Language and Textual History of the Syriac Bible: Collected Studies 9781463234959

Citation preview

Language and Textual History of the Syriac Bible

Texts and Studies

9 Series Editor Hugh Houghton

Editorial Board Jeff W. Childers Christina M. Kreinecker Alison G. Salvesen Peter John Williams

Text and Studies is a series of monographs devoted to the study of Biblical and Patristic texts. Maintaining the highest scholarly standards, the series includes critical editions, studies of primary sources, and analyses of textual traditions.

Language and Textual History of the Syriac Bible

Collected Studies

Jan Joosten

9

34 2013

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2013 by Gorgias Press LLC

All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2013

‫ܝ‬

9

ISBN 978-1-61143-891-8

ISSN 1935-6927

Library of Congress Cataloging-in-Publication Data Joosten, Jan. Language and textual history of the Syriac Bible : collected studies / by Jan Joosten. p. cm. -- (Texts and studies ; 9) Includes bibliographical references. ISBN 978-1-61143-891-8 1. Bible. Syriac--Criticism, Textual. 2. Syriac language. 3. Aramaic language. I. Title. BS14.J667 2013 220.4’36--dc23 2013040317 Printed in the United States of America

TABLE OF CONTENTS Table of Contents ..................................................................................... v Preface ....................................................................................................... xi Old Testament Peshitta ........................................................................... 1 General ....................................................................................................... 3 La Peshitta de l’Ancien Testament dans la recherche récente........... 5 La tradition manuscrite de la PAT ................................................ 6 La PAT et les Targoums................................................................. 9 Milieu et date de la PAT ............................................................... 12 La PAT et la Septante ................................................................... 16 Perspectives de recherche ............................................................ 18 Language .................................................................................................. 21 The Use of some Particles in the Old Testament Peshitta .............. 23 ‫‘ ܐܝܟܘ‬Why’ ..................................................................................... 25 ‫ ܓܝܪ‬in Rhetorical Questions ....................................................... 27 ‫ ܕܝܢ‬in Wish and Apodosis ............................................................ 29 ‫ ܥܕ‬Used as a Conjunction (+ Imperfect) ................................... 32 Conclusion ...................................................................................... 33 Greek and Latin Words in the Peshitta Pentateuch First Soundings........................................................................................ 35 Problems of Identification ........................................................... 36 Greek Words Slipping into the Manuscript Tradition............. 40 Loanwords or Foreign Words? ................................................... 42 Some Statistics................................................................................ 43 Further Questions ......................................................................... 45 Appendix: Check-List of Greek Words in the Peshitta Pentateuch ............................................................................. 46 Materials for a Linguistic Approach to the Old Testament Peshitta ............................................................................................ 49 Introduction ................................................................................... 49 The OTP is Written in Idiomatic Syriac .................................... 52

v

vi

LANGUAGE AND TEXTUAL HISTORY

The OTP is Written in Early Syriac............................................ 58 Non-Syriac Elements in the OTP............................................... 64 Conclusion ...................................................................................... 67 The Peshitta and other Translations of the Hebrew Bible .............. 69 Doublet translations in Peshitta Proverbs .......................................... 71 The Peshitta Version of Proverbs............................................... 72 A Survey of the Seven Doublet Translations............................ 75 Conclusions .................................................................................... 82 Greek Words shared by the Peshitta and Targums to the Pentateuch ...................................................................................... 85 The Evidence ................................................................................. 86 Agreements Due to Independent Borrowing from Greek ..... 88 Greek Words Going Back to an Exegetical Tradition ............ 92 The Greek Word is not an Obvious Equivalent ................ 92 The Greek Loan is Poorly Attested ...................................... 94 The Testimony of Symmachus .............................................. 96 Conclusions .................................................................................... 98 The Use of the Peshitta in Textual Criticism ..................................... 99 1 Sam xvi 6, 7 in the Peshitta Version...............................................101 Introduction .................................................................................101 1 Sam. xvi 6‫ܐܟܘܬܗ ܕܡܪܝܐ ܡܫܝܚܗ‬.............................................102 1 Sam. XVI 7 ‫ ܐܠ ܗܘܝܬ ܓܝܪ ܐܝܟ ܕܚܙܐ ܐܢܫܐ‬..........................104 The Coherence of the Peshitta Version of 1 Sam. xvi 6–7...106 The text of 1 Sam. xvi 6–7 .........................................................107 The Hebrew and Syriac text of Deut 1:44 ........................................109 A Problem of Style ......................................................................110 The Peshitta and its Vorlage ......................................................113 Conclusion ....................................................................................114 The Old Testament Peshitta in the Syriac New Testament...........115 The Old Testament in the New; The Syriac Versions of the New Testament as a Witness to the Text of the Old Testament Peshitta ......................................................................117 Tatian’s Use of the OTP ............................................................117 Acts and Epistles .........................................................................120 The Syriac New Testament’s Witness to the OTP ................121 Case Study: Rom 10:19 — Deut 32:21 ....................................123 Conclusion ....................................................................................126

TABLE OF CONTENTS

vii

Syriac Gospels .......................................................................................127 West Aramaic Elements in the Syriac Gospel Version ..................129 West Aramaic Elements in the Old Syriac and Peshitta Gospels..........................................................................................131 The West Aramaic Elements .....................................................133 Criteria .....................................................................................133 The List ...................................................................................134 Christian Loanwords .............................................................143 Jewish Expressions ................................................................144 Conclusions ..................................................................................146 Evaluation .....................................................................................146 A Different Syriac Dialect or the Translational Process? ..........................................................................147 Intention of the Translator or External Influence? .........149 A West Aramaic Gospel Tradition? ...................................150 The West Aramaic tradition: Tatian’s Fifth Source? ..............153 West Aramaic Elements in Diatessaric Quotations .........154 The Testimony of the Peshitta ............................................155 The Fifth Source of the Diatessaron ..................................156 Tatian’s Motivation for Using a Fifth Source ...................157 Two West Aramaic Elements in the Old Syriac and Peshitta Gospels..........................................................................................159 ‫“ ܨܠܝܒܐ‬Cross”; ‫“ ܨܠܒ‬to Crucify” ..........................................160 ‫“ ܬܪܥܣܪܬܐ‬the Twelve” ..............................................................162 West Aramaic Elements in the Syriac Gospels: Methodological Considerations .............................................................................165 West Aramaic Loanwords in the Syriac Christian Vocabulary ...........................................................................166 The Beginnings of Eastern Christianity ...................................172 Perspectives ..................................................................................173 « Le Père envoie le Fils »: La provenance occidentale d’une locution syriaque ..........................................................................175 La fonction des verbes ‫ ܫܠܚ‬et ‫( ܫܕܪ‬pa‘el) en syriaque............177 Les verbes ‫ ܫܠܚ‬et ‫ ܫܕܪ‬dans les Évangiles (Peshitta et Vetus Syra) ......................................................................................178 Autres exemples du verbe ‫“ ܫܠܚ‬Envoyer quelqu’un” ...........181 Conclusions ..................................................................................184

viii

LANGUAGE AND TEXTUAL HISTORY

The Aramaic Tradition of the Gospels and the Syriac Gospels ...187 La tradition syriaque des Évangiles et la question du « substrat araméén » ....................................................................189 Le problème du substrat sémitique des Évangiles .................189 Éléments archaïques préservés par la tradition syriaque .......193 Les origines de l’Église syriaque ................................................200 Quelques exemples......................................................................203 Conclusions ..................................................................................208 The Text of Matt 13:21a and Parallels in the Syriac Tradition ......211 Introduction .................................................................................211 The Syriac Versions of Matt 13:21a and Parallels. .................213 The Significance of the Syriac Variant .....................................216 The History of the Text of Matt 13:21a and Parallels ...........218 The Text-Critical Value of the Syriac Versions of the Gospels ....................................................................219 The Syriac Versions of the Gospels and the Old Testament Peshitta ..........................................................................................221 The Old Testament Quotations in the Old Syriac and Peshitta Gospels: a Contribution to the Study of the Diatessaron .....223 Introduction .................................................................................223 The Traces of OTP Text ............................................................228 The List .........................................................................................229 Clear examples of OTP dependence ..................................229 Additional cases of OTP dependence ................................232 Cases where OTP text has been obscured by a revisor .............................................................................235 Possible cases of harmonization with parallel passages ..........................................................................236 Conclusions ..................................................................................238 The distribution of traces of OTP text over P, C and S ...............................................................................238 The marks of correction towards the Greek in the OT quotations ...............................................................239 The cases where OTP = P/C/S = Greek .........................240 The Provenance of OTP Text in the OT Quotations...........240 OTP Text in OT Quotations: a Mark of Antiquity ...............240 Old Syriac and Peshitta: Revisions of an Older Text ............241 Old Syriac................................................................................241 Peshitta ....................................................................................242

TABLE OF CONTENTS

ix

Harmonistic Traits in the Text of the OT Quotations..........243 Harmonistic readings ............................................................243 Parallel variants ......................................................................244 Conclusions ..................................................................................245 Tatian’s Use of the OTP ............................................................245 Tatian’s Diatessaron and The Old Testament Peshitta ..................247 Traces of the OT Peshitta in Eastern Witnesses of the Diatessaron ..........................................................................248 Ephrem’s Commentary on the Diatessaron......................249 The Old Syriac and Peshitta Gospels .................................252 The Arabic and Persian Harmonies ...................................255 Conclusions ............................................................................256 Traces of the OTP in Eastern and Western Witnesses .........258 A List of Relevant Readings ................................................262 Discussion...............................................................................268 Traces of the OTP Attested only in the West ........................270 Conclusions ............................................................................274 Miscellanea.............................................................................................277 Odes de Salomon 7,3a Obervations sur un hellénisme dans le texte syriaque ................................................................................279 Éléments d’araméen occidental dans la version syriaque de Ben Sira .................................................................................................283 Date et milieu de la traduction syriaque ...................................284 La version syriaque de Ben Sira et la Peshitta .........................286 Éléments occidentaux .................................................................289 Conclusion ....................................................................................296 Sources ...................................................................................................299 Index of Biblical References ...............................................................301 Index of Syriac Words .........................................................................311

PREFACE The Syriac Bible is a fascinating field to which too little serious research has been devoted. This neglect, although hard to excuse, can easily be understood. Indeed, apart from being particularly interesting, the field also poses special challenges. A first hurdle is the Syriac language itself. Although Syriac is simply a dialect of Aramaic, and one for which good dictionaries and grammars exist, most students come to it only after having studied at least two other ancient languages — by which time they are often eager to pass on from philology to more immediately rewarding pursuits. The general depreciation of the ancient versions of the Bible is another factor of discouragement; most biblical scholars would rather deal with the “original texts.” Finally, investigations in this field are usually compartmentalized, being limited either to the Old Testament Peshitta, or to Syriac versions of the New Testament. To be sure, each part of the Syriac Bible holds its own part of marvel, but a really adequate approach demands that they be studied together and in relation to one another. The studies collected in the present volume seek to further our understanding of the Syriac Bible text through linguistic and philological investigation. Taken together, the studies point to the following general outline of the development of the Bible in the specific late-Aramaic dialect that was used in Upper Mesopotamia and has come to be known as Syriac. 1. The oldest part of this Bible is the Old Testament Peshitta. Most of this corpus was translated before the middle of the second century CE, although a few books, such as Proverbs and Chronicles, would appear to have been produced slightly later. The OT Peshitta started out as a Jewish work analogous to the Septuagint. The source text was a Hebrew one close to the Massoretic text, although not identical with it. The translators used many midrashic traxi

xii

LANGUAGE AND TEXTUAL HISTORY ditions of the type surfacing also in the Rabbinic Targums. In the first stage of the work, reflected in the Pentateuch, the translators consulted the Septuagint occasionally if at all. But in later books, notably Proverbs, influence of the Greek version becomes more palpable. 2. During the time the Hebrew Bible was being translated into Syriac, the first gospel traditions reached Mesopotamia. These traditions were formulated in an Aramaic dialect comprehensible in Mesopotamia but reflecting in several details its origin in Syria-Palestine. This form of the gospel would appear to have been relatively independent from the Greek Gospels that would later become canonical. To those who first embraced the Christian message in Edessa and its environs, the “New Testament” was known exclusively in this form. The early Christian community, many of whom may have been Jews, also adopted (or kept) the OT Peshitta as their Old Testament text. 3. From the second half of the second century onward, influence of the Greek-speaking Church on local Mesopotamian Christianity grew, and a translation of the New Testament writings became a necessity. The first Syriac “translation” of any part of the emerging Greek New Testament was Tatian’s Diatessaron. Tatian, around AD 170, did several things at once. He recognized the authority of the four “canonical” Gospels (he is in fact one of the earliest witnesses to the emergence of the fourfold gospel) and decided to translate them into his native Syriac. However, rather than to juxtapose the four Gospels with their potential contradictions, he welded them into one single and comprehensive harmony. In addition, he took account of existing traditions among the Syriac-speaking Christian community. OT quotations in the gospel text were not translated from Greek but given according to the text of the OT Peshitta. The Diatessaron also linked up with peculiar vocabulary and turns of phrase characteristic of the gospel tradition that had been in use among Syriacspeaking Christians earlier. 4. The Diatessaron was a huge success in the East. Until the beginning of the fourth century, it appears to have been dominant throughout the Syriac-speaking Church. How-

FOREWORD

xiii

ever, the strong and growing influence of the Western Church made it unsustainable in the long run. New Syriac translations were made of the Gospels that respected the fourfold format that the West used exclusively. The Old Syriac Gospels (probably third century) and the Peshitta Gospels (probably fourth century) are successive attempts at providing the Syriac Church with a fourfold gospel faithfully reproducing the Greek model. Nevertheless, neither the Old Syriac nor the Peshitta Gospels render the Greek texts “from the ground up.” Both versions reuse translation equivalents, renderings, and even whole verses of the Diatessaron. The Old Syriac translation generally stands closer to the Tatianic base than the Peshitta, but occasionally the latter preserves harmonistic elements absent from the former. The relation between the Old Syriac and Peshitta Gospels thus seems to be relatively indirect: what they have in common comes from the Diatessaron. The Diatessaron, the Old Syriac Gospels, and the Peshitta Gospels competed over a prolonged period. Eventually, however, only the Peshitta version survived, while the other two textual forms fell into disuse. The Old Syriac version is preserved in two rather distinct manuscripts, both of them fragmentary. As for the Diatessaron, no copy of its original Syriac text has been recovered so far, although the writing is attested in quotations and later translations. The picture emerging from these studies is not complete, of course. But it shows that the history of the Syriac “Old Testament” and that of the Syriac “New Testament” are intimately interwoven. Although the Peshitta OT and the Peshitta NT differ in date, historical background, and translation technique, they do cohere with one another, the latter corpus linking up with the former in many ways. Apart from the potential illumination of the textual history of the Syriac Bible, the papers touch upon various neighboring fields. Several studies contribute in one way or another to a better knowledge of the Syriac language, some of them specifically to its diachronic development. Textual criticism of the Hebrew Bible is another discipline that receives much attention. In the sphere of the New Testament, a few studies contain proposals that may be of interest for research on the tradition history of the gospel texts. Two studies are wholly devoted to writings that fall without the

xiv

LANGUAGE AND TEXTUAL HISTORY

confines of the Syriac Bible narrowly defined, one very short one to the Odes of Solomon, and a longer one to the Syriac version of the book of Ben Sira. The plan to collect some of my studies on the Syriac Bible, Old Testament and New Testament, was first conceived more than ten years ago. George Kiraz immediately embraced the idea and offered to publish the collection with Gorgias Press. Assembling the papers, each with its own Syriac fonts, into a single volume proved to be a huge task, however, and for many years the project lingered. In 2010, funds were made available by the Institut Universitaire de France, giving the editorial project a boost. My doctoral student, Christophe Bonnard, has prepared the manuscript for publication and prepared the index. My first teacher of Syriac and close friend, Shraga Assif, has spent many hours correcting the proofs. My sincere thanks go to the named individuals and institutions for their help, and to the Faculty of Protestant Theology of the University of Strasbourg and the Équipe d’Accueil 4378 for the researchfriendly environment that made this publication possible. I also thank the publishers of the journals and collective volumes where the papers originally appeared for granting permission to print them again in the present collection. Jan Joosten

LA PESHITTA DE L’ANCIEN TESTAMENT DANS LA RECHERCHE RÉCENTE*

*

La Peshitta de l’Ancien Testament (PAT) revêt une importance particulière à plusieurs égards: cette ancienne version faite directement sur un texte hébreu est d’une grande valeur pour la critique textuelle; de plus grâce à sa technique de traduction, qui suit sa source de près tout en ne pas la rendant de façon servile, elle est un témoin précieux pour l’histoire de l’exégèse; finalement, elle est le plus ancien corpus littéraire en syriaque avec une influence inégalée sur toute la littérature subséquente en cette langue. Pour toutes ces raisons la publication, à partir des années soixante,1 d’une édition critique du texte de la Peshitta, est une entreprise hautement louable.2 Pour la première fois depuis la publication de la Peshitta dans la polyglotte de Paris en 1645,3 les chercheurs disposeront d’un texte fiable, avec un apparat critique rendant compte de l’état de la tradition manuscrite. * Exposé fait le 16 février 1996 à Paris dans le cadre du cycle de conférences “La bible grecque des Septante”. Je tiens à remercier Piet Dirksen qui a lu et commenté une première mouture du texte. 1 Après des pourparlers au congrès IOSOT de Copenhague en 1953, le projet d’une édition critique de la Peshitta, à Leiden et sous la direction de P. A. H. de Boer, fut lancé pour de bon au congrès de 1959. A partir de 1961, les publications de l’Institut de la Peshitta ont paru avec une grande régularité. 2 The Old Testament in Syriac According to the Peshitta Version (Leiden, 1966–1997). 3 Les quelques tentatives d’édition partielle de la PAT antérieures à l’édition de Leiden ont été discutées brièvement dans M. H. GoshenGottstein, “Prolegomena to a Critical Edition of the Peshitta”, Scripta Hierosolymitana 8 (1961), 26–67, spécialement 27–29.

5

6

LANGUAGE AND TEXTUAL HISTORY

L’effort collectif produit autour de l’Institut de la Peshitta à Leiden a donné une nouvelle impulsion à l’étude — textuelle, linguistique, exégétique et historique — de la PAT.4 Ainsi d’anciens problèmes ont pu être résolus, ou du moins traités sous un autre jour, et de nouveaux problèmes abordés pour la première fois. L’intention du présent article est de rendre compte des principaux résultats atteints et des pistes ouvertes dans les recherches récentes sur la PAT.

LA TRADITION MANUSCRITE DE LA PAT Le travail d’édition de texte s’est naturellement traduit en une compréhension plus grande de l’histoire du texte de la PAT. L’analyse et la classification de bien plus de cent manuscrits indique que l’on peut globalement distinguer trois périodes.5 Il y a tout d’abord les manuscrits datant d’après le IXème siècle environ et qui représentent, avec de nombreuses variations entre eux, le textus receptus, c’est-à-dire la forme du texte sous laquelle la PAT est d’abord venue à la connaissance des biblistes en Occident.6 C’est en effet le 4 Une bibliographie très complète rendant compte du nouvel essor des études de la Peshitta a été réalisée par P. B. Dirksen, An Annotated Bibliography of the Peshitta of the Old Testament (Monographs of the Peshitta Institute 5; Leiden, 1989); avec une mise à jour dans P. B. Dirksen, A. van der Kooij, éd., The Peshitta as a Translation (Monographs of the Peshitta Institute 8; Leiden, 1995), 221–236. 5 La distinction globale des trois périodes a d’abord été formulée par M. D. Koster dans sa thèse sur les manuscrits syriaques du livre de l’Exode, The Peshitta of Exodus. The Development of its Text in the Course of Fifteen Centuries (Studia Semitica Neerlandica 19; Assen/Amsterdam, 1977). Elle a été acceptée par d’autres autorités telles que A. P. Hayman (éditeur du livre des Nombres dans l’édition de Leiden), S. P. Brock (éditeur du livre d’Ésaïe) et M. P. Weitzman. Voir la synthèse de M. D. Koster, “Peshitta Revisited. A Reassessment of its Value as a Version”, JSS 28 (1993), 235–268. 6 Voir pour la provenance de ce type de texte, et pour la ligne de démarcation entre les manuscrits anciens et récents, P. B. Dirksen, “East and west, old and young, in the text tradition of the Old Testament Peshitta” (Peshitta Institute Communication 19), VT 35 (1985), 468–484. Dans une lettre datée du 22 janvier 1996, Dirksen a exprimé quelques réticences à l’égard du terme texte receptus: dans le contexte de la PAT ce

LA PESHITTA DE L’ANCIEN TESTAMENT

7

texte d’un manuscrit du XVIIème siècle qui fut imprimé dans la polyglotte de Paris,7 reproduit avec peu de corrections dans la polyglotte de Londres (1657) et repris sous une forme légèrement retouchée dans l’édition de Samuel Lee (1823).8 Ce type de texte a été reconnu comme tardif et sensiblement éloigné du texte original de 9 la version syriaque. En conséquence, les manuscrits tardifs, qui sont pour la plupart marqués de ce profil, ne sont notés dans l’édition de Leiden que lorsqu’ils appuient une variante qui se trouve également dans des manuscrits plus anciens.10 L’état antérieur du texte est celui que l’on trouve dans les manuscrits occidentaux du VIème au IXème siècle environ. Le célèbre Codex Ambrosianus (7a1 dans l’édition de Leiden), un manuscrit non daté mais qui semble remonter au VIIème siècle, est pris comme texte de base dans l’édition de Leiden. Dans la littérature scientifique anglophone cet état du texte a été baptisé “Basic Textus Re-

terme n’implique ni une sanction ecclésiastique, ni un texte homogène. Par conséquent, Dirksen préfère parler en termes purement descriptifs: les manuscrits postérieurs au Xème siècle, etc. 7 Il s’agit du manuscrit 17a5 selon la numérotation de Leiden; ce manuscrit se trouve à la Bibliothèque Nationale à Paris. 8 A part l’édition de S. Lee, les éditions faites à Urmia (1852) et à Mossoul (1887) ont souvent été employées. La base manuscrite de ces éditions est mal connue; voir P. Dirksen, “The Urmia Edition of the Peshitta: the Story behind the Text”, Textus 18 (1995), 157–167. Il est incompréhensible que W. Strothmann ait choisi le texte d’Urmia comme base de ses concordances de la Peshitta du Pentateuque et des Prophètes (voir n. 62). 9 L’abandon du textus receptus et la reconnaissance de la valeur des manuscrits anciens ne se sont réalisés qu’au prix d’une lutte prolongée, ce qui rappelle les vicissitudes du texte du NT. Pour la PAT, les étapes principales de cette lutte ont été retracées par Goshen-Gottstein, “Prolegomena”, 30–31. 10 Voir “General Preface” in The Old Testament in Syriac According to the Peshitta Version, Vol 1,1 Genesis-Exodus (Leiden, 1977), VI–VII.

8

LANGUAGE AND TEXTUAL HISTORY

ceptus” (BTR),11 désignation commode quoique peu heureuse comme définition.12 Enfin, il est possible de remonter plus loin encore dans la tradition textuelle de la PAT. Depuis longue date les chercheurs avaient identifié un petit nombre de manuscrits transmettant des leçons particulières, souvent plus proches du texte massorétique. On avait parfois exprimé le soupçon que ces manuscrits aient été corrigés selon un manuscrit hébreu ou grec, et seraient donc sans valeur pour l’établissement du texte de la PAT. Les recherches de M. Koster sur le MS BM Add 14425 (5b1 dans l’édition de Leiden) et de M. Weitzman sur le Codex Florentinus (9a1 dans l’édition) ont établi de façon convainquante que ce soupçon n’était pas fondé.13 Les manuscrits en question sont les témoins authentiques d’un état du texte plus ancien, plus proche, par rapport au texte et à la phraséologie, de la source hébraïque. Ce résultat n’est plus guère contesté aujourd’hui, même si certains on conseillé la prudence: en effet à côté des leçons originales les manuscrits concernés contiennent parfois des corruptions évidentes.14 Afin de déterminer le plus ancien texte qu’on puisse atteindre, chaque leçon doit donc être étudiée séparément.15

11 Ce terme créé par Koster a été adopté par plusieurs autres éditeurs de la PAT. 12 Voir M. H. Goshen-Gottstein, “The Peshitta and its manuscripts. A Review” (Compte rendu de la thèse de Koster), BiOr 37 (1980), 13–16, en particulier 14, n. 8. 13 Voir Koster, The Peshitta of Exodus, et M. P. Weitzman, “The Originality of Unique Readings in Peshitta Ms 9a1” in P. B. Dirksen, M. J. Mulder, éd., The Peshitta: Its Early Text and History (Monographs of the Peshitta Institute 4; Leiden, 1988), 225–258. 14 Voir, p. ex., A. van der Kooij, “On the Significance of MS 5b1 for Peshitta Genesis” in The Peshitta: Its Early Text and History, 183–199; R. B. ter Haar Romeny, “Techniques of Translation and Transmission in the Earliest Text Forms of the Syriac Version of Genesis” in The Peshitta as a Translation, 177–185. 15 On peut déplorer avec Koster que la politique d’édition de l’Institut de la Peshitta à Leiden n’ait pas permis d’imprimer dans le texte les lectures qui sont jugées originales si celles-ci ne figurent que dans un manuscrit ancien. Voir Koster, The Peshitta of Exodus, 254.

LA PESHITTA DE L’ANCIEN TESTAMENT

9

Les manuscrits conservant ces leçons authentiques sont en partie très anciens (tel le manuscrit 5b1 et le palimpseste 5ph1, tous deux du Vème siècle). Cependant, c’est le Codex Florentinus, un manuscrit du IXème siècle, qui transmet le plus grand nombre de leçons originales, ce qui démontre le fait bien connu qu’un manuscrit relativement jeune peut conserver un texte relativement ancien.16

LA PAT ET LES TARGOUMS La PAT, et plus spécifiquement le Pentateuque, se situe dans la même tradition exégétique que les Targoums juifs. Le fait est bien établi: en effet depuis le siècle dernier, divers chercheurs ont collectionné des leçons et des interprétations que la PAT partage avec un ou plusieurs des Targoums.17 La question de la relation exacte entre la PAT et les Targoums reste cependant débattue. Schématiquement, deux types d’hypothèse ont été avancés pour expliquer la proximité entre ces versions. Les uns voient dans la PAT la traduction indépendante d’un texte hébreu, qui s’est cependant appuyée sur des traditions exégétiques juives existantes, lesquelles ont été préservées également dans les littératures midrashique et targoumique.18 Dans cette hypothèse, la PAT n’a subi aucune influence directe d’un Targoum et les points de contact s’expliquent plutôt

16 Voir

Weitzman, “The Originality”. La première collection étendue est celle de J. Perles, Meletemata Peschitthoniana, (Breslau, 1859). Un compte rendu très détaillé de l’histoire des recherches jusqu’au début des années 1980 est donné par P. B. Dirksen, “Syrian Christianity and Judaism” in M. J. Mulder, éd., Mikra. Text, Translation, Reading and Interpretation of the Hebrew Bible in Ancient Judaism and Early Christianity (Assen/Maastricht & Philadelphie, 1988), 255– 297, en part. 264ss. Plus récents sont K. Luke, “Targumisms in the Peshitta Version of the Pentateuch”, Bible Bhashyam 16 (1990), 52–64, et l’article de Brock cité ci-dessous dans la n. 27. 18 C’était déjà l’avis de Perles (voir la note précédente). L’étude la plus complète des parallèles entre la PAT et la littérature exégétique juive arrive également à cette conclusion; voir Y. Maori, The Peshitta Version of the Pentateuch and Early Jewish Exegesis (Jérusalem, 1995), 285–316, en part. 297–299. 17

10

LANGUAGE AND TEXTUAL HISTORY

par une dépendance commune envers les mêmes sources. 19 D’autres ont jugé que les correspondances verbales entre la PAT et les Targoums sont si frappantes qu’elles doivent être interprétées comme le résultat d’une influence directe. Pour ceux-ci, la PAT est un “proto-Targoum” (Urtargum) amputé de la plupart de ces élaborations explicatives et dont la langue a été complètement syriacisée.20 Il existe une théorie intermédiaire entre ces deux positions selon laquelle la base de la PAT aurait bien été un texte hébreu, que les traducteurs auraient compris à l’aide d’un certain nombre d’écrits exégétiques dont un Targoum.21 Les années récentes ont vu la majorité des chercheurs se rallier à la première hypothèse. Contre la théorie de l’influence directe du Targoum sur la Peshitta on a fait valoir qu’en dehors des ressemblances entre ces versions il y a aussi un grand nombre de différences, dont certaines sont suffisamment frappantes pour mettre en doute que les auteurs de la Peshitta aient connu un des Targoums existants (sous quelque forme que ce soit).22 En outre, il s’avère que la Peshitta contient des interprétations de type midrashique attestées dans l’ancienne littérature exégétique juive, mais non dans les

Ces sources ont pu être des traditions essentiellement orales; l’hypothèse n’implique pas une datation haute pour les écrits midrashiques existants (comme la Mekilta, Siphra, Siphré, etc.). 20 Cette hypothèse avait été formulée pour la première fois par I. Prager en 1875. Elle a été reprise et élaborée simultanément par P. Kahle et par A. Baumstark. Voir Dirksen, “The Old Testament Peshitta”, 265– 276. 21 Cette approche fut inaugurée par J. M. Schönfelder en 1869 dans le but exprès de démontrer l’antiquité du Targoum Onkelos. 22 Une liste de différences est donnée par P. Wernberg-Møller, “Some Observations on the Relationship of the Peshitta Version of Genesis to the Palestinian Targum Fragments Published by Professor Kahle, and to Targum Onkelos”, StTh 15 (1961), 128–180, en part. 152–163. Il faut cependant signaler que Wernberg-Møller reste attaché à l’hypothèse de l’influence du Targoum Onkelos sur la Peshitta. Pour les différences entre les deux versions, voir aussi W. F. Smelik, “The Targum of Judges”, OTS 36 (Leiden, 1995), 282–290. 19

LA PESHITTA DE L’ANCIEN TESTAMENT

11

Targoums.23 L’explication par l’influence directe du Targoum n’est donc ni nécessaire, ni suffisante. Un des meilleurs arguments en faveur de l’influence d’un Targoum avait été trouvé chez Aphrahat et Ephrem: en effet, certaines de leurs citations bibliques contiennent des éléments de type midrashique absents des manuscrits de la PAT. Ceci pourrait prouver que la version était à l’origine plus proche encore de la tradition targoumique.24 Il faut cependant abandonner ce type de raisonnement, pour deux raisons. D’une part, l’étude de la tradition manuscrite indique que l’état le plus ancien du texte se caractérise par une grande fidélité à l’hébreu, plutôt que par un “profil targoumique”.25 D’autre part, l’étude du texte biblique chez les Pères syriaques a démontré que ceux-ci ne citent pas toujours leurs sources de façon textuelle.26 Très souvent, les citations semblent être faites de mémoire, et suivant le cas une citation peut être erronée, adaptée au contexte littéraire, ou influencée par un texte parallèle. Dans certains cas il faut en outre tenir compte du fait qu’une tradition juive ait pu influencer un auteur syriaque sans passer par le biais de la PAT.27 Cependant, le débat n’est pas entièrement clos, et certains continuent à avancer des arguments en faveur d’une influence directe du Targoum sur la Peshitta.28

Voir M. P. Weitzman, “Peshitta, Septuagint and Targum” in R. Lavenant, éd., VI Symposium Syriacum 1992, (OCA 247; Rome, 1994), 51– 84, en part. 72–73. 24 Voir A. Vööbus, Peschitta und Targumim des Pentateuchs. Neues Licht zur Frage der Herkunft der Peschitta aus dem altpalestinischen Targum (Stockholm, 1958); Dirksen, “The Old Testament Peshitta”, 271–272. 25 Voir ci-dessus. 26 Voir l’étude fondamentale de R. J. Owens, The Genesis and Exodus Citations of Aphrahat the Persian Sage (Monographs of the Peshitta Institute 4; Leiden, 1983); Idem, “Aphrahat as a Witness to the Early Syriac Text of Leviticus” in The Peshitta: Its Early Text and History, 1–18. 27 Voir S. P. Brock, “A Palestinian Targum Feature in Syriac”, JSS 46 (1995), 271–282; Maori, The Peshitta Version of the Pentateuch, 27–28. 28 Voir en particulier l’étude très riche en données de J. C. de Moor, F. Sepmeijer, “The Peshitta and the Targum of Joshua” in The Peshitta as a Translation, 129–176. 23

12

LANGUAGE AND TEXTUAL HISTORY

MILIEU ET DATE DE LA PAT Même si elle ne doit pas être interprétée dans le sens d’une influence directe, la parenté entre les Targoums et la PAT semble indiquer que cette dernière trouve son origine dans des cercles juifs. La version syriaque est truffée d’interprétations et d’expressions idiomatiques que l’on trouve éga-lement dans les sources juives rabbiniques.29 Paradoxalement, cette traduction a cependant été préservée et transmise exclusivement dans un milieu chrétien. La PAT fait partie de la bible traditionelle des Églises d’expression syriaque, tandis que le judaïsme n’en a qu’une connaissance superficielle et de seconde main,30 mis à part la PAT des Proverbes qui a été adoptée dans la littérature targoumique.31 Cet état des choses a été interprété de différentes façons. Pour les uns, la PAT a son origine dans la diaspora juive de la Mésopotamie. En dehors des interprétations midrashiques, le fait que la version est basée sur un texte hébreu constitue un argument important en faveur de cette hypothèse. Il est en effet difficile d’imaginer comment des non-juifs auraient pu acquérir une connaissance suffisante de l’hébreu ancien pour mener à bien leur tâche. Cette version juive aurait, dans un deuxième temps, été adoptée par la jeune communauté chrétienne située en Orient, et — suite à cette adoption — répudiée par la communauté juive lors de la consolidation rabbinique du judaïsme à la fin du IIème siècle après Jésus-Christ.32 L’histoire de la version serait donc parallèle à celle de la Septante.

29 La plupart des recherches se sont limitées au Pentateuque; voir surtout Maori, The Peshitta Version of the Pentateuch. L’influence de l’exégèse juive est moins sensible, mais certainement présente, dans les autres livres. 30 Voir Weitzman, “Peshitta, Septuagint and Targum”, 81–83. 31 Le rapport entre la Peshitta et le Targoum du livre des Proverbes a été reconnu très tôt dans l’histoire de la recherche. La priorité de la Peshitta, qui a parfois été mise en doute, est démontrée de façon concluante par E. Z. Melammed, “The Targum on Proverbs”, Bar Ilan 9 (1972), 18–91. Je pense, avec Weitzman (voir la note précédente) et contre Brock (“A Palestinian Targum Feature”, 276), que l’emprunt de la Peshitta par les juifs est un fait tardif. Voir ci-dessous aux n. 51 et 52. 32 En général ce sont les savants défenseurs de l’origine targoumique de la PAT qui sont le plus attachés à cette position. Néanmoins, les deux

LA PESHITTA DE L’ANCIEN TESTAMENT

13

Pour d’autres, la version est l’œuvre de traducteurs chrétiens.33 Les éléments juifs que contient la version sont alors expliqués de plusieurs manières: les traducteurs auraient eu recours à des traditions juives qui leur étaient connues; 34 ou bien — comme Jérôme pour la Vulgate — ils auraient consulté des docteurs juifs; ou bien encore ils seraient eux-mêmes issus du Judaïsme.35 L’origine chrétienne de la PAT pourrait être confirmée par certains passages marqués par une interprétation influencée par le Nouveau Testament.36 La discussion entre les diverses positions, menée de façon intermittente, n’a pas toujours été des plus fructueuses.37 Les éléments juifs pourraient dériver d’une continuité entre le judaïsme et le christianisme syriaque, continuité qui est bien attestée en dehors de la question de la PAT;38 ils ne constituent donc pas un argument décisif en faveur d’une origine juive. D’autre part, d’éventuels éléments chrétiens pourraient être le fait de retouches chrétiennes dans une version originalement juive.

hypothèses sont indépendantes et, en repoussant la première, on n’enlève rien à la validité de la deuxième. 33 Le livre de B. J. Roberts, The Old Testament Text and Versions (Cardiff, 1951), a eu une grande influence à cet égard. Plus récemment, l’origine chrétienne de la PAT d’Ésaïe a été argumentée de façon détaillée par A. van der Kooij, Die alten Textzeugen des Jesajabuches (OBO 35; Fribourg, 1981), 273–284. Le dernier à s’être prononcé en faveur de cette position est H. J. W. Drijvers, “Syrian Christianity and Judaism” in J. Lieu et al., éd., The Jews among Pagans and Christians in the Roman Empire (Londres & New York, 1992), 124–146. 34 Voir Drijvers, “Syrian Christianity and Judaism”, 140. 35 Voir Van der Kooij, Die alten Textzeugen, 273–284. 36 Cependant, la plupart de ces lectures peuvent aisément être expliquées en fonction du texte hébreu ou des traditions juives, comme le montre M. P. Weitzman, “From Judaism to Christianity: The Syriac Version of the Hebrew Bible” in The Jews among Pagans and Christians in the Roman Empire, 147–173, en part. 169–171. 37 Voir Dirksen, “Syrian Christianity and Judaism”. 38 Voir J. Joosten, “West Aramaic Elements in the Syriac Gospels: Methodological Considerations” in VI Symposium Syriacum 1992, 101–109, en part. 107–108, avec renvois bibliographiques; Brock, “A Palestinian Targum Feature”.

