Sutu in the Amarna corespondence88 and in the single reference of Numb. 24 , 17 to the bane-Set, living in southern Transjordan where the S3sw appear. The available sources give the impression that the proper area of the Sutu bedouin was situated to the west and to the southwest of Mesopotamia, while the main older centres of the Aramaeans were located in Upper Mesopotamia. A distinction appears also in Assyrian royal inscriptions which do not seem to identfy the "Sutaean Ablamii" with the " Aramaean Ablamii". Adad-nirari I (1300 - 1270 B.C.), recording his achievements, mentions his victory over the Ab-la-mi-i Su-ti-i Ja-u­ ri89, where only one Sutaean tribe is meant, since Ablamii is a general designation of these nomadic populations, while Yauru is a Sutaean tribe, as shown by a Middle Assyrian deed said to have come from Tell 'Amiida: is-tu Su-ti-e l-ia-u-ra-ie90," from the Yauraean Sutaean". Much later, however, Sennacherib (704 -68 l B.C.) says: assub nagab 1"Ab-la­ me-e 1"Su-ti-i9 1 , "I uprooted all the Sutaean Ablamii", where Aramaeans are really meant, like in Sargon II's inscription which mentions 1 "A-ra-me H.I K. SETHE, Die Achtung feindlicher Fiirsten, Volker und Dinge auf altiigyptischen Tongefiij3scherben des Millleren Reiches, Berlin 1 926, e 4-6; G. PosENER, Princes et pays d"Asie et de Nubie, Bruxelles 1 940, E 52-53. "4 J.-R. KUPPER, op. cit. (n. 36), p. 86-95; M. HELTZER, op. cit. (n. 81 ), p. 57-75. "5 J.-R. KUPPER, op. cit. (n. 36), p. 96- 1 04; M. HELTZER, op. cit. (n. 8 1 ), p. 79-88. 6 " The Egyptian sources concerning the Shasu have been collected by R. GiVE0N, Les bedouins Sho.mu des documents egyptien.1· (Documenta et monumenta Orientis Antiqui 1 8), Leiden 197 1 . "7 Egyptian . *bi 'ar, "noble", hence "free", still attested in Syriac by bar-bi 're, "noble", and by the feminine M ')rtii 1 29 • The city name must thus derive from the tribal name. The medial 'alaf of Syriac does not appear in earlier texts, like the Story of AI:iiqar (brbrn) 1 30 or the Targum to Job from Qumran (brbryn) 1 3 1 • The correct vocalization bir is provided by the Galilean Aramaic spellings br-byryyh and br-byryn, "freeman", as well as by Christian Palestinian Aramaic br-bryn with a dot under the b, which indicates an i-vowel 132, thus bar-birin. In the 13th century, at the time when the letter found at Diir-Kurigalzu was sent from Assyria, some of the ljirana tribesmen were roaming in northern Mesopotamia (Subartu), while others were rambling along the Euphrates, between Mari and Siigu 1 3 3, before they shifted further to the south to reach finally the area of Sippar where the southern town of ljiran should be located. 125 Tigl. Ill, p. 159, line 5; cf. p. 1 60, line 9. See also S. SCHIFFER, Die Aramiier, Leipzig 1 9 1 1 , p. 2-3, and here below p. 446. It is doubtful, instead, whether the restora­ tion 1"iji-i-ra-[a-nu] in ABL 1468, r. 7 (RGTC 8, p. 162), is correct, since ljiranu is never written with a long first vowel (BRINKMAN, PKB, p. 271, n. 1743), except in Aramaic B 'I /fym (tablet published by T. KWASMAN, BSOAS 63 [2000], p. 274-280, line 9). 126 J.N. STRASSMAIER, lnschriften von Nabonitlus, Leipzig 1889, No. 505, 3; T.G. PINCHES, CT 56, London 1982, No. 239, line 3 ' . Cf. ZADOK, WSB, p. 284-285. 127 RGTC 8, p. 162. 1 2H ADD 2 10 = ARV 204, lines 2 1 and 30. The settlement was close to Diquqina which is identified with modern Ta'uq/fawuq; cf. E. FORRER, Die Provinzeintei/ung des assyrischen Reiches, Leipzig 1920, p. 43; PARPOLA, Toponyms, p. 164, cf. p. 105. 1 29 Compare Assyro-Babylonian yii 'iru and yirtu (AHw, p. 338b, 348a; CAD, lj, p. 3 1 and 200-20 1 ), Sabaic yyr, "noble" (Sabaic Dictionary, p. 64), and Arabic yayr, "good, excellent". IJO TAD III, C l . I , 2 16. 1 3 1 1 I QTgJob 39, 5 (col. XXXII, 4). 1 32 E.Y. KUTSCHER, Studies in Galilean Aramaic, Ramat-Gan 1976, p. 22. The faulty vocalization of the editors of I I QTgJob was corrected by M. SOKOLOFF, The Targum to Job from Qumran Cave XI, Ramat-Gan 1974, p. 152, but it is still followed by K. BEYER, Die aramiiischen Texte vom Toten Meer, Gtittingen 1 984, p. 585. rn 0. R. GURNEY, art. cit. (n. 120), p. 1 39 and 1 48, text No. I 0, lines 2 1-27.

ARAMAEAN TRIBES IN THE 1 3th CENTURY B.C.

47

The ljasam tribe, associated to the ljirana men, certainly originates from Upper Mesopotamia and it is named after tribal lands at ljasam, east of ljarran. The Mari archives mention ku•lja-sa-am 1 34 , which is the same mountainous area as the ku•A-sa-am, which an Old Babylonian itin­ erary situates at a two days' march east of ljarran, in the direction of the western ljabiir 1 35 • Shalmaneser Ill later crossed the ljasamu hills in his march from Assyria to Til-Barsip 1 36 and a town "'"/ja-sa-me is named in the ljarran Census 1 37 • The mountain was sacred and divinized, as shown by the Amorite theophorous name Zi-im-ri-lja-sa-am from Chagar Bazar 1 38 , dating back to the late 1 8th or the early 17th century B.C. The latest indirect attestation of the tribe can be found in the personal name lja-sa-me-e, attested in 682 B.C. in a document from the Mamu temple at Balawat 1 39, ancient Imgur-Enlil. An interesting case is presented in the mid-l 2th century by the Ruqabaeans, an Aramaean tribe which is usually associated with the area at the confluence of the Lesser Zab and of the Tigris, south of Ashur. Their known names are still Westsemitic in the 8th century and can be interpreted partly as Aramaic 1 40, despite the vicinity of their territory to the heartland of Assyria and the expected Assyrian cultural influence. They are attested in this region at the time of Ashur-dan II (934-9 1 2 B.C) 1 4 1 and this seems to have been their land at least as early as the 1 2th century B.C., at the time of Ninurta-tukulti-Ashur (ea. 1 1 33 B.C.) or of one of his immediate successors. In fact, a Middle Assyrian admin­ istrative text found at Nineveh mentions a sheikh Ru-qa-aa-ia 142 among 1 .i. J.-R. KUPPER, in ARM XVI/ I , Paris 1 979, p. 14. 1 .15 W.W. HALLO, The Road to £mar, in JCS 18 ( 1 964), p. 57-88 (see p. 75-76); cf. J. N. PosTGATE, ljasamu, in RLA IV, Berlin 1 972-75, p. 1 28; R. ZAD0K, On Some Upper Mesopotamian Toponyms, in NABU 1 998, p. 69-7 1 , No. 67 (see p. 70, No. 67:2). The mountain called Asiima in Syriac and Mediaeval Arabic sources seems to correspond to the present-day Kara�a Dag, more to the northeast; cf. R. PAYNE SMITH, Themurus Syri­ arn.1· I, Oxford 1 879, col. 406; M. AM0URoux-MoURAD, Le Comte d 'Edesse /098- 1 150 (BAH 1 28), Paris 1 988, p. 39 and 77. 1 .1o RIMA III, text A.0. 102. 1 , p. 1 0, line 83' ; text A.0. 1 02.2, p. 1 5, line 29. 1 .n SAA XI, 20 1 , II, 32. 1 .1s References in C.J. GADD, op. cit. (n. 99), p. 42; Ph. TALON, op. cit. (n. 99), p. 1 38. 1 .1 J.-Y. MoNCHAMBERT, le futur lac du Mayen Khabour: Rapport sur la prospection archeologique menee en / 983, in Syria 6 1 ( 1 984), p. 1 8 1 -2 1 8; D.J.W. MEIJER, A Survey in Northwestern Syria, Leiden 1 986, p. 49; K. KOHLMEYER, in MDOG 1 1 8 ( 1 986), p. 54-55.

50

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

may suggest that a confederation of Aramaic-speaking tribes came into existence sometime in the 1 3th century B.C. and that their collective name entered in the official Assyrian terminology towards the end of the 12th century B.C., when we first meet with the name "Aramaeans" in Mesopotamian sources, at the time of Tiglath-pileser I ( 1114- 1076 B.C.). The scattered evidence indicates at any rate that the Aramaeans were active players in the Near Eastern history for some two hundred years before they emerged as a threat against which Assyria and Babylonia had to defend themselves. This suggests that it was only as the conse­ quence of a widespread ecological and demographic crisis, beginning in the late 13th century, that the Aramaeans appeared as a definable force threatening the political establishment of the Near East. The existence of this crisis is revealed by Egyptian texts from the time of Merneptah ( 12 12- 1202 B.C.) and by a few Ugaritian letters. In his second year, Merneptah sent a huge gift of corn to Anatolia to alleviate a severe famine 1 6 1 and the king of Ugarit was asked in the same period to send 2,000 measures of grain to Ura in Cilicia 1 62 • The invasion of the so­ called "Sea Peoples", the end of the Hittite empire, and the Aramaean thrust into settled areas may all be consequences of the grim tribulation which had then afflicted Anatolia 1 63 • The abrupt end of historical records provided by the cuneiform archives deprives us of more con­ crete informations about the early Aramaean history. This relative obscurity comes to an end only in the late 1 0th century B.C., when Assyria begins to recover and Assyrian annals produce more detailed accounts of the military campaigns, while Hebrew historical literature begins to offer some information. The Aramaeans are then found settled and grouped in a number of federations or small states, centred on a capital city. 161 J.H. BREASTED, Ancient Records of Egypt, Chicago 1 906, vol. III, § 244. Cf. H. KLENGEL, op. cir. (n. 1 1 2), p. 298 and 3 1 0. 161 J. NoUGAYROL, in Ugaritica V, Paris 1 968, p. 105- 108, Nos. 33-34, and p. 323-324, No. 1 7 1 . For the location of Ura at the mouth of Gtiksu/Calycadnos and thus at the head of an important route from the Mediterranean to the plain of Konya, see Tiibinger Atlas des Vorderen Orients, Wiesbaden 1 977-9 1 , Map XVI; G.F. DEL MONTE - J. TISCHLER, Die Orts- und Gewiissernamen der hethitischen Texte (RGTC 6), Wiesbaden 1 978, p. 457-458; G.F. DEL MONTE, Supplement (RGTC 6/2), Wiesbaden 1 992, p. 1 79; J.D. HAWKINS, The Hieroglyphic Inscription of the Sacred Pool Complex at Hallusa (Siidburg) (Studien zu den Bogazktiy-Texten. Beiheft 3), Wiesbaden 1 995, p. 56; H. KLENGEL, op. cir. (n. 1 1 2), p. 240, n. 45 1 . 163 R.D. BARNETT, in CAH lU2, 3rd ed., Cambridge 1 975, p. 360-36 1 , also considers the possibility that HERODOTUS, History I, 94, records a popular souvenir or legend based on this severe famine that would have lasted for eighteen years.

THE ETYMOLOGY OF THE NAME "ARAM"

51

4. THE ETYMOLOGY OF THE NAME "ARAM"

The etymology of the name "Aram" referring to the Aramaeans has not been established until now. The suggestion, sometimes made in the past, that the East-Tigridic toponym Aramu of the third millennium B.C. passed to the Aramaeans who could have settled east of the Tigris during a certain time 1 64 is at least improbable, since we first meet with the so­ called "Aramaeans" in the Gazira and on the fringes of the Syrian Desert. It is also unlikely that a Hurrian onomastic element or proper name has become the tribal or collective name of Aramaic-speaking nomads or seminomads. We can even wonder whether the name was really at first geographical or personal, or even divine 165 , since no indication favours such hypotheses. Moreover, we should question the assumption that the original name of the Aramaeans was Aram with two short a vowels 1 66 • In fact, J.-R. Kupper has rightly observed that the noun "Aramaean" is writ­ ten A-ra-mi in the inscription of Ashur-bel-kala, Tiglath-pileser I's son 167 , and that this spelling contradicts the Assyrian dialectal vowel har­ mony which is generally respected by the scribes until the time of Sargon II and consists in assimilating a short a-vowel to the final vowel of the word, thus A-ru-mu, A-ri-mi 1 68 • The logical deduction would be that this 164 R.T. O'CALLAGHAN, Aram Naharaim (AnOr 26), Rome 1 948; N. SCHNEIDER, art. cit. (n. 1 6). 165 A geographic origin of the name is assumed, for instance, by H. GRIMME, Mohammed. Die weltgeschichtliche Bedeutung Arabiens, Mtinchen 1 904, p. 1 5, who derives "Aram" from rum, "to be high", and thinks of the highland Nagd in northern Saudi Arabia. R.T. O'CALLAGHAN, op. cit. (n. 1 64), p. 95-96, refers to a place name Arma, which does not occur in the text quoted (= RIMA I, text A.0.77.3, p. 1 90- 1 9 1 ). There is also the quite original hypothesis of H. Barth, the distinguished German geogra­ pher, historian and linguist of the mid- l 9th century, and one of the greatest explorers of Africa. H. BARTH, Voyages et decouvertes dans /'Afrique septentrionale et centra/e pen­ da/11 les annees /849 a /855, Paris-Bruxelles 1 860-6 1 , vol. IV, p. 1 68 and n. I , suggests that the name T - aram-t, given by the Tuareg to the whole country on the left bank of the Niger river, from Tombouctu (Timbuktu) to Gogo (Mali), derives from Aram, the sup­ posed homeland of the Berbers in the Middle East. M. STRECK, Ober die iilteste Geschichte der Aramiier, in Klio 6 ( 1 906), p. 1 85-235 (see p. 1 97), thought of a divine name. 166 W.F. ALBRIGHT, in CAH IU2, 3rd ed., Cambridge 1 975, p. 532. Cf. Th. NbLDEKE, Die Namen der aramiiischen Sprache und Nation, in ZDMG 25 ( 1 87 1 ), p. 1 1 3- 1 3 1 ; E. FORRER, in Rf.A I, Berlin-Leipzig 1 928, p. 1 3 1 . 167 RIMA II, text A.0.89.6, p. 98, line 7 ' . This spelling corresponds to the Babylonian and Sii!)u spellings A-ram or A-ra-mu/mi attested in the 8th century B.C. A-ram: Nippur IV, Nos. 4, 23; 1 5, 8; 1 8, 8; 27, 1 6.20; 62, 8; 96, 25; 1 04, 5; 1 05, 6. A-ra-mulmi: Nippur IV, No. 47, 5 ; RIMB II, text B.6. 1 4. 1 , p. 1 2 1 , line 42 ' ; text S.0. 1 002.2, p. 295, lines 20, 2 1 ; p. 298, line 1 4 ' ; text S.0. 1 002.3, p. 30 1 , lines 1 1 ', 1 2 ' . 1 6H J.-R. KUPPER, op. cit. (n. 36), p . I 1 6, n. I .

52

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

vowel was originally long in Aram but it could alternate dialectally with a stressed i (Arim), since the Greek language borrowed the name of the Aramaeans in the 9th or 8th century B.C. under the form 'Apiµot 1 69• Thus, referring to Typhon who personified the Mount Saphon, the pre­ sent Gebel al-' Aqra', Jlias II, 783 celebrates the Storm-god Zeus who is thundering and striking the earth with ligthnings around Typhon' s den Elv 'Apiµotc;, i.e. "among the Aramaeans" or "in the land of the Ara­ maeans". Hesiod follows Homer in his Theogony 304, and also Virgil's Aeneid IX, 715-716 is inspired by the same Homeric passage. Despite the form 'Apiµot, already Posidonius of Apamea (ea. 1 35-ca. 50 B.C.), quoted by Strabo, understood perfectly that Jlias II, 783 was referring to the Aramaeans populating the whole of Syria 1 70 • As for the initial a of 'Aram/'Arim, which is never reduced in ancient transcriptions, not even when the name receives an ethnic or plural affix, it has to be long or semi-long, thus 'Aram/'Arim. Both long vowels were preserved in Syriac 'Aram, with the eventual change a > o to 'Orom in western dialects. Now, the word 'aram is identical with the Arabic "broken" plural 'a ram, "white antelopes", in older languages "wild bulls" or "buffa­ los" 1 7 1 . Since the singular is ri'm in Classical Arabic, rim(um) in Akka­ dian, and rayma or rema in Aramaic, a variant "broken" plural * 'ar'im > * 'arim, eventually reflected in Greek 'Apiµot, is also conceivable in the light of Ge'ez 'ansart, plural of nasr < *nisr, "eagle"; 'abqal, plural of baql, "mule"; •a��al, plural of �a�l. "cup", etc. The existence of sim­ ilar formations of "broken" plurals can be inferred also for early stages of Aramaic, since we ought to admit the use of "broken" plurals in the Tell Fagariya inscription from the mid-9th century B.C. 1 72, unless we are ready to accept the presence of unexplainable mistakes in this text. If the true etymology of 'aram is "wild bulls", such an appellation of a people or a tribe implies a totemic social and religious structure, an essen­ tial peculiarity of which is the association of groups of persons or clans with groups of animals belonging to the same species and constituting the totem species. This conception is usually connected with a belief in some sort of kinship between the clan and the totem. Although worship of the totem never occurs - animal worship has little in common with totemism - it is not unusual for the totemic group to believe that mem­ bers of the totem species - the wild bulls in the case contemplated 169

E. LIPINSKI, art. cit. (n. 76), p. 40-4 1 . cf. I, 2, 34; XII, 8, 1 9; XIII, 4, 6. see W. FISCHER, op.cir. (n. 79), p. 24, § 4 1 b. SA/0 II, p. 54-56, 63, 68, 77-78.

170 S TRABO, Geography XVI, 4, 27; 1 7 1 For the phonetical development, 1 72

THE ETYMOLOGY OF THE NAME "ARAM"

53

Bull represented on an orthostat from Zincirli, 8th century B .C. (istanbul Arkeoloji Miizeleri, Inv. No. 7709).

assist the group in some ways. Our knowledge of the religion and the social organization of the early Aramaeans is very limited, but we know for sure that their main god was the Storm-god Hadad 1 73 and that this god is often represented in the Syro-Hittite art of the l0th-8th centuries B.C. standing on the back of a bulJ l74. Although this iconographic pattern was inspired by the Hittite religious art, it expresses the belief that the wild bull, that became the visible support of the anthropomorphic Storm-god, assists the "Aramaean" totemic group. This connection between the ety­ mology and the religious iconography can hardly be considered as acci­ dental, but the lack of further information does not allow us to answer 1 71 J.C. GREENFIELD, Hadad, in DDD, Leiden 1 995, col. 7 1 6-726 with literature. See also here below, p. 626-636. rn See, for instance, ANEP, Nos. 50 1 (Arslan Ta�) and 53 1 (Til-Barsip), reproduced also by A. PARROT, Assur (L'univers des formes), Paris 1 969, p. 76, Fig. 84, and p. 79, Fig. 89; cf. here below, p. 1 9 1 . One should also mention the stela of Kamanis, king of Carchemish (8th century), found at Cekke, northeast of Aleppo (ANEP, No 500), and the stela in the Bible Lands Museum, Jerusalem, first published in O.W. MUSCARELLA (ed.), Ladders to Heaven, Toronto 1 98 1 , and l.iinder der Bibel, Mainz a/R 1 98 1 , No. 227. For these monuments in general, see W. ORTHMANN, Untersuchungen zur spiithethitischen Kunst (Saarbriicker Beitrage zur Altertumskunde 8), Bonn 1 97 1 , Chapt. Ill.

54

ARAMAEAN PRE-HISTORY AND PROTO-HISTORY

various questions, for instance whether eating the bull was regarded as a sacrament, the bull being considered as of kin to the worshippers of the Storm-god. The proposed etymology of 'Aram has a bearing on the origins of the Aramaeans, for oxen, bisons, buffalos, in a wild state, haunt either prairies and woodlands or swampy marches where they love to wallow. The valleys of the Middle Euphrates and of its tributaries, as well as the forests and the plain at the foot of the Tiir 'Abdin, are therefore the regions where the "broken" plural 'Aram, at first designating herds of wild animals, is likely to have originated as a tribal name.

CHAPTER II ARAM AND THE HEBREW FOREFATHERS A remarkable feature of the ancient Hebrew profession of faith in Deut. 26, 5 is the asserted Aramaean origin of the nation's forefather: 'Aramml 'obed 'iibl, "my father was a roaming Aramaean" 1 • The Greek translators of the Septuagint offer an adaptation of this clause that reveals its misinterpretation in Jewish Hellenistic circles: :Eupiav ant�aA.EV 6 7tatT) p µou, "my father left Syria". This clause was proba­ bly understood as reference to Abraham's migration from ljarran to Canaan (Gen. 1 2, 4-5) and to Egypt (Gen. 1 2, 10). Instead, Targum Onqelos relates this passage to Jacob's flight from Laban's mansion (Gen. 3 1 ): "Laban the Aramaean wanted to ruin my father and he went down to Egypt"2 . The meaning and the grammatical function of 'obed in the clause under consideration suggest a quite different meaning. In fact, since the text is not written in Aramaic, one cannot assume that 'bd is the perfect 'obed of the af'el, as supposed in Targum Onqelos and rightly challenged by Ibn Ezra3 • The problem lays with the word 'obed itself that modem authors translate by "wandering", "perishing", "starv­ ing", even "fugitive"4 • While the latter connotation of 'bd does not occur in Hebrew texts5 , the two other interpretations are based on the actual 1 A survey of various interpretations was made by M.A. BEEK, Das Problem des aramiiischen Stammvaters (Deut. XXVI, 5), in Oudtestamentische Studien VIII, Leiden 1 950, p. 1 93-2 1 2. See now R.C. STEINER, The "Aramean " of Deuteronomy 26: 5 - Peshat and Derash, in Tehillah le-Moshe, Winona Lake 1 997, p. 1 27- 1 38. For a general assess­ ment of the meaning and function of the "profession of faith" in Deut. 26, 5-9, see N. LoHFINK, Studien zum Deuteronomium und zur deuteronomistischen Literatur l, Stuttgart I 990, p. 263-290. 2 A. SPERBER, The Bible in Aramaic I, Leiden 1 959, p. 333; cf. F. DREYFUS, "L'Arameen \'/illlait tuer mon pere ": L 'actualisation de Dr 26, 5 dans la tradition juive et la tradition chretie1111e, in M. CARREZ et al. (eds.), De la T6rah au Messie: Melanges H. Cazel/es, Paris 1 98 I , p. 1 47- 1 6 1 . This understanding of the text would retlect its original sense according lo S. NORIN, £in Aramiier, dem Umkommen nahe - ein Kerntext der Forschung und Tra­ dition, in Scandinavian Journal of the Old Testament 8 ( 1 994), p. 87- 1 04. .1 IRN EZRA, Commell/ary on the Pentateuch, first printed at Naples in 1 488. 4 See the discussion by B. OTZEN, 'iif:Jaq>a6ava mentioned by PToLEMY, Geography V, 17, 5. The latter, located south of the lj:abiir, cannot be identified with the Basileia of the Parthian Stations, probably the pre­ sent-day Zalabiya, as still maintained by M.-L. CHAUMONT, art. cit. (n. 25), p. 83-84; J. LAUFFRAY, Jjalabiya-Zenobia, place forte du Limes oriental et la Haute-Mesopotamie au VI' siecle I (BAH 1 19), Paris 1983, p. 79; T. GNOLI, I papiri dell 'Eufrate. Studio di geografia storica, in Mediterraneo Antico 2 ( 1 999), p. 32 1-358, in particular p. 325, 327, 350. Instead, it must be the Appadana which appears several times as App, Appad, Appadan in documents from Dura Europos and is mentioned in Pap. Dura 608 between Gazica ('Acnxa of the Parthian Stations) and Dura, on the right bank of the Euphrates; see the references in C.8. WELLES - R.0. FINK - J.F. GILLIAM, The Excavations at Dura-Euro­ pos. Final Report V, Part I. The Parchments and Papyri, New Haven 1959, p. 40 and 44 1. Besides, it is mentioned in Greek as Aq>qm6ava or Anaoava in the accounts of Nabuchelus: P.V.C. BAUR - M.I. ROSTOVTZEFF - A.R. BELLINGER (eds.), The Excavations

ARAM AND THE HEBREW FOREFATHERS

71

The Middle ljiibiir seen from Tell Raqa'i.

Apadna of the Notitia Dignitatum58 • It is mentioned also in Caliph 'Abd

al-Malik' s time (A.O. 68 5-70 5) and the poet Mohammed an-Nawag'i (A.O. 1386 -1455) mentions this settlement with A�-�uwwar59, located further north. One can assume that the city name is ancient and that it gave its name to Paddan- 'Aram60, but the forms with a prefixed a- obvi­ ously witness the influence of the Old Persian word apadiina, that des­ ignates a palatial residence, and they may indicate that the place name is not anterior to the 5th century B.C. The obvious conclusion would then

at Dura-Europos. Preliminary Report of Fourth Season of Work, October 1930 - March

1931, New Haven 1 933, p. 96, No. 22 1 ; p. 1 19- 1 20. No. 240. Finally, there are two Appadana in the area of the ljiibiir and of the Middle Euphrates, not three or four. 58 Not. Dign. Or. XXXVI, 8. It should be stressed that a second city Apadna is men­

tioned in Not. Dign. Or. XXXV, 1 3 (Apatna in 25), but it is situated more to the north, possibly near Mardin; cf. here below, p. 146, n. 80. DusSAUD, Topographie, p. 483, recuses both identifications, but proposes no arguments. 59 A. MUSIL, op. cit. (n. 29), p. 82-83, n. 46. For Mohammed an-NawiigI, author of various anthologies, see I. KRATSCHKOWSKY, al-Nawiidji, in Encyclopedie de / 'Islam, nouv. ed., VII, Leiden-Paris 1 993, p. 1 042. 60 It is quite possible that a relation was seen between this geographical name and the Assyro-Babylonian word paddiinu, "road". In Aramaic, instead, paddiin is a "plough", or a "yoke of (plough-)oxen", or the name of a square measure corresponding to about 0,42 ha or an "acre" and related to the surface ploughed in a determinate time. The meanings "yoke of oxen" and square measure ("feddan") are attested also in Arabic for the noun Jaddiin. See also below, p. 73

72

ARAM AND THE HEBREW FOREFATHERS

be that the so-called "Priestly" passages of the Pentateuch mentioning Paddan- 'A ram date from the Achaemenian period, possibly from the late 5th century B.C. As a matter of fact, such a date corresponds roughly to the usual dating of the last distinguishable stage in the for­ mation of the Pentateuch. In this general perspective, Paddan should stand for the city name Appadana, the relative frequency of which would have prompted the addition of the qualification "of Aram". This expla­ nation implies that the "Priestly" scribes in Jerusalem regarded the region of Tell Fiden as representative for Aram. Such a view is under­ standable because of the importance of the IJabur area for the Ara­ maeans, but the historical past of the site should eventually be better known to buttress the link of this particular place with the appellation Paddan- 'Aram. According to H. Ktihne, Tell Fiden corresponds to Suru in the Neo­ Assyrian inscriptions of the early 9th century B.C. 6 1 • This appellation Aramaic sur, "wall" - does not exclude another name, more precise. Besides, the identification is not certain, because not all the ancient sites along the Lower IJabur have been taken into account, especially the site of Al-Bu�ayra at the point where the IJabur joins the left bank of the Euphrates, flowing from the foothills62 • The possibility remains thus that Tell Fiden preserves the name of ancient Paddiin, shortened from Apiidana. Its older Assyrian name may instead have been quite different, for instance "Dur-Adad-nirari", which was founded in the early 8th cen­ tury by Nergal-eres with fifteen villages in the land of Lage according to the stela from Tell ar-Rima�63 • Concrete and firm information can be provided only by excavations on Tell Fiden. Paddan- 'Aram may be linked tentatively not only with Tell Fiden, on the right bank of the IJabur, but also with a borough close to IJarran. 61 H. KOHNE, Zur Rekonstruktion der Feldziige Adad-niriiri II., Tukulti-Ninurta II. und Assurna�irpal II. im ljiibiir-Gebiet, in BaM 1 1 ( 1 980), p. 44-70 and map, in particular p. 6 1-62. This identification is considered as "acceptable" by M. LIVERANI, Studies on the Annals of Ashurnasirpal II, 2: Topographical Analysis, Roma 1992, p. 33. For another identification, see here below, p. 87-89. One could add here that other "'"Pa-da-an occur in the Mura�ii archives (V. DoNBAZ - M.W. STOLPER, Istanbul Murasu Texts, istanbul 1997, No. 59, 3) and in Middle Babylonian texts, perhaps to be located southeast of the Diyala; cf. K. NASHEF, op. cit. (n. 48), p. 2 13. 62 See below, p. 86-89. 63 S. DALLEY, A Stela of Adad-nirari Ill and Nergal-eres from Tell al Rimah, in Iraq 30 ( 1 968), p. 139- 1 53 and Pl. XXXIX-XLI = RIMA Ill, text A.0. 104.7, p. 2 1 1 , line 19. Cf. also the improved decipherment by M. WEIPPERT, Die Feldziige Adadnararis Ill. nach Syrien. Voraussetzungen, Verlauf, Folgen, in ZDPV 108 ( 1992 [ 1993]), p. 42-67 (see p. 60-62).

ARAM AND THE HEBREW FOREFATHERS

73

The ltinerarium Egeriae, composed around 400 A.O., reports Egeria' s pilgrimage to a large borough (uicus) called Fadana which, according to the local tradition, has been the mansion (uilla) of Laban the Ara­ maean64 . Egeria there saw the well and the stone which Jacob rolled off the mouth of the well to water Rachel's sheep (Gen. 29, 10). She also saw the place where Laban' s household gods had been kept before Rachel has stolen them (Gen. 3 1, 1 9). This borough was at a distance of six miles (about 9 km) from ljarran (Charrae) and therefore can safely be identified with Tell Feddan North and Tell Feddan South, seen by E. Sachau in 1 879 and located by him at a distance of two hours walk west of ljarran65 . One might be tempted to assume that the local tradition had originated from the biblical account of Gen. 29-3 1, 2 1, that ends with the mention of Paddan-Aram (Gen. 31, 18). Now, the ljarran Census mentions a town uruBa-da-ni near ljarran66 . Since b/p was one phoneme in Neo-Assyrian, as shown e.g. by the variant spellings Ar-pa-a-a and Ar-ba-a-a, "Arab", while p changed later into f67 , the uicus.. . Fadana of the ltinerarium Egeriae and Tell Feddan may be identical with the uruBa­ da-ni near ljarran. One can just wonder whether this could be the origin of the appellation Paddan-Aram. Today there is only one good well in the vicinity of ljarran, that known as Bir Yaqub, Jacob' s Well, about I km northwest of Altmba�ak. the site of ancient ljarran. At this well occurred the fateful meeting of Jacob and Rachel according to the pre­ sent-day tradition. Finally, there is the possibility that Ba-da-ni, Fadana, Feddan, and Fiden are place names related to the Aramaic and Arabic noun faddiin, "acre", used in the broader sense of "arable land", just like "acre" derives from Latin ager. In this case, Paddan-Aram may just mean "the arable land of Aram", "the ground of Aram", like sadeh 'J\riim in Hos. 12, 13. Contacts between Upper Mesopotamia and Palestine by the caravan tract described above would have a particular significance for the Patri­ archal traditions of the Bible if David's scribe was called Say-Si ', "Gift of the Moon-god", and was native from the ljarran area, as we have argued elsewhere in consideration of the divine name Si ' borne by the Aramaean Moon-god of ljarran68 . Since two sons of Say-Si ' exercised 64 ltinerarium Egeriae XXI, in ltineraria et alia geographica (CCSL 175), Tumhout I 965, p. 65. Fadana appears in XXI, 26. 65 E. SACHAU, Reise in Syrien und Mesopotamien, Leipzig 1883, p. 222. M SAA XI, 2 1 3, III, l l ' - 12 ' . 6 7 LIPINSKI, Semitic, § 1 1 . 4 and § I I . I . 6" SA/0 II, p. 180- 1 8 1.

74

ARAM AND THE HEBREW FOREFATHERS

the same function of king' s scribes under the reign of Solomon (I Kings 4, 3), the influence of the ljarrii.nian traditions may have been quite strong in Jerusalem in the literary field. Another source of influences exercised by Upper Mesopotamia and Aram-Naharaim were the so-called "Hittite" wives of Solomon (I Kings 1 1 , l ), a qualification that could refer in that period to the Aramaean as well as to the Neo-Hittite states of northern Syria and Upper Mesopotamia. It is quite possible that the biblical story of Isaac' s mar­ riage with a spouse brought from the ljarrii.n area (Gen. 24, I O) was partly inspired by a princely marriage that the writer witnessed at the Court of Jerusalem. The fact is that Rebecca' s name Rbqh is Old Ara­ maic and derives from the root rb 6 of the root Jwf, attested in Syriac by lhe verb sii/, nJsu/, "to be burnt", and in Arabic by the noun suwai, "fire". "" RIMA II, text A.O. I O I . I , p. 2 13, line 8. "' H. K0HNE, art. cit. (n. 36), p. 63. "" Ibid., p. 69. The distance covered daily by an army is usually considered as inferior 10 25 km, but the Greek mercenaries of Cyrus the Young travelled about 5.5 parasangs a

88

LAQE

Basalt stela (90 cm high) representing the Storm-god while killing the mythical serpent. It was erected by Tukulti-Ninurta II (890-884 B .C.) at Tell al-'Asara to commemorate his father's, Adad-nirari II (9 1 1 -89 1 B.C.), and his own victorious campaigns in the land of Laqe (Aleppo Museum, Inv. No. 3 1 65 1 1 ll .

TERRITORY

89

Ashumasirpal II's account of another campaign, between 877 and 867 B.C., probably early in that period, also seems to indicate that AI­ Bu�ayra is the site of Sur of Bet-ljalupe. Taking rafts made especially at Sur, where the ljabur flows into the Euphrates, the king went up the river as far as the great defile of ljanuqa, above J:lalabiya: "I made my own boats in Sur, I set out for the bank(s) of the Euphrates, (and) I went up as far as the gorge of the Euphrates", eleppetirnd sa raminiya ina urusuri etapas ana put 1dPuratte G!ibat adi binqi SQ 1dPuratte attarid69 • He further records that he burnt Laqaean villages on the banks of the Euphrates "from the mouth of the ljabur as far as �ibbatu"70• �ibbatu is apparently the same place as Si-ib-na-tim (genitive) in the Mari texts, which lies in the province of Saggariitum, halfway between Terqa/Serqu and Dur­ Yatidun-Lim7 1 , thus north of the mouth of the ljabur and probably on the western bank of the Euphrates. This campaign of Ashumasirpal II started from Sur, that seems therefore to be situated at the mouth of the ljabiir, at AI-Bu�ayra. This place may have been called Sur already in the Old Babylonian period. According to a Mari letter, certain works have been executed "in the Valley of Dur-Yatidun-Lim", i.e. near the confluence of the Euphrates and of the ljabur, "from Pa-Su-ri-im to Sa-ba-ra-ta-a"72 • Considering the particular location of these works near modem AI­ Bu�ayra, Pa-Su-ri-im must be the mouth of the ljabur at Sur, as sug­ gested by its name "Mouth of Sur". Since this region passes from history towards the end of the 8th century B.C., we do not know the circum­ stances in which the site has been destroyed, so that only ruins remained in the 5th century B.C. This site has been rebuilt and resettled in Hellenistic times. Michael the Syrian relates that Seleucus II Callinicus (246-226 B.C.) built a town called Karkis on the ljabur73 • This is a shortened form of the city name which appears in the trilingual inscription of Sapor I from Naqs-i Rustam, day, i.e. about 28 km, sometimes more; cf. V. MANFREDI, op. cit. (n. 55), p. 25-26 and passim. A summary of the available evidence from the Middle East can be found in G.I. DAVIES, The Significance of Deuteronomy I, 2 for the Location of Mount Horeb, in PEQ 1 1 1 ( 1 979), p. 87- 1 0 1 , especially p. 89-97. 69 RIMA II, text A.0. 1 0 1 . 1 , p. 2 14, lines 29-30. 70 RIMA II, text A.0. 1 0 1 . 1 , p. 2 1 4, lines 3 1 -32. 71 ARM XXVI, 1 6, 25-27; 1 63, 7'- 1 1 ' . 72 ARM I I , 1 07, 22-25, where Sa-ger-ra-ta-a should b e corrected i n Sa-ba-ra-ta-a; cf. 1.-M. DURAND, Villesfantomes de Syrie et autres lieu:,;, in MARI 5 ( 1 987), p. 1 99-234 (see

p. 229).

73 MICHAEL THE SYRIAN, Chronicle, ed. by J.-8. CHABOT, Chronique de Michel le Sy rien, patriarche jacobite d 'Antioche ( J /66-1 / 99), Paris 1 899- 1 906, vol. IV, p. 78.

90

LAQE

where the Sassanid king mentions KopKoucrirova, Parthian Krksy', among the places he conquered during his second campaign74, probably in 256 A.D. The name used by Latin writers was Circessum, Cercusium or Circesium75 • Syriac sources generally transcribe it by Qarqusyon or Qarqisyon76, while Jewish Aramaic texts refer to the city as Qirqasyon77 . The meaning of the place name was no more understood, but judging from the Parthian spelling it was possibly *Karki-Sai ', "Fortified city of S ai '", the North-Arabian protective god of the caravans78 • A. Musil was of the opinion that the same site is called at.xya and Na�aya0 in the Parthian Stations79. This opinion cannot be sustained any more, since the first place belonged in the 2nd century A.D. to the hyparchy of 'Iapoa, while the second one was situated in the hyparchy of ra�at..dv80 • Nevertheless, the two places were very close to each other, since no distance is indicated in the Parthian Stations. As argued above, aAtya was probably situated at the outlet of an ancient canal, north of the IJabiir, while Na�aya0 was a K0>µ61to11.1� close to the ford where "the troops were crossing over to the Roman territory". Now, this ford of the Euphrates must correspond to the site of the mediaeval pon­ toon bridge at Dayr Ba�Ir (Al-Bu�ayra)8 1 , rather than to the ford just east 74 E. HONIGMANN - A. MARICQ, Recherches sur les Res Gestae Divi Saporis, Bruxelles 1953, p. 13, line 17; p. 1 47, No. 27, and p. 153; A. MARICQ, Res Gestae Divi Saporis, in Syria 35 ( 1958), p. 295-360 and Pl. XXIII-XXIV (see p. 3 1 1, line 17). 75 Roman attestations are related already to the emperors Gordian III (A.O. 238-242) and Decius (A.O. 249-25 1 ). For Gordian III' s burial place, see EUTROPIUS, Breviarium ab Urbe condita IX, 2 (Miles ei tumulum vigesimo milliario a Circesso . . . aedificavit), and Iuuus CAPITOLINUS, in Historia Augusta XXXIV (apud Circesium castrum). For this sub­ ject, see now M.J. JOHNSON, The Sepulcrum Gordiani at Zaitha and Its Significance, in Latomus 54 ( 1995), p. 141- 144. For the time of Decius, see the Paschal Chronicle, ed. by L. DINDORF, Chronicum Paschale (CSHB), Bonn 1832, p. 504 (ab Arabia et Palaestina usque ad Circesium castrum). 76 R. PAYNE SMITH, Thesaurus Syriacus II, Oxford 190 1, col. 3760. See also E.W. BROOKS (ed.), Vitae virorum apud Monophysitas ce/eberrimorum (CSCO 7-8), Louvain 1907, vol. 7, p. 6 1 :8 = vol. 8, p. 40:4; MICHAEL THE SYRIAN, op. cit. (n. 73), vol. II, p. 475 = vol. IV, p. 448, col. I, 6; BARHEBRAEUS, Chronicon Syriacum, ed. by P. BEDJAN, Paris 1 890, p. 38. 77 JASTROW, p. 1426a. See further here below, p. 9 1, n. 87. 7 8 This is an abridged form, used also in onomastics, of the North-Arabian divine name �ai ' al-Qawm, literally "companion of the kinsfolk". Cf. M. HbFNER, Sai ' al-qaum, Sai ' haq-qaum, in H.W. HAUSSIG (ed.), Gotter und Mythen im Vorderen Orient (Worter­ buch der Mythologie U I ), Stuttgart 1965, p. 465-466. 79 A. MUSIL, op. cit. (n. 32), p. 334. V. MANFREDI, op. cit. (n. 55), p. 12 1 - 122, identi­ fies Na�aya0 with AI-Bu�ayra. Ro C.B. WELLES - R.O. FINK - J.F. GILLIAM, op. cit. (n. 5 1 ), p. 1 1 5, No. 20, 2, and p. 1 28, No. 25, 2 1. • 1 See above, p. 85.

TERRITORY

91

of AI-Mayagin82 . In this hypothesis, Na�aya0, "by which flows the ljabiir river", should be located south of the confluence83, possibly at the ruin mound of Saf'at az-Zirr84, where much of the visible building mate­ rial was already carried away by 19 1 5 85 • This site is about 10 km to the north of the ford of AI-Mayagin. The place name Na�aya0 is Aramaic and derives from the root nbg, "to spring up"; it probably means "spring", unless it is "hillock", like its Arabic counterpart nabaka(tun). As for the name of the new settlement of AI-Bu�ayra, it goes back to Aramaic be- '�ir, "place (house) of pressed (grapes/olives)"86, and has no relation to the earlier names of the settlement87 • About 10 km northeast of Al-Bu�ayra, there is an ancient mound on the right bank of the ljabiir. It is called Tell al-Gubn and, after A. Musi1 88 , it was first signalized by M.E.L. Mallowan89 • Judging from its location and from Ashur-bel-kala's ( 1 073-1056 B.C.) account of his K2 A. MUSIL, Palmyrena. A Topographical Itinerary, New York 1928, p. 235-236, describes this ford: " . . . beside a large island, the Euphrates turns almost east. The chan­ nel there is broad and the current slow, and therefore the river is easy to cross. The con­ venient location of this ford also proves its importance", as it conducted one across the Euphrates to the road leading along the ljabiir into the interior of Mesopotamia. For his 0ljal-la-al-la- ' should be related most likely to Arabic ljalila, "sweetheart" 195 • After that Laqe seems to pass from history except for the mention of AaKEitTJ KCOµT) at Dura Europos, in the first century B.C. 1 96 •

194 ABL 520 = SLA 43, 1 5 . 195 Its masculine equivalent ljalil is well represented in ancient North-Arabian anthro­ ponomy (HARDING, Arabian Names, p. 228) and occurs already as mlja-al-li-li-i, with the suffix -i, in the 5th century B.C.: H.V. HILPRECHT - A.T. CLAY, BE IX, Philadelphia 1 898, No. I 09, I ; cf. ZADOK, WSB, p. 23 1 -232. 1 96 See above, p. 95 with n. 1 1 9.

CHAPTER IV NISIBIS AND THE TEMANITES Although the Lage tribes have created an Aramaean-Arabian federa­ tion on the Lower ljiibiir, just south of Diir-katlimmu, recent research suggests that the Aramaeans did not manage to seize the whole Middle ljabiir valley in the period of their greatest expansion in Upper Mesopotamia, between the death of Tukulti-Ninurta I (1 205 B.C.) and the reign of Ashur-dan II (934-9 1 2 B.C.) in Assyria. Local dynasties, with probable Assyrian roots, are attested then at Tell Taban (Tiibete), Tell Bderi (Diir-Assur-ketfi-leser), and Tell 'Agiiga (Sadikanni), while Tell as-Sayg I:Iamad (Diir-katlimmu) certainly subsisted as an Assyrian provincial centre, though it appears to have been threatened by the Aramaeans in the 1 1 th century B.C., as well as "Magrisu of the [Ri]ver-land", situated near Hassake 1 • On the contrary, the Aramaeans of Bet-Bagyiin and the Temanites managed to settle and to seize power on the Upper ljiibiir, with Goziin, Nisibis, and Gidara as their main centres. Nisibis, modem Nusaybin, was the centre of an Aramaean state which was conquered by Adad-nirari II in 896 B.C. and annexed to Assyria2 • The Aramaeans of this region are called "Temanites" in Assyrian sources, a name that certainly derives from Aramaic tymn, "south", "southem"3, and implies a distinction between the clans settled south of 1 Cf. here above, p. 99 with n. 1 33, as well as H. KOHNE, The Assyrians on the Middle Euphrates and the Jfiibur, in M. LIVERANI (ed.), Neo-Assyrian Geography, Roma 1 995, p. 69-85 and Pl. 1-11, especially p. 72-76; DION, Arameens, p. 36-38; M.G. MASETII­ RouAULT, Syriens et Assyriens dans la Djezire, X/Veme-/Xeme siecle av. J. -C., in M. LEBEAU (ed.), Subartu IV/2, Tumhout 1 998, p. 223-242, especially p. 232-236; H. KOHNE, Tall Seb lfamad - The Assyrian City of Dur-Katlimmu: A Historic-Geographical Approach, in T. MIKASA (ed.), Essays on Ancient Anatolia in the Second Millennium B.C., Wiesbaden 1 998, p. 279-307, in particular p. 282-284. However, the area was largely pop­ ulated by Aramaeans in the Late Neo-Assyrian period, as shown by more than one hun­ dred Aramaic tablets found at Diir-katlimmu in 1 998 (communications of H. KOHNE and W. ROLUG at the 2nd Conference on Aramaic Argillary Texts, Tiibingen, 1 8-20 March 1 999 ) and by the Aramaic name Magda] either of Diir-katlimmu or of a neighbouring town (Marqada?): H. KOHNE - A. LUTHER, Tall Seb /famad/Dur-Katlimmu/Magdalu?, in NABU 1998, p. 1 06- 1 09, No. 1 1 7. See also here above, p. 98-99 and n. 1 32- 1 33. 2 RIMA II, text A.0.99.2, p. 1 50- 1 52, lines 62-8 1 . J E. DHORME, Recueil Edouard Dhorme, Paris 1 95 1 , p. 223, n . 3 . It i s unlikely that these Te-ma-na-a-a, Te-man-na-a-a, Te-man-na-a-ia have something in common with the

I JO

NJSIBIS AND THE TEMANITES

the Tiir 'Abdin and another group, called " northern". The Temanites seem to have been a large people that settled in the 10 th century B.C. and was the most formidable foe of the Assyrians at the time of Adad­ nirari II (9 1 1 -89 1 B.C.), who names three of their principal leaders, viz. Niir-Hadad, Mamli, and Miiquru. The same sources seem to indicate, however, that the Temanites did not manage to create a confederation of clans or a kingdom. 1. TERRITORY

Nisibis lies on the borderland between the Tiir 'Abdin range and the plain at the point where the Cagcag (Gaggag) passes through a narrow canyon and enters the plain. It occupies a strategic position and, during centuries, it formed a frontier fortress and played the role of a trading town on the main route from Upper Mesopotamia to the West4 • It is sur­ prising therefore that the name of the city does not appear before the first millennium B.C. This name, uruNa�lblna in Assyrian5 and N;,�lbln in Syriac, is Aramaic and most likely designated something that is "set up", either " sacred pillars" in the plural or " sacred pillar" with the ending -in < -ayn of place names6 , *na�lbln. This name suggests that the Iron Age city developed around a Semitic shrine with dressed sacred stones, that had to be a meeting point for the Aramaean tribesmen of the region. The site may have been called differently before the Aramaean occupation of the area, perhaps Nawar or Nawali, with a famous shrine of the Storm­ god7, possibly marked by dressed sacred stones and still reflected in the name Niir-Hadad of the Aramaean ruler of Nisibis. But it is also possi­ ble that the site gained its importance only with the Aramaeans and that oasis of Tayma' in northwestem Arabia, as suggested by J. LEWY, The Late Assyro-Baby­ lonian Cult of the Moon and Its Culmination at the Time of Nabonidus, in HUCA 19 ( 1945-46), p. 405-489 (see p. 422). 4 The prosperity of the city in the Middle Ages, down to the 13th century, results clearly from writings of Arab authors; cf. G. LE STRANGE, The Lands of the Eastern Caliphate, Cambridge 1930, p. 94-95. 5 PARPOLA, Toponyms, p. 258-259. Cf. M.P. STRECK, Na$ibina, in RIA IX/3-4, Berlin 1999, p. 185- 186. 6 LIPINSKI, Semitic, § 29.54 and 67. 16. 7 The existence of an original northern Nawar, probably identical with Nawali/u (Hur­ rian r/1), has to be reckoned with in the vicinity of Nisibis in the late third and the second millennia 8.C.; cf. D. MATTHEWS - J. EIDEM, Tell Brak and Nagar, in Iraq 55 ( 1993), p. 201-207, especially p. 204-205. The identification of Nisibis with Sunli was proposed formerly by W.J. VAN LIERE, Capitals and Citadels of Bronze-Iron Age Syria in Their Relationship to Land and Water, in AAS 13 ( 1963), p. 109- 122 (see p. 120).