14

LANGUAGE AND TEXTUAL HISTORY

De nouveaux travaux ont balisé le chemin vers une solution à ce problème fondamental. Suite aux études de Brock et de Petersen, le présent auteur a tenté de démontrer que la PAT a été employée dans les citations vétérotestamentaires du Diatessaron de Tatien.39 Si cette hypothèse est admise, cela implique que la PAT non seulement existait, mais avait atteint un certain prestige au sein de la jeune Église syrophone avant que Tatien ne compose son œuvre, vers l’an 170.40 En effet, si Tatien a cru bon de remplacer le texte grec des citations de l’AT par une version reprise directement dans la PAT — ce qui impliquait parfois d’importantes divergences dans le texte de l’évangile — il n’a pu le faire que sous la pression d’un texte connu et reconnu par son auditoire.41 L’antériorité de la PAT par rapport au Diatessaron nous donne un repère chronologique précieux, car il est à peu près certain que l’harmonie évangélique de Tatien fut la première tentative de rendre les évangiles grecs en syriaque.42 Ayant ce repère chronologique le problème historique peut être résolu: il est en effet inimaginable que l’Église orientale ait fait traduire la plupart des livres de l’AT avant — et même bien avant — d’entreprendre une traduction des évangiles.43 On conclura que la PAT est d’origine juive pré-chrétienne. La réception d'une version juive de l’AT par l’Église syrophone s’accorde bien avec l’idée que cette Église aurait pris son existence Voir J. Joosten, “The Old Testament Quotations in the Old Syriac and Peshitta Gospels. A Contribution to the Study of the Diatessaron”, Textus 15 (1990), 55–76. 40 Le fait semble être admis par R. A. Taylor, The Peshitta of Daniel (Monographs of the Peshitta Institute 7; Leiden, 1994), 321, et par Brock, “A Palestinian Targum Feature”, 275. 41 Le principe de suivre un texte établi dans la “traduction” des citations bibliques est partout en évidence dans la littérature syriaque; voir, p. ex., A. Baumstark, Biblica 16 (1935), 257–299; pour le même principe dans l’Antiquité classique, voir G. Kennedy, “Classical and Christian Source Criticism” in W. O. Walker, éd., The Relationship among the Gospels. An Interdisciplinary Dialogue (San Antonio, 1978), 125–155, en part. 146. 42 Voir, pour tout ce qui touche au Diatessaron, W. L. Petersen, Tatian’s Diatessaron. Its Creation, Dissemination, Significance and History in Scholarship (Supplements to Vigiliae Christianae 25; Leiden, 1994). 43 Voir F. C. Burkitt, Early Eastern Christianity (Londres, 1904), 70– 74. 39

LA PESHITTA DE L’ANCIEN TESTAMENT

15

dans une matrice juive, ce qui malgré une contestation récente demeure une hypothèse des plus vraisemblables.44 Ce premier résultat peut cependant être affiné. Dans une série d’articles, M. Weitzman a insisté sur le fait que la traduction de la PAT a dû s’échelonner sur une période assez longue. Les différents livres de l’AT n’ont pas tous été traduits par la même personne ni à la même date; par une étude du vocabulaire et de la technique de traduction on peut tenter de reconstituer le développement de la version. Selon Weitzman, l’ordre chronologique de la traduction aurait suivi d’assez près l’ordre du canon juif: d’abord le Pentateuque, puis les livres historiques et prophétiques, finalement les Écrits. Pour ce qui concerne la question du milieu d’origine, il est à noter que les livres traduits plus tardivement, comme par exemple les Psaumes, manifestent une proximité plus grande avec la pensée chrétienne que les livres traduits plus tôt, comme le Pentateuque qui reste très proche du judaïsme.45 Ceci pourrait indiquer que la PAT trouve son origine dans une communauté juive en voie de christianisation.46 La théorie de Weitzman a pu être vérifiée par une étude de la Peshitta des Proverbes.47 Le livre des Proverbes, incontestablement un des derniers livres du canon juif à être traduit en syriaque, se caractérise par une pensée très proche du christianisme représenté par la littérature syriaque ancienne, notamment en ce qui concerne l’appréciation positive de la pauvreté. Il est tentant de conclure que la traduction de ce livre s’est faite au moment où la communauté concernée avait embrassé le christianisme et pris connaissance des principaux écrits du Nouveau Testament.48 La PAT serait donc essentiellement une version juive, commencée avant l’an 150 environ et achevée au début du IIème siècle, 44 Voir Brock, “A Palestinian Targum Feature”, 282, contre Drijvers, “Syrian Christianity and Judaism”. 45 Voir M. P. Weitzman, “The Origin of the Peshitta Psalter” in J. A. Emerton, S. C. Reif, éd., Interpreting the Hebrew Bible: Essays in Honour of E. I. J. Rosenthal (Cambridge, 1982), 277–298. 46 Voir M. P. Weitzman, “From Judaism to Christianity”. 47 Voir J. Joosten, “Doublet Translations in Peshitta Proverbs” in The Peshitta as a Translation, 63–72. 48 La traduction de Prov. 11,31 suit la version de 1 Pierre 4,18; l’influence du NT sur l’AT, à l’inverse de ce qui est habituel, est rendue certaine par le vocabulaire employé.

16

LANGUAGE AND TEXTUAL HISTORY

alors que la communauté productrice avait subi des contacts profonds avec le christianisme. Ce résultat est provisoire et demande à être confirmé par d’autres études. Cependant, la façon différenciée d’aborder la question du milieu et de la date de la PAT, en tenant compte de la chronologie interne de la version, constitue sans doute le mérite durable de cette approche.

LA PAT ET LA SEPTANTE L’Osrhoène — la région syrophone dont la PAT est originaire — faisait partie de la zone d’influence de la langue et de la culture grecque.49 Puisque, dans cette zone, la Septante s’était partout imposée comme version grecque faisant autorité, la question d’une éventuelle influence de celle-ci sur la PAT se pose tout naturellement. Même si la PAT a été faite sur un texte hébreu, ce qui est certain,50 n’est-il pas logique de supposer que les traducteurs aient consulté la Septante, surtout dans les cas où l’hébreu était difficile ou incompréhensible? Par le passé cette question a généralement reçu une réponse affirmative, quoique les voix contraires n’aient pas fait défaut.51 Le premier acquis des années récentes résulte d’une plus grande attention au problème méthodologique posé par la question de l’influence d’une version biblique sur une autre.52 Les correspondances entre la PAT et la LXX contre le texte hébreu massorétique sont légion, mais elles ne doivent pas forcément être interprétées toutes dans le sens d’une influence directe. Pour commencer, 49 Voir H. J. W. Drijvers, “Hatra, Palmyra und Edessa”, ANRW II 8 (1977), 799–906, en part. 888–889. 50 Rappelons que tous les livres canoniques de la PAT ainsi que le livre de Ben Sira ont été traduits d’un texte hébreu. La plupart des autres livres apocryphes furent traduit du grec, mais la chose n’est pas certaine pour Judith et Tobie. 51 Un bref état de la question est dressé par Weitzman, “Peshitta, Septuagint and Targum”, 51–57. 52 Voir J. Lund, The Influence of the Septuagint on the Peshitta: A Reevaluation of Criteria in Light of Comparative Study of the Versions in Genesis and Psalms, thèse non publiée (Jérusalem, 1988); P. B. Dirksen, “The Peshitta and Textual Criticism of the Old Testament”, VT 42 (1992), 376–390; Weitzman, “Peshitta, Septuagint and Targum”.

LA PESHITTA DE L’ANCIEN TESTAMENT

17

un bon nombre de correspondances remontent sans doute à un texte hébreu divergent du texte massorétique et employé comme Vorlage des deux versions. Ensuite, il faut tenir compte de la possibilité d’une polygénèse:53 une difficulté dans le texte hébreu a pu être résolue indépendamment de façon semblable dans la Septante et dans la PAT. Finalement, une variante commune aux deux versions peut être due à une tradition exégétique connue par les traducteurs. Par ailleurs, même si l’influence de la Septante est certaine, il faut encore distinguer entre emploi direct par le traducteur et retouches ultérieures par un correcteur. Deux critères généraux formulés par Weitzman permettent d’avancer avec plus de certitude. Pour établir qu’une variante commune est due à l’influence de la LXX sur la PAT il faut qu’elle remplisse deux conditions. Premièrement, elle ne doit pas être attestée dans d’autres versions ou écrits exégétiques, dans lequel cas on pourrait expliquer la correspondance par l’existence d’une tradition exégétique reflétée indépendamment par les deux versions. Deuxièmement, la variante ne doit pas s’expliquer logiquement par le contexte, de peur qu’il s’agisse d’une simple coïncidence. L’application de ces critères permet un progrès substantiel dans la réponse à la question qui nous occupe. Il s’avère que dans toute une série de livres, tels le Pentateuque et le livre de Job, 54 l’influence de la Septante est sporadique sinon inexistante. Par contre, d’autres livres, tels le livre de Ruth ou des Proverbes, 55 portent les signes évidents de cette influence. Notons encore, quoique le fait n’ait pas été suffisamment exploré, que l’influence de la LXX semble s’accroître au fur et à mesure du développement de la PAT. Le livre de la Genèse, qui a certainement été l’un des premiers à être traduit en syriaque, ne contient pratiquement aucune variante dont 53 Le

terme est repris de l’article de Weitzman. Pour la Genèse, voir Dirksen, “The Peshitta and Textual Criticism of the Old Testament”; pour le Pentateuque en général, voir Maori, The Peshitta Version of the Pentateuch, 317–318; pour Job, voir Weitzman dans les actes du congrès IOSOT, J. A. Emerton, Congress Volume Cambridge 1995, Leiden, 1997, 381–399. 55 Pour Ruth, voir Weitzman, “Peshitta, Septuagint and Targum”; Proverbes, voir Joosten, “Doublet Translations”. 54

18

LANGUAGE AND TEXTUAL HISTORY

l’explication doit avoir recours à la LXX.56 D’autre part, le livre des Proverbes, dont la date relativement tardive a été discutée cidessus, est marqué par cette influence à un tel point que l’on est en droit de se demander si ce livre est une traduction de l’hébreu ou un “collage” des textes hébreu et grec en version syriaque.57 Les autres livres se situent en général entre ces deux extrêmes.58

PERSPECTIVES DE RECHERCHE C’est dans les domaines discutés ci-dessus que les études peshittoniennes ont été marquées du progrès le plus significatif. Cela étant dit, beaucoup d’hypothèses restent à vérifier dans le détail. Des solutions durables ne pourront, à mon sens, être atteintes que par un accroissement des études linguistiques et philologiques. Une réponse définitive aux questions du développement historique de la version, de sa relation avec d’autres versions bibliques et, de là, de son milieu d’origine passera nécessairement par une étude de sa syntaxe, de son vocabulaire et de sa technique de traduction. Des travaux préliminaires existent déjà dans ces domaines. Une étude de la syntaxe du Pentateuque syriaque a été entreprise par I. Avinery dans sa thèse de doctorat, dont les principaux résultats ont été publiés dans une série d’articles.59 Un petit nombre de travaux d’autres auteurs pourraient aussi être cités.60 La recherche lexicale de la PAT est moins riche; à vrai dire, nous ne disposons ici que de

56 Voir n. 54. Weitzman, “Peshitta, Septuagint and Targum”, 57–58, trouve un indice de l’influence de la LXX dans la Peshitta de Gen. 4,8, mais puisque l’interprétation concernée est attestée dans d’autres écrits exégétiques son argumentation est peu convainquante; voir Maori, The Peshitta Version of the Pentateuch, 235–238. 57 Voir Joosten, “Doublet Translations”, 72. 58 Voir A. Gelston, The Peshitta of the Twelve Prophets (Oxford, 1987), 160–177; Taylor, The Peshitta of Daniel, 311 (influence de la Septante), 311– 312 (influence de Théodotion). 59 La liste complète de ces articles peut être trouvée dans Dirksen, Annotated Bibliography. 60 Voir, p. ex., G. Goldenberg, “Bible Translations and Syriac Idiom” in The Peshitta as a Translation, 25–39.

LA PESHITTA DE L’ANCIEN TESTAMENT

19

quelques contributions plutôt modestes.61 La publication de concordances sur la plupart des livres de la PAT est venue faciliter ce travail.62 La technique de traduction de la version a surtout été étudiée dans les limites de chaque livre particulier.63 Les travaux qui dépassent le stade du relevé des différences entre le texte massorétique et la version syriaque — additions, omissions, transpositions, etc. — sont généralement consacrés aux correspondances lexicales entre la source et la traduction.64 Il n’existe que très peu de recherches concernant l’aspect syntaxique de la traduction de l’hébreu en syriaque.65 Il subsiste donc là un vaste champ à défricher. Une question qui n’a pas encore été abordée, et qui ne l’est que rarement dans la littérature récente, est celle de l’emploi de la PAT dans la critique textuelle de l’Ancien Testament. La raison de cette réticence semble être qu’il existe un consensus virtuel soutenant que la source hébraïque de la traduction syriaque était généralement très proche du texte massorétique. Par conséquent, la plupart des exégètes hésitent à retenir des variantes attestées par la PAT si celles-ci ne sont pas confirmées par d’autres témoins tex61 Voir K. Beyer, “Der reichsaramäische Einschlag in der ältesten syrischen Literatur”, ZDMG 116 (1966), 242–254; Y. Maori, “Midrashic Influence on the Peshitta’s Choice of Words”, Tarbiz 46 (1976), 212–230. Voir également l’étude riche en données non-dépouillées de de Moor et Sepmeijer, “The Peshitta and the Targum of Joshua”. 62 Dans la série Göttinger Orientforschungen, W. Strothmann, a publié des concordances sur les livres suivants: Ecclésiaste (1973), Psaumes (1976), Prophètes (1984), Pentateuque (1986); un index sur les Apocryphes et le Deutérocanoniques (1988). Voir cependant n. 8 ci-dessus. Une concordance basée sur le texte édité à Leiden est en préparation, voir A. van der Kooij, “Peshitta Progress Report” in The Peshitta as a Translation, 219. 63 Voir, p. ex., H. M. Szpek, Translation Technique in the Peshitta to Job. A Model for Evaluating a Text with Documentation from the Peshitta to Job (SBL.DS 137; Atlanta, 1992). 64 Voir van der Kooij, “On the Significance of MS 5b1”; P. G. Borbone, “Correspondances lexicales entre Peshitta et TM du Pentateuque. Les racines verbales.” in The Peshitta as a Translation, 1–16. 65 On peut trouver quelques indications dans les articles de Avinery; voir Dirksen, Annotated Bibliography.

20

LANGUAGE AND TEXTUAL HISTORY

tuels. Cependant, l’établissement d’un texte relativement fiable, et la reconnaissance de la PAT comme une version relativement ancienne et indépendante devrait incliner les chercheurs vers une révision de ce jugement. Dans la pratique une grande prudence sera toujours nécessaire, mais du point de vue théorique il est certain que la PAT contient des leçons dues à un texte hébreu qui divergeait du texte massorétique. Entre ces variantes qui n’ont pas leur origine dans le processus de traduction il pourra se trouver des leçons originales. La critique textuelle devra nécessairement en tenir compte.66 Pour terminer j’exprimerai un souhait. Les richesses de la PAT en tant que témoin textuel et en tant qu’interprétation ancienne du texte biblique sont pour beaucoup rendues inaccessibles par l’obstacle de la langue syriaque. Ne serait-il pas opportun d’envisager une traduction annotée de ce texte vénérable, sur le modèle des traductions du Targoum et de la Septante qui sont actuellement en cours de parution? 67 Un tel ouvrage pourrait contenir, en une des langues scientifiques modernes, une traduction du texte de la PAT avec un relevé des déviations par rapport au texte massorétique et un commentaire sur les déviations les plus importantes.

66 Voir J. Joosten, “1 Samuel xvi 6, 7 in the Peshitta Version”, VT 41 (1991), 226–233. 67 Pour le Targoum, voir la série The Bible in Aramaic, éditée par B. Grossfeld et al.; pour la Septante, la série La Bible d’Alexandrie, éditée par M. Harl et al.

THE USE OF SOME PARTICLES IN THE OLD TESTAMENT PESHITTA** In an article dealing with the text of the Old Testament Peshitta M. H. Goshen has summed up some main traits of the Syriac Old Testament: “There is no other Syriac text which can compare in importance with the Old-Testament Peshitta. Not only is this the one direct non-midrashic translation of the Hebrew Bible into a language closely related to Hebrew and therefore invaluable to the study of the Bible text; it is also practically the oldest and best-attested text in Syriac, with an influence second to none on the development of that language.”1 Such a judgment warrants intensive study, not only of the text, but also of the language and translation technique of the Old Testament Peshitta. However, very little work has been done on the language of the Old Testament Peshitta. There are no specialized studies of its vocabulary,2 and research on its syntax is still in an * I wish to thank G. Marqis who has helped me to improve the English of the present article, and S. Asif who has made helpful remarks concerning the OT Peshitta. 1 M. H. Goshen-Gottstein, “Prolegomena to a Critical Edition of the Peshitta,” Scripta Hierosolymitana VIII, Ch. Rabin, ed. (Jerusalem, 1961), 26–67. 2 The only studies of aspects of OT Peshitta vocabulary are those which try to trace non-Syriac vocabulary in the OT Peshitta: K. Beyer, “Der reichs-aramäische Einschlag in der ältesten syrischen Literatur,” ZDMG 116 (1966), 242–254, gives examples of influence of Official Ara-

23

24

LANGUAGE AND TEXTUAL HISTORY

embryonic stage.3 This lack of research confronts textual critics who want to utilise this version with two sets of problems. Firstly, because we have no exact knowledge of the language of the Peshitta version, we may encounter a word, expression or syntagm whose meaning is unclear. Secondly, because the Syriac of the OT has not yet been studied thoroughly enough to categorize the language (classical Syriac, Old Syriac, Jewish Syriac, or “Hebraized” Syriac) we do not always know how to approach linguistic problems of the OT Peshitta. The present article makes a modest contribution to a better knowledge and a more exact characterization of the language of the OT Peshitta. Four particles will be studied with respect to their use and function in the OT Peshitta, in relation to matters of language and translation technique. The syntactic function of these particles will be defined, and it will be demonstrated that the use of each of these particles is stylistic, not related to the Hebrew Vorlage or Syriac syntax. Thus the use of these particles exemplifies the general translation technique of the OT Peshitta, which, though adhering closely to its Hebrew Vorlage, is not exaggeratedly literal. The authors of the OT Peshitta knew how to make use of nuances in their target language beyond the implications of the source language. At the same time, the syntax of these particles assists us in categorizing the language of the OT Peshitta. It will be shown that the idiomatic use of these particles is not limited to the OT Peshitta, but is also found in other early Syriac works. Thus, in the use of these particles, the OT Peshitta reflects the Syriac style of the oldest period known to us. maic; Y. Maori, “Midrashic Influence on the Peshitta’s Choice of Words,” Tarbiz 46 (1976), 212–230, brings some interesting instances of JewishAramaic influence. The question of non-Syriac influences in the OT Peshitta is important but should not obscure the fact that, generally spoken, the Peshitta is written in idiomatic Syriac. 3 The only work I know of is the doctoral thesis of I. Avinery, Syntaxe de la Peshitta sur le Pentateuque (Jerusalem, 1973) [Hebrew], and the articles based on it, in Leshonenu 38 (1973–74), 220–4 [Hebrew]; JNES 34 (1975), 123–7; Israel Oriental Studies 5 (1975), 45–6; Semitica 25 (1975), 105– 9; Afroasiatic Linguistics 3 (1976), 108–9; Le Muséon 90 (1977), 421–6; JSS 22 (1977), 48–9.

USE OF SOME PARTICLES

25

‫‘ ܐܝܟܘ‬WHY’ Already the native Syriac grammarians noted that while ‫ܐܝܟܐ ܗܘ‬ always means ‘where’, ‫ ܐܝܟܘ‬may mean either ‘where’ or ‘why’.4 This statement was taken over by the modern dictionaries. But nowhere do we find noted that ‫ ܐܝܟܘ‬meaning ‘why’ is quite rare and almost entirely limited to the OT Peshitta. Also, upon examining the evidence, the particle turns out to have a very specific, idiomatic function. The occurrences from the OT Peshitta are the following: Num. 22:37 1 Sam. 1:8 1 Sam. 15:19 1 Sam. 26:15 Ezek. 18:19 Job 3:16

‫ܗܐ ܫܕܪܬ ܥܠܝܟ ܠܡܩܪܝܟ ܐܝܟܘ ܐܠ ܐܬܝܬ ܠܘܬܝ‬ ‫ܚܢܐ ܠܡܢܐ ܒܟܝܐ ܐܢܬܝ ܘܐܝـܟܘ ܐܠ ܠܥܣܬܝ ܘܠܡܢܐ‬ ‫ܒܝܫ ܠܒܟܝ‬ ‫ܐܝܟܘ ܐܠ ܫܡܥܬ ܒܩܠܗ ܕܡܪܝܐ ܐܐܠ ܐܬܦܢܝܬ‬ ‫ܥܠ ܒܙܬܐ‬ ‫ܗܐ ܓܢܒܪܐ ܐܢܬ ܘܡܢܘ ܐܟܘܬܟ ܒܟܠܗ ܐܝܣܪܝܠ‬ ‫ܐܝܟܘ ܐܠ ܢܛܪܬ ܠܡܪܟ ܡܠܟܐ‬ ‫ܐܝܟܘ ܐܠ ܡܬܦܪܥ ܒܪܐ ̈ܚܛܗܐ ܕܐܒܘܗܝ‬ ‫ܐܝܟܘ ܐܠ ܗܘܝܬ‬

These examples clearly show that ‫ ܐܝܟܘ‬means ‘why’ (translating Hebrew ‫ל ׇמה‬,‫ ׇ‬and in Ezek. 18:19 ‫דּוּע‬ ַַ ‫)מ‬. ַ Moreover, it is always followed by a negative clause.5 It was not, however, necessary to use ‫ ܐܝܟܘ‬in this syntactic situation, as is shown by the places where we find ‫ ܠܡܢܐ ܐܠ‬+ a negative clause (Gen. 12:18, Num. 11:11, 2 Sam. 16:17, 19:25, Job 3:10). Apparently the use of ‫ ܐܝܟܘ‬as opposed to ‫ ܠܡܢܐ‬depended on the choice of the translator. See Thesaurus syriacus, R. Payne Smith, ed. (Oxford, 1879), s.v. See especially 1 Sam. 1:8 where ‘why’ appears three times, but only in the second clause is the particle ‫ ܐܝܟܘ‬used – followed by ‫ ܐܠ‬+ a negative clause. 4 5

26

LANGUAGE AND TEXTUAL HISTORY

This use of ‫ ܐܝܟܘ‬occurs sporadically outside the OT Peshitta, apparently only in the oldest writings.6 We find the particle a few times in the Old Syriac version of the Gospels: Mat. 16:11CS Mrk. 11:31S Luk. 19:23C

‫ܐܝܟܘ ܐܠ ܡܣܬܟܠܝܢ ܐܢܬܘܢ ܕܐܠ ܗܘܐ ܥܠ ܠܚܡܐ‬ ‫ܐܡܪܬ ܠܟܘܢ‬ ‫ܐܝܟܘ ܐܠ ܗܝܡܢܬܘܢ ܒܗ‬7 ‫ܐܝܟܘ ܐܠ ܐܪܡܝܬ ܟܣܦܝ ܥܠ ܦܬܘܪܐ‬

‫ ܐܝܟܘ‬in these verses reflects the same usage as in the OT Peshitta,

used as ‘why’ (it translates διὰ τί in all the cases except Mat. 16:11 πῶς and followed by ‫ ܐܠ‬+ a negative clause. In all these instances the Peshitta of the New Testament has corrected the particle, in Mat. 16:11 to ‫ ܐܝܟܢ‬in the other texts to ‫ܠܡܢܐ‬.8 This probably indicates that at a later stage of Syriac ‫ ܐܝܟܘ‬was no longer understood in the sense ‘why?’. We find the particle also in the very old texts contained in the Spicilegium Syriacum:9 Spic., p. 3, l. 4 ‫ܐܝܟܘ ܐܠ ܥܒܕܢ ܐܠܗܐ ܕܐܠ ܢܚܛܐ ܘܢܬܚܝܒ‬ “Why did God not make us so that we wouldn’t sin and be found guilty?” Spic., p. 27, l. 3 ̈ ‫ܐܝܟܘ ܐܠ ܥܒܕܢܝ ܐܠܗܐ ܕܠܗ ܦܠܚܬ ܕܝܢ ܘܐܠ‬ ‫ܠܨܠܡܐ‬

6 I have used the edition of F. C. Burkitt, Evangelion da-Mepharreshe (Cambridge: University Press, 1904), vol. I–II, for the Curetonian (C), and the edition of A. S. Lewis, The Old Syriac Gospels (London, 1910) for the Sinaitic (S) recension.

7 Also

in the parallel text Luk. 20:5CS.

That the Peshitta of the New Testament is basically a revised and corrected edition of an older text of the Old Syriac type is generally recognized, see B. M. Metzger, The Early Versions of the New Testament, Their Origin, Transmission and Limitations (Oxford, 1977), 60f. and the literature cited there. 9 Spicilegium syriacum, W. Cureton, ed. (London, 1885). 8

USE OF SOME PARTICLES

27

“Why did God not make me so that I would serve him and not the effigies?” Finally, the particle is also used by Ephraem:10 Prose Refutations11 vol. 1, 79:36 ‫ܩܢܛ ܗܘܐ ܓܝܪ ܡܢ ܫܪܝܪܘܬܗ‬ ‫ܕܟܝܢܐ ܕܕܠܡܐ ܢܟܣܝܘܗܝ ܘܐܐܠ ܐܝܟܘ ܐܠ ܐܫܬܟܪܬ‬ ̈ ‫ܒܐܕܢܝܗܘܢ ܬܫܥܝܬܐ ܗܦܟܝܬܐ‬ “For he was afraid of the truth of Nature lest it should refute him. But if not, how (?) was the perverse tale not disgraced in their ears?” Contra Haereses12, p. 115 ‫ܕܡ ܟܝ ܐܚܪܝܢ ܠܒܫܟ ܘܐܝܟܘ ܕܐܠ ܐܪܓܫܬ‬ “Or maybe someone else has possessed you and why did you not feel it?” I have not found the particle ‫ܐܝܟܘ‬, ‘why?’ in later Syriac literature. The fact that the correctors of the NT Peshitta felt obliged to correct it is at least an indication that it belongs to the oldest stratum of Syriac only.

‫ ܓܝܪ‬IN RHETORICAL QUESTIONS The student of Syriac is familiar with the particle ‫ ܓܝܪ‬which like γάρ means ‘for, because’ and usually takes the second position in the clause. This particle occurs in the OT Peshitta, albeit rarely (e.g. Exod. 23:9, Deut. 10:9, 18:2, Neh. 8:9). Equally rare are the cases where ‫ ܓܝܪ‬does not mean ‘for, because’, but has a different function: Gen. 4:9 ‫ܢܛܘܪܗ ܐܢܐ ܓܝܪ ܕܐܚܝ‬ 10 The

examples from Ephraem are listed in the Thesaurus syriacus and in the Lexicon syriacum of C. Brockelman (Halle, 1928). 11 S. Ephraim’s Prose Refutations of Mani, Marcion and Bardaisan, C. W. Mitchell, ed., vol. 1–2 (London, 1912 and 1921). The English translation is Mitchell’s. 12 E. Beck, Des heiligen Ephraem des Syrers Hymnen contra Haereses, (Leuven, 1957), vol. 1. Another example occurs in the same hymn, ibid., 119.

28

LANGUAGE AND TEXTUAL HISTORY

Isa. 36:10 ‫ܗܫܐ ܓܝܪ ܒܠܥܕ ܡܢ ܡܪܝܐ ܣܠܩܬ ܐܠܪܥܐ ܗܕܐ ܠܡܚܪܒܘܬܗ‬ Jer. 13:12 ‫ܡܕܥ ܓܝܪ ܐܠ ܝܕܥܝܢܢ ܕܟܠܗܘܢ ܓ̈ܪܒܐ ܡܬܡܠܝܢ ܚܡܪܐ‬ Job 1:9 ‫ܣܪܝܩܐܝܬ ܓܝܪ ܕܚܠ ܐܝܘܒ ܐܠܠܗܐ‬ Job 6:22 ‫ܐܠܘ ܓܝܪ ܐܡܪܬ ܠܟܘܢ ܗܒܘ ܠܝ‬ Job 21:29 ‫ܐܠ ܓܝܪ ܫܐܠܬܘܢ ܠܥܒܪܝ ܐܘܪܚܐ‬ All these clauses are rhetorical questions (introduced in Hebrew by -ַ‫[ ה‬Gen. 4:9, Isa. 36:10, Jer. 13:12, Job 1:9], ‫[ ה ִכי‬Job 6:22] and ַ‫הלא‬ [Job 21:29]) which require the answer ‘no’ (or ‘yes’ in case the question contains a negation). It is difficult to say whether this use of ‫ ܓܝܪ‬imitates a certain use of γάρ or whether, as I would rather think, it reflects the original meaning of this Semitic particle. 13 What is certain is that this use of ‫ ܓܝܪ‬is highly idiomatic: the particle does not correspond to any formal equivalent in the Hebrew and was freely added to enliven the style. Outside the OT Peshitta I have noted the following instances of this usage: Apoc. Bar. 21:16 ‫ܐܘ ܐܠ ܡܕܡ ܐܝܟ ܡܐ ܕܐܝܬܝܢ ܗܘܐ ܓܝܪ ܡܢ‬ ‫ܩܕܝܡ ܗܫܐ ܐܠ ܐܝܬܝܢ‬ “Or is it not so that as we were before we are now no more?”14 Acts of Thomas, vol. 1, p. 214, l. 1415 ‫ܕܝܠܟ ܓܝܪ ܐܠ ܣܦܩܝܢ ܠܟ‬

‫ ܓܝܪ‬is probably akin to Arabic jayr which presumably means ‘yes’, ‘truly’, ‘I know’ or sim., see Lexicon syriacum, s.v. 14 The text is difficult and may not be totally in order, see the note to the passage in the edition of M. Kmosko in Patrologia syriaca, pars 1, tomus 3 (Paris, 1926). 15 Apocryphal Acts of the Apostles, vol. 1–2, ed. by W. Wright (London, 1865). 13

USE OF SOME PARTICLES

29

Wright translates: “For thine own suffice thee not”, but the context shows clearly that the correct rendering is “Do not thine own suffice thee?” Cur. Anc. Doc. 16, p. 56, l. 9 ‫ܐܝܬ ܓܝܪ ܫܢܝܐ ܕܡܫܬܐܠ‬ “Is there (ever) a madman that is interrogated?”

‫ܕܝܢ‬

IN WISH AND APODOSIS

What is true for ‫ ܓܝܪ‬is true for ‫ܕܝܢ‬: it is of such current usage in the connection of phrases (imitating δέ), that one rarely pays attention to it. In the meaning ‘but, and’ ‫ ܕܝܢ‬is fairly rare in the OT Peshitta (e.g. Exod. 21:21, 32:19, Deut. 18:14, Isa. 23:15, Ezek. 37:16, Hos. 1:7, 2:10, Ps. 11:3, 13:6), but alongside this usage we also find ‫ ܕܝܢ‬with a different meaning, which cannot be derived from δέ:17 Gen. 43:10 Num. 11:29 Ps. 119:5

̈ ‫ܘܐܦ ܐܠܘ ܐܠ ܐܫܬܘܚ̈ܪܢ ܟܒܪ ܕܝܢ ܗܦܟܢ ܕܬܪܬܝܢ‬ ‫ܙܒܢܝܢ‬ ̈ ‫ܡܢ ܕܝܢ ܥܒܕ ܟܠܗ ܥܡܗ ܕܡܪܝܐ‬ ‫ܢܒܝܐ‬ ̈ ‫ܐܫܬܘܦ ܕܝܢ‬ ‫ܡܬܩܢܢ ܐܘ̈ܪܚܬܝ‬

Gen. 31:27 ‫ܠܡܢܐ ܐܬܛܫܝܬ ܠܡܥܪܩ ܘܐܬܓܢܒܬ ܡܢܝ ܘܐܠ ܚܘܝܬܢܝ ܫܕܪܬܟ ܕܝܢ‬ ‫ܒܚܕܘܬܐ‬ Job 10:18f ‫ܠܡܢܐ ܡܢ ܡܪܒܥܐ ܐܦܩܬܢܝ ܣܦܬ ܕܝܢ ܘܥܝܢ ܐܠ ܚܙܬܢܝ‬ ‫ܘܗܘܝܬ ܕܝܢ ܐܝܟ ܕܐܠ ܗܘܝܬ ܘܡܢ ܟܪܣܐ ܠܩܒܪܐ ܕܝܢ ܐܘܒܠܘܢܝ‬ Here ‫ ܕܝܢ‬is used in wishes (Num. 11:29, Ps. 119:5) or in apodoses (Gen. 43:10, 31:27, Job 10:18), uncapable of fulfillment. This use of ‫ ܕܝܢ‬is attested mainly in stereotyped expressions like ‫ܟܒܪ ܕܝܢ‬,18 ‫ܡܢ‬ ‫ ܕܝܢ‬+ perfect (= Hebrew ‫)מי ֵיתּן‬, ִ 19 ‫ܐܫܬܘܦ ܕܝܢ‬,20 and ‫ܠܘܝ ܕܝܢ‬.21 Now 16 Ancient

Syriac Documents, W. Cureton, ed. (London, 1848). Lexicon syriacum, s.v. 18 Gen. 43:10; Job 3:13. 19 Num. 11:4; 11:29; Jud. 9:29; 2 Sam. 19:1; Isa. 27:4; Jer. 8:23; 9:1; Ps. 55:7; Job 13:5; 13:15; 14:13; 19:23; 23:3; 29:2; 31:31; 31:35. 20 Gen. 17:18; Num. 14:2; 20:3; Deut. 5:26; Ps. 119:5; Job 9:33. 17 See

30

LANGUAGE AND TEXTUAL HISTORY

and then, however, it is used freely, i.e. outside these fixed expressions (as in Gen. 31:27 and Job 10:18).22 We should not hesitate to see in these cases a survival of the original Semitic meaning of ‫ܕܝܢ‬: ‘then, thereupon’.23 To the translators of the OT Peshitta this meaning was still well-known as is shown especially by those verses in which the particle is used freely. The fixed expressions noted above with the “Semitic” particle ‫ ܕܝܢ‬are found sporadically in Syriac writers from different periods. On the other hand, the free use of it (i.e. outside these stereotyped expressions), which is found in the OT Peshitta, is probably limited to the oldest writings. It occurs a few times in the Old Syriac version of the New Testament:24 Mat. 17:20S

Mat. 23:23CS Luk. 19:23CS Luk. 19:42C

‫ܐܠܘ ܐܝܬ ܗܘܬ ܒܟܘܢ ܗܝܡܢܘܬܐ ܐܝܟ ܦܪܕܬܐ‬ ‫ܕܚܪܕܐܠ ܐܡܪܬܘܢ ܕܝܢ ܠܛܘܪܐ ܗܢܐ ܕܐܫܬܢܐ ܘܢܫܢܐ‬ ‫ܗܠܝܢ ܕܝܢ ܥܒܕܬܘܢ ܘܗܠܝܢ ܕܝܢ ܐܠ ܫܒܩܬܘܢ‬ ‫ܠܡܢܐ ܐܠ ܐܪܡܝܬ ܟܣܦܝ ܥܠ ܦܬܘܪܐ ܘܐܢܐ ܕܝܢ‬ ‫ܐܬܝܬ ܬܒܥܬ ܕܝܠܝ ܒܪܒܝܬܗ‬

‫ܕܐܦܢ ܒܗܢܐ ܝܘܡܐ ܝܕܥܬܝ ܕܝܢ ܫܠܡܟܝ‬ As in the OT Peshitta, ‫ ܕܝܢ‬is used in the Old Syriac version of the NT, in wishes (Mat. 23:23, Luk. 19:42) and in apodoses to conditional clauses (Mat. 17:20S, Luk. 19:23).25 In all the cases cited

1 Sam. 20:14; Isa. 48:18; Job 6:2. Gen. 31:27; Deut. 28:67; Ps. 41:6; Job 10:18f; Lev. 10:18. 23 ‫ ܕܝܢ‬is an original Semitic particle, related to Biblical Aramaic ‫א ַדיִן‬. ֱ For the use of ‫ ܕܝܢ‬in an apodosis compare the use of Hebrew ‫ אזַ י‬in Ps. 124:1–5, and of Arabic ‫ إذأ‬in Koran 12:14. 24 The Semitic use of ‫ ܕܝܢ‬in the Old Syriac Gospels is treated by Burkitt, Evangelion da-Mepharreshe, vol. II, 74. 25 In addition to the cases cited there are 2 cases of ‫ܟܒܪ ܕܝܢ‬, Mat. 11:21CS and its parallel Luk. 10:13CS, and one case of ‫ ܡܢ ܕܝܢ‬+ perfect, Mrk. 16:3S. In these cases the particle was preserved in the Peshitta (in a stereotyped formula!). 21 22

USE OF SOME PARTICLES

31

above the NT Peshitta has omitted ‫ܕܝܢ‬, which indicates that the correctors did not recognize the genuine Semitic use of the particle. From other writings the following instances have been gleaned of the free use of ‫ ܕܝܢ‬in wish and apodosis: Spic., p. 27, l. 3 ‫ܐܝܟܘ ܐܠ ܥܒܕܢܝ ܐܠܗܐ ܕܠܗ ܦܠܚܬ ܕܝܢ ܘܐܠ‬ ̈ ‫ܠܨܠܡܐ‬ For translation see above p. 24. Acts of Thomas26, vol. 1, p. 286, l. 12

̈ ‫ܡܢ ܕܝܢ ܚܕܐ ܡܢ‬ ‫ܥܝܢܝ‬ ̈ ̈ ‫ܥܘܪ‬ ‫ܘܥܝܢܝܟܝ ܕܝܢ ܕܝܠܟܝ ܚܝ̈ܪܢ ܗܘܝ ܒܝ ܒܥܝܕܗܝܢ‬ “Would that someone had blinded one of my eyes and that your eyes looked on me as they used to.”

Ahiqar 27, p. ‫ܡܐ‬, l. 1

‫ܐܠܘ ܒܩܐܠ ܪܡܐ ܡܬܒܢܐ ܗܘܐ ܒܝܬܐ‬ ‫ܚܡܪܐ ܕܝܢ ܬܪܝܢ ̈ܒܬܝܢ ܒܢܐ ܗܘܐ ܒܚܕ ܝܘܡܐ‬ “If a house could be built with a loud voice, the donkey would build two houses in a day.”

Aphrahat28, vol. 1, p. 108, l. 20 ‫ܕܚܠܬܝ ܕܝܢ ܡܢ ܗܕܐ ܐܘܢ ܐܝܙܒܠ‬ ‫ܘܗܝܕܝܢ ܛܢܬܝ ܛܢܢܐ ܕܥܘܐܠ‬ “Would that you had feared this, o Jezebel, and then had been zealous about injustice.” The first clause in Acts of Thomas uses ‫ ܕܝܢ‬in a stereotyped expression (‫ ܡܢ ܕܝܢ‬+ perfect), but the use of ‫ ܕܝܢ‬in the apodosis shows that to the poet the primitive meaning of the particle was well known.

The preceding and following clauses have the same structure, which provides 4 more cases of ‫ܕܝܢ‬. 27 The Story of Ahikar from the Syriac, Arabic, Armenian, Ethiopic, Greek and Slavonic Versions, F. C. Conybeare et al., 2nd ed. (London, 1871). 28 Patrologia syriaca, vol. 1 (Paris 1894), vol. 2 (Paris, 1907), J. Parisot, ed. 26

32

LANGUAGE AND TEXTUAL HISTORY

‫ ܥܕ‬USED AS A CONJUNCTION (+ IMPERFECT) The preposition ‫ ܥܕ‬is not usually, in Syriac, employed as a conjunction. To express ‘until’, ‫ ܥܕܡܐ ܕ‬is used, including the OT Peshitta,

even where the Hebrew has ‫ עד‬+ imperfect (or + perfect).29 At the same time a few cases are evidenced of what is probably an older syntagm, namely ‫ ܥܕ‬+ imperfect. Here are the examples from the OT Peshitta: Isa. 36:17 ‫ܥܕ ܐܬܐ ܘܐܕܒܪܟܘܢ ܐܠܪܥܐ ܕܐܝܟ ܐܪܥܟܘܢ‬ Cant. 2:17 (= Cant. 4:6) ‫ܥܕ ܢܦܘܓ ܝܘܡܐ‬ Dan. 2:9 Dan. 11:36

‫ ܐܬܪܥܝܬܘܢ ܠܡܐܡܪ ܩܕܡܝ ܥܕ ܙܒܢܐ ܢܥܒܪ‬...‫ܡܠܬܐ‬ ‫ ܥܕ ܢܫܠܡ ܪܘܓܙܐ‬...‫ܢܥܒܕ ܡܠܟܐ ܐܝܟ ܨܒܝܢܗ‬

This is not an imitation of the Vorlage because in none of these places does the Hebrew (or, in Dan. 2:9, the Aramaic) text have ‫עד‬ + imperfect. The syntagm ‫ ܥܕ‬+ imperfect is extremely rare also outside the OT Peshitta. It seems to be found only in old works. Here are the examples that have come to my attention: Acts of Thomas, vol. 1, p. 275, l. 8 ‫ܚܕܪܘܗܝ ܕܐܫܦܙܗ ܫܪܝܬ ܥܕ ܢܢܘܡ ܘܥܕ ܢܫܟܒ ܘܡܢܗ ܠܡܪܓܢܝܬܝ‬ ‫ܐܫܩܠܗ‬ “Around his lodging I dwelt, until he would sleep and I could take from him my pearl.” Ahiqar p. ‫ܠܛ‬, l. 4 “Until you complete yours days.” Aphrahat, vol. 1, p. 97, l. 11

‫ܥܕ ܝܘܡܝܟ ܬܫܡܐܠ‬

̈ ‫ܐܝܬ ܓܝܪ ܕܨܐܡ ܡܢ ܠܚܡܐ‬ ‫ܘܡܝܐ‬ ‫ܥܕ ܢܟܦܢ ܘܢܨܗܐ‬

29 There are some cases of ‫ ܥܕ‬+ participle, e.g. Jud. 19:8; Isa. 26:20; Hos. 10:12.

USE OF SOME PARTICLES

33

“For some fast from bread and water until they get hungry or thirsty.” Aphrahat, vol. 2, p. 8, l. 19 ‫ܥܕ ܢܫܠܡ ܙܒܢܐ ܕܣܝܡ ܠܥܠܝܐ‬ “Until the time set by the Most High is completed.”

CONCLUSION We cannot base any far-reaching conclusions on what are after all a few instances of “marginal syntax”. What we can say is that what may appear at first sight to be idiosyncrasies of the OT Peshitta are actually marks of genuine Syriac style of the oldest period known to us. In the use of these particles the OT Peshitta apparently reflects the oldest stratum of Syriac literature.

GREEK AND LATIN WORDS IN THE PESHITTA PENTATEUCH FIRST SOUNDINGS1 Greek words in Syriac texts document the extent to which the Greek language invaded the local Semitic culture, while at the same time they give an indication of the type of Hellenistic influence to which Syriac speakers were exposed. Although the Hellenization of the East may be well understood in general, 2 the study of linguistic influences may contribute in assessing the impact of Hellenism on specific groups and social strata within Mesopotamian society. In this regard, the Greek words in the Peshitta OT have a claim to particular attention. As a matter of fact, it is somewhat amazing that neither of the two large scale studies of Greek words in Syriac — A. Schall’s Studien über griechische Fremdwörter im Syrischen 3 and S. Brock’s “Greek words in the Syriac Gospels”4 — includes treatment of the OT Peshitta.5 The OT Peshitta, in all likelihood dating Thanks are due to S. P. Brock, whose comments on an earlier version of my check-list and on points of detail have preserved me from several errors and oversights. 2 See, e.g., H. J. W. Drijvers, East of Antioch. Studies in Early Syriac Christianity (London, 1984), and other studies by Drijvers. 3 A. Schall, Studien über griechische Fremdwörter im Syrischen (Darmstadt, 1960). 4 S. P. Brock, “Greek words in the Syriac Gospels (vet and pe),” Le Muséon 80 (1967), 389–426. 5 More modest studies on Greek words in Syriac are: L. Haefeli, Stilmittel bei Afrahat, dem persischen Weisen (Leipzig, 1932), 190–195 (a list of Greek words in Aphrahat); F. Altheim, R. Stiehl, “Griechische Fremdwörter im Syrischen” in Idem, Die Araber in der Alten Welt I (Berlin, 1964), 608–617 (criticisms of Schall); S. P. Brock, “Some Aspects of 1

35

36

LANGUAGE AND TEXTUAL HISTORY

from the second half of the second century AD, 6 is older than most of the Syriac inscriptions treated by Schall,7 and much older than any other literary text in that language, including the versions of the Gospels. Moreover, although it is a translation, it is not one from Greek, and as such it is unlikely to contain any ad hoc transcriptions of Greek words which troubled the translator. 8 A study of Greek words in the OTP is bound to yield interesting results, both with regard to the history of the Syriac language as a whole and with regard to the milieu of the specific corpus, which, given its importance and the obscurity of its origins, will be quite welcome. The present inquiry is limited to some preliminary considerations. It is based on an exhaustive collection of the Greek words occurring in the Peshitta Pentateuch.

PROBLEMS OF IDENTIFICATION Establishing a definitive collection of Greek and Latin words occurring in the Peshitta Pentateuch is a time-consuming business. Moreover, one is never entirely sure to have gathered all the Greek and Latin words contained in the corpus, and only those. A first Greek Words in Syriac,” in A. Dietrich, Synkretismus im syrisch-persischen Kulturgebiet (Abh. Ak. Wiss. in Göttingen 96; Göttingen, 1975), 80–108 (methodological considerations); A. Tal, The Language of the Targum of the Former Prophets and its Position within the Aramaic Dialects (Tel Aviv, 1975), 181–186 (Greek words in Targum Jonathan and in Syriac); J. F. Healey, “Lexical Loans in Early Syriac: A comparison with Nabataean Aramaic,” Studi epigrafici e Linguistici 12 (1995), 251–262; M. Lattke, “Die griechischen Wörter im syrischen Text der Oden Salomos,” Aram 5 (1993 [appeared in 1996]), 285–302; S. P. Brock, “Greek Words in Syriac: Some General Features,” Scripta Classica Israelica 15 (1996), 251–262, with more literature on p. 252, n. 8. 6 See the discussion on the date of the OT Peshitta in J. Joosten, “La Peshitta de l’Ancien Testament dans la recherche récente,” Revue d’Histoire et de Philosophie Religieuses 76 (1996). 7 Although, of course, the manuscripts of the Peshitta are much later than the inscriptions. 8 This phenomenon seems to occur even in the OS Gospels: see Brock, “Greek Words,” 410. In later translations from Greek it is much more frequent.

GREEK AND LATIN WORDS

37

obstacle in studying Greek words in Syriac is posed by problems of identification. Greek and Syriac are genetically unrelated and consequently entirely divergent with regard to lexicon and morphology. This fact considerably facilitates the identification of Greek words in Syriac.9 Nevertheless, in a number of cases it is difficult to attain certainty concerning the etymology of the lexeme in question. A few examples will illustrate various reasons why the derivation may not be certain: ܳ ‫ܦܝ‬ ܺ ‘thurible’ has generally been taken to be a loan from a) ‫ܪܡܐ‬ Greek πυρεῖον ‘firepan’.10 Although this identification seems likely,11 there is an important formal divergence between the two words, which led Brockelmann to propose a different Greek word: πύρωμα.12 If nevertheless we hold on to the earlier view, we must suppose the Greek word to have deteriorated in the mouth of Syriܺ ‫ܐ‬ ac speakers. Other examples of deterioration include: ‫ܣܬܝܪ‬ ܰ ܳ ܺ στατήρ; ‫ ܠܡܦܐܕܐ‬λαμπάς (through contamination with a Semitic ܳ ‫ ܩ ܰܠ‬- κέλλα; ܳ ‫ ܺܣ‬- ἄσημον; ‫ܦܝ ܳܠ ܳܣܐ‬ ܺ φιάλη; ‫ܝܬܐ‬ word?); ‫ܐܡܐ‬ ܽ ‫ ܰܩ‬χαράκωμα. ‫ܠܩܘ ܳܡܐ‬ ܳ ‫‘ ܰܩܪܟ‬a precious stone’ is obviously of foreign proveb) ‫ܕܢܐ‬ nance; scholars have connected it with Greek χαλκηδών.13 The 9 In an earlier study I undertook to identify West Aramaic words in Syriac texts, where the close relationship of the two languages makes certain identification of foreign words almost unfeasible. See J. Joosten, “West Aramaic Elements in the Syriac Gospels: Methodological Considerations” in R. Lavenant, ed., VI Symposium Syriacum 1992 (OCA 247; Rome, 1994), 101–109. 10 Payne-Smith, Thesaurus, 3122. ܳ (1 ܳ ‫ܟܪ‬ 11 The interchange of n and m after r is attested also in ‫ܡܬܐ‬ Kgs 6:36), if this is indeed a loan from Greek κορώνις; see Tal, Language, 181. 12 Brockelmann’s proposal is to be rejected because πύρωμα does not express the required meaning. 13 See, e.g., J. Payne Smith, A Compendious Syriac Dictionary (Oxford, 1903), 519. The orthography, however, seems rather to favour a connection with Greek Καρχηδών, Carthage. The gem called “Carthage”, in Ex 28:19 (and 39:12), may owe its existence to the identification of Hebrew ‫ — ַתּ ְר ִשיש‬which designates both a gem and a place in the Western Mediterranean — with Carthage (as evidenced in the Septuagint of Isa 23:1, 6, 10, 14; Ez 27:12; 38:13). The gem ‫ ַתּ ְר ִשיש‬does indeed occur in

38

LANGUAGE AND TEXTUAL HISTORY

problem in this case is that the word is poorly attested in Greek. It is a hapax legomenon found only in the Greek text of Revelation 21:19 — not a very satisfying attestation.14 In other examples, lack ܰ ‫‘ ܐ‬red ̈ܺ ‫ܣܦ‬ of attestation may render the derivation unfeasible: ‫ܝܢܝܩܐ‬ 15 cloth’ (?) may reflect ‘Ισπανικός ‘Spanish’, but since the Greek (or Latin) word is not found in the required sense, it has not been included in our list. ܳ c) ‫‘ ܛܪܛܩܠ‬gridiron’ — if it is a loan from Latin craticula ‘small grate’,16 we have here another case of formal deterioration. A more serious problem, however, is the implication that the word has been borrowed directly from Latin. It is usually held that Latin words came to Syriac via Greek,17 and this seems to have been the ܳ ܽ , ‫ܓܝܘ ܳܢܐ‬ ܳ ‫ܩ ܰܠ‬, ܽ ‫ܠ‬, ‫ܝܬܐ‬ case for the other Latin words in our list (‫ܛܪܘܐܠ‬ ܳ ܽ ‫) ܳܩ‬. Craticula, however, seems to be unattested ܳ ‫ܩ‬, ‫ ܰܩܪܕܐܠ‬, ‫ܪܘ ܳܟܐ‬ ‫ܣܛܐ‬ 18 in Greek. ܰ ‘allow’ has never been derived from Greek ἄφες. It is d) ‫ܐܦܣ‬ true that few of the Greek words in Syriac are verbs (the only other Ex 28, though not in v. 19 but in v. 20, where the Peshitta transliterates the word. In the Targumic tradition, ‫ – כדכדינה‬apparently reflecting the same Greek word – occurs in v 18. Since the ancients seem not to have had a clear notion of the identity of the twelve gems listed in Ex 28:17–20, it is not unlikely that two different renderings of ‫ ַתּ ְר ִשיש‬should have ended up in the Peshitta version. Καρχηδών itself, of course, reflects Semitic qrt ḥdšt. The form of the word in Syriac clearly shows it to derive from Greek, however, and not from Phoenician directly. 14 See W. W. Reader, “The Twelve Jewels of Revelation 21:19–20: Tradition History and Modern Interpretations,” JBL 100 (1981), 433–457; Reader brings many interesting parallels to the list of gems in Rev 21:19, but none of his other Greek lists attests Καλκηδών. 15 See, however, the suggestion of R. P. Gordon, “The Gladius Hispaniensis and Aramaic ’ISPANIQE”, VT 35 (1985), 496–500. It is also possible that the word reflects Greek φοῖνιξ. 16 The identification seems likely for semantic reasons: craticula is indeed a gridiron (etymologically and materially). 17 See Brock, “Aspects”, 90. 18 It does not figure in S. Daris, Il lessico latino nel greco d’Egitto [The Latin vocabulary in Greek language of Egypt] (Papyrilogica Castroctaviana 3; Barcelona, 1971); nor in E. A. Sophocles, Greek Lexicon of the Roman and Byzantine Periods. From BC 146 to AD 1100, (New York & Leipzig, 1888).

GREEK AND LATIN WORDS

39

ܰ

ܺ cases occurring in the Pentateuch are ‫ܓܠܦ‬, ‫ܬܛܦܝܣ‬ ‫)ܐ‬.19 Neverthe20 less, the distribution of the word in Syriac, and the fact that it cannot be connected to a Semitic root, 21 are weighty arguments favouring the identification. ܺ ‫‘ ܣ‬fine flour’ has often been e) On the other hand, ‫ܡܝ ܳܕܐ‬ viewed as a loan from Greek σεμίδαλις ‘fine flour’,22 but it probably isn’t. As Landsberger was able to show it is really the other way round:23 the word is a genuine Semitic one and was at some time borrowed into Greek. The example reminds us that, although Syriac loanwords are very rare in Greek,24 a number of Semitic words had invaded the Greek language at an early period.25

Also remarkable is that the imperative should have provided the form which was borrowed. In a private communication, S. Brock has pointed to a possible parallel: ‫ ܬܪܣܘ‬from θαρσέω (see Brock, “Greek Words”, 402). 20 The occurrence of the word in Lev 19:29 is the result of an accident: it doesn’t reflect the Hebrew text, and seems to have come about by erroneous interpretation of Targumic ‫פסס‬. The only other occurrence of the verb in the OT Peshitta seems to be in Eccl 2:20; elsewhere ‘to permit’ is rendered as ‫ܫܒܩ‬. 21 The connection with the root ‫‘ פסס‬to throw lots’ is suspect. 22 See, e.g., M. D. Koster, “Peshitta Revisited. A Reassessment of its Value as a Version”, JSS 38 (1993), 235–268, here 247. 23 See B. Landsberger, “Zur Mehlbereitung im Altertum”, Orientalische Literaturzeitung 25 (1922), 337–344, here 343f. 24 Socio-linguistic studies have shown that in a bilingual community, the “high” or prestige language remains relatively untainted by any systematic influences from the relatively “low” language, whereas the “low” language tends to undergo heavy influence from the “high” language, especially with regard to the lexicon. See, e.g., the remarks on Aramaic and Greek in 1st century Palestine in G. H. R. Horsley, New Documents Illustrating Early Christianity, vol. 5 (New South Ryde, 1989), 5–40. 25 For a review of the debate on these loans, see M. Masson, “À propos des critères permettant d’établir l’origine sémitique de certains mots grecs,” Comptes Rendus du Groupe Linguistique d’Études ChamitoSémitiques 24–28 (1979–84), 199–231 (with literature). 19

40

LANGUAGE AND TEXTUAL HISTORY

GREEK WORDS SLIPPING INTO THE MANUSCRIPT TRADITION When the problems of identification have been cleared, another problem comes to the fore: one cannot always be sure that the Greek word is part of the original translation. In some cases, a word of Greek origin appears to have been introduced into the text by later scribes or correctors, instead of a more original, genuinely Syriac, expression.26 ܳ ‫ — ܟ‬both meaning, of ܺ supplanting ‫ܣܦܐ‬ The case of ‫ܣܐ ܳܡܐ‬ course, silver — is notorious. In the books of Genesis and Exodus the oldest manuscript, 5b1 in the Leiden edition, translates Hebrew ‫ ֶּכ ֶּסף‬throughout with Syriac ‫ܟܣ ܳܦܐ‬.27 All the other manuscripts ܳ ‫ ܟ‬when the word refers to money, but they have preserved ‫ܣܦܐ‬ ܺ ܳ substitute ‫ ܣܐܡܐ‬whenever the reference is to silver as metal. As has been shown convincingly by Koster, 5b1 reflects an older stage of the text in this respect. This means that the use of the Greek ܺ has to be ascribed to a later corrector or editor of the word ‫ܣܐ ܳܡܐ‬ text of Genesis and Exodus. Another example of this process may perhaps be found in the translations of Hebrew ‫ע ַט ֵלף‬, ‘bat’ (?), in Lev 11:19. One old manuscript of the Peshitta, 6b1, reads the genuine Syriac equivalent ܰ ܽ ‫ܦܪ‬ ‫ܚܕܘ ܳܕܐ‬ , ‘bat’, whereas all other manuscripts have the loanword ܰ ܳ ‫‘ ܛܘܣܐ‬peacock’. In view of Weitzman’s demonstration to the effect that no manuscripts of the Peshitta are tainted by revision according to the Hebrew, 28 the suggestion may be advanced that 6b1’s reading is original and that all the other manuscripts reflect a correction, which introduced a Greek loanword. If this is correct, one would further have to say that in the parallel passage, ܳ ‫ ܰܛ‬, the original Deut 14:17, where all Syriac manuscripts read ‫ܘܣܐ‬

Koster, “Peshitta Revisited”, 247. See M. D. Koster, The Peshitta of Exodus (SSN 19; Assen, 1977), ܳ ‫ܣ‬. 70–72; there are two exceptions: in Ex 11:2; 12:35, 5b1 has ‫ܐܡܐ‬ 28 M. P. Weitzman, “The Originality of Unique Readings in Peshitta Ms 9a1,” in P. B. Dirksen, M. J. Mulder, eds., The Peshitta: Its Early Text and History (Monographs of the Peshitta Institute 4; Leiden, 1988), 225– 258. 26 27

GREEK AND LATIN WORDS

41

reading, which would have been identical to the one in Leviticus,29 has been lost from the manuscript tradition entirely. This hypothesis is admittedly a bit tenuous; but it would solve another problem: it is extremely difficult to envisage a Jewish scholar — such as the translator of the Peshitta Pentateuch undoubtedly was — including the peacock, which may be eaten, 30 in a list of impure animals.31 The corrector who for some reason substituted “peacock” for “bat” probably operated at a later period, when the Peshitta had become the Bible of the Syriac-speaking Church.32 There are several more cases where a Greek word seems to be secondary in the textual tradition.33 Given that the oldest manuscripts of the OTP are centuries younger than the original translation, it is possible that some Greek words supported by all the manuscripts have been introduced in the text at a later stage. However this may be, all these words have nevertheless been included in the list, on the understanding that they are connected to the history of the version.

29 The list is substantially the same in the Peshitta in Leviticus and in Deuteronomy. 30 See Talmud Bavli Hullin, 116b. I thank A. Shinan for this reference. 31 On this point I disagree with J. Emerton, “Unclean Birds and the Origin of the Peshitta,” JSS 7 (1962), 204–211, here 210f. According to Emerton, the translator could have been a Jew who was negligent, or ill ܰ he does not ܳ ‫;ܛ‬ informed. Y. Maori also argues for the originality of ‫ܘܣܐ‬ see a problem in ascribing this rendering to a Jewish translator, see Y. Maori, The Peshitta Version of the Pentateuch and Early Jewish Exegesis, Jerusalem, 1995 [Hebrew], 333. 32 I am not at all sure how to evaluate the variant ταων/ταον, which occurs in the margins of the Codex Coislinianus (7th C) and of a Leipzig manuscript (10th C) of the LXX, at Lev 11:18 (not the same verse). The LXX reading in v. 19 is νυκτερίδα (νυκτικόραξ in Deut); the LXX does not include the peacock in its list of unclean animals. ܳ ‫ܓ‬, v.l. ‫ܠܡܐ‬ ܳ ‫;ܨ‬ 33 Note the following cases: Deut 4:16 ‫ܢܣܐ‬ ܰ Num 9:14 ܳ ܽ ܽ ܽ ܺ ܳ ܳ ܳ ܳ ܳ ܳ ‫ܩܐ‬ ܳ ܺ ‫ ܢܡܘܣܐ‬v.l. ‫ ;ܦܘܩܕܢܐ‬Exod 28:31 ‫ ܦܪܙܘܡܐ‬v.l. ‫;ܦܪܝܣܐ‬ Gen 42:3 ‫ܪܣܐ‬ ܳ ܽ ܳ v.l. ‫ܣܩܘܒܐܠ‬. It is not certain that in all these examples the Greek word is secondary and the Semitic word original.

42

LANGUAGE AND TEXTUAL HISTORY

LOANWORDS OR FOREIGN WORDS? The frequency of Greek words occurring in the Peshitta Pentateuch is very uneven. Many of them are used a number of times. ܽ ‫‘ ܳܢ‬law’ and ‫ܒܘ ܳܬܐ‬ ܽ ‫‘ ܩ ܺܐ‬ark, chest’ are very For instance, ‫ܡܘ ܳܣܐ‬ common and occur in different books of the Pentateuch. Some of these more frequent words even function as the equivalent of more ܰ than one Hebrew word: ‫‘ ܓܠ ܦ‬carve, engrave’ is used to render the ܽ ‫ܳܢ‬ Hebrew roots ‫( פתח‬e.g. Ex 25:9) and ‫( פסל‬e.g. Deut 7:25); ‫ܡܘ ܳܣܐ‬ is Hebrew ‫( תו ָֺרה‬e.g. Lev 7:1), ‫ חק‬or ‫( ֻח ׇקה‬e.g. Gen 47:26; Lev 18:3) and ‫( ִמ ְש ָפט‬e.g. Gen 40:13). The distribution of these words in the Peshitta Pentateuch, and the fact that they are common also in other Syriac texts, suggest that these are indeed loanwords in the full sense of the word. Although they are of foreign extraction, they have been fully adopted into Syriac, and should, for all practical purposes, be considered Syriac words, liable to be used in speech and in writing whenever the specific meaning attributed to them within the Syriac language system needed to be expressed. ܳ ܰ ܽ ‫ܐ‬ ܳ ‫ܓܘ‬ ܰ ‘field’ or ‫ܝܪܐ‬ Other words — like ‫ܪܣܐ‬ ‫‘ ܗܡ‬hostage’ — occur more rarely or even just once. In some cases the low frequency of these words may simply be due to the fact that the meaning they express was not needed in other passages. In other cases, it seems the word in question just started to penetrate into the language, and was still competing with a word of Semitic stock. Thus Hebrew ‫שַים‬ ִַ ‫‘ ִמגְ ָר‬common land’ is translated three times as ܽ ‫ܐ‬ ܰ in Num 35:2, 3, 4, but elsewhere as ‫ܫܛ ̈ܚܐ‬, a Semitic ‫ܓܘ̈ܪܣܐ‬ ݂ word apparently representing a more conservative usage. 34 A third possibility which should be considered is that some of the less frequent words are not really loanwords at all, but foreign words, 35 used in a specific passage because the translators could not find in their own language a satisfactory equivalent of the Hebrew term. ܳ ܳ ܰ This would seem to be the case of the word ‫ܦܬܪܐ‬ ‫ܐ‬, if indeed it ܳܽ ܳ ܽ Other examples: ‫ܛܪܘ ܳܐܠ‬ competes with Semitic ‫ܒܘܪܐ‬ ‫ܺܦ ܳܝܠ ܳܣܐ ;ܙ‬ ܳ See also below n. 39. competes with ‫ܡܓ ܳܣܐ‬. 35 To illustrate the distinction two examples from modern English may suffice: whereas English “regime” is to be considered a loanword from French, couloir, as in: “the couloirs of French diplomacy were rife with self-gratifying whispers”, is a foreign word, i.e. a French word occurring in an English text. 34

GREEK AND LATIN WORDS

43

reflects Greek φθορά ‘perdition, decay’ as is suggested by Brockelmann.36 The term is used just once, in Num 11:20, where it ܳ ܳ ܰ renders the obscure Hebrew word ‫זָ ָרא‬. Since ‫ܦܬܪܐ‬ ‫ ܐ‬does not seem to occur in later Syriac literature, it may perhaps be regarded as a foreign word: a Greek word employed ad hoc by the translator to express a specific nuance which he perceived in the Hebrew term.37

SOME STATISTICS Keeping the foregoing cautionary remarks in mind, it may now be stated that the Peshitta Pentateuch uses about 50 words of Greek or Latin provenance. The figures for the individual books are: 12 in Genesis, 30 in Exodus, 11 in Leviticus, 14 in Numbers, 10 in Deuteronomy.38 The relatively high number in Exodus is partly due to the subject matter of the book, which is particularly rich in technical terms. Note should also be taken, however, of several Greek words in Exodus for which other books supply native equivalents. Thus, to give one example, Hebrew ‫‘ ֶּח ֶּדר‬room, bedroom’ is rendered by ‫( ܰܩܝܛ ܽܘ ܳܢܐ‬reflecting Greek κοιτών) in Ex 7:28, while in 36 The feminine gender of the word is in favour of this derivation. In any case, a connection between the Syriac lexeme and Jewish Aramaic ’ptr/ptrn ‘weak, lean’ is extremely unlikely. ܽ in Ex 32:4, where it renܳ ‫ܛܘ‬ 37 Another interesting example is ‫ܦܣܐ‬ ders Hebrew ‫ח ֶּרט‬,ֶּ an obscure term designating a tool used to fashion the ܽ here refers to the mold in ܳ ‫ܛܘ‬ golden calf. I would suggest the word ‫ܦܣܐ‬ which the golden calf was cast; ‘mold’ is one of the meanings of Greek τύπος. If this is the case, then it seems the Greek word is used here differently than in later Syriac texts, where it means ‘figure, form, manner’ or ‘exemplar’ — meanings which are not really applicable in Ex 32:4. Perhaps, therefore, we may conclude and say that the translator, having understood the Hebrew text to mean that Aaron fabricated the golden calf in a mold, expressed this notion by the help of a Greek word of this meaning. It is to be noted that the same Greek word occurs in this place in Tg Neofiti (see n. 47 below). 38 Brock’s figures for the different books are slightly different (Brock, “Some Aspects”, 85, 86): Gen 10; Ex 26; Lev 10; Num 17; Dt 12. Since Brock has given a list only for the book of Genesis (where he does ܳܽ ܳ ‫ )ܣ‬I don’t know exactly where the other dinot include ‫ܪܘܐܠ‬ ‫ ܒ‬and ‫ܐܡܐ‬ vergencies arise.

44

LANGUAGE AND TEXTUAL HISTORY

Gen 43:30 and Deut 32:25 the Syriac synonym ‫ ܰܬ ܳܘ ܳܢܐ‬is used for the same Hebrew word. There are five examples of this phenomenon,39 which suggests, firstly, that the book of Exodus was translated by a different person than the other books of the Pentateuch, and, secondly, that the translator of Exodus was, for some reason, more prone than his colleagues to use words of Greek provenance.40 In comparison with other Syriac literature, the number of loanwords in the Peshitta Pentateuch is low. The Syriac Gospels — a much shorter text — contain 84 loanwords from Greek or Latin according to the study of Brock. This may partly be explained by the fact that the Gospels, unlike the Pentateuch, were translated from Greek, but the big increase of Greek loanwords also owes something to the development of the Syriac language.41 Words which might have been used in the Peshitta Pentateuch had they ܳ ܳ ‫ܬܪ‬ been at the command of the translators include ‫ܘܢܘܣ‬ θρόνος ܽ ܳ ‫ܠ‬ ܳ ‫ܓܠܘ‬ ܺ ‫‘ ܺܕ ܰܝ‬covenant’,44 ‫ܣܛܐ‬ ‘throne’,42 ‫ܣܩ ܳܡܐ‬ ‘coffin’,43 ‫ܬܝ ܺܩܐ‬ 45 ‘thief’. Their absence tends to indicate that Syriac had not yet borrowed these terms when the Pentateuch was translated. ܺ ‫ ܐ‬in Ex 21:32 (elsewhere Hebrew 39 The other examples are ‫ܣܬܝܪ‬ ܳ ‫;ܡ‬ ܳ ‫ܦܪܘ‬ ܽ ܰ ‫ܣܬ ܳܕܐ‬ ‫ ֶּש ֶּקל‬is rendered ‫ܬܩܐܠ‬ in Ex 12:7, 22, 23 (elsewhere ܳ ܳ ܽ ܽ ‫ ܳܩ‬in Ex 14:6 ܰ ܳ ‫ ܩܪܕܐܠ ;)ܐܣܩܘܦܬܐ‬in Ex 27:3; 38:3 (elsewhere ‫;)ܫܚ ܳܐܠ‬ ‫ܪܘ ܳܟܐ‬ ܰ ܳ (elsewhere ‫ ;)ܪܟܒܐ‬some of these words are exceptional even in Exodus. The phenomenon should be seen as a special case of the phenomenon indicated above, at n. 34. 40 Or perhaps one should say: more prone to experiment with words of Greek origin. For, as is indicated in the preceding note, the Greek word sometimes stands alone in Exodus as well. Another possibility is that these Greek words are due to a later corrector. 41 Note that in several cases Greek loanwords occur in the Syriac text of the Gospels which do not correspond to the word used in the Greek (see Brock, “Greek Words”, 394, 395, 402, 406, 412, 414, 419, 421, 422). 42 E.g. in Gen 41:40; Ex 11:5; 12:9; Deut 17:18. 43 Notably in Gen 50:26. 44 This word is common in other books of the OTP but it is never used in the Pentateuch. 45 In Ex 22:1, 6, 7.

GREEK AND LATIN WORDS

45

FURTHER QUESTIONS The preliminary interrogations treated in the present paper have, somewhat regrettably perhaps, taken up all the time allotted to me. The more important questions — like: how did these foreign words come to be used in this specific text? or: what do these words teach us about the milieu which produced the Peshitta Pentateuch? — have hardly been touched upon. To my mind, it was necessary first to pay some attention to these preliminary considerations before any farther-reaching conclusions could be aimed at. Allow me, at the end of this paper briefly to indicate some subjects which further exploration of the Greek and Latin words in the Peshitta Pentateuch would have to deal with. Firstly, the question of the influence of the Septuagint should be considered. Could it be, as some have suggested,46 that Greek words were taken over from the Greek version, which was consulted by the translators of the Hebrew text? The question needs to be asked if we want to discuss the provenance of these words. Secondly, the overlap between the Greek words in the Peshitta and the Greek words in the Aramaic Targums needs to be assessed. The Peshitta is known to share exegetic traditions with the Targums, among which traditions concerning the rendering of single words.47 Just such a tradition may have transmitted the “foreign words” which, as stated above, should perhaps be distinguished from true loanwords among the Greek words in the Peshitta Pentateuch.48

46 Discussion (with literature) in J. Lund, The Influence of the Septuagint on the Peshitta. A Re-Evaluation of Criteria in Light of Comparative Study of the Versions in Genesis and Psalms (doctoral dissertation; Jerusalem, 1988), 22– 26, 186–199; see also M. P. Weitzman, “Peshitta, Septuagint and Targum” in R. Lavenant, ed., VI Symposium Syriacum 1992 (OCA 247; Rome, 1994), 51–84, here 59f. And see J. P. Brown, “The Septuagint as a Source of the Greek Loan-Words in the Targums,” Biblica 70 (1989), 194–216. 47 See Weitzman, “Peshitta, Septuagint and Targum,” 65–70; and see also Maori, Peshitta (see n. 31 above), 227–245. 48 See n. 37 above. Other Greek words which may derive from a ܺ ‫ܓܝ‬ ܽ̈ ‫ ܠ‬in Num 24:24 (see Tg Neofiti); ‫ܣܝ ܰܪܣ‬ ܺ in Targumic tradition include ‫ܘܢܝܢ‬ ܳ ܳ Ex 28:32; 39:23 (see Tg Neofiti); in ‫ ܩܠܝܬܐ‬Num 25:8, 8 (see Tg Neofiti).