HISTORY

Ill

Middle Jjabiir.

a nearby town, like Amasaki or Nabula, was the main regional centre in an earlier period8 • 2. HISTORY

The first explicit mention of the Temanites is encountered in the report of Adad-nirari Il's campaign against Niir-Hadad, ruler of Nisibis, in 90 1 B.C. 9 • The Assyrian king claims a victory over the Aramaeans "defeated from Pauza (urupa- '-zi) to Nisibis (uruNa-�i-pi-na)" 1 0• The name of the first city is spelt urupa-u-za in the annals of Ashur-bel-kala (1073 -1056 ) 1 1 and this name is certainly related to the Nahar Baw�a of the Syriac sources 1 2 • This river corresponds either to the upper course of the Cagcag H K. KESSLER, Untersuchungen zur historischen Topographie Nordmesopotamiens nach kei/schriftlichen Quel/en des I. Jahrtausends v. Chr. (BTAVO 8/26), Wiesbaden

1 980, p. 208-209. For Nabula, identified tentatively with Nawar and Nawali/u, see below, p. 15 2 and n. 121. ., RIMA II, text A.0.99.2, p. 149, lines 39-4 1. '" RIMA II, text A.0.99.2, p. 149, line 41. 1 1 RIMA II, text A.0.89.7, p. 102, lines 8-9a. 12 W. WRIGHT, Catalogue of the Syriac Manuscripts in the British Museum, London 187 1 , vol. I, p. 274-275, and index s.v. 'brhmy '.

1 12

NISIBIS AND THE TEMANITES

(Gaggag), where about twelve fortified sites have been recognized on either side of the valley 1 3 , or to one of the channels formed by the Cagcag when it enters into the plain 14. In any case, the town was situated not far from Nisibis, where Niir-Hadad took refuge 1 5. Adad-nirari II's victory was by no means complete and a new campaign was required in 900 B.C. Adad-nirari II claims again to have gotten the victory, but the concrete results appear to be rather meagre. He only managed to seize the small town of Yaridu and to secure a great quantity of grain 1 6. The name of Yaridu ("ru/a-ri-di) means "market" 1 7, but its precise location is uncer­ tain 1 8. Then Adad-nirari annexed the town of Saraku, which became an Assyrian supply centre 1 9. Its exact location is unknown. After the campaigns of 901 and 900 against Niir-Hadad of Nisibis, Adad-nirari II made a deep westward thrust in 899 to IJuzirina, across the Ba1Ib20. The annals recount the capture of the city and the repair of its wall, along with the submission of towns at the foot of the Tiir 'Abdin (Mount Kasiyari), which had been seized previously by Mamli the Temanite. They also record that Adad-nirari has taken over the lat­ ter's "palaces" for himself. A doubt has been cast on the identity of this city of IJuzirina with Sul­ tantepe2 1 , 1 6 km southeast of Urfa, because it is situated some 200 km to 1 3 H. ERKANAL, Mardin, in RLA VII, Berlin 1 987-90, p. 358-359 (see p. 358a). 14 W. WRIGHT, op. cit. (n. 1 2), identifies N:ihar Baw�a with the Zerglin river ('Ap�aµwv), but his only argument is the passage of THEOPHYLACT SIMOCATTES, Historiae II, I , £1ti ,o BiPm; 1tapayivE,Ul fv0a 6 1tomµos 1tapappEi mu 'Ap�aµwv. This argu­ ment not only implies the identification of BiPas with Baw�a. but also assumes that the clause fv0a . . . belongs to the original text and gives the correct localization of BiPas. Now, the correctness of these assumptions is highly problematic, as shown by L. DILLEMANN, Haute Mesopotamie orientate et pays adjacents (BAH 72), Paris 1 962, p. 294. 15 RIMA II, text A.0.99.2, p. 1 49, line 4 1 . 16 RIMA II, text A.0.99.2, p. 149, lines 42-44. 17 JASTROW, p. 596b. 18 J. SEIDMANN, Die lnschriften Adadniriiris II. (MAOG 6/1 ), Leipzig 1 935, p. 65-66, suggests an identification with Redwan, which I don't manage to find on a map. Instead, K. KESSLER, op. cit. (n. 8), p. 2 1 5-2 1 7 , proposes to identify Yaridu with Aridu, later known as a town of the Gozlin province (SAA I, 233, r. I ; SAA XII, I , 1 6). In fact, the apheresis of y- is attested, but Aridu is usually located about 65 km west of Nisibis (see here below, p. 1 26), which seems to be a too big distance in the context of Adad-nirari II's campaigns against Nisibis. 19 RIMA II, text A.0.99.2, p. 1 49, line 44. 20 RIMA II, text A.0.99.2, p. 149- 1 50, lines 45-48. 2 1 The question has been discussed in detail and solved in a positive sense by E.I. GORDON, The Meaning of the Ideogram dKASKALKUR = "Underground Water Course " and Its Significance for Bronze Age Historical Geography, in JCS 2 1 ( 1 967[ 1 9691), p. 7088 (see p. 85-88). J.N. PosTGATE, Ijuzirina, in RLA IV, Berlin 1 972-75, p. 535-536, still seems to consider this identification as not finally proven.

nl.3 1 UK Y

J l .)

. •

KIZlltepe Nusaybin

W. Radd Hassake

I

G. 'Abd al- 'Aziz

• Bderi

Taban

H. David de/.

50

0

1600 m 1300 m 1000 m 650 m 300 m

3. Eastern Gazira.

1 14

NJSIBIS AND THE TEMANITES

the west of Nisibis, the focus of Adad-nirari II's military activity in the years 901 -896 B.C. While the existence of two cities called ljuzirina is a possibility, the immediate connection established by the Assyrian annals between the capture of ljuzirina by Adad-nirari II and the gift of two apes received by him from Bet-'Adini22 points to the area of Urfa and Sultantepe. The ending -lna < -ayna of place names suggests a Northsemitic or Westsemitic origin of the toponym ljuzirina23 , but this ending is missing in Old Babylonian lja-zi- ri, referring most likely to the same place, and it is replaced later by the Aramaic ending -a, echoed in the Latin form Hostra of the Peutinger Table24 . It does not result from Adad-nirari II's annals that ljuzirina was one of the towns occupied by the Temanites25 . Its political status is thus uncertain. An important centre of the Temanites was Gidara, which had been conquered by the Aramaeans under the reign of Tiglath-pileser II (967935 B.C.)26. This city is usually located southwest of Mardin27 , but no evidence seems to support this assumption. Instead, a German military map from the time of the First World War records a place name Bughedra, apparently *Abu-Gidara, about 20 km south of Mardin and some 35 km northwest of Nusaybin. Bughedra may preserve an ancient toponym, since the place is close to several mounds called, on the same map, Tell Dekuk, Muqbil al-FawqanI and al-Ta�tanI, Tell Sehir, Tell al­ Helif, all within a radius of about 5 km from Bughedra28 • The city name Gidara is Westsemitic29, but the Aramaeans were calling it Ra-qa-ma-tu or Ra-dam-ma-te according to the Assyrian annals30. These variants can­ not go back to a name *Ra 6. These developments are common among speak­ ers of Aramaic. 52 See above, p. 122. 53 RIMA Ill, text A.O. I 02.2, p. 22, line 80; text A.0. 102.6, p. 36, line 22; text A.0. 102.8, p. 45, line 13' ; text A.0. 102. 10, p. 52, line 15; text A.0. 102. 14, p. 65, line 56; text A.0. 1 02. 16, p. 75, line 29. 54 Gen. 1 1. 24-28; Josh. 24, 2; I Chron. I , 26-27. 55 See, for example, M. LICHTENSTEIN, Terah, in El XV, Jerusalem 1971, col. 10 131 0 1 4. 56 Gen. 1 1, 3 1-32; at the age of 145 according to the Samaritan Pentateuch.

128

GOZAN OR BET-BAGYAN AND BALilj

situated in the general vicinity of tfarran, although it is not mentioned in the tfarran Census. As for Sa�lala, it should be identified with the very large mound of Tell Sa�lan57 . 2. HISTORY

The first mention of Gozan occurs in the annals of Adad-nirari II of Assyria (9 1 1-89 1 B.C.). After three campaigns that ended with the con­ quest of Nisibis58 , Adad-nirari II advanced in 894 B.C. as far as Gozan and "entered Sikkiin at the source of the tfiibiir", where he received the tribute of Abisalam, "son of Bagyiin": "numerous chariots, teams of horses, silver, gold, the wealth of his palace"59• The qualification miir Ba­ bi-a-ni does not mean that mA-bi-sa-la-mu was the son and the immedi­ ate successor of the founder of the dynasty. It only implies that he belonged to the lineage of the Aramaean rulers of Bet-Bagyiin, the ori­ gins of which must go back at least to the 10th century B.C. The annals of Ashumasirpal II (883-859 B.C.) do not mention the name of the miir Ba-bi-a-ni who paid him a heavy tribute in 879 and around 870 B.C.60• The name of Goziin does not occur in the inscriptions of Shalmaneser III, with the exception of one erroneous mention in the Monolith Inscrip­ tion I, 2861 • The explanation of this fact is provided by a clause in Ashur­ nasirpal II' s annals and by the bilingual inscription on the statue from Tell FalJariya, that must date from the mid-9th century B.C. Although neither a historical event nor a known Assyrian ruler are mentioned in this inscription, both Hadd-yi!'i, the statue's donor, and his father Samas-nun are styled in the cuneiform text sakin uruGuziini, "governor of Goziin", which means that Goziin was an Assyrian province at that time62 • Now, 57 G. Doss1N, Le site de Tuttul-sur-Bali!J, in RA 68 ( 1974), p. 25-34 (see p. 26, n. 4). The alleged Tell Solola, proposed by K. KESSLER, op. cit. (n. 18), p. 20 1-202), does not exist as such. This name probably derives from 'Ayn Sliiq, the source of a small tributary of the Giillab river, according to L. DILLEMANN, op. cit. (n. 18), p. 168, n. 2. Sa-a!J-la-la is already attested in the Old Babylonian period: W.W. HALLO-, The Road to Emar, in JCS 18 ( 1964), p. 57-88 (see p. 64, line 34). Neo-Assyrian references are mentioned in PARPOLA, Toponyms, p. 2 1 1, sub "Kitlala". 58 See above, p. 1 1 1- 1 16. 59 RIMA II, text A.0.99.2, p. 153, lines 1 00- 104. 60 RIMA II, text A.0. 1 0 1 . 1 , p. 203, line 22; p. 2 16, lines 57-58; cf. p. 244, line 73. In SA/0 II, p. 23, the year "889" should be corrected in "879". 61 Cf. SADER, £tats arameens, p. 9 with n. 10. 62 RIMA II, text A.0. 101.2004, p. 391, lines 8-9 and 19-20. For this inscription, see SAID II, p. 1 9-8 1, with previous literature.

HISTORY

1 29

all are agreed that Goziin was an independent state up to 894 , when Adad­ nirari II received the tribute of Abisalam of Goziin, and a local ruler paid tribute to Ashurnasirpal II still in 879 and around 870 B.C. The second mention of this tribute, that consisted of chariots, equipage, horses, silver, gold, tin, bronze, bronze vessels, is followed by a complementary clause: "I took the chariots, the cavalry, the infantry of the mar Ba-bi-a-ni with me (and) I departed from Bet Ba-bi-a-ni"63 • This means that the troops of Goziin were incorporated in the Assyrian army and that the Aramaean state was thus changed practically into an Assyrian province without any reported fight, although its ruler was further considered as a "king". In fact, both Hadd-yi!'i and his father bear the title of "king of Goziin" in the Aramaic version of the Tell Fabarlya inscription64 . This is conceivable only in the period subsequent to a quiet change of the form of govern­ ment, and the editors of the Tell Fa!Jarlya inscription have rightly pro­ posed to identify Samas-nun with the eponym of the eighteenth year of Ashurnasirpal II, i. e. 866 B.C. 65 . The chief-town of the governor's province is not indicated in the Eponym Canon66, but his place in the sequence of eponyms corresponds to the one assigned to the governors of Goziin in the 8 th century, namely in the eighteenth year of each Assyrian king. This reasoning implies that Goziin became de facto an Assyrian province under Ashurnasirpal II, around 870 B.C., but that its governors kept the royal title of the former Bet-Bagyiin dynasty in their relations with their Aramaean subjects. One might surmise that the eponym of the eighteenth year of Shalmaneser III, i.e. 84 1 B.C., was Hadd-yi!'i whose name means " Haddu is my help". In fact, the Akkadian name of the eponym of 84 1 is dIM-ARI:fUs-ni, i.e. Adad- remanni67 , what means "Adad, show pity on me!" This quite common type of Akkadian proper names may have been used as an Assyrian adaptation of Hadd-yi!'i's Aramaic name. Incidentally, no other eponym of that period bears a name with the theophorous element Hadel/Adad. This historical reconstruction explains why Goziin is not even men­ tioned in the annals of Shalmaneser III, whereas the western neighbour of the former Bet-Bagyiin is attacked by the Assyrian king in 853 B.C. 63 64

RIMA II, text A.O. I Q I . I , p. 2 1 6, line 58. SA/0 II, p. 48, lines 6-7 and 1 2- 1 3.

65 A . Aeou-AssAF - P. BORDREUIL - A. R. MILLARD, La statue de Tell Fekherye et son inscription bilingue assyro-arameenne, Paris 1 982, p. 1 03- 1 05. 66 MILLARD, Eponyms, p. 26 and 1 1 9. 67 Ibid., p. 28. The name of the eponym's province ends in [ . . . ]-na, which in fact sug­ gests Guzana; cf. I.L. FINKEL - J.E. READE, Assyrian Eponyms, 873-649 BC, in Or 67

( 1 998 ), p. 248-254 (see p. 249).

1 30

GOZAN OR BET-BAGY AN AND BAL!lj

The Assyrian army crossed the Tigris and marched without meeting any resistance to conquer and pillage the towns of Gi-am-mu in "the land of the BalTIJ"68 . The name of this ruler is Aramaic, namely *G'y 'm, "Exalted is the forefather", and his country constitutes, therefore, another Aramaean state. According to the Monolith Inscription, the Assyrian army did not even need to fight: "They became afraid of the terror emanating from my position as overlord, says Shalmaneser III, as well as of the splendour of my fierce weapons, and killed their lord (EN) Giammu with their own weapons. I entered the towns Saglala and Til­ sa-Turagi and brought my gods into his palaces. I performed a festival in his palaces. I opened his treasury, inspected what he has hidden; I car­ ried away as booty his possessions, bringing (them) to my town Ashur, and I departed from Saglala"69• There is no allusion in this text to a town called BalTIJ, although this city name occurs frequently in Neo-Assyrian sources7° . This may signify that the state of BalTIJ has not been cancelled from the map in 853 B.C. and that its main city BalTIJ, possibly Tell 'Abya97 1 , has not been conquered by the Assyrians. Shalmaneser III's army would have followed a route more to the south to reach the very large mound of Tell Saglan, that should be identified with Saglala72 • The proposed reconstruction of the history of Gozan in the 9th cen­ tury B.C. cannot be considered as valid without assigning a determined period to the reign of Kapara, whose inscriptions appear on several stat­ ues and orthostats of Tell 1:falaf73 • The time of Kapara himself has to be clearly distinguished from the date of the reliefs which have been reused and reinscribed by Kapara74 . Since none of them shows any Neo-Assyr­ ian influence whatever, a date before the Assyrian occupation of the region in the time of Ashumasirpal II is to be preferred for a large num­ ber of the reliefs75 . Instead, a date in the 9th century is clear for Kapara's 6H RIMA III, text A.O. I 02.2, p. 22, lines 78b-8 l a; text A.0. 1 02.6, p. 36, lines 1 9-22; text A.0. 1 02.8, p. 45, lines 1 2'b- 1 4'a; text A.0. 1 02. 1 0, p. 52, lines 1 3b- 1 5 ; text A.0. 1 02. 1 4, p. 65, lines 54b-56; text A.0. 1 02. 1 6, p. 75, lines 28-30. 69 RIMA III, text A.0. 1 02.2, p. 22-23, lines 79-8 1 ; cf. ANET, p. 278b. 70 PARPOLA, Toponyms, p. 65. 71 See above, p. 1 22. 72 See above, p. 1 28 and n. 57. 73 These inscriptions have been edited by B. MEISSNER, Die Keilschrifttexte auf den steinernen Orthostaten und Statuen aus dem Tell Halaf, in Festschrift Max Freih. von Oppenheim (AfO, Beih. I ), Berlin 1 933, p. 7 1 -79. See also A. MooRTGAT, op. cit. (n. 4), Pl. 1 1 0- 1 1 4, 1 30- 1 3 1 , 1 33. A transliteration is given also by SADER, £tats arameens, p. 1 1 - 14. See here, e.g., p. I 3 1 . 74 A . MOORTGAT, op. cit. (n. 4), p. 1 6. 75 I.J. WINTER, North Syrian Ivories and Tell Halaf Reliefs: The Impact of Luxury Goods upon "Major" Arts, in A. LEONARD - B. BEYER WILLIAMS (eds.), Essays in Ancient

HISTORY

13 I

Basalt statue of a queen, 2.73 m high (with pedestal), from the entrance of the bet-biliini, with Kapara's cuneiform inscription: "Palace of Kapara, son of l:lagyiin. What my father and my grandfather did not do in their lifetime, I did. Whoever removes (my) name and sets (his) name, may his 7 sons be burnt in front of Hadad, may his 7 daughters be given up to lshtar as prostitutes. 'Abd-' ilTm, the skilled man, has written". (National Museum of Aleppo, photographed before restoration).

1 32

GOZAN OR BET-BAGY AN AND BALllj

own palace, the bet-ljiliini, since the ivories it contained, especially the female figures wearing a crown decorated with rectangles placed side by side, are exact parallels of Syrian pieces found at Nimrud and dated to the 9th century B.C. 76• If the destruction of the bet-ljiliini may be related to the Assyrian campaign of 808, as R. Naumann proposed77 , a date in the second half of the 9th century for its construction is then very prob­ able. This means that the reign of Kapara is to be placed after the gov­ ernorship of Samas-niir"i and of Hadd-yi!' i. In any case, Kapara is not styled "king of Gozan" in his own inscriptions, but "king of ku'PA-LID-E", a country name that should be read Ba-li8 -e and that refers most likely to the "land of the BalI!J."78 , mentioned in Shalmaneser III's Monolith Inscription79 in connection with the king's campaign to the west in 853 B .C. The sign E indicates the pharyngeal � here80, while PA is used with the value bd, as in Kapara's name, which should be read Ka-bd-ra, i.e. Kabbiirii ', "the Greatest"8 1 • Kapara, son of I:Iagyan, was most likely a later successor of Giammu, who was killed by his own subjects when Shalmaneser III was approaching in 853, and he obviously managed to extend BalI!}'s territory to the east in the latter part of Shalmaneser III's reign, probably around 830 B.C. The situation in the following years is far from being clear, although the turtiinu and palace herald Belu-lu-balat, eponym in 814, is called "governor of Tabete, ljarran, ljuzir"ina, Our, Qipan, Zallu, and Bali!}" on Ashur stela 4482 • Does this long list imply that these cities and the whole Bali!} area were actually in Assyrian hands at that time or that the palace herald was appointed turtiinu in order to restore the Assyrian power in that region? The same question must be raised in relation to Bet-' Adini, Civilization Presented to Helene J. Kantor (SAOC 47), Chicago 1 989, p. 32 1 -337. The earlier discussion of this problem is summarized by SADER, £tats arameens, p. 37-4 1 . 76 R.D. BARNETT, A Catalogue of the Nimrud Ivories, 2nd ed., London 1 975, Pl. LXX and LXXI; ID., Ancient Ivories in the Middle East and Adjacent Countries (Qedem 1 4), Jerusalem 1 982, Pl. 44, and A. MooRTGAT, op. cit. (n. 4), p. 1 3, Fig. 1 1 and 1 2. 77 R. NAUMANN, op. cit. (n. 4), p. 38 1 . 78 E. LIPINSKI, in WO 20-2 1 ( 1 989-90), p. 302. 79 RIMA III, text A.0. 1 02.2, p. 22, line 79 ( idKASKAL.KUR.A); cf. text A.O. I 02.6, p. 36, lines 20 and 23 ( i t reflected in this hypothesis by the spelling Ktk of the Sefire treaties does not create a serious problem, since the dialect of these inscriptions already attests this phonetic shift in the case of the verb yrt < yr[, "to inherit" 1 4 • The 7 THEODORET OF CYRRHUS, Religious History XXIV, 5, 7-8, in op. cit. (n. 4), p. 144. 8 P. CANIVET, op. cit. (n. 5), § 149, p. 203 with n. 187. 9 DussAUD, Topographie, p. 470 with n. 5, and Map XII, C2, where Nigere is an error for Niyere. See also PTOLEMY V, 14, 10; P. CANIVET, op. cit. (n. 5), p. 203, n. 186. 10 THEODORET OF CYRRHUS, Religious History XXV, 2, 13- 19, in op. cit. (n. 4), p. 140. 1 1 P. CANIVET, op. cit. (n. 5), § 149, p. 203. An Islamic wali has often taken the place of the tomb of an anachoret; cf. B. CARRA DE VAUX, Wali, in Encyc/opedie de / 'Islam IV, Leiden 1934, p. 1 168- 1 170. 12 KTU 4.3 19, 1-2 and 6.3. The actual whereabouts of this docket are unknown. For earlier discussions of this docket, see M. LIVERANI, Storia di Ugarit (Studi semitici 6), Roma 1962, p. 154- 155. J.-P. VITA, El Ejercito de Ugarit, Madrid 1995, p. 136, consid­ ers the Kfkym as mercenaries, but other explanations are possible. 13 See, in general, E. VON SCHULER, Die Ka.fkiier. Ein Beitrag zur Ethnographie des a/ten Kleinasien, Berlin 1965. 14 Sefire I, C, 24. One might add the spelling 'tr of the theophorous element 'Aftar in the name of Mati '-' Il's father, but this case is to be explained rather by the assimila­ tion ft > tt, cf. J. BLAU, Marginalia Semitica II, in Israel Oriental Studies 2 ( 1972), p. 57-82 (see p. 72-73). Blau's attempt to explain yrt in a similar fashion seems to us, at present, to be too complicated, although we have considered it as a good alternative