46

LANGUAGE AND TEXTUAL HISTORY

With regard to those words — undoubtedly the great majority — which will have to be regarded as genuine loanwords, one may further ask whether the borrowing happened on a popular and oral level, or whether it is the result of intellectual activities, such as translation from Greek to Syriac, and happened essentially on a written level. Here, the numerous formal deteriorations in Greek loanwords would seem to favour the first possibility. Finally, it will be particularly instructive to inquire into the domains of life to which Greek and Latin loanwords refer. This type of questioning will assist in defining the precise impact of Hellenism on the community which produced the Peshitta Pentateuch. For the true significance of loanwords is that they are cultural symbols indicative of cultural contacts and influences.

APPENDIX: CHECK-LIST OF GREEK WORDS IN THE PESHITTA PENTATEUCH ܽ ‫ܐ‬ ܳ ‫ܓܘ‬ ܰ - ἀγρός Num 35:2, 3, 4. ‫̈ܪܣܐ‬ ܳ ܰ ‫ ܐܣܘܛ‬- ἄσωτος Deut 21:20. ܺ ‫ ܐ‬- σχῆμα Num 18:7. ܳ ‫ܣܟ‬ ‫ܝܡܐ‬ ܳ ‫ ܐܣܩܦܐ‬- σκύφος Gen 44:2; et alibi. ܺ ‫ ܐ‬- στατήρ Ex 21:32. ‫ܣܬܝܪ‬ ܳ ܳ ܰ ‫ܦܬܪܐ‬ ‫ ܐ‬- φθορά Num 11:20. ‫ ܳܒܣܣ‬- βάσις Ex 25:31; et alibi. ܽ ‫ ܒ‬- βήρυλλος Gen 2:12; et alibi. ‫ܪܘ ܳܐܠ‬ ܰ ‫ ܓܠܦ‬- γλύφω Ex 28:9; et alibi. ܳ ‫ ܓ‬- γένος Gen 1:11; et alibi. ‫ܢܣܐ‬ ܳ ܺ ‫ ܗ‬- ἡμίνα Ex 29:40; et alibi. ‫ܡܝܢܐ‬ ܳ ܰ ‫ ܗܡܝܪܐ‬- ὅμηρος Num 21:29. ܳ ‫ ܰܙ‬- ζεῦγος Gen 45:22. ‫ܘܓܐ‬ ܳ ܺ - τήγανον Lev 2:5; 6:21(14); 7:9. ‫ܛܐܓܢܐ‬ ܰ ܳ ‫ ܛܘܣܐ‬- ταῶς Lev 11:19; Deut 14:17. ܽ - τύπος Ex 32:4. ܳ ‫ܛܘ‬ ‫ܦܣܐ‬ ܳ ‫ ܛܟ‬- τάχα Ex 32:30; Num 23:3. ܽ - τροũλλα, trulla Num 4:7. ‫ܛܪܘ ܳܐܠ‬ ܳ ‫ ܛܪܛܩܠ‬- craticula Lev 2:7; 7:9. ܽ ‫ ܟܪܘ ܳܡܐ‬- χρῶμα Ex 24:10. ܽ ‫ ܠ‬- λεγιών, legio Num 24:24. ‫ܓܝܘ ܳܢܐ‬ ܳ ‫ ܠܓܢܐ‬- λάγανον Ex 29:23; et alibi. ܺ ‫ܡܦܐ ܳܕܐ‬ ‫ ܰܠ‬- λαμπὰς Gen 15:17; Ex 20:18. ܳ ‫ ܠܩܢܐ‬- λεκάνη Lev 8:11; et alibi.

GREEK AND LATIN WORDS

ܳ ‫ܡܘ‬ ܽ - μοχλός Deut 3:5; et alibi. ‫ܟܐܠ‬ ܳ ܽ ‫ ܰܡܪܣܘܦܐ‬- μαρσύπιον Deut 25:13. ܽ ‫ ܳܢ‬- νόμος passim. ‫ܡܘ ܳܣܐ‬ ܺ ܳ ‫ܣܐܡܐ‬ - ἄσημόν passim. ܰܺ ‫ ܣܝܪܣ‬- σειρά Ex 28:32; 39:23. ܺ - φιάλη Ex 25:29; et alibi. ‫ܦܝ ܳܠ ܳܣܐ‬ ܳ ܺ - πυρεῖον Num 4:14; et alibi. ‫ܦܝܪܡܐ‬ ܺ ‫ ܐܬܛܦܝܣ‬- πεῖσαι Gen 24:8; 34:15, 22, 23; Deut 13:9. ܰ - ἄφες (ἀφίημι ) Lev 19:29. ‫ܐ ܦܣ‬ ܽ ܰ ‫ܦܪ‬ ܳ ‫ܘܣܬܕܐ‬ - παραστάδα (παραστάς) Ex 12:7, 22, 23. ܳ ܽ ܳ ‫ ܦܪܙܘܡܐ‬- περίζωμα Gen 3:7; et alibi. ܳ ‫ ܰܦ‬- πόρπη Ex 35:11. ‫ܪܦܐ‬ ܽ ‫ ܰܦ‬- πρόσωπον Deut 4:37. ‫ܪܨܘ ܳܦܐ‬ ܳ ܽ ‫ܩܐ‬ ܺ - κιβωτός passim. ‫ܒܘܬܐ‬ ܺ - καιρός Gen 42:4, 38; Ex 21:22, 23. ܳ ‫ܩܐ‬ ‫ܪܣܐ‬ ܳ ܰ ‫ ܩܕܣܐ‬- κάδος Ex 16:3; 27:3; 38:3. ܽ ‫ ܰܩ‬- κοιτών Ex 7:28; Num 8:3. ‫ܝܛܘ ܳܢܐ‬ ܳ ܳܺ ‫ ܩܝܬܪܐ‬- κιθάρα Gen 4:21. ܳ ‫ ܩ ܳܠ‬- κέλλα (or κέλλιον), cella Num 25:8. ‫ܝܬܐ‬ ܽ ‫ ܰܩ‬- χαράκωμα Deut 20:20. ܳ ‫ܠܩܘܡܐ‬ ܳ ‫ ܩܣܛܐ‬- ξέστης, sextarius Ex 16:33. ܳ ‫ ܰܩ‬- κασία, casia? Ex 30:24. ‫ܣܝܐ‬ ܳ ‫ ܰܩ‬- καλδάριον, calidarium Ex 27:3; 38:3. ‫ܪܕܐܠ‬ ܽ ‫ ܳܩ‬- καροũχα, carruca Ex 14:6. et Sym Is 66:20 ‫ܪܘ ܳܟܐ‬ ܳ ‫ ܰܩܪܟܕܢܐ‬- χαλκηδών or καρχηδών? Ex 28:19; 39:12. ܺ ‫ܗܛܢܐ‬ ‫ ܪ‬- ῥητίνη Gen 37:25; 43:11.

47

MATERIALS FOR A LINGUISTIC APPROACH TO THE OLD TESTAMENT PESHITTA**1 INTRODUCTION1 A notable by-product of the recent flowering of linguistic studies on Syriac are a number of tools and analyses directed toward the language of the Old Testament Peshitta. Express mention may be made of Avineri’s study on the syntax of the Peshitta Pentateuch,2 * This article is a revised version of a paper read at a workshop on the New English Annotated Translation of the Syriac Bible, held at the Peshitta Institute in Leiden on 4–5 February 1999. Thanks are due to T. Muraoka who acted as a respondent to this paper as well as to other participants in the seminar for questions and remarks. I am grateful also to David Phillips for correcting my English style. 1 Abbreviations: Acts of Thomas = W. Wright, Apocryphal Acts of the Apostles (London: Williams & Norgates, 1871); Ancient Documents = W. Cureton, Ancient Syriac Documents (London: Williams & Norgates, 1848); Aphrahat = J. Parisot, Patrologia Syriaca, vol. 1 (Paris: Firmin Didot, 1894), vol. 2 (Paris: Firmin Didot, 1907); Drijvers = H. J. W. Drijvers, Old Syriac (Edessean) Inscriptions (Leiden: Brill, 1972); Spicilegium = W. Cureton, Spicilegium syriacum (London: Williams & Norgates, 1885). The Peshitta of the New Testament (P) is quoted after the edition of the Bible Society, the Old Syriac Gospels (C = Curetonian, S = Sinaiticus) after the edition of A. S. Lewis, The Old Syriac Gospels (London, 1910). 2 See I. Avinery, Syntaxe de la Peshitta sur le Pentateuque (dissertation Hebrew University, Jerusalem, 1973) [Hebrew]. Note also a number of articles by Avineri, some of which are based on the dissertation; for a complete list of these articles, see P. B. Dirksen, An Annotated Bibliography of the Peshitta of the Old Testament (MPI 5; Leiden: Brill, 1989), with an update in P. B. Dirksen, A. van der Kooij, eds., The Peshitta as a Translation (MPI 8; Leiden: Brill, 1995), 221–236.

49

50

LANGUAGE AND TEXTUAL HISTORY

and of P. J. Williams’s on the syntax of Peshitta Kings.3 Important, too, is the concordance to the Peshitta Pentateuch recently published by the Peshitta Institute,4 and other concordances published by Strothmann.5 A new project to be launched by the Peshitta Institute goes by the promising name of CALAP: Computer assisted language analysis of the Peshitta. Important as these studies and tools devoted to the language of the OTP are, their enumeration also brings home how much is lacking. There is no specialized dictionary of the OTP. And research on its vocabulary is still in an embryonic stage. 6 There is no grammar of the OTP. One may think that this is asking too much. But specialized dictionaries and grammars do exist for the Septuagint, as well as a host of extensive lexical and grammatical studies. If it weren’t for these works, Septuagint studies would be severely handicapped. This perceived lack in Peshitta studies, however grating, cannot, of course, be remedied within the space of a medium sized paper. Rather, what will be attempted is a brief typology of the language of the OTP.7 What kind of Syriac do we find in this version? How does it relate to other varieties of Syriac? Where does it stand 3 See P. J. Williams, Studies in the Syntax of the Peshitta of 1 Kings (PhD thesis, Cambridge, 1997); see the abstract in Tyndale Bulletin 49 (1998), 183–186. 4 See P. G. Borbone, J. Cook, K. D. Jenner, D. M. Walter, Concordance. 1. The Pentateuch (The Old Testament in Syriac 5; Leiden: Brill, 1997). 5 In the Göttinger Orientforschungen series, W. Strothmann has published concordances to Ecclesiastes (1973), Psalms (1976), Prophets (1984), Pentateuch (1986), Bet Mautbe (1995). 6 Among pioneering efforts are those of J. C. de Moor & F. Sepmeijer, “The Peshitta and the Targum of Joshua” in P. B. Dirksen and A. van der Kooij, eds., The Peshitta as a Translation (MPI 8; Leiden: Brill, 1995), 129–176; Y. Maori, The Peshitta Version of the Pentateuch and Early Jewish Exegesis (Jerusalem: Magnes Press, 1995), 227–245; M. P. Weitzman, “Lexical Clues to the Composition of the Old Testament Peshitta” in M. J. Geller et al., eds., Studia Aramaica: New Sources and New Approaches (JSStS 4, Oxford: University Press, 1995), 217–246. 7 As far as possible, the focus will be on language as opposed to translation technique. The question of the influence of the Hebrew Vorlage on the style of the Syriac translation falls outside the scope of this study.

A LINGUISTIC APPROACH

51

in the spectrum of Aramaic dialects? An answer to these questions, even a tentative one, should provide a framework within which to approach linguistic problems as they may arise in dealing with the OTP. Before launching into the presentation, two cautionary remarks need to be made. Firstly, it is of course a simplification to speak of the language of the Peshitta. Linguistic differences exist between the books of the Bible; some of these are due to differences in the technique of individual translators, but others do seem to indicate different stages of the language.8 The Peshitta was not translated in one sitting. Between the Pentateuch and, say, Proverbs as much as a century (not much more, it would seem) may have passed. Some books, moreover, notably Chronicles, have a particular linguistic profile.9 While all this is true, the linguistic diversity within the OTP should not be exaggerated. The language of the OTP may be considered to be a relative unity, particularly when it is viewed in comparison with other types of Syriac such as the language of the Old Syriac Gospels, or the language of early Syriac inscriptions. Second, the text-critical dimension should be taken into account. The text of the OTP, as several monographs and the Leiden edition have shown, is remarkably stable but not monolithic. Many of the variants, especially those occurring in such manuscripts as 5b1 and 9a1, are relevant to linguistic analysis. And the possibility remains that even the oldest manuscripts do not give us the original Peshitta text.10 In principle, the language of the manuscripts is not to be confused with the language of the original text. Again, although these remarks are entirely correct, they need to be taken See Weitzman, “Lexical Clues”. Another book the Syriac translation of which is quite idiosyncratic, linguistically speaking, is Ben Sira. The present study is limited to the translation of the canonical books, however. 10 This has been pointed out repeatedly by M. J. Koster. A very interesting possible source for primitive Peshitta readings not transmitted by the manuscripts are the Syros readings signaled by Eusebius of Emesa and other Church Fathers; see, now, R. B. ter Haar Romeny, A Syrian in Greek Dress. The Use of Greek, Hebrew and Syriac Biblical Texts in Eusebius of Emesa’s Commentary on Genesis (Traditio Exegetica Graeca 6; Leuven: Peeters, 1997). 8 9

52

LANGUAGE AND TEXTUAL HISTORY

with a pinch of salt. In practice, there are very few systematic changes of a linguistic nature between the earlier and the later manuscripts of the Peshitta. Attention to text-critical questions is apt to nuance the linguistic typology, not to overturn it.

THE OTP IS WRITTEN IN IDIOMATIC SYRIAC Any description of the language of the Peshitta has to start from the observation that the version is written in good, idiomatic Syriac. This is often taken for granted.11 After all, the Peshitta comes to us in Syriac characters, as part of Syriac literature, a heritage of the Syriac Church. Since, however, a few non-Syriac elements will be pointed out below, it is worthwhile to linger a little over the basic and far-reaching agreement of the language of the Peshitta with that of other Syriac texts.12 Syriac is one of the Aramaic dialects that first attained written expression during the first two centuries AD.13 With the breakdown of the Persian Empire, Imperial Aramaic had lost its official transnational status. And although Aramaic continued as a vehicular language for a long time (some say until the Islamic conquest),14 different local dialects ascended to official and literary use. The earliest Syriac inscriptions date from the first century CE.15 For See, e.g., Th. Nöldeke, Die alttestamentliche Literatur (Leipzig: Quant & Hänsel, 1868), 264. 12 For the other texts employed for comparative purposes, see the abbreviations above. 13 A fundamental study on the emergence of classical Syriac is L. Van Rompay, “Some Preliminary Remarks on the Origins of Classical Syriac as a Standard Language. The Syriac Version of Eusebius of Caesarea’s Ecclesiastical History” in G. Goldenberg, S. Raz, eds., Semitic and Cushitic Studies (Wiesbaden: Harrassowitz, 1994), 70–89 (with literature). Note also the earlier study of Beyer quoted below in n. 35. 14 See, e.g., D. Boyarin, “An Inquiry into the Formation of the Middle Aramaic Dialect” in Y. L. Arbeitman & A. R. Bomhard, eds., Bono Homini Donum: Essays in Historical Linguistics in Memory of J. Alexander Kerns, II (Amsterdam: J. Benjamins, 1981), 613–649, esp. 641. 15 A full study of these inscriptions will be found in H. J. W. Drijvers and J. F. Healey, The Old Syriac Inscriptions of Edessa and Osrhoene. Texts, Translations and Commentary (Leiden: Brill, 1999). Since this work was not 11

A LINGUISTIC APPROACH

53

roughly the same period, we have written attestations of Nabatean, Judean, Palmyrene, Edessene (Syriac) and Hatran Aramaic. In all these dialects, one observes a development from early texts in which the Imperial Aramaic heritage is still quite visible to later texts with increasing distinctive traits. Thus it has been pointed out that early Syriac shows significantly more affinity with early Western dialects than does later Syriac with Galilean Aramaic or with Christian Palestinian Aramaic.16 The emergence of Syriac as a literary language is, of course, a very complex process.17 What is important for us, is that the OTP fits squarely within the Syriac dialect as it is attested in the early period. The orthography of the OTP manuscripts conforms to the remarkably stable orthography of other Syriac texts. 18 The same may be said for the morphology: non-paradigmatic forms are extremely rare in the OTP (for some exceptions, see below). Admittedly, the orthography and at least some of the morphology may have been regularized, or ‘syriacized’, between the making of the version and the writing of the oldest manuscripts (dating from the fifth century). It is not impossible, for instance, that the prefix of the third person masculine singular was yudh when the Peshitta was first translated (i.e. ‫ܝܩܛܘܠ‬, as in the inscriptions well into the second century), and that it was later changed throughout to nun.19 The vocabulary and syntax of the OTP, too, are to a very great extent shared with other Syriac texts. 20 Since it is hard to imyet available to me while preparing the present study, the Syriac inscriptions are quoted according to the earlier collection in Drijvers (see n. 1). 16 See, e.g., Van Rompay, “Preliminary Remarks”. 17 One may suspect differences between the spoken language and the Imperial Aramaic heritage as maintained in the literary language; moreover, it is quite possible that several local spoken dialects played a role. See Drijvers and Healey, Old Syriac Inscriptions, 32–34. 18 This is not to deny that orthographic variations occur in the Peshitta manuscripts. For lists of such variations, see, e.g., M. D. Koster, The Peshitta of Exodus. The Development of its Text in the Course of Fifteen Centuries (Studia Semitica Neerlandica 19; Assen: Van Gorcum, 1977). 19 For the date of the Peshitta, see below in the next section. 20 Note that the reverse is not always true: many words and syntagms found in other Syriac texts are absent from the OTP (see below, e.g., on copular ‫)ܐܝܬ‬.

54

LANGUAGE AND TEXTUAL HISTORY

agine a systematic regularization of the vocabulary and the syntax of the version, these data may be more significant than the orthographical and morphological ones. As was stated above, too little systematic research has been done in these domains. But a few striking examples, encountered somewhat haphazardly in the course of sundry investigations, may show the strength of the evidence. Let us first look at some examples regarding the vocabulary: 1. A good example of peculiarly Syriac semantics is provided by the verbs ‫ ܫܠܚ‬and ‫ܫܕܪ‬, both to be translated ‘to send’. Whereas the first is used when the direct object is the message sent, the second is used when the direct object is an object or a person. This distinction is observed throughout the OTP, for example 2 Kgs 14:8: ...‫ ܘܐܡܪ ܬܐ‬...‫ ܐܝܙܓܕܐ ܠܘܬ ܝܗܘܐܫ‬...‫ܗܝܕܝܢ ܫܕܪ ܐܡܘܨܝܐ‬ Then Amaziah… sent messengers to Jehoash… saying: “Come…”; and 2 Kgs 14:9: ...‫ ܘܐܡܪ ܚܘܚܐ ܕܒܠܢܒܢ‬...‫ ܐܠܡܘܨܝܐ‬...‫ܘܫܠܚ ܝܗܘܐܫ‬ And Jehoash… sent word to Amazia: “A thistle on Lebanon…”21 The same distinction occurs in other early Syriac texts, 22 for example Lk 7:6 S: ‫ܗܐ ܫܪܪ ܠܘܬܗ ܩܢܛܪܘܢܐ ܗܘ ܠ̈ܪܚܡܘܗܝ ܘܫܠܚ ܠܗ ܡܪܝ‬ Behold the centurion sent his friends to him and he sent to him: “Lord, do not trouble yourself…” To be sure, the individual lexemes ‫ ܫܠܚ‬and ‫ ܫܕܪ‬are attested in other Aramaic dialects as well. But the allocation of semantic functions observed here is, I think, found only in Syriac.23 Other passages where both verbs occur: 1 Kgs 5:1; 2 Kgs 5:6–7; more examples may be found in the article quoted in the following note. 22 See J. Joosten, “Le Père envoie le Fils. La provenance occidentale d’une locution syriaque,” Revue de l’Histoire des Religions 214 (1997), 299– 309. 23 For a similar distinction in Imperial Aramaic, see M. L. Folmer, The Aramaic Language in the Achaemenid Period: A Study in Linguistic Variation (Leuven: Peeters, 1995), 652–657. 21

A LINGUISTIC APPROACH

55

2. Striking, too, are a number of specifically Syriac idioms: - ‫‘ ܐܙܠ ܠ‬to be fit for’, e.g. ‫“ ܐܠ ܐܙܐܠ ܗܘܬ ܠܥܒܕܐ‬it wasn’t fit for any work” Ezek. 15:5 (‫ ;)לא יעשה למלאכה‬see Gen. 25:32; 27:46; Num. 32:1, 4; Prov. 17:16; Job 30:2. In other Syriac writings: Mt 5:13 P ‫“ ܠܡܕܡ ܐܠ ܐܙܐܠ‬it is good for nothing”; see Lk 14:35 PCS; Acts of Thomas, 291, 4–5. - ‫“ ܐܝܟܢܐ ܫܡܟ‬what (lit. ‘how’) is your name” Gen. 32:28. In other Syriac writings: Ancient Documents, 77, l. 19; Mk 5:9 P. - ‫‘ ܚܫܩܒܘܠ‬together’ Isa. 65:7. In other Syriac writings: Drijvers 26, l. 4; Spicilegium 44, l. 4.24 - ‫“ ܡܚܐ ܪܝܚܐ‬an odour spread” Isa. 34:3; Joel 2:20; see Am. 4:10. In other Syriac writings: Acts of Thomas, 43, l. 24.25 - ‫“ ܡܐܠ ܒܠܒܗ‬he comforted him” Gen. 34:3 and very often in the OTP. In other Syriac writings: Jn 11:19; 1 Thess. 2:11.26 - ‫“ ܡܦܨܝ ܐܢܐ ܡܢ‬I am innocent with regard to” Josh. 2:19; 2 Sam. 14:9; Job 33:9. In other Syriac writings: Ancient Documents, 75, l. 18. - ‫“ ܐܠ ܬܬܩܛܥ ܠܟܘܢ‬don’t be disheartened” Hab. 2:3 (‫)חכה לו‬. In other Syriac writings: 1 Thess. 3:3; Heb. 6:12; 10:38, 39; Aphrahat, I, p. 176, l. 3.27 - ‫“ ܪܡܐ ܒܝܬ‬he sowed discord between” Prov. 6:19. In other Syriac writings: Aphrahat, I, p. 677, ll. 9–10. - ‫“ ܬܬܘܝܟ ܢܦܫܟ‬you will regret” Prov. 5:11. In other Syriac writings: Mt 21:29, 32 CS; 2 Cor. 7:8.28

See R. Köbert, “Heisst syr. ḥašqbol duplex?,” Biblica 67 (1986), 555–556; J. F. Healey, “Syriac ḥašqbol : A Further Note,” Biblica 68 (1987), 258. 25 See R. Payne-Smith, ed., Thesaurus Syriacus (Oxford: Clarendon Press, 1879–1901), col. 2065. 26 See Payne-Smith, Thesaurus, col. 2118. 27 See Payne-Smith, Thesaurus, col. 3583. 28 See Payne-Smith, Thesaurus, col. 4397. 24

56

LANGUAGE AND TEXTUAL HISTORY

3. In its own way, a study of Greek and Latin loanwords sheds light on the continuity between the OTP and other Syriac texts.29 Such loanwords are fairly prominent in the OTP, as they are in Syriac in general. More specifically, the Greek words occurring in the OTP are almost invariably found, with the same meaning, in other Syriac texts as well.30 Regarding syntax, the following phenomena seem to be both specifically Syriac and common to the OTP and other texts: 4. ‫‘ ܐܝܟܘ‬why’ In Syriac, the grapheme ‫ ܐܝܟܘ‬must usually be analysed as ‫ܐܝܟܐ‬ ‘where’ followed by the enclitic third masc. singular pronoun. In a number of instances, however, ‫ ܐܝܟܘ‬has to be taken to mean ‘why’; it is then always followed by a negative sentence. 31 This peculiar syntax is found in the OTP (1 Sam. 1:8): ‫ܠܡܢܐ ܒܟܝܐ ܐܢܬܝ ܘܐܝܟܘ ܐܠ ܠܥܣܬܝ ܘܠܡܢܐ ܒܝܫ ܠܒܟܝ‬ Why are you crying, and why don’t you eat, and why is your heart sad? The same syntax is found in the Old Syriac Gospels and in other early Syriac texts, for example Spicilegium p. 3, l. 4: ‫ܐܝܟܘ ܐܠ ܥܒܕܢ ܐܠܗܐ ܕܐܠ ܢܚܛܐ ܘܢܬܚܝܒ‬ Why did God not make us so that we would not sin and be guilty? 5. ‫ ܕܝܢ‬for irrealis 32 The particle ‫ܕܝܢ‬, reflecting Greek δέ and meaning ‘but, and’, is known by anyone who has studied Syriac. A less well-known hom-

29 See J. Joosten, “Greek and Latin Words in the Peshitta Pentateuch. First Soundings” in R. Lavenant, ed., VII Symposium Syriacum (OCA 256; Rome: Pontificio Istitùto Biblico, 1998), 37–47. 30 An exception must be made for some rare words which may have come to the Peshitta through a Targumic tradition, see Joosten, “Greek and Latin Words,” p. 46, n. 48. 31 See J. Joosten, “The Use of Some Particles in the Old Testament Peshitta,” Textus 14 (1988), 175–183. In this article the attestations of this particle in the Peshitta and in other texts are discussed. 32 Joosten, “Some Particles”.

A LINGUISTIC APPROACH

57

onym is found a number of times in the OTP, for example Job 10:18: ‫ܠܡܢܐ ܡܢ ܡܪܒܥܐ ܐܦܩܬܢܝ ܣܦܬ ܕܝܢ ܘܥܝܢ ܐܠ ܚܙܬܢܝ‬ Why did you bring me forth from the womb? Would that I had died and no eye had seen me. In this clause ‫ ܕܝܢ‬is not the equivalent of Greek δέ but a genuinely Syriac development of Aramaic ‫‘ ֱא ַַדיִן‬then, thereupon’. While this Semitic ‫ ܕܝܢ‬lives on in Syriac expressions like ‫‘ ܐܫܬܘܦ ܕܝܢ‬would that’, its independent use, as in the example, is not found after the early period, for example Aphrahat, I, p. 108, ll. 20–21: ‫ܕܚܠܬܝ ܕܝܢ ܡܢ ܗܕܐ ܐܘܢ ܐܝܙܒܠ ܘܗܝܕܝܢ ܛܢܬܝ ܛܢܢܐ‬ Would that you had feared this, o Jezebel, and then had been zealous. 6. The negation of the non-verbal clause33 Non-verbal clauses in the OTP are negated with ‫ ܐܠ‬or with ‫ܗܘܐ‬ ‫ܐܠ‬. When ‫ ܐܠ ܗܘܐ‬is used, it is conjugated and accords with the subject, for example Ex. 4:10: ‫ܐܠ ܗܘܝܬ ܓܒܪܐ ܕܡܡܠܐܠ‬, “I am not a man of speech”. In this type of clause (of which about 80 examples exist in the OTP), the perfect of the verb ‫ ܗܘܐ‬expresses neither past tense nor becoming. Exactly the same syntax is found in the OS and Peshitta Gospels, in the Acts of Thomas and in other early Syriac texts, for example Jn 16:32 P: ‫ܐܠ ܗܘܝܬܝ ܒܠܚܘܕܝ ܕܐܒܝ ܥܡܝ ܗܘ‬ I am not alone for my Father is with me. Other syntactic phenomena could be listed.34 These few examples must suffice to show that the Syriac-ness of the Peshitta is not merely a matter of script and orthography: the language of the OTP is Syriac to the bone. This conclusion may not be very earth shattering. But sometimes it is necessary to state the obvious. In what is probably the 33 See J. Joosten, “The Negation of the Non-verbal Clause in Early Syriac,” JAOS 112 (1992), 584–588. 34 Such as: the construction of prepositions; the syntax of ‫ ;ܟܠ‬the functions of the infinitive absolute (particularly where it is absent in the Hebrew).

58

LANGUAGE AND TEXTUAL HISTORY

most recent previous effort to categorize the OTP’s language, K. Beyer presented the OTP as a “syriacized” revision of an Official Aramaic Targum.35 In my view a more penetrating linguistic analysis makes this theory — which has also been discredited on other grounds36 — unlikely. Our conclusion does need to be nuanced, however. The fact that the Peshitta is written in idiomatic Syriac should not cause us to lose sight of the distinctive character of its language. Linguistic continuity with other Syriac texts may run deep, but there are many differences as well. The single most important factor here is the early date of the OTP. The Syriac of the version is early Syriac — earlier, in fact, than all the other literary texts preserved in this language (although some inscriptions precede the OTP). So let us turn now to this aspect of our subject.

THE OTP IS WRITTEN IN EARLY SYRIAC There seems to be a growing consensus among Peshitta scholars that the OTP must date, in its larger part, from before 150 CE or so.37 A decisive consideration is the strong likelihood that the OTP was quoted in the Diatessaron. 38 If Tatian, who produced his Diatessaron around 170 CE, replaced the Old Testament quotations of the Gospels with their equivalent in the Peshitta he must have done so under the pressure of a text well-known and well-recognized by the community to whom his work was addressed. 35 K. Beyer, “Der reichsaramäische Einschlag in den ältesten syrischen Literatur,” ZDMG 116 (1966), 242–254, esp. 252–253. 36 See the case studies on this question in P. V. M. Flesher, ed., Targum and Peshitta (Targum Studies 2; Atlanta: Scholars Press, 1998). 37 See the review of the recent debate in J. Joosten, “La Peshitta de l’Ancien Testament dans la recherche récente,” Revue d’Histoire et de Philosophie Religieuses 76 (1996), 385–395, esp. 389–392. See now M. P. Weitzman, The Syriac Version of the Old Testament. An Introduction (Cambridge: University Press, 1999); Michael Weitzman’s long awaited book appeared when the present paper was in the latest stage of editing. 38 See J. Joosten, “The Old Testament Quotations in the Old Syriac and Peshitta Gospels,” Textus 15 (1990), 55–76, where earlier studies by S. P. Brock and W. L. Petersen are noted.

A LINGUISTIC APPROACH

59

Linguistic analysis confirms this early dating of the OTP. Let us again review some eclectic, but representative, examples. From the lexical domain, the following items appear to be relevant: 1. For ‘city’ the OTP uses the old Aramaic word ‫ܩܪܝܬܐ‬.39 Unless I err, this usage is not to be found in any other Syriac text. Especially revealing is the allusion in Mt 5:35 to Ps. 48:3: ‫ܩܪܝܬܐ ܗܝ‬ ‫ܕܡܠܟܐ ܪܒܐ‬, “It is the city of the Great King”. In the Old Syriac (Cureton and Sinaiticus) and in the Peshitta Gospels this phrase turns up as: Mt 5:35 PCS, ‫ܡܕܝܢܬܐ ܗܝ ܕܡܠܟܐ ܪܒܐ‬. Elsewhere in quotations and allusions to the OT, the NT versions follow the OTP.40 In this passage, however, ‫ ܩܪܝܬܐ‬was not considered to be the mot juste. Indeed, in Syriac texts from the third century onward, ‫ ܩܪܝܬܐ‬has the meaning ‘village’, or even ‘field’.41 The semantic development which resulted in the substitution of ‫ ܡܕܝܢܬܐ‬for ‫ ܩܪܝܬܐ‬with the meaning ‘city’ can already be traced to the OTP itself. Whereas the Peshitta Pentateuch uses ‫ܩܪܝܬܐ‬ exclusively, the historical books, prophets and Psalms use both it and ‫ܡܕܝܢܬܐ‬, while Proverbs and some other books which appear to have been translated relatively late have only the latter.42 2. For ‘to love’ and ‘love’ the OTP almost invariably uses the words ‫ ܪܚܡ‬and ‫ܪܚܡܬܐ‬. In later Syriac texts these words are used side by side with the words ‫ ܐܚܒ‬and ‫ܚܘܒܐ‬. Interestingly, attestations of ‫‘ ܚܘܒܐ‬love’ do turn up in Prov. 15:17 and Cant. 1:3, that

See Weitzman, “Lexical Clues,” 218–221. See Joosten, “Old Testament Quotations” (where, however, allusions are not dealt with). 41 In a document dated from the year 240 CE, the opposition between ‫‘ ܡܕܝܢܬܐ‬city’ and ‫‘ ܩܪܝܬܐ‬village’ is already well established, see J. Teixidor, “Deux documents syriaques du IIIe siècle après J.-C., provenant du Moyen Euphrate”, CRAI 1990, 144–163, esp. 147 (document A, scriptura exterior: line 5 ’RHY B’RS MDYNT’ ‘Orhai, la ville place-forte’ versus line 8 MYHRW QRYT’ ‘le village de Mihru’). 42 In Peshitta Proverbs, ‫ ܩܪܝܬܐ‬is not even used to render the Hebrew cognate ‫קריה‬, see Weitzman, “Lexical Clues,” 219. 39 40

60

LANGUAGE AND TEXTUAL HISTORY

is, in books that on other grounds appear to have been translated relatively late into Syriac.43 3. The study of Greek words too bears out the archaic character of the Peshitta’s language.44 The Peshitta Pentateuch contains a total of about 50 words of Greek or Latin origin (some of which may have crept in later into the manuscript tradition). This is a rather low number. The Old Syriac Gospels, a much shorter text, contain 85 such words according to the studies of S. P. Brock.45 Although several factors need to be taken into account, the disparity almost certainly indicates a diachronic development, from an earlier stage relatively untainted by Greek influence to a later stage where such influence became much stronger. More concretely, several well-known Syriac words of Greek origin, such as ‫‘ ܬܪܘܢܘܣ‬throne’ and ‫‘ ܠܣܛܐ‬thief’, seem to have been unknown to the translators of the OT. Secondly, a few morphological archaisms may be pointed to. As was stated above, these are quite exceptional: 4. A very interesting example pointed out by M. Weitzman is the occurrence of the demonstrative pronoun ‫ ܗܠܘܟ‬in 1 Chron. 9:1 (9a1 only) and Esth. 1:5 (10f1 only).46 This form occurs in the OS Gospels and in some other early Syriac texts,47 while it is absent from later texts. The fact that ‫ ܗܠܘܟ‬is found in single manuscripts only suggests that scribes tended ‘to normalize’ or ‘modernize’ such forms, substituting the regular form ‫ܗܢܘܢ‬. This would explain,

43 The Pael ‫ ܚܒܒ‬and the adjective ‫ ܚܒܝܒ‬do occur in the OTP, however. For ‫ ܚܘܒܐ‬in Judg. 5:14, see Weitzman, Syriac Version, 294. 44 See Joosten, “Greek and Latin Words”. 45 See S. P. Brock, “Greek words in the Syriac Gospels (vet and pe),” Le Muséon 80 (1967), 389–426; see also S. P. Brock, “Greek Words in Syriac: Some General Features,” Scripta Classica Israelica 15 (1996), 251–262, esp. 252, n. 7. 46 See Weitzman, Syriac Version, 257. 47 See Th. Nöldeke, Kurzgefasste syrische Grammatik (reprint Darmstadt: Wissenschaftliche Buchgesellschaft, 1977 [1898]), p. 46, § 67, n. 2; F. C. Burkitt, Evangelion da-Mepharreshe, vol. II (Cambridge: University Press, 1904), 42; Van Rompay, “Preliminary Remarks,” 76.