TERRITORY

223

older spelling of the city name would accordingly have been *Ksk, since s is normally used at Sefire to express the interdental !In any case, the identity of the two toponyms Ktk and KintKa is not purely casual, as shown by the important stipulation of Stela III, 23-27 from Sefire 1 5 • It concerns the territory of Tl 'ym which was restored by the treaty to the state of Ktk after having been occupied by the king of Arpad. In fact, TI 'ym has to be identified with the Late Roman borough of TtAAiµa 1 6, that should be located in the montainous region around later K6ppo� 1 7, about 20 km northwest of 'Aziiz and 25 km northwest of Ki.-nKa. It is one of the eight boroughs which Theodoret mentions in his Letter 8 1 , recording that he had converted many Marcionites there in the first years of his office as bishop of Cyrrhus 1 8 • The former identification of Tl 'ym with Tallyiyum 1 9, repeatedly men­ tioned in the Mari letters of the 1 8th-1 7th centuries B.C. 20, cannot be sustained for philological and geographic reasons. In fact, not only the correspondence between a North- and Westsemitic 'aleph and an Old Babylonian ba would be quite unusual2 1 , but even the final m of Talbiiyum is just a mimation, whereas the final m of Tl 'ym belongs to the place name22 • Besides, the location of Talbiiyum in the northwestem region of Al-Gazira, in Upper Mesopotamia23 , makes this town geo-

solution; cf. E. LIPINSKI, North-West Semitic Inscriptions, in OLP 8 ( 1 977), p. 8 1 - 1 1 7 (see p. 98). 15 TSSI II, p. 50. 16 THEODORET OF CYRRHUS, Religious History XXII, I , I , in op. cit. (n. 4), p. 1 24. 17 DUSSAUD, Topographie, p. 47 1 ; P. CANIVET, op. cit. (n. 5), § 1 42, p. 1 94- 1 95 with n. 1 62. - For a description and a history of Cyrrhus, one may consult E. FREZOULS, Cyrrhus et la Cyrrhestique jusqu 'a la fin du Haut-Empire, in ANRW IU8, Berlin 1 977, p. 1 64- 1 92 and Pl. I-VIII. The name Qal 'at NabI Huru of the site results from the loss of the initial q of Qurus; cf. DussAUD, Topographie, p. 47 1 . I H THEODORET OF CYRRHUS, Religious History XII, I , 1 -3, in op. cit. (n. 4), p. 1 24; cf., ID., Lei/er 8 1 , ed. by Y. AZEMA, Theodore/ de Cyr: Correspondance II (SC 98), Paris 1964, No. 8 1 , p. 196 ff. 19 This identification was first proposed by M. NOTH, Der historishe Hintergrund der ln.l'chriften von sefire, in ZDPV 77 ( 1 96 1 ), p. 1 1 8- 1 72 (see p. 1 56- 1 57), reprinted in M. NOTH, Aufsiitze zur biblischen Landes- und Altertumskunde, Neukirchen I 97 1 , vol. II, p. 1 6 1 -2 1 0 (see p. 1 95-196). This hypothesis was then developed by H. CAZELLES, Tai 'ayim, Ta/a et Mu.yur, in Hommages a Andre Dupont-Sommer, Paris 1 97 1 , p. 1 7-26. See also A. LEMAIRE - J .-M. DURAND, Les inscriptions arameennes de Sfire et / 'Assyrie de Shamshi-ilu, Geneve-Paris 1 984, p. 66-72. 2 ° Cf. ARM XVU I , Paris 1 979, p. 34. The same place name, written Ta-al-ba-wi-imki in the genitive, occurs in a text from Chagar Bazar: Ph. TALON, Old Babylonian Texts from Chagar Bazar (Akkadica. Supplementum 1 0), Brussels 1997, No. 45, 3. 21 This is admitted by A. LEMAIRE - J.-M. DURAND, op. cit. (n. 1 9), p. 67-68. 22 M. KREBERNIK, rec. in ZA 74 ( 1 984), p. 1 59. 2·1 J .-M. DURAND, le.I' Anciens de Talhayum , in RA 82 ( 1 988), p. 97- 1 I 3. In the light of a Mari letter, Durand had previously suggested to locate Talbayum in the neighbourhood

224

KITTIK OR BET-�ULLOL

graphically unfit to become a bone of contention between the dynasties of Kittik and of Arpad. The obvious question is whether the state and the city of Kittik are attested in other Iron II sources than the Sefire treaties. If the original pronunciation of their name was *Ki!Jik, as suggested by the Ugaritian docket, the ancient Aramaic spelling of the toponym was *Ksk and its transcription into Neo-Assyrian ought to be *K-s-ku. Now, the geo­ graphic location of this state and its name pair uruKas-ka-a-a or kurKas­ ka-a-a in the inscriptions of Tiglath-pileser III24 and kurKas-ku or kurKa­ as-ku in the texts of Sargon 1125 exceedingly well and it would be unwise to restrain from their identification under the pretence of taking the vocalization into account. In fact, the vocalization is identical with that of ku,Kas-ka(-a)-ia in the annals of Tiglath-pileser 126 , where this ethnic name refers to the Kaska peoples of northeastern Anatolia who had crossed the Kurdish mountains in huge numbers ("20,000 men") and were making their way down the Tigris valley towards Nineveh. These tribesmen, repelled and vanquished by Tiglath-pileser I in the first years of his reign, have nothing in common with the state of Kasku located west of the Middle Euphrates and mentioned almost four centuries later by Neo-Assyrian scribes who have vocalized its Aramaic name * Ksk as if its inhabitants where the Kaskaean moun­ taineers known from old records. This practice was quite common among Neo-Assyrian scribes who were calling ljatti the regions west of the Euphrates, Melubba the land of Egypt under the Kushitic dynasty, Aljlamu the Aramaean tribesmen, etc. The transposition was particularly easy in the case of * Ksk with all the consonants recalling the name of the Kaskaeans. In consequence, the name of the ruler mDa-di-ilu kurKas-ka- a-a, who paid tribute to Tiglath-pileser III in 738 B.C. according to the lat­ ter's annals and to the stela from Iran27 , does not need any more to be of Emar (A. LEMAIRE - J.-M. DURAND, op. cit. [n. 19), p. 69-72), but he now proposes a different reading of the concerned passage: J.-M. DURAND, Documents epistolaires du pa/ais de Mari II (LAPO 1 7), Paris 1998, p. 272. 24 Tigl. Ill, p. 68, line I ; p. 89, line 6; p. 1 08, line 1 5 . 25 Sargon II, p . 33 and 290, Cyl. 1 5; p . 6 3 and 303, Bull 2 1 ; p . 7 6 and 308, R . XIV 1 6; p. 262 and 359, Pavem. IV 36. The restoration of the name on p. 1 28 and 324, Ann. 220, is not justified, but the name appears also in Sargon II's inscription at Tang-i Var: G. FRAME, The Inscription of Sargon II at Tang-i Var, in Or 68 ( 1 999), p. 3 1 -57 and Pl. I-XVIII, see line 22. References in PARPOLA, Toponyms, p. 203. Besides, see here below, p. 23 1 . 26 RIMA II, te)(t A.0.87. 1 , p. 1 7, line 1 00. 27 Tigl. /II, p. 89, line 6; p. I 08, line 1 5.

TERRITORY

225

interpreted as an Anatolian personal name28 . It is an Aramaic name that means "(Ha)dad is god" and it confirms the Aramaean origins of the dynasty ruling in Kittik. Tiglath-pileser III's stela dating probably from 737 B.C. uses the determinative KUR with most ethnic names appearing in the list of the tributary kings29, but the annals seem to make a subtile distinction between KUR and URU, and mention rnDa-di­ i-lu uruKas-ka-a-a30, referring apparently to the capital city, the actual Tell Sayb Ri' a� or Yel Baba. Dad-'il was probably the immediate successor and perhaps the son of Bar-Ga'ya, the king acting in the Sefire treaties. He might have been also the last king of this Aramaean state which was annexed by Sargon II to Assyria, possibly in 7 1 1 B.C., the year in which neighbouring Gur­ gum was annexed3 1 • Judging from the parallelism of Bny Gs or Byt Gs and of Byt Sil in the Sefire treaties (I, B, 3; II, B, I 0), Byt Sil is to be understood as the dynastic name of the kingdom of Kittik32 • This would imply that its real or legendary founder, or the eponym of the ruling dynasty, was a certain Sulul or Sullu/33, who is not known from other sources. One would expect to find an older mention of a king of Kittik in the Zakkiir inscription referring probably to events in 796 B.C. The constel­ lation of names in the inscriptions of Tiglath-pileser III, Sargon II, and Zakkiir suggest restoring [wmlk Ksk/Ktk wm�nth], "and the king of Kit­ tik and his army", on side A, 7-8, of the Zakkiir stela, but this is just a conjecture34 • One is also inclined to suppose that a king of Kittik was intended among the adversaries mentioned globally in the inscriptions of Shalmaneser III, but the state of Kittik may have not yet been in exis­ tence in the mid-9th century B.C. In fact, Shalmaneser III specifies that he conquered, in 858 B.C., "the towns Taia, ljazazu, Nulia, (and) Butamu, which (belong) to the country Pattin"35 • Now, these cities lay on the later territory of Kittik or in its immediate surroundings. 2K This is regularly done, sometimes with a note of caution, following B. LANDSBERGER,

Sam'a/, Ankara 1948, p. 16- 17, n. 34, and E. VON SCHULER, op. cit. (n. 13), p. 68, 93 and n. 34, also by referring to E. LAROCHE, Les noms des Hittites, Paris 1966, p. 181, No. 1307. 29 Tigl. l/1, p. I 06 and I 08. The only exceptions are Tyre and Carchemish .

.1o Tig/. Ill, p. 89, line 6; also p. 68, line I . · 1 See below, p. 231 with n. 63 . .1 2 A. LEMAIRE - J .-M. DURAND, op. cit. (n. 19), p. 56-66. However, one cannot follow lhese authors when they identify Byt $11 with the house of the turtiinu Shamshi-ilu. .1.1 The name m$u-lu-lu is attested in the Neo-Assyrian period: APN, p. 206b; SAA I, 87, 6. Cf. E. LIPINSKI, rec. in OLZ 8 1 ( 1986), col. 353 . .14 See below, p. 303. ·15 RIMA 111, text A.O. I 02.2, p. 17, line I I . 1

226

KITIIK OR BET-�ULLOL

Lines 9- 12 of Stela I, B from Sefire use a figurative style to describe the large tenitory where the stipulations of the treaty ought to be pro­ claimed. The regions, mountains, and cities are mentioned not at ran­ dom, but following a pattern based on the four cardinal points and expressed by the repetition of the formula mn ("from") + place name + w 'd ("and as far as") + place name + w ("and") + place name: South ➔ North: [mn] Qrqr w 'd Y'd[y w]Bz. West ➔ East: mn Lbnn w 'd Yb[rd w . . . ]. North ➔ South: [mn 'm]q w 'd 'rw wM[,l"]W. East ➔ West: mn Bq 't w 'd Ktk [w 'rpd].

We are following A. Lemaire' s restoration for the first place name36 , since the slight trace of the initial letter suits a qof as well as an 'ayn, while the fourth letter, usually read w, does not have the shape of the clearly preserved waw signs of the inscription, and corresponds rather to r. Y'd[y] indicates the Aramaean kingdom of S am'al, in the north, and Bz may refer to Boz Tepe or Boz Dag, a peak of the Amanus range (alt. 2,240 m), 18 km east of the shoreline of the Gulf of iskenderun and about 45 km southwest of Zincirli, the site of the capital of S am'al. The place should thus be distinguished from Bazu, possibly the Azraq oasis. We accept the usual restoration Yb[rd], since this city lies to the east of the Lebanon ranges, and we read further [ 'm]q37 , the state of 'Umq in the north, while 'Arra ( 'rw) and Man�uate (M[�]w) would indicate the south38 . The Bq 't cannot be simply identified with the Lebanese Beqa', since this word means "valley" in general. Besides, the mention of Lebanon in the second phrase of the section seems to indicate that Bq 't refers to another region39 . The Euphrates valley seems to be the most appropriate place indicating the east, while Ktk and presumably [ 'rpd] would stand for the west. This schematic listing of the territories concerned by the Sefire treaties seems to c;onfirm the northwestern situation of the kingdom of Ktk and to support the identification of the city of Ktk with KitnKa. The location of the kingdom of Kittik in northwestern Syria, where it bordered on territories inhabited by Anatolian populations, is further confirmed by the invocation of the divine couple of Nkr and Kd 'h as 16 A. LEMAIRE - J. -M . DURAND,

op. cit. (n. 1 9), p. 8 1 -82. 17 The restoration "Yab[rud]" is suggested mainly by later attestations of the place name: L. J ALABERT - R. MouTERDE, /GLS V, Paris 1 959, p. 308 with bibliography. The restoration "['Um]q"' is proposed by A. LEMAIRE - J .-M . DURAND, op. cit. (n. 1 9), p. 83. lR See below, p. 294 and 304-2 10. 9 J.A. FITZMYER. op. cit. (n. 2), p. 1 06. -'

TERRITORY

227

witnesses to the treaty concluded by Bar-Ga'ya with Mati'-'11, king of Arpad40. In fact, Nkr is a phonetic variant of the name of the Lycian and prob­ ably South-Anatolian Apollo, called Natr(i)- in the personal name Natrbbijemi that translates 'A1to11.11.6cSoto k not only in the cluster kb < tb42 , but also in krltr, as shown by the name Krupsse/i- transcribed 0p(n11t l but also by l > dlt, as shown e.g. by hieroglyphic Luwian (a)daman, "name", corresponding to Hittite Laman (cf. Latin nomen). We can assume therefore that Kd'h stands for Kuda '+h of the Aramaic feminine, since the very element Kula is used in Anatolia as a feminine personal name (e.g. rouAac;, Kouoa, K ouA.a53) and seems thus to designate a female divine principle or entity. The probable presence of *Nakri and of *Kuda'a among the deities of Kittik seems to imply a dependence from Carchemish. Beside the men46 V. HAAS, Geschichte der hethitischen Religion, Leiden 1 994, p. 243-244. For a syn­ thesis of the opinions concerning this problem see M. GIORGIERI - C. MORA, Aspetti de/la regalita ittita nel XIII secolo a. C. (Biblioteca di Athenaeum 32), Como 1 996, p. 77-79. 47 M. POPKO, Religions of Asia Minor, Warsaw 1 995, p. 1 52- 1 54. 4 H L. ZGUSTA, op. cit. (n. 43), p. 207-208, § 500. 49 FGH IIA, p. 346-347, § 90, 44. 50 Ph.H.J. Houw1NK TEN CATE, The luwian Population Groups of licia and Cilicia Aspera during the Hellenistic Period, Leiden 1 96 1 , p. 1 24, with former literature; E. LAROCHE, op. cit. (n. 28), p. 28 1 -282. 51 Ph.H.J. HOUWINK TEN CATE, op. cit. (n. 50), p. 1 50- 1 5 1 ; L. ZGUSTA, op. cit. (n. 43), p. 1 37, § 233; p. 252, § 7 1 8; p. 253, § 726; E. LAROCHE, op. cit. (n. 28), p. 96-97, Nos. 608-6 1 3 . 5 2 H. KRONASSER, Vergleichende Laut- und Formenlehre des Hethitischen, Heidelberg 1 956, § 72a; E. LAROCHE, Comparaison du louvite et du lycien II, in Bulletin de la Societe de linguistique 55 ( 1 960), p. 1 55- 1 85, especially p. 1 8 1 ; L. ZGUSTA, op. cit. (n. 43), p. 143, § 252 and n. 57b; H. KRONASSER, Etymologie der hethitischen Sprache l, Wies­ baden 1 966, p. 6 1 -64, § 50; G. NEUMANN, lykisch, in Altkleinasiatische Sprachen, Leiden 1 969, p. 358-396, especially p. 376-377; A. HEUBECK, lydisch, ibid., p. 397-427, espe­ cially p. 405; M. PoPKO, Ludy i jezyki staroiytnej Anatolii, Warszawa 1 999, p. 88, I I O, 1 1 3, 1 24. As examples one can mention Lametru = Demeter, Lygdamis = Tugdamme, Labarna = Tabarna, KavoauA.T)c; = xnrawata, "king". See also F. STARKE, u,barna, in RLA VI, Berlin-New York 1 980-83, p. 404-408. 5 .1 L. ZGUSTA, op. cit. (n. 43), p. 1 37, § 233- 1 ; p. 252, § 7 1 8-2; p. 253, § 726-3.

TERRITORY

229

Neo-Hittite orthoslat from Carchemish with a musical scene, second half of the 8th Century B.C. (Ankara Museum, Inv. No. 1 1 9).

tion of Anatolian divine figures, this dependence is suggested by the geographic position of Kittik. It is uncertain where the northern frontier of Kittik should be placed. It is quite probable that the area of Kilis, ancient Cilliza54, belonged to Kittik, but it is unlikely that its territory was reaching as far as Gaziantep, Crusaders' Hantab and ancient 'Ayn Tab, near the sources of the Sagiir (Sacir Suyu)55 , that may have belonged to Bet-'Adini or to Gurgum. On the west, the montainous region around later Kupp°';, with Tillima as the main centre, was recognized as Kittik' s territory by the ;, The city is probably mentioned in ND. 1 0005, 1 3 (CTN III, 86), where one man is supposed to serve Sam'al and ""'Ka/-zi, that cannot be here Kilizu in central Assyria. ;; The ancient remains have been removed when the mediaeval castle was built; cf. A. ARCH! - P.E. PECORELLA - M. SALVINI, Gaziantep e la sua regione (Incunabula Graeca 48), Roma 1 97 1 , p. 67, No. 70, and Pl. XXIV, Fig. 88. The actual name of Gaziantep prefixes the Turkish word gazi, "hero", to the ancient name 'Ayn Tab that seems lo be Aramaic and to mean "good source".

230

KITIIK OR BET-�ULLOL

Sefire treaties. The state was bounded on the southwest by 'Umq and later by Arpad, as on the south. It certainly had a common border with Carchemish on the east, as indicated by the hieroglyphic Luwian inscrip­ tion from Cekke56, only l l km east of Yel Baba. The eastern boundary was, on the evidence of Cekke, the upper Quwaiq valley. The territory of Kittik was thus small and the importance of Kittik in the Sefire treaties must be ascribed exclusively to the Assyrian policy aiming at weakening Arpad. In an earlier period and after the destruction of Bet­ ' Adini, the influence of the city-state of Carchemish was certainly preva­ lent in the area, but the region was bound on the west and the southwest by another Neo-Hittite state, namely 'Umq/Pattin. These western rela­ tions may explain why the name of the king of Kittik in the mid-8th cen­ tury B.C. occurs in Cilicia as a patronymic Bapyma� as late as the Hel­ lenistic period, namely in Povoav Bapymou57 • 2. HISTORY

Although Bar-Ga' ya, king of Kittik, does not mention his ancestors in the Sefire treaties, the dynastic name Bet-�ullul (Byt Sil) of the country occurs in two passages of the stelae58 . The personal name Su-lu-lu is attested in the Neo-Assyrian period59 and, despite the similarity of the names, nothing indicates so far that it is historically related to the land of Izalla60, which never appears with the determinative of personal names. Nothing is known about the origins of this dynastic name and any spec­ ulation would be dangerous. The Sefire treaties indicate that Bar-Ga'ya and his possible predecessors have ruled in the concerned area for some time, that they obviously enjoyed the Assyrian protection in the first part 56 This inscription is the foundation charter of a city Kamana by Kamanis, king of Carchemish in the mid-8th century B.C., who had acquired the land by purchase from the men of the town Kanapu. Cf. J.D. HAWKINS - A. MORPURGO-DAVIES, Buying and Selling in Hieroglyphic Luwian, in Serta lndogermanica. Festschrift fur G. Neumann, Innsbruck 1 982, p. 9 1 - 1 05, especially p. 93 and 96. 57 R. HEBERDEY - A. WILHELM, Reisen in Kilikien, Wien 1 896, No. 1 3 1 ; L. ZGUSTA, op. cit. (n. 43), p. 1 20, § 1 50-3, where the reconstruction BapyatO(; should be corrected into Bapymm;. 58 Stela I, B, 3 and Stela II, B, I 0. Cf. A. LEMAIRE - J.-M. DURAND, op. cit. (n. 1 9), p. 59-60, although there is no relation between this dynastic name and Shamshi-ilu. 59 SAA I, 86, 6; APN, p. 206b. 60 E. LIPINSKI, rec. in OLZ 8 1 ( 1 986), col. 353, contrary to A. LEMAIRE - J.-M. DURAND, op. cit. (n. I 9), p. 60-66. For lzalla, see J.N. PosTGATE, /zalla, in RLA V, Berlin 1 976-80, p. 225-226.

HISTORY

23 1

of the 8 th century B.C., and that Dad-'il, a Bar-Ga'ya's successor, paid tribute to Tiglath-pileser III in 738 B.C.6 1 • The region of Tillima (Tl 'ym), lost to the kings of Kittik in the first part of the 8 th century B.C., was restored to Bar-Ga'ya probably around 754 B.C., although different datings have been proposed for the single Sefire stelae62 . Stela III fixing the boundary between Kittik and Arpad and retroceding Tillima to Bar-Ga'ya's dynasty may in fact be anterior or posterior to Stelae I and II, but the difference may just represent one year or two. The scribal and stylistic similarities between the stelae do not favour the hypothesis of longer periods separating their engraving on the same type of stone and apparently by the same stone-cutter. The events of the last third of the 8 th century B.C. are unknown in the present state of our knowledge and we can only assume that the small kingdom of Kittik, probably called Kasku, was annexed to Assyria by Sargon II, possibly in the same year 7 1 1 of the annexation of Gurgum63 . In fact, the inscriptions of Sargon II mention Kasku before Tabalu64 , before Tabalu and ljilakku65 , or between these two countries66, thus implying a western location of the territory. Later traces of that Ara­ maean state are limited to a personal name like mKas-ka-a-a61 •

61

Tigl. Ill, p. 68, line 1 2; p. 89, Ann. 27, line 6; p. 1 08, line 1 5 . 1'2 A. LEMAIRE - J.-M. DURAND, op. cir. (n. 1 9), p . 57-58. 63 Sargon II, p. 64 and 304, Bull 26; p. 76 and 308, R. XIV I 0- 1 1 ; p. I 3 1 - I 32 and 325-326, Ann. 238-24 1 ; p. 2 1 8 and 348, Displ. 86-89; p. 26 I and 359, Pavem. IV 28-3 1 . 64 Sargon II, p. 33 and 290, Cyl. 1 5 . 65 Sargon II , p . 6 3 and 303, Bull 2 1 -22; p . 7 6 and 308, R. XIV 1 6; p. 262 and 359, Pavem. IV 36. The restoration of the name of Kasku on p. 1 28 and 324, Ann. 220, is not justified. The same must be said about the parallel passage in Sargon Il's annals from 7 1 1 B.C., despite A. FUCHS, Die Annalen des Jahres 711 v. Chr. nach Prismenfragme/1/en aus Ninive UII([ Assur (State Archives of Assyria Studies 8), Helsinki 1 998, p. 49, line 14'. 66 G. FRAME, art. cir. (n. 25), line 22. 67 CTN III, 34, 28.

CHAPTER X SAM'AL OR *YU'ADDI At the foot of the eastern flank of the Amanus range (Nur Daglan) lies the small Aramaean state of Sam'al or *Yu'addI with its capital on the present-day site of Zincirli Hoyilk (Turkey). The German excava­ tions of this mound in 1888 -1902 have uncovered the Aramaean city 1 , surrounded by a circular defence wall, with an acropolis comprehending palaces which yielded a number of sculptures bearing Phoenician or Aramaic inscriptions. Some sculptures have been discovered also at Ger�in2, 7 km northeast of Zincirli, - one being a statue of Hadad with an Aramaic inscription of Panamuwa 1 3. This was probably the site of the royal necropolis of Sam'al4 • Besides, an inscribed stela was found in 1888 at Ordekbumu5 , some 12 km south of Zincirli. The badly damaged inscription is written in Aramaic monumental script and its language is most likely Semitic6 • However, three hieroglyphic Luwian signs appear 1 F. VON LuscttAN et al., Ausgrabungen in Sendschirli l - V (Mitteilungen aus den ori­ entalischen Sammlungen der Koniglichen Museen zu Berlin 1 1 - 1 5), Berlin 1 893- 1 943. 2 F. VON LUSCHAN, Fiinf Bildwerke aus Gerdschin, in Ausgrabungen in Sendschirli l, Berlin 1 893, p. 44-54. ·' The original publication of the inscription was prepared by E. SACHAU, Die Inschrift de.1· Kiinigs Panammii von Sam 'al, in Ausgrabungen in Sendschirli l, Berlin 1 893, p. 55-84. More recent treatments of the inscription can be found in P.-E. DION, La langue de Ya 'udi, Waterloo 1 974, and J. TROPPER, Die Inschriften von Zincirli (ALASP 6), Miinster 1993, with an improved interpretation of several passages in SA/0 II, p. 203-2 1 1 . 4 B. LANDSBERGER, Sam 'al, Ankara 1 948, p. 66, n. 1 67; H. NIEHR, Zurn Totenkult der Kiinige von Sam 'al im 9. und 8. Jh. v. Chr., in SEL 1 1 ( 1 994), p. 57-73 (see p. 70). 5 F. VON LUSCHAN, in Ausgrabungen in Sendschirli IV, Berlin 1 9 1 1 , p. 328-330. Cf. ESE Ill, p. 1 92-206 and Pl. XIII-XV. The stela dates from the 9th or 8th century B.C.; cf. W. ORTHMANN, Unter.mchungen zur spiithethitischen Kunst (Saarbriicker Beitriige zur Ahertumskunde 8), Bonn I 97 1 , p. 529; H. GENGE, Nordsyrisch-siidanatolische Reliefs l, Kopenhagen 1 979, p. I O 1 - 1 02. " Although M. LIDZBARSKI (ESE Ill, p. 1 92-206), J. FRIEDRICH (Kleinasiatische Sprach­ denkmiiler, Berlin 1 932, p. 38), and P. MERIGGI (Manua/e di eteo geroglifico II. Testi: 2a e 3a serie, Roma 1 975, p. 2 1 9-22 1 , No. 287) do not consider the inscription as Semitic, but rather Anatolian, B. LANDSBERGER (op. cit. [n. 4], p. 64-65, n. 1 66) and H. SADER (£tats arameen.1·, p. 1 72) rightly recognize its probably Aramaic features. Several words and phrases are clearly Semitic and some of them are paralleled in Panamuwa I's inscription on the Hadad statue: 1111 vbnv. . . 111b hlbbh, "the land he should (re)build . . . the succession should yield" (line 2; �f. SA/0 II, 204-205); ys 'd, "he should sustain" (line 3); nbk y11. . . lrkh '/ '/h, "a spring of wine . . . for the god Rakkab-'11" (lines 4-5); brkb '/ J ' yn lym zb, "with Rakkab-'11 a sheep; the wine !lowed daily" (lines 7-8); '.fm kyn mt h ' nb .fm.r '/h, "guilt-offering ('!), when the Sun-god has yielded the wine of this land" (line 1 0).

p.

234

SAM'AL OR *YU'ADDI

above the scene represented on the stela7 and the inscription below the bas-relief contains a few religious terms of Anatolian origin, like zrmly (line 3) and 'rmly (line 6), that are adjectives or names of agent derived from the divine names "Sarruma" 8 and "Arma", the Luwian Moon-god9 • Another stela, found also in 1888 at Karabun;lu, 5 km north of Zincirli, bears a hieroglyphic Luwian inscription 1 0 • These inscriptions witness the mixed Aramaean-Hittite culture of the kingdom, where the ruling dynasty, as known from the Phoenician inscription of Kilamuwa 1 1 , a late contemporary of Shalmaneser III, was largely Aramaean, although Kilamuwa and some of his successors, as Qarli, Panamuwa I, and Panamuwa II, bore Anatolian names. The local use of the Neo-Hittite sculptural tradition, especially in the rendering of the Storm-god and of the ruler 1 2 , underscores the close relationship between Neo-Hittite and Aramaean art, which may well parallel wider social and political links. The Aramaean city was founded towards the end of the 10th century B.C. on the site of an older townlet, which may have been called Usa in the Hittite period 1 3 , but was most likely abandoned by ESE III, Pl. XIII. 8 The spelling zrmly with z is related to one of the Ugaritic spellings of the divine name, namely (jrm. For this god, cf. E. LAROCHE, Le dieu anatolien Sarruma, in Syria 40 ( 1 963), p. 277-302; M. PoPKO, Religions ofAsia Minor, Warsaw 1 995, p. 97-98, 1 1 4- 1 15, 1 65- 1 66. For the later period, see also P.H.J. HoUWINK TEN CATE, The Luwian Population Groups of Lycia and Cilicia Aspera during the Hellenistic Period, Leiden 1 96 1 , p. 1 34- 146. 9 The name 'rmly is used in apposition to kbb in lkbb 'rmly, apparently "for the Lunar Kubaba". The goddess Kubaba is associated to the Moon-god of ljarriin in the hiero­ glyphic Luwian inscriptions from Sultanhan, Kayseri, and Karaburun; cf. J.D. HAWKINS, Kubaba. A. Philologisch, in RLA VI, Berlin 1 980-83, p. 257-26 1 , in particular p. 260a; ID., Kubaba at Karkamis and Elsewhere, in AnSt 3 1 ( 1 98 1 ) p. 1 47- 1 76. For Arma, see E. LAROCHE, Divinites lunaires d'Anatolie, in RHR 1 48 ( 1 955), p. 1 -24; ID., La lune chez les Hittites et Les Hourrites, in La lune. Mythes et rites (Sources orientales 5), Paris 1 962, p. 1 1 7- 1 25; P.H.J. HouwINK TEN CATE, op. cit. (n. 8), p. 1 3 1 - 1 34, with former literature. 10 P. MERIGGI, op. cit. (n. 6), p. 94 and Pl. XVI, No. 1 46. Cf. W. ORTHMANN, Karabur­ clu, in RLA V, Berlin 1 976-80, p. 40 1 b, with former literature. 1 1 KAI 24 = TSS/ III, 1 3. The vocalization "Kulamuwa" adopted by DION, Arameens, p. 99 ff., in consideration of F. STARKE, Untersuchungen zur Stammbildung des keilschrift-luwischen Nomens (Studien zu den Bogazkoy-Texten 3 1 ), Wiesbaden 1 990, p. 236 and n. 806 (cf. also ID., in Bi Or 46 [ 1 989], col. 668), is questionable, since the name "with an army's strength" should be spelt Klnmw, like in KTU 4.44, 2 1 .25, while the ele­ ment kila is attested as well in Anatolian onomastics. Besides, the name Kd.11.tµroi:ac; (L. ZGUSTA, Kleinasiatische Personennamen, Prag 1 964, p. 222, § 574) may just be a variant of Klmw, just like Panamuwata/i is a variant of Panamuwa; cf. P.H.J. HouwINK TEN CATE, op. cit. (n. 8), p. 1 67. 12 w. ORTHMANN, op. cit. (n. 5), p. 238-239 and 291 -292; H. GENGE, op. cit. (n. 5), p. 48-50, 53-55, 84-86. 13 A small plaquette, discovered at Zincirli, bears three fragmentary lines of a cuneiform inscription, where W. SCHRAMM, "Usa " = Sam 'al, in Or 52 ( 1 983), p. 458-460, 7

SAM' AL OR *YU' ADDI

235

the time of the Aramaean settlement 1 4 . Therefore, the assumption that Sam'al and Y'dy are Anatolian place names 1 5 lacks any factual basis. The name of Sam'al, in Semitic "left hand", indicates the "North" and implies that the Aramaean settlers of the 1 0 century B.C. belonged to the northern branch of a tribe or group of clans. S am'al was probably the name of the city as well. As for Y'dy, that cannot be explained by a known Anatolian language, it must be a Semitic tribal name, derived from a personal name. It belongs to the same root as the Arabic verb 'dy, stem II of which ( 'addii) means "to lead". The personal name * Yu 'addi is probably attested in the Neo-Assyrian census of the district around ljarran, datable in the late 8th century B.C. We find there a patronymic mO-a-di-i 1 6, which was rightly related to Old Arabian per­ sonal names 1 7 . Although this name may be interpreted as Waddi, it is more likely that D stands there for initial yu- like in the next following name m O-a-si-i of the ljarran Census 1 8 , which is identical with Thamiidic Y's 1 y 1 9 /Yu 'assi/, and like in the name mO-a-a-te- ' designat­ ing the same person as mJa-ulu-telta- ' and standing for * Yuwaifi '20• The spelling in D- occurs in other names as well, like in mO-a-bu2 1 corre­ sponding to Wahb rather than to Y'b or Y'b22 • The root 'dy is attested also in the Old Arabian personal names 'dy ( 'Udaiy) and 'dyn, both �afaitic23. We suggest therefore that Y'dy should be vocalized * Yu 'addi, "he leads", a shortened name in which only the first element is pre­ served24. This reference to Old Arabian is strengthened by the name of has proposed to read the title "king of U�a", a city known from Hittite sources. The inscription would have been written after 1 200 B.C. for a local ruler, governing a small Luwian state before the Aramaean settlement. Because of the rude drawing of the inscrip­ tion, the authenticity of the object was questioned by H.Th. BOSSERT, Neues von Zincirli und Mara�. in Or 27 ( 1 958), p. 399-406 (see p. 402 ff.). i -1 There was no systematic excavation of the levels below those of the Aramaean period. 15 This suggestion was made by B. LANDSBERGER, op. cit. (n. 4), p. 22, n. 42; p. 36, n. 76; p. 40, n. 93. 16 SAA XI, 203, r. I, 1 2 ' . 1 7 ZADOK, WSB, p . 220. I H SAA XI, 203, r. I, 1 3' . 19 HARDING, Arabian Names, p . 655. 2 ° CF. M. WEIPPERT, Die Kiimpfe des assyrischen Kijnigs Assurbanipal gegen die Araber, in WO 7 ( 1 973-74), p. 39-85 (see p. 40, n. 6). 21 APN, p. 238a. 22 HARDING, Arabian Names, p. 65 1 , 654, 674. 21 Ibid., p. 33. 2 -1 This interpretation was already proposed by E. LIPINSKI, The God 'Arqii-Rashap in the Samal/ian Hadad Inscription, in M. SOKOLOFF (ed.), Arameans. Aramaic and the Ara­ maic Literary Tradition, Ramat-Gan 1 983, p. 1 5-2 1 (see p. 1 8). Very likely, the name

236

SAM'AL OR *YU ' ADDY

the third king of Sam'al who is called lfy ' in the Kilamuwa inscription, lfy in another inscription of Kilamuwa on a tiny sheath of gold25 , but

lja-a-ni, lja-a-nu, lja-ia-a-nu or lja-ia-ni in the annals of Shalmaneser III for 858, 857, and 853 B.C. 26 • The same name, spelt mlja-a-ia-a-nu, appears in a Neo-Babylonian letter from Nippur, going back to the mid8th century B.C.27 . This spelling undoubtedly represents the Semitic name /fayyiin which has an exact parallel in �afaitic, Sabaic, LiJ:tyanite, Nabataean, and Hatraean lfyn28 • In Greek inscriptions from the Syrian steppe, the name is transcribed as Ai:1av11(

AramGraf 29

Yaqiit, about A.D. 1225 , the fortress was ruined 1 36, and the site does not seem to have been occupied since then. It consists of two connected mounds, Qastiin al-Garbi and Qastiin as-Sarqi, that cover a surface of about 16 hectares, stretching for some 600 m from east-northeast to west-south­ west and being between 250 m and 300 m wide. The eastern acropolis raises more than 25 m above the surrounding ground, on the banks of the Nahr Qastiin. Without excavations, it is impossible to know how important this town was in Iron Age II, at the time of the Aramaean kingdom. Two mentions of Rg on the I:Iama slabs are followed respectively by qt'n 1 and qt�n�, where the 'n1 of the first example is fairly certain 1 37 • This is most likely the ancient name *Qatiin or *Qattiin of Qastiin. In fact, Qastiin can probably be identified with Qd-ta-an, which is men­ tioned in an Old Assyrian text from Ktiltepe 1 38 between Tu-ni-ip, proba136 YAQ0T, op. cit. (n. 7), vol. IV, p. 97. Cf. G. LE STRANGE, op. cit. (n. 8), p. 490. 1 17 OTZEN, Hama, p. 300, AramGraf 28-29. The second letter is very similar to the t of the lion-weights from Nineveh, that date from the last quarter of the 8th century B .C. m K. HECKER, Hitit (:ivi Yazmnm Kbkeni Hakkmda, in S . ALP (ed.), U/uslaras1 l. Hititoloji Kongresi Bildirileri ( 1 9-21 Tammuz /990), Ankara 1 993, p. 43-63 (see p. 53 ff., line 1 1 ). The editor suggests to identify this place with Qatna-Tell Mi�rife, but the spelling is unusual for this city name and the distance between the towns mentioned becomes bigger.

f:IAMATH AND LUGATH

274

bly Tell 'Asarne 1 39 , about 45 km south of Qastiin, and ljassu, very likely present-day I:Iass 1 40, 20 km east-southeast of Qastiin. Besides, a Hittite text from the time of Shuppiluliuma I seems to mention "'["Q]d-ta-an just before "'"Tit-ni-pa and after "'"/-di-bd-a 1 4 1 , that might be Idlib, 35 km northeast of Qastiin.

.-. · /) �

AramGraf 8

l

AramGraf 9

The words 'nn, written on a I:Iama slab about 10 cm from !$bh, and �nn, on another slab 142 , may be place names as well. In fact, a town ljanana is mentioned in the ljarran Census and both Anan and I:Ianin occur nowa­ days in Lebanon as toponyms. In Syria, Kafman ( t, since -r0 is a frequent Greek notation of geminated tt. 156 OTZEN, Hama, p. 302, AramGra/ 32. There is a striking similarity between the first letter and the cursive South Arabian s2 • 157 Cf. H.F. RussELL, art. cir. (n. 14), p. 63, and here above, p. 96. 1 18 Respectively OTZEN, Hama, p. 304, AramGra/ 34, and p. 30 1 , AramGra/ 30. 1 59 Cf. T. BAUZOU, Activite de la mission archeologique " Strata Dioc/etiana " en 1 990 cl /992, in Chronique archeologique en Syrie I ( 1 992 [ 1 997]), p. 1 36- 140 (see p. 1 37). 160 PToLEMY, Geography V, 1 8. These identifications and locations are proposed by DusSAUD, Topographie, p. 455-457. Despite some phonetic similarities, no connection can be established with the country Wadastani, "Water province", men­ tioned in the hieroglyphic Luwian inscriptions from MuI:irada and Qal 'at Sayzar, where the J:lama-Apamea road crosses the Orontes; cf. A.M. JASINK, op. cir. (n. 5), p. 1 04- 1 05 .

278

l:IAMATH AND LUGATH

AramGraf 33

AramGraf 34

AramGraf 30

The slabs from I:Iama date approximately from the same period as the archive of the virtually independent ruler of Siigu around 770 B.C. Now, one of the documents of this archive reports the ruler's victory over two thousand Aramaean tribesmen marching on Laqe and having a certain Ya'e, son of Balaam, as leader of the h,A-mat-a-a, one of the invading tribes 1 6 1 • This tribal name has been related to I:Iamath 162 , but it certainly refers to a seminomadic tribe in Babylonia 1 63 , not in the Syrian steppe, where the king of I:Iamath must have exercised a certain authority. If this is the case, as suggested by the presence of a I:Iamathite gover­ nor in the land of Laqe, by the royal correspondence with the ruler of Siigu, by Zakkiir's origin from 'Ana, and by the inscribed slabs from I:Iama, then one would expect that the I:Iamathite king also controlled the caravan route leading through the oasis of Palmyra to the land of Laqe. No explicit information is available so far, but one of the names appear1 6 1 A. CAVIGNEAUX - B. Kh. !SMAIL, Die Statthalter von Suhu und Mari im 8. Jh. v. Chr. anhand neuer Texte aus den irakischen Grabungen im Staugebiet des Qadisija­ Damms, in BaM 2 1 ( 1 990), p. 321 -456 and Pl. 35-38, in particular Nos. 2 and 3 = RIMB

II, text S.0. 1 002.2, p. 295, line 1 7 ; text S0. 1 002.3, p. 20 1 , line 8'. 162 D10N, Arameens, p. 62-63. 163 See here above, p. 1 06, and below, p. 425-428.

TERRITORY

279

f:limyarite relief from the 2nd or 3rd century A.D. with the inscription {1/bn, "the (sacred) precinct" (istanbul Arkeoloji Miizeleri, Inv. No. 76 1 1 ). The first letter {1 is identical with the last letter of AramGraf 33 and 34.

ing on the slabs of J:lama, namely ''w'tfm, may designate Begara, as sug­ gested above, while another one, Z'/ 164 , may refer to the district of kurA­ za-al-la in the Syrian steppe 1 65 . According to the Tell ar-Rimal:i stela, this area belonged to the province of Ra�appa 1 66 , and kurA-za-la-a-a occurs also in a royal appointment that may date from the reign of Ashurnasirpal II l 67 . The report of an Ashurbanipal' s campaign against the Arabs provides more information, since it mentions "'"A-za-al-la after "'"/a-ar-ki 1 68 , the present-day Arak, about 27 km east of Palmyra 1 69 • l Rab'o, 5 km north­ west of Ma�yaf 1 83 , where Iron Age remains have been identified, or to 1 7H OUSSAUD, Topographie, p. 239; A. CAQUOT, in G. TCHALENKO, op. cit. (n. 75), vol. 111, p. 68-69. 1 79 L.W. KING, op. cit. (n. 49), Pl. LXXV, band XIII, upper register. Cf. also ANEP, No. 365. I HO N. EGAMI - S. MASUDA (eds.), Tell Mastuma. Excavations in the ldlib District, Syria 1-1 1, Tokyo 1 984-88; Sh. WAKITA et al., Tell Mastuma - A Preliminary Report of the Excavations in ldlib, Syria. 1 993, in Bulletin of the Ancient Orient Museum 1 5 ( 1 994), p. 23-49; ID. et. al., Tell Mastuma, in BUNNENS, Syria, p. 509-530. IH I See here above p. 207 with n. 74. I Hl RIMA II, text A.O. I O I . I , p. 2 I 8, lines 78-82. The city is qualified as iii dannuti, while Kinalua/Kunulua is a iii .forruti; cf. Y. IKEDA, Royal Cities and Fortified Cities, in Iraq 4 1 ( 1 979), p. 75-87, in particular p. 76-77. The number of I:Iamathite "royal cities" in Shalmaneser III's annals (cf. ibid., p. 79-80) casts a doubt on the real significance of this distinction. IH J DussAUD, Topographie, p. 24 1 . It is impossible to localize Aribua at Gisr a�­ Sugiir, as proposed by M. LIVERANI, Studies in the Annals of Ashurnasirpal II, 2: Topo­ Rraphical Analysis, Roma 1 992, p. 76-77, since the fortress was a two days' march dis­ tant from the Sangura river (RIMA II, text A.0. 1 0 1 . 1 , p. 2 1 8, lines 80-8 1 ). Following the road on the left side of the Orontes, the Assyrian army passed between Mounts Saratinu and Qalpiinu, probably the Gebel Wastiini, east of the river, and the Gebel Qu�ayr, west

282

f:IAMATH AND LUGATH

the l:li�n al-1:fariba, described in the 12th century A.O. by Usama ibn MunqI 1 in Aramaic397 • These changes are not yet attested by the Assyrian name Man$uate, which still shows the basis n$W, semantically related to ll$$, in Arabic "to set up". It has long been recognized, in fact, that there is a semantic correlation between roots with second geminated consonant and those with third radical w or y398 . The Arabic parallel to 392 C.C. EDGAR, Zenon Papyri I (Catalogue generale des antiquites egyptiennes du Musee du Caire, Nos. 5900 1 -591 39), Le Caire 1 925, p. 1 1 3. 393 Cf. A. UNGNAD, in RLA II, Berlin-Leipzig 1 938, p. 427, CJ , Rs., line 1 9, cf. p. 447, sub Dananu; MILLARD, Eponyms, p. 5 1 , sub Dana.nu. 394 A. LEMAIRE - J.-M. DURAND, Les inscriptions arameens de Sfire et l 'Assyrie de Shamshi-ilu, Geneve-Paris 1 984, p. 83, rightly notice that only one letter is missing between m and w. Considering the geographical area that should be taken into account, a restitution [mn 'm]q w 'd 'rw w-m[�]w appears to be the most appropriate: "[from 'Um]q to 'Arro and to Ma[��]uwa", i.e. from the north to the south. Cf. above, p. 226. 395 M. VAN BERCHEM, Epigraphie des Assassins de Syrie, in JA, 9th ser., 9 ( 1 897), p. 453-50 1 , had already pointed out "qu'il s'agit d'un vieux nom indigene que Jes Arabes ont cherche a ramener a une forrne de leur grammaire et a une racine de leur dictionnaire" (p. 458, in footnote). 396 Ibid., p. 457-458, n. 2. 397 LIPINSKI, Semitic, respectively § 27.3 and § I I. 1 0. 398 This was first recognized by B. LANDSBERGER, Die Eigenbegrijjlichkeit der baby­ /onischen Welt, in lslamica 2 ( 1 926), p. 355-372 (see p. 362 ff.). This is very clear in the present case, since the same root exists in Egyptian with nominal derivatives mj and ncfw, both meaning "support, protection". It is evident, instead, that the place name can­ not derive from the root n�y. "to quarrel", as proposed by R. ZAD0K, Historical and Onomastic Notes, in WO 9 ( 1 977), p. 35-56 (see p. 56). In fact, a basic principle of