A LINGUISTIC APPROACH

61

more generally, why non-paradigmatic forms are so rare in the OTP. 5. The preposition ‫ ܥܠܘܝ‬occurs in Exod. 39:21; 40:36; Job 29:3, 4, seemingly as a variant of ‫‘ ܥܠ‬on, upon’. The word appears not to be found elsewhere in Syriac. It is well known from official Aramaic, and is found in Targumic Aramaic as well.48 6. K. Beyer has pointed out the archaic u- perfects ‫‘ ܐܟܘܡ‬to become black’ in Nah. 2:11 and ‫‘ ܩܦܘܕ‬to bristle’ in Job 7:5; 30:30 and Lam. 4:8.49 These seem to become obsolete in later Syriac, as is illustrated by a comment of Ephrem’s on Job 30:30: ‫ܗܘ… ܩܦܘܕ‬ …‫ܩܦܕ‬.50 Finally, a few syntactic features may be mentioned. 7. The use of the particle ‫‘ ܐܝܬ‬there is’ with suffixes as a quasi-verbal copula is well known,51 for example: Mt 6:22 P: ‫ܫܪܓܐ‬ ‫ܓܝܪ ܕܦܓܪܐ ܐܝܬܝܗ ܥܝܢܐ‬, “For the eye is the lamp of the body”. This type of syntax is widespread; it would be hard to name a Syriac text of any length where it does not occur.52 But the spread of occurrences suggests that the copular function of ‫ ܐܝܬ‬was a late development in Syriac, probably owing something to the influence of Greek ἐστιν.53 Now in the OTP this syntactic device is extremely rare. In the entire Pentateuch only two cases are found (Gen. 28:17; Deut. 1:10); there is one case in Judges (7:14), one in 1 Samuel (23:23), See A. Tal, The Language of the Targum of the Former Prophets and its Position within the Aramaic Dialects (Tel Aviv: University Press, 1975), p. 22, n. 37. 49 See Beyer, “Reichsaramäische Einschlag,” 252. 50 See Payne-Smith, Thesaurus, col. 3687. 51 See Th. Nöldeke, Syrische Grammatik, 232–233; T. Muraoka, “On the Syriac Particle ’it,” BiOr 34 (1977), 21–22. 52 E.g. in the Sinaitic Gospel of Mark there are 6 occurrences: Mk 8:27,29; 12:11, 25, 42; 15:16. Note, however, that the copula may have been absent from the original Old Syriac version of Matthew, as shown in J. Joosten, The Syriac Language of the Peshitta and Old Syriac Versions of Matthew (Leiden: Brill, 1996), 150–151. 53 See T. Muraoka, Classical Syriac for Hebraists (Wiesbaden: Harrassowitz, 1987), 66; Joosten, Syriac Language, 107. 48

62

LANGUAGE AND TEXTUAL HISTORY

two in Chronicles (1 Chron. 2:21; 2 Chron. 29:1 mss), two in Psalms (102:28; 139:8), one in Proverbs (8:2), one in Job (22:30), one in Daniel (2:39), one in Song of Songs (8:10) and two in Ecclesiastes (7:26; 8:13). Fourteen cases in the whole OTP (presuming the above list to be complete) is a low number indeed. It may be interpreted to mean that the copular use of ‫ ܐܝܬ‬was not yet well established in Syriac when the OT was translated. This strengthens the case for a date before the 3rd century.54 8. The periphrastic tense ‫ ܗܘܐ ܩܛܠ‬usually expresses a wish, advice or obligation of general applicability. This use is attested, of course, in the OTP. 55 But at the same time, a number of cases occur where this verbal syntagm expresses a different function, namely iterative action in the past,56 for example 1 Kgs 18:26: ‫ܘܗܘܘ ܩ̈ܪܝܢ ܒܫܡܗ ܕܒܥܐܠ ܡܢ ܨܦܪܐ ܘܥܕܡܐ ܠܛܗܪܐ‬ And they kept calling the name of Baal from morning till afternoon. Some of the examples of iterative ‫ ܗܘܐ ܩܛܠ‬correspond to periphrastic ‫ ָהיָה ק ֵטל‬in Hebrew (22 cases), thus suggesting a Hebraism. Other cases, however, do not correspond to a periphrastic tense (9 cases). In 1 Kgs 18:26, for instance, the Hebrew simply has ‫וַ יִ ְק ְראוּ‬. This indicates that the usage is not a Hebrew calque. And indeed, other early Syriac texts not translated from Hebrew provide a number of examples of iterative ‫ܗܘܐ ܩܛܠ‬. The usage is found in Ahiqar, Aphrahat, Apocalypse of Baruch, Odes of Solomon, the translation of Eusebius’ Church History and Judith.57 Note the following instance (Judith 15:11):

54 In the early Syriac inscriptions three cases occur, two of them in fragmentary contexts, see Drijvers, Old Syriac (Edessean) Inscriptions, 39,7; 55,1.12. These inscriptions are dated to the third century or later by Drijvers. 55 See, e.g., 1 Kgs 17:4; 22:25. 56 This usage is not recognised by the standard grammars of Syriac. See J. Joosten, “Biblical Hebrew weqatal and Syriac hwā qātel expressing repetition in the past,” Zeitschrift für Althebraistik 5 (1992), 1–14, esp. 9–12. 57 With the exception of the cases in Judith the occurrences are listed in the article quoted in the previous note.

A LINGUISTIC APPROACH

63

‫ܘܗܘܘ ܡܒܥ̈ܪܝܢ ܥܡܐ ܡܫܪܝܬܗܘܢ ܕܐܬܘ̈ܪܝܐ ܥܕܡܐ ܠܬܠܬܐ ܝܘܡܝܢ‬ καὶ ἐλαφύρευσεν πᾶς ὁ λαὸς τὴν παρεμβολὴν ἐφ᾽ ἡμέρας τριάκοντα

It seems that the usage is genuinely Syriac, albeit limited to old texts. The extreme scarcity of instances — 32 in the OTP, but only 14 outside of the OTP — may perhaps be explained by saying that the iterative use of the tense was dying out in early Syriac, with the OTP and the other writings mentioned representing successive stages in the process. The history of the Syriac language is largely uncharted. However, common sense suggests, and preliminary studies demonstrate,58 that Syriac, which was in active literary use for about a millennium (taking a narrow view), did develop over the centuries. The ever-growing influence of Greek is one important factor affecting the history of the Syriac language, but internal developments also played their part. The OTP stands at the very beginning of this development. Orthography and morphology may have been “modernized” to a certain extent by later scribes; but the vocabulary and the syntax seem to be largely representative of an early stage of the language. In light of what has been said above on the emergence of Syriac as a written language, two corollaries may be stated. An early date implies continuity with Imperial Aramaic (and Old Aramaic), and it implies relative proximity to contemporary Western Aramaic. Both circumstances are easily verified by means of the eight features listed above. It is to be expected that for other features, too, the closest analogues of the OTP will be found not in other Syriac writings, but in earlier Aramaic or Western Aramaic texts.

58 See, e.g., S. P. Brock, “Diachronic Aspects of Syriac Word Formation: An Aid for Dating Anonymous Texts” in R. Lavenant, ed., V Symposium Syriacum 1988 (OCA 236; Rome: Pontificio Istitùto Biblico, 1990), 321–330; Joosten, Syriac Language, 143–152; Joosten, “Negation of the Non-verbal Clause”.

64

LANGUAGE AND TEXTUAL HISTORY

NON-SYRIAC ELEMENTS IN THE OTP On the basis of what has been said up till now, one might conclude that the OTP is written in idiomatic early Syriac. This conclusion is, I think, essentially correct but it still needs nuancing. There is a foreign element in the OTP’s language, at least in as far as its vocabulary is concerned.59 Even if the Peshitta was never based on a running Aramaic targum, no one will deny that the Syriac version stands partly within the same exegetical tradition as the Jewish Targums. The influence of early Jewish interpretation is keenly felt in most of the books of the Syriac Bible. How this influence was exerted is of course another question. What is relevant to us is that the dependence on Jewish tradition sometimes induced the use of lexemes or expressions that were not originally Syriac, although some of them subsequently became so because of their occurrence in the Syriac Bible.60 It is in the nature of things that non-Syriac words borrowed in the OTP are usually of Western Aramaic origin.61 Because of the likelihood that the Syriac of the OTP stood quite close to contemporary Western Aramaic,62 it is usually extremely difficult to distinguish such Palestinian Aramaic terms from genuine, early Syriac ones.63 I will simply list a number of items for which the claim of Western Aramaic provenance can be made while realizing that

59 To be excluded here are the occasional transcriptions of Hebrew words, such as ‫ ܐܕܘܢܝ‬in 2 Chron. 20:17, ‫ ܐܠܘܗܝܡ‬in Gen. 6:1, and even ‫ ܐܗܝܗ ܐܫܪ ܐܗܝܗ‬in Exod. 3:14. 60 See Y. Maori, The Peshitta Version of the Pentateuch and Early Jewish Exegesis (Jerusalem: Magnes Press, 1995), 227–245. 61 In the case of words that are of Hebrew origin one will have to suppose that they came to the OTP via West Aramaic. The profound influence of Hebrew on the West Aramaic dialects is well known (see, e.g., Tal, Language of the Targum). 62 See above. 63 See the debate between Van Rompay and the present writer on West Aramaic elements in the Syriac gospel, see Van Rompay, “Preliminary Remarks,” 81.

A LINGUISTIC APPROACH

65

some of them could be merely archaic. The list is not intended to be exhaustive:64 1. ‫ܐܒܐ‬, ‘my father’ Gen. 22:7. For ‘my father’ and vocative ‘father’, Syriac usually has ‫ܐܒܝ‬. The form found in this verse corresponds to early Palestinian Aramaic (see Mk 14:36) and Mishnaic Hebrew ‫ ;אבא‬its use in the Peshitta of Gen. 22 may go back to an exegetical tradition.65 2. ‫ܐܘܪܝܬܐ‬, ‘the law (of Moses)’ is found a few times in 2 Kgs, and is regular in Chronicles for Hebrew ‫תורה‬. The specific function of the word indicates that it derives from Jewish Aramaic ‫אוריתא‬. 3. ‫ܐܪܘܢܐ‬, ‘the ark’ is a regular equivalent of Hebrew ‫( ארון‬together with ‫)ܩܐܒܘܬܐ‬. Although the word derives from Hebrew and translates the Hebrew etymon it may actually have penetrated into the terminology of the OTP through the intermediary of an Aramaic exegetical tradition (note that ‫ ארונא‬is regular too in the Targums).66 4. ‫ܓܝܘܪܐ‬, ‘proselyte’, 2 Sam. 1:13; Job 28:4; 1 Chron. 22:2; 2 Chron. 2:16; 9:15; 30:25.67 It is hard to escape the impression that this word goes back to Jewish Aramaic ‫‘ גיורא‬proselyte’. Formally the words are identical; moreover, the corresponding root carries a very different meaning in Syriac. 5. ‫ܚܣܕܐ‬, ‘grace’, found about 20 times in the OTP, sometimes only in one manuscript (9a1 in Kings) while other manuscripts have ‫ܛܝܒܘܬܐ‬.68 The word usually translates Hebrew ‫חסד‬, 64 Not included are the many striking cases in the Syriac version of Ben Sira (e.g. ‫‘ ܐܒܥ‬to do quickly’ in BSir. 36:7). 65 See J. Joosten, “West Aramaic Elements in the Old Syriac and Peshitta Gospels,” JBL 110 (1991), 271–289, esp. 274, n. 15. 66 According to Weitzmann (“Lexical Clues,” 225) the word is a calque from Hebrew. 67 For the distribution of the word, see Weitzmann, “Lexical Clues,” 227–229. Weitzmann signals one more occurrence where the word corresponds to Hebrew ‫נתינים‬, 1 Chron. 9:2, and two occurrences in Aphrahat’s biblical quotations. 68 For the distribution of the word, see Weitzmann, “Lexical Clues,” 224.

66

LANGUAGE AND TEXTUAL HISTORY

but in some instances corresponds to Hebrew ‫( חן‬Gen. 39:21; Esth. 2:15, 17; see Ruth 1:9). The word is not attested in Old or Imperial Aramaic, hence one may suspect it to be a Hebrew loanword in Palestinian Aramaic whence it penetrated into the OTP. 6. ‫ܚܣܝܕܐ‬, ‘faithful (?)’, Job 37:13. If ‫ ܚܣܕܐ‬derives from Western Aramaic, then ‫ ܚܣܝܕܐ‬must do so too. The word is attested a few times in other Syriac texts, however.69 7. ‫ܝܬ‬, The ‘nota accusativi’ occurs a number of times in the OTP, but is not really found in any other text. 70 It might derive from a Targumic tradition, but could just as well be considered an archaism.71 8. ‫ܡܡܘܢܐ‬, ‘money’, Ex. 21:30. Although attested in the Syriac versions of the New Testament (Mt 6:24; Lk 16:9, 11, 13), the word is almost certainly a loan from Jewish Aramaic ‫ממון‬, ‘money’. 9. ‫ܢܚܡ‬, ‘to comfort’, Job 42:6. In Syriac this verbs usually means ‘to resurrect’. The meaning ‘to comfort’ is attested in one other place, in the Syriac translation of Ben Sira 48:25. If this is a Hebrew loan in Western Aramaic, as I suspect, then its occurrence in Syriac must be due to borrowing. 10. ‫ܢܩܦ‬, Aphel ‘to surround’, Deut. 32:10.72 As was shown by Maori, the rendering of as ‫ ַיְס ְב ֶּבנְ הוּ‬by ‫ ܐܩܦܗ‬is probably due to Midrashic influence. 11. ‫ܫܟܠܠ‬, ‘to lay the foundation’, 1 Kgs 5:31; 6:37 (9a1 only); 16:34; Isa. 44:28; Job 4:19; Ezra 3:11. In Syriac, this verb means ‘to complete (esp. a building)’ and ‘to adorn’. In the OTP it is used in this sense, but also as an equivalent of Hebrew ‫‘ יסד‬to lay the foundation’; the latter meaning is found in Targumic Aramaic.73 Even although a few more examples could probably be found, the West Aramaic elements in the OTP could never be used as an See Wis. 10:15; Acts of Thomas, p. 275, l. 11. See Nöldeke, Syrische Grammatik, p. 217, n. 1. 71 See F. Rosenthal, Die aramaistische Forschung seit Th. Nöldeke’s Veröffentlichungen (Leiden: Brill, 1939), 201. 72 See Maori, Peshitta Version, 227. But see also Weitzman, Syriac Version, 154–155. 73 See, e.g., Targ. Onkelos and Neof. Exod. 9:18; Targ. Jonathan 1 Kgs 6:37. See Weitzman, Syriac Version, 277–278. 69 70

A LINGUISTIC APPROACH

67

argument for the Peshitta-as-Targum hypothesis. There is a mere sprinkling of cases. These can be sufficiently explained by admitting that some Jewish Palestinian expressions were handed on to the translators by an existing interpretative tradition.

CONCLUSION The reflections proposed in this paper are necessarily tentative. Only when the vocabulary and the grammar of the OTP will have been analysed in more depth will it be possible to categorize its language with precision. On the basis of research at my disposal, I have argued that the Syriac of the Peshitta may be defined by three linguistic relations. The most important connection is that with other varieties of Syriac; the OTP manifests a high degree of continuity with other Syriac texts, particularly with early texts. Secondly, the Syriac of the Peshitta, being early Syriac, stands relatively close to earlier Aramaic dialects, in particular to Imperial Aramaic. Thirdly, there is a small but non-negligible Western Aramaic element due to the borrowing of specific terms. What should be clear is that the call for further research into the OTP’s language is not spurious. Looking up a word or an expression in the standard lexica of Payne-Smith or Brockelmann does not suffice when approaching the OTP; and neither Nöldeke nor Duval cover the entire syntax. The differences between the Syriac of the OTP and the Syriac of other texts may not be as important as those between Septuagint Greek and Classical Greek,74 but they certainly justify setting the Peshitta apart for linguistic study. Such research is needed if we wish to obtain an adequate understanding of the OTP as a Syriac text.

74 Of course, these differences are of an entirely different nature as well, since Septuagint Greek represents a relatively late stage of the language.

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS The Peshitta version of the book of Proverbs has not often been in the center of scholarly interests. It has, however, been subjected to a very thorough and dependable study, published in the last century, by Hermann Pinkuss.1 In his introduction,2 Pinkuss signals the existence of a number of doublet translations, and he extensively discusses every single case in his running textual commentary, comparing them with the Masoretic Text, the LXX and the Targum. The research presented in the present paper links up with Pinkuss’ results and attempts to extend our comprehension of the text of Peshitta Proverbs as it lies before us. The conclusion this research leads to may be condensed into the following thesis: doublet translations do not indicate conflation of two originally distinct translations,3 nor extensive editing of an “Old Syriac” version of Proverbs; rather, they are typical of the working method of the author — possibly of the group of authors — who produced the version roughly as we know it today. Doublet translations, or doublets, are defined as parallel renderings corresponding to one and the same element in the Maso-

H. Pinkuss, “Die syrische Übersetzung der Proverbien,” ZAW 14 (1894), 65–141, 161–222. His study includes a discussion of the earlier literature. 2 Pinkuss, “Die syrische Übersetzung,” 106. 3 The possibility of doublet translations pointing to multiple translations in the Peshitta book of Job has recently been raised by H. M. Szpek, Translation Technique in the Peshitta to Job (SBL.DS 137; Atlanta: GA, 1992), 155, 160. 1

71

72

LANGUAGE AND TEXTUAL HISTORY

retic Text.4 In the present paper, doublet translations of single words or expressions will be excluded;5 only doublets that extend to verses or half-verses will be discussed. Seven such doublet translations occur in Peshitta Proverbs.

THE PESHITTA VERSION OF PROVERBS In order correctly to evaluate the double renderings of some verses, it is necessary first to consider a few general characteristics of Peshitta Proverbs. Like the Peshitta in general, it is abundantly clear that the author of Peshitta Proverbs made use of a Hebrew text. An indisputable testimony to this fact are the instances where the Hebrew was translated erroneously. In 19:17, the phrase ‫“ ַמ ְלוֵַה ַיהוה‬one who loans to the LORD” is rendered ‫“ ܕܡܬܠܘܐ ܠܡܪܝܐ‬he who joins himself to the Lord” — against the LXX version which renders δανίζει θεῷ. This and other examples show that the author of Peshitta Proverbs was working on a Hebrew text.6 This Hebrew text seems to have been very similar to the MT, though not identical with it as will be seen below in the discussion of 14:9 and 14:32. In addition to the Hebrew, however, the author also made use of the LXX text of Proverbs. This can be demonstrated from a) the fact that the Peshitta includes many of the additional verses of the LXX,7 b) the numerous instances where the Peshitta has followed an erroneous, and sometimes nonsensical, interpretation of the LXX,8 and c) from a few syntactical grecisms in the Peshitta text, 4 The identification of doublets implies some subjectivity; in some cases it is possible to claim that one of the apparently parallel translations is rather a rendering of an addition in the Greek. For this reason I have not included the apparent doublet in 22:10b. 5 For some examples, see Pinkuss, “Die syrische Übersetzung,” 114; add 8:11. 6 Other examples: 24:28; 25:22; 26:10. 7 For a list of additions in the LXX contained in the Peshitta, see Pinkuss, “Die syrische Übersetzung,” 108. 8 In a paper read at the VI Symposium Syriacum, held in Cambridge in August 1992, M. P. Weitzman has pointed out that one may not postulate a common tradition issuing in the LXX and the Peshitta unless a reading

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

73

such as ‫ ܡܢ ܐܬܪ ܕ‬in 22:27, corresponding to the LXX’s πόθεν.9 Influence of the Greek version is quite pervasive in Peshitta Proverbs, more so — if I may give my opinion on this matter — than in any of the other canonical books of the OT. One sometimes gets the impression that the author of Peshitta Proverbs accorded as much importance to the LXX version as to the Hebrew. This point is illustrated by a curious phenomenon. In several instances, the Peshitta rendering corresponds neither to the MT nor to the LXX, but looks like a free combination of the two. Thus, 18:8a ַ‫ִדּ ְב ֵריַנִ ְרגָ ן‬ ‫ ְכ ִמ ְת ַלה ִמים‬is rendered ‫“ ̈ܡܠܘܗܝ ܕܚܒܢܢܐ ̈ܪܡܝܢ ܠܗ ܒܒܝܫܬܐ‬the words of the slothful put him in a bad spot.” 10 At first sight, this may look like a very free rendering of MT. This is not so, however: the words ‫“( ܚܒܢܢܐ‬the slothful”) and ‫“( ̈ܪܡܝܢ‬they throw”) point to influence from the LXX’s ὀκνηροὺς καταβάλλει φόβος (“fear casts down the slothful”). According to Lagarde,11 this LXX rendering has nothing to do with the Hebrew of 18:8a, but corresponds to the Hebrew of 19:15 ‫מה‬ ַָ ‫ילַתּ ְר ֵדּ‬ ַַ ‫הַתּ ִַפ‬ ַ ‫עַַ ְצ ָל‬. By combining elements from the Hebrew and the Greek, the author of the Peshitta ended up with a text that is a translation of neither, a versional patchwork.12 Extensive dependence on the LXX is probably connected to the particular difficulty of the Hebrew text of Proverbs. At the same time, however, it constitutes an index of the relative lateness of Peshitta Proverbs compared to the Peshitta of the other books of the OT, where influence of the LXX is much less in evidence.13 a) is not attested in other sources and b) does not intrinsically make sense. Compare Pinkuss, “Die syrische Übersetzung,” 103. Examples of places where the Peshitta has followed a bad rendering in the LXX: 11:15; 17:9b; 23:34b; 17:12. 9 Similar examples: 11:2; 21:1; 23:30; 26:20. 10 For the renderings of Hebrew ‫נִ ְרגָ ן‬, see Pinkuss, “Die syrische Übersetzung,” 184. 11 P. de Lagarde, Anmerkungen zur griechischen Übersetzung der Proverbien (Leipzig, 1863), 58. 12 Other examples 9:6; 9:13; 27:16, and see Pinkuss, “Die syrische Übersetzung,” 104. 13 An item of vocabulary indicating relative lateness is the word ‫“( ܚܘܒܐ‬love”) in 15:17; in all the OT Peshitta, the Hebrew ‫ ַאה ָבה‬is ren-

74

LANGUAGE AND TEXTUAL HISTORY

The instances of influence from the Peshitta of other books are an additional indication that Proverbs was one of the last books of the OT to be translated into Syriac. Thus in 23:32, ‫יַיַפ ִרש‬ ְ ִַ‫וּכ ִצ ְפענ‬ ְ “and it stings like an adder (RSV)” is rendered ‫“ ܘܐܝܟ ܚܪܡܢܐ ܕܦܪܚ‬and as a flying basilisk”, taking up the Peshitta rendering of ‫ףַמעֹופֵַף‬ ְ ‫ ָש ַָר‬in Isa. 14:29; 30:6. I have collected five examples of this phenomenon, and other examples could probably be found.14 Finally, it is enlightening to consider the textual relationship between Peshitta Proverbs and the Syriac NT. Seven verses from Proverbs are quoted in the NT. 15 However, whereas the OT quotes in the Syriac versions of the NT usually follow the OT Peshitta text, this is not the case for the quotes from Proverbs. On the contrary, very few points of contact exist between the text of the quotes in the NT and the text of Peshitta Proverbs. In one case only does the text of an entire verse quoted in the NT correspond to the text of Peshitta Proverbs, namely 1 Peter 4:18 quoting Prov. 11:31. In this case, however, it is possible to prove that the text is original in the version of the NT and borrowed in the OT Peshitta. Indeed, the use of the root ‫ ܚܝܐ‬to render words meaning “to be saved” is totally uncharacteristic of the OT, but standard in the NT.16 Whereas the Syriac versions of the NT do not give indidered by ‫( ܪܚܡܬܐ‬but see Cant. 1:4); in the Syriac NT the word ‫ ܚܘܒܐ‬is common. 14 In 2:7 and 30:5 Hebrewַ‫ ָמגֵ ן‬is rendered by forms of the verb ‫ܣܝܥ‬ as is usual in other books of the Peshitta. In 24:5, the expression ‫ܓܢܒܪ‬ ‫ ܚܝܐܠ‬seems to be dependent on the Peshitta of Judg. 6:12 and the like. The interpretation of 30:10 seems to be influenced by Deut. 23:16 (compare, however, the LXX of 30:10). The wording of 7:11 seems to echo that of Deut. 21:20, see Weitzman, “From Judaism to Christianity: The Syriac Version of the Hebrew Bible,” in J. Lieu et al., eds., The Jews among Pagans and Christians in the Roman Empire (London & New York, 1992), 163. 15 Prov. 3:11f quoted in Heb. 12:5f; 3:34 in Jam. 4:6; 1 Pet. 5:5; 10:12 in Jam. 5:20; 1 Pet. 4:8; 11:31 in 1 Pet. 4:18; 25:21f in Rom. 12:20. 16 The noun ‫ܡܚܝܢܐ‬, “saviour”, occurs in Isa. 60:16 in the manuscript 9a1. For discussion, see M. P. Weitzman, “The Originality of Unique Readings in Peshitta MS 9a1,” in P. B. Dirksen and M. J. Mulder, eds., The Peshitta: Its Early Text and History (Leiden, 1988), 225–255, esp. 236.

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

75

cation of having used Peshitta Proverbs, the author of Peshitta Proverbs apparently did know the Syriac NT.17 We may perhaps conclude, therefore, that Peshitta Proverbs is later than the oldest Syriac translation of the NT.18 On the other hand, however, the fact that Peshitta Proverbs is quoted by Aphrahat warns us not to date it very much later than the other books of the OT Peshitta.19 In addition, several syntactical features of Peshitta Proverbs militate against ascribing too late a date to it.20 If the bulk of the OT Peshitta was produced before around 150 AD, as I have argued elsewhere,21 a date in the early third century would seem probable for Peshitta Proverbs.

A SURVEY OF THE SEVEN DOUBLET TRANSLATIONS The doublet translations in Peshitta Proverbs can be divided into two groups: two doublets correspond to doublet translations in the LXX, while in the other five cases the Greek has no doublet. Let us first consider some examples of the second group. ̈ 11:29a. The first rendering, ‫ܬܢܚܬܐ‬ ‫ܕܒܢܐ ܒܝܬܗ ܒܥܬܐ ܢܫܒܘܩ‬ ̈ ‫“( ܠܒܢܘܗܝ‬He who builds his house with deceit will bequeath groan17 In an earlier publication I asserted the likelihood that the Syriac text of Prov. 11:31 was taken integrally from 1 Pet. 4:18 by a revisor of the OT Peshitta. In view of other indication of the lateness of Peshitta Proverbs, this claim seems unnecessary to me now. 18 This would also seem to imply that the author of Peshitta Proverbs was a Christian. For another argument, see below at n. 31. A Christian origin and a late date for Peshitta Proverbs would fit in with the theory of Weitzman that the community that produced the OT Peshitta moved from Judaism to Christianity during the process of translating the OT. 19 See Pinkuss, “Die syrische Übersetzung,” 107. 20 The most significant of these is the fact that the particle ‫ ܐܝܬ‬+ suffix is not used as a copula corresponding to Greek εἶμι, the sole exception being 8:2. 21 J. Joosten, “The Old Testament Quotations in the Old Syriac and Peshitta Gospels,” Textus 15 (1990), 55–76. In this article it is argued that Tatian made use of the OT Peshitta when he created his Syriac Diatessaron. This would mean that the OT Peshitta attained a certain prestige before 170 AD or so.

76

LANGUAGE AND TEXTUAL HISTORY

ings to his children”), looks like a very free rendering of the Hebrew. Alternatively, this may have been an existing Syriac proverb that was inserted here by the translator, as is suggested by Pinkuss.22 An argument in favour of this view is that elsewhere the Hebrew root ‫ נחל‬is always rendered by forms of the verb ‫ܝܪܬ‬.23 Be this as it may, nothing in this verse betrays an influence from the LXX. Influence from the LXX is manifest, however, in the second rendering: ‫“( ܕܐܠ ܟܐܫ ܒܒ ݈ܝܬܗ‬he who sits not quietly at home”) echoes the ὁ μὴ συμπεριφερόμενος τῷ ἑαυτοῦ οἴκῳ (“he that deals not graciously with his own house”), and the plural in the direct object “winds” corresponds to ἀνέμους in certain LXX manuscripts. The second Syriac rendering is not a straightforward translation of the LXX, however. Most notably, the expression ̈ ‫“( ܢܦܠܓ‬he will divide among his children”) does not corre‫ܠܒܢܘܗܝ‬ spond to κληρονομήσει (“he will inherit”),24 but seems to attest influence from the first rendering on the second. This is particular̈ (“to his children”) occurly apparent from the addition of ‫ܠܒܢܘܗܝ‬ ring in both renderings. Both Lagarde and Pinkuss have judged the second rendering to be a later correction on the basis of the LXX. 25 In light of that rendering’s poor quality as a translation of the LXX, this point seems to me unproven. One would expect that a corrector attaching importance to the LXX text would adhere to it more closely. 14:9. Here it is the second rendering — “fools do sins, and the sons of the upright (do) delightfulness” — that corresponds most closely to the Hebrew. In the first half, the meaning of ‫ יָ ִליץ‬is guessed at, andַ ‫שם‬ ַָ ‫ ָא‬is taken as a general designation of sin. The second half renders the MT, except that the translator seems to

Pinkuss, “Die syrische Übersetzung,” 168. See Prov. 3:35; 8:21; 13:22; 14:18; 28:10. 24 Whether the Syriac verb ‫ ܟܘܫ‬is an adequate rendering of Greek συμπεριφέρομαι is questionable as well. The Syrohexapla renders it with the verb ‫ܫܦ‬, Aphel. 25 Lagarde, Anmerkungen, 38; Pinkuss, “Die syrische Übersetzung,” 168. 22 23

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

77

have read ‫“( בני‬sons”) instead of ‫“( בין‬among”). Nothing points to influence from the LXX. The first rendering corresponds to the LXX. Yet, ‫ܬܪܝܨܐܝܬ‬ ̈ ‫ܒܥܝܢ‬ (“they rightly seek”) is a poor equivalent for ὀφειλήσουσιν (“they need”).26 Pinkuss again believes that the rendering corresponding to the LXX is due to later revision.27 This view will be further discussed below. 14:23. The second rendering corresponds to the MT except for the following points: a) instead of ‫ בכל‬the translator seems to have read ‫ ;כל‬b) for ‫ יהיה‬the translator has read ‫יהוה‬, which he rendered ‫“( ܡܪܝܐ‬the Lord”) as is his usual practice; c) the addition “(the speech of the lips) of the wicked” typifies a “Targum-like” tendency of Peshitta Proverbs to add words in order to specify the meaning of the text. 28 Other differences between the Peshitta and the Hebrew are the result of free translation. This rendering is free of influence from the LXX. The first rendering is dependent on the LXX. Yet it certainly is not an exact translation of the Greek. Most notably, ἔνεστιν was parsed as ἕν ἔστιν and rendered ‫“( ܚܕܐ ܗܝ‬there is one thing”),29 and in the second stich subject and predicate were inverted. Other variations may be explained as loose renderings. However, the resulting text “In all your worrying there is (only) one thing that is helpful, and the one who is poor with regard to sustenance will be tranquil and merry” is almost the opposite of the LXX’s “With every one who is careful there is abundance; but the pleasure-taking and the indolent shall be in want.”30 This divergence seems to be significant. It has been noted by Weitzman that Peshitta Proverbs 26 Could it be that this expression was influenced by the last two words in the Hebrew ‫יםַרצֹון‬ ָ ‫?יְ ָש ִר‬ 27 Pinkuss, “Die syrische Übersetzung,” 175. 28 See, e.g., the following additions: “innocent (blood)” in 1:16; “even the wise” in 14:35; “unto God” in 21:13; “his misfortune” 22:16; “God” in 28:13. 29 This reading is found in some Greek manuscripts, see Lagarde, Anmerkungen, 48. 30 Here, as elsewhere in renderings of the LXX, I have followed the English translation of Brenton.

78

LANGUAGE AND TEXTUAL HISTORY

in some places introduces positive statements on poverty absent from the Hebrew and the Greek.31 This tendency may be an indication that the author of Peshitta Proverbs was a Christian; as is wellknown, a positive view on poverty typifies much early Syriac Christian literature. What is important for our purpose is that if this rendering is colored by the ideology of the translator, this makes it unlikely that it comes from a later corrector. To begin with, one should expect a later editor, who set out to correct the original Syriac version of Proverbs on the basis of the LXX, to follow the LXX. Secondly, since the tendency to introduce positive statements on poverty is typical of Peshitta Proverbs as a whole and since it also occurs in places that are not dependent on the LXX, we will do well to ascribe the first rendering of 14:23 to the original Syriac translator of the book.32 23:31. This case is complicated somewhat by the fact that the LXX apparently exhibits a double rendering of part of the verse. The second rendering in the Peshitta is closer to the Hebrew. The following points are worthy of note: a) the Peshitta has no element corresponding toַ ‫י־יִתּן‬ ֵ ‫ ִַכ‬, b) the last two words of the Hebrew, ‫יש ִרים‬ ָ ‫ְךַב ֵמ‬ ְ ‫יִת ַה ֵל‬, ְ receive a contextual rendering “but meditate righteousness”, c) account is to be taken of the possibility that the expression ‫[“( ܬܨܕ ܥܝܢܟ‬do not] fix your eye”) is a reflex of δῷς τοὺς ὀφθαλμούς in the LXX — even though these words render a different part of the Hebrew verse. 31 Weitzman, “From Judaism to Christianity,” 161f. Weitzman points to 22:7 and 8:21. Other examples include 11:16 (“wisdom” is substituted for “riches”); 28:6. See also 18:11; 12:9; 19:1, 22; 21:20. It must be noted, however, that positive statements on wealth in the Hebrew are generally correctly rendered, see e.g. 3:16; 10:4, 22; 22:4; 13:8; 14:24; ̈ (“the 15:15. The text of 24:25 in the Leiden edition, ‫ܡܣܟܢܐ ܢܬܒܣܡܘܢ‬ ̈ poor will delight”) is due to a printing error; read ‫ܡܟܣܢܐ ܢܬܒܣܡܘܢ‬ (“those who rebuke will delight”). 32 The rare word ‫ܥܘܡܪܐ‬, which I have rendered “sustenance”, is used in the same sense in 13:23 and 30:8, another indication that the author of the first rendering of 14:23 is responsible for all of Peshitta Proverbs.

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

79

The first rendering (23:30b in the Leiden edition) approximates to the first part of the LXX text of the verse. It is to be noted, however, that the first four words, ‫“( ܒܚܡܪܐ ܐܠ ܬܗܘܐ ܪܘܝ‬do not be drunken from wine”), are taken over from the translation of 23:20a, where they render the Hebrew ַ ‫יַבס ְב ֵאי־ַָייִן‬ ְ ‫ל־תּ ִה‬ ְ ‫“( ַ ַא‬be not among winebibbers”); the LXX of v. 20a has μὴ ἴσθι οἰνοπότης. This makes it unlikely that the first rendering of the Peshitta was a later interpolation, as is claimed by Pinkuss.33 It is improbable that a later editor would borrow words of the version he set out to correct, in order to render the Greek text which he considered more authoritative. The four cases discussed thus far allow us to draw some preliminary conclusions. All four cases are doublet translations in which one rendering corresponds more closely to the Hebrew and the other to the Greek. In 11:29a the rendering of the Hebrew precedes the rendering of the Greek, in the other cases the order is reversed. Since the Hebrew-based rendering is, as a rule, free of influence from the LXX, these cases may create the impression that an original, “Old Syriac”, translation of Proverbs based mainly or exclusively on the Hebrew was later interpolated with renderings informed by the LXX version.34 This model is contradicted, however, by certain characteristics of the Greek-based renderings. Firstly, it is typical of these renderings that they do not adhere closely to the text of the LXX. Moreover, where they diverge they sometimes conform to general tendencies of Peshitta Proverbs found also in places where the text does not reflect the Greek. Finally, these Greek-based renderings occasionally link up with the corresponding Hebrew-based reading, or even with the rendering of an earlier verse. Admittedly, these characteristics cannot be demonstrated in every case. The Greek-based renderings discussed above are sufficiently similar, however, to be considered a homogeneous group.