HISTORY

309

*man�uwatu, coined on the basis n��. is mina��atu(n), "raised platform", which indicates that the city name originally had a similar meaning. It perfectly suits the site of the fortress of Ma�yaf, built on the platform formed by a rocky spur of the hill that dominates the fortified borough. In conclusion, the toponymy strongly suggests the identification of Man�uate with Ma�yaf. This location is also supported by the lists of Assyrian provinces which mention together Man�uate, Sumur, ljaQarik, and Soba399 , or Man�uate and Sumur400 , or Soba, Man�uate, and ljaQarik401 . Besides, a lexical list from Ashurbanipal' s reign, which cannot be considered as an administrative document enumerating the provinces of the Empire, men­ tions I:Iamath, ljaQarik, and Man�uate in succession402 . The province of Man�uate seems thus to have common borders with the province of ljaQarik to the northeast, with that of Soba to the south, and with the province of Sumur to the southwest. It probably included the region of I:Iamath in the east, since I:Iamath as such does not appear as the chief­ town of an Assyrian province in the 7th century B.C. The territory of Man�uate itself would then correspond to the southern region of Gebel An�ar'iya and would have belonged most likely to the former kingdom of I:Iamath, being its western province403 • The identification of Man�uate with Ma�yaf suggests to connect the Assyrian campaign "as far as Man�uate" with the events reported on the stela of Zakkiir. Besieged in his capital ljaQarik by Bar-Hadad II, king of Damascus, and his allies, he was most likely rescued by an Assyrian army. This explanation has already been proposed by M. Lidzbarski. However, he dated these events from the year 772, when an Assyrian campaign to ljaQarik had taken place404 . But this date is too low to per­ mit a connection with the inscription of Zakkiir. A. Jepsen has then related the liberation of Zakkiir's capital with the Assyrian campaign to anthroponomy and toponymy is that the proper name must be originally meaningful as name of a concrete entity that belongs to a well defined category. One can hardly imag­ ine a city called "bone of contention" or the like. 199 SAA VII, 1 1 6, 7'- 1 0' . l:lamath is missing in this list and in the lists referred to in n. 400 and 40 1 . The city belonged then most likely to the province of Man�uate. The dis­ tance by road between l:lamath and Man�uate amounts to 42 km. 4 a in front of r in QUrr, "free-born", "noble", and that the name means "noble young man". In any case, this noun Qiirr or Qarr < QUrr is unrelated to "Hunian", as suggested by B. PODOLSKY, Notes on Hebrew Etymology, in /OS 18 ( 1998), p. 199-205 (see p. 20 1-202). B. MORITZ, art. cit. (n. 7 1), p. 199, explains the name as a metathesis of Arabic sbbr.

ljINDARU

457

mRa-a-pi-' 380, the Aramaic or Arabic active participle Riipi' , "healer" , "restorer". At least two of the four names are North-Arabian and thus reveal the mixed nature of the tribe, a fact which may explain its very generic name "nomads", Qiddiir. This name is uncompromisingly Ara­ maic, because the root Qdr does not have such appropriate meanings in Arabic381 , but the various groups fonning the tribe may have spoken two different languages, respectively Aramaic and Arabic. Such a situation is by no means unique. For example, on another continent, the Amba of Uganda are considered one tribe though they speak two mutually unin­ telligible languages. The appearance of ethnically diversified tribes in Babylonia is quite understandable, for one has to reckon with an Arab penetration in the region at least by the first half of the 8th century B.C.Js2 _ The tribal territory of the IJindaru at the time of Sargon II may already have been the area east of the Shan al-l:lay. In Sargon II's texts, the IJindaru are often associated to the Aramaean tribes Gambiilu, Ru'a' , and Puqiidu, in southeastem Babylonia383 . The early Neo-Babylonian letter from Nippur384 confirms their location near the territories of these tribes, where the IJindaru either had established permanent settlements, or were seasonally drawing their migratory orbits, which were deter­ mined by the grazing facilities for their herds. At the time of Sen­ nacherib, they were settled along the Lower Uqnu385 , close to the Persian Gulf. Sennacherib mentions them several times among the Aramaean tribes he has subjugated386• The IJindaru are also referred to in a few let­ ters of the Sargonid period387 • 380 Sargon II, p. 1 47, Ann. 286. 381 The root occurs with the meanings "to be thick", "to bring down", "to come down", and �iidir is "thick". 382 I. EPH'AL, an. cit. (n. 72), p. 108- 1 1 5 . The mid-8th century, proposed by St.W. COLE, op. cit. (n. 70), p. 34-42, is a too low date. More material is found in the archive published in Nippur IV. See, for instance, here above, p. 43 1 with n. 1 4 1 , concerning the tribe of Hamdan. 383 S�rgon II, p. 1 36, Ann. 258; p. 1 46, Ann. 28 1 -282; p. 1 47, Ann. 286; p. 1 95, Displ. 1 9; p. 227, Displ. 1 26- 1 27; p. 265, Pavem. IV 74 and 76. See also G. FRAME, an. cit. (n. 52), line 32. 384 Nippur IV, No. 1 3. 385 Sennacherib, p. 49, lines 1 3- 1 4. 386 Sennacherib, p. 25, line 47; p. 43, line 5 1 ; p. 49, line 1 3 ; p. 54, line 55; p. 57, line 1 5 ; Sanherib, p. 5 1 , line 1 3. 387 ABL 848, 8 (1"/ji-in-dar-a-a) and 1 1 (Jji-in-dar); 850, 8; 865,5; M. DIETRICH, CT 54, London 1 979, No. 85, 1 3 and 1 4. These references should possibly be complemented by the mentions of 1"/n-da-ra-a-a, if this happens to be a variant spelling for Jjindiiraya; cf. here above, p. 455-456 with n. 369.

ARAMAEANS IN BABYLONIA

458

The '"Da-mu-nu are the next tribe in the great Tiglath-pileser III' s list388, but their case is examined more conveniently together with that of Amlatu389. 16. DuNANU, NILQU, RADIYE

The people called 1"Du-na-nu390 or 1"Du-na-a-nu39 1 bear a name that is apparently identical with the personal name mou-na-nu of the chief of the Gambiilu tribe who opposed Ashurbanipal392 , but the Dunanu, unlike Guliis, are really one of the tribes of Babylonia labelled specifically by Tiglath-pileser III as Aramaean. They are clearly referred to as a tribe in the early Neo-Babylonian archive from Nippur393, where the man who seems to act as their sheikh bears the uncompromisingly Aramaean name mBa-lJi-a-nu394, i.e. Bagyiin, and is the namesake of the epony­ mous ancestor of the Aramaean tribe of Bet-Bagyan395• The tribal name 1 "Du-na-a-nu appears to be the "broken" plural of a noun dunnu, that is given as a synonym of dannu, "strong", in Assyro-Babylonian lexical lists396 . In fact, the pattern qutiil was used in early Westsemitic to form the collective plural of nouns belonging also to the qutl/qull stem397 , as can still be seen in Hebrew where the external plural in -im is superim­ posed on original qutiil forms398 , as in qi5diis-im, plural of *quds, "holy thing", or in si5riis-im, plural of *surs, "root". The Dunanu would thus be "the strong ones" or the like. The name of the tribe is of course older than its earliest attestations in the mid-8th century B.C. The territory of the Dunanu was probably situated at that time to the north or to the northwest of Nippur, since Bagyan (mBa-lJi-ia-nu), appar­ ently the chief of the tribe, asks the governor of Nippur for news about JHH

Tig/. Ill, p. 1 60, line 7; cf. p. 66, line 3. Neo-Assyrian references are collected in

PARPOLA, Toponyms, p. 98. JH9

See here below, p. 462-463. 390 Tigl. Ill, p. 360, line 7; cf. p. 42, line 6; p. 1 22, line 4; p. 1 94, line 6. WI Nippur IV, Nos. 60, 7; 6 1 , 8. wz APN, p. 7 1 b. m Nippur IV, Nos. 60, 7; 6 1 , 8. 394 Nippur IV, No. 60, 6b-9a. The same name still occurs in this archive under Nos. 1 3, I ; 2 1 , I ; I O I , I, and probably designates the same man. 395 See here above, p. I 1 9. 396 AHw, p. 1 77a; CAD, D. p. 1 84- 1 85. The root dnn is not restricted lo East Semitic. It occurs also in Amorite anthroponomy (dunnu) and in Ugaritic (dn). 97 Its use in Classical Arabic is almost restricted to singulars of the fii 'i/ and fa 'i/ stems designating human beings. · 9" LIPINSKI, Semitic, § 3 1 .28, with examples of the parallel qitiil pattern. 1

1

DUNANU, NILQU, RADiYE

459

Chaldaea and in particular about Mukin-zeri399 , the chief of the Chal­ daean tribe of B"it-Amukkani which was obviously living closer to Nip­ pur than to the residence of Bagyan. This favours the identification of the town uruPa-#-tu of the Dunanu, which was conquered by Tiglath­ pileser III400, with Bit-Ba�ftu, a settlement in the neighbourhood of Bor­ sippa40 1 . Tiglath-pileser III's enumeration of the conquered cities "from Dur-Kurigalzu, Sippar of Samas, . . . Pa�Itu of the Dunanu as far as Nip­ pur"402 confirms the northern location of the Dunanu tribe and seems to indicate that Pa�1tu was a rather important centre in the mid-8th century B.C. After Tiglath-pileser III, the Dunanu are not mentioned any more. Nothing is known about the tribe called 1"Ni-il-qu in Tiglath-pileser III's list4°3 • R. Zadok' s suggestion identifying this tribal name with the place name uruNiqqu ( > Nilqu) in the Sippar region404 cannot be proved, but perhaps offers the possibility of finding an acceptable meaning for the tribal name. However, another uruNiqqu should be taken here into account, viz. uru/ku'Niqqu sa ku•Tublias4°5 • In fact, Tubliash is the name of a stream and of a large region southeast of the Diyala river406 , near the Elamite border, and the next mentioned tribe should probably be located there as well. As for the meaning of the tribal name, one might thing of a "broken" plural of naqly, "pure", attested also in Aramaic, but this col­ lective plural is niqa ' in Classical Arabic. The 1"Ni-il-qu are followed by the "'Ra-di-e407, possibly related to the place name Rade near the Elamite border4°8 • One cannot explain this name by Arabic radu 'a, "to be bad", or radiya, "to perish". In Aramaic, however, the root riJda, "to go", has various meanings and derivatives that offer an acceptable sense either for a place name or a tribal name, in particular radayya, "running water", and radiy/radya, "plougher", with a plural radayya. The meaning "to plough" is attested already in Imperial 199 Nippur IV, No. 2 1 , 8- 1 2. 41x1 Tigl. Ill, p. 1 22, line 4; cf. p. 42, line 6. 4111 RGTC 8, p. 84. Cf. R. ZADOK, art. cit. (n. 37), p. 84. The Assyrian spelling with initial p can be explained by the loss of the phonemic distinction between the labials b and p in Neo-Assyrian; cf. SA/O I, p. 92, 1 1 2; LIPINSKI, Semitic, § 1 1 .4. 4112 Tigl. Ill, p. 1 22, lines 4-5. 401 Tigl. Ill, p. 1 60, line 7. 4114 R. ZADOK, art. cit. (n. 37), p. 70; RGTC 8, p. 243. 405 Tigl. Ill, p. 1 02, line 3 ' ; p. 1 24, line 1 7 ; p. 1 32, line 1 5 ' 1 ; p. 1 64, line 29; p. 1 66, line 35; cf. p. 72, line 1 2. 406 J .V. KINNIER WILSON, The Kurba 'ii Statue of Sha/maneser Ill, in Iraq 24 ( 1 962), p. 90- 1 1 5 (see p. 1 1 3 - 1 1 5). For Niqqu, see also BRINKMAN, PKB, p. 232 with n. 1 468. 4o7 Tigl. Ill, p. 1 60, line 7. 4ox ABL 28 1 = SLA 44, r. 1 6; ABL 1 464, 6; cf. RGTC 8, p. 257.

460

ARAMAEANS IN BABYLONIA

Aramaic, in the 5th century B.C.409, and there was even an angel of rain and field labour, called Ridyii4 1 0• The 1 "Ra-di-e are followed by a badly damaged name 1"Da-[x-x-x] and by the 1 "V-bu-lu4 1 1 •

17. 'UBOLU The ' Ubiilu tribe is counted among the Aramaean tribes of Babylonia in the inscriptions of Tiglath-pileser III, of Sargon II, and of Sen­ nacherib4 1 2. In Neo-Babylonian sources, the 'Ubiilu appear only in two letters from Nippur, in the mid-8th century4 1 3, and in one document from Ur4 1 4. Like other Aramaean tribal names4 1 5, ' Ubul is a "broken" plural corresponding to the singular 'iibil, the camel-driver or rather the man who takes care of the dromedaries, the camelherd. This word is not attested so far in Aramaic texts, but the man in charge of David's camels was called 'wbyl (co�tA) according to I Chron. 27, 30. This is an Hebrew form, characterized by the change ii > o; it goes back to an original 'iibil and therefore does not correspond to Arabic 'abbiil, "camel-driver". David's 'obll is explicitly identified in I Chron. 27, 30 as Ishmaelite, what the Targum has rightly translated by "Arab"4 1 6. True, a driver seated on a dromedary is already represented on a relief from Tell I:Ialaf dating from the 9th century B.C.417 and the camel-driver had to be des­ ignated in Aramaic by some word, but we cannot simply assume that 'iibil was an Early Aramaic term meaning "camel-driver". An Old Ara­ bian word suits the biblical context better. In any case, the regular "bro­ ken" plural of ' iibil can be 'ubft/4 1 8, and the name of the 'Ubiilu tribe can therefore mean "camel-drivers" or "camelherds". This noun was proba­ bly a denominative formation based on ibil(u), "camel(s)" or rather "dromedaries" in both Assyro-Babylonian and Arabic. The place names lbulu and lbule4 1 9 may reflect the frequent alternation i/u, or be 409

TAD I, A5.2, 4. JASTROW, p. 1 452a. 41 1 Tigl. l/l, p. 160, line 7; cf. p. 194, line 6. 412 For the attestations, see PARPOLA, Toponyms, p. 364. 4 1 3 Nippur IV, Nos. 32, 8 and 98, 17. 414 H.H. FtGULLA, VET V, London 1 953, No. 140, 8. See RGTC 8, p. 3 17. 4 1 5 See here above p. 430. 4 16 A. SPERBER, The Bible in Aramaic IVA, Leiden 1 968, p. 29. 4 1 7 ANEP, No. 1 88. 418 The plural 'ubbal is attested in the classical period. 4 1 9 PARPOLA, Toponyms, p. 172 (on the territory of the Puqiidu tribe); RGTC 8, p. 1 78. There is also an "'"Ubula in another region: Tigl. Ill, p. 126, line 31; p. 184, line [ 1 5']. 41 0

KARMA.'

46 1

prompted by the initial i of ibilu, or both, but they do not need to be con­ nected with the tribe. They may represent another derivative and, for instance, designate the place where camel caravans converge. Interestingly enough, the author of one of the earliest texts dealing with the ' Ubiilu, namely the writer of a letter from the archive in Nippur, assumes that the ' Ubiilu may have stolen the dromedaries of some peo­ ple from Uruk420. Another letter to the governor of Nippur is apparently concerned with a dispute over a parcel of land involving the sheikh of the ' Ubiilu and the sender of the letter42 1 . The land concerned may have been traditional pasture ground of the tribe, but then had attracted culti­ vators. The Assyrian royal inscriptions do not provide any particular information concerning the ' Ubiilu tribe, but its territory can safely be located east of the Tigris. The Ubal river of Dan. 8, 2.3.6 may still recall the name of the tribe in the 2nd century B.C., but then it cannot be the stream of the city of Susa, as stated in Dan. 8, namely the modem Sa'fir which is an ann of the Kerha river, the Ab-i Diz (river of Dizful), or the nearby Kariin river422• 18. KARMA'

In Tiglath-pileser III's list of the Aramaean tribes, the 'Ubiilu are fol­ lowed by "'Kar-ma- ' 423 . It is unlikely that there is any connection between this tribe and the uruKar-me-e of some Nineveh texts from the time of the Sargonids424• This town is undoubtedly identical with the Nestorian bishopric of Kanne, located on the left bank of the Tigris, somewhere opposite to Tikrlt; it is still attested in the 5th and 6th cen­ turies A.O. by the names of its bishops who had taken part in various councils425 • It is also unlikely that luKar-ma-' derives from a tribalized place name Karma', "the vineyard" or "barren hill", if the toponym is a derivative of Assyro-Babylonian karmu. One should rather assume that 420 Nippur IV, No. 32. Nippur IV, No. 98. 422 In Dan. 8, 2, the transcription ' wly either of the Assyro-Babylonian name Olaya or of the Greek name Eu)caioc; of the river of Susa is apposed to the name of Ubal, but it was not read by Theodolion. 423 Tigl. III, p. 160, line 7. 424 ABL 574, 6; SAA XI, 231, 6 ' . 425 For the bishops mentioned in A.O. 486, 497, 554, see E. SACHAU, Zur Ausbrei­ tung des Christentums in Asien (APAW. PH 19 19-I), Berlin 19 1 9, p. 26-38, in particular No. 2 1 . 421

462

ARAMAEANS IN BABYLONIA

Karma' is related to Arabic karim, "noble", pronounced with an initial unvoiced velar, contrary to Guriimu. Now, the classical "broken" plural of this North-Arabian noun is kurama', but unstressed short vowels fol­ lowing upon a syllable with short vowel were normally elided in the Eastern dialects, whereas they changed the quality of the preceding vowel. This combined process of vowel harmony and elision produced kilma for kalima, "word", �udqa for �aduqa, "dowry", fib ii is attested, e.g., by *qii qiiqiina, "judges". The final vowel -a of the plural morpheme is not heard in the spoken language. m The Damunu are mentioned in Sargon II, p. 227 and 350, Displ. 126; also p. 33 and 290, Cyl. 12; p. 65 and 304, Bull 29. 4 4 .1 Sargon II, p. 227-228 and 350, Displ. 129- 13 1 . 4.15 Sargon II, p. 65 and 304, Bull 29-30; Sennacherib, p. 49, lines 13-14. 4 6 -' Sargon II, references above, in n. 433; Sennacherib, p. 25, line 46-47; p. 43, line 48-5 1 ; p. 49, line 13; p. 54, lines 55-56; p. 57, lines 14- 15; ABL 846 = SLA 73, r. 9. rn Sennacherib, p. 43, lines 43-57; p. 48-49, lines 6- 15.

464

ARAMAEANS JN BABYLONIA

battle with them further, like with other tribes of the region, and to pro­ ceed to massive deportations438 . The Damiinu are still mentioned under the reign of Ashurbanipal439• They seem to be related somehow to the 1 "Ru-�a-pi440 and possibly to the IiWa-#-[ru?] 44 1 • The latter tribe is named in Tiglath-pileser III's list and seems to be North-Arabian as we11442, while the first one is known only from a single letter. A Qatabanic clan or tribe {j-R�fm is named in a South-Arabian inscription443 and this resemblance between the two names favours an Arabian origin for the 1"Ru-�a-pi as well. The Arabic plural ru�afii', "associates", suggests con­ sidering the name as the collective "broken" plural of ra�lplf, "firmly joined".

20. Ru' A'

AND QABI'

The next tribe named in the long Tiglath-pileser III's list is North-Ara­ bian likewise. In fact, the name 1"Ru- '-a444 or 1 "Ru-it '-a445 exactly corre­ sponds to the Arabic "broken" plural ru 'ii ', "shepherds", which is formed on the rarely used pattern Ju 'ii/446 • The cuneiform spelling IuRu-u-a-a447 probably reflects a w-glide between u and a, while 1"Ru-u- '-a- a448 , uruRu­ '-u- a449 , and IuRu- '-ulu/it '-a450, are a kind of lectiones conflatae. Instead, it does not appear that KUR Ru-ulu-a mentioned in Tiglath-pileser III's inscriptions45 1 has some relation to the name of the tribe. The proposed interpretation of the tribal name seems to be confirmed by a detail preserved in a letter from the time of Ashurbanipal, where the 438

Sennacherib, p. 25, lines 43-53; p. 55, line 60; p. 57, line 1 6. References to the correspondence in RGTC 8, p. 1 1 6. 440 ABL 846 = SLA 73, r. 3-4. 441 ABL 846 = SLA 73, r. 2. 442 See here above, p. 447. 443 RES 4 1 1 8, 2. 444 Tigl. Ill, p. 1 60, line 7. References to the Neo-Assyrian texts are collected in PARPOLA, Toponyms, p. 294-295. 445 Sargon II, p. 1 36, Ann. 258; p. 1 46, Ann. 28 1 ; p. 1 47, Ann. 286. 446 w. FISCHER, op. cit. (n. 1 30), p. 53, § 90. 441 ABL 287 = SLA 84, 6; ABL 1 468, r. 6; NBC 6 142, 1 3 (unpubl.: RGTC 8, p. 427; here below, p. 465, n. 459). 448 ABL 1 222, r. 6. 449 ABL 268 = SLA 253, 10. 450 Tigl. Ill, p. 1 22, line 12; p. 1 30, line 5; Sargon II, p. 1 95, Displ. 1 9; p. 227, Displ. 1 27; Pavem. IV 74; Sennacherib, p. 25, line 47; p. 43, line 49; p. 49, line 1 3; p. 54, line 56; p. 57, line 1 5 ; Sanherib, p. 5 1 , line 1 3. 451 Tigl. Ill, p. 72, line 6; p. 1 32, line 9; p. 1 64, line 32. 439

RU' A' AND QABI'

465

flocks of the temple of Uruk are said to graze freely with those of the Puqiidu tribe in uruRu- '-u-a452. One wonders whether this is a settlement inhabited by members of the Ru'a' tribe or a "shepherds' land". At any rate, the North-Arabian extraction of the tribe is corroborated by personal names. A sheikh of the Ru'a' named in the annals of Sargon II is called mAb-ba-ta-a453, i.e. 'Ab- 'ata, "The father is fierce". The root 'ata is well represented in Arabic, but it is not used in Aramaic, and the related bib­ lical name 'Attay454 is rightly considered as Arabic by M. Noth455 • In fact, it is encountered in Palmyra ('ty, A00ma)456 and at Dura Europos (A00at0(;)457, also in �afaitic and in Thamiidic458 • A witness of a legal document from Nippur, dated to 664 B.C., is called mNa-(e-ru and is identified as a member of the Ru'a' tribe459 . This name is clearly Arabic and means "keeper", "guard", like the substantivated active participle na(ir460• It is amply attested in Li�yanite (N(r), �afaitic (N?r), Nabataean (N(rw, Ntyrw)461 , and in Greek transcription (Na-rapO(;, Nonpo�, Non:po�)462• The Aramaic root is spelt n(r already in the Proverbs of A�iqar, composed probably in the 7th century B.C., but it does not appear in Aramaean anthroponomy before the 5th century B.C.463 . A Bit­ mNa-(e-ri appears a few years later in another document from Nippur464 , followed by uruBit-mNa-(e-er in Late Babylonian texts465 . Three other names of members of the Ru'a' tribe are known thanks to a letter from Ashurbanipal's time466• One of these men bears the ABL 268 = SIA 253, 8- 1 0. Sargon II, p. 1 47, Ann. 286. The verbal predicate of this name is related to an unknown root */Jty by R. ZADOK, The Jews in Babylonia during the Chaldean and Achaemenian Periods according to the Babylonian Sources, Haifa 1979, p. 20, who 452

453

seems to refer at the same time to the root bt', "to smite". 454 I Chron. 2, 35. 36; 1 2, 1 1 ; II Chron. 1 1 , 20. 455 M. NOTH, Die israelitischen Personennamen im Rahmen der gemeinsemitischen Namengebung, Stutgart 1928, p. 1 9 1 . 4S6 J .K. STARK, op. cit. (n. 1 36), p . 46, 1 08. 457 C.B. WELLES - R.0. FINK - J.F. GILLIAM, The Excavations at Dura-Europos. Final Report VII . Parchments and Papyri, New Haven 1959, No. 23, 2.7. 1 1 . 458 HARDING, Arabian Names, p. 442. 459 NBC 6 1 42, line 1 3 , quoted by J .A. BRINKMAN, op. cit. (n. 70), p. 1 2- 1 3, n. 46. 460 Nippur IV, p. 199, relates this name to Arabic n;r, "to look", but n;r and n{r go back to ntr. 46 1 HARDING, Arabian Names, p. 591, 592; A. NEGEV, op. cit. (n. 1 36), p. 43, No. 727. 462 H. WUTHNOW, op. cit. (n. 279), p. 82, 1 5 1 ; A. NEGEV, op. cit. (n. 1 36), p. 1 35, No. 2 1 77; p. 204. 4 63 R. ZADOK, op. cit. (n. 453), p. 77, 1 03. 464 Nippur IV, No. 95, 22. 4 65 RGTC 8, p. 98. 466 ABL 287 = SI.A 84, 4-6.

466

ARAMAEANS IN BABYLONIA

abridged Assyro-Babylonian name mRe-mut461 , while the second one, mlja-an-na-na, has a name that can be either Aramaic or Arabic468 • The third man, called mA-a-il-a- 'i, bears the North-Arabian name /fayy(a)­ 'iliihi, "as my god liveth!" This name is attested in Nabataean (/fy 'lhy) and in �afaitic (/fy 'lh)469• Besides, there are many examples of the name lfy 'l in �afaitic inscriptions470. As for the frequent comparison of Ru 'ii ' with biblical R 'w (Payau)47 1 , it is based on a confusion and does not offer any clue. Tiglath-pileser III mentions the Ru'a' tribe in connection with his campaigns in southeastern Babylonia472 . Sargon II, who again battled against this tribe, locates it on the banks of the Tigris473 . He claims that he has subjugated them474, but it was no decisive victory, since Sen­ nacherib had to fight again against the same tribes, inclusive the Ru'a' 475 . An Ashurbanipal's letter to the people of Nippur indicates that the relations between the Ru'a' tribesmen and the Assyrian authorities were not good at that time either476 • Cf. APN, p. 188. ZADOK, WSB, p. 1 18; ID., op. cit. (n. 453), p. 30. This is one of the best attested �afaitic names: HARDING, Arabian Names, p. 206. Its pronunciation was most likely iden­ tical with that of the Arabic adjective �anniin, "affectionate, compassionate", and would thus correspond perfectly to the spelling lja-an-na-na, with geminated first n. 469 A. NEGEV, op. cit. (n. 136), p. 28, No. 43 1; HARDING, Arabian Names, p. 209. Not knowing these parallels, the present writer has interpreted the name of the Ru'a' tribesman in the sense "Where is God?": E. LIPINSKI, Recherches ugaritiques. I. - Ay, un dieu ugaritique? - Etude onomastique, in Syria 44 ( 1967), p. 253-282 (see p. 275). Its translation by "Ea is my god" in PNA 1/1, p. 9 1, is very likely erroneous as well: Ea can hardly be used as theophorous element by Ru'a' tribesmen. The name should instead be related to the Amorite names of the type lja-a-ia-a-bu-um (E. LIPINSKI, art. cit., p. 260) and lja-a-iil-i-lu-u (ARM X, 146, 5), to the Hebrew phrases �ay- 'E/, "as God liveth !" (Job 27, 2), and �ay- Yhwh, "as Yahweh liveth !" (I Sam. 20,3; 25, 26; II Kings 2, 2.6), and to the later Nabataean name lfy'bdt from 'Avdat (Negev), which means "as (king) Obodas liveth !"; cf. RES 527; J.T. MILIK, Notes d'epigraphie et de topographie jordaniennes, in Studii Biblici Franciscani /iber annuus 10 ( 1959-60), p. 147- 184 (see p. 148). 470 HARDING, Arabian Names, p. 209; Safaitic Cairns, p. 571. The name occurs also in Ugaritic (lja-ia8-El, /fy 'il, lfyl: F. GRONDAHL, Die Personennamen aus Ugarit [Studia Pohl I ), Rom 1967, p. 16, 43, 66, 97, 137, 330, 385-386) and in the Bible (I Kings 16, 34), where it should not be interpreted as (A)�i'el. Besides, the Gambiilaean place name "'"lja-i/-a-a (Sargon II, p. 145, Ann. 279n) may be based on this personal name. 471 Gen. 1 1, 18-2 1; I Chron I, 25. The comparison was already made by F. DELITZSCH, Wo lag das Paradies ?, Leipzig 188 1, p. 238 ff. 47 2 Tigl. Ill, p. 1 22, line 12; p. 130, line 5; p. 160, line 7. 47J Sargon II, p. 195 and 343, Displ. 18- 19; p. 265 and 360, Pavem. IV 70-74. 474 Sargon II, p. 147 and 329, Ann. 286-287. 475 Sennacherib, p. 25, line 47; p. 43, line 49; p. 49, line 13; p. 54, line 56; p. 57, line 15; Sanherib, p. 5 1, line 13. 47,, ABL 287 = SLA 84, 4-9. 467

468

Ll'ITIA ' U

467

The next tribal name 1"Qa-bi- ' 477 can be interpreted in two different ways. Reading Qabyii ', we obtain the Syrian noun qebyii, "swamp", with a geographical connotation which could be referred to the swamps in southeastern Babylonia. Instead, if we read Qabi', - which is the more likely reading, - we have the passive participle "settled" of the Aramaic verb qaba', "to fix, to settle", which implies that the tribe in question had settled or represented the settled part of a larger group. The lack of any concrete information prevents us from further examining this case. 21. LI'ITTA'U

The 1"Qa-bi-' are followed by the 1"Li- '-ta-a-u478 , whose name is spelt also 1"Li- '-ta-u, 1 "Li- '-it-ta-a-a, 1 "Li- '-ta-a-a, 1"Li-ta-a-a479 ; one should probably add the Babylonian settlement of u"'Li-ta-mu480 and the per­ sonal or ethnic name mLi-ta-mi481 • Since the suffix -ii 'ul-iiwu is a typi­ cally Arabian morpheme, the base of the name should not be linked with the Aramaic element *li 't482 . Besides, one of the variant spellings indi­ cates that the t was geminated and this confirms the statement by Arab grammarians that there was a plural pattern fi 'illii 'u, which is not attested in the classical language483 , but was obviously not invented by the grammarians themselves. Since North-Arabian l is regularly expressed in cuneiform script by signs with t484, especially ta and at, the name of the tribe should most likely be considered as a "broken" plural *li 'illii 'u/*liyiyiiwu of the noun layf, "lion", which was used also as per­ sonal name485 . This pattern is of course different from the classical plural luyuf, "lions", but the apparent replacement of -ii'u by -iiwu 477 Tigl. Ill, p. 1 60, line 7; cf. p. 84, Ann. 7, line 3 (Qa-ab- '-e). Perhaps Qw' in Ez. 23,23 with spirantized h > w. m Tigl. Ill, p. 1 60, line 7; cf. p. 1 22, line 1 2; p. 1 30, line 6; p. 1 94, line 6; p. 1 96, line 6. 479 The references are collected in PARPOLA, Toponyms, p. 226; BRINKMAN, PKB, p. 270-27 1 , n. 1 738; W. RbLLIG, li'ta'u, in Rf.A VII, Berlin 1 987-90, p. 33. 480 RGTC 8, p. 2 1 3. 481 Babylonian tablet BM. 40548, dated to the 9th year of Eriba-Marduk (ea. 765 B.C.) and quoted by BRINKMAN, PKB, p. 270-27 1 , n. 1 738. 482 For this element, see ZADOK, WSB, p. 98-99. 4"1 w. FISCHER, op. cit. (n. 1 30), p. 52, § 90, Anm. I . 4"4 LIPINSKI, Semitic, § 1 3.9. m Hebrew in I Sam. 25, 44; II Sam. 3, 1 5. Arabic in W. CASKEL, op. cit. (n. 235), vol. II, p. 376b, s.v. Laif. In �afaitic names, layf is monophthongized to Lt HARDING, Arabian Names, p. 5 I 0, also in Thamiidic 'b/J, "My father is a lion" (ibid., p. 1 5).

468

ARAMAEANS IN BABYLONIA

(Babylonian -ta-mu/-ta-mi) reminds us of the 'ibdiil rules of Arab gram­ marians and strengthens the assumption that the name of the tribe is North-Arabian. Before Tiglath-pileser III' s reports of his campaigns against the allegedly Aramaean tribes in Babylonia, the mention of a certain mKa­ bi-tu mar mLi-ta-mi in a legal document from Nippur, dated ea. 765 B.C.486, is the earliest attestation of this tribe in Babylonia as yet. A Nimrud letter seems to present the 1 "Li- '-ta-mu-[(i,1)] as submissive vas­ sals of the Assyrian king487 , but Sargon II names them again among the tribes he has to fight on the banks of the Tigris, in southeastem Babylo­ nia488. According to a Nineveh letter, they were allies of the Chaldaean tribe of Bit-Yak:In489. Sennacherib mentions the 1"Li- '-ta-(a-)u again among the Aramaean enemies that he deported to Assyria490, but the stereotyped list of his inscriptions cannot be taken at its face value. The settlement and the lake of Li-ta-mu were almost in the suburbs of Babylon, on the Euphrates49 1 . Their name can be linked with the name of the tribe and with the mar mLi-ta-mi of the Nippur document from ea. 765 B.C., but it is hard to understand how a relation can be established between these North-Babylonian records and the presence of the tribe in the area of the Uqnii river. 22. MARUSU, 'AMMATU, ijAGGARANU

So far, only Tiglath-pileser III mentions the 1 "Ma-ru-su492, without providing any particular information. The first syllable ma- of the name may suggest that this is a toponym *Ma 'rus derived from the root 'rs, that gives 'arls, "hut", in Arabic and 'ars, "tribe", in Maghrebine Ara­ bic. There are other possibilities. In Tiglath-pileser III' s list, the 16Ma-ru-su are followed by the 1"A-ma­ 493 tu . The Aramaean tribe of 'Ammatu ('"A-mat-a-a) is first mentioned See here above, p. 423, n. 76, and also p. 467, n. 48 1 . H.W.F. SAGOS, art. cit. (n. 3 1 5), in Iraq 1 7 ( 1 955), No. IX, 12', p. 40 and Pl. IX; cf. No. IV, 9', p. 32 and PI. VI. 488 Sargon II, p. I 95 and 343, Displ. I 8- 19; p. 265 and 360, Pavem. IV 70-74. 489 ABL 436, 9. 490 Sennacherib, p. 25, line 49; p. 49, line 14; p. 54, line 56; p. 57, line 1 5 ; Sanherib, p. 5 1 , line 1 3. 491 RGTC 8, p. 2 1 3 . 492 Tigl. Ill, p . 1 60, line 7; cf. p . 1 30, line 6; p . 1 94, line 6; p . 1 96, line 6. 493 Tigl. Ill, p. 1 60, line 8. It is unlikely that the eastern toponym Amat, mentioned in Tiglath-pileser III's inscriptions, is related to the tribe: Tigl. Ill, p. 72, line 6 ( k"'A-ma-te); p. 1 06, line 33' (k "'A-[mat)). 486 487

MARUSU, 'AMMATU, IJAGGARA.NU

469

about 770 or 760 B.C. in the inscriptions of Ninurta-kudurri-u�ur494, the semi-independent ruler of Siibu on the Middle Euphrates. It is associated there with the great IJaiallii tribe, to which it was allied or linked by some kin ties, and it took part with the IJa�llii tribesmen in a great razzia against the land of Lage, but finished by being utterly defeated by Ninurta-kudurri-u�ur495 • The name of the tribe should probably be interpreted as a derivative * 'ammat of the noun 'amm, "ancestor", like Arabic 'amma(tun) that means "people, community". This word 'mt occurs once in a Palmyrene inscription where it must mean "community", "clan", or "family"496• Considering the participation of the 'Ammatu men in the raid against the land of Lage, reachable along the Siibu territory, their tribal land at that time should be located somewhere between the WadI Tat!iir and the Euphrates, east of the Siibu dependencies. This is confirmed by the fact that the sheikh of the 'Amrnatu could be named with his patronymic in the inscriptions of Ninurta-kudurri-u�ur, who obviously had good infor­ mations about this tribe. Its head was then called m1a-a-a-e or m1a-a- '-e, son of mBa-li-am-mu or mBa-la-am-mu491• As already mentioned above498, the name Y'y is attested in Minaic and must derive from the same root as Hebrew y'h and Arabic wa'a, "to remember". As for the patronymic, paralleled by the name 'Amml-ba'l of the Aramaean ruler of the Bet-Zammani in the 9th century B.C.499, it can be interpreted confi­ dently as Ba 'll- 'ammu, "the Ancestor is my lord". In consequence of the defeat inflicted upon the 'Ammatu tribesmen by Ninurta-kudurri-u�ur or for some other reasons, the tribe migrated to the southeast and crossed the Tigris. It is uncertain where Tiglath-pileser III encountered the 'Ammatu, because the land of Amat in Media, "opposite the Rua mountains"500, cannot be linked easily with the tribe. Under the reign of Sargon II, the Gambiilaean town of u"'A-ma-te, possi­ bly named after the Aramaean tribe, was on the bank of the Uqnii river, and it was attached with other subjugated cities of the region to the province of the governor of the Gambiilu land501 . The same city, called u"'Am-mat, may be mentioned also in a document found at Nippur and 494 RIMB II, text S.0. 1 002.2, p. 295, line 1 7 ; text S.0. 1 002.3, p. 301 , line 8'. 495 See here above, p. 106, 278, 427-428. 496 E. LIPINSKI, Unpublished Funerary Relief Busts from Palmyra, in OLP 29 ( 1 998 (20001), p. 63-73 (see p. 68). 497 RIMB II, text S.0. 1 002.2, p. 295, line 1 7 ; text S.0. 1 002.3, p. 301 , line 8 ' . 49H See here above, p. 428. 499 Cf. here above, p. 1 53. 500 Tigl. Ill, p. 72, line 6; p. 1 06, line 33'. 50I Sargon II, p. 1 49- 1 50 and 330, Ann. 292-294.

470

ARAMAEANS IN BABYLONIA

belonging to the archive of Nergal-iddina502 . Besides, a man named in a Neo-Babylonian text from Ur is designated as 1"A-mat-if-a503 , but the tablet is not dated and it is hardly datable by conjecture. The 'Ammatu are followed in Tiglath-pileser III's great list by the 1 "lja-ga-ra-a-nu 504 , whose name obviously derives from the Arabic base �agar, "stone", by addition of the suffix -an which mainly forms adjec­ tives and personal names. Nothing is known about this tribe, although it is named also in Sennacherib' s inscriptions505 , but S. Schiffer506 has related its name to Agranis, a city on the Euphrates mentioned by Pliny507 , and to Hagronia, often referred to in Talmudic writings508 . Despite the Talmudic spelling with h, this identification is worth consid­ eration for Arabic � corresponds to East Aramaic h509, and this was the case at least as early as the 2nd-4th centuries A.O., since West-Aramaic � in proper names may then appear as h in the Babylonian Talmud5 1 0. The Jewish Aramaic spelling Hgrwny' obviously reflects the change a > o, which indicates that the name was *Hagran, corresponding to Latin Agranis, with the loss of the second short vowel a of *lfaggaran5 1 1 • The town was on the Euphrates, southwest of Baghdad, near Nehardea. 23. NAQRI AND TAN'iYE

A fragment of Tiglath-pileser III's annals reporting the events of 745 B.C. and describing the territories annexed to Assyria at the end of the Babylonian campaign inserts "[the lands? of] the sheikhs of the Naqri and the Taniye (tribes)" between Diir-Kurigalzu and Sippar of Samas, on 502 RGTC 8, p. 22-23. 503 H.H. FIGULLA, VET IV, London 1 949, No. 1 40, 3; cf. RGTC 8, p. 22. 504 Tigl. Ill, p. 1 60, line 8. 505 Sennacherib, p. 25, line 48; p. 49, line 14; p. 54, line 56; p. 57, line 1 5 ; 506 S. SCHIFFER, op. cit. (n. 254), p. 1 26. 507 PUNY THE ELDER, Natural History VI, 1 20. soK A. NEUBAUER, La geographie du Talmud, Paris 1 868, p. 347-348; A. BERLINER, Beitraege zur Geographie und Ethnographie Babyloniens im Talmud und Midrasch, Berlin 1 883, p. 3 1 -32; J. OBERMEYER, Die Landschaft Babyloniens im Zeiralter des Ta/­ muds und des Gaonats, Frankfurt a/M 1 929, p. 265-270; A. OPPENHEIMER, Babylonia Judaica in the Talmudic Period, Wiesbaden 1 983, p. 1 34- 1 40. 509 For example, Arabic f:,akim is borrowed as hakim, "wise"; cf. R. MACUCH - E. PANOUSSI, Neusyrische Chrestomathie, Wiesbaden 1 974, p. 47. 510 E.Y. KUTSCHER, Studies in Galilean Aramaic, Ramat-Gan 1 976, p. 70. 5 1 1 The loss of the long a is not probable and, therefore, the identification of ljag­ gariinu with lja-gar-nu (or Ija-.M -nu), in Assyria (PARP0LA, Toponyms, p. 1 4 1 ), is unlikely. The geminated gg underlies the dissimilated form ljangaranu of a proper name (PNA 11/ 1 , p. 453); cf. Arabic f:,aggar, "stone mason", "stone cutter ".

NAQRI AND TANIYE

47 1

the one hand, and the cities of Kala'in on the Shumandar canal and of Pa�'itu of the Dunanu tribe, on the other: [ . . . 1"]na-sik-ki '"Na-aq-ri 1"Ta­ ni-e5 1 2. This seems to point to the area of Babylon and of Sippar5 1 3 , north or northwest of Nippur5 1 4. In fact, these tribes, which are not labelled as Aramaean, appear also together as '"Na-qa-ri it '"Ta-ni-e in a letter sent by Bagyan, apparently the sheikh of the Dunanu tribe5 1 5 , to the governor of Nippur5 1 6 in the mid-8th century B.C. Bagyan reports that both tribes have joint forces with his opponent Niiru and the tfalapi tribe. Slightly earlier perhaps, the Assyrians based at Calah had cap­ tured a sheep-stealing sheikh of the Naqiru and a band of brigands from the RuqaiJu tribe5 1 7 , which was located near the confluence of the Lesser Zab and of the Tigris5 1 8. All these elements point to a northern location of Naqri and Taniye, which apparently ranged from the Sippar area as far northeast as the Diyala and the Lesser Zab. Therefore it is unlikely that there is any relation between the Taniye tribe and ur"Dur-mTa-ni-e conquered by Sennacherib on the territory of the Chal­ daean tribe of Bit-Sa'a11i5 1 9. The Naqri and Taniye tribes are not labelled specifically as Aramaean, but the Naqiraean sheikh mia-da- '-Ilu '"nasiku sa '"Na-qi-ra-a-a520 bears a perfectly Aramaic name, which is not common in North-Arabian52 1 . The reading of his patronymic is uncertain522 , but the name of the tribe is related to the Aramaic noun naqr, niJqar, "pure"523 . Its singular seems to appear in Tiglath-pileser III's na-aq-ri and Bagyan' s na-qa-ri, while the oldest attestation from Calah has the "broken" plural formation naqiru, "the pure ones", paralleled in Classical Arabic by 'abidu(n), "servants", the collective plural of 'abdu(n). As for the '"Ta-ni-e, they are literally Tig/. Ill, p. 42, line 6; cf. p. I SO, lines 8-9. 5 1 1 RGTC 8, p. 382-383; cf. B. GRONEBERG, op. cit. (n. 269), p. 3 1 0, and K. NASHEF, 511

op. cit. (n. 69), p. 3 1 6. 5 1 4 See here above, p. 458-459. 5 1 5 See here above, p. 458. 1 1 h Nippur IV, No. 1 3, 1 9-20. 5 1 7 CTN II, 1 1 9. 5 1 8 See here above, p. 48-49, and E. FORRER, Die Provinzeinteilung des assyrischen Reiche.1·, Leipzig 1 920, p. 1 2 and 47. 5 1 9 Sennacherib, p. 53, line 40. 520 CTN II, 1 1 9, 9- 1 0, cf. line 2 1 . 521 Instead, the name i s frequent in Sabaic and in Hadramic; cf. HARDING, Arabian Names, p. 664. 522 CTN II, 1 1 9, 22: 111 0-ka-tur·'-me. 521 The Arabic verb naqara means "to pierce", hence "to offend", and this metaphor­ ical connotation is attested also in Epigraphic South Arabian. These meanings are not fit for a use of the root in anthroponomy.

472

ARAMAEANS IN BABYLONIA

"sworn men", as it seems. The Westsemitic verb tny appears to be a vari­ ant of Assyro-Babylonian tamu, "to swear"524• In Hebrew, this verb is used only in the ancient texts of Judg. 5, l l and 1 1 , 40, where it means something like "recall", "evoke", hence also "mourn" (Judg. 1 1 , 40). A legal connotation underlies the Aramaic use of the phrase tny 'mh in a fragment of land registry from the Persian period where the entry of one section is partly preserved525 : tny 'mh there seems to designate the offi­ cial category of "swearers of an oath". The formula corresponds to Assyro-Babylonian mamita tamu, "to swear an oath", with mamitu belonging to the same root as Hebrew and Aramaic '(y)mh, "awe", "oath". A tribe with a name like 1 "Ta-ni-e must be based on a fraternity sealed by a solemn oath. It is not mentioned any more after Tiglath­ pileser III.

24. GAMBOLU

The Aramaean tribe of 1"Gam-bu-lu is first mentioned in Sargon II' s inscription reporting the Assyrian campaign in southern Babylonia, about 7 1 0 B.C. It frequently occurs later in texts of Sennacherib, Esarhaddon, Ashurbanipal, and even Nebuchadnezzar 11526• Its name derives by dissimilation from the word gabbul, which means "modeller, kneader", hence "creator" in Syriac. Its concrete ancient meaning may have been "potter". Gambiilu was the principal Aramaean tribe of south­ ern and southeastern Babylonia, next to the Puqiidu, and its tribal terri­ tory, situated north of Bit-Yilin, near the Elamite border, was quite large, since Sargon II turned the conquered area into an Assyrian province with six districts: tfubaqanu, Tarbugati, Timassunu or Tibar­ sunu, Pasur, Ijintu, and Ijilmu527 • Diir-Abi-l)ara528, a stronghold on the 524 For the mln alternative, see LIPINSKI, Semitic, § 1 1 .7. 525 TAD III, C3.2 1 , 7, where the phrase is left without translation. 526 Sargon II, p. 433-435. References in PARPOLA, Toponyms, p. 128- 129, and RGTC 8, p. 1 37. 521 Sargon II, p. 143 and 328, Ann. 280-28 1 . The name of the first district, character­ ized by the affonnative -iin, seems to be based on the proper name of the eponymous ancestor of a clan, possibly North-Arabian 'bqn (HARDING, Arabian Names, p. 402). One wonders whether the second name is not a misspeling of UflJTargibiitu (PARPOLA, Toponyms, p. 348; RGTC 8, p. 306-307). ljiritu means "ditch", "canal", or "moat" (G. FRAME, op. cit. [n. 70), p. 290-291), and ljilmu may have various Aramaic and Arabic etymologies. Timassunu or Tibarsunu and Pa.fur do not seem to have a Semitic etymon. 528 The proper name mAD-IJa-ra should be read either Abi-1.Jara (cf. mA-bi-1.Ja-ri in ND. 2357, 7, r. 7, and 'b!Jr) or Ad(da)-1.Jara (cf. mu-1.Ja-ri, mlja-ra-dlM, etc.), in Aramaic Hd'r,

GAMBULU

473

Surappu river, probably the Riidgane-ye Cangiile, east of the Kut al­ 'Amara area, became the chief-town of the province and was renamed Diir-Nabii529. The main indigenous fortified centre was Sa-pl-dBel, located "between rivers"530 and considered by Esarhaddon as "the gate to Elam"53 1 • Sargon II deported 1 8,400 persons from Diir-Abi-gara and took a large booty there532 • The Gambiilaean sheikhs of the neighbour­ ing area at the Uqnii river submitted then to Sargon II who imposed a heavy tribute on them and annexed their land to the newly formed Assyrian provincesJJ. Five of the six preserved names of the submissive Gambiilaean sheikhs are Aramaic: mlja-za-llu, "God has seen", mlja-am-da-nu, "the desired one", m-za-bi-du, "given", mAm-ma-i1-[ . . . ], "the Ancestor . . . ", mA-a-sa-Am-mu, "the Ancestor has given"534• A few generations later, however, msa-am- '-Gu-nu535 , son of the Gambiilaean sheikh Bel-iqisa, bears a name with the Arabic theophorous element Ginn536, thus "the Genius has heard". "Haddu protected"; cf. D. HoMES-FREDERICQ - P. GARELLI - E. LIPINSKI, Archives d'un centre provincial de /'Empire assyrien, Bruxelles, forthcoming, No. 48, 6 with comments.