Pinkuss, “Die syrische Übersetzung,” 200. It is theoretically possible to conceive the opposite, namely that an original translation exhibiting influence of the LXX was later revised on the basis of a Hebrew text. This hypothesis is unlikely for the same reasons as the one held to by Pinkuss. 33 34

80

LANGUAGE AND TEXTUAL HISTORY

On the basis of these considerations, and taking into account the general character of Peshitta Proverbs, I would like to propose a different model. The author of Peshitta Proverbs was working on the Hebrew and the Greek simultaneously. Where he understood the Hebrew, he translated it (though sometimes quite freely); where the Hebrew was difficult, he followed the Greek. In many places, however, both Hebrew and Greek contained elements that looked interesting to him. When such was the case, he did not always choose one at the expense of the other: at times, he combined elements from both his sources into a “versional patchwork” as we have seen above. At other times he opted to translate both versions integrally — which procedure resulted in the doublet translations evaluated above. A fact tending to support this view is that the translational attitude evinced by the Hebrew-based and the Greek-based renderings is remarkably similar. Even if we allow for insufficient knowledge of the languages involved, and for defects in the manuscripts used, both types of rendering are rather free and loose. This is entirely in keeping with the impression one gleans from reading through the rest of Peshitta Proverbs. Rather than postulating a succession of two authors who approached their respective tasks in the exact same way, we may accept that one and the same author was responsible for the entire version.35 One may think it strange that a translator would include doublets in his translation. It is not impossible that the author of Peshitta Proverbs did, in fact, adopt only one of each pair of parallel renderings in the body of his text, while writing the other rendering in the margin. This procedure would explain the remarkable position of the second rendering of 14:32. 14:32. Peculiar to this example is the fact that the two renderings are nearly identical. The only difference is between ‫ܒܒܝܫܘܬܗ‬ (“in his evil”) in the first, and ‫“( ܒܚܟܡܬ ܢܦܫܗ‬in his own wisdom”) in the second rendering. It seems to me that Pinkuss is wrong 35 In his textual notes, Pinkuss, as we have seen, occasionally champions the view of extensive revision on the basis of the Greek. In his introduction, however, he seems to come down to a view close to the one defended here, see Pinkuss, “Die syrische Übersetzung,” 106f.

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

81

when he ascribes the second reading to influence from the LXX version of 14:35. Rather, ‫ ܒܚܟܡܬ ܢܦܫܗ‬corresponds to a different reading of the Hebrew of 14:32: not ‫ברעתו‬, but ‫בדעתו‬. Support for this hypothesis can be found in the expression‫ ܚܟܡܬܐ ܕܢܦܫܗ‬rendering Hebrewַ ‫ ַ ִַבינָ ְתָך‬in 3:5. Why this variant interpretation caused the entire verse to be written twice is not clear. The second rendering of 14:32 does not immediately follow the first as in the other examples, but occurs in the middle of v. 35. This suggests that at an early stage in the textual tradition of the book of Proverbs, the second part of the doublet did not figure in the text but in the margin. Later copyists incorporated the rendering in the wrong place in the text. On the analogy of this case it is possible to imagine that all the doublet translations discussed above originally consisted of the reading in the body of the text and an alternative rendering written in the margin.36 In the original lay-out of Peshitta Proverbs, these doublets would then have been comparable to the text-critical notes found in some modern Bible translations. The two examples of the second group, where the Syriac doublet corresponds to a doublet translation in the LXX, still remain to be considered. 14:22. The first rendering in the Peshitta looks like a very free translation of the Hebrew, while containing some elements dë (“the pendent on the first rendering in the LXX: a) the word ‫ܥܘܐܠ‬ wicked”) seems to have been introduced under the influence of ἄδικοι attested by several manuscripts of the LXX,37 b) the verb ‫“( ܛܥܐ‬to forget”38) is a mistaken rendering of the Hebrew, c) the ̈ may be due to influence from the Greek κακά. plural ‫ܒܝܫܬܐ‬

This may explain why the Greek-based rendering sometimes precedes and at other times follows the Hebrew-based rendering. 37 See Lagarde, Anmerkungen, 48. Possibly, however, the addition was added freely by the author of Peshitta Proverbs, see n. 28 above. 38 The verb also has the meaning “to err”, which would be close to the meaning of the Hebrew. However, the syntactical construction clearly indicates the sense “to forget”. 36

82

LANGUAGE AND TEXTUAL HISTORY

Pinkuss judges this first rendering to be original (“echt”), probably because of the points of contact with the Hebrew.39 The second rendering does not indicate any knowledge of the Hebrew, but corresponds to the second rendering of the LXX, of which it is a fairly exact translation. To my mind, this fact is not sufficient to prove that the second rendering is secondary, as seems to be held by Pinkuss. In light of the preceding discussion it is better to suppose that the double rendering was made by the original author of Peshitta Proverbs, who did not want to eliminate a possible rendering of the verse from his translation. 18:22. In this case it is reasonably clear that the two renderings go back to the same author. In the first, influence of the first rendering of the LXX is evident: a) the relative clause ‫“( ܕܡܫܟܚ‬whoever finds”) corresponds to ὃς εὗρεν, b) the addition of ‫ܛܒܬܐ‬ (“good”) also attests influence of the LXX, c) the rendering ‫ܡܩܒܠ‬ (“receives”) is closer to the Greek than to the Hebrew. Other elements, however, go back to the Hebrew: neither ‫“( ܨܒܝܢܐ‬goodwill”) nor ‫“( ܡܢ ܡܪܝܐ‬from the Lord”) can be explained from the Greek. The second rendering corresponds to the second rendering in the LXX. Yet, the second stich — “he that keeps an adulteress is foolish and ungodly” — which has no counterpart in the MT, is not taken up in the Peshitta. This shows that the author of the second rendering in the Peshitta, though translating from the Greek, did not lose sight of the Hebrew. The two renderings are linked by the conjunction ‫ܘ‬, and the doublet is rounded off by the addition of the words ‫ܡܢ ܒܝܬܗ‬ (“from his house”). All these points concur in showing that one and the same person is responsible for the entire doublet.

CONCLUSIONS The conclusions of the present survey have already been intimated above. Doublet translations might be called an epitome of the translational procedure found in Peshitta Proverbs. Working on the Hebrew and the Greek texts of Proverbs simultaneously, but en39

Pinkuss, “Die syrische Übersetzung,” 177.

DOUBLET TRANSLATIONS IN PESHITTA PROVERBS

83

dowed with no particular skill in dealing with either language, the author of Peshitta Proverbs produced a very loose translation that incorporated elements from both his sources. Given the numerous differences between the MT and the LXX, he often had to choose one version at the other’s expense. In some cases, however, he refused to make a choice, attempting rather to preserve the gist of both versions. The result was, of course, more often than not a text that rendered neither the Hebrew nor the Greek. In the context of this general procedure, doublet translations stand revealed as a translational technique designed to preserve all the textual material available. Rather than choosing in favour of one of the texts, and rather than combining them into one rendering, the author simply translated the verse twice, following first one text form and then the other. If the doublet translations do not always preserve the meaning of this textual material very well, the author’s ineptitude and, occasionally, his ideology are to blame. It is not impossible that one of each set of double translations (excepting 18:22) was intended to feature in the margin, as a type of textual note. The evidence for this view is quite slim, however.

GREEK WORDS SHARED BY THE PESHITTA AND TARGUMS TO THE PENTATEUCH One of Michael Weitzman’s qualities as a scholar was his ability to take a complicated and controversial theory and make it appear reasonable and straightforward by the sheer accumulation of converging facts dug out in the course of his wide-ranging researches. An instance of this is the theory of a “fund of Aramaic renderings”, transmitted through the ages and available to different circles of interpreters or translators of the Bible in antiquity.1 Complementing the less controversial explanations invoking shared linguistic reflexes and accidental agreement, the hypothesis of a fund of Aramaic renderings explains the many similarities between the Peshitta and the Targums while doing away with the need to postulate a direct family relationship between the two.2 The materials to be presented in the present paper can well be explained in the framework of Weitzman’s approach, which is thus shown to be an effective working hypothesis in dealing with one of the more complicated problems in biblical philology. Some sixteen words of Greek (and, in some cases, ultimately Latin) origin occur in the Peshitta Pentateuch and in one or more See M. P. Weitzman, “Peshitta, Septuagint and Targum,” in R. Lavenant, ed., VI Symposium Syriacum 1992 (OCA 247; Rome: Pontificio Istitùto Biblico, 1994), 51–84, esp. 65; reprinted in Idem, From Judaism to Christianity (JSStS 8; Oxford: University Press, 1999), 181–216. 2 Any theory stipulating a direct genetic relationship between the Peshitta and the Targums – i.e. with the Peshitta depending directly on a Jewish Targum or the Targums depending on the Peshitta or with Peshitta and Targum depending on a Palestinian proto-Targum – runs into insuperable problems, see Weitzman, “Peshitta,” 60–79. 1

85

86

LANGUAGE AND TEXTUAL HISTORY

of the Targums rendering the same Hebrew word in the same passage. Some of these agreements in the use of a Greek equivalent may be fortuitous, i.e., they merely indicate that Greek influence was strong on different Aramaic dialects of the Late Aramaic phase. Other agreements, however, cannot be explained in this way. They point to the existence of an exegetical tradition on given passages, providing Greek glosses for certain Hebrew technical terms. This tradition seems to have been known to the Syriac and the Aramaic translators independently. 3

THE EVIDENCE A list of the words discussed in this study follows. The Greek donor word is given first, followed by its Syriac and Aramaic forms and an approximate English translation. For each word, the Hebrew equivalent or equivalents are noted, MT chapter and verse references given, and the versions where the word is found are listed.4 Some doubtful cases will be discussed below. 1) βάσις ‫‘ בסיס ܒܣܣ‬base’ ‫ ירך‬Exod. 25:31; 37:17; Num. 8:4 PN ‫ כן‬Exod. 39:39 PON 2) βήρυλλος ‫‘ בורלא ܒܪܘܐܠ‬beryl’ ‫ שהם‬Gen. 2:12 ; Exod. 25:7; 28:9, 20; 35:9, 27; 39:6, 13 PO 5 Greek words occurring in the Peshitta and in one of the Targums rendering different words have not been included in the present study. For one example of this phenomenon, see J. Joosten, “χαλκηδών,” RHPR 79 (1999), 135–43. 4 Hebrew equivalents are referred to by chapter and verse in the MT. Where this differs from the AV’s chapter and verse division, the latter follows in brackets. Peshitta is indicated by P. As far as the Targums are concerned, only Onqelos (O), Neofiti (N) and the Samaritan Targum (S) have been considered. The Fragmentary Targum, the Genizah fragments and Pseudo-Jonathan have been checked, but were found to provide no independent information on the issue under discussion. The Christian Palestinian Aramaic version has been excluded on methodological grounds. Since it was made from a Greek source text, the Greek words occurring in it require a different approach. 5 The same Greek word is found in the Septuagint in Exod. 28:20. 3

GREEK WORDS IN PESHITTA AND TARGUMS

87

3) γλύφω ‫‘ גלף ܓܠܦ‬to carve, to sculpture’ ‫( פתח‬Piel) Exod. 28:9 et alibi PON6 4) καγκέλλον (< cancelli) ‫‘ קנקל ܩܪܩܠ‬latticed screen’ ‫ מכבר‬Exod. 27:4; 38:4, 30; 39:39 PN 5) κέλλα or κέλλιον (< cella) ‫‘ קהלא ܩܠܝܬܐ‬cell, alcove’ ‫ קבה‬Num. 25:8 PN 6) κοιτών ‫‘ קיטון ܩܝܛܘܢܐ‬inner room’ ‫ חדר‬Exod. 7:28; Num. 8:3 PN 7) κρίκος ‫קירכס ܩܘܪܩܣܐ‬ ‫ קרס‬Exod. 26:6, 11 PS7 8) λαμπάς ‫‘ למפד ܠܡܦܐܕܐ‬lamp, torch’ ‫ לפיד‬Exod. 20:18 PN8 9) λεγιών (< legio) ‫‘ לגיון ܠܓܝܘܢܐ‬legion’ ‫ צים‬Num. 24:24 PN9 10) πεῖσαι ‫‘ אטפס ܐܬܛܦܝܣ‬to be persuaded’ ‫ אות‬Gen. 34:15, 22, 23 PO 11) πόρπη ‫‘ פרף ܦܪܦܐ‬clasp, hook’ ‫ קרס‬Exod. 35:11 PON 12) σειρά ‫‘ סירה ܣܝܪܣ‬cord, lasso (?)’ ‫ תחרא‬Exod. 28:32; 39:23 PN 13) τήγανον ‫‘ טיגן ܛܐܓܢܐ‬frying pan’ ‫ מחבת‬Lev. 2:5; 6:14 (21); 7:9 PS10 14) τύπος ‫‘ טופסא ܛܘܦܣܐ‬mould, form’ ‫ חרט‬Exod. 32:4 PN 15) φιάλη ‫‘ פיילי ܦܝܠܣܐ‬dish, bowl’ ‫ קערה‬Exod. 25:29; Num. 7:13ff PN11

The Septuagint too uses this word in rendering this Hebrew verb. The Greek word is found in the same passages in the Septuagint, whence it has also penetrated into the Christian Palestinian Aramaic version. 8 The word also occurs in the Septuagint of Exod. 20:18. 9 The first part of this verse is translated doubly in Targum Neofiti. The Greek word occurs in the second part of the doublet. 10 The same word occurs in these passages in the Septuagint. 6 7

88

LANGUAGE AND TEXTUAL HISTORY 16) χαράκωμα ‫‘ כרכומין ܩܠܩܘܡܐ‬bulwark, siege works’ ‫ מצור‬Deut. 20:20 PON

This analysis shows that the Peshitta Pentateuch agrees in the use of a Greek word: - nine times with Targum Neofiti exclusively (1, 4, 5, 6, 8, 9, 12, 14, 15) - twice with Targum Onqelos exclusively (2, 10); - four times with Onqelos and Neofiti (1, 3, 11, 16); - twice with the Samaritan Targum exclusively (7, 13).12

AGREEMENTS DUE TO INDEPENDENT BORROWING FROM GREEK A brief review of the principal features of the use of Greek loanwords in Aramaic will establish the background against which these sixteen cases should be viewed. Greek words in Aramaic texts are, in general, an index of Hellenistic influence on the Ancient Near East.13 Such influence goes back a long way,14 but becomes a dominant factor only from the Hellenistic period onward. Alexander’s conquests established the Greek tongue in a dominant position — as the language of administration and warfare, and later of commerce, culture and learning — in much of the Aramaeophone area. As a result, Aramaic assimilated certain Greek elements. 15 A few The Greek word is frequent in the Septuagint, where, however, it renders Hebrew words other than ‫קערה‬. 12 The figures add up to 17, because n° 1 was counted twice, the Greek word corresponding to two different Hebrew equivalents. 13 See A. Tal, The Language of the Targum of the Former Prophets and its Position within the Aramaic Dialects [Hebrew] (Tel-Aviv: University Press, 1975), 175–86. 14 For Greek words in early Aramaic, see K. Kitchen, “The Aramaic of Daniel,” in D. J. Wiseman et al., eds., Notes on Some Problems in the Book of Daniel (London: Tyndale Press, 1965), 31–79; for Greek words in early Hebrew, see J. P. Brown, Israel and Hellas (BZAW 231; Berlin: De Gruyter, 1995). 15 Socio-linguistic studies have shown that in a bilingual community, the “high” or prestige language remains relatively untainted by any systematic influences from the relatively “low” language, while the “low” 11

GREEK WORDS IN PESHITTA AND TARGUMS

89

legal, military and commercial terms were borrowed initially, followed rapidly by words and expressions from other areas. By the Middle Aramaic language period (ca. 200 BCE to ca. 200 CE),16 when local dialects emerged as a means of written expression, all these dialects, Nabataean, Judaean, Palmyrene, Edessene and Hatran, incorporated Greek loanwords.17 The process persisted into the Late Aramaic phase (200 CE to 700 CE) where, in the words of J. Fitzmyer, a “mounting influx of Greek words and constructions into almost all dialects of the language” was evident. 18 For Syriac this increase in Greek loanwords has been described; there is also some literature on the influx into Targumic Aramaic.19 Some Greek words may have been borrowed by Aramaic before the dialects branched off into distinct written languages.20 Many others would naturally have been accepted, even at a later stage, by the different Aramaic dialects. In this light, it is not suprislanguage tends to be heavily influenced by the “high” language, with particular effect on the lexicon. See the material gathered in M. Silva, “Bilingualism and the Character of Palestinian Greek,” Biblica 61 (1980), 198– 219. 16 For the chronology of the phases of the Aramaic language I have followed J. A. Fitzmyer, “The Phases of the Aramaic Language”, in Idem, A Wandering Aramean: Collected Aramaic Essays (Missoula: Scholars Press, 1979), 57–84. 17 See J. F. Healey, “Lexical Loans in Early Syriac: A comparison with Nabataean Aramaic”, SEL 12 (1995), 75–84; J. Hoftijzer and K. Jongeling, Dictionary of the North-West Semitic Inscriptions, Part 1 and 2 (Handbuch der Orientalistik; Leiden: Brill, 1995). There is unfortunately no index of Greek words in the latter dictionary. 18 See Fitzmyer’s “Phases,” 62. 19 For bibliography on Greek words in Syriac, see J. Joosten, “Greek and Latin Words in the Peshitta Pentateuch. First Soundings,” in R. Lavenant, ed., VII Symposium Syriacum 1993 (OCA 256; Rome: Pontificio Istitùto Biblico, 1998), 37–47, esp. 37, n. 3, 4 and 5. For Greek words in the Targums, see Tal, Language, 175–86, and S. Krauss, Griechische und lateinische Lehnwörter im Talmud, Midrasch und Targum, Parts 1 and 2 (reprint Hildesheim: G. Holms, 1987 [1898]). 20 An example would be the Greek word στατήρ, ‘stater’, attested as a loanword in Official Aramaic, in Hatran and in Syriac.

90

LANGUAGE AND TEXTUAL HISTORY

ing that many Greek words are attested both in Syriac and in Western Aramaic. Nothing prevents the use of such shared loanwords in the respective Bible versions, and it is to be expected that, when they are so used, they would correspond to the same Hebrew word. Some at least of the agreements between Peshitta and Targums in the use of Greek words are probably due to this process. Such instances simply show that a given meaning was most naturally expressed, in both Syriac and Western Aramaic,21 by the Greek word in question. The most obvious example is that of the verb ‫ܓܠܦ‬/‫גלף‬, ‘to engrave’ (n° 3 in the above list). Although a loan from Greek,22 this word is, in Syriac as well as in Targumic Aramaic, the normal way to express the relevant meaning. The use of the term in the same passages, to render the same Hebrew word, is no more surprising than the agreement between Peshitta and Targums in the use of many indigenous Aramaic words.23 The agreement flows from the fact that Syriac and Jewish Aramaic are closely related dialects, and needs no explanation on the basis of mutual influence or dependence on common tradition. Several other agreements listed in the first section may confidently be ascribed to the linguistic proximity between Syriac and Targumic Aramaic. These include, for instance, ‫ܠܡܦܐܕܐ‬/‫למפד‬, ‘lamp, torch’ (n° 8),24 and ‫ܩܝܛܘܢܐ‬/‫קיטון‬, ‘inner room’ (n° 6): these

There is of course considerable dialectal variety among the different Targums. The term “Western Aramaic” is used here in a general way. 22 The Greek origin of this word has been contested by A. Schall, but still seems likely and has been upheld in several more recent publications. For the discussion see, e.g., Tal, Language (see n. 13 above), 177. 23 The high degree of verbal coincidence between Peshitta and Targum is well illustrated by the lists collected in J. C. de Moor and F. Sepmeijer, “The Peshitta and the Targum of Joshua,” in P. B. Dirksen and A. van der Kooij, eds., The Peshitta as a Translation (Monographs of the Peshitta Institute 8; Leiden: Brill, 1995), 129–176. Note, however, that De Moor and Sepmeijer believe that the data indicate dependence of the Peshitta on the Targum. 24 For this loanword in Syriac, see Joosten, “Greek and Latin Words,” 39. 21

GREEK WORDS IN PESHITTA AND TARGUMS

91

words are fairly common in both Syriac25 and Jewish Palestinian Aramaic.26 Their occurrence in one and the same passage, rendering the same Hebrew word, is unsurprising. A similar explanation almost certainly applies to the two examples involving the Samaritan Targum. The words ‫ܩܘܪܩܣܐ‬ /‫( קירכס‬n° 7)27 and ‫ܛܐܓܢܐ‬/‫( טיגן‬n° 13)28 are well attested in both Syriac and Western Aramaic, and this probably accounts for their use to render the same Hebrew word in the same passage. The question that naturally arises at this point is: if the occurrence of the Greek loanwords discussed above in both the Peshitta and the Targum is a consequence of the linguistic closeness of Syriac and the Aramaic dialect used in the relevant Targums, does this explanation also account for the remaining agreements between the Peshitta and one or more of the Targums? This approach is attractive because it fits well with the linguistic profile of the Aramaic versions, and dispenses with any complicated theory of mutual dependence on a non-attested source. There are, however, some weighty arguments against this simple solution, as will be demonstrated by the examples given in the following section.

Apart from the verses under discussion, ‫ ܠܡܦܐܕܐ‬occurs in the Old Testament Peshitta in Gen. 15:17; Judg. 7:16; Job 41:10; ‫ ܩܝܛܘܢܐ‬in Judg. 15:1; 1 Kgs 1:15; 2 Kgs 6:12. For occurrences of these words in other Syriac texts see R. Payne Smith, ed., Thesaurus Syriacus (Oxford: Clarendon Press, 1879–1897); C. Brockelmann, Lexicon syriacum (2nd ed.; Halle: M. Niemeyer, 1928) 26 See M. Sokoloff, A Dictionary of Jewish Palestinian Aramaic of the Byzantine Period (Ramat Gan: Bar Ilan Press, 1990). 27 In the Peshitta this word is the regular translation equivalent of Hebrew ‫טבעת‬. It is well attested in Jewish Palestinian Aramaic (see Sokoloff, Dictionary) and in Christian Palestinian Aramaic (see n. 7 above). 28 This word is used again in the Peshitta of Ezek. 4:3; Jer. 29:22. Information on Samaritan Aramaic is hard to come by, but the word is well attested in Jewish Palestinian Aramaic (see Sokoloff, Dictionary) and Mishnaic Hebrew. 25

92

LANGUAGE AND TEXTUAL HISTORY

GREEK WORDS GOING BACK TO AN EXEGETICAL TRADITION Two main arguments may be fielded for deriving some of the Greek words of the Peshitta and the Targums from an existing tradition. First, some of the Greek words listed are not obvious equivalents for the Hebrew words they render, thus suggesting that the choice may have been influenced by an exegetical tradition. Second, some of the words are rare or even unknown in the Aramaic dialects in which they occur, so that the theory of coincidental use developed in the second section is inapplicable. A further argument is that some of these words also occur in Symmachus’ version of the Pentateuch. The Greek Word is not an Obvious Equivalent Provenance from an exegetical tradition may be suggested, first, by the unexpected interpretation of the Hebrew word in question implied by the choice of the Greek term. Several instances have been identified and discussed by Y. Maori.29 A striking example is that of ‫ܠܓܝܘܢܐ‬/‫לגיון‬, ‘legion’, in Num. 24:24 (n° 9).30 The Hebrew of the first part of the verse, ‫וצים מיד כתים‬, is admittedly oracular; nonetheless, rendering so as to imply a prophecy concerning the coming of (presumably Roman31) legions indicates a leap in interpretation. The occurrence of this interpretation in both the Peshitta and the Palestinian Targums, where the theme is further developed, cannot be due to chance. An ancient Jewish exegetical tradition has clearly been adopted by both the Syriac translators and the Targumists. The Greek, and ultimately Latin, loanword ‫ܠܓܝܘܢܐ‬/‫ לגיון‬appears to have been part and parcel of this tradition. Other words implying a choice which may have been determined by an exegetical tradition include ‫ܛܘܦܣܐ‬/‫טופסא‬, ‘mould’

29 Y. Maori, The Peshitta Version of the Pentateuch and Early Jewish Exegesis [Hebrew] (Jerusalem: Magnes Press, 1995). 30 See Maori, Peshitta, 191f. 31 See Maori, Peshitta, 191.

GREEK WORDS IN PESHITTA AND TARGUMS

93

(n° 14),32 and ‫ܐܬܛܦܝܣ‬/‫אטפס‬, ‘to be persuaded’ (n° 10),33 both of which render Hebrew lexemes that proved difficult to early interpreters. The first renders the difficult Hebrew word ‫חרט‬, to which have been ascribed a variety of meanings including ‘stylus’,34 ‘cloak’,35 ‘mould’,36 and other interpretations. 37 The second corresponds to the Hebrew verb ‫אות‬, interpreted as ‘to become like’ in the Septuagint,38 ‘to get involved with’ in Targum Neofiti,39 ‘to agree, be in agreement with’ in a fragment from the Geniza,40 and ‘to be persuaded’ in other versions.41 In the light of this exegetical diversity, it is noteworthy that Peshitta and Targum arrived at the same interpretation in these passages. That this interpretation was expressed by means of the same Greek word is too striking to be due to chance. 32 See Maori, Peshitta, 265–66. Maori hesitates to ascribe the agreement of Peshitta and Targum to a common tradition. 33 This word is not treated by Maori, but it is signalled by S. P. Brock, “Some Aspects of Greek Words in Syriac” in A. Dietrich, ed., Synkretismus im syrisch-persischen Kulturgebiet (Abh. Ak. Wiss. in Göttingen 96; Göttingen, 1975), 80–108, esp. 85. 34 Thus in the Septuagint, γραφίς, on the basis no doubt of Isa. 8:1. 35 Thus in the first rendering of Targum Pseudo-Jonathan and in the Samaritan Targum, see 2 Kgs 5:23. 36 See also the rendering in the Vulgate: formavit opere fusorio. 37 Ibn Ezra takes the word to mean ‘form’, with explicit reference to Isa. 8:1 where the same interpretation is proposed. Yet another rendering is ‫ זיפא‬in Targum Onqelos, a word that is itself difficult to interpret. 38 See below. In 2 Kgs 12:9, however, the Septuagint renders the verb συμφωνέω ‘to agree’. 39 Neofiti uses the verb ‫ אתערב‬in Gen. 34:15, 22, 23, while a fragment from the Genizah uses it in v. 22 only (for vv. 15 and 23 in this fragment, see the following note). It is to be noted that in Qumran Hebrew the verb ‫ אות‬in CD 20.7 is equivalent to the verb ‫ התרעב‬in the parallels 1QS 7.24– 25 and 8.23. 40 The verb ‫ אשתוי‬is found in vv. 15, 23 in a fragment from the Geniza and in the Samaritan Targum; the Vulgate’s rendering in v. 23, adquiescamus, also belongs here. 41 Thus the Peshitta and Targum Onqelos (but see n. 45 below), and Symmachus in v. 22. For the latter, see below.

94

LANGUAGE AND TEXTUAL HISTORY

The Greek Loan is Poorly Attested A second reason for assuming an exegetical tradition underlying the choice of certain loanwords is that some of these words are not otherwise well attested in Syriac and/or Jewish Aramaic. A good example from Syriac is ‫( ܩܪܩܠ‬n° 4), unattested except in the biblical passages listed above, and in exegetical comments on these verses. Its etymology is contested: in Payne-Smith’s Thesaurus the word is connected with Latin craticula, ‘small grate’, while Brockelmann ventures a derivation from Greek κλῇθρα, ‘bars’. In the opinion of the present author, the occurrence of the word ‫ קנקל‬in the Palestinian Targums, in the same passages, is conclusive evidence of a derivation of both words from καγκέλλον, as already suggested by some native Syriac lexicographers.42 The formal deterioration is remarkable neither with loanwords in general nor with Greek words in Syriac in particular.43 Another example of a Greek word poorly attested in Syriac is ‫ܦܪܦܐ‬, ‘clasp’ (n° 11).44 In Jewish Aramaic, the most striking example is certainly ‫סירה‬, ‘cord’ (n° 12). This word seems to be unattested in Jewish Aramaic, except for its occurrence in rendering of Hebrew ‫ תחרא‬in Exod. 28:32 and 39:23 in the Palestinian Targums. Similar remarks could be made regarding ‫( אטפס‬n° 10).45 It is very unlikely that ‫ ܩܪܩܠ‬and ‫ ܦܪܦܐ‬were used in the Peshitta under the influence of the Targumic tradition while, conversely, ‫ סירה‬and ‫ אטפס‬were taken by the Targumists from the Syriac version, and coincidence is an unlikely explanation of such examples. It would be just plausible to suggest that, for instance, ‫ ܩܪܩܠ‬was a well-established loanword in early Syriac, and that only by chance is it not attested elsewhere. Such an explanation, however, is not enSee R. Payne Smith, ed., Thesaurus Syriacus, col. 3760. See Joosten, “Greek and Latin Words,” 38–39. 44 The word is attested also in some manuscripts of 1 Macc. 10:89 where the Greek text has the same word. 45 The form of the verb in Targum Onqelos to Gen. 34:15, 22, 23 is, as far as the present author can see, unique in Targumic Aramaic. It is also noteworthy that the meaning of the loanword in Jewish Aramaic is usually ‘to be pacified’, or a similar sense. If the meaning in Onqelos is ‘to be persuaded’, as in the Peshitta, the verb is also unique semantically. 42 43

GREEK WORDS IN PESHITTA AND TARGUMS

95

tirely convincing for any individual example, and becomes even less probable when it needs to be invoked repeatedly. A more likely explanation is that these words are not really loanwords at all, but rather foreign words handed down in an exegetical tradition. Translation was a convenient approach to the interpretation of difficult or rare Hebrew words such as ‫ קרס‬or ‫תחרא‬, a very expedient method was to translate them into another language.46 The prestige of the Greek language, and the richness of its vocabulary, may have led, some time before the end of the first century CE, to the compilation of a list of Greek glosses explaining specific Hebrew words in the Pentateuch. If indeed there was such a list, the translators of the Peshitta and the Targumists appear to have made use of it independently, employing the same Greek term to render the same Hebrew word.47 This hypothesis would explain the picture in Targum Neofiti to Num. 25:8, where a transliterated Greek word has apparently been misunderstood. MT ‫ויבא אחר איש ישראל אל הקבה‬ And he went after the man of Israel into the alcove. N ‫ועל בתר גברא בר ישראל לגו קהלא‬ And he went after the man of Israel into the qhl’. Sokoloff does not list this attestation of ‫ קהלא‬in a separate entry, implying that he regards it as an instance of the well-known homograph meaning ‘gathering, congregation’.48 There is no reason to think that earlier readers of the Targum understood the word differently. However, the meaning ‘congregation’ fits the context poorly, and diverges markedly from the Hebrew. In the light of the Peshitta’s ‫ܩܠܝܬܐ‬, it is very probably that the Targum, too, reflects This method is well illustrated by the Fragmentary Targum, manuscript V, to Exod. 27:4, where the Hebrew word ‫ מכבר‬is glossed by means of the Greek (or Latin) loanword ‫קנקל‬. 47 Some words occurring in this hypothetical list may have been taken up in the Peshitta but not in the Targums, or vice versa. For some other possible examples in the Peshitta, see Joosten, “Greek and Latin”, 46, n. 48. 48 See Sokoloff, Dictionary. 46

96

LANGUAGE AND TEXTUAL HISTORY

the Greek word κέλλα (or κέλλιον). The he in ‫ קהלא‬may originally have been meant as a transcription of Greek epsilon;49 alternatively, it may have been added, or substituted for another letter, by a scribe who did not correctly identify the word. However this may be, the text of the Targum may be accounted for by positing that it incorporated a Greek word in Hebrew characters, which was later misunderstood.50 The Testimony of Symmachus Further support for the view that some of the Greek words contained in Peshitta and Targums derive from an exegetical tradition comes from Symmachus’ version of the Pentateuch. Symmachus’ version is a Jewish revision of the Septuagint dating from the second century CE.51 This makes it more or less contemporary with the Peshitta and with Targum Onqelos.52 Targum Neofiti represents a later stage in the Targumic tradition, but all authorities agree that it often reflects archaic elements. It is therefore of interest that two of the words discussed above also occur in Symmachus. ‫ אות‬πείθομαι Gen 34:22 ‫אך בזאת יאותו לנו האנשים לשבת אתנו‬ LXX μόνον ἐν τούτῳ ὁμοιωθήσονται ἡμῖν οἱ ἄνθρωποι τοῦ κατοικεῖν μεθ᾽ ἡμῶν

As perhaps in ‫ < סנהדריון‬συνέδριον; see S. E. Fassberg, “The Orthography of the Relative Pronoun -‫ שה‬in the Second Temple and Mishnaic Periods,” Scripta Classica Israelica 15 (1996), 240–250. 50 Another possible instance of this phenomenon occurs in Targum Pseudo-Jonathan to Exod. 32:4, where we find the word ‫טופרא‬, ‘nail, pencil’, corresponding to the Hebrew word ‫חרט‬. This is probably a simple corruption of the word ‫ טופסא‬as found in Neofiti. 51 See A. Salvesen, Symmachus in the Pentateuch (JSS Monograph 15; Manchester, 1991). 52 For the date of the Peshitta, see M. P. Weitzman, The Syriac Version of the Old Testament. An Introduction (University of Cambridge Oriental Publications 56; Cambridge: University Press, 1999); for the date of Targum Onqelos, see W. F. Smelik, The Targum of Judges (OTS 36; Leiden: Brill, 1995). 49

GREEK WORDS IN PESHITTA AND TARGUMS

97

The verb ‫ אות‬occurs three times in Gen. 34, in verses 15, 22, 23. The Septuagint renders with ὁμοιόω, ‘to become like’, throughout. Symmachus’ version has εὐνοήσομεν, “we will be kindly disposed”, instead of ὁμοιωθησόμεθα, in v. 15, while in v. 22 it reads πεισθήσονται, “they will be persuaded”, for the Septuagint’s ὁμοιωθήσονται; Symmachus’ reading in v. 23 is not known. The rendering in v. 22 clearly agrees with that of the Peshitta and Targum Onqelos in all three verses. ‫ תחרא‬σειρά/σειρωτόν Ex 28:32 ‫והיה פי ראשו בתוכו שפה יהיה לפיו סביב מעשה ארג‬ ‫כפי תחרא יהיה לו לא יקרע‬ LXX καὶ ἔσται τὸ περιστόμιον ἐξ αὐτοῦ μέσον, ᾤαν ἔχον κύκλῳ τοῦ περιστομίου, ἔργον ὑφάντου, τὴν συμβολὴν συνυφασμένην ἐξ αὐτοῦ, ἵνα μὴ ῥαγῇ

The difficult ‫ כפי תחרא יהיה לו‬has been rendered in the Septuagint, apparently rather freely, as “woven together in the joining of the same piece” (Brenton). For the phrase συμβολὴν συνυφασμένην Symmachus and Theodotion substitute σειρωτόν ‘bound (?), twisted (?), hemmed (?)’.53 Although this Greek word is not otherwise attested, it is clearly related to the Greek word σειρά which is reflected by the rendering in the Peshitta and Targum Neofiti. It is impossible to prove that these suggestions are correct; nonetheless, the most likely explanation for the agreement of Peshitta, Targum and Symmachus in each of these two cases is that the rendering by means of the Greek word in question had been previously established in an exegetical tradition of which translators in each group availed themselves independently. These two examples thus support the hypothesis advanced above.54

53 The word seems to be a hapax legomenon in Greek, see H. G. Liddell and R. Scott, Greek-English Lexicon (Oxford: University Press, 1996 [repr. of the 9th edition]), 1589. 54 Another striking agreement between the Peshitta and Symmachus is the use of the Greek word χρῶμα, ‘colour’, to render Hebrew ‫ עצם‬in

98

LANGUAGE AND TEXTUAL HISTORY

CONCLUSIONS The sixteen instances where the Peshitta agrees with one or more of the Targums in the use of a Greek word cannot all be similarly explained. Several indicate the close linguistic relationship of Syriac with Western Aramaic dialects. The Greek loanwords involved have been thoroughly assimilated into the Aramaic dialects in question and in fact function as Aramaic words. The agreement between Peshitta and Targums here is no more remarkable than their agreement in the use of many indigenous Aramaic words. There are however a number of agreements between the Peshitta and the Jewish Targums which cannot be explained in this way, but must indicate an established exegetical tradition. It seems probable that this tradition, circulating in a Semitic milieu, listed Greek equivalents for a number of difficult words in the Hebrew Pentateuch. The renderings involved are, therefore not really loanwords in the precise sense of the term, but rather foreign words, i.e., Greek words in Semitic transcription employed in an Aramaic text. The occasional misunderstanding of these words during the later course of transmission supports this suggestion. Thus, the Greek words shared by Peshitta and Targums to the Pentateuch epitomize the relationship of these versions with one another. Because these versions are all more or less contemporary Aramaic translations of the same Hebrew text there is a relatively high degree of coincidental agreement. A number of agreements, however, are due not to chance but to the fact that the Syriac translators and the Jewish Targumists have drawn on the same treasure of traditional Jewish exegesis.

Exod. 24:10. This rendering appears to be unattested in the Targumic tradition.

1 SAM XVI 6, 7 IN THE PESHITTA VERSION INTRODUCTION The books of Samuel have enjoyed intensive and unremitting interest from the side of textual critics. Apart from the standard critical editions, the text of the books of Samuel was made the subject of several monographs and numerous articles.1 Yet it seems that not all the textual treasures hidden in the various witnesses to this text have been uncovered. The present paper deals with a set of two variants in the Peshitta version (P) which have not been given the necessary attention in earlier research. It has not been observed with sufficient accuracy that the P account of the anointing of David contains a different version of the dialogue between Samuel and YHWH at the appearance of Eliab (1 Sam. xvi 6–7). P offers two variant readings: in v. 6 ‫ ܐܟܘܬܗ ܕܡܪܝܐ‬is not quite equivalent to ‫אְך נֶּ גֶּ דַיהוה‬ ַַ of the Massoretic Text (MT); v. 7 ‫ ܐܠ ܗܘܝܬ ܓܝܪ ܐܝܟ ܕܚܙܐ ܐܢܫܐ‬diverges markedly fromַ ‫הַה ָא ַָדם‬ ָ ‫ ִכיַלאַא ֶּשרַיִ ְר ֶּא‬. Furthermore, as will be discussed below, these two variants lead to a coherent understanding of the passage different from MT. The main part of this paper is devoted to the analysis of the two variants. It will be argued that they are genuine variant readings and not instances of exegesis, translation technique or inner-Peshitta corruptions or alterations. Some of the works on the text of the books of Samuel are considered milestones in the history of the text-critical study of the OT: J. Wellhausen, Der Text der Bücher Samuelis (Göttingen, 1871); S. R. Driver, Notes on the Hebrew Text and the Topography of the Books of Samuel (Oxford, 1890); more recent are P. A. H. de Boer, Research into the text of I Samuel I–XVI (Amsterdam, 1938) and E. Tov, ed., The Hebrew and Greek Texts of Samuel (Proceedings of the IOSCS; Vienna, 1980; Jerusalem, 1980). 1

101

102

LANGUAGE AND TEXTUAL HISTORY

On the basis of the analysis, the Hebrew text which they presuppose will be reconstructed. In the last section the textual history of 1 Sam. xvi 6–7 will be discussed. Though this point remains very hypothetical, it will be argued that the text preserved in P is the original text of 1 Sam. xvi 6–7 to which MT is secondary.

1 SAM. XVI 6 ‫ܐܟܘܬܗ ܕܡܪܝܐ ܡܫܝܚܗ‬ Samuel’s reaction to the appearance of Eliab is according to MTַ‫ַאְך‬ ‫“ נֶּ גֶּ דַיהוה ַ ְמ ִשיחֹו‬Surely the LORD’s anointed is before Him” (Revised Standard Version). In P the words of Samuel are ‫ܐܟܘܬܗ‬ ‫ܕܡܪܝܐ ܡܫܝܚܗ‬. The analysis of these words requires an exact knowledge of the Syriac idiom of P, but it is not problematic: - -‫ ܐܟܘܬ‬is the regular form with suffix pronouns of the preposition ‫“ ܐܝܟ‬as, like”;2 ‫ ܐܟܘܬܗ‬means “like him”. - the syntagm: preposition-suffix + ‫ܕ‬-noun is normal, with longer prepositions, when the noun is definite and has an important status in the context; 3 ‫ ܐܟܘܬܗ ܕܡܪܝܐ‬means “like the Lord, similar to the Lord”. The whole clause must be rendered: “similar to the Lord is His Messiah”, “The Lord’s Messiah is like He himself.” In light of this it is clear that the meaning of the Syriac diverges from that of the MT. But how shall we account for P’s text? Firstly, it is unlikely that P is an interpretation of the MT. The difference in meaning between “he stands before Him” (MT) and “he is like Him” (P) is simply too great. Also, we would expect an interpretation to be more transparent than the text it purports to interpret, whereas in this verse P’s text stands more in need of commentary than the MT.4 Secondly, it is problematic to view P’s text See T. Muraoka, Classical Syriac for Hebraists (Wiesbaden, 1987), p. 42, § 64. 3 See Th. Nöldeke, Kurzgefasste syrische Grammatik (Darmstadt, 1977), p. 167, § 222. 2a. 4 This not so much on the linguistic level (we may suppose that the grammatical analysis of the clause constituted no problem to a Syriacspeaker), as on the level of meaning in the context. As I will argue below, 2

1 SAM XVI 6,7 IN THE PESHITTA VERSION

103

as a mistaken translation of the MT. Both ‫אְך‬ ַַ 5 and ‫נֶַּגֶּ ד‬6 are elsewhere translated without difficulty by the Peshitta-translator(s), and the MT gives no apparent occasion to deviate from the usual translation technique.7 Therefore I propose to look to the Hebrew Vorlage of P for an explanation. Whereas - ‫( ܐܟܘܬ‬i.e. ‫ ) ܐܝܟ‬is equivalent neither to MT’s ‫אְך‬, ַ nor to ‫נֶּ גֶּ ד‬,8 it does render a Hebrew expression which resembles ‫אְך נֶַּגֶּ ד‬ ַַ of the MT of 1 Sam. xvi 6, namely ‫כְַנֶַּגֶּ ד‬. This composite preposition occurs in Gen. ii 18, 20 in the expression ‫עזֶּ ר ְכנֶּ גְ דֹּו‬. ֵ In P this is rendered both times with ‫ܡܥܕܪܢܐ‬ ‫ ܐܟܘܬܗ‬, “a helper like him”. Therefore, to account for P’s text we may assume that the Syriac translator read ‫ ְכנֶַּגֶּ ד יהוה ְמ ִשיחֹו‬. This he interpreted, in consonance with the meaning of ‫ ְכנֶּ גֶּ ד‬in Gen. ii 18, 20, as “like the Lord is His Messiah”. Since the difference between ‫ ְכנֶּ גֶּ ד‬and ‫אְך נֶַּגֶּ ד‬ ַַ is small, it is possible that the reading reflected by P came about by accident: either the translator of P misread ‫ ַאְך נֶּ גֶּ ד‬as ‫כנֶּ גֶּ ד‬, ְ or the same mistake was made by a Hebrew scribe of P’s Vorlage.9 However, we must also take account of the possibility that the Hebrew text reflected by P is not due to a mistake. If we take the reconstructed Vorlage seriously, the reading turns out to make good sense in the context. ‫כנֶּ גֶּ ד‬, ְ as used in Gen. ii 18, 20, means ‫‘ ְמ ִשיחֹו‬corresponding to, suitable to, fit for’; it expresses both similarity and complethe text represented by P must be understood in the light of Gen. ii 18, 20. 5 ‫אְך‬ ַַ is rendered in Syriac by ‫ ܒܪܡ‬e.g. 1 Sam. i 23; xii 20; xviii 17; xxix 9; by ‫ ܐܐܠ‬e.g. 1 Sam. xii 24; xx 39; it is sometimes omitted in the translation, e.g. 1 Sam. xviii 8; xxi 5; xxv 21. 6 ‫ נֶּ גֶּ ד‬is usually rendered in Syriac by ‫“ ܩܕܡ‬before” e.g. Gen. xxxi 32, 37; Deut. xxxi 11; Josh. viii 35; 1 Sam. xii 3, 3; xv 30, 30; or by ‫ܠܘܩܒܠ‬ “opposite, over against”: Exod. x 10; xix 2; xxxiv 10; Num. xxv 4; Josh. iii 16; viii 11, 33. 7 To view P’s text as an inner-Syriac corruption is even more problematic. A rendering of MT would probably have read ‫(ܒܪܡ) ܩܕܡ ܡܪܝܐ‬ ‫ ܡܫܝܚܗ‬or‫ (ܒܪܡ) ܠܘܩܒܠ ܡܪܝܐ ܡܫܝܚܗ‬or something similar. This could hardly have corrupted to what we now read in P. 8 It usually renders Hebrew ‫כ‬ ְַ and ‫ ְכמֹו‬. 9 As I will argue below, this is not likely in view of the connection between the variant in v. 6 and that in v. 7, see last section.

104

LANGUAGE AND TEXTUAL HISTORY

mentarity. In light of this, the use of ‫ ְכנֶַּגֶּ ד‬in 1 Sam. xvi 6 amounts to expressing that the Messiah-to-be (Eliab, to Samuel’s mind) is suitable to YHWH, that he is worthy of Him.10 This is certainly the most concise way of describing the ideal Messiah. If Samuel, at sighting Eliab, wished to express his approval of what he thought was God’s choice, he should have employed no other words.

1 SAM. XVI 7 ‫ܐܠ ܗܘܝܬ ܓܝܪ ܐܝܟ ܕܚܙܐ ܐܢܫܐ‬ God rebukes Samuel saying “Do not look on his appearance or on the height of his stature, because I have rejected him”, then, according to MT: ‫הַה ָא ַָדם‬ ָ ‫ ִכיַלאַא ֶּשרַיִ ְר ֶּא‬. These last words are problematic, maybe they must be rendered “for (what matters is) not what man sees.”11 P reads in this place ‫ ܐܠ ܗܘܝܬ ܓܝܪ ܐܝܟ ܕܚܙܐ‬. ‫ܐܢܫܐ‬. Against, the analysis of these words is not exactly simple but it can be made without problems: - ‫ ܐܠ ܗܘܝܬ‬is the 1st person, so to speak, of ‫ ܐܠ ܗܘܐ‬, the regular negation of non verbal clauses in Syriac. In P this negation is always declined in accordance with the subject of the clause; e.g. 10 A similar interpretation, but without emendation of MT and without reference to P, was proposed by A. B. Ehrlich, Randglossen zur hebräischen Bibel, 3 Band (Leipzig, 1910), 222: “JHVH entsprechend muss wohl sein Gesalbter sein.” See also A. Schulz, Das erste Buch Samuel. Exegetisches Handbuch zum Alten Testament (Münster, 1919), 245–246, esp. 242: “Ei, zu Jahwe passt sein Gesalbter.” 11 The commentaries explore different text-critical solutions. Ehrlich, 222 proposes to read ‫ ַיָ ָשר‬instead of ‫א ֶּשר‬: “Der Mensch sieht die Sachen nicht richtig an.” Others, like Driver (see n. 1) and Dhorme, follow the LXX: ὅτι οὐχ ὡς ἐμβλέψεται ἄνθρωπος ὄψεται ὁ θεός. In my opinion the text-form represented by the Septuagint is a simplification of MT. F. M. Cross, “The Oldest Manuscripts from Qumran,” JBL 74 (1955), 147–172, esp. 166, has suggested, on the basis of a computation of the length of the line, that the fragmentary manuscript from Qumran, 4QSamb supports the longer reading of the Septuagint against MT. Since, however, none of the words under discussion are actually attested by 4QSamb (except for the initial ‫)כי‬, ִ all that can reasonably be said on the basis of the available evidence, is that 4QSamb probably points to a text longer than MT.

1 SAM XVI 6,7 IN THE PESHITTA VERSION

105

“I am not a prophet” is rendered ‫( ܐܠ ܗܘܝܬ ܢܒܝܐ‬Zech. xiii 5). ‫ܐܠ‬ ‫ ܗܘܝܬ‬does not imply past tense or ‘becoming’, but simply the negation of a non verbal clause. - ‫ ܓܝܪ‬corresponds toַ‫ ִַכי‬, but takes of necessity the second position in the clause. - ‫“ ܐܝܟ ܕܚܙܐ ܐܢܫܐ‬as man sees”. Thus the clause is to be rendered: “For I am not as man sees.” This text is not a translation of the MT, which does not contain even a hint of a 1st person pronoun (in what follows the 3rd person is used to refer to God). Neither is it clear how P could be an interpretation of MT. 12 Though the MT is not clear, it gave the translator no cause to introduce a 1st person (pronoun or verb). I submit, therefore, that here, as in v. 6, the Syriac must be explained as being based on a variant Hebrew text. The following is a possible retroversion of the Syriac into Hebrew:13 ַ ‫כי ַלא ַכאשר ַיראה‬ ‫האדם ַאני‬,14 “I am not similar to what man sees.”15 The interpretation of this clause in the context is not too difficult: I am not like anything man can see, therefore if you say that Eliab is “similar/complementary” (‫ ) ְכנֶַּגֶּ ד‬to me on the basis of what your eyes see, you are mistaken; my likeness is not a matter of beauty or physique.

P. A. H. de Boer, Research, 40 n. 1: “‫ ִכי לא‬is considered too pregnant and is circumscribed by ‫ܐܠ ܗܘܝܬ ܓܝܪ ܐܝܟ‬.” Unfortunately, de Boer has not explained in what way “For I am not as . . .” could be viewed as a circumscription for “(It) is not . . .” 13 Compare Zech. viii 11 ‫יםַה ִראשנִ יםַאנִ י‬ ָ ‫יָמ‬ ִ ‫אַכ‬ ַ ‫ ל‬which is rendered by ̈ ̈ P ‫ܩܕܡܝ ܐ‬ ‫ܕܒܝܘܡ ܬܐ‬ ‫ ܗܘܝܬ ܐܝܟ‬. ‫ ܐܠ‬. 14 This reconstructed Hebrew text diverges much more widely from its MT equivalent than the reconstructed Hebrew of v. 6. In the last section of this paper it will be discussed how the difference between MT and the hypothetical Hebrew Vorlage of P could have come about. 15 ‫שר‬ ֶּ ‫ ַכא‬is usually employed as a conjunction (see BDB s.v.), which would lead to a meaning “I am not as man sees”. However ‫ ְַכ‬may also retain its prepositional force, with ‫ א ֶּשר‬introducing an independent relative clause (e.g. Job xxix 25). In this case we may render the clause “I am not like what man sees”. This last sense is the preferable one. 12 Against

106

LANGUAGE AND TEXTUAL HISTORY

THE COHERENCE OF THE PESHITTA VERSION OF 1 SAM. XVI 6–7 According to the MT the sequence of events described in 1 Sam. xvi 6–7 is as follows: when Eliab steps forward, Samuel conjectures (‫אְך‬ ַַ ) that the one standing before him must be the Lord’s Messiah. His conjecture is then corrected by God on the grounds that what counts is not what man sees. In comparison with MT, the reconstructed Hebrew Vorlage of P considerably heightens the theological significance of these verses. When confronted with Eliab, Samuel — assuming that this is the one God has chosen, 16 and enthused by God’s choice — makes his capital statement: “Suitable17 to the Lord is His Messiah.” God then corrects Samuel’s outburst: it is not Eliab he has chosen. However, the statement itself is not rejected by God, his Messiah is suitable to Him, only: “I am not similar18 to what man sees”, i.e. my attributes are not accessible to visual perception. The relationship of similarity/complementarity between the Lord and his Messiah is not a matter of physical appearance, though it can be perceived by the heart.19 Thus the reconstructed Hebrew Vorlage of P offers a coherent version of the events, different from MT. The exchange between As is quite natural: Samuel came to anoint one of Jesse’s sons as the new king, and Eliab’s stature is kingly by all means. It has often been remarked that the description of Eliab is reminiscent of the description of Saul in 1 Sam. ix 2. 17 It is difficult to express in one word the full meaning of Hebrew ‫כנֶּ גֶּ ד‬, ְ especially since it must be understood in the light of Gen. ii 18, 20. A question which must be kept apart is what the Syriac translator wished to express with ‫ܐܟܘܬܗ ܕܡܪܝܐ‬, which seems to stress more the aspect of similarity. 18 English cannot express the linguistic link between the vv. 6 and 7. The preposition ְַ‫ כ‬in ‫ ַכא ֶּשר‬of v. 7 echoes the same preposition in ‫ ְכנֶּ גֶּ ד‬of v. 6: “Similar (and complementary) to the LORD is his Messiah” – “I am not similar to what man sees”. Likewise, in P ‫ ܐܝܟ‬of v. 7 echoes ‫ܐܟܘܬܗ‬ of v. 6. 19 With Ehrlich, 223, I am inclined to take ‫ לעינים‬and ‫ ללבב‬as referring to eyes and heart of the subject. 16

1 SAM XVI 6,7 IN THE PESHITTA VERSION

107

the prophet and YHWH prepares the stage for the one who has, in fact, been chosen by YHWH to be his Messiah: David, the man after (ְַ‫ )כ‬His heart.20 In view of this coherence between the two variants preserved by P, it is preferable to attempt to explain both readings together.

THE TEXT OF 1 SAM. XVI 6–7 In the first and second sections it was argued that the variants in P of 1 Sam. xvi 6–7 are due to the fact that P’s Hebrew Vorlage diverged from MT. Also an attempt was made to reconstruct this non-Massoretic Hebrew text. However, the task of the textual critic is not limited to the fairly mechanical steps of determining variant readings and reconstructing the variant Hebrew Vorlage. In the present case, it is necessary to define the relationship between MT and the reconstructed Hebrew Vorlage of P and, if possible, to judge which of the two has more faithfully preserved the original text of 1 Sam. xvi 6–7. In the following I will try to argue for priority of the text-form represented by P. The MT of 1 Sam. xvi 6–7, in the places treated in the present paper, has been viewed as problematic even without reference to the ancient versions. This has led scholars to propose a variety of interpretations or even conjectural emendations in order to obtain a satisfactory sense. In v. 6 ‫דַיהוהַמ ִשיחֹו‬ ְ ֶּ‫ ַאְךַנֶּ ג‬, is a strangely roundabout way of expressing: “This must be the one you sent me to anoint.”21 And in v. 7ַ‫ ִכיַ ַלאַא ֶּשרַיִ ְר ֶּאהַ ָה ָא ָדם‬, is simply obscure.22 20 See 1 Sam. xiii 14 ‫שַיהוהַלֹוַאישַכִ ְל ָבבֹו‬ ִ ‫ ִב ֵק‬. Compare the later interpretation in Acts xiii 22: “I have found in David the son of Jesse a man after my heart.” 21 Because of this difficulty many commentators follow F. Perles, Analekten zur Textkritik des Alten Testamentes (1895), 64, who proposed to read ‫“ ַנְ גִ יד‬prince” instead of ‫נֶּ גֶּ ד‬. P. Dhorme, Les livres de Samuel (Paris, 1910), 141, emends the text to ‫אכן הגיד יהוה משיחו‬. W. D. Caspari, Die Samuelbücher. Kommentar zum Alten Testament (Leipzig, 1926), 129, is even more radical in his treatment of the text. None of these authors base their proposals on versional evidence. Another way of dealing with the difficulty in MT is to interpret ‫ נֶּ גֶּ ד‬in the sense of ‫כנֶּ גֶּ ד‬, ְ see n. 11.

108

LANGUAGE AND TEXTUAL HISTORY

The problems of MT are overcome by the text represented by P. As I have tried to demonstrate above, the reconstructed Vorlage of P yields a very satisfactory sense in the context. However, since the text-form represented by P has an entirely different meaning from MT, there is no reason to view P as an attempt to interpret MT. Furthermore, this is not a case of lectio difficilior and lectio facilior. Both MT and P (i.e. its Vorlage) are difficult, but whereas MT remains opaque, P is capable of being intelligibly interpreted. The main argument for the priority of P’s text-form is that it is very difficult to account for P’s version if we must assume that it is secondary to MT. Admittedly, ‫ ְכנֶַּגֶּ ד‬could have been read (by the translator of P) or written (by the Hebrew scribe of P’s Vorlage) accidentally instead ofַ ‫אְך ַנֶַּגֶּ ד‬ ַַ in v. 6. However, the variant in v. 7 could not easily have been caused by a scribal mistake. Neither is it clear why, in v. 7, MT would have been altered consciously to the text witnessed to by P (see above, section on 1 Sam. xvi 7). And in any case, as was argued above, in view of the coherence of the two variants in P it is not helpful to explain one of them as a scribal lapsus and the second one in a different way. On the other hand, it is perhaps possible to see how the Vorlage of P might have been changed, consciously or semi-consciously, into MT. A scribe or a corrector did not understand the use of ‫ ְכנֶַּגֶּ ד‬in v. 6, or maybe, seeing that Samuel was speaking about Eliab, did not accept the statement. In any case, he wrote ‫אְך‬ ַַ instead of ְַ‫כ‬, changing the statement into a supposition, and otherwise altering its sense considerably. But now that the notion of similarity to YHWH had been purged from v. 6, the clause “I am not similar to” in v. 7 had become meaningless.23 Therefore it was changed too, by omitting ְַ‫כ‬ and by omitting the 1st person pronoun. What was left was a text which did not convey a clear meaning, yet it is that form which was preserved by MT. If this line of arguing is correct, then it would seem that MT is indeed secondary to the text preserved by P. In the absence of supporting arguments this point must remain quite hypothetical. 22 See n. 12 for some text-critical solutions which have been proposed earlier. 23 See especially the linguistic link discussed in n. 20.

THE HEBREW AND SYRIAC TEXT OF DEUT 1:44 The Old Testament Peshitta is without doubt the ancient version most neglected by textual critics. Beyond the questions of language and script, the unreliability of textual editions may be partially to blame for this sad state of affairs. The progression, however, of the Leiden edition should transform earlier practice: the text of the Old Testament Peshitta is now available, for almost all books of the Bible, in an edition based on the best manuscripts and presented in a way designed to facilitate its use in textual criticism. For this achievement, the Peshitta Institute and its present custodian, who is the laureate of this volume, deserve high praise. The following study intends to illustrate the potential value of the Peshitta for the textual criticism of the Hebrew text of the Bible. Deut 1:44 relates the catastrophic outcome of Israel’s attempt, against the express command of YHWH sanctioning their earlier refusal, to conquer the promised land:

‫שר‬ ֶַּ ‫םַכא‬ ַ ‫אתכֵַםַוַ יִ ְר ְדּפוַּ ֶּא ְת ֶּכ‬ ְ ‫רַההוּאַלִ ְק ַר‬ ַ ‫בַב ָה‬ ָ ‫יַהי ֵש‬ ַ ‫אַה ֱאמ ִר‬ ָ ‫וַ יֵ ֵצ‬ ַ‫מה׃‬ ַָ ‫ד־ח ְר‬ ָ ‫ירַע‬ ַ ‫םַב ֵש ִע‬ ְ ‫יַכתוַּ ֶּא ְת ֶּכ‬ ְ ַַ‫הַה ְדּב ִריםַו‬ ַ ָ‫ַתּע ֶּשינ‬ “Then the Amorites who lived in that hill country came out against you and chased you as bees do and beat you down in Seir as far as Hormah.”(RSV) At first sight the text of Deut 1:44a poses no problems. The verse can easily be translated, and fits the context well. In comparison with the earlier narration in Num 14:45 one notes that the sentence ‘and they chased you as bees do’ is added. Such an embellishment is entirely natural in Moses’ oral retelling of the event: the simile adds life to the story of this terrible turn of affairs. The Masoretic

109

110

LANGUAGE AND TEXTUAL HISTORY

text of the verse is in the main supported by the Sama-ritan Pentateuch,1 the Septuagint,2 the Targums,3 and the Vulgate.4

A PROBLEM OF STYLE There is, however, a stylistic reason to suspect the soundness of the transmitted Hebrew text. The use of the verb ‫“ עשה‬to do” in replacement of the verb of the main clause is unique in similes likening an action to the same action as proverbially attributed to a different subject.5 In English, and other European languages, the use of “to do” as a “pro-verb” in order to avoid repetition is entirely idiomatic. The Hebrew language, however, requires repetition of the identical verb. Consider the following examples from the Book of Deuteronomy: Deut 1:31 ‫ת־בנֹו‬ ְ ‫ישַא‬ ֶּ ‫א־א‬ ִ ‫יָךַכא ֶּשרַיִ ָש‬ ַ ‫ֹלה‬ ֶּ ‫הַא‬ ֱ ָ‫נְ ָשאָךַ יְהו‬ “The Lord your God bore you, as a man bears his son.” Deut 8:5 ַָ‫יָךַמיַ ְס ֶַּרך‬ ְ ‫ֹלה‬ ֶּ ‫הַא‬ ֱ ָ‫ת־בנֹוַיְהו‬ ְ ‫ישַא‬ ֶּ ‫רַא‬ ִ ‫ס‬ ֵַ ַ‫יַכא ֶּשרַיְ י‬ ַ ‫ִכ‬ “As a man disciplines his son, the Lord your God disciplines you.” Deut 28:29 ‫שַה ִעוֵּ רַ ָבא ֵפלַָה‬ ָ ‫ש‬ ֵַ ‫רַיְמ‬ ַ ‫ש‬ ֶַּ ‫יִםַכא‬ ַ ‫שַב ָצ ֳה ַר‬ ַ ‫ש‬ ֵַ ‫ַמ ַמ‬ ְ ‫ית‬ ָ ִַ‫וְ ָהי‬

1 The Samaritan Pentateuch exhibits several variants which make the verse agree with the parallel text in Num 14:45. Conversely, the simile of the bees is introduced in the Numbers passage, where it is absent from the MT. Such harmonisations are typical of the Samaritan text and are generally agreed to be secondary. 2 For the syntax of the Septuagint, see the article quoted below in n. 5. 3 A minor variant in some of the Targums will be discussed below. 4 A minor variant in the Vulgate will be discussed below. 5 On the syntax of similes, see J. Joosten, “Elaborate Similes – Hebrew and Greek. A Study in Septuagint Translation Technique,” Biblica 77 (1996), 227–236; T. L. Brensinger, Simile and Prophetic Language in the Old Testament (Mellen Biblical Press Series 43; Lewiston NY, 1996).

THE HEBREW AND SYRIAC TEXT OF DEUT 1:44

111

“And you shall grope at noonday, as the blind grope in darkness.” In all these examples, the simile contains an adverbial complement thus setting them apart from the simile in Deut 1:44. But even when there is no further complement, the main verb is repeated: Judg 7:5 ַ‫קַה ֶַּכ ֶּלב‬ ַ ‫יִםַכא ֶּשרַיָ ֹל‬ ַ ‫ן־ה ַמ‬ ַ ‫שֹונֹוַמ‬ ִ ‫ַכלַא ֶּשר־יָ ֹלקַ ִב ְל‬ “Every one that laps the water with his tongue, as a dog laps.”6 Amos 2:13 ‫הַה ְמ ֵל ָאהַלַָּה‬ ַ ַָ‫רַתּ ִעיקַ ָהעגָ ל‬ ָ ‫םַכא ֶּש‬ ַ ‫יכ‬ ֶּ ‫יקַתּ ְח ֵתּ‬ ַ ‫יַמ ִע‬ ֵ ‫הַאנ ִַכ‬ ָ ֵַ‫ִהנ‬ ‫ָע ִמיר‬ “Behold, I am pressed under you, as a cart is pressed that is full of sheaves.” (KJV) The repetition of the verb of the main clause is the rule, not only in Deuteronomy, but in the Hebrew Bible in general.7 Whereas such repetition is found 18 times, the substitution of ‫ עשה‬in the comparative clause is found only in our verse.8 In light of this rule, one would have expected the text of Deut 1:44a to read: ‫…וירדפו אתכם כאשר תרדפנה הדברים‬ “…and chased you as bees chase”9

6 See the French translation La Bible en français courant: “Ceux qui laperont l’eau avec la langue comme le font les chiens.” 7 See also: Exod 33:11; Num 11:12; Judg 7:5; 16:9; 2 Sam 19:4; 1 Kgs 14:10; 2 Kgs 21:13; Isa 25:11; 66:20; Jer 13:11; 43:12; Amos 2:13; 3:12; 9:9; Mal 3:17. 8 This statement is valid only for quasi-proverbial similes. Where a comparison is made to a specific event, the verb ‫ עשה‬may be used, see Deut 2:12. 9 The verb ‫ רדף‬may occur without explicit direct object: Gen 14:14; 1 Sam 30:10.

112

LANGUAGE AND TEXTUAL HISTORY

It is interesting to note that the Vulgate as well as some of the Targums has adapted their rendering of the verse to the dominant phrasing of the Hebrew bible: Vulgate persecutus est vos sicut solent apes persequi Targum Neofiti ‫ור ַדפו יתכון היך מה דרדפן א>ו)q>‘ étant habituel, ainsi que le changement du ‘ayin initial en aleph par dissimilation.27 En syriaque, le verbe habituel pour exprimer la hâte est ‫ܣܪܗܒ‬.28 On ne peut cependant exclure que le verbe ‫ܐܒܥ‬, exprimant un sens proche, ait été connu en syriaque ancien. L’isolement de la présente attestation pourrait être le fruit du hasard. Il est néanmoins plus plausible de supposer que le verbe ne représente pas ici la langue syriaque mais qu’il s’agisse d’un élément occidental ayant pénétré dans la version de Ben Sira. Notons que dans le Targum d’Onkelos, la racine ‫ אבע‬rend plusieurs fois la racine hébraïque ‫ חוש‬qui est celle du texte hébreu de Ben Sira 36,7.

‫“ ܕܩܝܩܬܐ‬fièvre hectique” Sir 26,26 ‫ܚܪܝܢܗ ܕܐܢܬܬܐ ܒܡܟܝܟܘܬܐ ܗܘ ܘܐܝܟ ܐܫܬܐ ܕܩܝܩܬܐ‬ ‫ܗܟܢܐ ܬܬܚܙܐ‬ La dispute d’une femme s’exprime dans l’humilité (?) et comme une fièvre hectique, voilà comment elle paraîtra. Ce verset ne figure que dans la version syriaque. L’expression ‫ ܐܫܬܐ ܕܩܝܩܬܐ‬a été rendue tenuis febricula, “une fièvre légère”, dans la polyglotte de Londres. Cette interprétation rassurante doit probablement être écartée. En effet, sur une amulette magique retrouvée dans le Golan, ‫ אשתה דקיקתה‬apparaît comme l’une des menaces contre laquelle la porteuse, Yaïta, a besoin d’être protégée. Voir DNWSI, 821; E. Y. Kutscher, “The Language of the ‘Genesis Apocryphon’: A Preliminary Study”, Scripta Hierosolymitana 4 (1965), 8, 11 (Ap. de la Genèse XX.9). 26 P. ex., Nb 32,17; 1 R 22,9; formes réfléchies: 1 S 23,26; 2 S 4,4; 2 R 7,15. Voir A. Tal, The Language of the Targum of the Former Prophets and its Position within the Aramaic Dialects (Tel-Aviv, 1975), 90. 27 P. Grelot, “On the root ‫עבק‬/‫ עבץ‬in Ancient Aramaic and in Ugaritic”, JSS 1 (1956), 202–205. 28 Voir, p. ex., Sir 5,11; 6,7; 42,23; 43,5. 25

LA VERSION SYRIAQUE DE BEN SIRA

291

Les éditeurs de l’amulette, Joseph Naveh et Shaul Shaked, identifient le mal concerné comme la fièvre hectique — sens repris par Michael Sokoloff dans son dictionnaire.29 Le nom ‫ דקיקתה‬se rencontre, apparemment avec le même sens, dans le Targum samaritain en Dt 28,22. Il paraît certain que l’expression en Ben Sira doive être rapprochée de ces autres attestations. La distribution des attestations suggère qu’il s’agit d’une expression d’origine occidentale empruntée dans ce texte syriaque. Si la compréhension de l’expression est exacte, il convient d’attribuer un sens plus sévère au verset. La querelle de la femme, si elle s’exprime d’abord humblement, paraîtra finalement comme une fièvre violente qui consume tout le corps.

‫“ ܠܚܕܐ‬très” Sir 1,29

̈ ‫ܘܒܣܦܘܬܟ ܬܗܘܐ ܙܗܝܪ ܠܚܕܐ‬

Et sois très prudent de tes lèvres. Hébreu: hiat. Grec: καὶ ἐν τοῖς χείλεσίν σου πρόσεχε Sir 51,24

‫ܢܦܫܟܘܢ ܬܗܘܐ ܨܗܝܐ ܠܚܕܐ‬ Votre âme sera très assoiffée Hébreu: B ‫ונפשכם צמאה מאד תהיה‬ Grec: καὶ αἱ ψυχαὶ ὑμῶν διψῶσι σφόδρα Le cas de ‫ܠܚܕܐ‬, “très”, représente l’un des occidentalismes les plus frappants dans notre texte. En syriaque, l’adverbe “très” s’exprime en général par ‫ܛܒ‬.30 L’emploi de ‫לחדה‬/‫ לחדא‬dans ce sens est réservé aux dialectes occidentaux: Qumran, Targums d’Onkelos et de Jonathan, araméen palestinien juif et chrétien.31

‫“ ܐܬܚܟܡ‬se faire connaître” 29 J. Naveh, S. Shaked, Amulets and Magic Bowls (Jérusalem, 1985), 49; M. Sokoloff, A Dictionary of Jewish Palestinian Aramaic of the Byzantine Period (Ramat Gan, 1990), 154. 30 Voir, p. ex., Sir 7,17; 25,2; 42,23; 47,24. 31 Voir Kutscher, Language, 12.

292

LANGUAGE AND TEXTUAL HISTORY Sir 12,8

̈ ‫ܒܛܒܬܗ ܪܚܡܐ ܘܐܠ ܢܬܛܫܐ‬ ̈ ‫ܒܒܝܫܬܗ ܣܢܐܐ‬ ‫ܐܠ ܢܬܚܟܡ‬ L’ami ne se fait pas connaître lorsqu’il fait le bien, de même l’ennemi ne sait rester caché lorsqu’il fait le mal. Hébreu: ‫לא יודע ַבטובה אוהב ולא יכוסה ברעה שונא‬ Grec: οὐκ ἐκδικηθήσεται ἐν ἀγαθοῖς ὁ φίλος καὶ οὐ κρυβήσεται ἐν κακοῖς ὁ ἐχθρός

Le verbe ḥkm, “connaître”, bien attesté dans les dialectes araméens occidentaux, n’est pas fréquent en syriaque.32 Cependant, les occurrences, dont deux se trouvent dans le livre de Ben Sira (15,19; 50,28), sont trop nombreux pour penser à un emprunt. Il en est autrement pour le cas cité de la forme réfléchie. La langue syriaque connaît le Ethpaal signifiant “devenir sage, être instruit”, “se montrer sage” ou bien “être connu charnellement”.33 Mais aucune de ces significations ne convient dans le verset cité cidessus. D’ailleurs, dans notre verset, le verbe est dépourvu du point supralinéaire, ce qui indique qu’il se trouve au Ethpeel — conjugaison qui ne semble pas être attestée ailleurs en syriaque.34 Or, dans l’araméen juif palestinien, une forme réfléchie s’emploie dans le sens de “être connu, se faire connaître”.35 Ce sens correspond à l’équivalent hébreu en Sir 12,8 et permet de comprendre le contexte de façon satisfaisante.36 Si cette interprétation est à retenir, il paraît raisonnable de reconnaître ici un autre élément occidental.

32 Le sens de “connaître charnellement” est attesté à partir de la Peshitta de l’AT, Gn 4,1. Comme le fait remarquer Payne-Smith dans le Thesaurus, p. 1265, le sens de “connaître” est presque entièrement réservé au participe (mais voir Sir 50,28). 33 Voir, p. ex., Sir 2,3; 6,32.33; 18,29; 4,32; 38,2.24.25.31; 39,6. 34 Brockelmann, dans son Lexicon syriacum, 231, signale l’existence du Ethpeel mais pas du Ethpaal pour ce verbe. 35 Voir notamment Sokoloff, Dictionary, 201. Il paraît difficile de distinguer le Ethpeel du Ethpaal comme le fait Sokoloff. 36 Notons cependant que le sens du texte syriaque n’est pas celui du texte hébreu: en hébreu le “bien” et le “mal” se réfèrent à la situation de celui à qui le verset est adressé (voir les commentaires).

LA VERSION SYRIAQUE DE BEN SIRA

293

‫“ ܚܫܝܟ‬misérable” Sir 4,1 ‫ܐܠ ܬܛܪܦ ܠܡܣܟܢܐ ܚܫܝܟܐ‬ Ne maltraite pas le pauvre misérable. Hébreu: ‫אל תדאיב נפש עני ַומר נפש‬ Grec: μὴ παρελκύσῃς ὀφθαλμοὺς ἐπιδεεῖς L’adjectif ‫ܚܫܝܟ‬, “obscur”, se rencontre en syriaque, mais ce sens ne semble pas approprié dans le présent contexte. Brockelmann37 propose le sens de “aveugle”, attesté seulement dans le présent passage. Au vu du contexte et des équivalents hébreu et grec, il paraîtrait plutôt que l’adjectif dénote ici la pauvreté et la misère. Or, un adjectif ‫חשיך‬, “pauvre, misérable”, est attesté dans la langue du Targum de Jonathan.38 Si ce sens est adopté, on est amené une fois de plus à conclure à la présence d’un élément occidental dans notre texte.