The quite common transcription Diir-AtlJara is thus erroneous, but the correct reading Abi-�ara was already proposed by M. STRECK, Keilinschriftliche Beitriige zur Geographie Vorderasiens l, in MVAG 1 1 ( 1 906), p. 203-246, in particular p. 220-22 1 . 529 Sargon II, p. 1 43 and 328, Ann. 279-280. For the identification of the Surappu river, see ibid., p. 423 and 459. 530 R. BORGER, op. cit. (n. 29), p. 1 05 and 226-227, col. VI, 23-24 = VII, 1 8- 1 9. 5 3 1 R. BORGER, Die lnschriften Asarhaddons, Konigs von Assyrien (AfO, Beih. 9), Graz 1 956, p. 53, line 83. 532 Sargon II, p. 1 40 and 327, Ann. 27 1 -272. 533 Sargon II, p. 1 4 1 - 1 43 and 328, Ann. 272a-e and 275-28 1 . 534 Sargon II, p. 1 4 1 , Ann. 272a-b. However, mA-a-sa-Am-mu appears as uruA-a-sa-mu in Sargon II, p. 1 48 and 329, Ann. 2791, and the name 'ws 1 'm, "the Ancestor has given", occurs so far only in Qatabanic: HARDING, Arabian Names, p. 85. The interpretation "Where is Sammu?" in PNA U I , p. 92, is unlikely, since Sammu is so far unknown. 535 APN, p. 1 9 1 a. There is a variant spelling msa-am-Gu-nu without '. The verbal predicate can be vocalized Iama' or Jami'. 536 The spelling -Gu-nu in the annals of Ashurbanipal does not reflect the Assyrian vowel harmony, that affects a short, unstressed a in an open syllable. Instead, it should be explained as a case of dialectal Arabic u against i, like in The noun qiryii derives from qir, "wall". H, KAI 202A = TSSI II, 5A, 10. Hb

H7

/bill.

R. NAUMANN, Architektur Kleinasiens von ihren Anfiingen bis zum Ende der hethi­ tischen Zeit, 2nd ed., Ttibingen 1 97 1 , Fig. 356. Cf. H. GASCHE, Mauer (mur). B. Archiio­ logisch, in RLA VII, Berlin 1 987-90, p. 591 -595, especially p. 593b. HH M.V. SETON WILLIAMS, art. cit. (n. 7 1 ), p. 80; ID., The Excavation.1· at Tell Rifa 'at, / 964: Second Preliminary Report, in Iraq 29 ( 1 967), p. 1 6-33 and Pl . V-X, especially p. 1 8-20.

530

SOCIETY AND ECONOMY

early 9th century B.C. on89 . Walls in Aramaean cities were generally built of bricks on low stone foundations. Besides projecting towers, also a citadel could reinforce the defensive system, as in Stratum VIII at Hazor90 and in the southern and highest section of the mound at I:Iama9 1 . Another important structure of the rampart was the city gate92, called fara ' in Aramaic (fr' > tr '). For instance, the ancient enclosure wall with three of its gateways was partially preserved at Tell ar-Rif'at (Arpad) and Arslan Ta� (l:ladattu). The city gates ought to be distinguished from the inner gateway leading to the palatial quarter of the city. The palace or the administrative centre usually occupied the highest place of the mound. It often contained a porticoed building with a central broad room, which is referred to as bet biliini. But there were other pub­ lic structures, such as temples, pillared buildings interpreted as store­ houses93, and water supply systems94. Public buildings were often deco­ rated with gateway lions and bulls, and with portal stone reliefs or simple orthostats95. Private dwellings are known in an insufficient way, because excavations are generally focused on temples and palaces. However, the Aramaean settlement of the late 9th century B.C. at Deir 'Alla (Phase IX) gives quite a good idea of a country town with its houses of mud bricks and flimsily constructed rooms96• On the other hand, a surprisingly well­ planned orthogonal city of the final years of Aram-Damascus appears in Stratum V of Tell as-Sa'1d1yeh97 . The city of this stratum was destroyed K9 Cf. here above, p. 1 1 6. 9 A. MAZAR, Archaeology of the Land of the Bible: 10,000-586 B. C.E., New York 1 990, p. 4 1 2-4 1 3 ; z. HERZOG, op. cit. (n. 73), p. 224-225. 91 E.F. FUGMANN, Hama I U I . L 'architecture des periodes pre-hellenistiques, Copen­ hague 1 958, Fig. 1 85. 92 For a general discussion of city gates, see Z. HERZOG, Das Stadt/or in Israel und in den Nachbarliindern, Mainz 1 986. 91 Z. HERZOG, The Storehouses, in Y. AHARONI (ed.), Beer-Sheba l, Tel Aviv 1 973, p. 23-30. 94 The water system at Hazor is attributed to Stratum VIII: Y. YADIN, Hazor, London 1 972, p. 1 72- 1 78; A. MAZAR, op. cir. (n. 90), p. 48 1 -482. The unique stepped and roofed passage to an underground water source at Tell as-Sa 'idiyeh was dated to the 1 2th cen­ tury, but it may be later: A. MAZAR, op. cir. , p. 40 1 , n. 2 1 . and p. 527. n. 1 3. 95 Zincirli: F. VON LUSCHAN (ed.), Ausgrabungen in Sendschirli II, Berlin 1 898, Fig. 47, 88, 90; vol. III, Berlin I 902, p. 209. Tell l:lalaf: R. NAUMANN, Tell Halaf II. Die Bauwerke, Berlin 1 950, Fig. 1 7. Arpad: M.V. SETON WILLIAMS, art. cir. (n. 7 1 ), p. 80-8 1 . Arslan Ta§: P. ALBENDA, The Gateway and Portal Stone Reliefs from Arslan Tash, in BASOR 27 1 ( 1 988), p. 5-30. 96 G. VAN DER Koou - M.M. IBRAHIM (eds.), Picking up the Threads . . . , Leiden 1 989, p. 82-89. The Aramaic Plaster Inscription belongs to this Phase IX (former M). 97 J.B. PRITCHARD, Tell es-Sa 'idiyeh: Excavations on the Tell, 1964- 1966, Philadelphia 1 985, p. 4-59, 75-80. See also Z. HERZOG, op. cir. (n. 73). p. 232-234.

PRODUCTION OF LUXURY GOODS

53 1

by a severe conflagration, probably during Tiglath-pileser III's cam­ paigns against Damascus in 733 and 732 B.C. The domestic quarter was made up of uniform three-room houses with an average area of 40-42 m2, arranged in rectangular blocks between parallel streets, about 1.5 m wide. Each house consisted of a broad-room in the rear and a front space divided unequally into two rooms by a row of three or four brick pillars. The entrance to the house led into the wider front room, which was prob­ ably roofed, while the narrower room had a stone-paved floor. A build­ ing used both as residence and as workshop for industrial activities, apparently for weaving and dyeing, was uncovered at Tell AJ:imar98 , and installations with an exceptional number of looms, indicating weaving activities on a trade scale, were found at Deir 'Alla in the middle of rooms used for domestic purposes99 . Although the Aramaean states were dotted with many localities called cities, these settlements did not resemble the mediaeval towns in being primarily centres of handicrafts and trade. Their population, as a rule, consisted principally of farmers who had banded together to live behind city walls for protection against raiders. Their livelihood was derived from cultivating their fields, vineyards, and orchards, for the most part located outside the city walls, to which they proceeded in the morning and from which they returned in the evening. Beyond the fields and vineyards, there also were some pastures where the farmers could main­ tain cattle, sheep, and goats, particulary for the purpose of producing milk and wool. Industrial production was usually in the hands of the royal administration or of artisans who were probably organized in clans or guilds or both. Also the Chaldaeans of southern Babylonia constituted a partly urban society. Thus, in the first part of his reign, Sennacherib claims to have conquered 88 Chaldaean walled cities and 820 villages 1 00• 4. PRODUCTION OF LUXURY GOODS

The Aramaean capital cities in northern Mesopotamia and Syria exhibited many characteristics of major Near Eastern urban centres where upper classes indulged in considerable luxuries in dwellings and personal attire. It is only by understanding the rise of these classes with its consequent increase in wealth and social differentiation leading to a 9H

G. BUNNENS, art. cit. (n. 69), p. 20-22 with a plan on p. 1 9. G. VAN DER Koou - M.M. IBRAHIM (eds.), op. cit. (n. 96), especially p. 88. 1 6 attested also in North Syria (R. ZADOK, WO 9 [ 1 977], p. 40-4 1 ). The mean­ ing of the theonym was recognized by B. LANDSBERGER, Sam 'al. Studien zur Entdeckung der Ruinenstiitte Karatepe, Ankara 1 948, p. 45-46. The hieroglyphic Luwian spelling Ba+ra/i-ki-ba-sa (E. LAROCHE, Les hieroglyphes hittites I, Paris 1960, No. 446) of the name of king Bar-Rakkab reflects the metathesis ra-ka-bi-(genitive) > ra-ki-ba-, rather than a change a > i (ZADOK, WSB, p. 253; C. RABIN, op. cit. [n. 1 1 1 ] , p. 90, 1 0 1 , also p. 1 05: ii > e). 1 26 Cf. R. DE VAUX. Les institutions de l 'Ancien Testament I, Paris 1 958, p. 1 88; vol. II, Paris 1 960, p. 23; L.B. COUROYER, L 'Exode et la bataille de Qadesh, in RB 97 ( 1 990), p. 321 -358 (see p. 340-345). See also M.A. L11TAUER - J.H. CROUWEL, Wheeled Vehicles and Ridden Animals in the Ancient Near East, Leiden 1 979, especially p. 99- 1 43, as well as G. DEL OLMO LETE, Notes on Ugaritic Semantics I, in VF 7 ( 1 975), p. 89- 1 02 (see p. 96- 1 02); B.A. MASTIN, Was the salts the Third Man in the Chariot?, in J.A. EMERTON (ed.), Studies in the Historical Books of the Old Testament (VTS 30), Leiden 1 979, p. 1 25- 1 54; W. FARBER - M.A. LITTAUER - J.H. CROUWEL, Kampfwagen, in RLA V, Berlin 1 976-80, p. 336-35 1 , in particular p. 34 1 -342, § 7b, and p. 348-350, § 6. 127 I.J. WINTER, Carved Ivory Furniture Panels from Nimrud: A Coherent Subgroup of the North Syrian Style, in Metropolitan Museum Journal 1 1 ( 1 976), p. 25-54 (see p. 52 with n. 1 08); St. DALLEY, Foreign Chariotry and Cavalry in the Armies of Tiglath-Pi/eser Ill and Sargon II, in Iraq 47 ( 1 985), p. 3 1 -48 (see p. 39-40). 128 E. LIPINSKI, Le culte du soleil chez les Semites occidentaux du /"' mil/enaire av. J.C., in OLP 22 ( 1 99 1 ), p. 57-72, especially p. 64-66. 1 29 ANEP, No. 1 65. Cf. BUNNENS, Syria, p. 249, Pl. 9: Arslan Ta§, 8th century. 125

616

RELIGION

personifies the heavenly vehicle. In any case, it is important to stress that the holy patron of the Sam'alian royal dynasty is not 'II, but Rakkab-'II, the driver 1 30 • In the 8th century, he is mentioned without the qualifica­ tion "'II" in the name of king Bar-Rakkab and he is called "Baal Rakkab of Sam'al" in the 7th century 1 3 1 . The predominant position of "the Char­ ioteer of 'II" is probably due to the fact that he was viewed as the young and active god, cracking his whip as if he was thundering and thus sur­ named Baal, like the weather-god. However, he is clearly distinguished from Hadad in the S am' alian inscriptions. Another divine name composed with the element ' II is ' II-Wer 1 32, the god of Zakkiir, king of I:Iamath and Lugath. There is no doubt that this is the Mesopotamian god Ilu-Mer, which the An = Anum list identifies with the Storm-god 1 33 . This equation probably reflects the syncretistic tendency of the later period, but the etymology of Me-er or Me-ir indi­ cates that Ilu-Me-ir is "the god of the cereal harvest" and that ltur-Me­ ir, the god of Mari, bears a name which expresses the idea that "the cereal harvest came back". In fact, the South-Arabian noun myr means "cereal harvest" 1 34 and Arabic mira still designates the "supplies". Although the change m > w does not create special problems in Semitic 1 35 , the Middle and Late Assyrian spelling dBe-er of this theonym seems at first sight not to obey normal phonetic rules of Semitic 1 36. However, there are occurrences in which b alternates with m, especially in northern Syria and in Ethiopia 1 37 . Thus, there seems to be no valid reason, at least, why me-ir/myr should be considered as a word borrowed from a substrate non-Semitic language. 1 30 KAI 24 = TSSI III, 1 3, 16; KAI 2 1 5 = TSSI II, 1 4, 22; cf. KAI 25 = TSSI III, 14, 4-6; KAI 2 1 6 = TSSI II, 1 5, 5. 1.1 1 See here above, p. 6 I 5, n. 1 25. 1 32 KAI 202 = TSSI 11, 5, A, I; B, [ 1 4] .20.23-[24]; cf. here above, p. 254-255, 300. The name most likely appears also as [' ]1wr on an ostracon found in 1997 at Tell Afis: S. MAZZONI, Une nouvelle stele d'epoque arameenne de Tell Afis (Syrie), in Trans­ euphratene 1 6 ( 1 998), p. 9- 1 9 and Pl. 1-11 (see p. 9 and 1 8). 133 E . (P.) DHORME, Melanges assyriologiques, in RA 8 ( 1 9 1 1 ), p. 97- 1 05 (see p. 97-98); G. Doss1N, Inscriptions de fondation provenant de Mari, in Syria 2 1 ( 1 940), p. 1 52- 1 69 (see p. 1 54-1 58); E. DHORME, Les religions de Babylonie et d'Assyrie, in "Mana" II, Paris 1 945, p. 1 26; ID., Recueil Edouard Dhorme, Paris 195 1 , p. 750 and 759; W.G. LAMBERT, The Pantheon of Mari, in MARI 4 ( 1 985), p. 525-539 (see p. 534-535 and 538-539). Cf. L.W. KING, CT XXV, London 1909, Pl. 17, 30; 20, 7; 0. SCHROEDER, Keilschrifttexte aus Assur verschiedenen lnhalts (WVDOG 35), Leipzig 1 920, No. 5 1 ,

r . 14; etc.

134 Sabaic Dictionary, p. 89. 1 35 LIPINSKI, Semitic, § I 1 .8; cf. § 32. 17. 1 3 6 W.G. LAMBERT, art. cit. (n. 1 33), p. 534. 1.11 LIPINSKI, Semitic, § 1 1 .6; cf. § 48.5.

RESHEPH AND RU$A

617

If ' 11-Wer was a corn god, like Dagan, it is quite natural that he should be identified with a god of thunderstorm, dispensing natural irrigation in the form of rain, although cereal crops of the Middle Euphrates valley depended mainly from artificial irrigation. The particular case of the dedication of the Zakkiir stela to ' 11-Wer must be explained by the king's origin from the land of 'Ana. The element 'll occurs further in the theonym 'lqn 'r!f of the Phoeni­ cian inscriptions from Karatepe, and it reappears under the Aramaic form 'lq(w)nr ' in Palmyrene inscriptions, where he is equated with Ilocr£18&v 138 . The Palmyrene spelling 'lqwnr ' i 39 with the vowel letter w suggests that this is no actual Aramaean deity 140, notwithstanding the substitution of r ' for 'r!f. The introduction of this cult in Palmyra and in Baalbek, where Latin and Greek transcriptions of the name have been found 1 4 1 , is likely to be attributed to Phoenicians 142• 5. RESHEPH AND RU$A

The Sam'alian inscription carved on the colossal statue of Hadad on behalf of Panamuwa I mentions the god Resheph, to whom the king seems to have ought a special gratitude 143 . The association of this deity to the group of 'II, Rakkab-'11, and Samas is thus comprehensible, the more so as Resheph was worshipped in northern Syria since the third millennium B.C. and played there an important role in the cult 1 44 • In line 11 of the inscription, however, 'rqrsp is mentioned instead of Rsp. Authors dealing with this text either do not propose any explanation of the compound divine name 145 or propose unsatisfactory interpretations, 138 See E. LIPINSKI, op. cit. (n. 33), p. 59-60 with further literature. u9 D.R. HILLERS - E. CUSSINI, op. cit. (n. 1 0 1 ), PAT 2780, 5. The other attestations have no vowel letter: ibid., PAT 22 1 9-2222, cf. p. 339a. 140 The change a > 6 is attested in Palmyra (see here above, p. 6 1 2, n. 1 0 1 ), as shown e.g. by the name of BM < Bal < Ba '/, but the form qwn in a theonym not attested in ear­ lier Aramaic is possibly influenced by Phoenician; cf. DNWSI, p. 1 0 1 5. 141 J.-P. REY-COQUAIS, IGLS V, Paris I 967, Nos. 2743 and 284 1 ; lo., Connaros le Puissant, in Syria 55 ( 1 978), p. 36 1 -370. 142 This is stressed already by J. HoFTUZER, Religio Aramaica, Leiden 1 968, p. 47. See also M. GAWLIKOWSKI, art. cit. (n. 1 02), p. 2647. 143 See here above, p. 6 1 4-6 1 5 with n. 1 23- 1 24. 14" E. LIPINSKI, op. cit. (n. 33), p. 1 79- 1 88, with further literature, complemented by I. CORNELIUS, The Iconography of the Canaanite Gods Reshef and Ba 'al (0B0 140), Fri­ bourg-Giittingen 1 994, but cf. E. LIPINSKI, Egypto-Canaanite Iconography of Reshef, Ba 'al, lforon, and Anal, in Chronique d 'Egypte 7 1 ( 1 996), p. 254-262. 145 Thus H. DONNER, in KAI II, p. 2 1 9: "Das Element 'rq hat sich bisher alien Deu­ tungsversuchen entzogen". The name is simply omitted in the poorly documented entry Resheph in DDD, Leiden 1 995, col. 1 324- 1 330.

6 18

RELIGION

as "favourite of Resheph", written with a prosthetic alef 46• The only convincing solution was suggested in 1 948 by B. Landsberger who related 'rq to the name of the Palmyrene god 'Ar�ii 147 , who is a North­ Arabian deity. Since Old Aramaic inscriptions, in general, and Sam' alian inscrip­ tions, in particular, use the grapheme q to indicate the emphatic fricative lateral j, corresponding to later Arabic cjiid 148 , it is quite obvious that 'rq is the North-Arabian god Ru9a, who was of paramount importance in the Arab pantheon 149 • The oldest attestation of this deity recognized so far is found in the annals of Esarhaddon that mention dRu-ul-da-a-a-u among the gods of the Arabs 1 50 . The prosthetic vowel appears in Herodotus' transcription of Ru9a's name as OPOTAAT, OPOTAA, or OPA TAA 1 5 1 , spellings that probably go back to OPOTAAI with the well-known AA/ AA 1 52 and /IT confusions. In the same period, the pros­ thetic vowel appears in the theophorous element of the name dAr-gu-u­ za-bad-du, found in the Murasii documents from the time of Darius 11 1 53 . The spelling dAr-gu-u of the divine name 'Arqii 1 54 indicates that the Old 146 Thus J .C.L. GIBSON, TSSI II, p. 7 1. It is not necessary to refer here to other, even less convincing explanations, and it goes without saying that the Sidonian place name •r� Rspm, "Land of Reshephs", designating a city quarter in Bodashtart's inscriptions, is no divine name and cannot have any relation to the Sam'alian god 'rqrsp. It occurs in the sequence ym, "sea", Imm, "heaven", ·r�. "land", mJ/, "royal residence (?)", Jd, "countryside". 147 B. LANDSBERGER, op. cit. (n. 125), p. 48, 49. See further E. LIPINSKI, The God 'Arqu-Rashap in the Samallian Hadad Inscription, in M. SOKOLOFF (ed.), Arameans, Ara­ maic and the Aramaic Literary Tradition, Ramat-Gan 1983, p. 15-21. 148 R. C. STEINER, The Case for Fricative Laterals in Proto-Semitic (AOS 59), New Haven 1977; LIPINSKI, Semitic, § 16.7-9. 149 J. HENNINGER, art. cit. (n. 58), p. 135- 1 36, n. 24, has collected the literature con­ cerning Ru9a. The number of attestations of Ru9a has increased in these last years. See now M. GAWLJKOWSKJ, art. cit. (n. 102), p. 262 1-2623. 150 R. BORGER, op. cit. (n. 73), p. 53, line 1 1, and p. 100, Mnm. B, line 1 1 . Cf. J. STARCKY, Petra et la Nabalene, in DBS VII, Paris 1966, col. 886- 1017 (see col. 990-99 1); A. CAQUOT, rec. in Syria 47 ( 1970), p. 1 90b. 151 HERODOTUS, History III, 8. Cf. Ph.-E. LEGRAND, Herodote, Histoires, Livre Ill (Collection Bude), Paris 1939, p. 42. The identification of OPOTA A T with Ru9a was already made by M. LJDZBARSKJ, ESE III, Giessen 1909- 15, p. 90-93. However the expla­ nation of the spelling found in some of the Greek manuscripts is not convincing. 152 The lateral articulation of the later (jiid accounts for the appearance of tl, which ought to be compared with the use of di for Arabic (jiid in relatively recent Malay bor­ rowings. 153 Cf. M.D. COOGAN, West Semitic Personal Names in the Mura.fa Documents (Har­ vard Semitic Monographs 7), Missoula 1976, p. 1 4 and 48 ("the theophorous element is perhaps to be identified with the deity of the city of Arqa"!); ZADOK, WSB, p. 230. 154 It is tempting to recognize it also in the 'rqw of KAI 2 15 = TSSI 11, 14, 1 3, but this word is most likely a verbal form (cf. DNWSI, p. 1 1 1). As for the alleged personal name 'rg in TAD I, A2.3, 2, quoted in ZADOK, WSB, p. 230, it should be read, 'dy, i.e. 'Adday.

RESHEPH AND RU$A

6 19

Aramaic grapheme q transliterated a voiced sound when it was used for the phoneme corresponding to Arabic t/,ad 1 55 • This is confirmed by the later Greek transcriptions of personal names like Pay11aoao0(;, Paytt­ �TJAoi;, PaytAoi; 1 56, where y stands for the same Aramaic phoneme. The prosthetic vowel, introduced by an alef, appears also in Palmyra where the spelling 'r�w of the theonym 1 57 reflects the Palmyrene scribal practice of rendering the Old Arabian ancestor of the later {iis often refers to swiftly flying birds250 and it can petfectly sug­ gest the rapid flash of lightning. Philo of Byblos did not understand the meaning of the phrase; he is quoted to have written: tou,ecrnv dcp' U\j/OU