‫“ ܢܚܡ‬consoler” Sir 48,24

̈ ‫ܐܠܒܝܐܠ‬ ‫ܘܒܪܘܚܐ ܕܓܢܒܪܘܬܐ ܚܙܐ ܐܚ̈ܪܝܬܐ ܘܢܚܡ‬ ‫ܕܨܗܝܘܢ‬ Dans un esprit de puissance, il (le prophète Ésaïe) vit les choses dernières et consola les endeuillés de Sion. Hébreu: ‫ברוח גבורה חזה אחרית וינחם אבלי ציון‬ Grec: πνεύματι μεγάλῳ εἶδεν τὰ ἔσχατα καὶ παρεκάλεσεν τοὺς πενθοῦντας ἐν Σιων

La racine ‫ ܢܚܡ‬porte en syriaque le sens de “ressusciter”. Le sens de “consoler, réconforter”, bien connu de l’hébreu et des dialectes occidentaux,39 n’y est pas attesté.40 Dans le verset auquel il est fait Brockelmann, Lexicon syriacum, 262. Voir Éz 18,12 (où le terme se trouve en parallèle avec ‫;)מסכין‬ Jr 52,15; És 54,11; de même dans Midrash Tehillim (Buber, éd.), 192. 39 On trouve une seule attestation de cette racine en araméen ancien (KAI 271 A 2), mais dans un contexte difficile, voir DNWSI, 725. Il faudra envisager la possibilité que le sens attesté en araméen occidental soit dû à l’influence du substrat hébreu. 40 Seule exception possible: Job 42,6 dans la Peshitta. 37 38

294

LANGUAGE AND TEXTUAL HISTORY

allusion ici, Ésaïe 40,1, la Peshitta traduit le verbe hébreu ‫ נחם‬par le verbe syriaque ‫ܒܝܐ‬. Pourtant, le contexte ainsi que l’équivalent hébreu et grec ne laissent aucun doute: le verbe signifie “consoler” dans le présent passage. On pourrait expliquer ce changement de sens en invoquant un hébraïsme: le traducteur aurait simplement transcrit le terme de son modèle dans sa traduction syriaque. De tels hébraïsmes lexicaux sont cependant rares dans la version syriaque de Ben Sira. Il paraît donc plus judicieux de supposer que le verbe problématique représente l’araméen occidental plutôt que l’hébreu.

‫“ ܩܘܒܠܐܠ‬obscurité” (?) Sir 23,19

̈ ‫ܘܚܙܐ ܐܘ̈ܪܚܬܐ ܕܟܠ‬ ‫ܕܒܢܝ ܐܢܫܐ ܘܡܣܬܟܠ ܡܐ ܕܒܚܫܘܟܐ‬ ̈ ‫ܕܥܒ‬ ‫ܕܝܗܘܢ‬ ‫ܩܘܒܠܐܠ‬ ̣ Et il voit les chemins des hommes et observe ce qui est dans l’obscurité, la face (leg. l’obscurité?) de leurs actions. Hébreu: hiat. Grec: ἐπιβλέποντες πάσας ὁδοὺς ἀνθρώπων καὶ κατανοοῦντες εἰς ἀπόκρυφα μέρη

Le nom ‫ܩܘܒܠܐܠ‬, “face, visage”, est bien attesté en syriaque. Cependant, l’interprétation ne satisfait guère dans le contexte et ne correspond à rien dans la version grecque (le texte hébreu étant malheureusement vacant). En face de ce problème, il paraît licite d’envisager l’intrusion d’un élément occidental. Or, il existe un mot ‫(קבל‬/‫ )קביל‬dans l’araméen des Targums d’Onkelos et de Jonathan ainsi que dans l’araméen palestinien juif, chrétien et samaritain signifiant “obscurité, ténèbres”. Si ce mot se lisait dans une version antérieure de ce verset, la forme textuelle actuelle pourrait ܳ ‫ܚ ܽܫ‬ s’expliquer comme une “syriacisation” du texte: le mot ‫ܘܟܐ‬ pourrait être une glose expliquant au lecteur syriaque le sens du mot occidental; dans un deuxième temps, le mot ‫ קבלה‬lui-même aurait reçu une forme syriaque impliquant un sens étranger au contexte. L’hypothèse peut paraître précaire, mais elle a le mérite d’expliquer la forme alambiquée du texte syriaque.

‫“ ܐܬܪܥܝ‬agréer” Sir 37,28 ‫ܡܛܠ ܕܐܠ ܗܘܐ ܟܠ ܡܐܟܘܠܬܐ ܛܒܐ ܘܐܠ ܟܠ ܢܦܫܐ‬

LA VERSION SYRIAQUE DE BEN SIRA

295

‫ܒܩܠܝܠ ܡܬܪܥܝܐ‬ Car ce n’est pas toute nourriture qui est bonne et toute âme ne se plait pas à peu de choses (?).41 Hébreu: ‫כי לא הכל ַלכל טוב לא כל נפש כל זן תבחר‬42 Grec: οὐ γὰρ πάντα πᾶσιν συμφέρει καὶ οὐ πᾶσα ψυχὴ ἐν παντὶ εὐδοκεῖ

En syriaque, le verbe ‫ ܪܥܝ‬signifie “penser, réfléchir” au Ethpeel et “se réconcilier” ou encore “avoir l’intention” au Ethpaal. Le sens d’“agréer, choisir”, n’est attesté que dans le présent passage. Pour exprimer ce sens, le syriaque emploie habituellement le verbe ‫ܨܒܝ‬ au Ethpeel.43 Par contre, ce sens est fréquemment exprimé par la racine ‫ רעי‬dans les Targums d’Onkelos et de Jonathan et dans les dialectes palestiniens juif, chrétien et samaritain. Notons encore deux détails: l’expression ‐‫ אתרעי ב‬traduit parfois le verbe hébreu ‫בחר‬, qui figure également dans le texte hébreu de notre passage;44 et l’expression ‐‫א ְת ַר ְעיַת נַ ְפ ֵשיּה ב‬, ֵ “son âme se plait de, il désire” est une tournure targumique connue.45 Dans l’ensemble ces faits font soupçonner qu’il s’agit ici une nouvelle fois d’un élément occidental dans la version syriaque de Ben Sira. Les huit éléments énumérés ci-dessus ont ceci en commun que leur attestation en syriaque est limitée aux passages de Ben Sira tandis qu’ils sont bien représentés dans l’araméen occidental. Sur le plan dialectal, c’est la langue des Targums d’Onkelos et de Jonathan qui offre le plus de parallèles, seuls l’expression ‫ ܐܫܬܐ ܕܩܝܩܬܐ‬et le verbe ‫ ܐܬܚܟܡ‬n’y sont pas représentés. Pour expliquer la présence de ces éléments dans la version syriaque de Ben Sira, plusieurs hypothèses différentes se présentent à l’esprit. La version a pu être réalisée dans une localité se trouvant à mi-chemin entre Édesse et Jérusalem. Ou bien le traducteur serait un savant d’origine palestinienne, migré en Mésopotamie mais ayant appris le syriaque de façon imparfaite. Au stade actuel de la Le sens du texte syriaque est difficile. Le traducteur a-t-il lu ‫ בקל‬au lieu de ‫?בכל‬ 42 Le manuscrit D donne un texte hébreu légèrement différent. 43 Voir, p. ex., Sir 51,13. 44 Voir, p. ex., Tg És 42,1. 45 Voir, p. ex., TgO Gn 34,3.8.; TgN Gn 49,6. 41

296

LANGUAGE AND TEXTUAL HISTORY

recherche, on ne saurait exclure aucune de ces possibilités. Deux phénomènes notés en passant ci-dessus privilégient cependant une autre piste. Premièrement, comme nous l’avons vu, l’élément occidental se trouve souvent isolé dans le livre de Ben Sira lui-même: on rencontre une fois le verbe ‫ܐܒܥ‬, “se hâter”, mais plusieurs fois le verbe habituel ‫ ;ܣܪܗܒ‬l’adverbe ‫ܠܚܕܐ‬, “très”, figure deux fois, mais son équivalent ‫ ܛܒ‬est plus fréquent; et ainsi de suite.46 Cet état de choses crée l’impression d’un texte de base révisé de façon imparfaite, c’est-à-dire, probablement: une version en araméen occidental révisée d’après les règles de la langue syriaque (l’inverse paraît moins probable). Deuxièmement, le cas de ‫ܩܘܒܠܐܠ‬, “face, visage”, qui cache vraisemblablement un mot occidental signifiant “obscurité”, suggère un procédé semblable. Dans ce cas précis, tout se passe comme si un texte en araméen occidental aurait été rendu lisible pour un lecteur syriaque, toutefois au détriment du sens. Peut-on aller plus loin? Il est tentant d’employer les données linguistiques pour essayer de résoudre une partie des problèmes de l’arrière-plan historique de la version. La traduction syriaque de Ben Sira aurait connu une rédaction en deux phases. L’écrit de base serait une sorte de Targum d’origine purement juive, rédigé en araméen occidental. Dans un deuxième temps, ce “Targum” aurait été revu et corrigé dans un milieu chrétien de langue syriaque.47 Comme cela a été dit précédemment, une genèse en deux phases expliquerait le profil étonnant, incorporant des motifs juifs et chrétiens, de la version. Toutefois, l’hypothèse ne peut être poursuivie dans ces pages.

CONCLUSION L’objectif de la présente étude a été d’avancer la connaissance de la traduction syriaque de Ben Sira en attirant l’attention sur la présence dans ce document d’un certain nombre d’occidentalismes. L’existence dans un texte syriaque de mots et d’expressions représentant l’araméen occidental est frappante — et suggestive. Elle Voir, ci-dessus, les notes 28, 30, 33, 42. Cette hypothèse rejoint un modèle longtemps soutenu, mais aujourd’hui abandonné, pour expliquer la genèse de la Peshitta de l’Ancien Testament. 46 47

LA VERSION SYRIAQUE DE BEN SIRA

297

pourrait fournir la clé d’une meilleure compréhension de la genèse et du milieu d’origine de la version syriaque du livre de Ben Sira. Le mélange d’éléments juifs et chrétiens pourrait être le résultat d’une réalisation en deux temps de cette version ancienne. Les recherches futures devront déterminer le bien-fondé de cette intuition. Je suis heureux d’offrir ces quelques pages au grand savant et à l’hôte généreux qu’est Moshe Bar Asher. Puisse le proverbe de Ben Sira (3,31) s’appliquer à lui: ‫פועלַטובַיקראנוַבדרכיו‬

‫ܕܥܒܕ ܕܫܦܝܪ ܥܬܝܕ ܒܐܘܪܚܗ‬

SOURCES “La Peshitta de l’Ancien Testament dans la recherche récente”, Revue d’Histoire et de Philosophie religieuses 76 (1996), 385–395. “The Use of some Particles in the Old Testament Peshitta,” Textus 14 (1988), 175–183. “Greek and Latin Words in the Peshitta Pentateuch. First Soundings” in R. Lavenant, ed., VII Symposium Syriacum (Orientalia Christiana Analecta 256; Rome, 1998), 37–47. “Materials for a Linguistic Approach to the Old Testament Peshitta,” Journal for the Aramaic Bible 1 (1999), 203–218. “Doublet Translations in Peshitta Proverbs” in P. B. Dirksen, A. van der Kooij, eds., The Peshitta as a Translation (Monographs of the Peshitta Institute 8; Leiden, 1995), 63–72. “Greek Words Shared by the Peshitta and Targums to the Pentateuch” in G. Greenberg and A. Rapoport, eds, Biblical Hebrew, Biblical Texts. Michael Weitzman Memorial Volume (Sheffield, 2001), 165– 177. “1 Samuel XVI 6, 7 in the Peshitta Version,” Vetus Testamentum 41 (1991), 226-233. “The Hebrew and Syriac Text of Deut 1:44,” in W. Th. van Peursen, R. B. ter Haar Romeny, eds., Text, Translation and Tradition. Studies on the Peshitta and its Use in the Syriac Tradition Presented to Konrad D. Jenner on the Occasion of his Sixty-Fifth Birthday (Monographs of the Peshitta Institute 14; Leiden, 2006), 65–69. “The Old Testament in the New. The Syriac Versions of the New Testament as Witness to the Text of the Old Testament Peshitta” in B. ter Haar Romeny, ed., The Peshitta : Its Use in Literature and Lit-

299

300

LANGUAGE AND TEXTUAL HISTORY

urgy. Papers Read at the Third Peshitta Symposium (Brill, Leiden, 2006), 99–106. “West Aramaic Elements in the Old Syriac and Peshitta Gospels,” Journal of Biblical Literature 110 (1991), 271–289. “Two West Aramaic Elements in the Syriac Gospels,” Biblische Notizen 61 (1992), 17–21. “West Aramaic Elements in the Syriac Gospels: Methodological Considerations” in R. Lavenant, ed., VI Symposium Syriacum 1992 (Orientalia Christiana Analecta 247; Rome, 1994), 101–109. “« Le Père envoie le Fils »: La provenance occidentale d’une locution syriaque”, Revue d’Histoire des Religions 214 (1997), 299–309. “La tradition syriaque des Évangiles et la question du « substrat araméen »”, Revue d’Histoire et de Philosophie Religieuse 77 (1997), 257– 272. “The Text of Matt 13,21a and parallels in the Syriac Tradition,” New Testament Studies 37 (1991), 153–159. “The Old Testament Quotations in the Old Syriac and the Peshitta Gospels: A Contribution to the Study of the Diatessaron,” Textus 15 (1990), 55–76. “Tatian’s Diatessaron and the Old Testament Peshitta,” Journal of Biblical Literature 120 (2001), 501–523. “Odes de Salomon 7,3a. Observations sur un hellénisme dans le texte syriaque,” Zeitschrift für die neutestamentlische Wissenschaft 89 (1998), 134–135. “Éléments d’araméen occidental dans la version syriaque de Ben Sira” in A. Maman, S. E. Fassberg, Y. Breuer, eds, Sha’arey Lashon. Studies in Hebrew, Aramaic and Jewish Languages Presented to Moshe Bar Asher, Vol. II, Mishnaic Hebrew and Aramaic (Jerusalem, Bialik, 2007), *42–*55.

INDEX OF BIBLICAL REFERENCES OLD TESTAMENT Genesis

1:11 2:12 2:18.20 2:24 3:7 4:8 4:9 4:21 12:18 15:17 22:7 24:8 25:32 27:46 28:17 31:27 32:28 34:3 34:15.22.23 37:25 38:25 39:21 40:13 41:40 42:3 42:4 42:38 43:10 43:11 43:30 44:2

45:22 47:26 50:26

46 46, 86 103 240 47 18 27–28 47 25 46 65 47 55 55 61 29–30 55 55 47, 87, 93, 94 47 179 66 42 44 41 47 47 29 47 44 46

Exodus

46 42 44

4:10 57 7:28 43 47, 87 12:7.22.23 44, 47 14:6 44, 47 16:3 47 16:33 47 20:18 46, 87 21:17 230, 233–234, 244, 265 21:21 29 21:22.23 47 21:30 66 21:32 46 23:9 27 24:10 46 25:7 86 25:9 42 25:29 47, 87 25:31 46, 86 26:6.11 87 27:3 47 27:4 87 28:9 46, 86, 87 28:19 47 28:20 86 28:32 45, 47, 87, 94, 97 29:23 46 29:40 46 30:24 47

301

302 32:4 32:19 32:30 35:9 35:11 35:27 37:17 38:3 38:4 38:30 39:6 39:12 39:13 39:21 39:23 39:39 40:36

Leviticus 2:5 2:7 6:14(21) 7:1 7:9 8:11 11:19 18:3 19:29

Numbers 4:7 4:14 7:13f 8:3 8:4 9:14 11:4 11:11 11:12 11:20 11:29 14:2 14:45 15:4

LANGUAGE AND TEXTUAL HISTORY 43, 46, 87, 96 29 46 86 47, 87 86 86 47 87 87 86 47 86 61 47, 87, 94 86–87 61

46, 87 46 46, 87 42 46, 87 46 40, 46 42 39, 47 46 47 87 47, 87 86 41 29 25 111 43, 46 29 29 109–110 103

15:38 18:7 19:18 21:29 22:37 23:3 24:24 25:8 32:1.4 35:2.3.4

145–146 46 271 46 25 46 45–46, 87, 92 45, 47, 87, 95 55 46

Deuteronomy 1:10 1:31 1:44 2:12 3:5 4:16 4:37 5:26 6:5 6:13 7:25 8:3 8:5 9:9.11.18 10:9 10:10 13:9 14:17 17:18 18:2 18:14 20:20 21:20 21:22.23 23:16 25:13 26:5 28:22 28:29 28:49 28:67 31:11

61 110 109, 111–114 111 47 41 47 29 261 271 42 229 110 139 27 139 47 40, 46 44 27 29 47, 88 46, 74 161 74 47 135 291 110 112 30 103

INDEX 32:10 32:21 32:25

Joshua 2:19 3:16 8:11.33 8:29 8:35 10:26

Judges

5:14 6:12 7:5 7:14 7:16 9:29 15:1 19:8

Ruth 1:9

1 Samuel 1:8 9:2 12:3 12:24 15:19 15:30 16:6.7 18:8 20:14 20:39 21:5 25:21 26:15 23:23 26:20 30:10

66 123–124 44 55 103 103 161 103 161 60 74 111 61 91 29 91 32 66 25, 56 106 103 103 25 103 101–108 103 30 103 103 103 25 61 112 111

303 2 Samuel 1:13 14:9 16:17 17:12 19:1 19:4 19:15 19:25

1 Kings 1:15 3:16 5:1 5:31 6:36 6:37 14:6 14:10 16:34 17:4 18:26 18:43 19:3 22:25

2 Kings 4:12.38 5:6.7 5:23 6:9 6:12 12:9 14:8 14:9 21:13

1 Chronicles 2:21 9:1 9:2 22:2 25:8

65 55 25 112 29 111 177 25 91 126 54 66 37 66 170 111 66 62 62 143 143 62 143 54 93 177 91 93 54 54 111 62 60 65 65 144

304 2 Chronicles 2:16 9:15 20:17 29:1 30:25

Ezra 3:11 6:11

Nehemiah 8:9

Judith 15:11

Esther 1:5 2:15.17 2:23 4:17

Job

1:9 3:10 3:13 3:16 4:19 6:2 6:22 7:5 9:33 10:18 13:5 13:15 14:13 19:23 21:29 22:30 23:3 28:4 29:2 29:3.4

LANGUAGE AND TEXTUAL HISTORY

65 65 64 62 65 66 161 27 62 60 66 161 178 28 25 29 25 66 30 28 61 29 29–30, 57 29 29 29 29 28 62 29 65 29 61

30:2 30:30 31:31 31:35 33:9 37:13 42:6

Psalms 1:3 8:3 11:3 13:6 22:19 41:6 48:3 55:7 78:2 78:24 89:4 91:11 91:12

102:28 106:23 118:22 118:23 119:5 124:1–5 139:8

Odes

6:8 7:2 7:3 10:3 15:6 17:14 25:1 33:4 37:3 42:15.17

55 61 29 29 55 66 66, 293 217 272 29 29 237 30 59 29 230, 260 268 206 236 229, 231, 235, 253, 260 62 206 231, 253, 266 233 29 30 62 280 280 279 280 280 280 280 280 280 280

INDEX Proverbs 3:5 3:11 3:35 5:11 6:19 8:2 8:21 11:29a 11:31 13:22 14:9 14:18 14:22 14:23 14:31 14:32 14:35 15:17 17:16 18:8a 18:22 19:15 19:17 22:27 23:20a 23:30b 23:31 23:32 28:10

81 74 76 55 55 62 76 75, 79 15, 74–75, 137 76 72, 76 76 81 77–78 288 72, 80–81 81 59 55 73 82–83 73 72 73 79 79 78 74 76

Ecclesiastes 2:20 7:26 8:13

Song of Songs

1:3 1:4 2:17 4:6 8:10

39 62 62 59 74 32 32 62

305 Sirach

1:8.9.10.17.20 288 1:20 287 1:29 291 2:3 292 3:12.13 288 3:31 297 4:1 293 4:32 292 5:11 290 6:7 290 6:32.33 292 7:17 291 11:12 284 12:8 292 14:2 288 15:19 292 16:24.27.29 288 18:29 292 21:14.16 288 21:27 287 22:12 288 23:15 288 23:19 294 24:22 288 25:2 291 26:9 287 26:26 290 28:1.2 288 29:15 288 30:11 288 31:22 288 32:2.23 288 32:20 287 33:13.22 288 34:24 287 35:9 288 36:7 65, 289–290 36:13 285 37:13 285 37:24 288 37:28 294 38:2.24.25.31 292 38:18 287

306 38:25 38:29 39:6 39:31 40:21 40:28 41:12 42:23 43:5 44:23 47:24 48:17 48:24 48:25 49:13 50:28 51:3 51:13 51:24

Isaiah

LANGUAGE AND TEXTUAL HISTORY 287 288 292 288 287 285 285 290–291 290 288 291 287 293 66 287 292 288 295 291

6:9 229, 264, 269 6:10 236, 269 7:14 232, 253, 255 8:1 93 8:23 232 9:1 233, 271 13:10 230, 253 14:29 74 23:1.6.10.14 37 23:15 29 25:6 137 25:11 111 26:20 32 27:4 29 30:6 74 34:3 55 36:10 28 36:17 32 37:17 177 40:1 294 40:3 231–232, 234–235, 253, 256, 267–268 40:4 231, 256

40:5 42:1 42:2 42:4 44:28 48:18 53:4 56:7 60:16 61:1 65:7 66:20 66:20

Jeremiah 8:23 9:1 13:11 13:12 17:8 29:22 43:12

Lamentations 4:8

Ezekiel 4:3 15:5 18:19 27:12 36:35 37:16 38:13

Daniel 2:9 2:39 11:36

Hosea 1:7 2:10 10:12

231 256 229 271 66 30 271 237, 244 74 273 55 111 47 29 29 111 28 217 91 111 61 91 55 25 37 145 29 37 32 62 32 29 29 32

INDEX Joel 2:3 2:20

145 55

307 Nahum 2:11

61

Habakkuk

Amos 2:13 3:12 4:10 7:10 9:9

111 111 55 177 111

2:3

Zechariah 9:9 13:5

55 230, 237, 251 105

Malachi

Jonah 1:6 2:1

138 139

3:1 3:17

233–234, 244, 256 111

10:38 10:40 11:10

160 141, 155, 180 233–234, 238, 244 179 194 30 256 229, 238 271 139, 154 229, 239 230, 235, 238– 239 204, 213–215, 218 238 194 145, 148 230, 233, 238, 244 135, 154 26 203 30 233 240 160 230, 238

NEW TESTAMENT Matthew 1:23 2:19 2:23 3:3 3:4 3:5 4:4 4:6 4:10 4:15 4:16 4:25 5:13 5:35 5:42 5:43 6:12 6:22 6:24 7:6 8:17 9:18 9:22 10:32

232, 238 230 232 232, 234–235, 238, 244 138, 148 233 229, 238 229, 235, 237, 238, 244 271 232, 238 233, 238, 270 233 55 59 285 271 191 61 66 191 271 230 286 135

11:16 11:17 11:21 12:18 12:19 12:20 12:40 13:14 13:15 13:21 13:35 14:26 14:36 15:4 15:13 16:11 16:18 17:20 19:1 19:5 19:21 21:5

308 21:13 21:16 21:29.32 21:42 22:25 22:28 23:5 23:23 23:37 24:40 24:29 26:14 26:47 27:19 28:6

Mark 1:2 1:3 2:14 3:8 4:8 4:12 4:17 5:9 5:34 6:49 7:10 7:26 7:34 9:48 10:1 10:21 10:41 10:52 11:17 11:31 12:10 12:11 14:36 15:24 15:40 16:1 16:6

LANGUAGE AND TEXTUAL HISTORY 237–238, 244 256, 272 55 233, 238, 244 230 163 145 30 170 138, 148 230, 238 162 162 179 139–140 233, 256 267 196 233 190 235 213–215 55 286 194 230, 234, 238 144, 150 174 230 233 160 163 286 237 26 265 234, 238 65, 174 237, 239 196 196 139, 140

Luke

1:32 232 2:14 141, 154, 196 2:35 142, 154 2:49 135 3:4 231, 234, 238 3:5 231, 238 3:6 231, 238 4:8 271 4:11 231, 239 4:10 235 4:18 141, 180, 273 7:2 230 7:6 54, 141, 171, 179 7:14 135–136 7:27 234, 238 7:50 286 8:13 213–214 8:48 286 9:23 160 9:35 136, 205 9:52 170 10:16 141, 155 11:4 191 11:10 140 12:28 138, 148 13:25 140 14:27 160 14:35 55 16:9.11.13 66 17:19 286 18:42 154, 286 19:23 26, 30 19:42 30 19:46 237–238 20:17 231, 238 23:35 205 23:39 160 23:43 145, 154 24:6 139, 140, 147

John 1:23

234, 244, 256, 266–267, 270

INDEX 1:28 233 1:34 136, 206 3:2 140 3:18 136, 149, 205 3:26 233 4:48 140 5:36 141, 180 5:37 180 6:31 234, 238, 268 6:32 135 6:71 162 7:29 141 7:35 144, 150 10:9 154 10:40 233 11:19 55 11:39 230 12:15 237–238 12:20 144, 150 12:37 140 12:40 236, 238–239, 243 13:20 141 16:32 57 17:23.25 141 18:10 196 19:6.15 160 19:24 237 20:24 162

Acts

5:28 13:22 13:29 14:9 16:1.3 23:10

260 107 160 286 145 177

309 Romans 1:16 8:15 10:19 12:20

1 Corinthians 1:11

2 Corinthians 7:8

Galatians 3:1 3:24

1 Thessalonians 2:11 3:3

Hebrews 6:12 10:38.39 12:2 12:5

1 Peter 2:9 2:24 4:8 4:18 5:5

Revelation 9:11 21:19

145 147 123–125 74 177 55 160 280 55 55 55 55 160 74 234 160 74 75, 137 74 281 38

‫‪INDEX OF SYRIAC WORDS‬‬ ‫‪135–136, 148‬‬

‫–‪65, 135, 147‬‬ ‫‪148, 154,180,‬‬ ‫‪184‬‬

‫ܐܪܢܐ‬

‫–‪ 136–137, 148‬ܒܚܝܪܐ‬ ‫‪149, 180, 205‬‬

‫ܐܒܐ‬

‫‪42, 46‬‬

‫ܐܓܘܪܣܐ‬

‫‪65‬‬

‫ܐܘܪܝܬܐ‬

‫‪55‬‬

‫ܐܙܠ ܠ‬

‫‪105‬‬

‫ܐܝ ܟ‬

‫‪25, 56‬‬

‫ܐܝܟܘ‬

‫‪55‬‬

‫ܐܝܟܢܐ ܫܡܟ‬

‫‪61‬‬

‫ܐܝܬ‬

‫‪61‬‬

‫ܐܟܘܡ‬

‫‪101–102‬‬

‫ܐܟܘܬ‪-‬‬ ‫ܡܐܡܪܐ‬

‫‪287‬‬

‫ܰܒ ܺܢܝ‬

‫‪46, 86‬‬

‫ܒܣܣ‬

‫‪46, 86‬‬

‫ܒܪܘܐܠ‬

‫‪27–28, 105‬‬

‫ܓܝܪ‬

‫‪65‬‬

‫ܓܝܘܪܐ‬

‫‪44‬‬

‫ܓܠܘܣܩܡܐ‬

‫‪42, 46, 87, 90‬‬

‫ܓܠܦ‬

‫‪145, 148, 150,‬‬ ‫‪154‬‬

‫ܓܢܬ ܥܕܢ‬

‫‪46‬‬

‫ܓܢܣܐ‬

‫‪207, 284‬‬

‫‪29–31, 56–57‬‬

‫ܕܝܢ‬

‫‪46‬‬

‫ܐܣܘܛ‬

‫‪44‬‬

‫ܕܝܬܝܩܐ‬

‫‪46‬‬

‫‪290, 295‬‬

‫ܕܩܝܩܬܐ‬

‫‪38‬‬

‫ܐܣܟܝܡܐ‬ ‫̈‬ ‫ܐܣܦܝܢܝܩܐ‬

‫‪60‬‬

‫ܗܠܘܟ‬

‫‪46‬‬

‫ܐܣܩܦܐ‬

‫‪46‬‬

‫ܗܡܝܢܐ‬

‫‪37, 46‬‬

‫ܐܣܬܝܪ‬

‫‪42, 46‬‬

‫ܗܡܝܪܐ‬

‫‪287‬‬

‫ܐܦܘܕܝܐ‬

‫‪46‬‬

‫ܙܘܓܐ‬

‫‪42, 46‬‬

‫ܐܦܬܪܐ‬

‫‪73‬‬

‫ܚܒܢܢܐ‬

‫‪65‬‬

‫ܐܪܘܢܐ‬

‫‪74, 137, 138,‬‬ ‫‪148, 154, 180‬‬

‫ܚܝܐ‬

‫‪144, 148, 150‬‬

‫ܐܪܡܝܐ‬

‫‪311‬‬

‫‪312‬‬

‫‪LANGUAGE AND TEXTUAL HISTORY‬‬ ‫‪47‬‬

‫ܡܘܟܐܠ‬

‫‪137, 166‬‬

‫ܡܚܝܢܐ‬

‫‪55‬‬

‫ܡܚܐ ܪܝܚܐ‬

‫‪291–292, 295‬‬

‫ܐܬܚܟܡ‬

‫‪55‬‬

‫ܡܐܠ ܒܠܒܗ‬

‫‪65‬‬

‫ܚܣܕܐ‬

‫‪66‬‬

‫ܡܡܘܢܐ‬

‫‪66‬‬

‫ܚܣܝܕܐ‬

‫‪47‬‬ ‫‪289–290, 296‬‬

‫ܡܪܣܘܦܐ‬ ‫ܰ‬ ‫ܐ ܰܒܥ‬

‫‪293‬‬

‫ܚܫܝܟ‬

‫‪55‬‬

‫ܚܫܩܒܘܠ‬

‫‪66, 293‬‬

‫ܢܚܡ‬

‫‪46, 87, 91‬‬

‫ܛܐܓܢܐ‬

‫‪140, 148, 180‬‬

‫ܢܝܣܐ‪/‬ܢܣܐ‬

‫‪40, 46‬‬

‫ܛܘܣܐ‬

‫‪42, 47‬‬

‫ܢܡܘܣܐ‬ ‫ܰ‬ ‫ܐܩܦ‬

‫‪46, 87, 92‬‬

‫ܛܘܦܣܐ‬

‫‪138, 148, 149‬‬

‫ܛܘܪܐ‬

‫‪66‬‬

‫–‪ܰ 140–141, 148‬‬ ‫ܐܩܫ‬ ‫‪149‬‬

‫‪46‬‬

‫ܛܟ‬

‫‪38, 46‬‬

‫ܛܪܘܐܠ‬

‫‪38, 46‬‬

‫ܛܪܛܩܠ‬

‫‪138–139, 148,‬‬ ‫‪154‬‬

‫ܝܘܡ‬

‫‪66‬‬

‫ܝܬ‬

‫‪66‬‬

‫ܫܟܠܠ‬

‫‪46‬‬

‫ܟܪܘܡܐ‬ ‫ܠ‬

‫‪37, 40, 47‬‬

‫ܣܐܡܐ‬

‫‪47, 87‬‬

‫ܣܝܪܣ‬

‫‪39‬‬

‫ܣܡܝܕܐ‬

‫‪142, 148–149,‬‬ ‫‪154‬‬

‫ܬܥܒܪܝܢܗ‬

‫‪32‬‬

‫ܥܕ‬

‫‪61‬‬

‫ܥܠܘܝ‬

‫‪280‬‬

‫‪37, 47, 87‬‬

‫ܦܝܠܣܐ‬

‫‪38, 46, 87, 92‬‬

‫‪55‬‬

‫ܡܦܨܝ ܐܢܐ ܡܢ‬

‫ܠܓܝܘܢܐ‬

‫‪46‬‬

‫‪37, 47‬‬

‫ܦܝܪܡܐ‬ ‫ܰ‬ ‫ܐܦܣ‬

‫ܠܓܢܐ‬

‫‪287‬‬

‫ܡܠܝܛ‬

‫‪291, 296‬‬

‫‪39, 47, 87, 93‬‬

‫ܐܬܛܦܝܣ‬

‫ܠܚܕܐ‬

‫‪139, 147–149‬‬

‫‪47‬‬

‫ܦܪܘܣܬܕܐ‬

‫ܠܝܬ‬

‫‪143, 148, 166‬‬

‫‪47‬‬

‫ܦܪܙܘܡܐ‬

‫ܬܠܡܝܕܐ‬

‫‪37, 46, 87, 90‬‬

‫‪47, 87, 94‬‬

‫ܦܪܦܐ‬

‫ܠܡܦܐܕܐ‬

‫‪44, 60‬‬

‫‪47‬‬

‫ܦܪܨܘܦܐ‬

‫ܠܣܛܐ‬

‫‪46‬‬

‫ܠܩܢܐ‬

‫‪38, 47‬‬

‫‪INDEX‬‬

‫‪313‬‬ ‫‪59‬‬

‫ܪܚܡܬܐ‬

‫‪160–162, 166‬‬

‫ܨܠܒ‪/‬ܨܠܝܒܐ‬

‫‪55‬‬

‫ܪܡܐ ܒܝܬ‬

‫‪42, 47, 65‬‬

‫ܩܐܒܘܬܐ‬

‫‪294‬‬

‫ܐܬܪܥܝ‬

‫‪47‬‬

‫ܩܐܪܣܐ‬

‫‪141, 148, 155,‬‬ ‫‪181, 197‬‬

‫ܐܪܥܘܬܐ‬

‫‪47‬‬

‫ܩܕܣܐ‬

‫ܰ‬ ‫ܪ ܰܩܕ‬

‫‪294, 296‬‬

‫ܩܘܒܠܐܠ‬

‫‪87, 91‬‬

‫ܩܘܪܩܣܐ‬

‫‪55‬‬

‫ܐܠܬܬܩܛܥ‬ ‫ܠܟܘܢ‬

‫‪43, 47, 87, 90‬‬

‫ܩܝܛܘܢܐ‬ ‫ܩܝܬܪܐ‬

‫‪194‬‬ ‫‪194‬‬

‫ܐܪܩܕ‬

‫‪54, 141, 155,‬‬ ‫‪171, 176–177,‬‬ ‫‪179, 182, 184‬‬

‫ܫܕܪ‬

‫‪287‬‬

‫ܫܚܘܪܬܐ‬

‫‪47‬‬

‫‪143, 148, 166,‬‬ ‫‪170–171‬‬

‫ܫܠܝܚܐ‬

‫‪37–38, 47, 87,‬‬ ‫‪95‬‬

‫ܩܠܝܬܐ‬

‫‪54, 141–142,‬‬ ‫–‪148, 155, 175‬‬ ‫‪185‬‬

‫ܫܠܚ‬

‫‪37, 47, 88‬‬

‫ܩܠܩܘܡܐ‬

‫‪38, 47‬‬

‫ܩܣܛܐ‬

‫‪287‬‬

‫ܡܬܝܟܐ‬

‫‪47‬‬

‫ܩܣܝܐ‬

‫‪55‬‬

‫ܬܬܘܝܟ ܢܦܫܟ‬

‫‪61‬‬

‫ܩܦܘܕ‬

‫‪287‬‬

‫ܡܬܝܪ‬

‫‪38, 47‬‬

‫ܩܪܕܐܠ‬

‫‪145–146, 148,‬‬ ‫‪150‬‬

‫ܬܟܠܬܐ‬

‫‪38, 47‬‬

‫ܩܪܘܟܐ‬

‫‪59, 138‬‬

‫ܩܪܝܬܐ‬

‫‪44, 60‬‬

‫ܬܪܘܢܘܣ‬

‫‪37, 47‬‬

‫ܩܪܟܕܢܐ‬

‫‪162‬‬

‫ܬܪܥܣܪܬܐ‬

‫‪87, 94‬‬

‫ܩܪܩܠ‬

‫‪47‬‬

‫ܪܗܛܢܐ‬

‫‪59‬‬

‫ܪܚܡ‬